神经行为学 AVIOR ANALYSIS in NEUROSCIENCE, Second Edition

Methods of
BEHAVIOR ANALYSIS
in NEUROSCIENCE
Second Edition
© 2009 by Taylor & Francis Group, LLC

Methods of
BEHAVIOR ANALYSIS
in NEUROSCIENCE
Second Edition
Edited by
Jerry J. Buccafusco
Medical College of Georgia
Augusta
Boca Raton London New York
CRC Press is an imprint of the

Taylor & Francis Group, an informa business

CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works

Printed in the United States of America on acid-free paper
10 9 8 7 6 5 4 3 2 1
International Standard Book Number-13: 978-1-4200-5234-3 (Hardcover)
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts
have been made to publish reliable data and information, but the author and publisher cannot assume
responsibility for the validity of all materials or the consequences of their use. The authors and publishers
have attempted to trace the copyright holders of all material reproduced in this publication and apologize
to copyright holders if permission to publish in this form has not been obtained. If any copyright material
has not been acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, trans-
mitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter
invented, including photocopying, microfilming, and recording, or in any information storage or retrieval
system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.
com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood
Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and
registration for a variety of users. For organizations that have been granted a photocopy license by the
CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Methods of behavior analysis in neuroscience / editor, Jerry J. Buccafusco. -- 2nd ed.

p. ; cm. -- (Frontiers in neuroscience)
Rev. ed. of: Methods of behavior analysis in neuroscience / edited by Jerry J.
Buccafusco. c2001.
Includes bibliographical references and index.
ISBN 978-1-4200-5234-3 (hardcover : alk. paper)
1. Neurosciences. 2. Nervous system--Diseases--Animal models. 3. Animal
behavior I. Buccafusco, Jerry J. II. Methods of behavior analysis in neuroscience. III.
Title. IV. Series: Frontiers in neuroscience (Boca Raton, Fla.).
[DNLM: 1. Behavior, Animal. 2. Neurosciences--methods. 3. Animals,
Laboratory--psychology. WL 100 M5925 2008]
RC343.M45 2008
616.8--dc22 2008038324
Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com

This book is dedicated to my wife Regina, to her parents
Claire and James, and to my parents Rose and Dominick — with
appreciation for their love and support, particularly during my early
academic years; and to our children Chris and Marty,
both writers in their respective fields.
Acknowledgment and appreciation also goes to my office manager

Vanessa Cherry for her assistance in the organization of the manuscripts
and for helping (often vainly) to keep us on schedule.
Finally, I would like to express my appreciation to our excellent animal

behavior technical assistants, Laura Shuster-Pearson, Nancy Kille and
Donna Blessing for their scientific prowess and for the kindness
they impart in working with our nonhuman primate and rodent subjects.

Contents
Series Preface............................................................................................................ix
Preface ......................................................................................................................xi
The Editor ................................................................................................................ xv
Contributors ...........................................................................................................xvii
Chapter 1 Transgenic Mouse Models of Alzheimer’s Disease: Behavioral

Testing and Considerations .................................................................. 1
Kathryn J. Bryan, Hyoung-gon Lee, George Perry, Mark A.
Smith, and Gemma Casadesus
Chapter 2 Cued and Contextual Fear Conditioning for Rodents ........................ 19

Peter Curzon, Nathan R. Rustay, and Kaitlin E. Browman
Chapter 3 Drug Discrimination .......................................................................... 39

Richard Young
Chapter 4 Conditioned Place Preference ............................................................ 59

Adam J. Prus, John R. James, and John A. Rosecrans
Chapter 5 Anxiety-Related Behaviors in Mice................................................... 77

Kathleen R. Bailey and Jacqueline N. Crawley
Chapter 6 Behavioral Assessment of Antidepressant Activity in Rodents....... 103

Vincent Castagné, Paul Moser, and Roger D. Porsolt
Chapter 7 Assessing Attention in Rodents........................................................ 119

Philip J. Bushnell and Barbara J. Strupp
Chapter 8 The Behavioral Assessment of Sensorimotor Processes in the

Mouse: Acoustic Startle, Sensory Gating, Locomotor Activity,
Rotarod, and Beam Walking............................................................ 145
Peter Curzon, Min Zhang, Richard J. Radek, and Gerard B. Fox
vii

viii Contents
Chapter 9 Intravenous Drug Self-Administration in Nonhuman Primates ...... 179

Leonard L. Howell and William E. Fantegrossi
Chapter 10 Contextually Induced Drug Seeking During Protracted

Abstinence in Rats............................................................................ 199
Jerry J. Buccafusco and Laura Shuster
Chapter 11 Operant Analysis of Fronto-striatal Function in Rodents................ 215

Máté D. Döbrössy, Simon Brooks, Rebecca Trueman, Peter J.
Brasted, and Stephen B. Dunnett
Chapter 12 Working Memory: Delayed Response Tasks in Monkeys ............... 247

Jesse S. Rodriguez and Merle G. Paule
Chapter 13 Spatial Navigation (Water Maze) Tasks ........................................... 267

Alvin V. Terry Jr.
Chapter 14 Water Maze Tasks in Mice: Special Reference to Alzheimer’s

Transgenic Mice ............................................................................... 281
Dave Morgan
Chapter 15 Behavioral Neuroscience of Zebrafish ............................................. 293

Edward D. Levin and Daniel T. Cerutti
Chapter 16 Caenorhabditis elegans Model for Initial Screening and

Mechanistic Evaluation of Potential New Drugs for Aging and
Alzheimer’s Disease......................................................................... 311
Yuan Luo, Yanjue Wu, Marishka Brown, and Christopher D. Link
Chapter 17 The Revival of Scopolamine Reversal for the Assessment of

Cognition-Enhancing Drugs ............................................................ 329
Jerry J. Buccafusco

Series Preface
Our goal in creating the Frontiers in Neuroscience Series is to present the insights of
experts on emerging fields and theoretical concepts that are, or will be, in the van-
guard of neuroscience. Books in the series cover genetics, ion channels, apoptosis,
electrodes, neural ensemble recordings in behaving animals, and even robotics. The
series also covers new and exciting multidisciplinary areas of brain research, such as
computational neuroscience and neuroengineering, and describes breakthroughs in
classical fields like behavioral neuroscience. We hope every neuroscientist will use
these books in order to get acquainted with new ideas and frontiers in brain research.
These books can be given to graduate students and postdoctoral fellows when they
are looking for guidance to start a new line of research.
Each book is edited by an expert and consists of chapters written by the leaders
in a particular field. Books are richly illustrated and contain comprehensive bibli-
ographies. Chapters provide substantial background material relevant to the particu-
lar subject. We hope that as the volumes become available, the effort put in by us,
the publisher, the book editors, and individual authors will contribute to the further
development of brain research. The extent to which we achieve this goal will be
determined by the utility of these books.
Sidney A. Simon, Ph.D.
Miguel A.L. Nicolelis, M.D.,Ph.D.
Series Editors
ix

Preface
In the six years since the publication of the first edition of Methods of Behavior
Analysis in Neuroscience there have been significant advances in the concepts and
methodology used in the assessment of animal behavior. The purpose of this second
edition is to provide the reader with updates to the technology, as well as to describe
some new approaches that are becoming standard in the field. Examples of the latter
are the chapters that assess behavior in zebra fish and in Caenorhabditis elegans.
This edition also includes chapters describing methods for assessing the cognitive
impairment associated with major psychotic disorders, as well as new approaches
to assessing cognitive impairment in aged mice and in mice impaired as a conse-
quence of high cerebral amyloid burden. Examples of the revival of the scopolamine
reversal model for assessing the clinical relevance of new cognition-enhancing drugs
rounds out the book.
In the first edition the lead article, “Choice of Animal Subjects in Behavioral
Analysis,” written by William J. Jackson, set the stage for the ensuing chapters.
The discussion mainly focused on the origin of the laboratory rat, historically one
of the most dominant animal species used in biomedical research, and on nonhu-
man primate species used in behavior research. Only a few short paragraphs were
dedicated to research with mice. In recent years, with the development of transgenic
and knockout mice, this species is now poised to dominate the field. In this edition
the stage again is set with a timely review of the use of transgenic mice as models
for Alzheimer’s disease. The chapter is balanced with a discussion of the strengths
and weaknesses of the model. In fact, one of the themes of this edition is the util-
ity of animal behavior methods in both basic and applied research. The timing of
the publication of this edition is particularly relevant in view of the vast number of
scientists who are returning from the past two decades of molecular and cellular
biology to look toward whole animal research to translate the most important of
their findings for clinical application. Since the neuroscientist trained in methodolo-
gies directed toward the molecular and cellular level does not often have experience
in the intricacies of animal behavioral analyses, there is often much time devoted to
analyzing complex literature, or to developing an approach de novo. Specialists who
are recognized experts in several fields of cognitive and behavioral neuroscience
have provided chapters that focus on a particular behavioral model. Each author has
analyzed the literature to describe the most frequently used and accepted version of
the model. Each chapter includes (1) a well-referenced introduction that covers the
theory behind, and the utility of, the model; (2) a detailed and step-wise methodol-
ogy; and (3) an approach to data interpretation. Many chapters also provide examples
of actual experiments that use the method.
This edition was designed as a reference manual for use by practicing scientists
with various levels of experience who wish to use well-studied behavioral approaches
in animal subjects to better understand the effects of disease, and to predict the
effects of new therapeutic treatments on human cognition. As with the first edition,
xi

xii Preface
there has been no attempt to cover all areas of animal behavior and sensory process-
ing, though together both volumes now provide significant coverage of the field.
These texts will help take the guesswork out of designing the methodology for many
of the most widely used animal behavioral approaches developed for the study of
brain disorders, drug abuse, toxicology, and cognitive drug development.
As a matter of convenience the topics have been arranged in chapter form, and
there is a false sense of security that each method described is the last word on the
subject. However, it is often not sufficient to employ only one of these methods to
assess the cognitive status of an animal. For example, when studying memory or
recall, it is prudent to use a test battery that can better provide a comfortable level
of interpretation of the effect of the perturbation applied to the subject. Spatial and
nonspatial tasks should be considered. If negative reinforcement is involved, such
as electrical shock, the animal should be tested for its response to pain. Drugs or
other manipulations that might alter pain sensitivity could give false impressions in a
shock-motivated memory task. Drugs that affect motor activity may alter maze activ-
ity or swimming behavior, and drugs that alter taste or appetite, or that induce GI
disturbances could affect food-motivated behaviors. Whenever possible, the animal
should be observed (at least initially) while performing the task. It is often surprising
to some investigators (this one included) to find the animal using a behavior to solve
the problem that was not considered in designing the task. A good example is the
mediating or non-mnemonic strategies that rats use to solve matching problems in
various operant paradigms. Most animals would rather use such strategies (such as
orientating to a proffered lever) to obtain food rewards than use memory. Whenever
possible, the authors have provided some of these pitfalls in their chapters, although
every possible contingency cannot be anticipated. Thus, it is in the best interest of the
investigator to use this book to help develop several strategies to understand the com-
plex behaviors of animals as they respond to drugs, new diets, surgical interventions,
or to additional or fewer genes. While danger in anthropomorphizing the behavior
of animals always exists, the investigator should feel some level of confidence that
much of the behavioral literature is replete with instances of high predictive value
for similar perturbations in humans. Of course, species and strain differences can
limit such interpretations. Mice are clearly not little rats, and rats are not nonhu-
man primates. Each species has a specific level of predictive value that should be
assessed. A final cautionary note is that investigators make every attempt to make
their experiments as reproducible as possible when studying animal behavior.
Handlers, experimenters, food, water, bedding, noise, and surrounding visual cues
are just a few of the factors that should be held constant when performing behavioral
studies. Inconsistency contributes mightily to response variability in a population,
and may even lead to a completely opposite behavior to the one expected.
I would like to express my sincere thanks to the many authors who contributed
these chapters. Their difficult task in preparing this information will make easier the
tasks of our readers in their own efforts to assess animal behavior. I would also like
to acknowledge the support (moral and technical) of the CRC Press staff, Senior Edi-
tor Barbara Norwitz and Senior Project Coordinator Jill Jurgensen, and the Methods
in Neuroscience Series Editors, Sidney Simon and Miquel Nicolelis. Finally, I would

Preface xiii
like to thank my office administrator, Vanessa Cherry, for her many contributions in
getting this book together for publication.

The Editor
Jerry J. Buccafusco, Ph.D., is Director of the Alzheimer’s Research Center in the
Department of Pharmacology and Toxicology of the Medical College of Georgia.
He holds the ranks of Professor of Pharmacology and Toxicology and Professor of
Psychiatry and Health Behavior. He holds a joint appointment as Research Pharma-
cologist and Director of the Neuropharmacology Laboratory at the Department of
Veterans Affairs Medical Center. Dr. Buccafusco is also president and CEO (and
founder, est. 03/01/2000) of Prime Behavior Testing Laboratories, Inc. (Evans, GA),
a contract research company for the preclinical evaluation of cognition-enhancing
therapeutic agents.
Dr. Buccafusco was classically trained as a chemist, receiving the M.S. degree
in inorganic chemistry from Canisius College in 1973. His pharmacological training
was initiated at the University of Medicine and Dentistry of New Jersey where he
received the Ph.D. degree in 1978. His doctoral thesis concerned the role of cen-
tral cholinergic neurons in mediating a hypertensive state in rats. Part of this work
included the measurement of several components of hypothalamically mediated
escape behavior in this model. His postdoctoral experience included two years at the
Roche Institute of Molecular Biology under the direction of Dr. Sydney Spector. In
1979 he joined the Department of Pharmacology and Toxicology of the Medical Col-
lege of Georgia. In 1989 Dr. Buccafusco helped found and became the director of the
Medical College of Georgia’s Alzheimer’s Research Center. The center hosts several
core facilities, including the Animal Behavior Center, which houses more than 50
young and aged macaque monkeys who participate in cognitive research studies.
Awards and honors resulting from Dr. Buccafusco’s research include the New
Investigator Award, National Institute on Drug Abuse, 1980; Sandoz Distinguished
Lecturer, 1983; Distinguished Faculty Award for the Basic Sciences, School of Med-
icine (Medical College of Georgia), 1988; Callaway Foundation of Georgia, Center
Grant recipient, 1989; and the Distinguished Alumnus Award, University of Medi-
cine and Dentistry of New Jersey, 1998. In 2008, Dr. Buccafusco was appointed
Veterans Administration Career Scientist, and he was the recipient of the American
Society for Pharmacology and Therapeutics’ Pharmacia-ASPET Award for Experi-
mental Therapeutics. Dr. Buccafusco also served as member of the Pharmacology
II Study Section of the National Institute on Drug Abuse from 1989–1991. He is a
member of the Scientific Advisory Board of the Institute for the Study of Aging,
New York, NY, and is a consultant to several pharmaceutical companies in the area
of neuropharmacology and drug discovery. Dr. Buccafusco holds memberships in
several scientific societies. In the professional society, the American Society for
Pharmacology and Experimental Therapeutics, he recently completed a three-year
term as (inaugural) Chairman of the Division of Systems and Integrated Pharmacol-
ogy. He also serves as Associate Editor (Neuro-Behavioral Pharmacology section)
for the Journal of Pharmacology and Experimental Therapeutics.
xv

xvi The Editor
Dr. Buccafusco has authored more than 200 research publications and book chap-
ters. Over the years these articles have received more than 3000 citations by other
authors. His research area includes the development of novel treatment modalities
for Alzheimer’s disease and related disorders. In 1988, his laboratory was the first
to report the cognitive enhancing action of low doses of nicotine in nonhuman pri-
mates. Since that time he has studied numerous novel memory-enhancing agents
derived from several pharmacological classes in this model. His most recent work
is directed at the development of single molecular entities that act on multiple CNS
targets to not only enhance cognitive function, but also to provide neuroprotection
or alter the disposition and metabolism of amyloid precursor protein. Dr. Buccafusco
has also studied the toxic effects of organophosphorus anticholinesterases used as
insecticides and chemical warfare agents. In particular, he has studied the behav-
ioral/cognitive alterations associated with low level, chronic exposure to such agents.
His work in the area of drug abuse has centered around the role of central cholinergic
neurons in the development of physical dependence on opiates, and in the expres-
sion of acute and protracted withdrawal behaviors. Most recently, his laboratory is
investigating the role of the immune system and in the production of autoantibodies
to G-amyloid and to the receptor for advanced glycation end products (RAGE) by
individuals with Alzheimer’s disease. These studies have been supported by continu-
ous federally sponsored grants and by several private foundations and commercial
interests.

Contributors
Kathleen R. Bailey, Ph.D.
Department of Psychology
Susquehanna University
Selinsgrove, Pennsylvania, USA
Peter J. Brasted, Ph.D.

Cardiff School of Biosciences
Cardiff University
Cardiff, Wales, UK
Simon Brooks, Ph.D.

Cardiff University
Cardiff, Wales, UK
Kaitlin E. Browman, Ph.D.

Neuroscience Research
Abbott Laboratories
Abbott Park, Illinois, USA
Marishka Brown, B.S.

Department of Pharmaceutical Sciences
School of Pharmacy
University of Maryland
Baltimore, Maryland, USA
Kathryn J. Bryan, Ph.D.

Department of Pathology
Case Western Reserve University
Cleveland, Ohio, USA
Jerry J. Buccafusco, Ph.D.

Alzheimer’s Research Center
Department of Pharmacology and Toxicology
and
Charlie Norwood Veterans Affairs Medical Center
Augusta, Georgia, USA
xvii

xviii Contributors
Philip J. Bushnell, Ph.D.

Neurotoxicology Division
National Health and Environmental Effects Research Laboratory
US Environmental Protection Agency
Research Triangle Park, North Carolina, USA
Gemma Casadesus, Ph.D.

Department of Neurosciences
Vincent Castagné, Ph.D.

Porsolt & Partners Pharmacology
Boulogne-Billancourt, France
Daniel T. Cerutti, Ph.D.

Psychology Department
California State University
East Bay, California, USA
Jacqueline N. Crawley, Ph.D.

Laboratory of Behavioral Neuroscience
National Institute of Mental Health
National Institutes of Health
Bethesda, Maryland, USA
Peter Curzon
Research Investigator
Abbott Laboratories
Máté D. Döbrössy, Ph.D.

Cardiff University
Cardiff, Wales, UK
Stephen B. Dunnett, Ph.D.

Cardiff University
Cardiff, Wales, UK
and
Stereotactic Neurosurgery
Laboratory of Molecular Neurosurgery
Universitätsklinikum Freiburg
Freiburg, Germany

Contributors xix
William E. Fantegrossi, Ph.D.

Yerkes National Primate Research Center
Emory University
Atlanta, Georgia, USA
Gerard B. Fox, Ph.D.

Advanced Technology
Abbott Laboratories
Leonard L. Howell, Ph.D.

Yerkes National Primate Research Center
Department of Psychiatry and Behavioral Sciences
and
Department of Pharmacology
Emory University
Atlanta, Georgia, USA
John R. James, Ph.D.

Department of Pharmaceutics
Virginia Commonwealth University
Richmond, Virginia, USA
Hyoung-gon Lee, Ph.D.

Edward D. Levin, Ph.D.

Department of Psychiatry and Behavioral Sciences
Duke University Medical Center
Durham, North Carolina, USA
Christopher D. Link, Ph.D.

Institute for Behavioral Genetics
University of Colorado
Boulder, Colorado, USA
Yuan Luo, Ph.D.

School of Pharmacy

xx Contributors
Dave Morgan, Ph.D.

Department of Molecular Pharmacology and Physiology
School of Biomedical Sciences
College of Medicine
University of South Florida
Tampa, Florida, USA
Paul Moser, Ph.D.

Merle G. Paule, Ph.D.

National Center for Toxicological Research
Division of Neurotoxicology
Jefferson, Arkansas, USA
George Perry, Ph.D.

College of Sciences
University of Texas at San Antonio
San Antonio, Texas, USA
Roger D. Porsolt, Ph.D.

Adam J. Prus, Ph.D.

Psychology Department
Northern Michigan University
Marquette, Michigan, USA
Richard J. Radek, M.S.

Abbott Laboratories
Jesse S. Rodriguez, Ph.D.

National Center for Toxicological Research
Division of Neurotoxicology
Jefferson, Arkansas, USA
John A. Rosecrans, Ph.D.


Contributors xxi
Nathan R. Rustay, Ph.D.

Abbott Laboratories
Laura Shuster
Charlie Norwood Veterans Affairs Medical Center
Mark A. Smith, Ph.D.

Barbara J. Strupp, Ph.D.

Division of Nutritional Sciences and Department of Psychology
Cornell University
Ithaca, New York, USA
Alvin V. Terry Jr., Ph.D.

Rebecca Trueman, Ph.D.

Cardiff University
Cardiff, Wales, UK
Yanjue Wu, Ph.D.

School of Pharmacy
Richard Young, Ph.D.

Department of Medicinal Chemistry
School of Pharmacy
Min Zhang, Ph.D.

Abbott Laboratories

1 Transgenic Mouse
Models of Alzheimer’s
Disease
Behavioral Testing and
Considerations
Kathryn J. Bryan, Hyoung-gon Lee,
George Perry, Mark A. Smith,
and Gemma Casadesus
CONTENTS
1.1 Introduction.....................................................................................................2
1.2 Behavioral Tests ..............................................................................................2
1.2.1 Spatial Memory Tasks .........................................................................3
1.2.1.1 The Morris Water Maze .........................................................3
1.2.1.2 Radial Arm Maze...................................................................3
1.2.1.3 Radial Arm Water Maze ........................................................4
1.2.2 Contextual Memory .............................................................................4
1.2.2.1 Fear Conditioning...................................................................4
1.2.2.2 Passive-Avoidance Learning ..................................................5
1.2.3 Working Memory/Novelty/Activity ..................................................... 5
1.2.3.1 Y-Maze ...................................................................................5
1.2.3.2 T-Maze....................................................................................6
1.2.3.3 Object Recognition................................................................. 6
1.2.3.4 Open Field ..............................................................................6
1.3 Transgenic Mouse Models of Alzheimer’s Disease........................................7
1.3.1 Amyloid-G Transgenic Mouse Models................................................. 7
1.3.2 Tau Transgenic Mouse Models ............................................................9
1.4 Concerns with Transgenic Mouse Models of Alzheimer’s Disease.............. 10
References................................................................................................................ 14

2 Methods of Behavior Analysis in Neuroscience, Second Edition
1.1 INTRODUCTION
One hundred years ago, the German psychiatrist and neuropathologist Alois
Alzheimer gave a lecture in which he identified a disease of the cerebral cortex1
that would ultimately bear his name: Alzheimer’s disease (AD). In individuals with
this condition, the cerebral cortex is thinner than normal and senile plaques, along
with neurofibrillary tangles (NFTs), are found in the brain.2 In the early 1980s, the
biochemical characterization of senile plaques in patients with Down’s syndrome
and AD led to the identification of amyloid-G (AG) peptide as a major component.
Thereafter, it was determined that AG is a product of the AG protein precursor (APP).
The importance of AG/APP in the pathogenesis of AD is evidenced by the fact that
genetic mutations in the APP gene invariably cause AD in cases with the early onset
familial form of the disease.3–5 The relationship between APP and AG caused the
research community to respond with quick enthusiasm for AG and laid the founda-
tion for the amyloid cascade hypothesis.4,6 The amyloid cascade hypothesis states
that mutations in APP (or other genes) lead to an increase in AG and that this then
leads to disease. While the original hypothesis6 posited AG fibrils as the major medi-
ator of the disease, a more recent incarnation of the hypothesis4 proposes smaller
oligomeric forms of AG as key. In both cases, AG is viewed as being important in
mediating the neuronal and synaptic toxicity that leads to the deterioration of cogni-
tion.7 Likewise, a steady influx of research began to elucidate the role of NFTs and
their principal protein component, phosphorylated tau, in the brain and how these
pathological entities related to the symptomatology of AD.8 While the pathological
significance of AG and NFTs in disease, as well as their interaction is still under
much discussion,9,10 the majority of investigators in the field are convinced that they
play fundamental roles in the onset and progression of AD. That said, other theo-
ries of AD, unrelated to NFTs and AG deposits, are also being actively pursued (for
review see11–15). Nevertheless, the development of transgenic mouse models of AD
over the last decade has primarily focused on the pathological markers (NFTs and
senile plaques), and such transgenic models have become promising tools to deci-
pher the mechanistic importance of tau phosphorylation and AG deposits, as well
their relationship between each other and the other pathological changes.
While seemingly obvious, it is important to remember that the validity of a
mouse model of disease is tightly linked to the ability of the animal to mimic the
signs of the disease—in the case of AD, cognitive decline. The aim of this review is
to discuss cognitive function in transgenic mouse models focused predominantly on
AG and tau models and, thereafter, the validity of these models to study AD and the
mechanistic questions that have arisen based on their behavioral phenotype.16,17
1.2 BEHAVIORAL TESTS

The most predominant and striking sign in an AD patient is the progressive decline
in cognition, primarily due to loss of neurons and synapses in the hippocampal for-
mation and related areas.18 As such, a “must have” feature of a valid AD-transgenic
model is the ability of the model to accurately reflect the behavioral changes observed
in human AD patients. To accurately interpret behavioral results from transgenic

Transgenic Mouse Models of Alzheimer’s Disease 3
mouse models of AD, it is important to intimately understand the behavioral tasks

that are most often used to test cognitive changes in mice, as well as what each cog-
nitive test is actually measuring. When examining cognition in animals, behavioral
tasks are typically divided into either associative or operant learning tasks. Associa-
tive learning tasks use cues in the environment to condition a specific response in the
animal. Operant learning tasks require the animal to make a particular response to a
specific stimulus in order to receive an outcome. Cognitive tasks are further divided
into groups by the type of memory being tested. The following are some of the most
often used tasks to determine cognitive changes in mouse models, transgenic or
otherwise.
1.2.1 SPATIAL MEMORY TASKS

1.2.1.1 The Morris Water Maze
The Morris water maze (MWM) is a particularly sensitive task to examine age-
related/AD-like deficits because it is highly specific for hippocampal function, one
of the first and most affected brain regions in AD.18 As a result, the MWM test is
one of the most common behavioral tasks used to determine hippocampal spatial
memory deficits.19 The test consists of placing the rodent in a circular tank filled with
cloudy water, which is used to motivate the animal to escape the water by swimming
to a hidden platform located right below the water’s surface. Over several days the
rodent learns to find the hidden platform by using spatial cues, such as posters or
taped objects strategically placed on the walls outside of the water maze, in the test-
ing room. Distance swam, latency to reach the platform, and swim speed, most often
recorded on video, are common measures of this test. The capacity of the animal to
retrieve and retain learned information or the flexibility to purge and relearn new
strategies can be determined using a probe trial and reversal trial. In the probe trial
the platform is taken out and the animals are allowed to swim in the pool. Time spent
in the region that previously contained the platform, crossings over the platform area,
and time to reach the platform location are measured. The reversal trial is identical
to the training trials, but in this case, the platform is switched to the opposite region
of the pool, testing the cognitive flexibility of the animal that is necessary to relearn
a new location. A cued version of this task, rendering the platform visible, can also
be used to measure nonspatial strategies as well as visual acuity.20 Variations include
the radial arm water maze (RAWM) or plus-shaped water maze.21
One desirable aspect of this task is that the motivating stimulus, i.e., escaping the
water, does not require the food or water deprivation that is common in other spatial
memory tasks. However, it has certain limitations as well, one of which is the fact
that the various components of memory, i.e., reference and working memories, can-
not be tested simultaneously.
1.2.1.2 Radial Arm Maze

One task that can accommodate simultaneous measurement of memory components
and has also been widely used to study spatial memory performance in rodents is the
radial arm maze (RAM). This maze consists of 8–17 equally spaced arms radiating

from a central platform, which the rodent has to enter in order to attain a food or
water reward placed in some of the arms. In this task, the animals guide themselves
using spatial cues around the room, with the goal to enter each arm only once to
receive the maximum amount of food or water rewards in the shortest period of time
and with the least amount of effort. This maze requires the use of working memory
to retain information that is important for a short time (within trial information), as
well as the use of reference memory to retain the general rules of the task across
days. Specifically, the animal must be able to remember which arms were baited
as well as which it already entered (working memory), but it also must know to
avoid non-baited arms across trials (reference memory), all of which takes place
by being able to successfully encode spatial information. However, while this task
permits the examination of both reference and working memory, major limitations
are the use of food or water deprivation in this task, as well as the presence of odor
confounds.22–24
1.2.1.3 Radial Arm Water Maze

A relatively new spatial memory task, the RAWM, has been designed to eliminate
the limitations of the above-mentioned tasks by combining the positive aspects of the
MWM and RAM. The difference between the MWM and RAWM is that performance
in the RAWM entails finding a platform that is submerged in water located in one of
several arms (6–8) in the water bath, compared to the classic MWM which only has an
open swim field. This makes the task a bit more difficult, but forces the animal to use
spatial cues and working memory (keeping track of the arms it has already visited) to
remember where the platform is located. Several variations of this task, using different
numbers of platforms and platform location organization, have been used to examine
spatial memory differences after pharmacological treatment25,26 and differences across
species,27 gender,28 and, importantly, models of AD.24,29
1.2.2 CONTEXTUAL MEMORY

1.2.2.1 Fear Conditioning
Freezing response, defined as a complete lack of movement, is the innate response of
rodents to fear. In a fear conditioning paradigm, the animal is placed in a box con-
taining a grid that delivers a mild aversive stimulus for two minutes. In the box, the
animal is presented with a tone (usually 80 dB) (conditioned stimulus) that is paired
with a mild shock (unconditioned stimulus) at the end of the trial with the result that
the tone elicits the freezing response. Repeated exposures are sometimes necessary
depending on the strain used or the interval time between the tone and the shock.
Some researchers use trace fear conditioning, which increases the time gap between
the tone and the shock in order to investigate prefrontal cortical activity. Here, the
animal is taken out of the box and returned 24 hr later to evaluate its learned aversion
for an environment associated with a mild aversive stimulus (context-dependent fear)
by measuring freezing behavior in the absence of tone or aversive stimulus. Cue-
dependent fear can be measured by placing the animal in a new box that is different

in color, shape, etc., and presenting it with the tone as it explores the new environ-
ment; freezing behavior associated with the tone is measured.
Fear conditioning is a widely used test to measure hippocampal-dependent asso-
ciative learning. This test is thought to be sensitive to emotion-associated learning
and therefore is a useful measure of amygdalar–hippocampal communication. Many
of the transgenic mouse models of AD display impairments in fear and anxiety,
which is primarily a function of the amygdala. The hippocampal function used in
fear conditioning may be different from learning in a spatial task.30–32
1.2.2.2 Passive-Avoidance Learning

In the passive-avoidance learning task, the animal must learn to avoid a mild aver-
sive stimulus, in this case darkness, by remaining in the well-lit side of a two-cham-
ber apparatus and not entering the dark where it receives the aversive stimulus. Note
that since rodents innately gravitate to darkness, the animal has to suppress this ten-
dency through pairing the negative stimulus with the desired compartment. Animals
that do not remember the aversive stimulus will cross over earlier than animals that
remember. Dependent measures include the median step-through latency (latency
to cross into the unsafe side) and the percentage of animals from each experimental
group that cross the threshold within an allocated time.20,33,34
1.2.3 WORKING MEMORY/NOVELTY/ACTIVITY

1.2.3.1 Y-Maze
This test is based on the innate preference of mice to alternate arms when explor-
ing a new environment. Various modifications are available with different levels of
difficulty and different demands on specific types of cognition. One version that is
particularly popular for the study of cognitive changes in AD transgenic models is
the spontaneous alternation version of the Y-maze. In this instance, test animals are
placed in a Y-shaped maze for 6–8 min and the number of arms entered, as well as
the sequence of entries, is recorded and a score is calculated to determine alternation
rate (degree of arm entries without repetitions). A high alternation rate is indicative
of sustained cognition as the animals must remember which arm was entered last to
not reenter it.35
A short-term memory version can also be carried out in which one arm of the
Y-maze is blocked and the subject is allowed to explore the two arms for 15–30 min.
The animal is then removed from the maze for a few minutes or up to several hours,
depending on the experimental manipulation, and then placed back into the maze,
this time with all arms open, to explore for 5 min. Animals with preserved cognitive
function will remember the previously blocked arm and will enter that one first on
the second trial. This test can also be repeated a week after the last trial with a delay
time of only 2 min between the trials in order to test long-term memory and the
time it takes the animal to relearn the task. Typically measured parameters include
the first arm entered, amount of time spent in each arm, and total number of arm
entries.35

1.2.3.2 T-Maze
T-maze tasks are incredibly well characterized and are widely used for cognitive
behavioral testing in both mice and rats. Animals are started at the base of the T
and allowed to choose one of the goal arms abutting the other end of the stem.
If two trials are given in quick succession, on the second trial the rodent tends to
choose the arm not visited before, reflecting memory of the first choice. This is
called “spontaneous alternation.” This tendency can be reinforced by making the
animal hungry and rewarding it with a preferred food if it alternates. Both spontane-
ous and rewarded alternations are very sensitive to dysfunction of the hippocampus,
and hence are sensitive to AD-like symptoms, but other brain structures are also
involved. Each trial should be completed in less than 2 min, but the total number of
trials required will vary according to statistical and scientific requirements.36
1.2.3.3 Object Recognition

The object recognition test is based on the natural tendency of rodents to investigate
a novel object instead of a familiar one, as well as their innate tendency to restart
exploring when they are presented with a novel environment. The choice to explore
the novel object, as well as the reactivation of exploration after object displacement,
reflects the use of learning and recognition memory processes. The available object-
recognition tasks to test cognition in rodents use different numbers of available
objects and environments in which the animals are tested, as well as types of config-
uration aimed to test spatial recognition and novelty, among other things. One par-
ticular object recognition task that is sensitive to age-related deficits is very suitable
to test AD-related deficits.37–39 In this task, a rodent is placed in a circular open field
filled with different objects (i.e., various plastic toys of different sizes and shapes)
for 6 min. After a series of trials, during which the animal has habituated to the con-
figuration and properties of the different objects, some of the objects are switched
from one location to another to assess spatial recognition. Subsequently, some of the
objects are replaced with new ones to evaluate novel object recognition. The time
spent exploring the open field (movement/inactivity) as well as number of times and
length of time inspecting each object over the different trials is calculated.
1.2.3.4 Open Field

The open field locomotion test is used primarily to examine motor function by
means of measuring spontaneous activity in an open field. The circular or square
open fields vary in size depending on the experiment and are divided into distinct
quadrants or sections. The animal is placed in the open field and the movements of
the animal are either videotaped or monitored by automated computer programs.
Rearing, line crosses, cleaning, general movement, number of lines crossed, prefer-
ence for particular sections, and/or fecal movements can all be calculated to examine
behavior and anxiety.40,41

1.3 TRANSGENIC MOUSE MODELS OF ALZHEIMER’S DISEASE

1.3.1 AMYLOID-G TRANSGENIC MOUSE MODELS
The first transgenic mouse model of AD, PDGF promoter expressing amyloid precur-
sor protein (PDAPP), was developed in 1995 by Games et al.42 and displays increased
human AG1–40 and AG1–42 that are 5–14 times higher than endogenous mouse AG.
The PDGF-driven mouse human amyloid precusor displays synaptic loss; reductions
in size of the hippocampus, fornix, and corpus callosum; and memory loss that is
comparable to that of human AD patients.43 PDAPP mice older than 6 mo have been
tested in the MWM, open field, radial arm maze, operant bar pressing, and visual
object recognition tasks and have significant memory impairments on all tasks
compared to age matched controls.29,44–46 For instance, when tested in the MWM,
PDAPP mice have significantly higher swim latencies in finding the platform than
controls. During open field trials and visual object recognition tasks, PDAPP mice
tend to exhibit high levels of motor activity and revisit already explored areas or
objects more often than control animals. PDAPP mice have further been tested in an
associative learning task—the fear conditioning task—which relies on the ability of
the animal to associate an auditory cue with a foot shock. After training, however,
both PDAPP and control animals display the same amount of freezing response after
the auditory cue is given, as well as when reexposed to the same training context.47
There are no studies to date that have examined how PDAPP mice perform if the
context is altered (to context B). A better indicator of cognitive deficits involving the
hippocampus compared to the auditory cue, which is primarily driven by the amyg-
dala, would be to perform this test in an altered context, such as a dark room with a
berry-scented odor, after initial training. No studies to date have explored different
contexts.
The deficits in cognition in the older PDAPP mice correlate with increased AG
and reductions in the hippocampus/brain ratio. However, in many cases, the same
cognitive deficits are also found in young (3–4 mo) animals in which AG deposits,
or hippocampal formation reduction, are not yet apparent.43 As such, these results
tend to not support the amyloid cascade hypothesis; however, the three types of mice,
C57Bl/6, DBA/2J, and Swiss-Webster, that are used to produce a PDAPP mouse,
are not the same for each study.43 This aspect will be further discussed, but it is
important to mention that this is a recurring problem in all transgenic mouse models
of AD. Another potential reason is that PDAPP mice tend to have lower body tem-
peratures, which may result in varying degrees of hypothermia during the MWM
task, which can produce amnesia in animals.4850 Although there are many theories
as to why young PDAPP mice perform like older PDAPP mice, the reason for the
inconsistencies in the literature is still unknown.
Shortly after the PDAPP mouse was developed, another human mutant APP
transgenic mouse model, which over-expresses the Swedish double mutant form of
APP695, was introduced as the Tg2576 mouse.51 Tg2576 mice are similar to the
PDAPP mouse in that they exhibit five times the level of endogenous murine APP
in the brain and, after 11 mo, develop plaque-like deposits of AG1–40 and AG1–42/43
in the frontal, temporal, and entorhinal cortices; hippocampus; presubiculum; and
cerebellum. Unlike the PDAPP mice, Tg2576 mice do not have significant synaptic

loss or reductions in hippocampal size. Tg2576 mice have been tested in many of the
same tasks as the PDAPP mice. For example, mice at 2, 6, 9, and 12 mo of age have
been tested in the MWM. In this regard, after 10 mo of age, this transgenic mouse
line demonstrates poor spatial memory retention and is unable to find a visible plat-
form after 2 and 4 days of training compared to controls.43,52 Tg2576 mice also tend
to explore the familiar arm of the Y-maze more than controls. As with the PDAPP
mice, Tg2576 mice are not always cognitively impaired. King and Arendash53 did
not see cognitive deficits in the MWM task in young or old animals, but did find that
the Tg2576 mice had sensorimotor deficits during the visual cue trials. However,
many studies that did not find a difference in cognition between Tg2576 mice and
controls did find a significant decline in memory if they eliminated animals that
showed visual and/or motor deficits.43,54,55
Tg2576 mice have also been tested in a variety of Pavlovian tasks, such as fear
conditioning. Older mice first trained in a salient context (context A) were then
divided into subgroups, one of which was again tested in context A and the other in a
novel context (context B). Based on the context-shift theory, normal animals perform
well when trained and tested in the same context, but show a decline in memory
when tested in the new context if they acquired the memory for the original training
cues.56 Conversely, Tg2576 mice performed well in both contexts, unaffected by the
change in cues, most likely because they were unable to remember the cues from the
original training context. Although, Tg2576 mice did not distinguish between con-
textual cues, they were able to learn the fear response when trained with a specific
cue, such as a sound or light. These results are consistent with the PDAPP mice.57
The accumulation of AG1–42 is dependent upon the cleavage of the G-secretase
and the L-secretase enzymes. Individual enzymes known as presenilins are involved
in L-secretase enzyme activity, and mutations in presenilins often lead to AG1–42
accumulation as found in AD patients.58 The first mouse model to examine the role
of presenilin 1 (PS1) was produced by Shen et al.59 The PS1 knockout mice were
deficient in PS1; however, they quickly died after birth. Massive neuronal loss and
hemorrhages were found in the brain. Today there are a few types of PS1 and PS2
transgenic mouse models that survive after birth. All of the models that lack the PS1
or PS2 gene demonstrate cognitive decline on the MWM and on object recognition
tasks, but compared to the Tg2576 animals, they are not severely impaired. In ani-
mals that over-express human PS1, high levels of AG1–42 were found, but without
accompanying plaque-like accumulations or behavioral alterations.43,60
The first multiple gene transgenic mouse model of AD was developed to alter
both the presenilins and the accumulation of human APP, today known as APP+PS1.
Compared to the Tg2576 animals, the APP+PS1 has levels of AG1–40 five times higher
by 6 mo of age.61 Young and old APP+PS1 mice have been tested in the Y-maze, ele-
vated plus maze, MWM, and RAWM. Both young and old animals display deficits
in the spontaneous alternation version of the Y-maze task, with fewer alternations
between arms on the Y-maze. However, in the other behavioral tests, young animals
tend to perform as well as controls, but by 15–17 mo of age, the APP+PS1 animals
showed spatial deficits in the MWM and RAWM, and increased activity in an open-
field test. This was one of the first transgenic mouse models that showed a strong
positive correlation between AG1–42 development and cognitive decline.61–63

A second multi-gene PS1/APP mouse model, known as the PSAPP mouse, was
developed from a different mutation in human PS1 (A246E) and was crossed with
the Tg2576 mouse. AG1–42 and plaque loads occurred as early as 7 mo, earlier than
in the Tg2576 mice. PSAPP mice perform similarly to the Tg2576 mice in a cued
fear conditioning paradigm. Notably, PSAPP mice perform well at this test if they
are given a cue, but are unable to distinguish altered contexts between training and
testing, suggesting a hippocampal deficit. During spatial MWM testing, PSAPP
mice have longer latencies to find the hidden platform, which is significantly cor-
related with the levels of insoluble hippocampal AG1–42.64
The CRND8 transgenic mouse is derived from an APP Swedish mutation and
V717F mice. Plaque formation develops in the hippocampus and cortex around 9
wk of age. They differ from the PDAPP and Tg2576 mice in that they have dense
core deposits and dystrophic neurites without hippocampal volume decreases. In an
MWM test, CRND8 mice perform worse than controls when tested after plaques
have developed.65 Hyde et al.66 confirmed that AG production occurs prior to the for-
mation of plaques, and therefore animals at the pre-plaque, early/mid-plaque stage,
and late-plaque stage were tested in the MWM. Pre-plaque animals perform as well
as controls; however, both early/mid-plaque and late-plaque animals have deficits in
swim time. Further, early/mid-plaque animals perform well on the probe trial, while
the late-plaque animals do not.66
A more recent transgenic mouse model is the PDGF-APPSw,Ind mouse, which
expresses the Swedish and Indiana APP mutations with increased BrdU and imma-
ture neuronal markers in the dentate gyrus and subventricular zone.67 This increased
neurogenesis is also found in the brains of patients with AD. While neurogenesis in
AD may be a result of increased injured neurons or the loss of neurons, the PDGF-
APPSw,Ind, however, do not display neuronal loss, indicating that another mechanism
is responsible for the neurogenesis.68
1.3.2 TAU TRANSGENIC MOUSE MODELS

Tau, a microtubule protein, is modified in AD, resulting in neuronal degeneration.
Tau transgenic mouse models have been designed to model the NFT pathology often
observed in AD. Early over-expressing tau transgenic mouse models demonstrate
motor deficits and cell loss in the spinal cord; however, they did not develop “true”
NFTs. Recently, a new tau transgenic mouse, P301S, was developed that demon-
strates progressive NFT formation and neuronal loss,69 and NFT formation is found
in the spinal cord, brainstem, cerebellum, diencephalon, and basal telencephalon
when the expression of FTDP-17–associated mutation P301L is increased under the
mouse prion promoter (JNPL3 line). Significantly, increased NFTs are correlated
with a decline in MWM performance.70 Other mouse models using the P301S FTDP-
17–associated mutation, which show human tau is expressed in the spinal cord and
hippocampus under the mouse Thy1.2 promoter, have phosphorylated tau, but do not
have NFTs.71
The rTg(tauP301L) 4510 mouse expresses the P301L mutation in tau associated
with frontotemporal dementia and develops NFTs in the neocortex and hippocam-
pus, which is consistent with other tau transgenic mice. When tested in the MWM,

cognitive decline can be seen as early as 4 mo of age in the rTg(tauP301L) 4510 mice
and this memory deficit is also accompanied by neuronal loss.72 The development of
the P301S mouse allows researchers to examine the effect of AG on NFT formation.
Injections of AG1–42 into the hippocampus of P301L mice leads to increased NFT
numbers in the amygdala and hippocampus.73 The best model for AD, however, com-
bines amyloid and tau pathology by crossing the APP, PS1, and P301L genes. The
result is the 3xTg-AD mouse model, which develops amyloid plaques that quickly
develop first in the neocortex and then spread to the hippocampus, and then develops
NFT in the hippocampus shortly after the appearance of amyloid pathology. The
3xTg-AD mice display long-term potentiation deficits and precede plaque and tangle
formation. A reduction in plaques and tangles occurs in response to immunotherapy
treatment with AG antibodies.74 Cognitive testing of both male and female 3xTg-AD
using the MWM and passive-avoidance tests displayed impairments by females and
males by 4.5 mo of age. There were no differences between controls and 3xTg-AD
on the object recognition task.75
1.4 CONCERNS WITH TRANSGENIC MOUSE

MODELS OF ALZHEIMER’S DISEASE
Transgenic mouse models allow us to examine the mechanisms involved in the devel-
opment of diseases such as AD. However, because of the large discrepancy in the
behavioral findings observed across the now plentiful number of AD mouse models,
a simple question that arises is whether we are really any closer today to determining
what these mechanisms are than when the first PDAPP mouse was produced. The
majority of AD research is carried out using animal models that have increased AG
levels compared to controls, and while AG pathology is mimicked in these models,
many other factors associated with AD pathology are not. For instance, as described
above, many of the transgenic models, such as the Tg2576 and PS1+APP mice, do
not have neuronal loss or larger ventricles, as would be expected in a true model of
AD.51,54 We also cannot disregard the many AD mouse models that have increases
in AG or APP, but do not demonstrate cognitive deficits.43,48,54,60 Inconsistencies in
the literature could be due to differences in the behavioral protocols, type of tests
that were conducted, age of the animals, the genetic background the transgenic ani-
mals were designed on, timing, sleep cycle of the animals, etc. Researchers often
use a standard behavioral protocol, but the age of the animals, environmental cues,
changes in researchers during the study, timing, techniques, handling, and time of
day are difficult to keep constant from one lab to another. Any and all of these fac-
tors can affect behavioral outcomes. Likewise, the background of animals used for
the transgenic mouse AD-model design influences how the animals will perform
on various behavioral tasks.76 For instance, Pugh et al.76 found differences in learn-
ing of the passive avoidance task and MWM in two strains of mice (FVB/N and
C57BL6/J) that are often used to engineer transgenic mouse lines. This information
is an important factor when designing experiments and evaluating cognition testing.
A lack of behavioral differences should not preclude the manipulated target from
playing a role in the disease.

Another argument that further complicates the use of animal models based
solely on APP and/or tau mutations is that other mechanisms may be at play.77 In this
regard, the possibility that AG production and tau hyperphosphorylation are compen-
satory responses to other pathogenic mechanisms such as cell cycle dysregulation or
oxidative stress has not been excluded.13,78 For example, oxidative stress as mea-
sured by 8-hydroxyguanosine (8OHG) and nitrotyrosine adduct formation, precedes
AG deposition by decades in Down’s syndrome and AD patients.79–83 Moreover, the
pathological lesions in the brains of patients with AD are associated with decreased
oxidative markers compared to histologically unaffected but vulnerable neurons.16
Similarly, in Down’s syndrome, 8OHG immunoreactivity increases significantly in
the teens and twenties, while AG burden only increases after age 30.79 Tau accu-
mulation may also be an indicator of an oxidative imbalance. Oxidative stress and
attendant modifications of tau byproducts of oxidative stress include Hydroxy-2,3-
nonenal (4-HNE) and other cytotoxic carbonyls, which may enable neurons modi-
fied by tau and neurofilament proteins to survive for decades.83
Mechanistic questions aside, the fact that studying AD via the use of mouse
models carrying specific familial mutations to pathological entities of the disease
(AG, tau hyperphosphorylation) may only provide a partial view rather than a com-
plete picture of this disease.16 As such, some stereological studies have suggested
that there may be little or no neuronal loss during “normal” aging, even though the
number of plaques is increased.84 This observation parallels that observed in many
of the transgenic mouse models. Importantly, like their human counterparts, these
mice show evidence of oxidative stress that precedes the AG deposits.85,86 Also, due
to the fact that AG may be an end product of an underlying cause of AD, researchers
using transgenic AD models may ultimately be examining a later stage of AD, when
cognitive decline is seen. Nevertheless, some reports of neuronal loss in various
transgenic AD models argue that AG is a bioactive substance. Furthermore, because
these models are based on mutations associated with early onset AD, careful evalu-
ation is needed to determine whether they provide a compelling analogy to sporadic
AD in humans, which comprises 95% of the cases. To address this issue, perhaps,
animal models of aging rather than mutation-specific models may afford a more
accurate picture of how all of these pathogenic entities interact for the development
and progression of AD.87,88
In conclusion, the development of transgenic models of AD may provide tools
to achieve an understanding of pathogenic mechanisms and develop new therapies.
The efforts in this respect with regard to AD have been monumental, with several
transgenic lines being available to researchers (Table 1.1). However, the validity of
these models is overwhelmingly based on the ability of over-expression of APP and
tau mutations to cause the pathological inclusions observed in the AD brain (plaques
and NFTs); however, work is still needed to transfer this validity to other events-asso-
ciated AD pathology. As such, AD transgenic mice differ in the timing and level of

12
TABLE 1.1
Transgenic Mouse Models of Alzheimer’s Disease
Model Gene Promoter Strain Pathology MWM RAM Fear Passive Y-Maze Object Open Field
Name Background Conditioning Avoidance Recognition
PDAPP hAPP, PDGF-G Swiss- Webster, Plaques, ptau Impaired, poor Impaired, Good Impaired, visited High levels of
Methods of Behavior Analysis in Neuroscience, Second Edition

V717f C57B6, DBA spatial memory, continued performance already motor activity,
unable to find to visit to auditory cue explored areas/ visited already
platform (6–9 same arms (11 mo) 47 objects (6, 9– explored areas
mo) 44 (3 mo) 45 10 mo) 45 and objects
(10 and 16
mo) 46
Tg2576 Swedish Hamster C57B6, SJL Plaques, ptau Impaired, poor Impaired, did Normal at 3 Impaired, Normal,
PrP spatial memory, not distinguish and 9 mo 89 explored explored new
unable to find between familiar arm objects more
platform (3 and training and more than than old
9 mo) 51,55 testing cues controls (10, objects 54
(1618 mo) 57 16–18 mo) 54
hPS1-2 Tg Swedish Hamster C57B6, Few plaques Not impaired (6– Impaired at
PrP PVBx129S6 17 mo) 43,60 15 mo,
continued
to visit
same arms
(12 mo) 43,60
APP+PS1 Swedish Hamster Swiss-Webster, Plaques Impaired, poor Impaired, did not Increased
PrP B6D2F1 spatial memory, alternate activity (15–
unable to find between arms 17 mo) 29,63
platform (15 (5–14 wk)29,61,63
mo) 90

Transgenic Mouse Models of Alzheimer’s Disease
PSAPP Swedish Hamster C3H/B6 Plaques Impaired, poor Impaired, did Impaired, prior
PrP spatial memory, not distinguish to amyloid
took longer time between accumulation
to find platform training and (3 mo) 61,62
(14 mo) 91 testing cues,
but performed
well if given
an auditory
cue (5 and 9
mo) 64
Tg Swedish Hamster C3H/C57B6 Plaques Impaired during
CRND8 PrP mid to late
plaque
development
(6–17 mo) 39,65
JNPL3 Tau PrP C57/BL Tau Impaired with
NFTs increasing levels
of pTau (5 mo)
70
rTg Tau PrP FVB/N Tau Impaired, little to

(tauP30 NFTs no retention of
1L) 4510 platform by 9.5
mo of age (4.5
mo) 69,72
3xTg-AD Swedish Thy1 Plaques Impaired (4.5 Impaired, No impairment
(PS1 Tau mo) 75 shorter (4.5 mo) 75
knockin) NFTs latencies to
cross to dark
side than
controls (4.5
mo) 75
13
AG, PS1/2, and tau accumulation, and not all of the animals demonstrate neuronal
cell loss, or hippocampal atrophy and ventricular enlargement. More importantly,
cognitive decline is not always correlated with AG deposits or NFT formation. AD
pathogenesis is likely a syndrome rather than a disease of specific mutations. There-
fore, full validation of an AD model will only be recognized when features of AD
beyond tau and AG are incorporated in the models.
REFERENCES
1. Alzheimer, A. 1907. Uber eine eigenartige Erkrankung der Hirnrinde. Allg. Zeitschr.
Psychiatr. 64:146–8.
2. Smith, M. A. 1998. Alzheimer disease. Int. Rev. Neurobiol. 42:1–54.
3. Glenner, G. G., and Wong, C. W. 1984. Alzheimer‘s disease and Down‘s syndrome:
Sharing of a unique cerebrovascular amyloid fibril protein. Biochem. Biophys. Res.
Commun. 122:1131–35.
4. Hardy, J., and Selkoe, D. J. 2002. The amyloid hypothesis of Alzheimer’s disease: Prog-
ress and problems on the road to therapeutics. Science 297:353–56.
5. Knowles, R. B., Gomez-Isla, T., and Hyman, B. T. 1998. AG associated neurophil
changes: Correlation with neuronal loss and dementia. J. Neuropathol. Exp. Neurol.
57:1122–30.
6. Hardy, J. A., and Higgins, G. A. 1992. Alzheimer’s disease: The amyloid cascade
hypothesis. Science 256:184–85.
7. Walsh, D. M., and Selkoe, D. J. 2004. Oligomers on the brain: The emerging role of
soluble protein aggregates in neurodegeneration. Protein Pept. Lett. 11:213–28.
8. Tiraboschi, P., Sabbagh, M. N., Hansen, L. A., et al. 2004. Alzheimer disease without
neocortical neurofibrillary tangles: ”A second look”. Neurology 62:1141–47.
9. Arriagada, P. V., Growdon, J. H., Hedley-Whyte, E. T., and Hyman, B. T. 1992. Neu-
rofibrillary tangles but not senile plaques parallel duration and severity of Alzheimer’s
disease. Neurology 42:631–39.
10. Roberson, E. D., Scearce-Levie, K., Palop, J. J., et al. 2007. Reducing endogenous tau
ameliorates amyloid beta-induced deficits in an Alzheimer’s disease mouse model. Sci-
ence 316:750–54.
11. Raina, A. K., Zhu, X., and Smith, M. A. 2004. Alzheimer’s disease and the cell cycle.
Acta Neurobiol. Exp. (Wars) 64:107–12.
12. Casadesus, G., Atwood, C. S., Zhu, X., et al. 2005. Evidence for the role of gonadotropin
hormones in the development of Alzheimer disease. Cell. Mol. Life Sci. 62:293–98.
13. Zhu, X., Raina, A. K., Perry, G., and Smith, M. A. 2004. Alzheimer’s disease: The two-
hit hypothesis. Lancet Neurol. 3:219–26.
14. Smith, M. A., Rottkamp, C. A., Nunomura, A., Raina, A. K., and Perry, G. 2000. Oxi-
dative stress in Alzheimer’s disease. Biochim. Biophys. Acta 1502:139–44.
15. Sayre, L. M., Perry, G., Atwood, C. S., and Smith, M. A. 2000. The role of metals in
neurodegenerative diseases. Cell. Mol. Biol. (Noisy-le-grand) 46:731–41.
16. Castellani, R. J., Lee, H. G., Zhu, X., et al. 2006. Neuropathology of Alzheimer disease:
Pathognomonic but not pathogenic. Acta Neuropathol. (Berl). 111:503–9.
17. Berg, L., McKeel, D. W. Jr., Miller, J. P., Baty, J., and Morris, J. C. 1993. Neuropatho-
logical indexes of Alzheimer’s disease in demented and nondemented persons aged 80
years and older. Arch. Neurol. 50:349–58.
18. West, M. J. 1993. Regionally specific loss of neurons in the aging human hippocampus.
Neurobiol. Aging 14:287–93.
19. Morris, R. 1984. Developments of a water-maze procedure for studying spatial learning
in the rat. J. Neurosci. Methods 11:47–60.

20. Lawlor, P. A., Bland, R. J., Das, P., et al. 2007. Novel rat Alzheimer’s disease models
based on AAV-mediated gene transfer to selectively increase hippocampal AG levels.
Mol. Neurodegener. 2:11.
21. Vloeberghs, E., Van Dam, D., D’Hooge, R., Staufenbiel, M., and De Deyn, P. P. 2006.
APP23 mice display working memory impairment in the plus-shaped water maze. Neu-
rosci. Lett. 407:6–10.
22. de Toledo-Morrell, L., Morrell, F., and Fleming, S. 1984. Age-dependent deficits
in spatial memory are related to impaired hippocampal kindling. Behav. Neurosci.
98:902–7.
23. Ikegami, S. 1994. Behavioral impairment in radial-arm maze learning and acetylcho-
line content of the hippocampus and cerebral cortex in aged mice. Behav. Brain Res.
65:103–11.
24. Morgan, D., Diamond, D. M., Gottschall, P. E., et al. 2000. A beta peptide vacci-
nation prevents memory loss in an animal model of Alzheimer’s disease. Nature
408:982–85.
25. Shear, D. A., Dong, J., Haik-Creguer, K. L., et al. 1998. Chronic administration of
quinolinic acid in the rat striatum causes spatial learning deficits in a radial arm water
maze task. Exp. Neurol. 150:305–11.
26. Bimonte, H. A., and Denenberg, V. H. 1999. Estradiol facilitates performance as work-
ing memory load increases. Psychoneuroendocrinology 24:161–73.
27. Hyde, L. A., Hoplight, B. J., and Denenberg, V. H. 1998. Water version of the radial-arm
maze: Learning in three inbred strains of mice. Brain Res. 785:236–44.
28. Bimonte, H. A., Hyde, L. A., Hoplight, B. J., and Denenberg, V. H. 2000. In two spe-
cies, females exhibit superior working memory and inferior reference memory on the
water radial-arm maze. Physiol. Behav. 70:311–17.
29. Arendash, G. W., Gordon, M. N., Diamond, D. M., et al. 2001. Behavioral assessment
of Alzheimer’s transgenic mice following long-term Abeta vaccination: Task specific-
ity and correlations between Abeta deposition and spatial memory. DNA Cell Biol.
20:737–44.
30. Fanselow, M. S. 1980. Conditioned and unconditional components of post-shock freez-
ing. Pavlov. J. Biol. Sci. 15:177–82.
31. Fanselow, M. S., and Tighe, T. J. 1988. Contextual conditioning with massed versus
distributed unconditional stimuli in the absence of explicit conditional stimuli. J. Exp.
Psychol. Anim. Behav. Process. 14:187–99.
32. Hamann, S., Monarch, E. S., and Goldstein, F. C. 2002. Impaired fear conditioning in
Alzheimer’s disease. Neuropsychologia 40:1187–95.
33. Senechal, Y., Kelly, P. H., and Dev, K. K. 2008. Amyloid precursor protein knockout
mice show age-dependent deficits in passive avoidance learning. Behav. Brain Res.
186:126–32.
34. McGaugh, J. L. 1966. Time-dependent processes in memory storage. Science
153:1351–58.
35. Jackson, L. L. 1943. V.T.E. on an elevated maze. J. Comp. Psychol. 36:99–107.
36. Deacon, R. M., and Rawlins, J. N. 2006. T-maze alternation in the rodent. Nature pro-
tocols 1:7–12.
37. Shukitt-Hale, B., Casadesus, G., Cantuti-Castelvetri, I., and Joseph, J. A. 2001. Effect
of age on object exploration, habituation, and response to spatial and nonspatial change.
Behav. Neurosci. 115:1059–64.
38. Casadesus, G., Shukitt-Hale, B., Cantuti-Castelvetri, I., Rabin, B. M., and Joseph, J. A.
2004. The effects of heavy particle irradiation on exploration and response to environ-
mental change. Adv. Space. Res. 33:1340–46.

39. Janus, C., Pearson, J., McLaurin, J., et al. 2000. A beta peptide immunization reduces
behavioural impairment and plaques in a model of Alzheimer’s disease. Nature
408:979–82.
40. Hall, C. S. 1934. Emotional behavior in the rat: Defecation and urination as measures
of individual differences in emotionality. J. Comp. Psychol. 18:385–403.
41. Hrnkova, M., Zilka, N., Minichova, Z., Koson, P., and Novak, M. 2007. Neurodegen-
eration caused by expression of human truncated tau leads to progressive neurobehav-
ioural impairment in transgenic rats. Brain Res. 1130:206–13.
42. Games, D., Adams, D., Alessandrini, R., et al. 1995. Alzheimer-type neuropathol-
ogy in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature
373:523–27.
43. Kobayashi, D. T., and Chen, K. S. 2005. Behavioral phenotypes of amyloid-based
genetically modified mouse models of Alzheimer’s disease. Genes, Brain, and Behav-
ior 4:173–96.
44. Chen, G., Chen, K. S., Knox, J., et al. 2000. A learning deficit related to age and beta-
amyloid plaques in a mouse model of Alzheimer’s disease. Nature 408:975–79.
45. Dodart, J. C., Meziane, H., Mathis, C., et al. 1999. Behavioral disturbances in trans-
genic mice overexpressing the V717F beta-amyloid precursor protein. Behav. Neurosci.
113:982–90.
46. Morgan, D. 2003. Learning and memory deficits in APP transgenic mouse models of
amyloid deposition. Neurochem. Res. 28:1029–34.
47. Gerlai, R., Fitch, T., Bales, K. R., and Gitter, B. D. 2002. Behavioral impairment
of APP(V717F) mice in fear conditioning: Is it only cognition? Behav. Brain Res.
136:503–9.
48. Justice, A., and Motter, R. 1997. Behavioral characterization of PDAPP transgenic
Alzheimer mice. Soc. Neurosci. Abstr. 23:1637.
49. Rauch, T. M., Welch, D. I., and Gallego, L. 1989. Hypothermia impairs performance in
the Morris water maze. Physiol. Behav. 46:315–20.
50. Richardson, R., Riccio, D. C., and Morilak, D. 1983. Anterograde memory loss induced
by hypothermia in rats. Behav. Neural Biol. 37:76–88.
51. Hsiao, K., Chapman, P., Nilsen, S., et al. 1996. Correlative memory deficits, Abeta
elevation, and amyloid plaques in transgenic mice. Science 274:99–102.
52. Takeuchi, A., Irizarry, M. C., Duff, K., et al. 2000. Age-related amyloid beta deposition
in transgenic mice overexpressing both Alzheimer mutant presenilin 1 and amyloid
beta precursor protein Swedish mutant is not associated with global neuronal loss. Am.
J. Pathol. 157:331–39.
53. King, D. L., and Arendash, G. W. 2002. Behavioral characterization of the Tg2576 trans-
genic model of Alzheimer’s disease through 19 months. Physiol. Behav. 75:627–42.
54. Arendash, G. W., and King, D. L. 2002. Intra- and intertask relationships in a behavioral
test battery given to Tg2576 transgenic mice and controls. Physiol. Behav. 75:643–52.
55. Westerman, M. A., Cooper-Blacketer, D., Mariash, A., et al. 2002. The relationship
between Abeta and memory in the Tg2576 mouse model of Alzheimer’s disease. J.
Neurosci. 22:1858–67.
56. Riccio, D. C., Richardson, R., and Ebner, D. L. 1984. Memory retrieval deficits based
upon altered contextual cues: A paradox. Psychol. Bull. 96:152–65.
57. Corcoran, K. A., Lu, Y., Turner, R. S., and Maren, S. 2002. Overexpression of hAPPswe
impairs rewarded alternation and contextual fear conditioning in a transgenic mouse
model of Alzheimer’s disease. Learn. Mem. 9:243–52.
58. Brunkan, A. L., and Goate, A. M. 2005. Presenilin function and gamma-secretase
activity. J. Neurochem. 93:769–92.
59. Shen, J., Bronson, R. T., Chen, D. F., et al. 1997. Skeletal and CNS defects in Presenilin-
1-deficient mice. Cell 89:629–39.

60. Spires, T. L., and Hyman, B. T. 2005. Transgenic models of Alzheimer’s disease:
Learning from animals. NeuroRx 2:423–37.
61. Holcomb, L., Gordon, M. N., McGowan, E., et al. 1998. Accelerated Alzheimer-type
phenotype in transgenic mice carrying both mutant amyloid precursor protein and pre-
senilin 1 transgenes. Nat. Med. 4:97–100.
62. Holcomb, L. A., Gordon, M. N., Jantzen, P., et al. 1999. Behavioral changes in trans-
genic mice expressing both amyloid precursor protein and presenilin-1 mutations: Lack
of association with amyloid deposits. Behav. Genet. 29:177–85.
63. Arendash, G. W., King, D. L., Gordon, M. N., et al. 2001. Progressive, age-related
behavioral impairments in transgenic mice carrying both mutant amyloid precursor
protein and presenilin-1 transgenes. Brain Res. 891:42–53.
64. Dineley, K. T., Xia, X., Bui, D., Sweatt, J. D., and Zheng, H. 2002. Accelerated plaque
accumulation, associative learning deficits, and up-regulation of alpha 7 nicotinic
receptor protein in transgenic mice co-expressing mutant human presenilin 1 and amy-
loid precursor proteins. J. Biol. Chem. 277:22768–780.
65. Chishti, M. A., Yang, D. S., Janus, C., et al. 2001. Early-onset amyloid deposition and
cognitive deficits in transgenic mice expressing a double mutant form of amyloid pre-
cursor protein 695. J. Biol. Chem. 276:21562-570.
66. Hyde, L. A., Kazdoba, T. M., Grilli, M., et al. 2005. Age-progressing cognitive
impairments and neuropathology in transgenic CRND8 mice. Behav. Brain Res.
160:344–55.
67. Jin, K., Galvan, V., Xie, L., et al. 2004. Enhanced neurogenesis in Alzheimer’s disease
transgenic (PDGF-APPSw,Ind) mice. Proc. Natl. Acad. Sci. U. S. A. 101:13363-367.
68. Casadesus, G., Zhu, X., Lee, H. G., et al. 2006. Neurogenesis in Alzheimer’s disease:
Compensation, crisis, or chaos? In The cell cycle in the central nervous system, ed. D.
Janigro, 359–70. Totowa: Humana Press.
69. Gotz, J., Chen, F., Barmettler, R., and Nitsch, R. M. 2001. Tau filament formation in
transgenic mice expressing P301L tau. J. Biol. Chem. 276:529–34.
70. Ramsden, M., Kotilinek, L., Forster, C., et al. 2005. Age-dependent neurofibrillary
tangle formation, neuron loss, and memory impairment in a mouse model of human
tauopathy (P301L). J. Neurosci. 25:10637–647.
71. Allen, B., Ingram, E., Takao, M., et al. 2002. Abundant tau filaments and nonapoptotic
neurodegeneration in transgenic mice expressing human P301S tau protein. J. Neuro-
sci. 22:9340–51.
72. Gotz, J., Chen, F., van Dorpe, J., and Nitsch, R. M. 2001. Formation of neurofibril-
lary tangles in P301l tau transgenic mice induced by Abeta 42 fibrils. Science
293:1491–95.
73. Oddo, S., Caccamo, A., Kitazawa, M., Tseng, B. P., and LaFerla, F. M. 2003. Amyloid
deposition precedes tangle formation in a triple transgenic model of Alzheimer’s dis-
ease. Neurobiol. Aging 24:1063–70.
74. Oddo, S., Caccamo, A., Shepherd, J. D., et al. 2003. Triple-transgenic model of
Alzheimer’s disease with plaques and tangles: Intracellular Abeta and synaptic dys-
function. Neuron 39:409–21.
75. Clinton, L. K., Billings, L. M., Green, K. N., et al. 2007. Age-dependent sexual
dimorphism in cognition and stress response in the 3xTg-AD mice. Neurobiol. Dis.
28:76–82.
76. Pugh, P. L., Ahmed, S. F., Smith, M. I., Upton, N., and Hunter, A. J. 2004. A behav-
ioural characterisation of the FVB/N mouse strain. Behav. Brain Res. 155:283–89.
77. Lee, H. G., Casadesus, G., Zhu, X., et al. 2004. Challenging the amyloid cascade
hypothesis: Senile plaques and amyloid-beta as protective adaptations to Alzheimer
disease. Ann. N. Y. Acad. Sci. 1019:1–4.

78. Nunomura, A., Perry, G., Pappolla, M. A., et al. 1999. RNA oxidation is a prominent
feature of vulnerable neurons in Alzheimer’s disease. J. Neurosci. 19:1959–64.
79. Nunomura, A., Perry, G., Pappolla, M. A., et al. 2000. Neuronal oxidative stress
precedes amyloid-beta deposition in Down syndrome. J. Neuropathol. Exp. Neurol.
59:1011–17.
80. Nunomura, A., Perry, G., Aliev, G., et al. 2001. Oxidative damage is the earliest event
in Alzheimer disease. J. Neuropathol. Exp. Neurol. 60:759–67.
81. Nunomura, A., Chiba, S., Lippa, C. F., et al. 2004. Neuronal RNA oxidation is a promi-
nent feature of familial Alzheimer’s disease. Neurobiol. Dis. 17:108–13.
82. Odetti, P., Angelini, G., Dapino, D., et al. 1998. Early glycoxidation damage in brains
from Down’s syndrome. Biochem. Biophys. Res. Commun. 243:849–51.
83. Long, J. M., Mouton, P. R., Jucker, M., and Ingram, D. K. 1999. What counts in brain
aging? Design-based stereological analysis of cell number. J. Gerontol. A. Biol. Sci.
Med. Sci. 54:B407–17.
84. Calhoun, M. E., Wiederhold, K. H., Abramowski, D., et al. 1998. Neuron loss in APP
transgenic mice. Nature 395:755–56.
85. Morsch, R., Simon, W., and Coleman, P. D. 1999. Neurons may live for decades with
neurofibrillary tangles. J. Neuropathol. Exp. Neurol. 58:188–97.
86. Smith, M. A., Hirai, K., Hsiao, K., et al. 1998. Amyloid-beta deposition in Alzheimer
transgenic mice is associated with oxidative stress. J. Neurochem. 70:2212–15.
87. Wei, X., Zhang, Y., and Zhou, J. 1999. Alzheimer’s disease-related gene expression in
the brain of senescence accelerated mouse. Neurosci. Lett. 268:139–42.
88. Butterfield, D. A., and Poon, H. F. 2005. The senescence-accelerated prone mouse
(SAMP8): A model of age-related cognitive decline with relevance to alterations of
the gene expression and protein abnormalities in Alzheimer’s disease. Exp. Gerontol.
40:774–83.
89. King, D. L., Arendash, G. W., Crawford, F., et al. 1999. Progressive and gender-depen-
dent cognitive impairment in the APP(SW) transgenic mouse model for Alzheimer’s
disease. Behav. Brain Res. 103:145–62.

2 Cued and Contextual
Fear Conditioning
for Rodents
Peter Curzon, Nathan R. Rustay,
and Kaitlin E. Browman
CONTENTS
2.1 Introduction...................................................................................................20
2.2 Contextual/Cued Fear Conditioning: Overview ........................................... 21
2.2.1 Contextual Fear Conditioning............................................................ 21
2.2.2 Cued Fear Conditioning..................................................................... 21
2.2.3 Delay and Trace Conditioning ........................................................... 21
2.3 Brain Areas Involved .................................................................................... 22
2.3.1 Amygdala ........................................................................................... 22
2.3.2 Hippocampus ..................................................................................... 22
2.3.3 Frontal/Ventromedial/Cingulate Cortex............................................ 22
2.4 Before Getting Started .................................................................................. 23
2.4.1 Types of Paradigms............................................................................ 23
2.4.1.1 Contextual/Cued Fear Conditioning .................................... 23
2.4.1.2 Contextual Conditioning ...................................................... 23
2.4.1.3 Delay/Cue Fear Conditioning...............................................24
2.4.1.4 Trace Fear Conditioning.......................................................24
2.4.1.5 Backward Trace Conditioning..............................................24
2.5 Sample Experiments .....................................................................................24
2.5.1 Delay Cued and Contextual Fear Conditioning.................................24
2.5.1.1 Day 1 ....................................................................................25
2.5.1.2 Day 2 ....................................................................................26
2.5.2 Trace Cued and Contextual Fear Conditioning .................................26
2.5.3 Contextual Fear Conditioning............................................................ 27
2.6 Data Analysis ................................................................................................ 27
2.7 Sample Data .................................................................................................. 27
2.8 Nonassociative Freezing Complications ....................................................... 29
2.9 Final Note...................................................................................................... 31
2.10 Addendum ..................................................................................................... 31
2.10.1 Available Equipment Options ............................................................ 31
2.10.2 Conditioning Chambers ..................................................................... 32
2.10.3 Considerations for the US (Shockers) ................................................ 33
19

2.10.4 Cued Conditioning Measurement ...................................................... 35

2.10.5 Measuring Freezing Behavior............................................................ 35
2.10.5.1 Hand Tallying Method ......................................................... 35
2.10.5.2 Activity Monitoring Method ................................................ 36
2.10.6 Animal Strain Considerations ........................................................... 36
2.10.7 Auditory Cue Considerations............................................................. 36
References................................................................................................................ 37
2.1 INTRODUCTION
Understanding what an animal learns when exposed to novelty is of great interest
to behavioral neuroscientists, but it can be challenging to understand what informa-
tion is acquired in a particular learning session. The behavior of an animal has to
be quantified using either visual or mechanical measures of a particular response.
One way of elucidating mechanisms involved in discrete learning sessions is to study
associative learning processes. Simplistically, associative learning is an adaptive
process that allows an organism to learn to anticipate events.
One form of associative learning that has been used in multiple species, includ-
ing humans, is eye-blink conditioning. The most common species used, the rab-
bit, has yielded interesting results, especially in identifying and elucidating the
involvement of the cerebral cortex. Similar procedures have been used in cats, rats,
and humans. Another form of associative learning that has gained popularity with
behavioral pharmacologists is fear conditioning. While the eye-blink procedure has
overlap with context/cue fear conditioning and in many cases yields similar results,
there are some basic differences between fear conditioning and eye-blink condition-
ing. One main difference is that eye-blink conditioning takes many more training
trials to establish. Fear conditioning has gained popularity, in large part as a result of
the need to characterize mutant mice and the effects of genetic alterations; therefore,
this chapter primarily focuses on fear conditioning.
Fear conditioning to either a cue or a context represents a form of associative
learning that has been well used in many species.1 The majority of the experiments
reported in the literature involve the mouse; however, there is also a generous propor-
tion of the literature devoted to the rat. There are also several reports in higher spe-
cies that are not covered in this chapter. In general any of the procedures described
in this chapter can be used for either the rat or the mouse.
The dependent measure used in contextual and cued (delay or trace) fear condi-
tioning is a freezing response that takes place following pairing of an unconditioned
stimulus (US), such as foot shock or air puff, with a conditioned stimulus (CS), a
particular context and/or such a cue. In the case of rats and mice, this US is generally
a foot shock. Obviously, if in a conditioning context one administers a foot shock
that is paired with a tone, there will be learning not only to the tone, but also to the
context. Two types of conditioning that are typically employed are delay or trace
conditioning. Delay conditioning refers to a situation in which the US is adminis-
tered to co-terminate with or occur immediately after the CS. Trace conditioning dif-
fers from delay conditioning in that the US follows an empty (“trace”) interval that
separates the cessation of the CS from the onset of the US. Trace conditioning adds

Cued and Contextual Fear Conditioning for Rodents 21
additional complexity to delay conditioning, as the time interval between the CS and
US requires the formation of a temporal relationship between the two stimuli.
In this chapter we discuss the various challenges inherent in this type of pro-
cedure in order to enable the experimenter to set the conditions to best answer the
questions being posed. One of the biggest advantages of cued and contextual fear
conditioning in the rodent is that they are forms of passive learning that can be used
in many strains of mice and rats, even when more pronounced motor deficits are
problematic in other learning assays. As a consequence of these procedural advan-
tages, contextual fear conditioning is gaining popularity, especially in the phenotyp-
ing of transgenic mice.
2.2 CONTEXTUAL/CUED FEAR CONDITIONING: OVERVIEW

In this section we review aspects of the different conditioning assays that are crucial
in conducting these tests.
2.2.1 CONTEXTUAL FEAR CONDITIONING

Contextual fear conditioning is the most basic of the conditioning procedures.
It involves taking an animal and placing it in a novel environment, providing an
aversive stimulus, and then removing it. When the animal is returned to the same
environment, it generally will demonstrate a freezing response if it remembers and
associates that environment with the aversive stimulus. Freezing is a species-spe-
cific response to fear, which has been defined as “absence of movement except for
respiration.” This may last for seconds to minutes depending on the strength of the
aversive stimulus, the number of presentations, and the degree of learning achieved
by the subject.
2.2.2 CUED FEAR CONDITIONING

Cued fear conditioning is similar to contextual conditioning, with one notable excep-
tion: a CS is added to the context. In order to separate context from cue conditioning
some investigators provide their subjects with a preexposure trial to the context with-
out a US. This then allows the animal to take in all the information about the context
without the presence of the cue. On a second exposure to the context, the CS is pre-
sented and the animal is better able to learn the CS association because the context is
not as accurate a predictor of shock as the CS (since the animal has previously experi-
enced the context in the absence of shock). However, preexposure to the context alone
is not sufficient to fully separate cue- and context-specific freezing behavior.
2.2.3 DELAY AND TRACE CONDITIONING

Although delay and trace conditioning differ procedurally in the presence (trace)
or absence (delay) of a time interval between the termination of the CS and US,
trace conditioning uses additional brain regions in order to establish the response.
Depending on the particular region of interest (or learning realm), researchers should
decide on the appropriate testing paradigm. The trace interval used can range from a

relatively short (2–5 sec) period, when only small learning differences in associative
learning between trace and delay conditioning is observed, to quite long (45–60 sec)
periods, when the association to the cue is very weak. Repeated training trials are
needed for trace conditioning in order for the association between the CS and US to
be formed. However, contextual learning remains strong.
2.3 BRAIN AREAS INVOLVED

The major brain areas shown to be involved in contextual and cued fear conditioning
include the amygdala, hippocampus, frontal cortex, and cingulate cortex.
2.3.1 AMYGDALA
Attempts to identify the contribution of individual amygdaloid nuclei demonstrate
that lesions to the lateral nucleus and central nucleus attenuated freezing to both con-
textual and auditory conditional stimuli, while lesions of the basal nuclei produced
deficits in contextual and auditory fear conditioning when the damage included ante-
rior lesions of the amygdala.2
Evidence suggests that the basolateral amygdala complex is a critical site for
fear conditioning. This observation stems, in part, from evidence demonstrating that
rodents with lesions to this neuroanatomical region demonstrate a lack of freezing in
the presence of cues previously paired with foot shock. An important caveat is that
some studies have suggested that an intact basolateral amygdala is not essential for
the formation and expression of long-term cognitive/explicit memory of contextual
fear conditioning,3 but may play more of an exclusive role in cue fear conditioning.
2.3.2 HIPPOCAMPUS
While learning of the context requires input from the hippocampus, especially dorsal
hippocampus and CA3, experiments have shown that this input is not necessary spe-
cifically for the learning of cue associations. It has been demonstrated, however, that
for trace conditioning, the hippocampus is required for learning the tone–shock asso-
ciation. Manipulating the interval or “gap” between the US and CS is one way studies
has isolated hippocampal involvement. As the trace interval is increased from very
short intervals of 1–2 sec to 15, 30, or 45 sec, the degree of associative cue learning is
reduced. Also, human subjects with damage in the hippocampus have been shown to
be able to acquire delay conditioning but are not able to acquire trace conditioning.4
2.3.3 FRONTAL/VENTROMEDIAL/CINGULATE CORTEX

The frontal/cingulate cortexes are areas of attentional learning and have been shown
to be involved in the acquisition of new memories.5 Consistent with this role, lesions
or pharmacological inactivation produce deficits in contextual conditioning.
Combined results indicate that there may be significant redundancy in the neu-
roanatomical regions mediating fear conditioning. The ability to dissect aspects of
memory is one of the advantages of this type of learning paradigm. To understand

some of the literature investigating the neuroanatomical underpinnings of contextual

conditioning, we suggest additional reading.5–9
2.4 BEFORE GETTING STARTED

It is important to establish the elements of the test to which the animal may respond.
If, in a given environment you deliver a shock (US) to the feet of a rat or mouse at the
termination of an audible sound cue (CS), the animal will learn that the testing envi-
ronment (context) is unpleasant. The animal will also learn that when the auditory
cue is presented there will be a shock in the near future. If this pairing is repeated,
the animal’s learning generally will be stronger, and when the animal is returned to
the conditioning context, will not only freeze to the context but also to the audible
cue. Freezing to this auditory cue will not be specific to the conditioning environ-
ment, and can also be observed in response to the cue in a totally new environment.
A challenge is dissociating how much of the learning is to the cue and how much is
to the context. One way would be a subtraction process where some animals are not
subjected to the audible cue, but as you will see later in this chapter, in some cases
the cue may interfere with contextual learning. Therefore, separation of context and
cue is ideal to establish and understand what the animal is learning.10
2.4.1 TYPES OF PARADIGMS

2.4.1.1 Contextual/Cued Fear Conditioning
It is important to understand the different methodologies and their implications
before selecting the method that is best for the given research needs (see Figure 2.1
for a schematic).
2.4.1.2 Contextual Conditioning

This occurs when an animal is placed in a new environment (chamber, cage, etc.) and
is presented with a US.
A. Context Conditioning US
Tone CS
B. Delay Conditioning US
Tone CS
C. Trace Conditioning US
Tone CS
D. Backward Trace US
Conditioning
FIGURE 2.1 Four basic conditioning paradigms illustrating the timing of US (aversive
stimulus) presentation.

2.4.1.3 Delay/Cue Fear Conditioning

This conditioning takes place when the aversive stimulus is presented at the end of a
cue (CS) (light, tone, odor), and is thus paired with the aversive stimulus (US).
2.4.1.4 Trace Fear Conditioning

Trace fear conditioning is similar to delay fear conditioning except the cue is pre-
sented for a period of time and terminated. Then following a short interval (100 msec
to 60 sec) an aversive stimulus is presented (Figure 2.1).
2.4.1.5 Backward Trace Conditioning

Backward trace conditioning is used as a control group to ascertain that the measure
of freezing in a trace-conditioning paradigm reflects learning of the association and
not some arbitrary freezing behavior. In this paradigm the CS is presented after the
US has already been presented for the trial. If the animal freezes to the tone when it
has been trained in the backward trace conditioning, it suggests that the freezing is
due to nonassociative factors, because the tone does not predict shock.
2.5 SAMPLE EXPERIMENTS

In the following section we detail two different procedures with variations that read-
ers may consider for conducting experiments in their laboratories.
General Considerations. Standard housing conditions are usually acceptable for
animals used in conditioning. The one caveat is that the mice should be calm and
healthy before testing. In the case of some strains, fighting is quite common in male
mice. This is especially prevalent in C57BL/6 and in some transgenic mice strains
such as the Tg2576. Many experimenters have opted to avoid males to overcome this
problem, although there are some differences in responses between male and female
mice. Generally, at least 8–12 animals per treatment group are needed to generate
statistical significance.
Note: Always give the mice ample habituation (60–90 min) to a novel environment
if you need to change the location of animals before testing.
2.5.1 DELAY CUED AND CONTEXTUAL FEAR CONDITIONING

2-Trial Delay Cued and Contextual Fear Conditioning. This is a standard procedure
shown to produce good cue learning and contextual learning.
Place the equipment in a quiet room. It is convenient to have an anteroom in
which to house the mice; however, this anteroom should be sound insulated from the
testing room so that the mice in their home cages are not exposed to any auditory
cues either before or after testing. If you are running more than one chamber in the
room at a time, unless the separate conditioning chambers are totally isolated, it is

important to have the stimuli synchronized so that any noise leakage and response to
shock will not interfere across animals.
2.5.1.1 Day 1
Set up the computer control programming of the equipment for conditioning so that
a house light illuminates the chamber continuously during testing.
Program a 120-sec habituation period before the first of two identical trials
begins. This allows the animal to explore briefly and to take in the aspects of the
chamber. A tone (auditory) cue is then presented, generally at a level of 70–80 dB
(we use 80 dB) for 15–30 sec. A mild foot shock is administered during the last
2 sec of the tone presentation and co-terminates with the tone. The foot shock is
generally 0.6 mA, (0.17–0.8 mA) for 1–2 sec. (The level you select will depend on
your shock source; an initial shock titration experiment may be advisable.) After the
shock presentation, an intertrial interval (60–210 sec) precedes a second identical
trial. Following the final shock presentation, the house light should remain on for
an additional 60 sec, to enable removing the mouse in a 30–60 sec time period after
the last trial.
In setting up for the experiment it is preferable to run a set of mice from a sin-
gle home cage all at once. This prevents previously tested mice from affecting the
behavior of cage mates. Therefore, as we have four training boxes in the condition-
ing room, we house our mice four per cage when possible. If desirable, mice can be
weighed and injected in the anteroom before bringing them into the room for the
conditioning session.
Before starting, wipe out the chamber with the same solution you are using to
clean the apparatus between animals to allow the first set of mice to experience the
same odors as the groups that follow. We use 70% isopropyl alcohol.
1. The first mouse is removed from the home cage and gently placed into the
conditioning chamber (repeat for the other mice in the cage). Start the train-
ing session for all the boxes in the room. The animals can be observed live
or recorded. If video recording is not available, the freezing can be scored
for any or all periods during training. Generally the mice will freeze when
the tone comes on at the start of the second trial since they have already
received one tone-shock pairing. These data can be used to assess rate of
acquisition and/or effect of drug treatment in the conditioning session.
2. At the end of the training, remove the mice. Keep in mind that the mice
have had a stressful experience and are likely to be more difficult to handle.
Use caution and handle them gently to avoid influencing the consolidation
process. Place the mice back in the home cage.
3. Between animals each cage is again cleaned/wiped out with the 70% alco-
hol solution and is readied for the next animal.
4. Try to disrupt animals as little as possible when moving them in and out of
the room.

2.5.1.2 Day 2
1. It is important to conduct the contextual testing as similarly to the training
session as possible to maximize context conditioning. This includes odor,
lighting, time of day, etc. This also maximizes differences of the novel
environment where changes are made to distinguish the environments. Sub-
jects should be well habituated to a holding area before testing if they are
moved. If they are housed in an anteroom, this saves time.
2. If the context testing is performed first, it is usual to do this at the same time
of day as the training, using the same habituation procedure.
3. Clean the chamber as before, then place the first mouse in the chamber with
the house light illuminated. For contextual conditioning testing, simply
place the mouse into the illuminated training chamber for 3–5 min; there
are no tone cues presented. The mouse can be observed for the presence or
absence of freezing response live or recorded for later analysis. The mice
should be removed promptly at the end of contextual testing and returned
to the original home cage.
4. The testing chamber should be cleaned out as on the conditioning day.
5. Allow approximately 30 min before transferring them to a new location for
cue testing.
6. If cue testing is being carried out in the same “altered conditioning cham-
ber,” it is very important to clean out the chamber thoroughly, and it is best
to use a novel odor in the chamber for subsequent cue testing
7. Another alternative is to transfer the mice to a novel test room and again
allow 60 min for habituation. The cue testing chambers should be distinct in
size, lighting intensity, background, floor texture, and odor (we use diluted
vanilla extract food flavoring wiped on the floor).
8. The mouse is placed in the chamber and allowed to habituate for 3 min. The
same intensity tone cue used in the conditioning session is then activated
for the next 3 min. One additional minute of recording without the cue is
taken before the animal is removed. Again the mouse freezing behavior can
either be captured live or recorded for later analysis. Using a Kinder Sci-
entific Motor Monitor, activity beam breaks are recorded and measures of
freezing are derived from a computer analysis. In our studies we have used
a criterion of fewer than three beam breaks in 3 or 5 sec as the criterion
for freezing. In addition, simply using beam breaks as an activity score can
also be useful.
2.5.2 TRACE CUED AND CONTEXTUAL FEAR CONDITIONING

Trace conditioning is carried out in a similar manner to delay conditioning. Differ-
ences in the procedures are related to the setup of the programming to run the con-
ditioning phase. Generally, trace conditioning requires more trials for the animals to
associate the cue with the US. Thus, for our trace-conditioning program we settled
on a five trial procedure often seen in the literature.
Computer Control. The conditioning session should start with a 60–120 sec
habituation period, followed by presentation of a tone cue for 30 sec. Then there is a

gap (trace interval) between the end of the CS and the start of the shock (US). This
trace interval can range from 2–60 sec. (Note: As the trace interval increases, the
association becomes more difficult to learn. A 15-sec trace interval with five repeat-
ing trials appears to differentiate between delay conditioning and allows for reason-
able associative learning.) Similar shock levels (0.5–0.75 mA) and duration (1–2 sec)
are used as in delay conditioning. Following each shock, a variable intertrial interval
of 90–210 sec occurs when only the house light is illuminated. A variable intertrial
interval is used between trials to minimize the possibility of the animals “expecting”
a new trial. Again, wait 30–60 sec after the final shock to remove the animals from
the chambers. Day 2 testing for freezing to the context and cue are carried out as in
delay conditioning.
2.5.3 CONTEXTUAL FEAR CONDITIONING

In cases where the experimenter is only interested in observing the fear response to
the training context, the two-trial cued and contextual fear parameters can be used.
In this case the tone (CS) is not presented in training. It would seem appropriate
that the cue testing after the exposure to the original context would be superfluous;
however, if the animals are then exposed to a novel context and freezing is measured,
this measure of nonassociative freezing can be factored in the contextual freezing
measure by simple subtraction.
Other Variations. The use of a variety of trial groups in which the CS and US are
not paired within an experiment (e.g., in backward trace or backward conditioning)
would show whether the learning taking place is a true measure of either delay or
trace conditioning. These groups are appropriate behavioral controls.
2.6 DATA ANALYSIS

The data analysis is fairly simple for the context conditioning portion. Animals
watched live or post-test can simply be timed for freezing individually and assessed
a time of freezing. For a “normal” animal, this would lie in the 60%–80% range,
however this will vary depending on the training paradigm and strain. As mentioned
previously, if several animals are viewed simultaneously, one way to score freezing
is to view each animal for time epochs and note whether freezing occurs or not.
2.7 SAMPLE DATA

Our initial experiments using delay and trace conditioning revealed some of the con-
siderations highlighted in the previous sections. We will present some of those data
and discuss some of these points. One would expect only mice that have received
shock in the initial training using a tone CS to show freezing to the original context.
Then, when in a novel environment for the cue test all the mice would show initial
increased activity, and when presented with the CS again only mice that have made
an association would demonstrate freezing. Also, the magnitude of the freezing
response would demonstrate the extent of that “learned association.”

Contextual Learning
180
*sig. difference from ns
150
Seconds Freezing
in a 5 min Test
120 *
90 *
60
30
0
ns t 30 0.17 mA t 30 0.35 mA
Group
(a)
Cue Learning
150 Group
* No shock
125 0.17 mA
Rest Time (s)
100 * 0.35 mA
75
*sig. increase over
50 no tone control
25
0
No tone Tone on
1–3 min 4–6 min
(b)
FIGURE 2.2 (A) A measure of contextual learning, showing the magnitude of context
freezing times in the original training context 24 hr post training without shock (ns), and that
freezing times increase in relation to the shock level. (B) A measure of cue learning, showing
the rest time (< 3 beam breaks in 5 sec) in a novel chamber before and after the presentation
of the original CS (tone).
To demonstrate that mice learn this association we ran three groups of C57BL/6
mice using a five trial trace-conditioning paradigm with a 30-sec trace interval. The
groups included a control group that did not receive shock during training and two
groups of mice receiving shock, one at 0.17 mA and one at 0.35 mA. All mice were
then tested 24 hr later in the conditioning chamber for contextual learning, followed
by placement in a novel chamber to assess cue learning. As can be seen in Figure 2.2,
the mice that did not receive shock during training did not freeze when assessed in
the 5 min context testing session in the conditioning chamber 24 hr following train-
ing, whereas the animals that received shock froze in a significant shock related
fashion. When later placed in the novel environment all mice showed some freezing
to the novel environment in the first 3 min before the cue was presented, with the
0.35 mA shock group showing slightly more freezing. Following presentation of the
original tone (CS) for the next 3 min there was a small increase in freezing that was

180
150
Seconds Freezing
in a 3 min Test
120
90
60
30
0
Harlan Jax Harlan Jax
Delay Trace
Figure 2.3 Depicts the difference in freezing times (mean, +/- SEM) between C57BL/6
mice obtained from Jackson Labs (Jax) and Harlan Sprague Dawley (Harlan) when trained
and tested for contextual conditioning 24 hr later.
not significant in the no-shock group (nonassociative freezing, which is discussed

later), but there was a significant increase in the groups receiving shock, which indi-
cates learning of the tone–shock association.
Also studied was a direct comparison between C57BL/6 mice from two different
suppliers, Jackson Labs and Harlan Sprague Dawley, Inc., keeping all environmental
conditions equal. The mice were trained with five trials consisting of a 30 sec CS
tone of 80 dB, and a shock (0.35 mA) administered for 2 sec, in either a delay con-
ditioning procedure (US in the last 2 sec of the CS) or in a 30-sec trace conditioning
procedure (US 30 sec following the termination of the CS).
As can be seen in Figure 2.3, the Harlan mice exhibited more freezing than the
Jax (Jackson Labs) mice in the original context; however, the Harlan and Jax mice
were similar when trained with five trials of trace conditioning.
These mice were also tested for cue conditioning in the Kinder Scientific Motor
Monitors. In order to clearly show some supplier differences, Figure 2.4 shows the
activity scores as beam breaks for the cue testing in a novel environment. It can be
seen in the delay conditioning paradigm that mice from both Harlan and Jax demon-
strate less activity when the tone is presented. However, the Jax mice recover their
activity levels once the tone is turned off during the last minute of the session (see
arrow), showing good associative learning; the Harlan mice do not show this. In
addition, trace conditioned mice do not demonstrate this recovery of activity when
the cue is turned off (minute 7). This could be interpreted as the demonstration that
the mice trained in the trace conditioning paradigm learned that “tone off” is as
good of a predictor of shock as is “tone on.” Also, Jax mice show more activity than
the Harlan mice in the first 3 min of the session.
2.8 Nonassociative Freezing Complications

It has been shown that exposing rodents to foot shock will sometimes produce gen-
eralized, or nonassociative freezing to unconditioned stimuli, such as when placed

Time in
Delay Trace Minutes
175
1
150 2 no tone
125 3
Activity Score
100 4
5 tone
75
6
50 7 no tone
25
0
Harlan Jax Harlan Jax
FIGURE 2.4 Illustration of differences in freezing times (mean, +/- SEM) between C57BL/6
mice obtained from Jackson Labs (Jax) and Harlan Sprague Dawley (Harlan) when trained
and tested for response to the cue when presented in a novel environment 24 hr later. The
arrow points to the rapid recovery of activity in the 7th min by Jax mice only when trained
in the delay paradigm.
in a novel environment.11 In our experience, we can see increased freezing to the

tone stimulus in the novel environment in animals that have received contextual
conditioning, i.e., those that have not been exposed to the tone during conditioning.
To demonstrate these nonassociative freezing effects of the tone in our trace con-
ditioning paradigm, we ran an experiment using groups of mice preexposed to the
tone CS, thereby establishing it as a neutral stimulus. The experimental setup is seen
below: the preexposure, conditioning, and context tests take place in the condition-
ing chamber, and the cue test takes place in a novel chamber.
Group Preexposure Conditioning Context Test Followed by Cue Test

Day 1 Day 2 Day 3
Tone—Paired Context Only Tone and Shock 5 min Context Cue Test with Tone
Tone—Unpaired Context and Tone Shock Only 5 min Context Cue Test with Tone
On day 1, an acclimation day, all mice were placed into the conditioning chamber.
The “tone—unpaired” group was exposed to five 30-sec CS (tone 80 dB), and
another “tone—paired” group of mice was exposed to the conditioning chamber
without the tone. On day 2, both groups were again placed in the conditioning
chamber. The unpaired tone group received five contextual conditioning (no tone)
trials. The paired tone group received five trials of trace conditioning with a trace
interval of 15 sec. The US was a 2-sec shock of 0.78 mA for both groups. This
design resulted in both groups receiving an equal number of exposures to the tone
(CS), the shock (US), and to the amount of time in the conditioning chamber. The

Context Test Cue Test

100 Tone Paired
100 Tone Unpaired
Percent Freezing
75
Percent Freezing
5 min Test
75
50
50
25 25
0 0
Tone Paired Tone Unpaired Pre-Cue Cue Post-Cue
FIGURE 2.5 Illustration showing that nonassociative freezing to the tone cue can be
reduced by preexposure to the tone in the unpaired group. Both groups had equal exposure
to the tone cue, but mice in the unpaired group received tone exposure on the day prior to the
foot shock conditioning. In the context test, mice trained with paired tone or unpaired tone
showed similar freezing times. However, in the cue test, the paired group showed increased
freezing to the tone.
only difference in the treatment of the two groups was the presence (paired) or
absence (unpaired) of the tone on day 2 when the shock and tone association was
presented. On day 3, both groups were tested first for contextual conditioning in the
conditioning context for 5 min, then in a novel chamber with 3 min of no tone, 3
min of tone, and 1 min recovery (no tone).
As can be seen in the Figure 2.5, both groups exhibited the same level of freez-
ing in the conditioning context; however, the mice with a paired association exhib-
ited a greater response when the CS was presented in the novel context.
Statistical Analysis. Generally all that is needed is either a one- or two-way
analysis of variance (ANOVA) with appropriate post hoc analysis comparing the
various treatment groups using either the raw or percent freezing scores of the con-
textual freezing. In addition, this analysis may be performed on the data from the
cued conditioning scores.
2.9 FINAL NOTE

We hope from reading this chapter the reader has gained a basic understanding of the
cued and contextual fear conditioning paradigms used in rodents, and comprehen-
sion of the difference between trace and delay fear conditioning. This chapter is just a
stepping-off point—there is a vast amount of literature that has been published on the
involvement of the various brain regions responsible for associative learning, as well
as the differences between trace and delay conditioning models. Using this informa-
tion, one should easily be able to set up equipment to carry out studies that will lead
to productive research. There is an extensive addendum to help you start testing.
2.10 ADDENDUM
2.10.1 AVAILABLE EQUIPMENT OPTIONS
There are several options when it comes to choosing the equipment necessary to
demonstrate a reliable cue and contextual fear response. The equipment does not

have to be terribly sophisticated and a number of labs that made their own chambers
are still using the custom equipment.
It is most important to be able to monitor the animals either visually (live or
through the use of a video system) or by incorporating a movement monitoring sys-
tem to reliably measure freezing behavior. Miniature video cameras are available,
(for example, CCTVOne) and easily located inside an isolation cubicle to monitor the
animal or record the session on tape or DVD. Recording of the behavior enables the
researcher to review and reanalyze an experiment, which can be advantageous. Some
laboratories perform live visual monitoring of animals during an experiment and
when testing multiple animals simultaneously, and will time sample each chamber.
2.10.2 CONDITIONING CHAMBERS

Chamber Size. The size of the chamber is not critical, but the chamber should be
constructed to be easily cleaned and have a way to easily view the subject. In addi-
tion, if the same chamber is being used for both contextual and cued conditioning, it
should be easily adaptable to making the context distinctly different when changing
from contextual to cued conditioning. Making the chamber different should involve
changes in the floor (e.g., from stainless steel bars to a solid plastic or equivalent),
and changes in the inside dimensions by the addition of a diagonal divider. In addi-
tion, it is helpful to change the odor by using a diluted food essence.
Equipment Examples. In this section we list some suppliers of fear conditioning
equipment (Table 2.1).
Kinder Scientific. Our lab uses equipment available from Kinder Scientific. The
conditioning chamber we use consists of one side of the active avoidance system and
is therefore used for contextual freezing testing (Figure 2.6). We run four chambers
at a time that are under the control of one computer to maximize throughput. Each
chamber is equipped with a video camera and the video output is fed into a splitter to
record multiple animals on one DVD recorder.
Coulbourn Habitest. Testing can also be achieved by using a Coulbourn Habitest
chamber for conditioning and contextual fear testing. Cue testing can then be car-
ried out in another chamber equipped with the same audible cue or by altering the
conditioning chamber.
Med Associates. Med Associates will sell a complete package for contextual fear
conditioning and can include an infrared video analysis system that detects freez-
ing (Figure 2.7). This analysis has been shown to be equivalent to visual scoring in
a paper by Contarino and colleagues.12 There are confounds to learning when both
cued and contextual conditioning are measured in the same chamber where the con-
ditioning took place.
San Diego Instruments. San Diego Instruments also provides a stand-alone
Freeze Monitor that uses photocells in a higher density than a normal activity cham-
ber arrangement. They also provide a separate Freeze Monitor context enclosure to
detect nonassociative freezing.
Clever Systems, Inc. Freeze Scan¥ is a software system for automatically detect-
ing freezing states in rodents and fulfills the demand for high throughput screening.
Freeze Scan¥ is a video-based tool that provides precise motion control for accu-

TABLE 2.1
Names and Addresses of Vendors
Company Name Company Contact Information
Clever Systems, Inc. 11425 Isaac Newton Square, Suite # 202
Reston, VA 20190.
Tel: (703) 787-6946
Fax: (703) 787-6684
www.cleversysinc.com
CCTVOne 509 Mercury Lane
Brea, CA 92821
Tel: (866) 582-2881
Fax: (714) 529-8599
www.cctvone.com
Coulbourn Instruments 7462 Penn Drive
Allentown, PA 18106
Tel: (610) 395-3771
www.coulbourn.com
Kinder Scientific 2655 Danielson Court, Suite 308
Poway, CA 92064
Tel: (858) 679-15
Fax: (858) 679-4811
Med Associates, Inc. PO Box 319
St. Albans, VT 05478
Tel: (802) 527-2343
Fax: (802) 527-5095
www.med-associates.com
San Diego Instruments 7758 Arjons Drive
San Diego, CA 92126-4391
Tel: (858) 530-2600
Fax: (858) 530-2646
www.sandiegoinstruments.com
rate freezing detection. Freeze Scan¥ accepts video taken from different views in
a confined chamber. It precisely detects the onset and completion of the freezing
behavior of a rodent. Its output is a sequential list of the occurrences of the freezing
behaviors. Further statistics can be analyzed from this output data. Freeze Scan¥ has
the capability to set the same intervals generated by the tone, shock, or light control
program. Clever Systems, Inc. can also provide the hardware for tone, shock, and
light control for use in your chamber, and thus Freeze Scan¥ can synchronously
work with the control program and provide accurate freezing state results based on
these intervals.
2.10.3 CONSIDERATIONS FOR THE US (SHOCKERS)

The level of shock produced as the US will vary to a great degree on the shock
source (manufacturer). In order to present a standard shock level to each mouse or rat
it is best to use a DC constant current source. Most commercially available shockers,

FIGURE 2.6 Kinder Scientific Learning and Memory Avoidance Systems.
FIGURE 2.7 MedAssociates Contextual Fear System including freezing detector.
such as those from Coulbourn Instruments, are square wave sources. This means
that in contrast to AC or sine wave sources, there is an instantaneous rise time and
off time that is more aversive at a lower current level. A scrambled shock is switched
between the bars of the floor; the current is applied separately to each bar in a span
of 8 bars that are sequenced over a 32-bar floor. The Kinder Scientific Active Avoid-
ance box comes with an internal square wave constant current source. In the case of
other chambers, external shock sources need to be supplied. Discrepancies in shock
levels used in different laboratories usually can be attributed to the type of shocker
used. The best measure is to actually observe the animal and gradually increase the

shock to a level where the animal first vocalizes and rapidly runs around the cage.
This level is lower than that where the animals start jumping to avoid the shock,
although some strains may show a hyperresponsivity to an initial shock.
Note: It is advisable to use an oscilloscope to measure and view the shock output
at the level of the grid in the test apparatus. This helps to reduce the possibility of not
obtaining a reproducible effect.
2.10.4 CUED CONDITIONING MEASUREMENT

The Kinder Scientific activity monitoring system has a board containing LED photo-
cells that surrounds the chamber and measures beam breaks surrounding a Plexiglas
chamber (7.5 in × 14 in) with a flat, white plastic tray floor that can be used for mea-
suring conditioned association. The ceiling contains a lamp and a Mallory Sonalert®
sound source identical to the ones used in training. We enclosed the chambers in a
box made of pink insulation and the chambers were adapted to allow video recording
of the animal behavior. You may also consider using the Kinder Scientific cued fear
conditioning chamber or any of the other conditioning setups. However, it must be
emphasized that the animal not only learns about the immediate context into which
it was placed when it received the aversive stimulus, but it also learns all the events
and places leading up to being placed in that context. To reduce any generalization
from the conditioning context to the novel context, it is preferable to carry out the
cue testing in a separate room with new visual, tactile, and olfactory cues.13 In our
laboratory, the assessment of cued conditioning is conducted in a totally different
sized room in the facility.
2.10.5 MEASURING FREEZING BEHAVIOR

Two options are available for measuring freezing behavior.
2.10.5.1 Hand Tallying Method

This method can be carried out in a few different ways. If the animals’ behavior
is recorded, individual animals can be observed and continuously monitored for
freezing. If it is desirable to monitor more than one animal simultaneously, then a
sampling system can be adopted. For example, one method is to observe each cage
sequentially for 5 sec and simply note whether freezing is occurring (yes or no) in
each chamber for each 5-sec bin. These data can be recorded on a chart. If six cages
were scored simultaneously for a 5-min session you would have 10 bins per animal.
If an animal demonstrated freezing in five of the bins, this would translate to a 50%
freezing score. Another similar approach is to monitor each animal once every 5–10
sec and note whether the animal is freezing at the exact moment of observation. Yes
or no scores can be tabulated in an identical way as in the above approach. These
sampling methods may not be as accurate as constant measuring, but if carried out
uniformly they will yield reproducible results. In fact, it has been shown that scores
obtained from these sampling methods correlate very highly with scores obtained
from continuous sampling.12

2.10.5.2 Activity Monitoring Method

This method takes into account distance traveled and/or the number of beams bro-
ken. In the Kinder analysis, the movement of the central part (centroid) of the animal
is taken into account to calculate a “rest time.” An animal is considered at rest when
there are fewer than a specific number of beam breaks in a specified period (the
numbers can be specified by the experimenter). These time epochs are then counted
up to give a rest time total in seconds. We consider an animal to be at rest or “freez-
ing” if there are fewer than three beam breaks in 5 sec. The total freezing score is
validated by independent observation and timing of the freezing response compared
to the computer analysis of the rest time.
San Diego Instruments has a similar system (Freeze Monitor) that uses a con-
centrated number of infrared photo beams measuring beam breaks. The analysis is
similar to that described above.
Infrared Video: A Fire-Wire® system is available from Med Associates, Inc. that
can be used to detect freezing in the animal. This is a computer-based system that
has been validated. See Anagnostaras and colleagues.14
Clever Systems also has a sophisticated online video analysis. This system uses
a similar method of pixel analysis of the streamed video frames.
Note: Whatever system you choose, it is always necessary to run a set of animals
to demonstrate that any instrumentation analysis is producing data that correlates
highly with those obtained using a visual scoring method.
2.10.6 ANIMAL STRAIN CONSIDERATIONS

Many different mouse strains have been used for cued/contextual fear conditioning.
Behavioral phenotype can affect the magnitude of the freezing behavior or the level
of associative learning. In some strains of mice retinal degeneration is a background
genetic defect that may or may not affect contextual freezing. In other cases, hearing
deficits develop as the animal ages. Some mouse strains may demonstrate reduced
responsiveness to foot shock. Therefore, when designing the experiment, the strain,
cue type or magnitude, and aversive conditioning levels should be considered. In
many studies that have reported optimal learning of the context and cue, researchers
have used the C57BL/6 mouse with a white noise, tone, or clicker cue, and 0.4–0.6
mA shock level. Comparison of not only strain but also animal supplier may also
yield different results and can be explored.15
In the case of rats (especially aged rats), quantification of the freezing response
as well as variability in audition and response to shock could also affect the behav-
ioral measures. Sprague Dawley rats tend to freeze more than other strains, and some
of these problems have been overcome by using transmitters to measure changes in
heart rate as the measure of conditioned fear.
2.10.7 AUDITORY CUE CONSIDERATIONS

The superficial selection of the CS for the apparatus appears to be wide ranging.
Studies have been conducted with clickers, white noise, and pure tones generally

between 70 and 85 dB. RadioShack makes a convenient meter that will give a good
reading of dB levels. It should be set on continuous with “A” weighting to allow for
mean dB levels. The noise-producing device is most effective if it can fill the cham-
ber without noise “dead spots” to ensure that subjects will receive the CS regardless
of their position in the chamber.
REFERENCES
1. Kim, J. J.. and Jung, M. W. 2006. Neural circuits and mechanisms involved in Pavlovian
fear conditioning: A critical review. Neurosci. Biobehav. Rev. 30(2):188.
2. Goosens, K. A., and Maren, S. 2001. Contextual and auditory fear conditioning are
mediated by the lateral, basal, and central amygdaloid nuclei in rats. Learn. Mem.
8(3):148–55.
3. Vazdarjanova, A., and McGaugh, J. L. 1998. Basolateral amygdala is not critical
for cognitive memory of contextual fear conditioning. Proc. Natl. Acad. Sci. USA
95(25):15,003–7.
4. Clark, R. E., and Squire, L.R. 1998. Classical conditioning and brain systems: The role
of awareness. Science 280(5360):77–81.
5. Pezze, M. A., and Feldon, J. 2004. Mesolimbic dopaminergic pathways in fear condi-
tioning. Progress in Neurobiology 74(5):301–320.
6. Anagnostaras, S. G., Gale, G. D., and Fanselow, M. S. 2001. Hippocampus and contex-
tual fear conditioning: Recent controversies and advances. Hippocampus 11(1):8–17.
7. Atallah, H. E., Frank, M. J., and O’Reilly, R. C. 2004. Hippocampus, cortex, and basal
ganglia: Insights from computational models of complementary learning systems. Neu-
robiology of Learning and Memory 82(3):253–267.
8. Gewirtz, J. C., McNish, K. A., and Davis, M. 2000. Is the hippocampus necessary for
contextual fear conditioning? Behavioural Brain Research 110(1–2):83–95.
9. Maren, S., and Holt, W. 2000. The hippocampus and contextual memory retrieval in
Pavlovian conditioning. Behavioural Brain Research 110(1–2):97–108.
10. Rudy, J. W., Huff, N. C., and Matus-Amat, P. 2004. Understanding contextual fear condi-
tioning: Insights from a two-process model. Neurosci. Biobehav. Rev. 28(7):675–685.
11. Balogh, S. A., and Wehner, J. M. 2003. Inbred mouse strain differences in the establish-
ment of long-term fear memory. Behavioural Brain Research 140(1–2):97–106.
12. Contarino, A., Baca, L. Kennelly, A., and Gold, L. H. 2002. Automated assessment of
conditioning parameters for context and cued fear in mice. Learn. Mem. 9(2):89–96.
13. White, N. M., and McDonald, R. J. 2002. Multiple parallel memory systems in the
brain of the rat. Neurobiol. Learn. Mem. 77(2):125–84.
14. Anagnostaras, S. G., Josselyn, S. A., Frankland, P. W., and Silva, A. J. 2000. Computer-
assisted behavioral assessment of Pavlovian fear conditioning in mice. Learn. Mem.
7(1):58–72.
15. Schimanski, L. A., and Nguyen, P. V. 2004. Multidisciplinary approaches for investi-
gating the mechanisms of hippocampus-dependent memory: A focus on inbred mouse
strains. Neurosci. Biobehav. Rev. 28(5):463–83.

3 Drug Discrimination
Richard Young
CONTENTS
3.1 Introduction................................................................................................... 39
3.2 Methods......................................................................................................... 41
3.2.1 Apparatus........................................................................................... 41
3.2.2 Subjects .............................................................................................. 41
3.2.3 Operant Training................................................................................ 41
3.3 Drugs as Stimuli............................................................................................ 43
3.3.1 Discrimination Training Procedure................................................... 45
3.3.2 Discrimination Data........................................................................... 45
3.3.2.1 Percent Drug Lever Responding .......................................... 45
3.3.2.2 Response Rate ...................................................................... 47
3.4 Applications .................................................................................................. 47
3.4.1 Stimulus Generalization .................................................................... 47
3.4.2 Test Considerations ............................................................................ 48
3.4.2.1 Dose Response .....................................................................48
3.4.2.2 Comparison of Results of Test Agents ................................. 48
Data Analysis, Interpretation, Examples...................................................... 49
3.4.3.1 Statistical Analysis ............................................................... 50
3.4.3.2 Examples of Complete, Partial, and No Substitution........... 50
3.4.3.3 Time Course ......................................................................... 52
3.4.3.4 Stimulus Antagonism ........................................................... 53
3.5 Summary.......................................................................................................54
References................................................................................................................ 56
3.1 INTRODUCTION
The psychoactive effect of a drug usually refers to a chemical agent that exerts an
action upon the central nervous system (CNS), alters brain function, and, conse-
quently, produces a temporary change in an individual’s mood, feelings, perception,
and/or behavior. Such agents may be prescribed as therapeutic medications or used
(or abused) as recreational drugs. In each case, the subjective effects produced by
such agents are generally not accessible to independent verification by an observer.
However, methods were developed about 50 years ago whereby human subjects could
self-rate their experiences on questionnaires after administration of a drug.1 Gener-
ally, these self-inventories require subjects to provide information about themselves
and are considered valuable because they venture “below the surface” to glean the
effect of a drug on an individual. Also, they are convenient because they (usually) do
39

not require the services of a group of raters or interviewers. Their chief disadvantage
may be that individuals might not completely understand the effect of the drug or
their drug “experience” and therefore might not always give an accurate report.
The drug discrimination (DD) paradigm is an assay of, and relates to, the subjec-
tive effect of drugs in nonhuman animals or humans. In a typical DD experiment,
there are four basic components: (1) the subject, (2) the dose of drug that exerts an
effect on the subject and precedes a response by the subject, (3) an appropriate (or
correct) response, and (4) presentation of reinforcement.
SUBJECT q DOSE OF DRUG q RESPONSE q REINFORCEMENT
The drug effect that “leads to” a behavioral event (i.e., particular response) and
signals that reinforcement is available is called the discriminative stimulus. A wide
variety of psychoactive drugs can serve as discriminative stimuli (see below). In
laboratory subjects, discriminative control by (usually) two treatments is established
through the use of reinforcement (reward). When subjects receive a dose of a drug,
it functions as a signal that prompts a correct behavioral response and results in the
presentation of a reward. In other words, the effect of the drug is used as a “help” or
“aid” to control appropriate behavioral responding by signaling that reinforcement
is (or will be) available. Subjects are usually trained to distinguish administration of
a particular dose of a particular drug (i.e., the training dose of a training drug) from
administration of saline vehicle (i.e., usually a 0.9% sodium chloride solution that
is often used as a solvent for many parenterally administered drugs). In a subject’s
course of training sessions, the dose of drug is administered (i.e., drug sessions)
and lever presses on the drug-designated lever (for that subject) produce reinforce-
ment. In other training sessions, saline is administered (i.e., vehicle sessions) and
responses on the (alternate) saline-designated lever produce reinforcement. The DD
procedure can be characterized as a highly sensitive and very specific drug detection
method that provides both quantitative and qualitative data on the effect of a training
drug in relation to the effect of a test (i.e., challenge) agent. Historically, DD studies
are linked by a common requirement that subjects must perform an appropriate (or
correct) response that indicates a distinction was made between drug and nondrug
conditions. As such, when employed with animals or humans, a subject’s response
permits an experimenter to determine if a drug effect has been “perceived.” An
excellent source of information on DD studies can be found at the Drug Discrimina-
tion Bibliography Web site (http://www.dd-database.org). The Web site, established
and maintained by Drs. Ian P. Stolerman and Jonathan B. Kamien, is funded by the
National Institute on Drug Abuse (NIDA) of the National Institutes of Health (NIH)
and contains close to 4000 DD references published between 1951 and the pres-
ent. The citations include DD abstracts, journal articles, reviews, book chapters, and
books. In addition, the Web site can be navigated to selectively retrieve references on
particular training drugs, drug classes, test drugs, authors, and DD methodologies.

Drug Discrimination 41
3.2 METHODS
3.2.1 APPARATUS
Behavioral experiments with animals are often conducted in testing environments
that eliminate or minimize the occurrence of extraneous events or conditions (e.g.,
loud sounds, lights, temperature changes, etc.). The experimental setting is also
designed to make more likely the occurrence of a particular behavior. For example,
placing a hungry rat in a small chamber in which a lever is the most prominent object
increases the likelihood that the animal will press the lever, which will result in the
delivery of a reward. Studies of DD are often conducted in standard two-lever oper-
ant chambers (Coulbourn Instruments, Model E10-10, Lehigh Valley, Pennsylva-
nia, USA, or Med Associates, Model ENV-008, St. Albans, Vermont, USA) housed
within light- and sound-attenuating outer chambers. Typically, one wall of each
operant chamber is fitted with two levers and a device, centered equidistant between
the levers, to deliver reinforcement. The reinforcement may be, for example, a 45-
mg food pellet (e.g., Noyes Precision Pellets® PJAI-0045, Research Diets, Inc., New
Brunswick, New Jersey, USA), sweetened condensed milk, or water (delivered in a
0.01 mL cup). An overhead 28-V house light illuminates each chamber. Solid-state
and computer equipment are used to record lever presses, program the delivery of
reinforcement, and record the number of reinforcements.
3.2.2 SUBJECTS
Table 3.1 shows that different species have been used as subjects in DD studies.
To date, the rat has been used most often as the experimental subject in DD cita-
tions. Also of interest is the number of studies that cited humans as the experimen-
tal subjects. It is noted that DD procedures for humans are similar to those used
for laboratory animals, but are adjusted to the uniqueness of humans. For example,
drugs are usually administered under double-blind conditions and money typically
serves as reinforcement for correct responses. In addition, many human DD stud-
ies include questionnaire data on subjective effects of the administered agent(s).12,13
In a DD study with animals or humans as subjects, however, the learning of a DD
involves appropriate responses for the presentation of reinforcement under the phar-
macological effect of different treatments.
3.2.3 OPERANT TRAINING

The discriminative stimulus effect of a drug is most frequently established via oper-
ant conditioning, a learning paradigm in which a subject emits a response that is
followed closely by reinforcement. In general, operant behavior is “controlled” by its
consequences. For example, a hungry rat may “act on” or “operate on” its environ-
ment and press a lever, which closes a switch and activates a food dispenser or liquid
dipper to produce a pellet in a tray or liquid (e.g., sweetened milk) in a small cup. The
operant is the behavior just prior to the reinforcement. In such a situation, the rat’s
press of the lever, which is followed by reinforcement, may result in an increased
probability of the animal pressing the lever in the future. In practice, however,

TABLE 3.1
Species Used as Subjects in Drug Discrimination Experiments
Species Citations Reference (Example)
Cat, Dog, Guinea Pig 1 each Kilbey and Ellinwood2
Cook et al.3
Hudzik et al.4
Gerbil 24 Jarbe et al.5
Human 262 Altman et al.6
Mice 97 Snoddy and Tessel7
Monkey 513 Schuster and Brady8
Pig 3 Carey and Fry10
Pigeon 360 Henriksson et al.10
Rat 2641 Barry11
Source: Data obtained from citations in Drug Discrimination Bibliography (http://www.
dd-database.org).
operant conditioning is the study of behavior maintained by schedules of reinforce-

ment, which are defined as the delivery of reinforcement to a subject according to
some well-defined rule. In DD applications, the animals’ opportunity to press a lever
under a schedule of reinforcement gives them, in effect, “communication” to the
investigator of “how a drug affects their CNS.” It is also noted that schedules of rein-
forcement are used with humans and the pattern of responding is generally similar
to those obtained with nonhuman animals.
An animal’s initial training in a DD two-lever operant task begins with “maga-
zine training,” which involves training the subject to eat from a food tray or drink
from a dipper cup and, consequently, for it to learn that the noise made by the acti-
vation of the (mechanical) delivery device indicates the imminent presentation of
“compensation.” At the beginning of the study, the experimenter teaches the rats to
press a lever for reinforcement with the technique of successive approximation or
“shaping.” The latter procedure involves the reinforcement of initial behavior that
may only be vaguely similar to the final desired response (i.e., lever pressing); rein-
forcement continues for variations in behavior that come closer to pressing the lever.
For example, an experimenter may begin with the presentation of reward (e.g., via
a hand switch that bypasses a schedule of reinforcement) to a rat every time it turns
toward a lever. Later, only movements that bring it closer to a lever are reinforced.
After the rat has learned to approach a lever, it is not reinforced until it touches the
lever. Eventually, the rat’s behavior will have been so shaped that it will readily
press a lever when put in the chamber. (It is noted, however, that when a relatively
large number of animals are used in an experiment, shaping can be a very time-
consuming procedure that requires much patience on the part of the experimenter.)
When every press of the lever is followed by reinforcement, the organism is said to

be on a continuous reinforcement schedule. As a consequence of such a schedule,

the number of reinforcements could become quite high and might lead to decreased
responding over time because of satiation. However, it is not necessary to reinforce
every response in order to maintain responding. That is, an animal can be reinforced
intermittently (i.e., part of the time). The intermittent schedule can be based on a
portion of responses or on a time interval. The two most common schedules of rein-
forcement are ratio and interval schedules, each of which can be fixed (unvarying) or
variable (random). In DD studies, the fixed ratio (FR) and variable (random) interval
(VI or RI) schedules of reinforcement are used extensively and are discussed here.
In an FR schedule the subject must complete a fixed number of responses in order
to obtain reinforcement. In an FR 10 schedule, for example, every 10th response
is reinforced. In VI schedules, the length of time elapsing before reinforcement is
delivered varies around the mean value specified by the schedule. On a VI 15 sec
schedule, reinforcement is available, on average, after 15 sec have elapsed since the
last reinforcement but may be available, for example, as shortly as 2 sec later, or not
until 60 sec have elapsed. The first response after a time interval has elapsed pro-
duces reinforcement for the organism. Once the rat learns to press on the right-side
lever and the left-side lever under the schedule of reinforcement (e.g., FR or VI), DD
training begins.
3.3 DRUGS AS STIMULI

Table 3.2 lists some of the drugs, from several different pharmacological and chemi-
cal classes, that have been shown to serve as stimuli. In most of these studies, either
the FR or VI schedule of reinforcement was used to establish the discrimination.
Although DD techniques have varied, it is typical that an organism is taught to
respond on one lever (e.g., right-side lever) when a dose of training drug is adminis-
tered before a training session and on another lever (e.g., left-side lever) when vehicle
(e.g., saline) is given. Correct responses are intermittently reinforced by delivery of
food (e.g., pellet, sweetened milk). The organism’s eventual learning of the correct
response in a typical two-lever choice task involves its determination that the effect
produced by the administration of the dose of the training drug on certain days is
distinct from that produced by the injection of vehicle on other days.
In general, there has been good agreement between species on (1) whether or not
a particular drug can function as a discriminative stimulus, and (2) results obtained
with test (or challenge) agents (see stimulus generalization tests below). The present
survey describes some training and test results from rats trained to discriminate
diazepam (3.0 mg/kg) from vehicle (one drop of Tween 80 per 10 mL distilled water)
under an FR 10 schedule of reinforcement. For these studies, 12 experimentally
naïve, male albino Sprague Dawley rats (Charles River Labs, Wilmington, Massa-
chusetts, USA) weighing 325–350 g at the beginning of the experiment were used.
Rats were housed individually and had free access to water, but were gradually food
restricted to approximately 80% of their free-feeding weights before training began.
The colony room was kept at a constant temperature (approximately 21–23°C) and
humidity (~ 50%); lights were turned on from 0600 to 1800 hr.

TABLE 3.2
A Partial List of Drugs that Have Been Used as the Discriminative Stimulus
in Drug Discrimination Experiments
Drug Drug Class or Mechanism of Action Reference (Example)
Amphetamine Stimulant Schechter and Rosecrans14
Apomorphine Dopamine receptor agonist Colpaert et al.15
Atropine Muscarinic antagonist Barry and Kubena16
Buprenorphine Partial agonist (μ-opioid receptor) Holtzman17
Buspirone Antianxiety Hendry et al.18
Caffeine Stimulant Carney and Christensen19
Cholecystokinin (Neuro) peptide hormone De Witte et al.20
Chlorpromazine Antipsychotic Goas and Boston21
Clozapine Antipsychotic Browne and Koe22
Cocaine Stimulant Jarbe23
Desipramine Antidepressant Shearman et al.24
Dextromethorphan Antitussive Holtzman25
Diazepam Antianxiety Young et al.26
Diphenhydramine Antihistamine Winter27
DOMa Hallucinogen Young et al.28
Ephedrine Agonist (adrenergic receptors) Young and Glennon29
Ethanol Stimulant/sedative Schechter3
Fenfluramine Appetite suppressant Goudie31
Fentanyl Opioid analgesic Colpaert et al.32
LSD b Hallucinogen Hirschhorn and Winter33
MDAc Designer drug Glennon and Young34
MDMAd Designer drug Glennon and Misenheimer3
Morphine Opioid analegesic Hirschhorn and Rosecrans36
Naloxone Antagonist (μ-opioid receptor) Carter and Leander37
Nicotine Nicotinic acetylcholine receptor agent Schechter and Rosecrans38
NMDA e Agonist (NMDA receptor) Willetts and Balster39
Pentazocine Opioid analgesic Kuhn et al.40
Pentobarbital Sedative Herling et al.4
PCP f Dissociative anesthetic Brady and Balster42
Pregnenolone (Neuro) steroid Vanover3
9
Δ -THC g Cannabinoid1 receptor agent Jarbe et al.44
Toluene Solvent (Abused by Inhalation) Rees et al.45
Source: Data obtained from citations in Drug Discrimination Bibliography (http://www.dd-database.
org).
a1-(2,5-dimethoxy-4-methylphenyl)-2-aminopropane
bLysergic acid diethylamide
c1-(3,4-methylenedioxyphenyl)-2-aminopropane (3,4-MDA)
dN-methyl-1-(3,4-methylenedioxyphenyl)-2-aminopropane
e(N-methyl-D-aspartic acid)
f1-(1-phenylcyclohexyl)piperidine (phencyclidine)
gΔ9-tetrahydrocannabinol

3.3.1 DISCRIMINATION TRAINING PROCEDURE

The dose-of-a-drug versus saline treatment paradigm constitutes the single most
widely used DD procedure and its properties have been documented best. In the
present example, the rats’ training sessions are preceded by an intraperitoneal (IP)
injection of either 3 mg/kg of diazepam (dose is based on weight of base) or vehicle
(one drop of Tween 80 per 10 mL of distilled water) with only the stimulus-appro-
priate lever present (i.e., left- or right-side lever). A pre-session injection interval
(PSII) of 15 min is used; during this interval the animals are in their home cages.
The route of administration and the PSII for the drug and its vehicle are typically
chosen on the basis of the known pharmacokinetic properties and/or behavioral
effect(s) of the drug. Training sessions are of 10 min duration, 5–7 days per week.
For a particular session, just one of the two levers (i.e., the treatment-appropriate
lever) is programmed to deliver reinforcement; presses on the incorrect lever have no
programmed consequence. For six of the rats, responses on the right-side lever are
reinforced after drug administration, while responses on the left-side lever are rein-
forced after vehicle administration; lever response conditions are reversed for the
remaining six rats. In addition, lever assignments for a particular operant chamber
are alternated (e.g., first animal in chamber 1 is assigned left-side lever as drug lever
and right-side lever as saline lever; second animal in chamber 1 is assigned right-
side lever as drug lever and left-side lever as saline lever, etc). The latter tactic is
important because it has been observed that rodents may learn to use olfactory hints
(or cues) that remain in the operant chamber from preceding animals.46 In addition,
diazepam or vehicle is administered on a random schedule with the constraint that
no more than two consecutive sessions with the drug or vehicle can occur; an equal
number of drug and vehicle sessions occur. The experimenter will note that initial
injections of the dose of training drug might hinder or disrupt the animals’ pressing
of the drug-designated lever. Animals should develop behavioral tolerance to the
disruptive effects of the drug (i.e., diazepam) and will, over time, perform the task.
Animals do not, however, develop tolerance to the stimulus effect of the training
dose of the training drug. If tolerance did develop, then the dose of the drug would
not continue to serve as a discriminative stimulus and the animals’ performance
would significantly decline.
3.3.2 DISCRIMINATION DATA

3.3.2.1 Percent Drug Lever Responding
An animal’s degree of progress in learning the DD is determined by an evaluation
of its distribution of presses on the two levers. In particular, an animal’s learning of
the discrimination can be evaluated either prior to, or up to, the delivery of the first
reinforcement. Thus, when an FR 10 schedule of reinforcement is used, DD learning
can be assessed for each subject by dividing the number of responses that occurred
on the drug-designated lever by the total number of responses that occurred on both
levers up to the delivery of the first reinforcement; percent of responses on the drug-
appropriate lever is then obtained by multiplying the value by 100. For instance,
assume that a rat has the right-side lever designated as the diazepam-appropriate
lever. On a Monday, the animal is injected with 3 mg/kg of diazepam, placed in its

assigned operant chamber, and proceeds to press the left-side lever 9 times and the
right-side lever 10 times; food reward (in this example) could be presented after the
10th right-side lever press. For this day, discriminative control would be assessed
at 53% diazepam-appropriate responding (i.e., × 100). On Tuesday, this same rat
is injected with vehicle, placed into its designated chamber, and presses the right-
side lever 4 times and the left-side lever 10 times; food is presented after the 10th
left-side press. On this day, discriminative control would be assessed at 29% diaz-
epam-appropriate responding (i.e., × 100). Alternatively, if the VI schedule of rein-
forcement is programmed, then discrimination performance is evaluated during
a short period (e.g., 2.5 min) of non-reinforced responding (referred to as extinc-
tion) at the beginning of a session; extinction sessions usually occur once or twice
per week. Each animal’s distribution of presses on the two levers is then evaluated
in the same manner as it is under the FR schedule of reinforcement. As might be
expected, the administration of drug or vehicle during initial training sessions under
either FR or VI schedules of reinforcement usually results in the animals dividing
their responses equally (e.g., 50% diazepam-appropriate responding after injection
of drug or saline) between the two levers (Figure 3.1). However, as training ses-
sions progress with drug and vehicle, the animals gradually learn to respond on the
drug-designated lever (i.e., percent of responses on the drug-designated lever is high
and percent of responses on the vehicle-designated lever is low) when given drug,
and on the vehicle-designated lever (i.e., percent of responses on the drug-desig-
nated lever is low and percent of responses on the vehicle-designated lever is high)
when given vehicle. In other words, the learning of a DD occurs gradually over time
100
% Diazepam-Appropriate
80
Responding
60 Diazepam (3.0 mg/kg)
40 Vehicle (1.0 ml/kg)
20
0
0 5 10 15 20 25 30 35 40 45 50 55 60
Session
FIGURE 3.1 Learning curve results of rats trained to discriminate the stimulus effect of 3
mg/kg (IP) of diazepam (closed squares) from 1 mL/kg of vehicle (open squares). Ordinate:
Mean (n = 12) percent (± SEM) of responses on the diazepam-appropriate lever after the
administration of diazepam or vehicle. Abscissa: Number of sessions. Note that the animals
gradually learn, as training sessions progress, to respond on the diazepam-designated lever
when administered drug (i.e., percent of responses on the diazepam-assigned lever is high
and percent of responses on the vehicle-designated lever is low) and on the vehicle-designated
lever when administered vehicle (i.e., percent of responses on the diazepam-assigned lever is
low and percent of responses on the vehicle-designated lever is high).

(Figure 3.1). A generally accepted guideline is that after 6 to 9 wk of training, ani-

mals (individually and, consequently, as a group mean) consistently make ≥ 80%
of their responses on the drug-appropriate lever after administration of drug (e.g., 3
mg/kg of diazepam) and ≤ 20% of their responses on the same lever after administra-
tion of vehicle (Figure 3.1).
3.3.2.2 Response Rate

In addition to the animals’ distribution of responses on the two levers under FR or VI
schedules of reinforcement, their response rate data (i.e., total number of responses
on both levers expressed as responses per second or minute) also can be calculated.
For example, animals’ (individual and/or group mean) response rate can be calcu-
lated under the FR schedule of reinforcement for a behavioral session (e.g., 15 min).
Alternatively, under the VI schedule of reinforcement, the total number of responses
made during the 2.5 min extinction session (or the entire session) can be recorded.
The animals’ response rate can be viewed as another indicator of the effect(s) of a
drug on behavior. In some cases, the animals’ response rate after the training dose
of the training drug is suppressed when compared to that of the vehicle. In other
instances, response rate data can assist the experimenter in the selection of (1) an
appropriate training dose of the training drug, and/or (2) a range of doses to be
examined for test drugs. Further, animals’ response rate can be an ancillary mea-
sure in cases where a test drug may, or may not, affect the main dependent variable
(i.e., percent responding on the drug-designated lever). Finally, this parameter can be
used in conjunction with an evaluation of the subjects’ general behavioral condition
(e.g., sedated, incapacitated, overly excited).
3.4 APPLICATIONS
3.4.1 STIMULUS GENERALIZATION
Stimulus generalization of the training dose of the training drug is said to occur to
the test drug if administration of the test substance results in the animals responding
on the drug-designated lever. It is noted, however, that the phrase “stimulus gener-
alization of the vehicle to the test agent” is not used when the animals respond on
the vehicle-designated lever after administration of the test treatment; typically, the
latter result would be characterized as “ the test agent induced vehicle-like respond-
ing.” In the present example, maintenance of the diazepam/vehicle discrimination
was ensured by continuation of training sessions that were intermingled between
stimulus generalization test sessions. Discrimination training sessions were con-
ducted with 3 mg/kg of diazepam or vehicle on the four days prior to a stimulus gen-
eralization test session. On at least one of those days, six of the animals received 3
mg/kg of diazepam and the other six rats received vehicle; percent diazepam-appro-
priate responding was then determined under the FR schedule of reinforcement as
described above. Animals not meeting the above criteria (i.e., ≥ 80% drug-appro-
priate responding after drug administration and ≤ 20% drug-appropriate respond-
ing after vehicle injection) were not used in that week’s stimulus generalization
test. During generalization investigations, test sessions were interposed between

discrimination training sessions. In these test sessions, the rats were given a test
treatment and then allowed to select one of the two levers in a 15-min session (FR
procedure). The lever on which the animal first totaled 10 responses was regarded
as the selected lever; percent diazepam-appropriate lever responding was calculated
as described above. Subsequent reinforcement was delivered for responses on the
selected lever according to the FR 10 schedule of reinforcement. Alternatively, if the
VI schedule of reinforcement was programmed, then the animals would have been
injected with test treatment, given a 2.5 min extinction session, and removed from
the operant chambers; subsequently, percent diazepam-appropriate lever responding
would have been calculated as described above.
3.4.2 TEST CONSIDERATIONS

3.4.2.1 Dose Response
A very important consideration in tests of stimulus generalization is the necessity of
a thorough dose-response investigation. An extensive literature review of DD tests
of stimulus generalization reveals that certain agents produce saline-like effects at
particular doses (usually relatively low doses), and disruption of behavior at some
higher doses. While an initial conclusion to the results of such a study may be that
there is a lack of stimulus generalization, it has been found in a number of instances
that a careful evaluation of additional doses (i.e., doses between the highest dose
that resulted in vehicle-like responding and the lowest dose that produced disrup-
tion of behavior) ultimately resulted in stimulus generalization. This has even been
observed with agents where the difference in vehicle-like and disruptive doses has
been quite small. Thus, several instances have been reported where doses of a chal-
lenge drug, administered in a logarithmic progression (e.g., 0.1, 0.3, 1, 3, 10 mg/kg),
resulted in saline-like responding at the lower doses and disruption of behavior at
the highest doses. However, an examination of doses between, for example, 3 mg/kg
and 10 mg/kg resulted in stimulus generalization. As such, these types of situations
appear to emphasize (1) that stimulus generalization to a test agent may occur within
a “narrow window of doses,” and (2) the sensitive and specific nature of the drug-
induced stimulus.
3.4.2.2 Comparison of Results of Test Agents

A preferred tactic is to evaluate doses of a challenge drug in drug-trained subjects
until either stimulus generalization or disruption of behavior (i.e., no responding)
occurs. If, for example, the highest test dose (e.g., dose X) of a challenge drug elicits
50% drug-appropriate responding, and, for some reason, the evaluation of higher
doses is precluded, it is not appropriate to conclude that the challenge drug is half
as potent as the training drug when, in fact, there has been no demonstration that
the two agents can produce a common effect. In this situation, comparisons can
only be made in a qualitative sense. That is, an appropriate conclusion that could be
stated is that the challenge agent is less effective than the training drug in producing
a training-drug–like effect (or, correspondingly, that it is less effective than some
other challenge drug which, at a dose below dose X, produced training-drug–like

effects). Likewise, if two challenge drugs produce partial generalization (e.g., 40%
and 60% training-drug–appropriate responding) at dose X, it should not be stated
with certainty that the second challenge drug is more potent than the first because
the possibility exists that one (or both) agent(s) may not exert a stimulus effect that is
common (i.e., complete generalization) to that of the training drug.
DATA ANALYSIS, INTERPRETATION, EXAMPLES

In general, the phenomenon of generalization involves engaging in previously learned
behaviors in response to new situations that resemble those in which the behavior
was first learned. In the DD procedure, subjects respond to other drug stimuli that
are more or less similar to those present during discrimination training. Stimulus
generalization studies are used to determine whether a discriminative stimulus will
generalize to (i.e., substitute for) other drugs. The rationale of this approach is that
an animal trained to discriminate a dose of training drug will display stimulus gen-
eralization only to agents having a similar effect, though not necessarily an iden-
tical mechanism of action. Thus, in the present example, stimulus generalization
is said to have occurred when the animals, after being administered a given dose
of a challenge drug, make ≥ 80% of their responses on the diazepam-appropriate
lever. Where stimulus generalization occurs, an effective dose 50% (ED50) value
can be calculated, which reflects the dose at which the animals would be expected
to make 50% of their responses on the diazepam-appropriate lever.47 In addition to
complete stimulus generalization, two other results might be encountered: partial
generalization and vehicle-appropriate responding. Partial generalization is said to
have occurred when the animals, after being administered a thorough dose-effect
test, make approximately 40%–70% of their responses on the diazepam-appropri-
ate lever. In this case, percent diazepam-appropriate lever responding is not fully
appropriate for either training condition. Data of this type are very difficult to inter-
pret. However, it has been posited that partial generalization may occur with a test
compound because there are pharmacological effects that are common to both the
training drug and the challenge drug. Complete stimulus generalization does not
occur, however, because the overlap of effects is incomplete. For example, one expla-
nation for a partial generalization result may be that low doses of a test compound
are similar to low doses of the training drug. However, as the dose of challenge drug
is increased, another kind of pharmacological effect emerges. A third type of test
result is that the administration of various doses of a challenge drug may result in
≤ 20% diazepam-appropriate responding. Such a result does not necessarily mean
that a challenge drug is inert (i.e., without a pharmacological effect), but does suggest
that the effect of the challenge drug is different from that produced by the training
drug. Finally, an important factor in the interpretation of any DD data is that results
must be considered in the context of the training drug. That is, the DD paradigm is
used to generate data that are only valid with respect to a particular dose of a given
training drug. The sensitivity and duration of effect of a stimulus are related to the
training drug and are time dependent. As such, dose response relationships repre-
sent relative, not absolute, relationships between challenge drugs and training drugs.
Finally, it should be recognized that when challenge drugs are being examined, the

data that are obtained relate to training-drug–like effects. For example, an investiga-
tion of the effect of a barbiturate in diazepam-trained animals does not provide data
on barbiturate activity; rather, the data reflect the diazepam-like effect of the barbitu-
rate (see example below). Thus, the result may or may not be the same as the effect of
diazepam or a series of barbiturates in, for example, pentobarbital-trained animals.
3.4.3.1 Statistical Analysis

The use of statistical analysis (e.g., analysis of variance, Fischer’s exact probability,
t-tests, etc.) with DD data can be very problematic. One major concern is the fail-
ure of statistical procedures to account for behavioral disruption (i.e., no respond-
ing) into the analysis. In particular, an animal that fails to press a lever after being
administered a dose of test agent in a stimulus generalization test cannot be assigned
a percent score; 0% drug-appropriate responding cannot be assigned because it has a
different meaning (i.e., the animal pressed the saline-appropriate lever). Some inves-
tigators argue that statistical analysis is robust enough to account for such missing
data. However, the percent drug-appropriate lever responding data are not missing
because the animals failed to appear for a test appointment. The data are lacking and
unavailable because the drug interfered with the ability of the animal to respond. A
more palpable description of such data is that the effect should be characterized as
“disruption.” Statistically, one could (and some investigators do) ignore the disrup-
tive effect of a dose of drug, use only the data from very few animals (sometimes n
= 1 or 2 out of 6 or 8 subjects that were tested) that respond (usually those few sub-
jects have responded to a high degree on the drug-designated lever), and statistically
conclude the occurrence of stimulus generalization. However, in such cases, it would
seem more appropriate and meaningful to characterize the effect as disruption rather
than to promote a statistical conclusion that may be misleading. In any case, the most
prudent approach to the presentation of stimulus generalization data in DD studies is
to account fully, by description of the effects on subjects and/or statistically for the
behavioral effect of the test agent in all subjects, individually and/or as a group.
3.4.3.2 Examples of Complete, Partial, and No Substitution

In rats trained to discriminate the benzodiazepine diazepam at 3 mg/kg from vehicle
at 1 mL/kg, the administration of lower doses of diazepam (i.e., construction of a
dose-response function for diazepam) led to progressively less responding on the
diazepam-appropriate lever (Figure 3.2); furthermore, an ED50 value (ED50 = 1.2 mg/
kg) was calculated. Moreover, several metabolites of diazepam were evaluated in
tests of stimulus generalization. Figure 3.2 shows that the diazepam stimulus gener-
alized in a dose-related manner to oxazepam (ED50 = 1.4 mg/kg), temazepam (ED50
= 1.4 mg/kg), and desmethyldiazepam (ED50 = 2.3 mg/kg). The latter results indi-
cate that, in comparison with diazepam, the metabolites are relatively potent behav-
iorally, and indicate the distinct possibility that the metabolites may contribute to
the stimulus effect of diazepam. Figure 3.2 also reveals that the diazepam stimulus
generalized to the barbiturate anxiolytic/sedative pentobarbital (ED50 = 4.5 mg/kg),
which illustrates the idea that animals trained to discriminate a dose of a training
drug can display stimulus generalization to an agent that exerts a similar behavioral

100 Diazepam
80 Desmethyldiazepam
Responding 60
Temazepam
Oxazepam
40 Pentobarbital
Buspirone
20
0
0.1 1 10
Drug Dose (mg/kg IP)
FIGURE 3.2 Results of stimulus generalization tests with diazepam (closed squares), des-
methyldiazepam (closed triangle), temazepam (closed inverted triangle), oxazepam (closed
diamond), pentobarbital (closed circle), and buspirone (open square) in rats trained to dis-
criminate 3 mg/kg of diazepam from vehicle. Ordinate: Mean (n = 12) percent (± SEM)
of responses on the diazepam-appropriate lever after the administration of the test agents.
Abscissa: Drug dose plotted on a logarithmic scale. Typically, a figure of response rates
would appear below this figure.
effect, although not necessarily through an identical mechanism of action; diazepam

and pentobarbital do not share the same mechanisms of action (see antagonism tests
below). In comparison, the administration of buspirone (0.3–3.0 mg/kg), a serotonin
5-HT1A receptor (partial) agonist anxiolytic agent that is structurally unrelated to
diazepam, produced only partial diazepam-appropriate lever responding (i.e., maxi-
mal 43% diazepam-appropriate lever responding), while the administration of doses
between 4 mg/kg and 10 mg/kg produced disruption of behavior (Figure 3.2; dis-
ruption data not shown). Thus, the diazepam stimulus may partially generalize to
buspirone because there may be some degree of pharmacological effects that is com-
mon to both diazepam and buspirone at low doses. Complete stimulus generaliza-
tion does not occur, however, because the overlap of effects is incomplete. Lastly,
the administration of S(+)-amphetamine (0.1–1.5 mg/kg), a CNS stimulant, to the
diazepam-trained animals produced vehicle-appropriate responding (i.e., maximal
18% diazepam-appropriate lever responding; data not shown in Figure 3.2), while
the administration of doses of 2–3 mg/kg produced disruption of behavior (i.e., no
responding; data not depicted in Figure 3.2). Since percent diazepam-appropriate
lever responding is fairly low (i.e., S(+)-amphetamine–induced responding on the
vehicle-designated lever), it can be stated that the stimulus effect produced by 3.0
mg/kg of diazepam is quite different from that produced by S(+)-amphetamine.
However, the fact that S(+)-amphetamine, and for that matter buspirone, can serve as
training drugs indicates that these are not inert substances. Thus, animals trained in
a DD task respond on the drug-designated lever only when administered a test agent
that produces some degree of effect that is similar to the training dose of the training
drug. If the test agent produces an effect that is “inert” or unlike that of the training
drug, then responding will occur on the vehicle-designated lever until doses of the
test agent are administered that disrupt lever response behavior by the animal (i.e.,
little or no responding). In a final comment, it is noted that results from stimulus gen-

eralization studies of test drugs under particular training agents have been consistent
across different species trained under different schedules of reinforcement.
3.4.3.3 Time Course

Once a dose of a drug has been established as a discriminative stimulus, tests can be
performed to determine its time course of actions. Such tests investigate the effects
of changing the PSII of the training dose of drug and the beginning of a test session.
In particular, the time course of any drug stimulus can be characterized by its latency
of onset of action, peak activity, and duration of effect. The latency of onset of action
refers to the time (or interval) between the administration of the training drug and
the first indications of a marked effects on drug-appropriate responding. The peak
activity of the drug refers to the time (or interval) that the drug exerts maximal per-
cent drug-appropriate responding (i.e., ~80%–100% drug-appropriate responding).
Lastly, the duration of action refers to the interval of time between onset of action
and the point of time (or interval) that the drug no longer exerts percent drug-appro-
priate responding that is notable (i.e., ~20% drug-appropriate responding). In the
present example, 3.0 mg/kg of diazepam was established as a discriminative stimu-
lus with a 15 min PSII; PSII intervals of 5, 10, 30, 45, 90, 120, 180, and 240 min
also were examined (Figure 3.3). The results indicated that the onset of effect of the
diazepam stimulus occurred between the PSIIs of 5 min and 15 min; peak activity
was exhibited from PSIIs of 10 min to approximately 90 min; and duration of action
occurred between PSIIs of 10 min and ~180 min. In addition, Figure 3.3 illustrates
the time course effects of the major metabolites of diazepam. These studies were
conducted with the dose of each metabolite that produced stimulus generalization in
the 3 mg/kg diazepam-trained animals: desmethyldiazepam (6 mg/kg), oxazepam
(3 mg/kg), and temazepam (3 mg/kg). An important difference among benzodiaz-
100
Diazepam
80 Desmethyldiazepam
Temazepam
Responding
60 Oxazepam
40
20
0 30 60 90 120 150 180 210 240

Pre-Session Injection Interval
FIGURE 3.3 Results of time course studies (i.e., stimulus generalization tests with various
pre-session injection intervals) with diazepam (3 mg/kg; closed squares), desmethyldiazepam
(6 mg/kg; open squares), temazepam (3 mg/kg; open triangle), and oxazepam (3 mg/kg; open
circle) in rats trained to discriminate 3 mg/kg of diazepam from vehicle. Ordinate: Mean (n =
12) percent (± SEM) of responses on the diazepam-appropriate lever after the administration
of the test agents. Abscissa: PSII. Typically, a figure of response rates would appear below
this figure.

epines is their pharmacokinetic properties. For example, in humans, diazepam is

considered to have a relatively rapid onset of action and a relatively long half-life.
In comparison, oxazepam and temazepam are considered to have slower onsets of
action and much shorter half-lives, relative to diazepam. The time course studies of
the stimulus effects of these agents in the diazepam-trained animals are not inconsis-
tent with the human data. In any case, it is clear that familiarity with the time-course
of action of the training drug or challenge compounds in tests of stimulus generaliza-
tion and/or stimulus antagonism is of great importance; drug responses should not be
measured too long, or short, after drug administration.
3.4.3.4 Stimulus Antagonism

An effective strategy to determine the mechanisms of action of psychoactive agents
is to study drugs that block their effects. In DD studies, the rationale of such an
approach is that the training dose of the training agent will only be blocked by recep-
tor antagonists that interfere with the mechanism of action of the drug. The results of
antagonism tests, as with generalization tests, typically fall into one of three catego-
ries: (1) complete antagonism (i.e., saline-appropriate responding); (2) partial antago-
nism (i.e., ~ 40%–70% drug-appropriate responding); and (3) no antagonism (i.e.,
≥ 80% drug-appropriate responding). Three strategies to study stimulus antagonism
can be employed. One approach can determine whether the stimulus effect of the
training dose of the training drug can be attenuated when various doses of an appro-
priate receptor antagonist are combined with the training dose of the training drug.
Thus, the rats trained to 3.0 mg/kg of diazepam were administered various doses of
the benzodiazepine receptor antagonist flumazenil prior to the administration of their
training dose of training drug. If a drug is an effective antagonist, then a dose-related
antagonism of the animals’ percent drug-appropriate responding should occur. Fig-
ure 3.4 shows that the administration of various doses of flumazenil (3–12 mg/kg)
prior to the injection of the 3 mg/kg training dose of diazepam was sufficient to
produce antagonism (i.e., responding ultimately occurred on the vehicle-designated
lever). In contrast, the administration of various doses of flumazenil prior to the injec-
tion of the dose of pentobarbital (10 mg/kg) that produced complete stimulus gener-
alization in these animals (see above) failed to produce antagonism (i.e., responding
occurred on the diazepam-designated lever). Lastly, when the animals were adminis-
tered doses of flumazenil (3–40 mg/kg) in control tests, they failed to respond on the
diazepam-designated lever. Taken together, these data support the idea that diazepam
and pentobarbital can induce a similar stimulus effect but the mechanism of action
can be differentiated by flumazenil, a benzodiazepine receptor antagonist.
In a second approach, the dose response of the training drug is determined in
both the presence and absence of a constant dose of the antagonist. If the antago-
nism is competitive, then the dose response of the training drug (in the presence
of the constant dose of the antagonist) should shift in a rightward and parallel
manner. Figure 3.5 (top figure) shows the dose-response effect of diazepam in the
absence (i.e., left dose-response function) and the presence (right dose-response
effect) of flumazenil (5 mg/kg). As can be seen, pretreatment of the animals with
flumazenil induced a rightward shift of the dose-response function of diazepam. In

100
80
Responding 60 Flumazenil and Vehicle

Flumazenil and DZP (3 mg/kg)
40 Flumazenil and PB (10 mg/kg)
20
0
1 10 100
Flumazenil Dose (mg/kg IP)
FIGURE 3.4 The effect of flumazenil administered alone (open squares), in combination
with 3 mg/kg of diazepam (DZP; open triangles), or in combination with 10 mg/kg of pen-
tobarbital (PB; open circles). The administration of various doses of flumazenil prior to the
injection of DZP produced a dose-related antagonism of the stimulus effect of DZP. In con-
trast, the administration of various doses of flumazenil prior to the injection of PB produced
no attenuation of the diazepam-like response of PB. Lastly, flumazenil, administered alone,
did not induce diazepam-appropriate responding. Ordinate: Mean (n = 12) percent (± SEM)
of responses on the diazepam-appropriate lever after the administration of flumazenil (alone
or) in combination with DZP or PB test. Abscissa: Flumazenil drug dose plotted on a loga-
rithmic scale. Typically, a figure of response rates would appear below this figure.
a third approach, various doses of the training drug can be combined with various
doses of the receptor antagonist. This approach will generate a series of training-
drug/antagonist dose-response curves and probably provide the most comprehensive
or detailed picture of the interaction between the agents. Figure 3.5 (bottom figure)
shows the dose-response effect of diazepam in the absence (i.e., left dose-response
function) and the presence (middle and far right dose-response effects) of flumazenil
(5 mg/kg and 12 mg/kg, respectively). Clearly, the dose-response functions of the
discriminative stimulus effect of diazepam were shifted rightward and these data
strongly indicate the presence of competitive antagonism.
3.5 SUMMARY
The DD assay is a behavioral procedure whereby an organism must recognize a
particular drug state, choose a correct response, and receive reinforcement. Most
often, subjects are presented with the choice of two levers: one response (i.e., press of
a left- or right-side lever) should be emitted in the presence of the training dose of a
drug and a similar response (i.e., press of the alternate lever) should be emitted in the
absence of the training drug. All other environmental conditions are held constant.
Overall, DD studies have involved different species (including humans), learning
paradigms (typically two-lever operant choice tasks that employ FR or VI sched-
ules of reinforcement), and either solid or liquid reinforcement. Most often, studies
have used rats that are trained to press levers on FR schedules of reinforcement
for food pellets. Many agents from different psychoactive drug or chemical classes
have been shown to serve as discriminative stimuli. Once trained, subjects can be

100
Vehicle and DZP
Responding 80 Flumazenil (5 mg/kg) and DZP
60
40
20
0
0.1 1 10
Diazepam Dose (mg/kg, IP)
100
Vehicle and DZP
80 Flumazenil (5 mg/kg) and DZP
Responding
Flumazenil (12 mg/kg) and DZP

60
40
20
0
0.1 1 10 100
Diazepam Dose (mg/kg, IP)
FIGURE 3.5 The effect of various doses of diazepam alone (DZP; closed squares) or in
combination with 5 mg/kg of flumazenil (open squares) in rats trained to discriminate 3
mg/kg of diazepam from vehicle (top figure). The bottom figure depicts the effects of various
doses of DZP alone (closed squares) or in combination with 5 mg/kg (open squares) or 12
mg/kg (open circles) of flumazenil. Ordinate: Mean (n = 12) percent (± SEM) of responses on
the diazepam-appropriate lever after the administration of 5 mg/kg or 12 mg/kg of flumazenil
in combination with various doses of DZP. Abscissa: Diazepam drug dose plotted on a loga-
rithmic scale. Typically, a figure of response rates would appear below this figure.
“asked” whether they recognize a novel agent as producing a stimulus effect similar
to that produced by the training dose of the training drug. Several factors can influ-
ence the results of such studies. For example, an important factor to be considered
is the choice of doses to be examined for a particular challenge compound under
investigation; the need for thorough dose-response investigations cannot be over-
emphasized because stimulus generalization can occur within a narrow window of
doses. Moreover, studies have shown that drugs that generalize (substitute, transfer)
to one another in tests of stimulus generalization in animals often produce similar
effects in humans. Lastly, antagonism studies have evaluated the effects of purported
receptor antagonists in combination with the training dose (or other doses) of the
training drug. Such studies can elucidate a neurochemical mechanism involved in
the discriminative stimulus properties of a drug. Taken together, the DD paradigm
can be characterized as a highly sensitive and relatively specific “drug detection”
assay that provides qualitative, quantitative, and mechanistic results of psychoactive
agents.

REFERENCES
1. Beecher, H. K. 1959. Measurement of subjective responses: quantitative effects of
drugs, 193–424. New York: Oxford University Press.
2. Kilbey, M. M., and Ellinwood, E. H. 1979. Discriminative stimulus properties of psy-
chomotor stimulants in the cat. Psychopharmacology 63:151–153.
3. Cook, L., Davidson, A., Davies, D. J., and Kelleher, R. G. 1960. Epinephrine, nor-
epinephrine, and acetylcholine as conditioned stimuli for avoidance behavior. Science
131:990–991.
4. Hudzik, T. J., Yanek, M., Porrey, T., et al. 2003. Behavioral pharmacology of AR-
A000002, a novel, selective 5-hydroxytryptamine1B antagonist. J. Pharmacol. Exp.
Ther. 304:1072–1084.
5. Jarbe, T. U. C., Johansson, J. O., and Henriksson, B. G. 1975. Delta-9-tetrahydrocan-
nabinol and pentobarbital as discriminative cues in the mongolian gerbil (Meriones
unguiculatus). Pharmacol. Biochem. Behav. 3:403–410.
6. Altman, J. L., Albert, J., Milstein, S. L., and Greenberg, I. 1976. Drugs as discrimina-
tive events in humans. Psychopharmacol. Commun. 2:327–330.
7. Snoddy, A. M., and Tessel, R. E. 1983. Nisoxetine and amphetamine share discrimina-
tive stimulus properties in mice. Pharmacol. Biochem. Behav. 19:205–210.
8. Schuster, C. R., and Brady, J. V. 1971. The discriminative control of a food-reinforced
operant by interoceptive stimulation. In Stimulus properties of drugs, ed. T. Thompson
and R. Pickens, 1:133–148. New York: Appleton-Century Crofts.
9. Carey, M. P., Fry, J. P. 1991. A behavioral and pharmacological evaluation of the dis-
criminative stimulus induced by pentylenetetrazole in the pig. Psychopharmacology
111:244–250.
10. Henriksson, B. G., Johansson, J. O., and Jarbe, T. U. C. 1975. Delta-9-tetrahydrocan-
nabinol produced discrimination in pigeons. Pharmacol. Biochem. Behav. 3:771–774.
11. Barry, H. III, 1968. Prolonged measurements of discrimination between alcohol and
nondrug states. J. Comp. Psychol. Psychol. 65:349–352.
12. Chait, L. D., Uhlenhuth, E. H., and Johanson, C. E. 1984. An experimental paradigm
for studying the discriminative stimulus properties of drugs in humans. Psychophar-
macology 82:272–274.
13. Kamien, J. B., Bickel, W. K., Hughes, J. R., Higgins, S. T., and Smith, B. J. 1993. Drug
discrimination by humans compared to nonhumans: Current status and future direc-
tions. Psychopharmacology 111:259–270.
14. Schechter, M. D., and Rosecrans, J. A. 1973. D-amphetamine as a discriminative cue:
Drugs with similar stimulus properties. Eur. J. Pharmacol. 21:212–216.
15. Colpaert, F. C., Niemegeers, C. J. E., Kuyps, J. J. M. D., and Janssen, P. A. J. 1975.
Apomorphine as a discriminative stimulus, and its antagonism by haloperidol. Eur J.
Pharmacol. 32:383–386.
16. Barry, H. III, and Kubena, R. K. 1972. Discriminative stimulus characteristics of alco-
hol, marihuana and atropine. In Drug addiction 1: Experimental pharmacology, ed. J.
M. Singh, L. Miller, and H. Lal, 3–16. New York: Futura.
17. Holtzman, S. G. 1997. Discriminative stimulus effects of buprenorphine in the rat.
Psychopharmacology 130:292–299.
18. Hendry, J. S., Balster, R. L., and Rosecrans, J. A. 1983. Discriminative stimulus proper-
ties of buspirone compared to central nervous system depressants in rats. Pharmacol.
Biochem. Behav. 19:97–101.
19. Carney, J. M., and Christensen, H. D. 1980. Discriminative stimulus properties of caf-
feine: Studies using pure and natural products. Pharmacol. Biochem. Behav. 13:313.

20. De Witte, P. H., Swanet, E., Gewiss, M., Goldman, S., Roques, B., and Vanderhaeghen,
J. 1985. Psychopharmacological profile of cholecystokinin using the self- stimulation
and drug discrimination paradigms. Ann. NY Acad. Sci. 448:470–487.
21. Goas, J. A., and Boston, J. E. 1978. Discriminative stimulus properties of clozapine and
chlorpromazine. Pharmacol. Biochem. Behav. 8:235–241.
22. Browne, R. G., and Koe, B. K. 1982. Clozapine and agents with similar behavioral and
biochemical properties. In Drug discrimination: Applications in CNS pharmacology,
ed. F. C. Colpaert and J. L. Slangen, 241–254. Amsterdam: Elsevier.
23. Jarbe, T. U. C. 1978. Cocaine as a discriminative cue in rats: Interactions with neuro-
leptics and other drugs. Psychopharmacology 59:183–187.
24. Shearman, G. T., Miksic, S., and Lal, H. 1978. Discriminative stimulus properties of
desipramine. Neuropharmacology 17:1045–1048.
25. Holtzman, S. G. 1994. Discriminative stimulus effects of dextromethorphan in the rat.
26. Young, R., Glennon, R. A., Brasse, D. A., and Dewey, W. L. 1986. Potencies of diaz-
epam metabolites in rats trained to discriminate diazepam. Life Sci. 39:17–20.
27. Winter, J. C. 1985. Sedation and the stimulus properties of antihistamines. Pharmacol.
28. Young, R., Glennon, R. A., and Rosecrans, J. A. 1980. Discriminative stimulus proper-
ties of the hallucinogenic agent DOM. Commun. Psychopharmacol. 4:501–506.
29. Young, R., and Glennon, R. A. 1998. Discriminative stimulus properties of (-)ephed-
rine. Pharmacol. Biochem. Behav. 60:771–775.
30. Schechter, M. D. 1974. Effect of propranolol, d-amphetamine and caffeine on ethanol
as a discriminative cue. Eur. J. Pharmacol. 29:52–57.
31. Goudie, A. J. 1977. Discriminative stimulus properties of fenfluramine in an operant
task: An analysis of its cue function. Psychopharmacology 53:97–102.
32. Colpaert, F. C., and Niemegeers, C. J. E. 1975. On the narcotic cuing action of fentanyl
and other narcotic analgesic drugs. Arch. Int. Pharmacodyn. Ther. 217:170–172.
33. Hirschhorn, I. D., and Winter, J. C. 1971. Mescaline and lysergic acid diethylamide
(LSD) as discriminative stimuli. Psychopharmacologia 22:64–71.
34. Glennon, R. A., and Young, R. 1984. MDA: A psychoactive agent with dual stimulus
effects. Life Sci. 34:379–383.
35. Glennon, R. A., and Misenheimer, B. R. 1989. Stimulus effects of N-monoethyl-1-
(3,4-methylendioxyphenyl)-2-aminopropane (MDE) and N-hydroxy-1-(3,4-methylene-
dioxyphenyl)-2-aminopropane (N-OH MDA) in rats trained to discriminate MDMA
from saline. Pharmacol. Biochem. Behav. 33:909–912.
36. Hirschhorn, I. D., and Rosecrans, J. A. 1974. A comparison of the stimulus effects
of morphine and lysergic acid diethylamide (LSD). Pharmacol. Biochem. Behav.
2:361–366.
37. Carter, R. B., and Leander, J. D. 1982. Discriminative stimulus properties of naloxone.
38. Schechter, M. D., and Rosecrans, J. A. 1972. Nicotine as a discriminative cue in rats:
Inability of related drugs to produce a nicotine-like cueing effect. Psychopharmacolo-
gia 27:379–387.
39. Willetts, J., and Balster, R.,L. 1989. Effects of competitive and non competitive N-
methyl-D-aspartate (NMDA) antagonists in rats trained to discriminate NMDA from
saline. J. Pharmacol. Exp. Ther. 251:627–633.
40. Kuhn, D. M., Greenberg, I., and Appel, J. B. 1976. Stimulus properties of the narcotic
antagonist pentazocine similarity to morphine and antagonism by naloxone. J. Phar-
macol. Exp. Ther. 196:121–127.
41. Herling, S., Valentino, R. J., and Winger, G. D. 1980. Discriminative stimulus effects of
pentobarbital in pigeons. Psychopharmacology 71:21–28.

42. Brady, K. T., and Balster, R. L. 1981. Discriminative stimulus properties of phen-
cyclidine and five analogues in the squirrel monkey. Pharmacol. Biochem. Behav.
14:213–218.
43. Vanover, K. E. 1997. Discriminative stimulus effects of the endogenous neuroactive
steroid pregnanolone. Eur. J. Pharmacol. 327:97–101.
44. Jarbe, T. U. C., Henriksson, B. G., and Ohlin, G. C. 1977. Delta-9-THC as a discrimina-
tive cue in pigeons: Effects of delta-8-THC, CBD and CBN. Arch. Int. Pharmacodyn.
Ther. 228:68–72.
45. Rees, D. C., Knisely, J. S., Jordan, S., and Balster, R. L. 1987. Discriminative stimulus
properties of toluene in the mouse. Toxicol. Appl. Pharmacol. 88:97–104.
46. Extance, K., and Goudie, A. J. 1981. Inter-animal olfactory cues in operant drug dis-
crimination procedures in rats. Psychopharmacology 73:363–371.
47. Finney, D. 1952. Probit analysis. 183–197. London: Cambridge University Press.

4 Conditioned
Preference
Place
Adam J. Prus, John R. James,

and John A. Rosecrans
CONTENTS
4.1 Introduction................................................................................................... 59
4.2 Research Design and Methodological Considerations..................................60
4.3 Training and Testing ..................................................................................... 63
4.4 Drug Studies Using the Conditioned Place Preference Paradigm ................64
4.5 The Mesolimbic Dopamine System Is Important for Conditioned Place
Preference......................................................................................................64
4.6 Mechanisms Mediating Conditioned Place Preference for Common
Drugs of Abuse ............................................................................................. 65
4.6.1 Opiates ............................................................................................... 65
4.6.2 Psychostimulants................................................................................66
4.6.3 Nicotine.............................................................................................. 67
4.6.4 Ethanol ............................................................................................... 67
4.6.5 MDMA............................................................................................... 68
4.6.6 Delta-9-THC and Endocannabinoids ................................................. 68
4.7 Conditioned Place Preference Versus Self-Administration .......................... 69
4.8 Summary....................................................................................................... 70
Acknowledgments.................................................................................................... 70
References................................................................................................................ 70
4.1 INTRODUCTION
The conditioned place preference paradigm is a standard preclinical behavioral
model used to study the rewarding and aversive effects of drugs. Although a number
of different designs and apparatuses are used in this model, the basic characteristics
of this task involve the association of a particular environment with drug treatment,
followed by the association of a different environment with the absence of the drug
(i.e., the drug’s vehicle). A common variation of this design consists of a three-com-
partment chamber with the outer compartments being designed to have different
characteristics (e.g., white vs. black walls, pine vs. corn bedding, horizontal grid vs.
cross-grid flooring). The center compartment has no special characteristics and is
not paired with a drug, and the gates between the compartments can be opened to
allow an animal to pass freely between them. During training, an animal (typically
59

TABLE 4.1
Common Neuroscience Techniques Used in Conditioned Place Preference
Research
Technique Example References
Lesioning 64, Spyraki et al. 1982b
Knockout mice 31
Microdialysis 53, Duvauchelle et al. 2000a, 55
Microinfusion 32, Rezayof et al. 2007b, 47
Neurotransmitter depletion 55
Strain comparisons 13
a rat or mouse) is given an injection of a drug with potentially rewarding or aversive

properties, and is then placed into one of the outer compartments for several minutes.
On the following day, the rat is injected with the drug’s vehicle and then placed in the
opposite compartment. Generally, these daily sessions alternate between drug and
vehicle for 2 or 3 days each. Afterward, a test session is conducted, which consists
of placing the animal in the center compartment and then, after opening the gates
to both of the outer compartments, recording the time the animal spends in each of
the outer compartments during the session. A conditioned place preference (CPP) is
found if the animals spend significantly more time in the drug-paired compartment
versus the vehicle-paired compartment. On the other hand, if the animals spend
significantly more time in the vehicle-paired compartment versus the drug-paired
compartment, then this is considered a conditioned place aversion (CPA). Typically,
drugs of abuse, such as cocaine, produce CPP, and drugs that elicit aversive effects,
such as lithium chloride, produce CPA. As with other behavioral models used in
pharmacology research, the behavioral effects of drugs used in the CPP paradigm
depend on species, strain, route of administration, time interval of drug administra-
tion, dose concentration, and the CPP apparatus used. Many drugs of abuse produce
both CPP and CPA, depending on the dose administered. In drug-dependent ani-
mals, withdrawal effects generally produce CPA. Because the CPP paradigm gen-
erally provides a reliable indicator for studying the rewarding effects of drugs that
require relatively little training compared to self-administration paradigm, the CPP
paradigm has been commonly used in conjunction with standard neuroscience tech-
niques to elucidate the subjective effects of drugs (Table 4.1).
4.2 RESEARCH DESIGN AND

METHODOLOGICAL CONSIDERATIONS
Although this chapter focuses primarily on studying the effects of drugs of abuse
in the CPP model, CPP has also been established with food, copulatory activity,
and other rewarding stimuli. The ability of a stimulus, whether it be a drug, food,
etc., to produce a preference for the associated environment is generally considered
a process governed by Pavlovian, i.e., classical or respondent, conditioning. Using
the rewarding effects of a drug as an example, repeatedly pairing the rewarding

Conditioned Place Preference 61
effects of the drug (unconditioned stimulus, using Pavlov’s terminology) with certain
stimuli (i.e., those contained within an environment) would be expected to result in
the extension of these rewarding effects to the properties of these previously neu-
tral stimuli. Thus, the drug-paired environment eventually serves as a conditioned
stimulus (CS). Although it is not known if the compartment would actually serve as
a CS in its truest sense (i.e., elicit rewarding effects similar to those produced by the
drug of abuse), dopamine (DA) levels in the nucleus accumbens have been found to
be elevated when rats are placed in the drug-paired environment, compared to the
nondrug-paired environment.1
Although the three-compartment chamber described above is a common appa-
ratus used in CPP research (Figure 4.1), other apparatuses vary from this design
by having a different number of compartments (e.g., two or four compartments),
assessing place preference within an open field, or allowing for the association of
the interoceptive effects of drugs with a unique environment. Although all of these
approaches have been used to study CPP, an important consideration in the choice of
an apparatus is the decision to have a “forced choice” (i.e., the animal must choose
the drug-paired side or the nondrug-paired side) or an “unforced choice” (i.e., the
animal can remain in a compartment or other area of the apparatus that has not been
associated with drug or vehicle) (Figure 4.2, top panel). Thus, a two-compartment
apparatus would require a forced choice, whereas a three-compartment area could
offer a central choice area between the experimental chambers. Although commonly
used, the central concern of using a forced choice procedure is the potential of a bias
for the compartment the animal was placed in during the test session.
Another important consideration in CPP research is the use of biased versus
unbiased research designs. These research designs are used to take into consideration
the fact that subjects may have an initial preference for a particular compartment of
the apparatus. For instance, if subjects were assessed for place preference in a two-
compartment apparatus prior to conditioning, some subjects would spend more time
in compartment A, whereas other subjects would spend more time in compartment
B. In an unbiased CPP study, the assignment of a particular compartment for pair-
ing with a drug is determined by the researcher, regardless of the preference of each
subject for either compartment prior to conditioning (see Figure 4.2, bottom panel).
FIGURE 4.1 Standard two- (left) and three-chamber (right) shuttle boxes used to study con-
ditioned place preference in rodents. Source: From Med-Associates, Inc., with permission.

Methods of Behavior Analysis in Neuroscience, Second Edition
Unforced Choice Forced Choice

A B
A B ?
Stay or
?
OR
A B
Stay or
Biased Unbiased
A B
A B
Most Preferred Least Preferred
Vehicle Drug
Vehicle Drug
?
FIgure . Unforced versus forced choice procedures (top panel) and biased versus
unbiased designs (bottom panel) used in the conditioned place preference paradigm. In an
unforced choice procedure, the subject is placed in a central choice area between the parts
of the apparatus used for conditioning. In the forced choice procedure, a central choice area
is not used. In the biased design, the baseline preference of the subjects for each part of
the apparatus is assessed before condition sessions are begun. Each subject’s least preferred
compartment is paired with the drug, and the most preferred compartment is paired with the
drug’s vehicle. In the unbiased design, assignment of drug or vehicle pairing with each com-
partment is made regardless of each subject’s baseline preference.
In a biased design, the preference of each individual subject for a particular environ-
ment prior to conditioning is assessed first by placing the animals in the apparatus,
and then by assessing the amount of time the subjects spend in each compartment.
The least-preferred compartment for each subject is then assigned to be the drug-
paired compartment. Depending on the design used in a CPP study, different results

100 Baseline Nicotine Paired with 100 Baseline Nicotine Paired with
Preference Most Preferred Side Preference Least Preferred Side
80 80
Percent Preference
Percent Preference
60 60
*
40 40
20 20
0 0
Least Most Least Most Least Most Least Most
(a) (b)
FIGURE 4.3 Effects of nicotine (0.4 mg/kg) in the conditioned place preference paradigm
in a biased design with (A) nicotine paired with the most preferred side and (B) nicotine
paired with the least preferred side. Baseline preferences for each compartment were assessed
prior to conditioning. *p < 0.05 for the baseline preference versus preference after pairing
with nicotine. Source: This figure was produced from data reported in Calcagnetti, D. J., and
Schechter, M. D. 1994. Nicotine place preference using the biased method of conditioning.
Prog. Neuropsychopharmacol. Biol. Psychiatry 18:925.
may occur. For example, early CPP2 studies with nicotine found discrepant find-
ings between laboratories, which included no effect, CPP, or CPA. In an attempt
to clarify these discrepancies, Calcagnetti and Schechter2 tested nicotine for CPP
by first assessing the most and least preferred sides of a three-chamber shuttle box.
After baseline preferences were assessed, half of the rats were assigned nicotine for
the least preferred side and the remaining half were assigned nicotine for the most
preferred side. Nicotine produced a CPP with the least preferred side, but failed to
develop a CPP or CPA for the most preferred side (Figure 4.3). Consequently, ran-
domly assigning compartments to be paired with nicotine without assessing baseline
preferences may not result in a CPP.
Other important methodological procedures that should be considered in CPP
research include the drug’s time course, the number of conditioning sessions, and
the sensory modalities used to discriminate between environments. Generally, drugs
that have a slow onset and a long duration of action (e.g., phenobarbital) are not good
reinforcers, and consequently, may not readily establish CPPs. For drugs that have
potent rewarding properties, fewer conditioning sessions will be needed to estab-
lish a CPP (e.g., amphetamine), whereas drugs with weaker rewarding properties
may require more conditioning sessions (e.g., nicotine). Finally, sensory modalities
should be appropriate for the species being used. For example, visual cues are a poor
choice for albino rats, whereas olfactory cues are an excellent choice for these rats.
Tactile and auditory cues are also good choices when using rodents.
4.3 TRAINING AND TESTING

Several days of free access to all environments allows the animal to habituate to
the apparatus, eliminating novelty as a confounding variable. Baseline data should
be determined as an average amount of time spent in each chamber over 3–5 days.

The length of time necessary to determine baselines depends on the environmental

differences between apparatus chambers. Generally, distinct environments require
fewer baseline sessions, whereas less distinct or ambiguous environments require
more baseline sessions.
The testing of subjects is performed in a non-manipulated state. The percentage
of time spent in each chamber is tabulated during the test session. If the animal spends
more time in the chamber associated with treatment, the researcher can infer that the
experimental manipulation had a rewarding effect on the affective state of the animal.
If the opposite were to occur, then the researcher can infer that the experimental
manipulation had an aversive effect on the affective state of the animal. Designs that
include a novel chamber (e.g., the center chamber of a three-chamber shuttle box) do
not allow free access to the novel chamber during the habituation period.
4.4 DRUG STUDIES USING THE CONDITIONED

PLACE PREFERENCE PARADIGM
The CPP paradigm has been widely used in pharmacology, behavioral science, and
neuroscience research. A recent database search of Pubmed (http://www.pubmed.
gov) using the keywords “conditioned place preference” yielded 1398 results. The
CPP paradigm has not simply been used as a screening tool for drug abuse potential,
but has been used to study neurotransmitters, brain areas, genes, signaling path-
ways, and other mechanisms mediating the rewarding (or aversive) effects of drugs.
The drugs studied in CPP have been the subject of many reviews.3,4 Generally, psy-
chostimulants and opiates reliably produce a CPP in this paradigm. For example,
systemic administration of cocaine, amphetamine, and nicotine have been found to
produce a CPP after two or three pairings in rats and mice.5–14 In addition, CPP has
been established with opiates, such as morphine, heroin, and buprenorphine, as well
as drugs from other classes, including the CNS depressants ethanol and diazepam,
the cannabinoid receptor agonist delta-9-tetrahydrocannabinal (THC), and the adre-
noceptor agonist clonidine.15–19,8,20–32 Many of these compounds also produce CPA,
depending on the dose administered. For example, nicotine produces CPP in rats
that are administered doses between 0.4 and 0.8 mg/kg., whereas higher doses are
reported to produce CPA.5,33 Similar findings have been shown with other drugs,
such as morphine and the psychostimulant apomorphine.34–39
4.5 THE MESOLIMBIC DOPAMINE SYSTEM IS

IMPORTANT FOR CONDITIONED PLACE PREFERENCE
Although these drugs differ in their CNS effects, the majority of CPP-producing
drugs affect the mesolimbic DA system, which consists of DA pathways that origi-
nate in the ventral tegmental area (the A10 region in rodents) and terminate in limbic
system structures, including the nucleus accumbens and hippocampus. Therefore,
DA D2 receptor antagonists, such as haloperidol and metoclopramide, have been
found to block CPP or CPA produced by systemically administered amphetamine,
cocaine, morphine, and heroin.40–45 Moreover, direct injection of psychostimulants
and opiates into the ventral tegmental area or the nucleus accumbens also produces

CPP, whereas direct injection of amphetamine or morphine into other areas, such as
the prefrontal cortex, caudate, or amygdala, fails to produce CPP or CPA.46–52 In rats
that developed a CPP for cocaine after conditioning, significantly greater elevations
in DA levels were found in the nucleus accumbens after vehicle injection when rats
were placed in the cocaine-paired compartment, as opposed to the vehicle-paired
compartment.53,54 However, DA levels in the prefrontal cortex have also been found
to be elevated in rats placed in an amphetamine-paired compartment after several
days of conditioning, and selective depletion of prefrontal cortical norepinephrine
prevents amphetamine- and morphine-induced CPP and amphetamine- and mor-
phine-induced DA release in the nucleus accumbens.1,55,56 Selective lesioning of DA
terminals using 6-hydroxydopamine in the ventral pallidum, another target region of
mesolimbic DA neurons, has been shown to attenuate the development of cocaine-
induced CPP.57 CPP has also resulted from morphine infusion into the hippocam-
pus.58 Thus, although the nucleus accumbens is an important region that mediates
the effects of drugs of abuse, other limbic structures, as well as structures that medi-
ate limbic system function, may alter the ability of drugs of abuse to elicit CPP.
4.6 MECHANISMS MEDIATING CONDITIONED PLACE

PREFERENCE FOR COMMON DRUGS OF ABUSE
4.6.1 OPIATES
As presented earlier, opiates, such as morphine, are capable of producing CPP in
rats and mice. The ability of opiate drugs to elicit a CPP depends, in part, on the
release of dopamine from mesolimbic DA neurons, because morphine- and heroin-
induced CPP can be blocked by the DA D2 receptor antagonists haloperidol and
metoclopramide. Furthermore, direct injection of opiates into the ventral tegmental
area or nucleus accumbens also produces CPP.40,42,46–48,49,51 Opiates are known to
enhance mesolimbic DA neuronal firing and ultimately release DA into the nucleus
accumbens, which has been shown to be caused by a disinhibition of DA neurons in
the ventral tegmental area through attenuating gamma-aminobutyric acid (GABA)
release in the ventral tegmental area. The μ-opioid receptor has long been identified
as the key receptor for mediating the subjective effects of opiates, and is also likely
responsible for mediating opiate-induced CPP. The opioid receptor partial agonist
buprenorphine has been shown to produce CPP and increase locomotor activity in
wild type (WT) mice, but not in μ-receptor knockout (KO) mutant mice; whereas
amphetamine has been shown to produce CPP and increase locomotor activity in
both WT and μ KO mice.31 Moreover, agonists selective for μ-opioid receptors can
reinstate CPP in formerly morphine-dependent rats trained in the CPP procedure.59
Other neurotransmitters may also be important for the development of CPP
by opiates. CPP produced by morphine has been shown to be attenuated after pre-
treatment with the cannabinoid CB1 receptor antagonist SR-141716.60 In addition
to potential modulation of morphine-induced CPP by cannabinoid receptors, intra-
ventral tegmental area infusions of nicotine and the acetylcholinesterase inhibitor
neostigmine have been found to facilitate CPP produced by an otherwise ineffective
dose (0.5 mg/kg) of morphine. Intraventral tegmental area infusions of the musca-

rinic receptor antagonist atropine and the nicotinic receptor antagonist mecamyla-
mine have been found to prevent CPP produced by an effective dose of morphine
(5.0 mg/kg), suggesting that the cholinergic system may also mediate the rewarding
effects of morphine and other opiates.32 Local administration of the glutamate ion
channel agonist N-methyl-D-aspartate (NMDA) into the amygdala has been shown
to potentiate morphine-induced CPP, whereas the glutamate ion channel blocker
MK-801 has been shown to attenuate morphine-induced CPP.61
4.6.2 PSYCHOSTIMULANTS
As noted above, psychostimulants such as amphetamine and cocaine often produce
a robust CPP, and these effects also depend upon limbic system functioning, par-
ticularly on the release of DA into the nucleus accumbens. Although psychostim-
ulants as a class tend to produce rewarding effects, as demonstrated in CPP and
self-administration studies, drugs in this class vary greatly as to the mechanisms
mediating reward. Apomorphine, a psychostimulant and classic DA receptor ago-
nist, has been reported to produce CPP in numerous studies, as well as CPA at rela-
tively high doses.34,41,62,63 The rewarding effects of apomorphine are potentiated in
the CPP model by pretreatment with the DA D3 receptor agonist 7-OH-DPAT, yet in
the same study pretreatment with 7-OH-DPAT was found to attenuate the rewarding
effects of cocaine.63 Moreover, systemic administration of the typical antipsychotic
drug and D2 receptor antagonist haloperidol, as well as 6-OH-DA lesions in the
nucleus accumbens, failed to attenuate cocaine-induced CPP.64 However, another
study reported that haloperidol did prevent cocaine-induced CPP when cocaine was
administered intravenously.44 The D1 receptor antagonist SCH23390 also has been
shown to block cocaine-induced CPP in both male and female rats.65 Cocaine, unlike
apomorphine and other psychostimulants, with the exception of amphetamine,
which is a competitor with DA for the vesicular DA transporter, is an inhibitor of
DA, norepinephrine, and serotonin transporters; although the dopamine transporter
(DAT) inhibition is generally thought to be most important for the rewarding effects
of cocaine. However, DAT KO mice exhibit a CPP for cocaine and are still found to
self-administer cocaine.66,67 However, cocaine-induced CPP in DAT KO mice may
be due to compensatory changes in DA systems in the DAT KO mice, given that
cocaine-induced CPP is not found in a triple mutant DAT KO mouse line that results
in a relatively cocaine-insensitive DAT that still transports DA.68 Amphetamine-
induced CPP is also abolished in DAT KO mice, but can be blocked by haloperi-
dol.4,41,43 The rewarding effects of amphetamine may be mediated, at least in part, by
serotonin receptors, given that the amphetamine-induced CPP is also blocked by the
5-HT2A/2B/2C receptor antagonist ritanserin and by the 5-HT reuptake inhibitors zem-
ilidine and fluoxetine.69–71 Despite differences between cocaine and amphetamine
in the CPP paradigm, both cocaine and amphetamine have been reported to elevate
cocaine- and amphetamine-regulated transcript (CART) mRNA levels after acute
administration in the nucleus accumbens and ventral tegmental area, and bilateral
intraventral tegmental area injections of the CART peptide fragment 55-102 have
been found to produce CPP.72–74 Moreover, CART 55-102–induced CPP was blocked
by systemic administration of haloperidol.

4.6.3 NICOTINE
Systemic administration of nicotine has been shown to produce both CPP and CPA
in rodents through stimulation of nicotinic acetylcholine receptors (nAChrs).5,75,76
Both CPP and CPA is observed at low and high doses, respectively, after intra-ventral
tegmental area infusions of nicotine, which can be blocked by both the B4C2 nAChr
antagonist DHbeteE and the B7 nAChr antagonist methyllycaconitine (MLA).52
Moreover, MLA pretreatment shifted nicotine-induced CPP to CPA in this study.
In nicotine-dependent rats, pairing withdrawal effects induced by administration of
the nonselective nAChr antagonist mecamylamine produces CPA, which can be pre-
vented upon coadministration of the 5-HT3 receptor antagonist ondansetron.77 The
nAChr antagonist epibatidine produces a relatively weak CPP when administered
systemically alone, but also produces a CPA at higher doses, and further nAChr
studies have revealed that C2, but not B7, nAChrs are necessary to establish nicotine-
induced CPP.78,79
The ability of nicotine to produce CPP differs markedly between strains of rats.
For example, nicotine CPP has been established in the Lewis strain of rat, but not
in the Fischer-344 strain of rat.10,13 Further individual differences in susceptibility
to nicotine dependence have been shown within the same strain of mice, in which
a single injection of nicotine (0.75 mg/kg) was found to either increase or decrease
locomotor activity, and subsequent testing for nicotine CPP found that nicotine pro-
duced CPP in mice that had increased locomotor activity after nicotine administra-
tion, and that nicotine produced a lower degree of CPP in mice that had decreased
locomotor activity after nicotine administration.80
4.6.4 ETHANOL
Ethanol, when administered alone, produces CPP and CPA in rodents, with lower
doses producing CPP and higher doses producing CPA.21,24,29,81 Receptor mecha-
nisms found to mediate ethanol’s CNS effects, GABAA, NMDA, and 5-HT3 recep-
tors, have also been tested to determine which receptors mediate reward using the
CPP paradigm. Ethanol-induced CPP was shown to be attenuated when ethanol was
coadministered with the competitive NMDA receptor antagonist CGP-37849, but not
when coadministered with the noncompetitive NMDA receptor antagonists MK-801
and ketamine, nor with NMDA subunit antagonists.82 Again, the mesolimbic DA
pathway appears critical for the rewarding effects of ethanol, since ethanol-induced
CPP can be potentiated by systemically administered heroin, and can be attenu-
ated by intra-accumbens administration of D2 receptor antagonists, such as fluphen-
azine.83 Ethanol has been found to potentiate the effects of cocaine in the CPP model
by shifting high cocaine doses from producing CPP to CPA and by increasing CPP
induced by lower cocaine doses.84,85
In the liver, ethanol is broken down by alcohol dehydrogenase to acetaldehyde,
which in turn is broken down by aldehyde dehydrogenase to acetic acid. When
acetaldehyde accumulates, symptoms of acetaldehyde syndrome may occur, which
include nausea, headache, and vomiting. In the CPP paradigm, acetaldehyde has
been shown to produce CPP, but not CPA, including doses that approached the lethal
limit.81 Intriguingly, deactivation of acetaldehyde by d-penicillamine prevents etha-

nol-induced CPP, whereas ethanol-induced CPA is unaffected, adding to evidence

from other studies that acetaldehyde may mediate the rewarding effects (e.g., eupho-
ria) of ethanol.86–89
4.6.5 MDMA
3,4-Methylenedioxymethamphetamine (MDMA) has been shown to readily estab-
lish a CPP in rodents.90–92 The ability of MDMA to produce a CPP may be caused by
effects on the mesolimbic DA pathways, based on a microdialysis study, which found
that doses of MDMA that produced CPPs also significantly elevated levels of DA
and lowered levels of the DA metabolite DOPAC in the nucleus accumbens.93 More-
over, MDMA-induced CPP is attenuated upon pretreatment with the 5-HT3 recep-
tor antagonist MDL72222 and tropisetron.94,95 MDMA-induced CPP has also been
found to be diminished by pretreatment of the cannabinoid CB1 receptor antagonist
SR141716A and the opioid antagonist naltrexone.95 In adolescents, the neurotoxic
effects of MDMA appear to be diminished, suggesting that the MDMA receptors
become more prominent in later development, perhaps during puberty. In a study by
Fone et al.,96 cocaine produced CPP in adolescent rats previously treated for three
consecutive days with MDMA, whereas cocaine failed to produce CPP in adolescent
rats treated for seven days with the MDMA vehicle. Aberg et al.97 found a similar
effect in adolescent rats, and interestingly, these effects were reversed in adult rats;
MDMA-pretreated rats exhibited a diminished CPP for cocaine compared to vehicle
pretreated rats. The rewarding effects of MDMA have also been shown to be poten-
tiated when coadministered with delta-9-THC, the principle psychoactive ingredi-
ent in cannabis. Robledo et al.98 found that doses of delta-9-THC (0.3 mg/kg) and
MDMA (3.0 mg/kg) that did not produce CPP when administered alone, did produce
CPP when coadministered.
4.6.6 DELTA-9-THC AND ENDOCANNABINOIDS

Delta-9-THC, the psychoactive ingredient in smoked cannabis, has been shown to
produce CPP under certain conditions. Initial findings of THC CPP were reported
in rats using doses that ranged from 2.0–4.0 mg/kg, but a study that came out soon
after reported that a THC CPP was not demonstrated by a 1.5 mg/kg dose in rats.99,100
In fact, the later study reported a CPA to THC following a 15 mg/kg dose, and that
the cannabinoid CB1 receptor antagonist SR141716A produced a CPP.100 A CPA to
THC has also been found in mice.25,101 Recently, the rewarding effects of THC in the
CPP paradigm have been shown in mice after administering an injection of THC (as
a priming dose) 24 hr prior to beginning several daily conditioning sessions with
THC. This modified procedure resulted in a CPP for THC.25,102 In this methodologi-
cal variation, the initial administration of THC may have resulted in tolerance to the
dysphoric, or otherwise aversive, effects of THC, thus enabling the rewarding effects
of THC to become more salient during the following conditioning trials with THC.
However, the inability of some studies to establish a CPP for THC may have been
a result of the doses used, since Braida et al.103 reported that THC, at doses rang-
ing from 0.075–0.75 mg/kg, produced a CPP in Wistar rats. The opioid system may
mediate the effects of THC in the CPP paradigm, because CPP for THC has been

shown to be attenuated in μ/I-opioid receptor double-KO mice, and that chronic

administration of THC produces cross-tolerance to the rewarding effects of mor-
phine in the CPP paradigm.102,104
The endogenous cannabinoid anandamide has not been well characterized in
the CPP paradigm, but doses up to 16.0 mg/kg have failed to produce a CPP, despite
coadministration of a protease inhibitor to lengthen the half-life of anandamide.105
However, the anandamide transport inhibitor AM404 was shown to produce a CPP
in rats raised in an enriched environment, suggesting that anandamide may be capa-
ble of producing a CPP under certain conditions.106
4.7 CONDITIONED PLACE PREFERENCE

VERSUS SELF-ADMINISTRATION
Another common model for assessing the rewarding properties of drugs is the self-
administration paradigm. As the name suggests, this paradigm consists of recording
the number of times an animal produces a response (e.g., a lever press) that results
in an infusion of drug, which is usually given intravenously. The self-administration
paradigm is an important tool for screening drugs for abuse potential and to eluci-
date the rewarding effects of drugs. Although the conditioned place preference and
self-administration paradigms both measure the rewarding properties of drugs, there
are important differences between these two models (Table 4.2). First, although both
CPP and self-administration studies are sensitive to the rewarding effects of many of
TABLE 4.2
Comparison of Conditioned Place Preference and Self-Administration
CPP Self-Administration
Affective drug properties Rewarding and aversive effects Rewarding effects
Behavioral training Classical conditioning Operant conditioning
No surgery required Requires catheter implantation
Usually 1 wk of training Can conduct tests to substitute for,
Can conduct tests to block effects block, or alter motivation for
of drug training drug
Drug administration Drug injected before session Drug administered after response
by subject
Drug classes Psychostimulants Psychostimulants
Opiates Opiates
Equipment Generally a shuttle box (two or Operant chamber, syringe pump for
three chamber), open field, or drug administration, liquid swivel,
maze, but many other variations and animal harness
used
Experimental design Between groups Within subjects
Species Usually rats and mice Rats, mice, monkeys and
sometimes pigeons

the same drugs, including psychostimulants and opiates, some drugs produce CPP
but may not be self administered (e.g., LSD, buspirone, and pentylenetetrazole),
while others are self administered but do not produce CPP (e.g., pentobarbital and
phencyclidine).4 Second, the preponderance of CPP studies have used only rats and
mice, whereas self-administration studies have been conducted in monkeys, rats,
mice, and pigeons. Third, the mechanisms that mediate drug-induced CPP and self-
administration of a drug may be different. For example, D2 receptor antagonists have
minimal effects on the ability of cocaine to produce CPP, whereas D2 antagonists
readily attenuate self-administration for cocaine.11 Finally, an important contrast
between these two paradigms is the difference in methodological procedures. Unlike
the CPP paradigm, the self-administration paradigm requires surgical implantation
of a catheter, usually for intravenous drug administration, and an extensive operant
training history. Moreover, in CPP, the subjective effects of the drug are present prior
to the task, whereas in the self-administration paradigm, a subject is learning a task
where responses produce near-immediate effects from drug administration. The lat-
ter appears to be most similar of these two models to drug use in humans.
4.8 SUMMARY
The CPP paradigm is a useful tool for studying the affective properties of drugs, and
is routinely used in concert with standard research techniques in neuroscience. Most
drugs of abuse elicit a CPP in rats and mice, and the neural substrates of these effects
can often be traced to the mesolimbic DA system. Alternative models for assess-
ing the rewarding effects of drugs (e.g., self-administration) do not always produce
similar results, and therefore, researchers should be careful when evaluating results
based on the behavioral model they are using in their study.
ACKNOWLEDGMENTS
The authors wish to thank Med Associates, Inc. for providing the images used in
Figure 4.1 and Juan Rodriguez for producing the illustrations used in Figure 4.2.
REFERENCES
1. Lin, S. K., Pan, W. H., and Yeh, P. H. 2007. Prefrontal dopamine efflux during exposure
to drug-associated contextual cues in rats with prior repeated methamphetamine. Brain
Res. Bull. 71:365–371.
2. Calcagnetti, D. J., and Schechter, M. D. 1994. Nicotine place preference using the biased
method of conditioning. Prog. Neuropsychopharmacol. Biol. Psychiatry 18:925–933.
3. Hoffman, D. C. 1989. The use of place conditioning in studying the neuropharmacol-
ogy of drug reinforcement. Brain Res. Bull. 23:373–387.
4. Bardo, M. T., and Bevins, R. A. 2000. Conditioned place preference: What does it
add to our preclinical understanding of drug reward? Psychopharmacology (Berl.)
153:31–43.
5. Fudala, P. J., Teoh, K. W., and Iwamoto, E. T. 1985. Pharmacologic characteriza-
tion of nicotine-induced conditioned place preference. Pharmacol. Biochem. Behav.
22:237–241.

6. Tuazon, D. B., Suzuki, T., Misawa, M., and Watanabe, S. 1992. Methylxanthines (caf-
feine and theophylline) blocked methamphetamine-induced conditioned place prefer-
ence in mice but enhanced that induced by cocaine. Ann. NY Acad. Sci. 654:531–533.
7. Masukawa, Y., Suzuki, T., and Misawa, M. 1993. Differential modification of the
rewarding effects of methamphetamine and cocaine by opioids and antihistamines.
Psychopharmacology (Berl.) 111:139–143.
8. Suzuki, T., and Misawa, M. 1995. Sertindole antagonizes morphine-, cocaine-, and
methamphetamine-induced place preference in the rat. Life Sci. 57:1277–1284.
9. Suzuki, T., Sugano, Y., Funada, M., and Misawa, M. 1995. Adrenalectomy potenti-
ates the morphine—but not cocaine-induced place preference in rats. Life Sci. 56:
PL339–344.
10. Horan, B., Smith, M., Gardner, E. L., Lepore, M., and Ashby, C. R. Jr. 1997. (-)-Nico-
tine produces conditioned place preference in Lewis, but not Fischer 344 rats. Synapse
26:93–94.
11. Bardo, M. T., Valone, J. M., and Bevins, R. A. 1999. Locomotion and conditioned place
preference produced by acute intravenous amphetamine: Role of dopamine receptors
and individual differences in amphetamine self-administration. Psychopharmacology
(Berl.) 143:39–46.
12. Busse, G. D., and Riley, A. L. 2004. Cocaine, but not alcohol, reinstates cocaine-
induced place preferences. Pharmacol. Biochem. Behav. 78:827–833.
13. Philibin, S. D., Vann, R. E., Varvel, S. A., et al. 2005. Differential behavioral responses
to nicotine in Lewis and Fischer-344 rats. Pharmacol. Biochem. Behav. 80:87–92.
14. Ettenberg, A., and Bernardi, R. E. 2007. Effects of buspirone on the immediate positive
and delayed negative properties of intravenous cocaine as measured in the conditioned
place preference test. Pharmacol. Biochem. Behav. 87:171–178.
15. Asin, K. E., and Wirtshafter, D. 1985. Clonidine produces a conditioned place prefer-
ence in rats. Psychopharmacology (Berl.) 85:383–385.
16. Brown, E. E., Finlay, J. M., Wong, J. T., Damsma, G., and Fibiger, H. C. 1991. Behav-
ioral and neurochemical interactions between cocaine and buprenorphine: Implications
for the pharmacotherapy of cocaine abuse. J. Pharmacol. Exp. Ther. 256:119–126.
17. Suzuki, T., Shiozaki, Y., Masukawa, Y., Misawa, M., and Nagase, H. 1992. The role of
mu- and kappa-opioid receptors in cocaine-induced conditioned place preference. Jpn.
J. Pharmacol. 58:435–442.
18. Cervo, L., Rossi, C., and Samanin, R. 1993. Clonidine-induced place preference is
mediated by alpha 2-adrenoceptors outside the locus coeruleus. Eur. J. Pharmacol.
238:201–207.
19. Suzuki, T., Funada, M., Narita, M., Misawa, M., and Nagase, H. 1993. Morphine-
induced place preference in the CXBK mouse: Characteristics of mu opioid receptor
subtypes. Brain Res. 602:45–52.
20. Gaiardi, M., Bartoletti, M., Bacchi, A., Gubellini, C., and Babbini, M. 1997. Motiva-
tional properties of buprenorphine as assessed by place and taste conditioning in rats.
21. Matsuzawa, S., Suzuki, T., Misawa, M., and Nagase, H. 1999. Involvement of dopamine
D(1) and D(2) receptors in the ethanol-associated place preference in rats exposed to
conditioned fear stress. Brain Res. 835:298–305.
22. McFarland, K., and Ettenberg, A. 1999. Haloperidol does not attenuate conditioned
place preferences or locomotor activation produced by food- or heroin-predictive dis-
criminative cues. Pharmacol. Biochem. Behav. 62:631–641.
23. Cheer, J. F., Kendall, D. A., and Marsden, C. A. 2000. Cannabinoid receptors and
reward in the rat: A conditioned place preference study. Psychopharmacology (Berl.)
151:25–30.

24. Matsuzawa, S., Suzuki, T., and Misawa, M. 2000. Ethanol, but not the anxiolytic drugs
buspirone and diazepam, produces a conditioned place preference in rats exposed to
conditioned fear stress. Pharmacol. Biochem. Behav. 65:281–288.
25. Valjent, E., and Maldonado, R. 2000. A behavioural model to reveal place preference
to delta 9-tetrahydrocannabinol in mice. Psychopharmacology (Berl.) 147:436–438.
26. Braida, D., Pozzi, M., Cavallini, R., and Sala, M. 2001. Conditioned place preference
induced by the cannabinoid agonist CP 55,940: Interaction with the opioid system.
Neuroscience 104:923–926.
27. Robinson, L., Hinder, L., Pertwee, R. G., and Riedel, G. 2003. Effects of delta9-THC
and WIN-55,212-2 on place preference in the water maze in rats. Psychopharmacology
(Berl.) 166:40–50.
28. Walker, B. M., and Ettenberg, A. 2003. The effects of alprazolam on conditioned place
preferences produced by intravenous heroin. Pharmacol. Biochem. Behav. 75:75–80.
29. Busse, G. D., Lawrence, E. T., and Riley, A. L. 2005. The effects of alcohol preexpo-
sure on cocaine, alcohol and cocaine/alcohol place conditioning. Pharmacol. Biochem.
Behav. 81:459–465.
30. Simpson, G. R., and Riley, A. L. 2005. Morphine preexposure facilitates morphine
place preference and attenuates morphine taste aversion. Pharmacol. Biochem. Behav.
80:471–479.
31. Marquez, P., Baliram, R., Kieffer, B. L., and Lutfy, K. 2007. The mu opioid receptor is
involved in buprenorphine-induced locomotor stimulation and conditioned place pref-
erence. Neuropharmacology 52:1336–1341.
32. Rezayof, A., Nazari-Serenjeh, F., Zarrindast, M. R., Sepehri, H., and Delphi, L. 2007.
Morphine-induced place preference: Involvement of cholinergic receptors of the ven-
tral tegmental area. Eur. J. Pharmacol. 562:92–102.
33. Le Foll, B., and Goldberg, S. R. 2005. Control of the reinforcing effects of nicotine
by associated environmental stimuli in animals and humans. Trends Pharmacol. Sci.
26:287–293.
34. Best, P. J., Best, M. R., and Mickley, G. A. 1973. Conditioned aversion to distinct envi-
ronmental stimuli resulting from gastrointestinal distress. J. Comp. Physiol. Psychol.
85:250–257.
35. van der Kooy, D., Swerdlow, N. R., and Koob, G. F. 1983. Paradoxical reinforcing prop-
erties of apomorphine: Effects of nucleus accumbens and area postrema lesions. Brain
Res. 259:111–118.
36. Bechara, A., and van der Kooy, D. 1985. Opposite motivational effects of endogenous
opioids in brain and periphery. Nature 314:533–534.
37. Bechara, A., Zito, K. A., and van der Kooy, D. 1987. Peripheral receptors mediate
the aversive conditioning effects of morphine in the rat. Pharmacol. Biochem. Behav.
28:219–225.
38. Papp, M. 1988. Different effects of short- and long-term treatment with imipramine on
the apomorphine- and food-induced place preference conditioning in rats. Pharmacol.
39. Zito, K. A., Bechara, A., Greenwood, C., and van der Kooy, D. 1988. The dopamine
innervation of the visceral cortex mediates the aversive effects of opiates. Pharmacol.
40. Schwartz, A. S., and Marchok, P. L. 1974. Depression of morphine-seeking behaviour
by dopamine inhibition. Nature 248:257–258.
41. Spyraki, C., Fibiger, H. C., and Phillips, A. G. 1982. Dopaminergic substrates of
amphetamine-induced place preference conditioning. Brain Res. 253:185–193.
42. Spyraki, C., Fibiger, H. C., and Phillips, A. G. 1983. Attenuation of heroin reward in
rats by disruption of the mesolimbic dopamine system. Psychopharmacology (Berl.)
79:278–283.

43. Mithani, S., Martin-Iverson, M. T., Phillips, A. G., and Fibiger, H. C. 1986. The effects
of haloperidol on amphetamine- and methylphenidate-induced conditioned place pref-
erences and locomotor activity. Psychopharmacology (Berl.) 90:247–252.
44. Spyraki, C., Nomikos, G. G., and Varonos, D. D. 1987. Intravenous cocaine-induced
place preference: Attenuation by haloperidol. Behav. Brain Res. 26:57–62.
45. Hoffman, D. C., and Beninger, R. J. 1989. The effects of selective dopamine D1 or D2
receptor antagonists on the establishment of agonist-induced place conditioning in rats.
Pharmacol. Biochem. Behav. 33:273–279.
46. Phillips, A. G., and LePiane, F. G. 1980. Reinforcing effects of morphine microinjec-
tion into the ventral tegmental area. Pharmacol. Biochem. Behav. 12:965–968.
47. van der Kooy, D., Mucha, R. F., O’Shaughnessy, M., and Bucenieks, P. 1982. Reinforc-
ing effects of brain microinjections of morphine revealed by conditioned place prefer-
ence. Brain Res. 243:107–117.
48. Phillips, A. G., LePiane, F. G., and Fibiger, H. C. 1983. Dopaminergic mediation of
reward produced by direct injection of enkephalin into the ventral tegmental area of the
rat. Life Sci. 33:2505–2511.
49. Glimcher, P. W., Giovino, A. A., Margolin, D. H., and Hoebel, B. G. 1984. Endogenous
opiate reward induced by an enkephalinase inhibitor, thiorphan, injected into the ven-
tral midbrain. Behav. Neurosci. 98:262–268.
50. Carr, G. D., and White, N. M. 1986. Anatomical disassociation of amphetamine’s
rewarding and aversive effects: An intracranial microinjection study. Psychopharma-
cology (Berl.) 89:340–346.
51. Bozarth, M. A. 1987. Neuroanatomical boundaries of the reward-relevant opiate-recep-
tor field in the ventral tegmental area as mapped by the conditioned place preference
method in rats. Brain Res. 414:77–84.
52. Laviolette, S. R., and van der Kooy, D. 2003. The motivational valence of nicotine in
the rat ventral tegmental area is switched from rewarding to aversive following block-
ade of the alpha7-subunit-containing nicotinic acetylcholine receptor. Psychopharma-
cology (Berl.) 166:306–313.
53. Duvauchelle, C. L., Ikegami, A., Asami, S., Robens, J., Kressin, K., and Castaneda,
E. 2000. Effects of cocaine context on NAcc dopamine and behavioral activity after
repeated intravenous cocaine administration. Brain Res. 862:49–58.
54. Duvauchelle, C. L., Ikegami, A., and Castaneda, E. 2000. Conditioned increases in
behavioral activity and accumbens dopamine levels produced by intravenous cocaine.
Behav. Neurosci. 114:1156–1166.
55. Ventura, R., Cabib, S., Alcaro, A., Orsini, C., and Puglisi-Allegra, S. 2003. Norepi-
nephrine in the prefrontal cortex is critical for amphetamine-induced reward and meso-
accumbens dopamine release. J. Neurosci. 23:1879–1885.
56. Ventura, R., Alcaro, A., and Puglisi-Allegra, S. 2005. Prefrontal cortical norepineph-
rine release is critical for morphine-induced reward, reinstatement and dopamine
release in the nucleus accumbens. Cereb. Cortex 15:1877–1886.
57. Gong, W., Neill, D., and Justice, J. B. Jr. 1997. 6-Hydroxydopamine lesion of ventral
pallidum blocks acquisition of place preference conditioning to cocaine. Brain Res.
754:103–112.
58. Corrigall, W. A., and Linseman, M. A. 1988. Conditioned place preference produced by
intra-hippocampal morphine. Pharmacol. Biochem. Behav. 30:787–789.
59. Lenard, N. R., Daniels, D. J., Portoghese, P. S., and Roerig, S. C. 2007. Absence of
conditioned place preference or reinstatement with bivalent ligands containing mu-
opioid receptor agonist and delta-opioid receptor antagonist pharmacophores. Eur. J.

60. Singh, M. E., Verty, A. N., McGregor, I. S., and Mallet, P. E. 2004. A cannabinoid
receptor antagonist attenuates conditioned place preference but not behavioural sensi-
tization to morphine. Brain Res. 1026:244–253.
61. Rezayof, A., Golhasani-Keshtan, F., Haeri-Rohani, A., and Zarrindast, M. R. 2007.
Morphine-induced place preference: Involvement of the central amygdala NMDA
receptors. Brain Res. 1133:34–41.
62. Swerdlow, N. R., Swanson, L. W., and Koob, G. F. 1984. Electrolytic lesions of the sub-
stantia innominata and lateral preoptic area attenuate the “supersensitive” locomotor
response to apomorphine resulting from denervation of the nucleus accumbens. Brain
Res. 306:141–148.
63. Khroyan, T. V., Fuchs, R. A., Beck, A. M., Groff, R. S., and Neisewander, J. L. 1999.
Behavioral interactions produced by co-administration of 7-OH-DPAT with cocaine or
apomorphine in the rat. Psychopharmacology (Berl.) 142:383–392.
64. Spyraki, C., Fibiger, H. C., and Phillips, A. G. 1982. Cocaine-induced place preference
conditioning: Lack of effects of neuroleptics and 6-hydroxydopamine lesions. Brain
Res. 253:195–203.
65. Nazarian, A., Russo, S. J., Festa, E. D., Kraish, M., and Quinones-Jenab, V. 2004. The
role of D1 and D2 receptors in the cocaine conditioned place preference of male and
female rats. Brain Res. Bull. 63:295–299.
66. Sora, I., Wichems, C., Takahashi, N., et al. 1998. Cocaine reward models: Conditioned
place preference can be established in dopamine- and in serotonin-transporter knock-
out mice. Proc. Natl. Acad. Sci. USA 95:7699.–7704.
67. Medvedev, I. O., Gainetdinov, R. R., Sotnikova, T. D., Bohn, L. M., Caron, M. G., and
Dykstra, L. A. 2005. Characterization of conditioned place preference to cocaine in
congenic dopamine transporter knockout female mice. Psychopharmacology (Berl.)
180:408–413.
68. Chen, R., Tilley, M. R., Wei, H., et al. 2006. Abolished cocaine reward in mice with a
cocaine-insensitive dopamine transporter. Proc. Natl. Acad. Sci. USA 103:9333–9338.
69. Kruszewska, A., Romandini, S., and Samanin, R. 1986. Different effects of zimelidine
on the reinforcing properties of d-amphetamine and morphine on conditioned place
preference in rats. Eur. J. Pharmacol. 125:283–286.
70. Nomikos, G. G., and Spyraki, C. 1988. Effects of ritanserin on the rewarding properties
of d-amphetamine, morphine and diazepam revealed by conditioned place preference
in rats. Pharmacol. Biochem. Behav. 30:853–858.
71. Takamatsu, Y., Yamamoto, H., Ogai, Y., Hagino, Y., Markou, A., and Ikeda, K. 2006.
Fluoxetine as a potential pharmacotherapy for methamphetamine dependence: Studies
in mice. Ann. NY Acad. Sci. 1074:295–302.
72. Douglass, J., McKinzie, A. A., and Couceyro, P. 1995. PCR differential display identi-
fies a rat brain mRNA that is transcriptionally regulated by cocaine and amphetamine.
J. Neurosci. 15:2471–2481.
73. Koylu, E. O., Couceyro, P. R., Lambert, P. D., and Kuhar, M. J. 1998. Cocaine- and
amphetamine-regulated transcript peptide immunohistochemical localization in the rat
brain. J. Comp. Neurol. 391:115–132.
74. Kimmel, H. L., Gong, W., Vechia, S. D., Hunter, R. G., and Kuhar, M. J. 2000. Intra-
ventral tegmental area injection of rat cocaine and amphetamine-regulated transcript
peptide 55-102 induces locomotor activity and promotes conditioned place preference.
J. Pharmacol. Exp. Ther. 294:784–792.
75. Fudala, P. J., and Iwamoto, E. T. 1986. Further studies on nicotine-induced conditioned
place preference in the rat. Pharmacol. Biochem. Behav. 25:1041–1049.
76. Le Foll, B., and Goldberg, S. R. 2005. Nicotine induces conditioned place preferences
over a large range of doses in rats. Psychopharmacology (Berl.) 178:481–492.

77. Suzuki, T., Ise, Y., Mori, T., and Misawa, M. 1997. Attenuation of mecamylamine-pre-
cipitated nicotine-withdrawal aversion by the 5-HT3 receptor antagonist ondansetron.
Life Sci. 61:PL249–254.
78. Janhunen, S., Linnervuo, A., Svensk, M., and Ahtee, L. 2005. Effects of nicotine and
epibatidine on locomotor activity and conditioned place preference in rats. Pharmacol.
79. Walters, C. L., Brown, S., Changeux, J. P., Martin, B., and Damaj, M. I. 2006. The
beta2 but not alpha7 subunit of the nicotinic acetylcholine receptor is required for nico-
tine-conditioned place preference in mice. Psychopharmacology (Berl.) 184:339–344.
80. Schechter, M. D., Meehan, S. M., and Schechter, J. B. 1995. Genetic selection for nico-
tine activity in mice correlates with conditioned place preference. Eur. J. Pharmacol.
279:59–64.
81. Quertemont, E., and De Witte, P. 2001. Conditioned stimulus preference after acetalde-
hyde but not ethanol injections. Pharmacol. Biochem. Behav. 68:449–454.
82. Boyce-Rustay, J. M., and Cunningham, C. L. 2004. The role of NMDA receptor bind-
ing sites in ethanol place conditioning. Behav. Neurosci. 118:822–834.
83. Walker, B. M., and Ettenberg, A. 2007. Intracerebroventricular ethanol-induced con-
ditioned place preferences are prevented by fluphenazine infusions into the nucleus
accumbens of rats. Behav. Neurosci. 121:401–410.
84. Busse, G. D., and Riley, A. L. 2002. Modulation of cocaine-induced place preferences
by alcohol. Prog. Neuropsychopharmacol. Biol. Psychiatry 26:1373–1381.
85. Busse, G. D., Lawrence, E. T., and Riley, A. L. 2004. The modulation of cocaine-
induced conditioned place preferences by alcohol: effects of cocaine dose. Prog. Neu-
ropsychopharmacol. Biol. Psychiatry 28:149–155.
86. Brown, Z. W., Amit, Z., and Rockman, G. E. 1979. Intraventricular self-administration
of acetaldehyde, but not ethanol, in naive laboratory rats. Psychopharmacology (Berl.)
64:271–276.
87. Smith, B. R., Amit, Z., and Splawinsky, J. 1984. Conditioned place preference induced
by intraventricular infusions of acetaldehyde. Alcohol 1:193–195.
88. Rodd, Z. A., Bell, R. L., Zhang, Y., et al. 2005. Regional heterogeneity for the intracra-
nial self-administration of ethanol and acetaldehyde within the ventral tegmental area
of alcohol-preferring (P) rats: Involvement of dopamine and serotonin. Neuropsycho-
pharmacology 30:330–338.
89. Font, L., Aragon, C. M., and Miquel, M. 2006. Ethanol-induced conditioned place pref-
erence, but not aversion, is blocked by treatment with D-penicillamine, an inactivation
agent for acetaldehyde. Psychopharmacology (Berl.) 184:56–64
90. Bilsky, E. J., Hui, Y. Z., Hubbell, C. L., and Reid, L. D. 1990. Methylenedioxymetham-
phetamine’s capacity to establish place preferences and modify intake of an alcoholic
beverage. Pharmacol. Biochem. Behav. 37:633–638.
91. Bilsky, E. J., Hubbell, C. L., Delconte, J. D., and Reid, L. D. 1991. MDMA produces a
conditioned place preference and elicits ejaculation in male rats: A modulatory role for
the endogenous opioids. Pharmacol. Biochem. Behav. 40:443–447.
92. Daza-Losada, M., Ribeiro Do Couto, B., Manzanedo, C., Aguilar, M. A., Rodriguez-
Arias, M., and Minarro, J. 2007. Rewarding effects and reinstatement of MDMA-
induced CPP in adolescent mice. Neuropsychopharmacology 32:1750–1759.
93. Marona-Lewicka, D., Rhee, G. S., Sprague, J. E., and Nichols, D. E. 1996. Reinforcing
effects of certain serotonin-releasing amphetamine derivatives. Pharmacol. Biochem.
Behav. 53:99–105.
94. Bilsky, E. J., and Reid, L. D. 1991. MDL72222, a serotonin 5-HT3 receptor antagonist,
blocks MDMA’s ability to establish a conditioned place preference. Pharmacol. Bio-
chem. Behav. 39:509–512.

95. Braida, D., Iosue, S., Pegorini, S., and Sala, M. 2005. 3,4 Methylenedioxymethamphet-
amine-induced conditioned place preference (CPP) is mediated by endocannabinoid
system. Pharmacol. Res. 51:177–182.
96. Fone, K. C., Beckett, S. R., Topham, I. A., Swettenham, J., Ball, M., and Maddocks, L.
2002. Long-term changes in social interaction and reward following repeated MDMA
administration to adolescent rats without accompanying serotonergic neurotoxicity.
97. Aberg, M., Wade, D., Wall, E., and Izenwasser, S. 2007. Effect of MDMA (ecstasy) on
activity and cocaine conditioned place preference in adult and adolescent rats. Neuro-
toxicol. Teratol. 29:37–46.
98. Robledo, P., Trigo, J. M., Panayi, F., de la Torre, R., and Maldonado, R. 2007. Behav-
ioural and neurochemical effects of combined MDMA and THC administration in
mice. Psychopharmacology (Berl.). 195:255–264.
99. Lepore, M., Vorel, S. R., Lowinson, J., and Gardner, E. L. 1995. Conditioned place pref-
erence induced by delta 9-tetrahydrocannabinol: Comparison with cocaine, morphine,
and food reward. Life Sci. 56:2073–2080.
100. Sanudo-Pena, M. C., Tsou, K., Delay, E. R., Hohman, A. G., Force, M., and Walker, J.
M. 1997. Endogenous cannabinoids as an aversive or counter-rewarding system in the
rat. Neurosci. Lett. 223:125–128.
101. Hutcheson, D. M., Tzavara, E. T., Smadja, C., et al. 1998. Behavioural and biochemical
evidence for signs of abstinence in mice chronically treated with delta-9-tetrahydro-
cannabinol. Br. J. Pharmacol. 125:1567–1577.
102. Castane, A., Robledo, P., Matifas, A., Kieffer, B. L., and Maldonado, R. 2003. Cannabi-
noid withdrawal syndrome is reduced in double mu and delta opioid receptor knockout
mice. Eur. J. Neurosci. 17:155–159.
103. Braida, D., Iosue, S., Pegorini, S., and Sala, M. 2004. Delta9-tetrahydrocannabinol-
induced conditioned place preference and intracerebroventricular self-administration
in rats. Eur. J. Pharmacol. 506:63–69.
104. Jardinaud, F., Roques, B. P., and Noble, F. 2006. Tolerance to the reinforcing effects of
morphine in delta9-tetrahydrocannabinol treated mice. Behav. Brain Res. 173:255.
105. Mallet, P. E., and Beninger, R.J. 1998. Delta9-tetrahydrocannabinol, but not the endog-
enous cannabinoid receptor ligand anandamide, produces conditioned place avoidance.
Life Sci. 62:2431–2439.
106. Bortolato, M., Campolongo, P., Mangieri, R. A., Scattoni, M. L., Frau, R., Trezza, V.,
La Rana, G., Russo, R., Calignano, A., Gessa, G. L., Cuomo, V., Piomelli, D. 2006.
Anxiolytic-like properties of the anandamide transport inhibitor AM404. Neuropsy-
chopharmacology 31:2652–2659.
107. Bardo et al. 1999.

5 Anxiety-Related
Behaviors in Mice
Kathleen R. Bailey and Jacqueline N. Crawley
CONTENTS
5.1 Introduction................................................................................................... 78
5.2 General Methodological Considerations....................................................... 79
5.3 Paradigms...................................................................................................... 79
5.3.1 Open Field Exploration Test .............................................................. 79
5.3.1.1 Equipment ............................................................................80
5.3.1.2 Procedure .............................................................................80
5.3.1.3 Analysis and Interpretation .................................................. 81
5.3.1.4 Sample Results ..................................................................... 81
5.3.2 Elevated Plus-Maze/Elevated Zero-Maze ......................................... 81
5.3.2.1 Subjects ................................................................................ 83
5.3.2.2 Equipment ............................................................................ 83
5.3.2.3 Procedure .............................................................................84
5.3.2.4 Analysis and Interpretation ..................................................84
5.3.2.5 Sample Results ..................................................................... 85
5.3.3 Light n Dark Exploration Test ......................................................... 85
5.3.3.1 Subjects ................................................................................ 87
5.3.3.2 Equipment ............................................................................ 87
5.3.3.3 Procedure ............................................................................. 87
5.3.3.5 Sample Results ..................................................................... 88
5.3.4 The Social Interaction Test ................................................................ 88
5.3.4.1 Subjects ................................................................................90
5.3.4.2 Equipment ............................................................................90
5.3.4.3 Procedure .............................................................................90
5.3.4.5 Sample Results ..................................................................... 91
5.3.5 Novelty-Induced Hypophagia ............................................................ 91
5.3.5.1 Subjects ................................................................................92
5.3.5.2 Equipment ............................................................................ 93
5.3.5.3 Procedure ............................................................................. 93
5.3.5.5 Sample Results ..................................................................... 93
5.4 Conclusion..................................................................................................... 95
References................................................................................................................ 95
77

5.1 INTRODUCTION
Human anxiety disorders are broadly grouped according to symptomology and
responsiveness to pharmacological and psychological treatment.1,2 Generalized anx-
iety disorder and panic disorder are the two primary classifications of pathological
anxiety in humans. The distinguishing feature of generalized anxiety disorder is
a pervading sense of unrealistic worry about everyday life situations. In contrast,
panic attacks constitute the primary symptom of panic disorder. These events are
characterized as sudden, extreme fear accompanied by autonomic nervous system
arousal.3
Similar changes in physiological indicators and behavioral responses to fear and
painful stimuli in humans and other animals suggest the possibility of homologous
or analogous, ethologically motivated defensive responses4–10 In the description of
human anxiety disorders, the concepts of “state” and “trait” anxiety have a long
history. However, it is only recently that these concepts have been suggested as a
means of differentiating situational anxiety-like behavior in rodents from anxiety
that transcends the situation and is an enduring condition in the animal.11 The former
is the focus of the rodent behavioral tests reviewed in this chapter. Procedures are
designed to trigger ethologically relevant conflict or conditioned behaviors. The lat-
ter is most often associated with selective breeding, e.g., the high versus low anxiety-
related traits in the high anxiety-related behavior (HAB) versus low anxiety-related
behavior (LAB) rats,12 inbred mouse strains such as BALB/c, and mice with relevant
targeted gene mutations.13,14
In an attempt to model human pathological anxiety in rodents, a wide range of
behavioral testing paradigms have been developed.8,15–19 Many of these tests induce a
fearful response through an aversive event or anticipated aversive event. Others inte-
grate an approach–avoidance conflict designed to inhibit an ongoing behavior that
is characteristic for the animal, such as contrasting the tendency of mice to engage
in exploratory activity or social investigation against the aversive properties of an
open, brightly lit, or elevated space. The premise that basic physiological mecha-
nisms underlying fear in rodents can be equated to similar mechanisms operating in
humans provides a degree of face validity for these paradigms.7,9,10 In rodents, these
responses are deemed appropriate and adaptive for the current conditions, whereas
in humans, anxiety disorders constitute maladaptive or pathological responses to the
existing situation. Further exploration of rodent neuroanatomy and neurochemistry
involved in fear extinction and inhibition of conditioned fear could offer important
insights into effective targets for novel pharmacological treatment of pathological
human anxiety.20
Although rats have been the rodent of choice for much of the preclinical research
on anxiety-like behavior, recent technical advances in molecular genetics have placed
the mouse in the forefront of neuropsychiatric research.7,13,21–26 This has resulted in
the adaptation of many well-validated behavioral tests of anxiety from rats to mice,
with varying degrees of success. This chapter offers a sampling of well-established
tests of anxiety-like behavior in mice that use an ethological conflict. The interested
researcher is directed to the classic source literature for more information on other

Anxiety-Related Behaviors in Mice 79
types of anxiety-based tests, including conditioning paradigms,27–29 punishment-

induced conflict tests,18,19,30,31 developmental models,32,33 and aversive tests.5,34
5.2 GENERAL METHODOLOGICAL CONSIDERATIONS

Several excellent papers have identified important factors that careful researchers
will want to consider when designing experiments to assess anxiety-like behaviors
in mice.35–37 The behavioral paradigms described in this chapter are suitable for most
inbred mouse strains. In addition, mice with targeted genetic mutations that do not
alter exploratory drive, motor ability, or recognition memory are also suitable test
subjects for these paradigms. Other factors to consider when designing experiments
assessing anxiety-like behavior include, but are not limited to, the experimental his-
tory of test subjects, prior test exposure, differences in exploratory motivation, and
whether the test is to be conducted as part of a test battery or administered as a single
behavioral assessment.9,38–42 Mice are social animals and are typically group-housed
(four to five per cage) in same-sex home cages. Special circumstances (e.g., aggres-
sive strains or mice with head mounts or other surgical interventions) may require
single-housing prior to testing. Environmental conditions in the animal housing
rooms should be as quiet as possible and consistent across experiments to control for
extraneous variables that can significantly alter physiological and behavioral indica-
tors of stress.37,43 Food and water are typically ad libitum unless the experimental
design requires restriction. Avoid behavioral testing on days when, as a part of nor-
mal animal husbandry, home cages are scheduled for changing. Cage change typi-
cally causes an increase in general activity and stress levels.44–46 Researchers should
fully describe any special circumstances pertaining to housing and care of test sub-
jects in their experimental methods. To ensure consistency of experience prior to the
test session, subjects are brought to the testing room or a common staging area, in
their home cages, at least 1 hr prior to the start of behavioral testing. Individual mice
can then be transported singly, in clean cages, into the testing apparatus. Test room
lighting, temperature, and noise levels should be consistent for all subjects. Behav-
ioral testing equipment described in this chapter is usually cleaned thoroughly with
a solution of mild soapy water at the end of a test session. Prior to running the first
subject on a day of testing, the experimenter wipes the behavioral equipment with
70% ethanol. After each subject completes its test session, fecal boli and urine are
removed, surfaces are wiped with 70% ethanol, and the test chamber is allowed to
dry completely before starting another subject. Most studies use a minimum of 12–15
animals per experimental group to insure sufficient power for statistical analysis.22
5.3 PARADIGMS
5.3.1 OPEN FIELD EXPLORATION TEST
Originally introduced as a measure of emotional behavior in rats,8 open field explo-
ration has proven to be equally successful with mice.47 The test provides a unique
opportunity to systematically assess novel environment exploration, general locomo-
tor activity, and provide an initial screen for anxiety-related behavior in rodents.48

In addition, repeated exposure or extended session length provides a method for

assessing habituation to the increasingly familiar chamber environment. It has been
suggested that two factors influence anxiety-like behavior in the open field. The
first is social isolation resulting from the physical separation from cage mates when
performing the test. The second is the stress created by the brightly lit, unprotected,
novel test environment.17,48
5.3.1.1 Equipment
Although several different shapes have been used as rodent open field arenas,49,50
the most common design for mice is a large square chamber ranging in size from 28
× 28 cm to 56 × 56 cm. Chamber walls and floor can be plastic or wood but many
automated systems use transparent Plexiglas. The open field arena is divided into
a grid of equally sized areas by infrared photocell beams or lines drawn on the
chamber floor for visual scoring of activity by the experimenter. Automated systems
such as the VersaMax Animal Activity Monitoring System with Analyzer software
(AccuScan Instruments, Inc., Columbus, Ohio, USA), SmartFrame open field sys-
tem with Motor Monitor control and software (Lafayette Instruments, Lafayette,
Indiana, USA), Open Field Activity System MED–OFA-MS (Med Associates, Inc.,
St. Albans, Vermont, USA), and Photobeam activity system–open field (San Diego
Instruments, San Diego, California, USA), record each beam break as one unit of
exploratory activity, similar to manual scoring of each line crossed.
5.3.1.2 Procedure
Transport acclimated mice to the test room singly, if only one test chamber is avail-
able, or as a group in the home cage, if several automated chambers are available for
testing. Place each mouse in the center of a chamber. If the experimenter intends to
remain in the testing room, care should be taken to be as distant and unmoving as
possible once the test session has started. Sudden motion or noise can greatly affect
exploratory activity. Mice are allowed to freely explore the chamber for the dura-
tion of the test session. Each line crossed or photocell beam break is scored as one
unit of activity. For assessing novel environment exploration, a 5-min test length is
typical. If the researcher is interested in examining habituation to an increasingly
familiar environment, a 30-min test session is recommended. Mice are allowed to
freely explore the test arena for the entire session duration. Upon completion of the
test, return the mouse to the home cage. In addition to horizontal units of activ-
ity, rearing behavior, defecation, and grooming activity can also be scored. These
parameters provide measures of general physical motor abilities and level of interest
in the novelty of the environment.
Rodents will typically spend a significantly greater amount of time exploring
the periphery of the arena, usually in contact with the walls (thigmotaxis), than
the unprotected center area. Mice that spend significantly more time exploring the
unprotected center area demonstrate anxiolytic-like baseline behavior. The center
area of the chamber can be defined by the experimenter as a proportion of the overall
test arena size. Many software systems allow the researcher to designate this cen-
ter area, as well as multiple other regions of the test chamber, to track exploratory

activity. When the open field arena size is 40 × 40 cm2, the center region size is often
designated as 20 × 20 cm2.51,52
5.3.1.3 Analysis and Interpretation

Open field exploration results are generally analyzed using repeated-measures anal-
ysis of variance (ANOVA) for longer sessions when the researcher is interested in
comparing levels of exploratory activity over the duration of the session. The session
can be divided into time bins (e.g., 5 min) and changes in exploratory behavior can
be compared across the length of the session. By contrast, if the experimenter’s only
interest is novel environment exploration, then a one-way ANOVA can be run using
the total scores across the test session for each behavioral measure (e.g., vertical
activity, horizontal activity, total distance, and center time). A 5-min test session is
often sufficient to capture the critical components of general exploratory locomotion.
The most commonly used measure of overall exploratory/locomotor activity is cur-
rently the total distance traveled. Although horizontal activity appears to be record-
ing a similar measure, in fact, the equipment records every beam break including
those not associated with ambulatory activity (e.g., repetitive head movements). In
contrast, the calculation of total distance includes constraints that exclude units of
activity that are generated by these repetitive beam breaks. Time spent investigat-
ing the central region of the chamber can be reported as a percent of total session
length for both the short and longer habituation versions of this test. Alternatively,
center time can be examined in 5-min bins over the duration of the 30-min session
to examine changing patterns of anxiety-related behavior.
5.3.1.4 Sample Results

Figure 5.1A–C illustrates open field activity for galanin receptor subtype 2 (GalR2)
null mutant mice.53 Behavioral measures reported include total distance traveled,
horizontal beam breaks, and time spent exploring the center area of the chamber.
There were no effects of genotype on horizontal activity or total distance traveled
(all p comparisons > 0.05). Males were significantly more active than females on
horizontal activity and total distance traveled in the arena (p = 0.0009 and p =
0.0042, respectively). Galanin null mutant mice spent less time exploring the central
area of the chamber compared to wild type (WT) littermates, but this did not reach
significance (p = 0.0714).
5.3.2 ELEVATED PLUS-MAZE/ELEVATED ZERO-MAZE

This well-established paradigm has a long and successful history in assessing anxi-
ety-like behavior in mice.22,42,54–56 The test takes advantage of the natural tendency of
mice to explore novel environments. The mouse is given the choice of spending time
in open, unprotected maze arms or enclosed, protected arms, all elevated approx-
imately 1 m from the floor. Mice tend to avoid the open areas, especially when
they are brightly lit, favoring darker, more enclosed spaces. This approach–avoid-
ance conflict results in behaviors that have been correlated with increases in physi-
ological stress indicators.52 In contrast, administration of benzodiazepines and other

800 M +/+ 450
Horizontal Activity-Cumulative Beam Breaks

M +/–
Total Distance-Cumulative Beam Breaks
700 M –/– 400

F +/+
F +/– 350
600
F –/–
300
500
250
400
200
300
150
200 100
100 50
1-5 6-10 11-15 16-20 21-25 26-30 1-5 6-10 11-15 16-20 21-25 26-30
Minutes Minutes
(a) (b)
12
10
8
% Center Time
0
+/+ +/– –/–
(c)
FIGURE 5.1 Effect of GalR2 mutation on open field exploration. There were no significant
genotype differences on horizontal activity or total distance traveled (p > 0.05). Males were
significantly more active than females on horizontal activity and distance traveled (all p com-
parisons < 0.01). Examination of open field center time as a preliminary screen for anxiety-
like behavior revealed no sex differences, thus data from males and females were combined.
GalR2 -/- spent less time exploring the center of the open field than their wild type littermates
(p = 0.0714), although this trend did not reach significance. N = 22 +/+, 23 +/-, 17 -/-. Data are
shown as mean + standard error of the mean. Source: Reprinted from Bailey, Pavlova, Rohde,
Hohmann, and Crawley. 2007. Galanin receptor subtype 2 (GalR2) null mutant mice display
an anxiogenic-like phenotype specific to the elevated plus-maze. Pharmacol. Biochem. and
Behav. 86:13, with permission from Elsevier.

anxiolytic treatments results in increased exploration of the open arms, without

affecting general motivation or locomotion.42,55,57,58
5.3.2.1 Subjects
The primary requirements for subjects performing this test are normal ambulatory
ability and average levels of exploratory drive. Mice that spend prolonged time in the
center start area, enter only partially into one arm of the maze without transition-
ing through, or do not explore the entire maze, may confound the interpretation of
behavioral data for a group. In these cases, data may primarily reflect physical motor
abilities that are minimally relevant to anxiogenic or anxiolytic traits. It is important
to note this type of behavior during the test session, as it may be necessary for later
identification of outliers. Strains that consistently demonstrate very low levels of
exploratory behavior (e.g., AJ, some 129 substrains) should be avoided.
5.3.2.2 Equipment
Conceptually the equipment design has remained virtually unchanged for mice since
its introduction.59–61 However, there have been substantial alterations and modifica-
tions in the materials and specific details of the maze construction. The apparatus
consists of two sets of opposing arms approximately 30 × 5 cm extending from
a central (5 × 5 cm) region. Two arms are enclosed with 15-cm high walls. The
remaining two arms are open. Differences in maze construction include wood con-
struction versus Perspex or other similarly smooth material. Some researchers have
provided a slightly raised lip (0.25 cm) on three sides of the open arms to minimize
falls. Walls of the enclosed arms may be transparent, opaque, or dark. While a con-
sensus has not been reached about the advantages or limitations of wall transpar-
ency, researchers may want to consider the impact of these different materials on
light levels within the arm.62 Ideally, minimizing variability in external factors (e.g.,
light level differences) will increase replication across labs and simplify interpreta-
tion of behavioral findings.
The elevated zero-maze offers a conceptually identical behavioral test that elimi-
nates the ambiguous center start area of the elevated plus-maze (EPM).63 In the plus-
maze, test subjects will often remain in the central start area, or return to it regularly,
thereby spending considerable amounts of time in a region of the maze that is consid-
ered ambiguous in the evaluation of anxiety-related behavior. The elevated circular
runway alternates equally sized, open, brightly lit areas and enclosed, dark arc areas.
The uninterrupted nature of the open versus closed segments of the circular runway
mitigates the concerns surrounding the central start area of the plus-maze. Similar
behavioral measures are scored for this version of the test during the 5-min session.
Scoring from a videotaped session minimizes environmental variables introduced
by the presence of the investigator that may impact anxiety-related behaviors.
Technological advances have been introduced as a means of standardizing the
EPM paradigm, including automated tracking and scoring software (e.g., Noldus
Ethovision video tracking, Hamilton-Kinder infrared photobeam tracking). Con-
cerns have been raised10,42 about the sensitivity of automated systems for detect-
ing measures of ethologically relevant risk assessment behaviors and the utility of

scoring many additional behavioral indices as factors to explain anxiety-related behav-

ior of animals.64,65 Inconsistent results with anxiolytic compounds and a desire for
more targeted therapeutic treatments suggests that scoring additional, ethologically
relevant behavioral indicators (e.g., head dipping, stretch-attend postures) may pro-
vide more sensitive measures of the effects of new anxiolytic compounds.10,65 It
remains to be determined whether current tracking and scoring software can accu-
rately and consistently detect these additional behavioral indices in the wide variety
of inbred strains and transgenic and knockout (KO) lines currently being studied.
5.3.2.3 Procedure
Subjects are generally group-housed (four to five per cage) in same-sex home
cages. Home cages are brought to the testing room or a common staging area 1 hr
prior to testing. Transport mice singly in clean cages to the apparatus or testing
room. Room level lighting should be consistent for all subjects. Mice generally
avoid brightly lit areas, therefore high illumination levels would be expected to
increase anxiety-like behaviors. Care is taken to avoid light levels that are high
enough to restrict the natural exploratory tendency of mice. Pilot studies will assist
in determining the most appropriate illumination level from those reported in the
literature.67–70 Each subject is placed in the central area of the maze with open
access to any arm. Mice are allowed to freely explore the maze for 5 min. The
number of arm entries and the amount of time spent in the open and closed arms
are recorded. These can be recorded manually by a highly trained observer, or
by an automated photo beam sensor recording system. The session can also be
recorded using any one of the currently available video tracking systems for sub-
sequent scoring. There are advantages and limitations to each of these methods.
The obvious advantage to scoring from a recorded test session is the ability to
minimize errors and recheck the reliability of the scoring at a later time. Simi-
larly, photo beam recording systems remove the subjective interpretations by the
experimenter. For researchers with limited resources, however, these systems may
be cost prohibitive. Two advantages of manual scoring by highly trained observers
are lower equipment costs and identifying unusual or ethologically relevant behav-
iors that might go undetected by automated systems.

EPM results are generally analyzed using between-subjects ANOVAs followed by
Newman-Keuls post-hoc comparisons when a significant ANOVA value is obtained.
There are several factors that should be considered when interpreting EPM results.
Strains or treatment groups that show unusually high or low time spent in the open
arms may do so for reasons other than anxiety-related behavior. For instance, exten-
sive time spent in the open arms may reflect a group that displays very low levels
of exploratory activity. The arm they first transition into, closed or open, is where
they remain. In addition, specific pharmacological treatments, background strain
differences, genetic mutations, or environmental factors can impact locomotor activ-
ity, exploratory behavior, or behavioral motivation for novelty.25,38,39,44,71,72 Finally,

behavior in the EPM is influenced by prior handling, exposure to previous behavioral

testing paradigms, or repeated experience in the plus-maze.10,54,56,73–77
Repeated testing was thought to have no significant impact on measures of anxi-
ety behavior.59,61 However, recent studies suggest that prior test experience increases
open arm avoidance behavior and alters the effectiveness of anxiolytic drugs in
reducing open arm avoidance.57,75,77–84 The change in pharmacological efficacy has
been termed “one-trial tolerance”78 and has been alternately explained as reflecting
a change in the state of the benzodiazepine receptor or the gamma-aminobutyric
acid (GABA)A receptor complex,58,83 a change in the anxiety state manifested in trial
1 compared to trial 2,84 or a change in the underlying mechanism triggering the
behavior from one of unconditioned fear avoidance to learned avoidance based on
prior exposure to the situation.85 These concerns can best be addressed by ensuring
that test subjects are experimentally naïve and receive minimal handling. For situa-
tions in which retest in the EPM is necessary, Adamec and Shallow86 have developed
a test–retest protocol that appears to prevent the increase in open arm avoidance
generally exhibited in trial 2. They suggest a 3-wk interval between test sessions and
moving the maze to a novel test room for the second session.

Figure 5.2A–D provides an example of the behavioral measures most commonly
reported in the literature.60,61,87 In this experiment mice with a null mutation in the
galanin receptor subtype GalR2 were tested in several complementary approach–
avoidance paradigms designed to assess anxiety-like traits. Previous research has
implicated the neuropeptide galanin in rodent emotionality.52,88–91 The results from
two independent cohorts of mice missing the galanin subtype-2 receptor indicate an
anxiogenic phenotype in the EPM. Both cohorts spent significantly less time explor-
ing the open arms (Figure 5.2A) and made fewer open arm entries (Figure 5.2B).
Importantly, the number of overall arm transitions (Figure 5.2D) did not significantly
differ compared with WT littermates.53 Although not shown in the figure, some labs
are now reporting additional ethologically relevant behaviors in their experimental
results, including head dips and stretch-attend postures. These behaviors have been
described as a means for the animal to actively assess dangers within the specific
testing environment and are characterized as risk assessment behaviors.6,42
5.3.3 LIGHT n DARK EXPLORATION TEST

The light n dark exploration test, developed by Crawley and Goodwin,16 was a pre-
cursor to the EPM and provides another means of examining anxiety-like behavior
in rodents. As with the EPM, the subject is exposed to a novel environment with pro-
tected (dark compartment) and unprotected (light compartment) areas. The inherent
conflict between exploratory drive and risk avoidance is thought to inhibit explo-
ration.16,92,93 Most mice naturally demonstrate a preference for the dark, protected
compartment. The key measure for assessing anxiety-related behavior in this design
is a change in willingness to explore the illuminated, unprotected area, reflected
in increases or decreases in the number of transitions between the compartments,
and in time spent in each compartment, during a 10-min test session. Treatment

60 70
50 60
% Time Spent in Open Arms
50
40 *
% Open Entries
40
30
* 30
20
* 20
10 10
0 0
+/+ +/– –/– +/+ +/– –/– +/+ +/– –/– +/+ +/– –/–
Cohort 1 Cohort 2 Cohort 1 Cohort 2
(a) (b)
12 20
10
* 16
8
# Closed Entries
# Total Entries
12
6
8
4
4
2
0 0
+/+ +/– –/– +/+ +/– –/– +/+ +/– –/– +/+ +/––/–
Cohort 1 Cohort 2 Cohort 1 Cohort 2
(c) (d)
FIGURE 5.2 Anxiogenic-like phenotype of GalR2 knockout mice on the elevated plus-
maze. Two independent cohorts of GalR2 -/- displayed an anxiogenic-like phenotype com-
pared to their +/+ littermates in the elevated plus-maze. GalR2 -/- mice spent significantly (*p
< .05) less time in the open arms (A) and made fewer entries into the open arms (B) than +/+
mice. The -/- mice in experiment 1 made significantly (*) more entries into the closed arms
(C), while total arm entries were similar across genotypes (D), suggesting that less explora-
tion of the open arms did not reflect lower overall exploratory behavior. Cohort 1 N = 22 +/+,
23 +/-, 19 -/-; Cohort 2 N = 14 +/+, 12 +/-, 17 -/-. Data are shown as mean + standard error
of the mean. Source: Reprinted from Bailey, Pavlova, Rohde, Hohmann, and Crawley. 2007.
Galanin receptor subtype 2 (GalR2) null mutant mice display an anxiogenic-like phenotype
specific to the elevated plus-maze. Pharmacol. Biochem. and Behav. 86:13, with permission
from Elsevier.
with anxiolytic drugs increased the number of transitions between the two compart-
ments, without altering the preference of the mice to spend more time in the dark
compartment.16,93 This increase in exploratory activity is interpreted as a release of
exploratory inhibition.16

5.3.3.1 Subjects
Similar to the EPM, careful consideration should be given to testing specific inbred
strains of mice and mice with genetic manipulations that inhibit locomotor activity
or interfere with novelty-seeking behavior.
5.3.3.2 Equipment
The chamber is constructed from a standard polypropylene rat cage (44 × 21 × 21
cm) divided into two unequal compartments by a dark partition with a small aperture
(13 × 5 cm) located in the bottom center. The smaller compartment (14 cm) is painted
black and covered by a hinged lid. The larger compartment (28 cm) is uncovered
with transparent sides and is brightly lit from above by fluorescent room lighting.
Transitions between the compartments are electronically recorded by four sets of
photocells mounted in the partition opening. Entry into the dark compartment trig-
gers a timer that records the duration of time spent in the dark compartment.
5.3.3.3 Procedure
Transport acclimated mice to the test room or test apparatus singly, in clean cages.
The mouse is placed centrally into the larger, brightly illuminated compartment fac-
ing away from the partition. Mice are allowed to freely explore the chamber for 10
min while transitions and time spent in the dark compartment are recorded. After
completion of the test, return mice to the home cage. Unlike the EPM, some previ-
ous testing experience with this, or other behavioral tests, does not appear to alter
behavioral performance.40,92,94

The number of transitions and the time spent in the dark compartment are analyzed
using one-way ANOVAs and Newman-Keuls post-hoc comparisons when indicated.
Mice exhibiting higher levels of anxiogenic-like behavior will make fewer transi-
tions between the brightly illuminated, open area and the dark, enclosed compart-
ment. Many laboratories also use time in the dark or, reciprocally, time in the light
as a measure of anxiogenic-like behavior.51,95–97 Recently some laboratories have
included time spent in risk assessment as another measure of anxiety-related behav-
ior.97 Risk assessment includes a stretch-attend posture in which the head and fore-
paws extend into the lighted area but the remainder of the body stays in the dark
compartment. This test has been shown to be very sensitive to the anti-anxiety–like
effects of benzodiazepines. Benzodiazepines increase transitions between the com-
partments without affecting general locomotor activity. Investigators should use cau-
tion when interpreting results of new compounds on anxiety-like behavior until they
have been screened for nonspecific locomotor effects in a separate apparatus such as
an automated open field.


Figure 5.3A and B illustrate the two most commonly reported behavioral measures
from the light n dark exploration test. Figure 5.3C illustrates risk assessment, a
measure appearing more often in recent literature as another indicator of anxiety-
related behavior in rodents.5,6,42 The study97 illustrated in Figure 5.3 examined the
effects of two centrally administered neuropeptides, NPY and galanin, on anxiety-
related behavior in C57BL/6J mice. Mice receiving two different doses of NPY made
more transitions between the two chambers and spent more time in the light com-
partment than controls or galanin-treated mice. In addition, Figure 5.3C illustrates
that NPY-treated mice spent significantly less time engaged in risk assessment than
controls or galanin-treated mice.97 Thus, neuropeptide Y, but not galanin, produced
an anxiolytic-like action when centrally administered to mice.
5.3.4 THE SOCIAL INTERACTION TEST

The social interaction test, developed by File and Hyde,98 provided the first test of
anxiety-like behavior that focused on ethologically relevant concepts. The test elimi-
nated the need to introduce aversive or appetitive conditions. In addition, the design
of the social interaction test is suitable for use with naïve animals. Pairs of male rats
are allowed to freely interact in an arena while time spent interacting is recorded as
the dependent measure. Interaction time for each of the rats in the pair is directly
impacted by the behavior of the partner animal. Therefore, the pair counts as one
unit for data collection purposes. If the design of the experiment involves one rat
receiving treatment while the other serves as a control, then interaction time initiated
by the treated rat is the appropriate dependent measure. Anxiolytic-like behavior is
inferred if social interaction time increases and general motor activity remains unaf-
fected. Conversely, decreased time spent engaging in social behavior would indicate
anxiogenic-like behavior.
Manipulating environmental conditions allows the researcher to induce varying
levels of anxiety in the test subject. The arena is either familiar or novel and illumi-
nation levels can range from bright to dim. These conditions can be characterized as
low anxiety inducing when the environment is familiar (F) and illumination levels
are low (L), versus high anxiety inducing when lighting conditions are bright and the
test arena is unfamiliar (U). The remaining two conditions, low illumination, novel
arena and high (H) illumination, familiar arena, result in moderate baseline anxiety
levels.99 The ability to systematically increase or decrease baseline anxiety levels has
proven especially useful for screening novel pharmaceutical compounds developed
for treating anxiety. Higher baseline anxiety levels, induced by the HU condition,
are well suited for detecting the effects of anxiolytic compounds. Conversely, robust
anxiogenic effects can be detected in the LF condition, when it is expected that the
greatest amount of time would be spent in social interaction.99 It should be noted that
adult female rats failed to increase time spent in social interaction as a function of
increasing familiarity of the test environment,100 suggesting that some environmen-
tal manipulations have different salience for the social behaviors of each sex.101
Although originally designed for rats, modified versions of this test have
been used relatively successfully to evaluate anxiety-like behavior in mice.102,103 It

500
*** ***
Time Spent in the Light

400
Compartment (sec) 300
200
100
0
Veh 0.5 GAL 1 GAL 0.5 NPY 1 NPY
(a)
75 * *
Light-dark Transitions
50
25
0
(b)
100
Time of Risk Assessment (sec)
75
50
**
25 **
0
(c)
FIGURE 5.3 Light n dark exploration. Mice treated with NPY at an icv dose of 0.5 and
1.0 nmol spent significantly more time in the brightly lit open area (A) and made significantly
more transitions (B) between the two compartments than the vehicle-treated (deionized water)
control group. Attempts to enter the light compartment, termed risk assessment, were signifi-
cantly lower in NPY-treated mice than vehicle-treated mice (C). The neuropeptide galanin
did not produce any significant effects at similar doses in this test. N = 8–13 per treatment
group. *p < 0.05, **p < 0.01, ***p < 0.001. Data are shown as mean + standard error of the
mean. Source: Reprinted from Karlsson, Holmes, Heilig, and Crawley. 2005. Anxiolytic-like
actions of centrally administered neuropeptide Y, but not galanin, in C57BL/6J mice. Phar-
macol. Biochem. and Behav. 80:431, with permission from Elsevier.

should be noted that effects demonstrated in mice are less consistent than those
exhibited by rats. Of the manipulated variables, light level appears to have the great-
est impact on anxiety in mice,104,105 while familiarity of the test arena, similar to
the response of female rats, does not provide consistent changes in anxiety level in
mice. In singly housed mice, similar to the effect seen in rats, anxiolytics reverse the
inhibition of social interaction induced by brighter lighting.102,106
5.3.4.1 Subjects
Inconsistent findings with female mice would indicate that this test is more suitable
for testing male social behavior. Young male mice of approximately the same weight
(< 4 g difference) are the preferred subjects. Noticeably aggressive, dominant, group-
housed mice should not be used, as this could significantly impact the sociability of
the isolate mouse.
5.3.4.2 Equipment
The novel cage environment can be a standard polypropylene rat cage or clear Plexi-
glas chamber that is unfamiliar to the subjects before acclimation. Recording equip-
ment is mounted above the cage at a distance that provides complete coverage of the
arena but does not interfere with the test environment.
5.3.4.3 Procedure
Social interaction is tested between pairs of mice that are either singly housed for
3–6 wk or group housed. Test pairs can involve one group-housed and one isolate
mouse, or two unfamiliar, group-housed mice. Singly housing mice has been dem-
onstrated to increase social investigation.105,106 Isolate mice are acclimated to the
testing cage (size ranges 30 × 25 × 17 cm, 20 × 30 × 20 cm) for 30 min prior to test-
ing. At the end of the acclimation period a group-housed mouse is introduced for a
4-min test period. In the case of pairs of unfamiliar, group-housed mice, each mouse
is given a 10-min acclimation session in the test cage on the two days prior to the
experiment. On day 3 the pair of mice is placed into the test cage for the 10-min test
session.103 Test sessions are recorded and scored at a later time. As the test was origi-
nally developed, the mean total time engaged in social behaviors is scored, analyzed,
and reported.17 An alternative to this method is to score categories of behavior for
each treatment group including aggressive (attack, aggressive unrest), fearful (vigi-
lant posture, escape and defense activity), social (following, social sniffing, over-
under climbing), and locomotion (rearing, walking during cage investigation) and
report the mean number of events in each category.102 Scorers should be blind to any
experimental treatment. Inter-rater reliability values are determined for a sampling
of the tested mice by multiple scorers. If the experimental design includes evaluating
the effects of anxiolytic ligands, illumination levels can be increased to inhibit base-
line social investigation. Conversely, low illumination levels (< 20 lux) may enhance
social investigation, providing a method for exploring anxiogenic effects on baseline
social interactions.


Mean time spent in social interaction is the most reported parameter of social
behavior. Active social behavior of the subject mouse is scored, including following,
sniffing, and climbing on or under the other mouse. The means for two groups are
analyzed using an unpaired Student’s t-test. If lighting levels have been manipulated,
or more than two groups are included, then analyze the means using an ANOVA fol-
lowed by Neuman-Keuls post-hoc test when indicated.

Time spent interacting with an unfamiliar partner is the primary measure reported as
a measure of sociability in mice. Figure 5.4 illustrates a study that examined social
investigation in vasopressin 1a receptor (V1aR) KO and WT control mice.103 Previ-
ous findings had implicated this receptor in modulating social recognition memory.
The mean amount of time spent engaged in social behavior is shown in Figure 5.4 for
pairs of mice with a null mutation in the V1aR KO compared to WT pairs. V1aR KO
pairs spent significantly less time in social interaction than WT pairs.
5.3.5 NOVELTY-INDUCED HYPOPHAGIA

Rodents encountering a desirable food in a novel environment will consume very
limited quantities after considerable investigation. Mice tend to avoid exploration of
novel open environments, yet are motivated to approach and consume palatable food.
150
120
Social Interaction Time (sec)
90 *
*
*
60
30
(8) (8)
0
WT V1aR KO
FIGURE 5.4 Time spent in social investigation in the social interaction test of V1aR knock-
out (KO) mice and wild type mice. V1aR KO mice spent significantly less time in social
interactions (***p < 0.0001) compared to wild type mice. N = 8 pairs of each genotype. Data
are shown as mean + standard error of the mean. Source: Reprinted from Egashira, Tanove,
and Matsuda et al. 2007. Impaired social interaction and reduced anxiety-related behavior in
vasopressin V1a receptor knockout mice. Behavioral Brain Research 178:125, with permis-
sion from Elsevier.

This inhibition of feeding behavior has been termed hyponeophagia and is robust in
both rats and mice. The response is unconditioned, requires no training, and can be
elicited in food-deprived or satiated animals by substituting a highly palatable food
source for regular chow. Treatment with a variety of drugs used to manage anxiety in
humans reliably reverses this decrement in feeding, reducing the latency to the first
taste and increasing the total amount of food consumed (for review see107,108). Sev-
eral factors have been found to influence baseline levels of hyponeophagia in mice,
including the genetic background of inbred strains, long durations of isolate housing,
and specific genetic mutations that affect anxiety-related behaviors.109–112
Several methodological concerns have been raised with hyponeophagia-based
testing. One is the failure of many designs to include a comparison of food con-
sumption in the home cage environment.107 Investigators should report equivalent
assessment measures of feeding behavior (latency and total consumed) in both the
novel and home cage environments to determine the contribution of the independent
variable to any observed differences.107 Another possible confound is the potential
impact of drug treatments or genetic manipulations on factors unrelated to anxi-
ety. Drugs targeting serotonergic function selectively decrease feeding behavior and
alter macronutrient intake.113,114 Experimental protocols that incorporate food depri-
vation may compound these appetite-related effects, potentially masking or exacer-
bating anxiety-like measures. Substituting a familiar, highly palatable food in the
home cage and unfamiliar environment minimizes some of these methodological
problems.107 In the home cage mice quickly approach and ingest the food. In the
novel environment they show a marked increase in latency to first taste the familiar
food.108 In addition, Dulawa and Hen107 suggest using higher illumination levels for
the novel environment to optimize hyponeophagia levels.
In their modified model, Dulawa, Holick, Gundersen, and Hen (2004)115 propose
reporting measures of both latency and total food consumed in the novel and home
cage environments. When latency alone is reported, home cage scores may be very
low, making it extremely difficult to detect manipulations expected to enhance appe-
tite. This modified model provides some advantages over older versions, including
improved sensitivity and reliability of the test results by assessing two behavioral
measures, and increasing the likelihood that the test will discriminate treatments
that enhance, as well as decrease, feeding behavior. However, as with most designs,
there are a few limitations to note, including training the mice to consume the highly
palatable novel food and single housing animals immediately prior to testing.
5.3.5.1 Subjects
Mice ranging in age from juvenile to older adult can be tested in this paradigm.
As mentioned above, attention should be given to the background strain, housing
arrangements, and genetic mutations designed to influence emotionality, as these
may alter baseline levels of feeding behavior. Depending on the independent vari-
ables of interest in the experimental design, group-housed mice should be singly
housed for at least 5 days prior to testing.

5.3.5.2 Equipment
Standard mouse cages of identical size can be used for the home cage and novel
cage environments. In the novel environment condition, cages can be either free
of bedding or have new bedding. One option for a highly palatable food source is
diluted (3–1) sweetened condensed milk (Carnation), although other food may be
substituted. Lighting in the home cage condition is dim (~50 lux). The illumination
level for novel cage testing is very bright (~1200 lux) and the table area under the test
cage is lined with white paper.
5.3.5.3 Procedure
Singly housed mice are trained to consume the palatable food source by introducing it
to them in their home cage for 30 min daily over three consecutive days. Diluted con-
densed milk in plastic serological pipettes (10 mL) with attached sippers and rubber
stoppers are mounted to the wire cage lid. Mice are allowed access for 30 min daily.
On the fourth day mice are tested in the home cage condition. Remove mice from the
cage while the pipette is installed on the cage lid. This maintains a consistency in the
handling procedure for the two (home versus novel cage) experimental conditions.
Commence testing as soon as mice are returned to the cage. Record the latency to the
first lick and the total volume consumed in 5-min intervals across the 30-min session.
Note any mice that do not consume any condensed milk. They should be excluded
from further testing as they failed the training protocol. On day 5, position the pipette
in the wire lid of the novel cage and place the mouse into the novel cage environment.
Record latency and total volume consumed as previously described.

Comparison of the total volume of food consumed across the 30 min can be analyzed
using a between-subjects repeated measures ANOVA. Although the initial 5-min
period may provide sufficient information for assessing anxiolytic effects of treat-
ment, the sensitivity of this initial period for distinguishing anxiogenic effects is
less certain.115 One-way ANOVA may be used for comparing means for total food
consumed. Latency data generally violate several assumptions of the ANOVA test;
therefore, violations of these tenets should be examined. It may be necessary to trans-
form or truncate the data, according to statistical convention, prior to analysis.116

The graphs presented in Figure 5.5A and B illustrate the behavioral measures gen-
erally reported in novelty-induced hypophagia testing, latency to the first lick, and
total volume of food consumed in the initial 5-min period.115 This study examined
the effect of chronic fluoxetine treatment (29 days) on the latency and volume of
food consumed in a familiar home cage environment versus a novel environment for
BALB/cJ mice, a highly anxious strain. Fifteen mice per group received tap water
laced with one of three fluoxetine doses or tap water only. Fluoxetine decreased the
latency to the first lick at all doses in the novel cage, but had no effect on latency to
drink in the home cage (Figure 5.5A). In addition, in the novel environment fluoxetine


!

")(+

"&#)*%'%$

FIGURE 5.5 Novelty-induced hypophagia. The effects of a novel cage on latency to con-
sume, and the amount consumed, of a familiar and palatable snack are shown for BALB/c
mice. The difference in latency to consume (A) in the home cage, (B) in a novel cage, and
(C) amount consumed in the first 5 min in the home cage and a novel cage, for BALB/c mice
receiving 0 (n = 13), 10 (n = 13), 18 (n = 12), or 25 (n = 14) mg/kg/day chronic fluoxetine
treatment, *p < 0.05 vs. control group with ANOVA, is shown. Data are shown as mean +
standard error of the mean. Source: Reprinted from Dulawa, Holick, Gundersen and Hen.
2004. Effects of chronic fluoxetine in animal models of anxiety and depression. Neuropsy-
chopharmacology. 29:1327, with permission from Elsevier.

decreased overall food consumption compared to the home environment. However,

in both the home and novel cages, the 18 mg/kg dose increased food consumption
over that of the 0 and 10 mg/kg doses. Although 25 mg/kg did increase food con-
sumption in both conditions, serum levels of mice in the home cage were more than
twice that observed in humans. In this study the 10 and 18 mg/kg doses produced
anxiolytic effects in the novelty-induced hypophagia test.115
5.4 CONCLUSION
Several ethologically relevant tests of anxiety-like behavior have been presented
as a representative sampling of the broader collection of assays designed to assess
anxiety-related behavior in mice in the field of behavioral neuroscience. Space limi-
tations and methodological specificity necessitated limiting the scope of the pres-
ent work to this smaller subset of anxiety-related behavioral tests. The interested
researcher seeking additional tests that directly assess anxiety-like behavior may
wish to explore the following excellent paradigms: stress-induced hyperthermia, a
measure of the effect of stress (handling, temperature measurement) on body tem-
perature;117 the mouse marble-burying test, a modification of the shock-probe bury-
ing test for rats;118,119 the open field emergence test;52 fear conditioned startle and
light enhanced startle;27,120,121 and the Vogel conflict test.19 Investigators seeking an
in-depth characterization of anxiety-related behaviors in a mutant line of mice are
encouraged to conduct two or more of these well-validated assays to strengthen the
interpretation of their findings.
REFERENCES
1. Nutt, D. J. 1990. The pharmacology of human anxiety. Pharmacology & Therapeutics
47 (2):233–66.
2. Weiss, S. J. 2007. Neurobiological alterations associated with traumatic stress. Perspec-
tives in Psychiatric Care 43 (3):114–22.
3. American Psychiatric Association. 2000. Diagnostic and statistical manual of mental
disorders: DSM-IV-TR. Washington, DC: American Psychiatric Association.
4. Blanchard, R. J., and Blanchard D. C. 1989 Attack and defense in rodents as ethoex-
perimental models for the study of emotion. Progress in Neuro-Psychopharmacology
& Biological Psychiatry 13:S3–S14.
5. Blanchard, R. J., Griebel, G., Henrie, J. A., and Blanchard, D. C. 1997. Differentiation
of anxiolytic and panicolytic drugs by effects on rat and mouse defense test batteries.
Neurosci. Biobehav. Rev. 21 (6):783–89.
6. Blanchard, D. C., Griebel, G., and Blanchard, R. J. 2003. The mouse defensive test bat-
tery: Pharmacological and behavioral assays for anxiety and panic. European Journal
of Neuroscience 463:97–116.
7. Cryan, J. F., and Holmes, A. 2005. The ascent of mouse: Advances in modelling human
depression and anxiety. Nature Reviews: Drug Discovery 4 (9):775–90.
8. Hall, C. S. 1934. Emotional behavior in the rat. I. Defecation and urination as mea-
sures of individual differences in emotionality. Journal of Comparative Psychology 18
(3):385–403.
9. Ohl, F. 2005. Animal models of anxiety. Handbook of Experimental Pharmacology
(169):35–69.

10. Rodgers, R. J., Cao, B. J., Dalvi, A., and Holmes, A. 1997. Animal models of anxiety:
An ethological perspective. Brazilian Journal of Medical and Biological Research 30
(3):289–304.
11. Belzung, C., and Griebel, G. 2001. Measuring normal and pathological anxiety-like
behaviour in mice: A review. Behavioural Brain Research 125:141–49.
12. Landgraf, R., and Wigger, A. 2002. High vs. low anxiety-related behavior rats: An
animal model of extremes in trait anxiety. Behavior Genetics 32 (5):301–14.
13. Finn, D. A., Rutledge-Gorman, M. T., and Crabbe, J. C. 2003. Genetic animal models
of anxiety. Neurogenetics 4 (3):109–35.
14. Gross, C., Zhuang, X., Stark, K., et al. 2002. Serotonin1A receptor acts during develop-
ment to establish normal anxiety-like behaviour in the adult. Nature 416:396–400.
15. Borsini, F., Lecci, A., Volterra, G., and Meli, A. 1989. A model to measure anticipatory
anxiety in mice? Psychopharmacology 98 (2):207–11.
16. Crawley, J., and Goodwin, F. K. 1980. Preliminary report of a simple animal behavior
model for the anxiolytic effects of benzodiazepines. Pharmacology, Biochemistry, and
Behavior 13 (2):167–70.
17. File, S. E. 1980. The use of social interaction as a method for detecting anxiolytic activ-
ity of chlordiazepoxide-like drugs. Journal of Neuroscience Methods 2 (3):219–38.
18. Slotnick, B. M., and Jarvik, M. E. 1966. Deficits in passive avoidance and fear condi-
tioning in mice with septal lesions. Science 154 (3753):1207–8.
19. Vogel, J. R., Beer, B., and Clody, D. E. 1971. A simple and reliable conflict procedure
for testing anti-anxiety agents. Psychopharmacologia 21 (1):1–7.
20. Fendt, M., and Fanselow, M. S. 1999. The neuroanatomical and neurochemical basis of
conditioned fear. Neurosci. Biobehav. Rev. 23 (5):743–60.
21. Crawley, J. N. 1999. Behavioral phenotyping of transgenic and knockout mice: Experi-
mental design and evaluation of general health, sensory functions, motor abilities, and
specific behavioral tests. Brain Research 835 (1):18–26.
22. Crawley, J. N. 2007. What’s wrong with my mouse?: Behavioral phenotyping of trans-
genic and knockout mice. 2nd ed. Hoboken, NJ: Wiley-Liss.
23. Crawley, J. N., Belknap, J. K., Collins, A., et al. 1997. Behavioral phenotypes of inbred
mouse strains: Implications and recommendations for molecular studies. Psychophar-
macology 132:107–24.
24. Crawley, J. N., and Paylor, R. 1997. A proposed test battery and constellations of spe-
cific behavioral paradigms to investigate the behavioral phenotypes of transgenic and
knockout mice. Hormones and Behavior 31:197–211.
25. Holmes, A. 2001. Targeted gene mutation approaches to the study of anxiety-like
behavior in mice. Neurosci. Biobehav. Rev. 25 (3):261–73.
26. Weiss, S.M., Lightowler, S., Stanhope, K. J., Kennett, G. A., and Dourish, C. T.
2000. Measurement of anxiety in transgenic mice. Reviews in the Neurosciences 11
(1):59–74.
27. Davis, M. 1979. Morphine and naloxone: Effects on conditioned fear as measured with
the potentiated startle paradigm. European Journal of Pharmacology 54 (4):341–47.
28. Davis, M. 1990. Animal models of anxiety based on classical conditioning: The condi-
tioned emotional response (CER) and the fear-potentiated startle effect. Pharmacology
& Therapeutics 47 (2):147–65.
29. Davis, M. 1992. The role of the amygdala in fear-potentiated startle: Implications for
animal models of anxiety. Trends in Pharmacological Sciences 13 (1):35–41.
30. Aron, C., Simon, P., Larousse, C., and Boissier, J. R. 1971. Evaluation of a rapid tech-
nique for detecting minor tranquilizers. Neuropharmacology 10 (4):459–69.
31. Pinel, J. P., and Treit, D. 1978. Burying as a defensive response in rats. Journal of Com-
parative and Physiological Psychology 92 (4):708–12.

32. Hofer, M. A. 1973. Maternal separation affects infant rats’ behavior. Behavioral Biol-
ogy 9 (5):629–33.
33. Plotsky, P.M., and Meaney, M. J. 1993. Early, postnatal experience alters hypotha-
lamic corticotropin-releasing factor (CRF) mRNA, median eminence CRF content
and stress-induced release in adult rats. Brain Research: Molecular Brain Research 18
(3):195–200.
34. Archer, T., Sjödén, P. O., and Nilsson, L. G. 1984. The importance of contextual ele-
ments in taste-aversion learning. Scandinavian Journal of Psychology 25 (3):251–57.
35. Wahlsten, D. 2001. Standardizing tests of mouse behavior: Reasons, recommendations,
and reality. Physiology & Behavior 73:695–704.
36. Wahlsten, D., Rustay, N. R., Metten, P., and Crabbe, J. C. 2003. In search of a better
mouse test. Trends in Neurosciences 26:132–36.
37. Würbel, H. 2001. Ideal homes? Housing effects on rodent brain and behaviour. Trends
in Neurosciences 24:207–11.
38. Crabbe, J. C. 1986. Genetic differences in locomotor activation in mice. Pharmacology,
Biochemistry, and Behavior 25:289–92.
39. DeFries, J. C., Gervais, M. C., and Thomas, E. A. 1978. Response to 30 generations of
selection for open-field activity in laboratory mice. Behavior Genetics 8 (1):3–13.
40. McIlwain, K. L., Merriweather, M. Y., Yuva-Paylor, L. A., and Paylor, R. 2001. The
use of behavioral test batteries: Effects of training history. Physiology & Behavior 73
(5):705–17.
41. Paylor, R., Spencer, C. M., Yuva-Paylor, L. A., and Pieke-Dahl, S. 2006. The use
of behavioral test batteries, II: Effect of test interval. Physiology & Behavior 87
(1):95–102.
42. Rodgers, R. J., and Dalvi, A. 1997. Anxiety, defence and the elevated plus-maze. Neu-
rosci. Biobehav. Rev. 21 (6):801–10.
43. Elliott, B. M., and Grunberg, N. E. 2005. Effects of social and physical enrichment on
open field activity differ in male and female Sprague-Dawley rats. Behavioural Brain
Research 165 (2):187–96.
44. Bailey, K. R., Rustay, N. R., and Crawley, J. N. 2006. Behavioral phenotyping of trans-
genic and knockout mice: Practical concerns and potential pitfalls. ILAR Journal/
National Research Council, Institute of Laboratory Animal Resources 47 (2):124–31.
45. Meijer, M. K., Sommer, R., Spruijt, B. M., van Zutphen, L. F., and Baumans, V. 2007.
Influence of environmental enrichment and handling on the acute stress response in
individually housed mice. Laboratory Animals 41 (2):161–73.
46. Van Loo, P. L., Van der Meer, E., Kruitwagen, C. L., Koolhaas, J. M, Van Zutphen, L.
F., and Baumans, V. 2004. Long-term effects of husbandry procedures on stress-related
parameters in male mice of two strains. Laboratory Animals 38 (2):169–77.
47. Christmas, A. J., and Maxwell, D. R. 1970. A comparison of the effects of some benzo-
diazepines and other drugs on aggressive and exploratory behaviour in mice and rats.
Neuropharmacology 9 (1):17–29.
48. Prut, L., and Belzung, C. 2003. The open field as a paradigm to measure the effects of
drugs on anxiety-like behaviors: A review. European Journal of Pharmacology 463
(1–3):3–33.
49. Ernsberger, P., Azar, S., and Iwai, J. 1983. Open-field behavior in two models of genetic
hypertension and the behavioral effects of salt excess. Behavioral and Neural Biology
37 (1):46–60.
50. Kafkafi, N., Lipkind, D., Benjamini, Y., Mayo, C. L., Elmer, G. I., and Golani, I. 2003.
SEE locomotor behavior test discriminates C57BL/6J and DBA/2J mouse inbred strains
across laboratories and protocol conditions. Behavioral Neuroscience 117 (3):464–77.

51. Hefner, K., Cameron, H. A., Karlsson, R. M., and Holmes, A. 2007. Short-term and
long-term effects of postnatal exposure to an adult male in C57BL/6J mice. Behav-
ioural Brain Research 182 (2):344–8.
52. Holmes, A., Kinney, J. A., Wrenn, C. C. et al. 2003. Galanin GAL-R1 receptor null
mutant mice display increased anxiety-like behavior specific to the elevated plus-maze.
Neuropsychopharmacology: Official Publication of the American College of Neuro-
psychopharmacology 28 (6):1031–44.
53. Bailey, K.R., Pavlova, M. N., Rohde, A. D., Hohmann, J. G., and Crawley, J. N. 2007.
Galanin receptor subtype 2 (GalR2) null mutant mice display an anxiogenic-like phe-
notype specific to the elevated plus-maze. Pharmacology, Biochemistry, and Behavior
86 (1):8–20.
54. Pellow, S., and File, S. E. 1986. Anxiolytic and anxiogenic drug effects on exploratory
activity in an elevated plus-maze: A novel test of anxiety in the rat. Pharmacology,
Biochemistry, and Behavior 24 (3):525–29.
55. Rodgers, R. J., Johnson, N. J., Carr, J., and Hodgson, T. P. 1997. Resistance of experi-
entially induced changes in murine plus-maze behaviour to altered retest conditions.
Behavioural Brain Research 86 (1):71–77.
56. Bertoglio, L. J., and Carobrez, A. P. 2002. Prior maze experience required to alter
midazolam effects in rats submitted to the elevated plus-maze. Pharmacology, Bio-
chemistry, and Behavior 72 (1–2):449–55.
57. Gonzalez, L.E., and File, S. E. 1997. A five minute experience in the elevated plus-maze
alters the state of the benzodiazepine receptor in the dorsal raphe nucleus. The Journal
of Neuroscience: The Official Journal of the Society for Neuroscience 17 (4):1505–11.
58. Handley, S. L., and Mithani, S. 1984. Effects of alpha-adrenoceptor agonists and antag-
onists in a maze-exploration model of “fear”-motivated behaviour. Naunyn-Schmiede-
berg’s Archives of Pharmacology 327 (1):1–5.
59. Pellow, S., Chopin, P., File, S. E., and Briley, M. 1985. Validation of open:closed arm
entries in an elevated plus-maze as a measure of anxiety in the rat. Journal of Neurosci-
ence Methods 14 (3):149–67.
60. Lister, R. G. 1987. The use of the plus-maze to measure anxiety in the mouse. Psycho-
61. Hagenbuch, N., Feldon, J., and Yee, B. K. 2006. Use of the elevated plus-maze test with
opaque or transparent walls in the detection of mouse strain differences and the anxio-
lytic effects of diazepam. Behavioural Pharmacology 17:31–41.
62. Shepherd, J. K., Grewal, S. S., Fletcher, A., Bill, D. J., and Dourish, C. T. 1994. Behav-
ioural and pharmacological characterisation of the elevated “zero-maze” as an animal
model of anxiety. Psychopharmacology 116 (1):56–64.
63. Wall, P. M., and Messier, C. 2000. Ethological confirmatory factor analysis of anxi-
ety-like behaviour in the murine elevated plus-maze. Behavioural Brain Research 114
(1–2):199–212.
64. Wall, P. M., and Messier, C. 2001. Methodological and conceptual issues in the use of
the elevated plus-maze as a psychological measurement instrument of animal anxiety-
like behavior. Neurosci. Biobehav. Rev. 25 (3):275–86.
65. Borsini, F., Podhorna, J., and Marazziti, D. 2002. Do animal models of anxiety predict
anxiolytic-like effects of antidepressants? Psychopharmacology 163 (2):121–41.
66. Griebel, G., Belzung, C., Perrault, G., and Sanger, D. J. 2000. Differences in anxiety-
related behaviours and in sensitivity to diazepam in inbred and outbred strains of mice.
Psychopharmacology 148 (2):164–70.
67. Haller, J., Varga, B., Ledent, C., Barna, I., and Freund, T. F. 2004. Context-dependent
effects of CB1 cannabinoid gene disruption on anxiety-like and social behaviour in
mice. The European Journal of Neuroscience 19 (7):1906–12.

68. Patti, C.L., Kameda, S. R., Carvalho, R. C., et al. 2006. Effects of morphine on the
plus-maze discriminative avoidance task: Role of state-dependent learning. Psycho-
pharmacology 184 (1):112.
69. Rodgers, R.J., Boullier, E., Chatzimichalaki, P., Cooper, G. D., and Shorten, A. 2002.
Contrasting phenotypes of C57BL/6JOlaHsd, 129S2/SvHsd and 129/SvEv mice in
two exploration-based tests of anxiety-related behaviour. Physiology & Behavior 77
(2–3):301–10.
70. Galani, R., Duconseille, E., Bildstein, O., and Cassel, J. C. 2001. Effects of room and
cage familiarity on locomotor activity measures in rats. Physiology & Behavior 74
(1–2):1–4.
71. Holmes, A., and Rodgers, R. J. 1998. Responses of Swiss-Webster mice to repeated
plus-maze experience: Further evidence for a qualitative shift in emotional state? Phar-
macology, Biochemistry, and Behavior 60 (2):473–88.
72. Mitchell, H. A., Ahern, T. H., Liles, L. C., Javors, M. A., and Weinshenker, D. 2006.
The effects of norepinephrine transporter inactivation on locomotor activity in mice.
Biological Psychiatry 60 (10):1046–52.
73. Rodgers, R. J. 1997. Animal models of “anxiety”: Where next? Behavioural Pharma-
cology 8:477–96. Discussion 497–501.
74. Rodgers, R.J., Lee, C., and Shepherd, J. K. 1992. Effects of diazepam on behavioural
and antinociceptive responses to the elevated plus-maze in male mice depend upon
treatment regimen and prior maze experience. Psychopharmacology 106 (1):102–10.
75. Rodgers, R.J., Johnson, N. J., Cole, J. C., Dewar, C. V., Kidd, G. R., and Kimpson, P.
H. 1996. Plus-maze retest profile in mice: Importance of initial stages of trail 1 and
response to post-trail cholinergic receptor blockade. Pharmacology, Biochemistry, and
Behavior 54 (1):41–50.
76. Rodgers, R.J., and Shepherd, J. K. 1993. Influence of prior maze experience on behav-
iour and response to diazepam in the elevated plus-maze and light/dark tests of anxiety
in mice. Psychopharmacology 113 (2):237–42.
77. Treit, D., Menard, J., and Royan, C. 1993. Anxiogenic stimuli in the elevated plus-maze.
Pharmacology, Biochemistry, and Behavior 44 (2):463–69.
78. File, S.E. 1990. One-trial tolerance to the anxiolytic effects of chlordiazepoxide in the
plus-maze. Psychopharmacology 100 (2):281–82.
79. Lee, C., and Rodgers, R. J. 1990. Antinociceptive effects of elevated plus-maze exposure:
Influence of opiate receptor manipulations. Psychopharmacology 102 (4):507–13.
80. Griebel, G., Moreau, J. L., Jenck, F., Misslin, R., and Martin, J. R. 1994. Acute and
chronic treatment with 5-HT reuptake inhibitors differentially modulate emotional
responses in anxiety models in rodents. Psychopharmacology 113 (3–4):463–70.
81. Bertoglio, L.J., and Carobrez, A. P. 2000. Previous maze experience required to
increase open arms avoidance in rats submitted to the elevated plus-maze model of
anxiety. Behavioural Brain Research 108 (2):197–203.
82. Bertoglio, L.J., and Carobrez, A. P. 2002. Behavioral profile of rats submitted to session
1-session 2 in the elevated plus-maze during diurnal/nocturnal phases and under differ-
ent illumination conditions. Behavioural Brain Research 132 (2):135–43.
83. Bertoglio, L.J., and Carobrez, A. P. 2002. Anxiolytic effects of ethanol and phenobar-
bital are abolished in test-experienced rats submitted to the elevated plus maze. Phar-
macology, Biochemistry, and Behavior 73 (4):963–69.
84. Carobrez, A. P., and Bertoglio, L. J. 2005. Ethological and temporal analyses of anxi-
ety-like behavior: The elevated plus-maze model 20 years on. Neurosci. Biobehav. Rev.
29 (8):1193–1205.
85. File, SE. 1993. The interplay of learning and anxiety in the elevated plus-maze. Behav-
ioural Brain Research 58 (1–2):199–202.

86. Adamec, R., and Shallow, T. 2000. Effects of baseline anxiety on response to kindling
of the right medial amygdala. Physiology & Behavior 70 (1–2):67–80.
87. Hogg, S. 1996. A review of the validity and variability of the elevated plus-maze as an
animal model of anxiety. Pharmacology, Biochemistry, and Behavior 54 (1):21–30.
88. Karlsson, R. M., and Holmes, A. 2006. Galanin as a modulator of anxiety and depres-
sion and a therapeutic target for affective disease. Amino Acids 31 (3):231–39.
89. Ogren, S. O., Kuteeva, E., Hökfelt, T., and Kehr, J. 2006. Galanin receptor antago-
nists: A potential novel pharmacological treatment for mood disorders. CNS Drugs 20
(8):633–54.
90. Swanson, C.J., Blackburn, T. J., Zhang, X. et al. 2005. Anxiolytic- and antidepres-
sant-like profiles of the galanin-3 receptor (Gal3) antagonists SNAP 37889 and SNAP
398299. Proceedings of the National Academy of Sciences of the United States of
America. 102 (48):17,489–94.
91. Wrenn, C.C., and Holmes, A. 2006. The role of galanin in modulating stress-related
neural pathways. Drug News & Perspectives 19 (8):461–67.
92. Blumstein, L. K., and Crawley, J. N. 1983. Further characterization of a simple, auto-
mated exploratory model for the anxiolytic effects of benzodiazepines. Pharmacology,
Biochemistry, and Behavior 18 (1):37–40.
93. Crawley, J. N. 1981. Neuropharmacologic specificity of a simple animal model for the
behavioral actions of benzodiazepines. Pharmacology, Biochemistry, and Behavior 15
(5):695–99.
94. Crawley, J. N. 1985. Exploratory behavior models of anxiety in mice. Neuroscience &
Biobehavioral Reviews 9:37–44.
95. Jacobson, L. H., Bettler, B., Kaupmann, K., and Cryan, J. F. 2007. Behavioral evalua-
tion of mice deficient in GABA-sub(B(1)) receptor isoforms in tests of unconditioned
anxiety. Psychopharmacology 190 (4):541–53.
96. Karl, T., Burne, T. H. J., and Herzog, H. 2006. Effect of Y-sub-1 receptor deficiency on
motor activity, exploration, and anxiety. Behavioural Brain Research 167 (1):87–93.
97. Karlsson, R. M., Holmes, A., Heilig, M., and Crawley, J. N. 2005. Anxiolytic-like
actions of centrally-administered neuropeptide Y, but not galanin, in C57BL/6J mice.
Pharmacology, Biochemistry, and Behavior 80 (3):427–36.
98. File, S.E., and Hyde, J. R. 1978. Can social interaction be used to measure anxiety?
British Journal of Pharmacology 62 (1):19–24.
99. File, S.E., and Seth, P. 2003. A review of 25 years of the social interaction test. Euro-
pean Journal of Pharmacology 463 (1–3):35–53.
100. Johnston, A. L., and File, S. E. 1991. Sex differences in animal tests of anxiety. Physiol-
ogy & Behavior 49 (2):245–50.
101. File, S. E. 2001. Factors controlling measures of anxiety and responses to novelty in the
mouse. Behavioural Brain Research 125:151–57.
102. Krsiak, M., and Sulcova, A. 1990. Differential effects of six structurally related ben-
zodiazepines on some ethological measures of timidity, aggression and locomotion in
mice. Psychopharmacology 101 (3):396–402.
103. Egashira, N., Tanoue, A., Matsuda, T., et al. 2007. Impaired social interaction and
reduced anxiety-related behavior in vasopressin V1a receptor knockout mice. Behav-
ioural Brain Research 178 (1):123–27.
104. de Angelis, L., and File, S. E. 1979. Acute and chronic effects of three benzodiazepines
in the social interaction anxiety test in mice. Psychopharmacology 64 (2):127–29.
105. Lister, R. G., and Hilakivi, L. A. 1988. The effects of novelty, isolation, light and etha-
nol on the social behavior of mice. Psychopharmacology 96 (2):181–87.
106. Krsiak, M., Sulcova, A., Donat, P., Tomasikova, Z., Dlohozkova, N., Kosar, E., et al.
1984. Can social and agonistic interactions be used to detect anxiolytic activity of
drugs? Progress in Clinical and Biological Research 167:93–114.

107. Dulawa, S. C., and Hen, R. 2005. Recent advances in animal models of chronic anti-
depressant effects: The novelty-induced hypophagia test. Neurosci. Biobehav. Rev. 29
(4–5):771–83.
108. Merali, Z., Levac, C., and Anisman, H. 2003. Validation of a simple, ethologically
relevant paradigm for assessing anxiety in mice. Biological Psychiatry 54 (5):552–65.
109. Jennings, K. A., Loder, M. K., Sheward, W. J., et al. 2006. Increased expression of the
5-HT transporter confers a low-anxiety phenotype linked to decreased 5-HT transmis-
sion. The Journal of Neuroscience: The Official Journal of the Society for Neurosci-
ence 26 (35):8955–64.
110. Santarelli, L., Gobbi, G., Blier, P., and Hen, R. 2002. Behavioral and physiologic effects
of genetic or pharmacologic inactivation of the substance P receptor (NK1). The Jour-
nal of Clinical Psychiatry 63:11–7.
111. Trullas, R., and Skolnick, P. 1993. Differences in fear motivated behaviors among
inbred mouse strains. Psychopharmacology 111 (3):323–31.
112. Võikar, V., Polus, A., Vasar, E., and Rauvala, H. 2005. Long-term individual housing
in C57BL/6J and DBA/2 mice: Assessment of behavioral consequences. Genes, Brain,
and Behavior 4 (4):240–52.
113. Heisler, L. K., Kanarek, R. B, and Gerstein, A. 1997. Fluoxetine decreases fat and
protein intakes but not carbohydrate intake in male rats. Pharmacology, Biochemistry,
and Behavior 58 (3):767–73.
114. Leibowitz, S. F., Alexander, J. T., Cheung, W. K., and Weiss, G. F. 1993. Effects of
serotonin and the serotonin blocker metergoline on meal patterns and macronutrient
selection. Pharmacology, Biochemistry, and Behavior 45 (1):185–94.
115. Dulawa, S. C., Holick, K. A., Gundersen, B., and Hen, R. 2004. Effects of chronic fluox-
etine in animal models of anxiety and depression. Neuropsychopharmacology: Official
Publication of the American College of Neuropsychopharmacology 29 (7):1321–30.
116. Kirk, R. E. 1995. Experimental design: Procedures for the behavioral sciences. Pacific
Grove: Brooks/Cole Publishing Company.
117. Bouwknecht, A. J., Olivier, B., and Paylor, R. E. 2007. The stress-induced hyperthermia
paradigm as a physiological animal model for anxiety: A review of pharmacological
and genetic studies in the mouse. Neurosci. Biobehav. Rev. 31 (1):41–59.
118. Njung’e, K., and Handley, S. L. 1991. Evaluation of marble-burying behavior as a model
of anxiety. Pharmacology, Biochemistry, and Behavior 38 (1):63–67.
119. Treit, D., Pinel. J. P., and Fibiger, H. C. 1981. Conditioned defensive burying: A new
paradigm for the study of anxiolytic agents. Pharmacology, Biochemistry, and Behav-
ior 15:619–26.
120. Kehne, J. H., Cassella, J. V., and Davis, M. 1988. Anxiolytic effects of buspirone and
gepirone in the fear-potentiated startle paradigm. Psychopharmacology 94 (1):8–13.
121. Walker, D. L., and Davis, M. 2002. Light-enhanced startle: Further pharmacological
and behavioral characterization. Psychopharmacology 159 (3):304–10.

6 Behavioral Assessment
of Antidepressant
Activity in Rodents
Vincent Castagné, Paul Moser,
and Roger D. Porsolt
CONTENTS
6.1 Introduction................................................................................................. 103
6.2 Methods....................................................................................................... 106
6.2.1 Animal Subjects............................................................................... 107
6.2.2 Equipment ........................................................................................ 107
6.2.3 Procedure: Forced Swimming Test in the Rat (Protocol 1)............. 108
6.2.4 Procedure: Forced Swimming Test in the Mouse (Protocol 2) ....... 109
6.2.5 Procedure: Tail Suspension Test in the Mouse (Protocol 3)............ 109
6.3 Typical Applications ................................................................................... 110
6.4 Analysis and Interpretation......................................................................... 110
6.5 Representative Data .................................................................................... 111
6.6 Comparison with Related Procedures......................................................... 114
References.............................................................................................................. 114
6.1 INTRODUCTION
Depression is one of several disorders affecting mood, along with mania, hypoma-
nia, and bipolar disorders. The present chapter focuses on behavioral assessment of
antidepressant action in animals with a focus on simple tests performed in rodents.
Many of the primary symptoms of depression (depressed mood, low self-esteem,
guilt, difficulty in concentration, suicidal ideation, thoughts of death) are by their
nature difficult to model in animals. This problem is further confounded by their
unknown etiology. Several theories have been proposed1 but most theories of depres-
sion concur in suggesting that stressful life events play an important role. There is
also a small genetic component, as demonstrated by substantially increased risk in
families with heritability being estimated at between 40% and 70%, leading to a
much greater incidence than observed in the general population, which is neverthe-
less very high at around 10%.2
If little is known about the etiology of depression, even less is known about mania
and bipolar disorders. The genetic component appears to be greater than for unipolar
depression.3 Modeling the cycling, recurrent nature of bipolar disorder in animals
103

has not even been attempted. There are, however, some models for mania that pres-
ent an interesting pharmacology, in particular the combined amphetamine-chlordi-
azepoxide hyperactivity model, although the few publications on these models and
their lack of reproducibility from one laboratory to another4–7 make an overview of
their utility difficult. They will not be further discussed in this chapter.
The clinical diagnosis of depression requires the presence of several “core”
symptoms (depressed mood, decreased pleasure) often accompanied by more vari-
able symptoms such as irritability, changes in weight, sleep disturbance, feelings of
guilt, poor concentration, thoughts of death, suicidal ideation, etc. It is clearly not
possible to reproduce in animals all symptoms observed clinically. Table 6.1 shows
the principal symptoms observed in depressed patients and suggests analogous signs
that can be observed in animals. These signs can be used as dependent variables
(end point measures) allowing behavioral assessment in different animal models of
depressive states.
TABLE 6.1
Human Symptoms of Depressive States, Animal Behavioral Signs,
Preclinical Tests
Human Symptom Behavioral Sign Preclinical Test
Depressed mood Resignation Forced swimming
Tail suspension
Learned helplessness
Decreased pleasure Anhedonia Sucrose consumption
ICSS
Sexual behavior
Novelty seeking
Chronic mild stress
Irritability Aggressiveness Muricidal behavior
Social behavior
Olfactory bulbectomy
Changes in weight Body weight Body weight
Food and water intake
Sleep disturbance Sleep architecture EEG
Circadian rhythms
Psychomotor disturbance Locomotor activity Activity meter
Impulsivity DRL
Feelings of guilt No sign identified Not applicable
Poor concentration No sign identified Not applicable
Suicidal ideation No sign identified Not applicable
Thoughts of death No sign identified Not applicable
Note: DRL, differential reinforcement of low rate; EEG, electroencephalograph; ICSS, intracranial
self-stimulation.

Behavioral Assessment of Antidepressant Activity in Rodents 105
Several of the behavioral signs presented in Table 6.1 are, however, amenable
to preclinical testing. Measures thought to be related to resignation (often termed
“behavioral despair” or “learned helplessness”) are used as the main behavioral
parameter in screening tests for antidepressant activity (forced swim and tail sus-
pension tests), as well as in the learned helplessness model. In the first two tests,
immobility induced by exposure to an inescapable aversive situation (forced swim-
ming or suspension by the tail) serves as an indicator of resignation. In the learned
helplessness model, animals (generally rats) are exposed to inescapable foot shocks
and show “helplessness” by subsequently failing to learn to escape when the envi-
ronment is modified to allow escape.8
The forced swim and tail suspension procedures are best viewed as simple tests
for antidepressants rather than as models of depression, because the dependent vari-
able (immobility) is a direct reaction to the test itself and does not persist outside the
test situation. There is no obvious induction of a “depressive state,” although there
are elements of construct validity (stressful inducing conditions, decreased behav-
ioral output). The learned helplessness procedure, where prior exposure to the aver-
sive stress induces a more long lasting change in that animals are subsequently less
able to learn appropriate escape responses, can be considered closer to a model of
depression.8,9 The above procedures have nonetheless been used not only to assess
potential antidepressant activity of test substances, but also to study possible neuro-
biological substrates of depression.9–11 The most obvious difference between these
tests is the duration and frequency of the initiating factors. Prolonged and repeated
stress is probably necessary for inducing a lasting change that could be construed as
a “depressive state.”
The decreased sensitivity and lack of interest in pleasure observed in depressed
patients has some analogy to anhedonia as measured in animals.12 Anhedonia can
be assessed by a variety of tests including the consumption of palatable food (such as
sucrose), intracranial self-stimulation (ICSS), preference for novel objects or situations,
or frequency of sexual interactions.13 Several of these tests have been used to assess
the effects of chronic mild stress and olfactory bulbectomy.13,14 Preference for sucrose
is the most widely used measure of anhedonia.15 Other tests for anhedonia are techni-
cally challenging (ICSS) and are thus less widely used. Of all the available models, the
chronic mild stress procedure possesses the greatest number of attributes of clinical
depression, including putative inducing conditions and a wide variety of long-lasting
behavioral changes. Rats (or mice) submitted to a series of mild stressors, such as food
and water deprivation, soiled cages, and light cycle shifts, show clear and enduring
signs of anhedonia (absence of preference for palatable foods or for novel objects,
higher thresholds for ICSS, lowered sexual activity) and other signs (decreased food
and water intake, weight loss, decreased locomotion, sleep disturbance). On the other
hand, chronic mild stress procedures are very time consuming—a single study could
last 2–3 months16 —are frequently subject to methodological bias, and are reportedly
difficult to reproduce from one laboratory to another.17–18
The olfactory bulbectomy model in rats also induces several long-lasting behav-
ioral changes (increased locomotor activity, passive avoidance deficit, mouse kill-
ing, and intra-specific aggressiveness as observed in dyadic social interaction tests),
together with a variety of neurochemical changes.14–20 Although most of these bear

little direct relation to the clinical symptoms of depression, it is of more concern for
this model that there is no clear analogy between the inducing conditions (olfactory
bulbectomy) and the kind of life events thought to induce or favor depressive states
in humans.21 The usefulness of the olfactory bulbectomy model therefore resides
largely on its predictive validity, in that most clinically effective antidepressants
show activity in the test.14,22
Another approach to assessing the potential antidepressant action of novel sub-
stances is to look at their effects on different behavioral signs that are observed in
clinical depression, but are not necessarily linked to an induced “depressive” state
in the animal. Although problems of body weight loss or gain feature prominently
in depression, and tests for assessing changes in food/water intake or body weight
gain present no major technical difficulty, no specific effects of antidepressants on
these parameters have been described.13,23 Sleep architecture, which is comparable
between humans and animals,24 can be studied by electroencephalographic (EEG)
analysis,25 or more simply by measurement of circadian changes in locomotor activ-
ity.26,27 On the other hand, although it is known that antidepressants affect sleep
architecture in rats,28 there are no data demonstrating the specificity of such changes
to antidepressant action. Few data are available on sleep disturbance in animal mod-
els of depressive states.14,25,29
Another behavior, the capacity of animals to repress a response over a predefined
duration, which is assessed by the differential reinforcement of low rate (DRL) oper-
ant schedule, is thought to represent a measure of impulsivity.30 An abundant amount
of literature31 has shown that numerous antidepressants show a characteristic profile
in this test (moderate decreases in the number of responses accompanied by clear
increases in the number of reinforcements), which can been interpreted as suggest-
ing anti-impulsive activity. It is less clear whether anti-impulsivity characterizes
clinical antidepressant activity.
The brief review presented above indicates the complexity of modeling depres-
sion in animals,18,32,33 in particular the low construct validity of available models.11,34
The problem is less severe for antidepressant testing, where the lack of construct
validity is tempered by an increase in predictive validity.35 The procedures selected
for the following sections (forced swim and tail suspension) represent a compromise
in that they possess high predictive validity but also elements of construct validity.
Furthermore, they do not present any major technical difficulty, are rapid to execute,
and generate data that are highly reproducible.
6.2 METHODS
Rodents forced to swim in small enclosures (cylinders) from which there is no escape
rapidly become immobile after an initial period of vigorous activity.36 Initially,
immobility was interpreted as evidence they had learned that escape was impos-
sible and had given up hope. Immobility was therefore given the name “behavioral
despair.” It has subsequently been shown in numerous laboratories that immobility
is reduced by a wide range of clinically active antidepressant drugs.37 As a conse-
quence, this simple test is now widely used to screen novel substances for potential
antidepressant activity. The following paragraphs describe the basic protocol (forced

swimming test, protocol 138) for examining drug effects in the rat, an equivalent pro-
cedure in the mouse (protocol 239), and the conceptually related tail suspension test,
where immobility is induced by suspending mice by the tail (protocol 340).
6.2.1 ANIMAL SUBJECTS

A large number of rodent strains have been used in the following procedures. Differ-
ent strains of rats display different durations of immobility and variable sensitivity to
antidepressants in the forced swimming test.41,42 Likewise, marked strain differences
have been described in the forced swimming and the tail suspension tests in the
mouse.43–45 To control the variability between different experiments, we recommend
using the same rodent strains within a specific laboratory.
House animals in standard plastic cages (usually 41 × 25 × 15 cm, four to six
rats per cage, or 25 × 19 × 13 cm, 10 mice per cage) containing wood shavings, and
provide free access to a standard rodent diet and tap water, except during the test.
Maintain the animals under strictly controlled environmental conditions (usually
21°C ± 3°C on a standard light-dark cycle with illumination from 0700 to 1900 hr).
Note 1: We recommend that the animals are delivered to the laboratory at least 5
days before the experiment, and are placed in the experimental room at least 60 min
before the test. Experiments should be performed during the light phase of the cycle,
although it is also possible to perform the test under dim red light during the dark
phase of the cycle. In this latter case the light-dark cycle should also be reversed.
Note 2: All protocols using live animals must first be reviewed and approved by an
Institutional Animal Care and Use Committee (IACUC) or must conform to govern-
mental regulations regarding the care and use of laboratory animals.
6.2.2 EQUIPMENT
1. Forced Swimming Test in the Rat (Protocol 1). Transparent Plexiglas cylin-
ders (20 cm in diameter × 40 cm high) containing water (25°C) to a depth
of 13 cm (made in house or obtained from local commercial suppliers).
Opaque screens for visually separating cylinders.
2. Forced Swimming Test in the Mouse (Protocol 2). Transparent Plexiglas
cylinders (13 cm in diameter × 24 cm high) containing water (22°C) to a
depth of 10 cm. Opaque screens for visually separating cylinders.
3. Tail Suspension Test in the Mouse (Protocol 3). Automated tail suspension
apparatus (e.g., Tail Suspension Test System, Bioseb, France) consisting of
plastic enclosures (20 × 25 × 30 cm) fitted with a ceiling hook connected to
a strain gauge and computer assembly.46,47 Without an automated apparatus
it is possible to perform the test using standard laboratory chronometers.
Note: All tests should be performed blind with coded solutions to avoid bias in
evaluating the animal’s behavior. Decoding of treatment group codes should be per-
formed after all evaluations have been completed.

6.2.3 PROCEDURE: FORCED SWIMMING TEST IN THE RAT (PROTOCOL 1)

In two sessions separated by 24 hr, rats are forced to swim in a cylinder from which
they cannot escape. The first 15-min session is conducted prior to drug administra-
tion and without behavioral recording. This prior habituation session ensures a stable
and high duration of immobility during the 5-min test session, usually performed
24 hr later. In the standard procedure, rats are administered the test substance three
times: 24 hr (i.e., immediately after the first session), 4 hr, and 60 or 30 min before
the test (the last pretreatment time depending on the route of administration). Two
or three test substance administrations before the test provide more stable phar-
macological results than a single administration. Control animals receive the same
number of administrations of vehicle.
A variation of the protocol is to insert repeated treatments between the two
sessions. In these cases, the interval between the two sessions usually has to be
increased. We have used intersession intervals of up to 15 days without observing
any change in the behavior during the second session. Repeated administration in
animals is more comparable to the clinical situation where the therapeutic effect of
the majority of antidepressants appears only after several weeks of treatment.48
Equip the experimental room with white neon ceiling lights (standard light-
ing). Set up two transparent cylinders separated visually from each other by opaque
screens. On day 1, at least 60 min before the beginning of the habituation session,
mark the animals and randomly assign them to a drug treatment. All animals within
a cage receive the same treatment. Weigh two animals individually, then place one
rat in each of the two cylinders for 15 min (habituation session). No scoring of immo-
bility is performed during the habituation session. Remove the rats from the cylin-
ders, dry them with a cloth towel, and place them into a cage adjacent to their home
cage. Immediately after the habituation session, treat the first group of two rats with
the appropriate treatment (the first pretest administration), and place them back in
their home cages. Change the water in the cylinders after every three rats. When the
day 1 session is completed, return the animals to the colony room and provide food
and water ad libitum.
On the test day, administer the test substance 4 hr prior to the session and 30 min
(for intraperitoneal or subcutaneous injection) or 60 min (for oral administration)
prior to the session. Test two animals simultaneously in adjacent cylinders separated
by an opaque screen. Observe their behavior for 5 min. Score the duration of immo-
bility by summing the total time spent immobile (i.e., the time not spent actively
exploring the cylinder or trying to escape from it). Included within the time spent
immobile are the short periods of slight activity where the animals just make those
movements necessary to maintain their heads above water.
Note: A standard forced swimming test using 30 rats (five treatment groups of N =
6) requires two consecutive afternoons to perform the two sessions, with the morn-
ing of day 2 reserved for the 4-hr pretest drug administration. If necessary, one
extra group can be tested within the same time frame (i.e., the maximum number
of rats tested per experiment is 36). Experiments requiring more animals should be
performed in separate sub-experiments, with the same number of animals from each
treatment group being tested in each sub-experiment.

6.2.4 PROCEDURE: FORCED SWIMMING TEST IN THE MOUSE (PROTOCOL 2)

In a single session, mice are forced to swim in a narrow cylinder from which they
cannot escape. Equip the experimental room with white neon ceiling lights (stan-
dard lighting). Set up two transparent cylinders separated visually from each other
by opaque screens. At least 60 min before testing, mark the animals and randomly
assign them to a drug treatment. All animals within a cage receive the same treat-
ment. Weigh two mice individually, administer the test substance 30 min (for intra-
peritoneal or subcutaneous injection) or 60 min (for oral administration) prior to the
test and place them back in their home cages. Test two animals simultaneously in
adjacent cylinders separated by an opaque screen. Score immobility during the last
4 min of the 6-min test session by summing the total time spent immobile (i.e., the
time not spent actively exploring the cylinder or trying to escape from it). Included
within the time spent immobile are the short periods of slight activity where the ani-
mals just make those movements necessary to maintain their heads above water.
Whereas mice show a high frequency of exploratory and escape-directed behav-
iors during the first 2 min of the test session, the last 4 min is the time during which
the animals show the most immobility. The first 2 min of the session can be used for
preparing other animals.
Note: A standard forced swimming test using 50 mice (five treatment groups of N =
10) requires a morning or an afternoon. If necessary, two extra groups can be tested
within the same time frame (i.e., the maximum number of mice tested per experi-
ment is 70). Experiments requiring more animals should be performed in separate
sub-experiments, with the same number of animals from each treatment group being
tested in each sub-experiment.
6.2.5 PROCEDURE: TAIL SUSPENSION TEST IN THE MOUSE (PROTOCOL 3)

This protocol describes a procedure in mice that is conceptually related to the forced
swimming test, except that immobility is induced by suspending the mice by the tail.
After initially trying to escape by engaging in vigorous movements, mice rapidly
become immobile. The duration of immobility is reduced by a wide variety of anti-
depressants. This procedure has several advantages over the forced swim procedure
(protocol 2). No hypothermia is induced and the animals resume normal sponta-
neous activity immediately after the test. No special post-experimental treatment
(rubbing down, maintenance in a warmed environment) is required. The procedure
readily lends itself to automation, permitting testing of a greater number of mice
simultaneously.
Equip the experimental room with white neon ceiling lights (standard light-
ing). With the automated tail suspension apparatus (we use the TST System, Bioseb,
France), six mice are tested simultaneously. Weigh the mice and administer the test
substance 30 min (for intraperitoneal or subcutaneous injection) or 60 min (for oral
administration) prior to the test and place the mice back in their home cages. The
different treatments should be administered to individual animals in a fixed rotation
to ensure a regular distribution of the different treatments over time. Our automated
apparatus provides randomization sequences, permitting balanced distribution over

time and over the different positions in the apparatus. Wrap adhesive tape around
the animal’s tail in a constant position three quarters of the distance from the base of
the tail. Suspend the animals by passing the suspension hook through the adhesive
tape so that the animal hangs with its tail in a straight line. Measure the duration of
immobility continuously for 6 min. If an automated testing apparatus is not avail-
able, the duration of immobility can be measured using separate chronometers for
each animal.
Note: Our automated procedure (TST System, Bioseb, France) permits testing of
six animals simultaneously, with all animals being placed in the apparatus before
starting the measurement. For nonautomatic observation, the same observer can
comfortably observe two animals simultaneously. Whatever the configuration, the
animals should be visually shielded from one another during the test.
Note: A standard tail suspension test in the mouse using our automated device
(five treatment groups of N = 12) requires an afternoon. If necessary, three extra
groups can be tested within the same time frame (i.e., the maximum number of mice
tested per experiment is 96). If an automated device cannot be used, the throughput
becomes similar to the forced swimming test in the mouse (i.e., 70 animals can be
tested during the same session).
6.3 TYPICAL APPLICATIONS

The forced swimming test in the rat and the mouse, and the tail suspension test in the
mouse are widely used for early behavioral screening of antidepressants.49,50 Some
false positives (mainly excitatory substances) and some false negatives (mainly sero-
tonin reuptake inhibitors) have been described (see Section 6.5). As a consequence,
we recommend using all three procedures to evaluate new test substances, instead
of relying on a single procedure.34 This reduces the number of false positives and
false negatives and thereby enhances the efficiency of drug discovery programs.
The forced swimming and the tail suspension tests can also be used to character-
ize the phenotype of different strains of animals, including transgenic mice. 51 In
this respect, the forced swimming and the tail suspension tests can also be used as
research tools for investigating the neurobiological bases of depressive states.
6.4 ANALYSIS AND INTERPRETATION

Compare data from treated groups with data from the control group using unpaired
Student’s t-tests (two tailed), although other statistical evaluations (e.g., analysis
of variance followed by post-hoc tests) can also be used. For initial screening, we
strongly recommend two-by-two t-test comparisons of treated groups with control.
Although increasing the risk of a type I error (false positive), there is a decreased
risk of a type II error, i.e., missing a potential drug effect at a particular dose (false
negative), as can happen with more global analyses including all treatments.
Antidepressants decrease the duration of immobility in the forced swimming
test in the rat and the mouse (protocols 1 and 2), and in the tail suspension test in the

mouse (protocol 3). Sedative/myorelaxant substances are generally inactive or even

increase the duration of immobility in the different tests.
6.5 REPRESENTATIVE DATA

Pharmacotherapy of depressed patients uses various classes of antidepressants that
generally target central monoaminergic systems.52,53 Besides monoamine oxidase
inhibitors (MAOIs), the majority of antidepressants belong to different classes of
monoamine reuptake inhibitors, including selective serotonin reuptake inhibitors
(SSRIs), selective norepinephrine reuptake inhibitors (NRIs), and mixed reuptake
inhibitors (SNRIs).54
The etiology of depression is, however, insufficiently understood to limit discov-
ery efforts to substances with clearly identified targets.1 Although the involvement of
central monoaminergic systems in antidepressant action is widely accepted, the long
delay between the initiation of monoamine-based treatments and the first clinical
effects suggests that more complex mechanisms are involved.53 In addition, studies
showing that multiple central neurotransmitter systems are implicated in depression
suggest that non-monoamine-based treatments may represent potentially interest-
ing new therapeutic approaches.55 The major problem with current pharmacological
treatments is that they either fail to produce complete recovery or induce unwanted
side effects. Thus there is urgency for the development of new pharmacological
treatments.
The data presented in Table 6.2 show the effects of diverse substances after i.p.
administration in the forced swimming test in the rat. All substances were adminis-
TABLE 6.2
Effects of Diverse Substances in the Forced Swimming Test in the Rat
Substance Change in immobility (versus vehicle control group) (%)
Dose (mg/kg)
0.5 1 2 4 8 16 32 64
Imipramine NT NT NT NT -17 -23 -48*** -69***
Fluoxetine NT NT NT NT -19 +9 -38* -56*
Desipramine NT NT NT NT -23*** -44*** -45*** -29
Venlafaxine NT NT NT NT -6 -10* -30 84***
8-OH-DPAT -41* -73** NT -100*** NT NT NT NT
Flesinoxan NT +7 NT -64** NT -82** NT NT
Idazoxan NT +7 NT -24* NT NT NT NT
Note: Animals were individually placed in a cylinder (height = 40 cm, diameter = 20 cm) containing 13
cm water (25°C) for 15 min on the first day of the experiment (session 1) and were then put back
in the water 24 hr later for a 5-min test (session 2). The duration of immobility during the 5-min
test was measured. The test substances were administered i.p. 24 hr, 4 hr, and 30 min before the
test (session 2). The duration of immobility was comprised between 150 and 250 sec in the
vehicle control group. Data shown as mean ± SEM. (N = 6 per group). NT, not tested. *p < 0.05;
**p < 0.01 and ***p < 0.001. Student’s t-tests, as compared with vehicle controls.

TABLE 6.3
Effects of Diverse Substances in the Forced Swimming Test in the Mouse
Dose (mg/kg)
0.5 1 2 4 8 16 32 64
Imipramine NT NT NT -4 -22*** -45*** -60*** -100***
Fluoxetine NT NT NT NT -2 -12* -25*** -49***
Desipramine NT NT NT NT -22*** -45*** -60*** -100***
Venlafaxine NT NT NT NT -18*** -32*** -86*** -100***
Clobazam NT NT NT NT +10 +8 +12* +13*
Clozapine NT +2 0 -1 0 NT NT NT
Nicotine NT -12* NT NT NT NT NT NT
8-OH-DPAT -1 NT -54*** NT -27** NT NT NT
Flesinoxan NT +2 NT +13 NT +44** NT NT
Idazoxan NT -6 NT -26*** -30*** NT NT NT
Buspirone NT +24 NT +12 NT +10 NT NT
Alnespirone NT -18 NT -50*** NT -77*** NT NT
Note: Animals were individually placed in a cylinder (height = 24 cm, diameter = 13 cm) containing 10
cm water (approximately 22°C) from which they cannot escape. The mice were placed in the
water for 6 min and the duration of immobility during the last 4 min was measured. The test sub-
stances were administered i.p. 30 min before the test. The duration of immobility was comprised
between 160 and 220 sec in the vehicle control group. Data shown as mean ± SEM. (N = 10 per
group). NT, not tested. *p < 0.05; **p < 0.01 and ***p < 0.001. Student’s t-tests, as compared with
vehicle controls.
tered 24 hr, 4 hr, and 30 min before the test. Dose-dependent decreases in the dura-
tion of immobility are observed with imipramine (8–64 mg/kg), fluoxetine (32 and
64 mg/kg), desipramine (8–32 mg/kg, but not 64 mg/kg), and venlafaxine (16–64
mg/kg). Some serotonergic substances targeting the 5-HT1A receptor also decrease
immobility. This was observed for 8-OH-DPAT (0.5–4 mg/kg), flesinoxan (4 and 16
mg/kg), and idazoxan (4 mg/kg).
administration in the forced swimming test in the mouse. All substances were admin-
istered 30 min before the test. Dose-dependent activity is observed for imipramine
(8–64 mg/kg). Monoamine reuptake inhibitors show clear activity, generally more
marked than in the rat, as observed with fluoxetine (16–64 mg/kg), and desipramine
and venlafaxine (8–64 mg/kg). Clobazam (32 and 64 mg/kg) increases immobility
consistently with its sedative/myorelaxant effects. Clozapine (1–8 mg/kg) is devoid
of activity. Nicotine decreases immobility at 1 mg/kg. Serotonergic substances tar-
geting the 5-HT1A receptor have variable effects in the mouse. 8-OH-DPAT (2 and
8 mg/kg), idazoxan (4 and 8 mg/kg), and alnespirone (4 and 16 mg/kg) decrease
immobility, whereas buspirone (1–16 mg/kg) is devoid of activity, and flesinoxan
increases immobility at 16 mg/kg.

TABLE 6.4
Effects of Diverse Substances in the Tail Suspension Test in the Mouse
Dose (mg/kg)
0.5 1 2 4 8 16 32 64
Imipramine NT NT NT NT -44 -73* -78** NT
Fluoxetine NT NT NT NT -58** -39* -20 -46*
Desipramine NT NT NT NT -75** -64** -53* -62**
Venlafaxine NT NT NT NT -64** -70** -87*** -95***
Clobazam NT NT NT NT +167*** +127*** +100** NT
Clozapine NT +62 +73 +168*** +294*** NT NT NT
Nicotine -17 +5 -12 NT NT NT NT NT
8-OH-DPAT +28 NT +38 NT +154*** NT NT NT
Flesinoxan NT +26 NT +135*** NT +176*** NT NT
Buspirone NT +17 NT +58 NT +103** NT NT
Alnespirone NT +66 NT +109** NT +110** NT NT
Note: Animals were suspended by the tail and the duration of immobility was recorded automatically
for 6 min using a computerized device (Bioseb TST). Six mice were studied simultaneously.
The test substances were administered i.p. 30 min before the test. The duration of immobility
was comprised between 60 and 120 sec in the vehicle control group. Data shown as mean ±
SEM. (N = 10 or 12 per group). NT, not tested. *p < 0.05; **p < 0.01 and ***p < 0.001. Stu-
dent’s t-tests, as compared with vehicle controls.
administration in the tail suspension test in the mouse. All substances were admin-
istered 30 min before the test. Decreases in immobility are observed for imipra-
mine (8–32 mg/kg), and fluoxetine, desipramine, and venlafaxine (all over the dose
range 8–64 mg/kg). Clobazam (8–32 mg/kg) and clozapine (4 and 8 mg/kg) increase
immobility. Nicotine (0.5–2 mg/kg) does not affect immobility. 5-HT1A agonists
increase immobility in the tail suspension test. This is observed for 8-OH-DPAT (8
mg/kg), and buspirone, flesinoxan, and alnespirone (all at 4 and 16 mg/kg).
Taken together, these results confirm the specificity of the forced swimming
test toward antidepressant substances.36 In our hands, the mouse version appears
somewhat more sensitive to serotonin reuptake inhibitors. The weak but significant
activity of nicotine at 1 mg/kg is consistent with the fact that excitatory substances
may constitute false positives in the forced swimming test.41,56 Nevertheless, anti-
depressant-like activity for nicotine has also been described in the mouse57 and in
the rat.58 The tail suspension test is sensitive toward a wide variety of antidepressant
substances that are clearly distinguished from other psychotropic substances such as
anxiolytics, neuroleptics, and other diverse agents.47,59 It is interesting to note that the
tail suspension test in the mouse appears to be more sensitive to the sedative activity
(increase in immobility) of 5-HT1A agonists, in contrast to the forced swimming

test in the rat, which detects mainly their antidepressant-like activity (decrease in
immobility).
6.6 COMPARISON WITH RELATED PROCEDURES

The methods of testing clearly influence the results of the tests. Some studies sug-
gest absence of activity of serotonin reuptake inhibitors using the standard forced
swimming test.60,61 Modifications of the original procedure, including measurement
of active behaviors such as swimming and climbing and increasing the water depth,
have been claimed to facilitate detection of SSRIs.62 Although this scoring method
is now used by several laboratories,63 other data cast doubt on a clear behavioral dis-
tinction of the effects of SSRIs and NRIs.64 Since the standard forced swimming test
allows detection of multiple classes of antidepressants, we prefer to limit the behav-
ioral analysis to the measure of immobility, which is simpler to estimate than the
measure of active behaviors. Extensive training of the observers to score immobility
is needed even in the standard version of the forced swimming test.10
REFERENCES
1. Berton, O., and Nestler, E. J. 2006. New approaches to antidepressant drug discovery:
Beyond monoamines. Nat. Rev. Neurosci. 7:137–51.
2. Sullivan, P. F., Neale, M. C., and Kendler, K. S. 2000. Genetic epidemiology of major
depression: Review and meta-analysis. Am. J. Psychiatry 157:1552–62.
3. Machado-Vieira, R., Kapczinski, F., and Soares, J. C. 2004. Perspectives for the devel-
opment of animal models of bipolar disorder. Prog. Neuropsychopharmacol. Biol. Psy-
chiatry 28:209–24.
4. Okada, K., Oishi, R., and Saeki, K. 1990. Inhibition by antimanic drugs of hyperac-
tivity induced by methamphetamine-chlordiazepoxide mixture in mice. Pharmacol.
5. Cao, B. J., and Peng, N. A. 1993. Magnesium valproate attenuates hyperactivity
induced by dexamphetamine-chlordiazepoxide mixture in rodents. Eur. J. Pharmacol.
237:177–81.
6. Lamberty, Y., Margineanu, D. G., and Klitgaard, H. 2001. Effect of the new antiepilep-
tic drug levetiracetam in an animal model of mania. Epilepsy Behav. 2:454–59.
7. Arban, R., Maraia, G., Brackenborough, K., et al. 2005. Evaluation of the effects of
lamotrigine, valproate and carbamazepine in a rodent model of mania. Behav. Brain
Res. 158:123–32.
8. Telner, J. I., and Singhal, R. L. 1984. Psychiatric progress: The learned helplessness
model of depression. J. Psychiatr. Res. 18:207–15.
9. Vollmayr, B., and Henn, F. A. 2001. Learned helplessness in the rat: Improvements in
validity and reliability. Brain Res. Protoc. 8:1–7.
10. Petit-Demouliere, B., Chenu, F., and Bourin, M. 2005. Forced swimming test in mice:
A review of antidepressant activity. Psychopharmacology (Berl.) 177:245–55.
11. Cryan, J. F., and Slattery, D. A. 2007. Animal models of mood disorders: Recent devel-
opments. Curr. Opin. Psychiatry 20:1–7.
12. Anisman, H., and Matheson, K. 2005. Stress, depression, and anhedonia: Caveats con-
cerning animal models. Neurosci. Biobehav. Rev. 29:525–46.
13. Willner, P. 2005. Chronic mild stress (CMS) revisited: Consistency and behavioural-
neurobiological concordance in the effects of CMS. Neuropsychobiology 52:90–110.

14. Song, C., and Leonard, B. E. 2005. The olfactory bulbectomised rat as a model of
depression. Neurosci. Biobehav. Rev. 29:627–47.
15. Papp, M., Willner, P., and Muscat, R. 1991. An animal model of anhedonia: Attenuation
of sucrose consumption and place preference conditioning by chronic unpredictable
mild stress. Psychopharmacology (Berl.) 104:255–59.
16. Papp, M. 2000. Models of affective illness: Chronic mild stress. In Current Protocols
in Pharmacology, eds. S. J. Enna and M. Williams. Chap. 5.9, 1-8. Hoboken, NJ: John
Wiley & Sons, Inc.
17. Porsolt, R. D. 1997. Historical perspective on CMS model. Psychopharmacology (Berl.)
134:363–64.
18. Nestler, E. J., Gould, E., Manji, H., et al. 2002. Preclinical models: Status of basic
research in depression. Biol. Psychiatry 52:503–28.
19. Slattery, D. A., Markou, A., and Cryan, J. F. 2007. Evaluation of reward processes in an
animal model of depression. Psychopharmacology (Berl.) 190:555–68.
20. Wang, D., Noda, Y., Tsunekawa, H., et al. 2007. Behavioural and neurochemical fea-
tures of olfactory bulbectomized rats resembling depression with comorbid anxiety.
Behav. Brain Res. 178:262–73.
21. Holmes, P. V. 2003. Rodent models of depression: Reexamining validity without
anthropomorphic inference. Crit. Rev. Neurobiol. 15:143–74.
22. Kelly, J. P., Wrynn, A. S., and Leonard, B. E. 1997. The olfactory bulbectomized rat as
a model of depression: An update. Pharmacol. Ther. 74:299–316.
23. Willner, P. 1997. Validity, reliability and utility of the chronic mild stress model of depres-
sion: A 10-year review and evaluation. Psychopharmacology (Berl.) 134:319–29.
24. Dürmüller, N., Scherschlicht, R., and Porsolt, R. D. 2000. Vigilance-controlled quanti-
fied EEG in safety pharmacology. In Current Protocols in Pharmacology, eds. S. J.
Enna and M. Williams. Chap. 10.6, 1-27. Hoboken, NJ: John Wiley & Sons, Inc.
25. Cheeta, S., Ruigt, G., van, Proosdijt, J., and Willner, P. 1997. Changes in sleep architec-
ture following chronic mild stress. Biol. Psychiatry 41:419–27.
26. Vanderwolf, C. H. 1992. The electrocorticogram in relation to physiology and behav-
ior: A new analysis. Electroencephalogr. Clin. Neurophysiol. 82:165–75.
27. Benstaali, C., Mailloux, A., Bogdan, A., Auzeby, A., and Touitou, Y. 2001. Circadian
rhythms of body temperature and motor activity in rodents their relationships with the
light-dark cycle. Life Sci. 68:2645–56.
28. Sanchez, C., Brennum, L. T., Storustovu, S. I., Kreilgard, M., and Mork, A. 2007.
Depression and poor sleep: The effect of monoaminergic antidepressants in a pre-clini-
cal model in rats. Pharmacol. Biochem. Behav. 86:468–76.
29. Overstreet, D. H., Friedman, E., Mathe, A. A., and Yadid, G. 2005. The Flinders sensi-
tive line rat: A selectively bred putative animal model of depression. Neurosci. Biobe-
hav. Rev. 29:739–59.
30. Seiden, L. S., Dahms, J. L., and Shaughnessy, R. A. 1985. Behavioral screen for antide-
pressants: The effects of drugs and electroconvulsive shock on performance under a dif-
ferential-reinforcement-of-low-rate schedule. Psychopharmacology (Berl.) 86:55–60.
31. O’Donnell, J. M., Marek, G. J., and Seiden, L. S. 2005. Antidepressant effects assessed
using behavior maintained under a differential-reinforcement-of-low-rate (DRL) oper-
ant schedule. Neurosci. Biobehav. Rev. 29:785–98.
32. Willner, P. 1991. Animal models as simulations of depression. Trends Pharmacol. Sci.
12:131–36.
33. Porsolt, R. D., and Lenegre, A. 1992. Behavioural models of depression. In Experimen-
tal Approaches to Anxiety and Depression, ed. J. M. Elliott, D. J. Heal, C. A. Marsden,
73–85. Chichester: Wiley.
34. Castagné, V., Porsolt, R. D., and Moser, P. 2006. Early behavioral screening for antide-
pressants and anxiolytics. Drug Dev. Res. 67:729–42.

35. Willner, P. 1991. Methods for assessing the validity of animal models of human psy-
chopathology. In Animal models in psychiatry, Ed. A. Boulton, G. Baker, M. Martin-
Iverson, Chap. 1, 1-23. Clifton, NJ: The Humana Press.
36. Porsolt, R. D., Le Pichon, M., and Jalfre, M. 1977. Depression: A new animal model
sensitive to antidepressant treatments. Nature 266:730–32.
37. Borsini, F., and Meli, A. 1988. Is the forced swimming test a suitable model for reveal-
ing antidepressant activity? Psychopharmacology (Berl.) 94:147–60.
38. Porsolt, R. D., Anton, G., Blavet, N., and Jalfre, M. 1978. Behavioural despair in rats:
A new model sensitive to antidepressant treatments. Eur. J. Pharmacol. 47:379–91.
39. Porsolt, R. D., Bertin, A., and Jalfre, M. 1977. Behavioral despair in mice: A primary
screening test for antidepressants. Arch. Int. Pharmacodyn. Ther. 229:327–36.
40. Steru, L., Chermat, R., Thierry, B., and Simon, P. 1985. The tail suspension test: A
new method for screening antidepressants in mice. Psychopharmacology (Berl.)
85:367–70.
41. Porsolt, R. D., Bertin, A., and Jalfre, M. 1978. “Behavioural despair” in rats and mice:
Strain differences and the effects of imipramine. Eur. J. Pharmacol. 51:291–94.
42. Lopez-Rubalcava, C., and Lucki, I. 2000. Strain differences in the behavioral effects
of antidepressant drugs in the rat forced swimming test. Neuropsychopharmacology
22:191–99.
43. Porsolt, R. D., Chermat, R., Lenegre, A., Avril, I., Janvier, S., and Steru, L. 1987. Use
of the automated tail suspension test for the primary screening of psychotropic agents.
Arch. Int. Pharmacodyn. Ther. 288:11–30.
44. Ripoll, N., David, D. J., Dailly, E., Hascoet, M., and Bourin, M. 2003. Antidepres-
sant-like effects in various mice strains in the tail suspension test. Behav. Brain Res.
143:193–200.
45. David, D. J., Renard, C. E., Jolliet, P., Hascoet, M., and Bourin, M. 2003. Antidepres-
sant-like effects in various mice strains in the forced swimming test. Psychopharma-
cology (Berl.) 166:373–82.
46. Steru, L., Chermat, R., Thierry, B., et al. 1987. The automated Tail Suspension Test:
A computerized device which differentiates psychotropic drugs. Prog. Neuropsycho-
pharmacol. Biol. Psychiatry 11:659–71.
47. Cryan, J. F., Mombereau, C., and Vassout, A. 2005. The tail suspension test as a model
for assessing antidepressant activity: Review of pharmacological and genetic studies in
mice. Neurosci. Biobehav. Rev. 29:571–625.
48. Katz, M. M., Tekell, J. L., Bowden, C. L., et al. 2004. Onset and early behavioral effects
of pharmacologically different antidepressants and placebo in depression. Neuropsy-
49. Porsolt, R. D., Brossard, G., Hautbois, C., and Roux, S. 2000. Models of affective ill-
ness: Forced swimming and tail suspension tests in rodents. In Current Protocols in
Pharmacology, eds. S. J. Enna and M. Williams. Chap. 5.8, 1-9. Hoboken, NJ: John
Wiley & Sons, Inc.
50. Porsolt, R. D., Brossard, G., Hautbois, C., and Roux, S. 2001. Rodent models of depres-
sion: Forced swimming and tail suspension behavioral despair tests in rats and mice. In
Current Protocols in Neuroscience, eds. S. J. Enna and M. Williams. Chap. 8.10, 1-10.
Hoboken, NJ: John Wiley & Sons, Inc.
51. Porsolt, R. D. 2000. Animal models of depression: Utility for transgenic research. Rev.
Neurosci. 11:53–58.
52. Stafford, R. S., MacDonald, E. A., and Finkelstein, S. N. 2001. National patterns of
medication treatment for depression, 1987 to 2001. Prim. Care Companion, J. Clin.
Psychiatry 3:232–35.
53. Slattery, D. A., Hudson, A. L., and Nutt, D. J. 2004. Invited review: The evolution of
antidepressant mechanisms. Fundam. Clin. Pharmacol. 18:1–21.

54. Walter, M. W. 2005. Monoamine reuptake inhibitors: Highlights of recent research

developments. Drug Dev. Res. 65:97–118.
55. Slattery, D. A., and Cryan, J. F. 2006. The role of GABAb receptors in depression and
antidepressant-related behavioural responses. Drug Dev. Res. 67:477–94.
56. Panconi, E., Roux, J., Altenbaumer, M., Hampe, S., and Porsolt, R. D. 1993. MK-801
and enantiomers: Potential antidepressants or false positives in classical screening
models? Pharmacol. Biochem. Behav. 46:15–20.
57. Suemaru, K., Yasuda, K., Cui, R., et al. 2006. Antidepressant-like action of nicotine in
forced swimming test and brain serotonin in mice. Physiol. Behav. 88:545–49.
58. Vazquez-Palacios, G., Bonilla-Jaime, H., and Velazquez-Moctezuma, J. 2004. Anti-
depressant-like effects of the acute and chronic administration of nicotine in the rat
forced swimming test and its interaction with fluoxetine. Pharmacol. Biochem. Behav.
78:165–69.
59. Porsolt, R. D., Chermat, R., Lenegre, A., Avril, I., Janvier, S., and Steru, L. 1987. Use
of the automated tail suspension test for the primary screening of psychotropic agents.
Arch. Int. Pharmacodyn. Ther. 288:11–30.
60. Detke, M. J., Rickels, M., and Lucki, I. 1995. Active behaviors in the rat forced swim-
ming test differentially produced by serotonergic and noradrenergic antidepressants.
61. Porsolt, R. D., Bertin, A., Blavet, N., Deniel, M., and Jalfre, M. 1979. Immobility
induced by forced swimming in rats: Effects of agents which modify central catechol-
amine and serotonin activity. Eur. J. Pharmacol. 57:201–10.
62. Detke, M. J., and Lucki, I. 1996. Detection of serotonergic and noradrenergic antide-
pressants in the rat forced swimming test: The effects of water depth. Behav. Brain Res.
73:43–46.
63. Cryan, J. F., Valentino, R. J., and Lucki, I. 2005. Assessing substrates underlying the
behavioral effects of antidepressants using the modified rat forced swimming test. Neu-
rosci. Biobehav. Rev. 29:547–69.
64. Kelliher, P., Kelly, J. P., Leonard, B. E., and Sanchez, C. 2003. Effects of acute and
chronic administration of selective monoamine re-uptake inhibitors in the rat forced
swim test. Psychoneuroendocrinology 28:332–47.

7 Assessing
in Rodents
Attention
Philip J. Bushnell and Barbara J. Strupp
CONTENTS
7.1 Introduction................................................................................................. 119
7.2 Multiple-Choice Serial Reaction Time Tasks ............................................. 120
7.2.1 Introduction...................................................................................... 120
7.2.2 Materials and Methods .................................................................... 122
7.2.3 Preparation of the Subjects .............................................................. 123
7.2.4 Training Steps .................................................................................. 123
7.2.5 Alternative Methods ........................................................................ 124
7.2.6 Testing Mice in the 5-CSRTT.......................................................... 124
7.2.7 Apparatus and Methodology for the 3-Choice Variant ................... 125
7.2.8 Data Analysis and Notes.................................................................. 125
7.3 Signal Detection Tasks with Blank Trials................................................... 128
7.3.1 Introduction...................................................................................... 128
7.3.2 Materials .......................................................................................... 129
7.3.4 Training Steps .................................................................................. 130
7.4 Attentional Set-Shifting .............................................................................. 131
7.4.1 Introduction...................................................................................... 131
7.4.2 Sand-Digging Task: Materials and Methods ................................... 134
7.4.4 Training Steps .................................................................................. 135
7.5 Selecting a Test Method .............................................................................. 136
Disclaimer.............................................................................................................. 137
References ............................................................................................................. 137
Appendix: Names and Addresses of Vendors Discussed in the Text .................... 143
7.1 INTRODUCTION
“Attention” refers to a variety of hypothetical constructs by which the nervous sys-
tem apprehends and organizes sensory input and generates coordinated behavior.
Although it has been a subject of psychological investigation since William James
introduced it to the field in the late 19th century, systematic assessment of atten-
tion in animals has a shorter history. As with any unobservable cognitive process,
119

assessment of attention requires quantification of an observable phenomenon, such

as the behavior of the animal or the electrical activity of its nervous system. To
the extent that these events can be measured objectively, attention can be inferred
equally readily in any animal species, including humans or other primates, rats,
mice, or birds.1
As James2 pointed out, attention is not a unitary phenomenon, but rather a term
that subsumes several different varieties of attentional processes. In the present dis-
cussion, we focus on three such processes: the ability to sustain attention over time,
the ability to attend selectively to a subset of environmental information while fil-
tering out extraneous stimuli, and the ability to shift attentional set. Accordingly,
this chapter discusses three behavioral approaches to assessing attention in rodents.
These approaches include multiple-choice serial reaction time tests that can be
arranged to assess both sustained and selective attention, signal detection tests with
blank trials that focus on sustained attention, and attentional set-shifting procedures.
For each of these approaches, we present a commonly used method and then discuss
design and analytic procedures that can help determine whether observed changes
in performance can be attributed to the target attentional construct (see Sections
7.2–7.4). Section 7.5 discusses some guidelines for task selection. The appendix lists
suppliers for necessary equipment.
7.2 MULTIPLE-CHOICE SERIAL REACTION TIME TASKS

7.2.1 INTRODUCTION
In 1983, Robbins and colleagues3 developed a test for assessing attention in rats
based on a test for human subjects that was originally ascribed to Leonard,4 and
is still in use.5 The rat method was called the “5-choice serial reaction time test”
(5-CSRTT) and has since been widely applied for exploring the neurobiology of
normal attentional processes and dysfunctions associated with disease states. In the
prototypical application, a rat or mouse faces five openings (or ports) in a horizontal
array along one curved wall of a test chamber, and a food cup with a clear plastic
door is located on the wall behind it. The animal initiates a trial by opening the
food cup door. After a short delay, a visual signal is presented, consisting of a brief
illumination of one of the five ports. If the animal then breaks a photobeam at the
opening of the port, a food pellet (or fixed volume of liquid reinforcer) is delivered
into the food cup.
A 3-choice variant of this type of task has been developed by Strupp and col-
leagues.6–11 This task is similar in concept to the 5-CSRTT and taps similar func-
tions but uses a slightly different apparatus. The most important difference is that the
rat is not required to turn around and obtain the reinforcer at the back wall; the pellet
dispenser delivers the reinforcer under the center response port.
In these tasks, the animal must maintain attention to the array of ports in order
to detect the signal and respond correctly. Accurate responding thus requires atten-
tion in both the temporal and spatial domains. In addition, because responses prior to
cue presentation (premature responses) are tallied as errors and terminate the trial,

Assessing Attention in Rodents 121
the task also places demands on inhibitory control, which permits inferences about
effects on impulsivity. The location, duration, and timing (pre-cue delay) of the
visual cue can be varied across trials, enabling independent assessments of sustained
attention as well as impulsivity. Varying the duration of cue illumination allows one
to parametrically manipulate the demands on sustained attention. Selective attention
may also be tapped by presenting distracting stimuli on some trials during the inter-
val between trial onset and cue presentation.
An impressive accumulation of studies over the past 25 years using the 5-
CSRTT has substantially increased understanding of the neural substrates underly-
ing sustained and selective attention, as well as inhibitory control.12,13 These studies
have generally used selective lesions or pharmacological manipulations of ascend-
ing monoaminergic systems. In general, accuracy of responding on the basic task
appears to depend upon cortical acetylcholine, and speed of responding is mediated
by mesolimbic dopamine. Auditory distractors are particularly disruptive to rats with
loss of ceruleocortical norepinephrine, and adequate forebrain serotonin appears to
be necessary to suppress premature responding. Further work with both methods
has illuminated conditions known—or suspected—to cause deficits in attention and
inhibitory control, such as attention deficit hyperactivity disorder,14 prenatal cocaine
exposure,7,11,15 and early childhood lead exposure.8,10
Research into the genetic underpinnings of attention has been facilitated by new
techniques to manipulate the mouse genome, which has stimulated the development
of behavioral methods for assessing attention in mice. Humby et al.16 first showed
that mice could be trained to perform the 5-CSRTT, and demonstrated the sensitiv-
ity of two mouse strains to parametric manipulations and the muscarinic cholinergic
antagonist scopolamine. Since that time, a number of studies have employed genetic
manipulations to examine the influence of affective states17,18 and neurochemical
pathways19–21 on sustained attention. This task has also proven to be a valuable tool
for studying murine models of genetic disorders in which attentional dysfunction is
prominent; examples include attention deficit hyperactivity disorder,20,22 fragile X
syndrome,23 and Down syndrome.24
In addition to its use in its original form for assessing visuo-spatial sustained
attention, variations on the method have been used for a number of interesting pur-
poses. For example, true “serial reaction time”—that is, the accuracy and speed
of responding to sequentially-presented stimuli—has also been modeled in rats
using illuminated nosepoke ports. This method focuses on the analysis of sequen-
tial behaviors per se, rather than of control of behavior via attention to temporally
unpredictable stimuli.25–27
To probe attention in terms of the Pearce–Hall model of attention,28 Holland’s
group29,30 modified the 5-CSRTT method to dissociate effects of the information
value of the cue, using continuous reinforcement for responses to two ports and
partial reinforcement for responding to two other ports in the five-choice apparatus.
Responses to the fifth port were never reinforced. Trials were paced by the experi-
menter, not the rat, to maintain an appropriate balance of trial types. Asymptotic
performance was more accurate to continuously reinforced ports, but new learning
(involving discriminative auditory cues) was more rapid to partially reinforced ports.
Lesion studies using this behavioral method showed that cholinergic projections

from the nucleus basalis magnocellularis to the amygdala central nucleus, medial
prefrontal cortex, and posterior parietal cortex support performance of the task.
7.2.2 MATERIALS AND METHODS

The following section describes the apparatus and methods commonly used for the
5-CSRTT for rats. Some modifications that pertain to the 3-choice variant for rats
(discussed above), as well as those for the mouse version of the 5-CSRTT are noted
briefly at the end of this section.
Subjects. Rats and mice, male and female, of several strains and varieties, can
learn to perform these multiple-choice serial reaction time tasks. The animal must
be mildly hungry at the time of testing, which can be arranged by a number of stan-
dard methods. 31 Methods are available to maintain adult rodents at a constant body
weight.32 See section 7.2.3, “Preparation of the Subjects,” below.
Apparatus. The 5-CSRTT requires equipment that was originally built in the
Laboratory of Experimental Psychology, Cambridge, UK.3 It has since become
commercially available from a number of manufacturers, including Campden
Instruments, Ltd.; Lafayette Instrument Co.; Med Associates, Inc.; PanLab, S.L.;
and TSE Systems. The test chambers for rats are roughly the size of a standard
operant conditioning chamber, with dimensions approximately 25 × 25 cm hori-
zontally, and roughly 30 cm in height. In place of response levers, a series of five
or nine openings, each about 2.5 × 2.5 cm in size, are arranged along a curved
rear wall of the chamber, about 2 cm from the floor. Each opening is bisected by
a photobeam, which is used to detect entry of the animal’s nose into the opening.
A light mounted inside each opening is illuminated briefly on each trial to serve
as a signal, and the animal is trained to poke its nose into the illuminated opening
to receive food, which is delivered into a cup mounted in the wall opposite to the
openings. The food cup must either be accessed via a hinged door with a micro-
switch, or bisected by a photobeam, so that nosepokes to retrieve the food can be
detected. A dispenser is needed to deliver food pellets or liquid food to the animal.
A house light is also needed for general illumination of the chamber. Mouse cham-
bers are designed the same way, but scaled down in size, with smaller openings
for nosepokes and food delivery. Both solid (dustless pellets) and liquid (diluted
condensed milk) reinforcers have been used.
A computer and interface for programming the stimulus events and recording the
animals’ responses are also necessary. Systems are available commercially for PCs.
Programming the procedures can be accomplished in a number of ways, including
state notation software for Windows-based systems, and several graphical-display-
based programming systems. Sources of these systems (detailed in the appendix)
include Campden Instruments, Ltd.; Lafayette Instrument Co.; Med Associates, Inc.;
PanLab, S.L.; and TSE Systems. A repository of open-code programs for MedState
notation software is available at http://www.mednr.com/.
Calibration devices should include a photometer for measuring the intensity of
the light under various stimulus conditions and a sound level meter for measuring the
intensity of the white noise and any auditory distractors that might be employed.

7.2.3 PREPARATION OF THE SUBJECTS

1. Ensure that the animals are motivated (hungry). One method is to deter-
mine their free-feeding body weight, and then reduce that weight by about
15%. Do not deprive the animals completely of food, but reduce their daily
allotment such that target body weights are achieved within 5–10 days. If
the animal is fully grown, this target weight can be maintained for the
remainder of the experiment. On the other hand, if the animal is still grow-
ing, allow it to grow in parallel with free-fed animals to a maximum level
(e.g., 350 g for an adult male Long-Evans rat and 250 g for an adult female
L-E rat).
2. Adapt the animals to the handling procedures31 and to the food pellets or
liquid reinforcers that will be used to reinforce responding in the chambers.
Commercially available precision 45-mg pellets are appropriate for adult
male rats; use 25- or 12-mg pellets for small rats or mice. This latter step
can be accomplished by offering the animals the new food each day in their
home cages or in a holding cage for several days prior to beginning train-
ing. This adaptation will obviate possible bait-shyness that may accompany
introduction of a novel food. The following procedure describes methods
for using food pellets.
7.2.4 TRAINING STEPS

1. Cover the response openings. Place the hungry animal in the chamber and
turn on the house light. Provide food pellets in the food cup. On the first day,
begin with 10–20 pellets in the cup, and allow the animal 30 min to explore
the chamber and collect the food. On the three following days, deliver the
pellets singly at 30-sec intervals and ensure that the animal retrieves the
pellets and consumes them.
2. Remove the covers from the openings. Turn on the house light and deliver
a single food pellet to start the session and begin the first trial when the
animal retrieves the pellet. A trial involves illuminating the signal light in
one opening (selected at random) after an inter-trial interval (ITI) of 2 sec
and recording the nosepokes made by the animal into the five openings.
When the animal pokes its nose into the illuminated opening, turn off the
signal light and deliver a single reinforcer. (An auditory cue may be help-
ful in informing the rat that food has been delivered, if the action of the
dispenser is quiet.) If the animal pokes its nose into a dark opening, turn
off the house light for a 2-sec timeout period and do not deliver food. Reset
the timeout each time the animal pokes its nose into a dark opening. Begin
a new trial when the animal pokes its nose into the food cup after either
timeout or food delivery. Present signals in each opening an equal number
of times in each session, and select at random the opening to illuminate on
each trial. Criterion: 100 reinforced nosepoke responses in a 30-min test
session.

3. Repeat step 2, but place a time limit on the signal light and the response
period (called a “limited hold”). Allow the animal to initiate each trial as
above, and use a 2-sec delay (pre-cue delay) before illuminating a signal.
Illuminate each opening for 60 sec and set a 60-sec limited hold after the
signal period. If the animal pokes its nose into the illuminated opening dur-
ing this 2-min period, deliver a food pellet and count a correct response. If
the animal pokes its nose into an unlit opening during this period, turn off
the signal light and house light and count an error of commission. If the ani-
mal fails to make a nosepoke response in this period, turn off the signal light
and the house light and count an error of omission. If the animal makes
a nosepoke into any opening during the pre-cue delay, turn off the house
light, count a premature response, and restart the same trial. Sessions for
rats commonly terminate after either 100 correct responses in a 30-min test
session, or a 100-trial session with 80% correct responding (see below).
4. Repeat step 3, progressively shortening the signal duration and limited
hold, ending with a signal duration of 0.5 sec and a limited hold of 5 sec.
Lengthen the pre-cue delay to 5 sec and the timeout period to 3 sec during
these steps. A stable baseline of about 80% correct responding with about
15% omissions should be achieved in about 30 training sessions.
5. After the basic rules have been learned, it is useful to vary the duration of
the pre-cue delay across trials within each session (e.g., 0, 3, 5, and 9 sec for
rats; 0, 2, and 4 sec for mice). Similarly it is useful to vary the duration of
the visual cue across trials within the session (as discussed below).
7.2.5 ALTERNATIVE METHODS

1. The apparatus may be modified so that the food is dispensed on the same
side of the apparatus as the response ports.17 This arrangement requires
more hardware but facilitates training. Another advantage of this setup,
as noted above for the 3-choice variant, is that the animal does not have to
turn around and traverse the chamber to obtain the reinforcer. The animal
can thereby maintain attentional focus on the response ports throughout the
testing session, rendering it more similar to human tests of attention.
2. Trials may be paced by the experimenter rather than the animal29,30,33,34
by starting a new trial at some fixed or variable time after the animal’s
choice response on the previous trial, rather than at the time that the animal
retrieves the reinforcer (as is done in the standard procedure). This proce-
dure gives the experimenter control over the timing of events in the test,
and removes any potential influence that the animal might exert on the trial
pacing by its speed of retrieving reinforcers.
7.2.6 TESTING MICE IN THE 5-CSRTT

DeBruin et al.35 provide a systematic, step-by-step approach to training mice in this
task, based on their previous work with rats.23

7.2.7 APPARATUS AND METHODOLOGY FOR THE 3-CHOICE VARIANT

The 3-choice serial reaction time task used by Strupp and colleagues is identical
in concept to the 5-CSRTT, but uses a 3-choice chamber originally developed by
Eichenbaum for olfactory discrimination tasks.37 Briefly, the apparatus is comprised
of a small testing alcove containing three response ports, separated from a larger
waiting area by a metal guillotine-type door that closes at the end of each trial, and
then opens to initiate the next trial following an inter-trial interval. Light emitting
diodes (LEDs), one positioned above each of the three response ports, provide the
target cue. A correct response (a nosepoke into the port under the illuminated LED)
is rewarded with a 45-mg Noyes pellet, delivered onto the floor of the chamber via
an opening beneath the center port. As noted above, this feature obviates the need
for the animal to turn around to retrieve the reward as required in the 5-CSRTT. This
feature of the 3-choice variant makes the task more similar to vigilance tasks used for
humans and nonhuman primates and, consequently, facilitates extrapolations from
the animal findings to the target human population.38 This task has provided impor-
tant information on the lasting cognitive and affective changes produced by prenatal
cocaine exposure,7,9,11,39–42 early postnatal lead exposure,8,10,43–47 prenatal or postna-
tal hyperphenylalaninemia (models of maternal and classic phenylketonuria),48,49 as
well as the attentional role of the ceruleocortical noradrenergic system.6
7.2.8 DATA ANALYSIS AND NOTES

Parametric Manipulation of Cue Characteristics: As noted above, schedule param-
eters may be manipulated to place greater or lesser challenges on various aspects
of attention as well as inhibitory control. Temporal challenges to attending include
lengthening and/or shortening the pre-cue delay, or making it variable across trials
rather than constant. In addition, shortening the duration of the signal is commonly
claimed to increase the “attentional load” of the task.50 In contrast, dimmer signals
have been used to challenge visual detection of the signals, a manipulation that has
been claimed to differ in nature from reducing its duration.51
By presenting olfactory or auditory distractors on some trials within a session,
one can assess selective attention. Systematically varying these parameters (e.g.,
duration of the pre-cue delay, cue duration, presence or absence of olfactory distrac-
tors) within a given testing session is often an effective means of gaining insight into
the integrity of specific functions, because information is then provided concerning
the particular conditions under which the subjects succeed and fail. This approach
can often effectively exclude alternative explanations for poor performance, and
thereby specify the nature of the impairment.
Evaluating Performance as a Function of the Outcome of the Prior Trial: These
multiple-choice reaction time tasks not only provide indices of various attentional
functions and inhibitory control, but they can also provide measures of arousal and/or
emotion. One available index of arousal and/or emotion within the context of perfor-
mance in these multiple-choice tasks is the animals’ reaction to committing an error.
Several dependent measures in these tasks have been found to vary significantly as
a function of the outcome of the previous trial. Specifically, on trials following an
error, the animals take longer to enter the testing alcove at trial onset, take longer

to make a response, and are more likely to commit all types of errors: premature
responses, inaccurate responses (responding after cue onset but to an incorrect port),
and omission errors (missing the cue). This pattern—increased response latency and
increased error rate on post-error trials—likely reflects an emotional response to the
error (for discussion, see23,24). Thus, the degree of disruption produced by commit-
ting an error provides a useful index of emotion or arousal. This type of analysis has
revealed functionally important deficits in rat models of early developmental expo-
sure to toxicants such as lead8 or cocaine,7,15 and in murine models of Down syn-
drome24 or fragile X syndrome.23 In some cases, such as the Down syndrome model,
the greater reactivity of the mutant mice to committing an error became apparent
only as a result of coding videotapes of the mice performing the task (see24).
Different Types of Errors: Several types of errors are possible in these tasks, the
delineation of which can shed light on the nature of group differences. Nosepokes
into the ports prior to cue presentation (premature responses) terminate the trial
and are tallied as errors. The percentage of such responses can provide an index of
impulsivity or inhibitory control. Trials on which the animal initiates the trial but
then does not make a nosepoke into one of the response ports within a specified
time after trial onset, scored as omission errors, suggest that the animal missed the
cue due to impairment of sustained attention. Inaccurate responses (responses made
after cue onset but to a port that had not been illuminated) are also indicative of
lapses in attention. The most basic measure of accuracy is the ratio of the number
of correct responses divided by the total number of trials. Another useful measure
is to calculate the accuracy of the animal given a response at the correct time; this
measure is calculated as the number of correct responses divided by the number of
“timely” responses (trials on which the animal responded within the limited hold,
i.e., excluding premature responses and omission errors).
Clues regarding the nature of the dysfunction are also often provided by catego-
rizing the types of errors committed, and then evaluating each error type as a func-
tion of these various parameters (delay before cue onset, cue duration, trial block
[portion of session]), as well as the outcome of the previous trial, as discussed above.
For example, in this visual attention task, we found that adult male rats exposed to
cocaine in utero committed more omission errors than controls only on trials in the
final third of the testing session that occurred after an error.9,15 These animals were
not impaired in this final portion of the session on trials that followed a correct
response, or earlier in the session, regardless of prior trial outcome. This pattern
implicates the additive effects of impairments in two areas: sustained attention and
emotion regulation.
Use of Distractors to Assess Selective Attention: The task may be modified to
assess selective attention by presenting irrelevant auditory3 or olfactory stimuli7,8,23
during the interval between trial onset and cue presentation while the animal is
waiting for the cue. These distracting stimuli lead to an increase in premature and
inaccurate responses. It is best to present the distractors on a minority of the trials in
a session, so that they are surprising, and therefore maximally disruptive. This pro-
cedure also minimizes habituation to the distracting stimuli. Interestingly, in studies
with olfactory distractors, the distractors seem to produce the greatest disruption in

performance when presented 1 sec following trial onset, regardless of whether the
cue is presented after a 2 or 3 sec delay.7
Two different indices are useful for assessing the effect of the distractors. First,
the effect of the manipulation of interest (e.g., lesion, drug treatment, genetic manip-
ulation) on selective attention can be assessed by comparing performance on tri-
als with distractors (distraction trials) to performance on trials without distractors
(non-distraction trials). However, the distractors may disrupt performance on the
non-distraction trials as well as the distraction trials, due to heightened arousal or
emotion. To ascertain whether the manipulation of interest alters this putative effect
(which may be thought of in terms of emotion or arousal regulation), performance
on the non-distraction trials of the distraction task can be compared to performance
on a baseline task that is identical in terms of light cue presentation parameters
but does not include distractors. Interestingly, which of these two measures will
be more sensitive in any given case depends on the nature of the dysfunction seen
in the experimental group. For example, in a mouse model of fragile X syndrome,
which in humans is characterized by impairments in attention and arousal regula-
tion, the mutant mice differed from controls in terms of the generalized disruption
produced by the distractors; performance on the distraction trials did not differ
between groups.23
Latency Measures: Several latency measures are also informative. Response
latency (the time between onset of the signal and a correct response) on correct trials
provides a measure of information processing speed. Food retrieval latency (the time
between delivery of a food pellet and the animal’s entry into the food cup) provides
a measure of motivation. Similarly, in task variants in which trial onset is indicated
by the opening of a door at the dipper alcove (e.g., see23,24), the latency to respond to
the dipper alcove after the door is raised provides another index of motivation, and
also an index of the emotional reaction to the outcome (correct or incorrect) of the
prior trial. All of these latency measures may, however, be influenced by changes in
motoric function. Therefore, it is important to determine whether all of these latency
measures are altered or only certain ones. For example, if correct response latency
is slowed but alcove latency and dipper latency are not altered, the most parsimoni-
ous interpretation is that information processing speed is slowed; the fact that alcove
latency and dipper latency are normal allows one to exclude an impairment of motor
function.
Varying the Probability of Reinforcement: The probability of reinforcement for
correct responses has been manipulated as a way to control the predictive validity
of selected cue-port stimulus complexes to test the associability of these cues with
new learning.29 Strupp and colleagues have also used periodic reward omission as a
means of assessing reaction to non-reward (e.g., emotion or affect regulation) in both
rats and mice using a similar task.47

7.3 SIGNAL DETECTION TASKS WITH BLANK TRIALS

7.3.1 INTRODUCTION
The ability of human subjects to report the occurrence of rare and unpredictable
signal events over prolonged periods of time has been extensively characterized.52,53
Accurate detection of such signals is assumed to depend upon maintaining attention
to the task over time, and many of the factors that affect performance of humans on
these tasks have been systematized. Sustained attention tasks comprise an impor-
tant and sensitive component of neurobehavioral test batteries used for assessing the
effects of drugs in humans, e.g., benzodiazepines,54 stimulants,55 and ethanol.56
A major problem for tests of sustained attention involves quantifying and mini-
mizing the false alarm rate. That is, a subject can successfully report many signals
simply by responding frequently—though doing so will generate a large number
of erroneous reports that a signal had occurred (false alarms). Human subjects
can be instructed not to respond in this manner and will normally withhold most
false alarms; animals can be trained to do so as well. However, manipulations that
increase or decrease overall “responsivity” or response rate are difficult to interpret
if no independent measure of the false alarm rate is obtained.
Better estimates of the false alarm rate can be obtained by counting responses
to specified non-signal events (blank trials). This approach has been used with both
fixed and retractable response levers. The task described below employs a discrete-
trial, two-lever approach that requires rats to report the occurrence or nonoccurrence
of a single, brief, centrally located signal. Thus, two retractable levers are inserted
into the test chamber after a variable period of time to “ask” the rat to report whether
a brief signal was presented during that period. If a signal was presented (“signal
trial”), a press on one lever produces food and a press on the other lever produces
a short timeout period without food. If no signal has occurred (“blank trial”), the
converse contingencies apply. Because the levers are retracted between trials, no
presses can occur during the inter-trial interval (ITI). Because an explicit response
is required on each trial, the proportions of hits and false alarms (P[hit] and P[fa])
can be calculated in relation to the total number of completed signal and blank trials,
respectively.
Validation of this method includes both studies of the effects of parameters
known to affect human sustained attention, and pharmacological and neurobiological
manipulations. Parametric studies include observations that signal intensity, signal
rate, and the type of task all affect response accuracy,57–59 as predicted from stud-
ies of vigilance in humans.52 Thus the parameters that affect the behavior of rats in
this task closely parallel those that affect sustained attention in humans. In addition,
three of the variables that control the behavior of rats in this task (signal intensity,
trial presentation rate, and whether detection of a single stimulus or discrimination
between two stimulus classes is required) have been shown experimentally to control
the behavior of humans in this task.
Pharmacological studies have shown dose-related impairment of signal detection
in this task after a variety of nicotinic drugs,59,61–68 d-amphetamine,59 the muscarinic

drugs pilocarpine and scopolamine, and the F2-adrenergic compounds clonidine and
idazoxan.65 Further, the influence of cholinergic projections from the basal forebrain
to the cortical mantle in sustained attention has been described in a series of elegant
studies.69–74 This work has led to advances in understanding the neurobiology of sus-
tained attention75,76 and hypotheses regarding the role of attention in addictive behav-
ior.63 The method has also been used to characterize the acute effects of organic
solvents58,78–80 and other neurotoxic chemicals.81–83
This method has also been enhanced by systematic manipulation of the post-
signal interval to engage working memory as well as attention.84 Using this hybrid
task, the effects of scopolamine and mecamylamine, drugs often presumed to impair
working memory, were shown to affect attention. Martin et al.85 trained wild type
and lurcher mice to perform this task, and determined that the effects on perfor-
mance in the lurchers were due to motoric rather than attentional deficits.
7.3.2 MATERIALS
Subjects. Rats and mice can perform the task. See section 7.2.3, “Preparation of the
Subjects” above.
Apparatus. Assemble one or more standard operant conditioning chambers
equipped at minimum with a signal light, a food cup and food pellet dispenser, and
two retractable response levers. A loudspeaker for presentation of masking noise
may also be used. The two retractable levers should be mounted on either side of
the food cup. Mount the signal light immediately above one of the levers at the
start of training, and later move it to the top center of the wall above the food cup
when the rat has learned the response rule required for the task. This equipment
can be purchased from one of the vendors of behavioral test systems listed in the
appendix.
Assign the lever below the signal lamp as the “signal” lever and the other lever
as the “blank” lever. Set up half of the chambers with the signal lever on the left
and the other half with it on the right. Counterbalance all treatments for signal lever
position.
A computer and interface for programming the stimulus events and recording
the animals’ responses are also necessary. Commercially available hardware and
software systems are available, as described above for the 5-CSRTT.
Calibration devices should include a photometer for measuring the intensity of
the light under various stimulus conditions and a sound level meter for measuring the
intensity of the white noise.

Same as above for the 5-CSRTT.


1. Shape the rats to press the signal lever for food by autoshaping,86,87 by
long (e.g., overnight) sessions with a continuous reinforcement schedule in
effect, or by hand shaping. If an autoshaping procedure is used, turn on the
signal lamp whenever the lever is extended into the chamber, and turn it off
when the lever retracts (either when the rat presses it, or after 15 sec without
a press). If overnight sessions are used, be sure to provide adequate water.
Criterion: one session of 50 reinforced responses on this lever.
2. Shape the rats to press the other (“blank”) lever by the same means. How-
ever, do not turn on the signal lamp when shaping responses on this lever.
Criterion: one session of 50 reinforced responses on this lever.
3. Begin training using trials in which both levers are extended into the cham-
ber simultaneously on each trial. Light the signal lamp in half the trials
(“signal” trials) and not in the other half (“blank” trials). Retract both levers
as soon as one is pressed. Deliver a food pellet after a press on the signal
lever in a signal trial and after a press on the blank lever in a blank trial.
Turn off all lights for 3 sec after a press on the signal lever in a blank trial
and after a press on the blank lever in a signal trial. In signal trials, turn on
the signal light 2 sec before extending the levers, and leave it on until the
rat presses a lever. Use correction trials to reduce the likelihood of position
habits: repeat the conditions presented in each trial that terminates in an
incorrect response, up to a maximum of three such correction trials. If the
rat makes three consecutive errors, extend only the other (correct) lever in
the fourth trial to force a correct response. Criterion: two 100-trial sessions
with overall accuracy of 80% or better.
4. Remove the correction trials and increase the total number of trials to 120.
Criterion: one 120-trial session with overall accuracy of 80% or better.
5. Turn off the signal when the levers extend (rather than when the rat makes
a response) and increase the total number of trials to 150. Criterion: one
150-trial session with overall accuracy of 80% or better.
6. Reduce the duration of the signal from 2 sec to 0.3 sec in gradual steps (e.g.,
1.5 sec, 1.0 sec, 0.7 sec, 0.5 sec, and 0.3 sec). The onset of the signal should
occur 2 sec before insertion of the levers in all cases, leaving an empty
period between offset of the signal and insertion of the levers. Increase the
total number of trials in stages to 240. Criterion: one session with overall
accuracy of 80% or better at each signal duration.
7. Move the signal lamp from its position above the signal lever to a position
at the top of the panel, centered between the levers (above the food cup).
Retrain to criterion accuracy.
8. Make the interval after the signal offset variable (select values of 2, 3, or 4
sec randomly on each trial). Maintain accuracy at the 80% criterion.
9. Make the interval before the signal onset variable. (Begin with a list of rela-
tively short and homogeneous values, and work up to a list of values ranging
from less than 1 sec to about 25 sec, selected randomly on each trial. A con-

stant-probability list, such as that provided by Fleschler and Hoffman88 is

recommended for the final stage.) Maintain accuracy at the 80% criterion.
10. Vary the strength of the signal. Either the intensity or duration may be var-
ied. Varying the intensity is preferred, but requires digital control of the
voltage provided to the signal lamp (through a digital-to-analog converter).
At least three signal strengths should be used, preferably more (up to seven).
Administer at least 20 trials at each signal strength (10 signal and 10 blank).
Maintain accuracy at the 80% criterion.

The proportion of correct detections of the signal (P[hit]) should increase with
increasing signal strength. The signal strength should be adjusted so that the weak-
est signal produces a P(hit) about equal to the guessing rate, and the strongest signal
produces a P(hit) of about 1.0. The guessing rate is given by the proportion of errors
on blank trials, or false alarms (P[fa]). P(fa) should be independent of signal strength
and range from about 0.10–0.20.
A wide range of signal strength values improves the consistency of the baseline
from day to day, and allows one to differentiate between changes in attention and
visual function.65 That is, poor attending to the signals should cause an increase in
P(fa) and a decrease in P(hit) at all signal strengths where P(hit) exceeds P(fa). In
other words, the P(hit) by signal strength gradient should shift downward. In con-
trast, a change in the ability of the rat to see the signal should produce a horizontal
shift in the P(hit) by signal strength gradient, so that P(hit) is altered only for signals
of intermediate intensity; in addition, P(fa) should not change.
P(hit) and P(fa) can be used to calculate signal detection indices of sensitivity
and bias by any of a number of methods.88–91 However, interpretation of these derived
measures depends upon the particular assumptions upon which their calculation is
based, and explanation of their meaning invariably requires reference to the values
of P(hit) and P(fa) from which they were derived. Thus the advantages of deriving
signal detection indices—which involves trading one pair of measures (P[hit] and
P[fa]) for another pair of more derived measures (sensitivity and bias)—are generally
outweighed by the effort required to calculate and explain these derived measures.
Response time may also be measured as the latency between insertion of the
lever and the rat’s response. This variable provides an index of motor function simi-
lar to a simple reaction time, because rats typically choose which lever to press
during the time interval after the signal by positioning themselves in front of one of
the levers and pressing it during its insertion into the chamber. Response time typi-
cally does not vary with signal intensity, but does tend to be shorter for hits and false
alarms than for misses and correct rejections.58
7.4 ATTENTIONAL SET-SHIFTING

7.4.1 INTRODUCTION
Another aspect of attentional function frequently assessed in human neuropsycholog-
ical testing batteries is attentional set-shifting, commonly indexed by the Wisconsin

Card Sorting Task (WCST), the primary clinical index for frontal lobe dysfunction.
This function can also be tested by the extradimensional shift (EDS) task which is
part of the Cambridge Neuropsychological Test Automated Battery (CANTAB), a
testing battery originally developed for the assessment of cognitive function in elderly
and dementing patients,93 but now also widely used to test patients with Alzheimer’s
disease and other forms of dementia, basal ganglia disorders including Parkinson’s
disease, Korsakoff syndrome, depression, and schizophrenia, as well as children
with learning difficulties or autism (see94). Notably, versions of the EDS paradigm
have been developed for nonhuman primates and rodents (described below).
In this paradigm, which includes a series of tasks, the subject is first trained
to respond to one stimulus dimension (e.g., odor) of a multidimensional compound
stimulus and is then required to respond instead to a previously irrelevant dimen-
sion (e.g., texture). This shift from one stimulus dimension to another defines an
EDS. Insight into the nature of the dysfunction is provided by comparing the rate
of mastering the EDS to the rate of mastering an intradimensional shift (IDS), in
which two novel stimuli are presented, but the predictive dimension is the same
as in the original discrimination. If the subject has formed an attentional set, the
mastery of the IDS is more rapid than for the original discrimination, and mastery
of the EDS is slower than for the IDS. The EDS phase requires cognitive flexibility
(to shift attention from the previously predictive dimension to the newly predictive
dimension), and associative ability (to figure out the new contingencies), as well as
selective attention (to attend selectively to the new predictive dimension while ignor-
ing the previously predictive dimension). In a typical study, reversals of the correct
and incorrect cues within a dimension are also commonly introduced following both
the IDS and the EDS, to determine the extent to which behavior is controlled by the
dimension as opposed to the specific exemplars of the dimension. (Further discus-
sion of methodological issues can be found in95,96.)
As discussed by Chudasama and Robbins,94 the ED/ID set-shifting test can serve
several functions. First, it provides a sensitive index of frontal lobe dysfunction,
based not only on recent empirical evidence from lesion studies,96,97 but also on the
fact that it taps the primary function required for successful performance on the
WCST, the primary clinical index for frontal lobe dysfunction.92 Second, it allows
one to distinguish between two levels of cognitive flexibility: perseveration to a spe-
cific exemplar (tapped by the reversal learning task in this series) versus inflexibility
with respect to shifting attention from one perceptual domain to another (i.e., atten-
tional set-shifting). An attractive feature of this task series is that parallel versions
have been devised for testing monkeys,98 rats (e.g., see41,45,96), and mice,99 thereby
providing an opportunity to integrate clinical findings with human subjects (e.g.,
from the CANTAB) with information concerning the neural and neurochemical sys-
tems that underlie specific aspects of task performance.
Both operant and sand-digging versions of the set-shifting paradigm have been
described for rodents. In the following sections, we discuss these two versions and
outline the key advantages and disadvantages of each, to aid the reader in deciding
which task would be preferable for achieving specific goals while considering tem-
poral and fiscal constraints. We describe the sand-digging version of the paradigm

in preference to the operant method because of its relative efficiency in time and
equipment.
The sand-digging EDS method developed by Brown and colleagues96 consists of
a series of seven tasks, including a simple discrimination,50 a compound discrimi-
nation, a reversal of the CD (R1), an IDS, and an EDS, and a reversal of the EDS
(R3). Two sessions are required: an initial training session, in which the animals
are trained to dig for bits of food hidden in bowls of digging medium, followed by a
second test session in which seven discrimination, reversal, and shift tasks are given
in sequence. The task is detailed below.
There are several attractive features of this task. First, it does not require expen-
sive operant equipment and can be set up quickly. Second, the entire series of seven
tasks can be completed in a single session, representing a considerable savings in
time relative to the operant EDS task version. Third, this task series includes novel
stimuli at each stage, i.e., a “total change design.” This feature, which aids in inter-
pretation of results (discussed in96,97) is more difficult to implement in operant set-
ups. Finally, the task has been validated as an index of frontal lobe dysfunction for
rats based on the fact that, in both rats and primates, medial prefrontal lesions impair
EDS but not reversal learning, and orbitofrontal cortex lesions impair reversal learn-
ing but not EDS learning.96,97,100
Despite these assets, there are two drawbacks to the sand-digging EDS task rela-
tive to the operant version. First, it cannot be automated because it requires hands-on,
trial-by-trial administration by an experimenter. Second, as presently configured, it
does not enable one to determine the basis of an observed alteration in EDS learn-
ing rate. That is, if the EDS task is the only task in the series that is impaired by
the experimental manipulation (i.e., no group differences are observed in original
learning, IDS, or reversals), it is possible to conclude that the manipulation being
tested (e.g., a lesion or a drug) specifically impaired the rate at which the rat mas-
tered an EDS, but the nature of the cognitive change remains ambiguous. This is
because there are at least two possible reasons for a selective EDS impairment: (1) an
impaired ability to shift attention from the previously predictive cues (inflexibility
with regard to attentional set); and (2) impaired selective attention (an inability to
filter out the previously predictive cues.
In contrast, because acquisition of the operant EDS task is more prolonged, it is
possible to demarcate different phases of learning based on the subject’s patterns of
responding. These phases include a perseverative phase, characterized by repetitive
responding to the previously correct stimulus; a subsequent chance phase, reflecting
trial-and-error responding, with inconsistent patterns of correct responding; and a
post-chance phase, in which accuracy exceeds chance and increases steadily toward
criterion. Changes in the durations of the specific phases can shed light on the nature
of the impairment. For example, an experimental group that exhibits a significant
lengthening of the initial perseverative phase, with later learning phases of normal
length, is likely to suffer from cognitive inflexibility. In contrast, a group that exhibits
an elongated post-chance phase, combined with normal IDS performance, is likely
to be impaired in selective attention, unable to focus selectively on the new predic-
tive dimension as a result of being distracted by the previously predictive cues.

Previous research involving a rodent model of prenatal cocaine exposure illus-

trates how phase analysis of this type of task series can shed light on the integrity
of these specific cognitive functions. In this study, the animals were administered
a series of olfactory discrimination and reversal tasks followed by two EDS tasks.41
The first and third EDS tasks required a shift from the olfactory to the spatial dimen-
sion; the second EDS task required a shift from spatial to olfactory cues. Analyses
of learning rate (errors to criterion) demonstrated that the cocaine-exposed (COC)
animals were significantly impaired in the two spatial EDS tasks, but not in the
olfactory EDS task. The fact that the COC animals exhibited slower learning than
controls only late in the task (after the subjects had made eight consecutive correct
responses) suggests that the deficiency was not related to being inflexible in shifting
attentional set, as they did not show perseverative responding to the dimension that
was previously predictive. Rather, the increased error rate (relative to controls) seen
later in the task suggests impairment of selective attention (i.e., difficulty ignoring
the irrelevant cues).
7.4.2 SAND-DIGGING TASK: MATERIALS AND METHODS

Below we describe the sand-digging EDS task series developed for96 and also adapted
for mice.99 This method entails easily mastered tasks that do not require expensive
equipment, and therefore may be accessible to investigators on a tight budget in
terms of time or money. However, the interpretive limitations associated with the
rapid learning, in terms of being able to identify the specific nature of the dysfunc-
tion, should also be kept in mind.
Subjects. Rats and mice can perform the task. The animals must be hungry;
methods for maintaining appetitive motivation are discussed above. The method
described here is designed for rats and can be scaled down for mice.
Apparatus. A large plastic rodent cage (e.g., 40 × 70 × 18 cm) with clear plastic
dividers and a set of digging bowls (e.g., ceramic pots, 7 cm diameter) are needed.
The cage is divided along its long axis for one-third of its length with a permanent
vertical plastic panel, and a single digging bowl is placed on either side of this panel.
This panel serves to prevent rapid movement of the animal from one bowl to the
other. A second, removable panel is placed across the short axis of the cage blocking
entrance to the divided area, and serves to prevent the animal from starting the trial
prematurely.
Digging bowls are covered on the sides and rim with various materials that differ
in texture (e.g., sandpaper, cloth, or wax paper) and are filled with digging medium
of various textures (e.g., sawdust, sand, or tea leaves). The media are also scented
(e.g., with cumin, cinnamon, or cloves). Six exemplars of each stimulus dimension
(texture, medium, and odor) are necessary for the complete design.

Rats require habituation to the apparatus and to digging in bowls for food (e.g., a
small piece of sweetened cereal as bait). During a single 60-min session, the rat
first learns to dig for bait, which is replaced in the bowl every 5 min as the animal
retrieves it. Next, the rat is given three simple discrimination tasks, in which bait is

placed in one of two bowls, which differ along one of the three stimulus dimensions
(texture, odor, or medium). The exemplars used in this session are not used further.
Each rat is trained to a criterion of six consecutive correct choices, where a choice is
defined as the first bowl that the rat digs in.

In the second session, the rats are given a series of seven discriminations using three
different pairs of stimulus exemplars of each dimension: a simple discrimination
(SD), a compound discrimination (CD), a reversal of the CD (R1), an IDS, a second
reversal (R2), an EDS, and a reversal of the EDS (R3). Each rat receives the tests
in the same order, and is tested with one relevant stimulus dimension and one irrel-
evant dimension. In the first five tests, the relevant dimension is always the same; it
changes at test 6, the EDS.
In each test, the bowls are baited and placed on either side of the permanent bar-
rier at the end of the cage. The removable barrier is set in place, and the rat is placed
in the test box on the side of the removable barrier opposite the bowls. A trial begins
with removal of the barrier and ends when the rat obtains the bait or makes an error.
The first four trials of each test are “discovery trials” in which the rat is permitted
to dig in each bowl to find the bait. Beginning on the fifth trial, the rat is permitted
to eat the bait if it chooses the baited bowl first. If it chooses the unbaited bowl first,
an error is scored and the trial is terminated. The rat is trained to a criterion of six
consecutive correct choices.

The stimulus dimensions used and the correct and incorrect exemplars must be
distributed systematically across individual subjects so that, when averaged across
animals, learning scores are not biased by the relative difficulty of the specific dis-
criminations used in each test. In contrast, treatment groups should be matched for
stimulus conditions at each stage of testing, to prevent confounding of differences in
stimulus discriminability or dimensional salience with the treatment. See Birrell and
Brown96 for further details of the design.
For control animals, the rate of learning (errors or trials to criterion) should
be faster for the ID than for the ED. Note that the rat is presented with two
multidimensional stimuli in both of these tasks; if the animal had not formed an
attentional set (predisposing it to focus on one perceptual dimension over others),
then these two tasks would be solved at exactly the same rate and, indeed in this
method, a particular set of cues must be used for the ED for some animals and the
ID for others, ruling out the possibility that a particular set of cues is easier to master
than others.
The inclusion of the five task types in the series (simple and compound discrimi-
nation, IDS, EDS, and reversals) aids in interpreting the nature of observed group
differences. For example, if the simple and compound discrimination tasks do not
reveal group differences, but the experimental group is impaired on the IDS, the
reversals, and/or the EDS tasks, then one can exclude the possibility that the observed
impairment on one or more of these latter tasks is a result of decreased motivation,

impaired ability to perform the motoric demands of the task, or impaired sensory
acuity. If the experimental group is unimpaired on all tasks except for the reversals
(as seen following lesions of the orbitofrontal cortex100), then one can conclude that
the experimental group has a specific deficit in the ability to inhibit responses to a
previously rewarded object within a given perceptual domain. Finally, if the deficit is
limited to the EDS phase, one can conclude that the manipulation specifically altered
the ability to shift attentional set; the intact IDS phase allows one to rule out forma-
tion of the attentional set as the locus of the impairment. However, as noted above, an
impaired ability to shift attentional set could be due to either inflexibility in shifting
attentional set or impaired selective attention.
7.5 SELECTING A TEST METHOD

The several methods described here represent commonly used approaches to assess-
ing attention in rodents. Others are also available; a more comprehensive treatment of
them was compiled by Bushnell,1 and surely more have been devised in the interim.
The choice of method should ultimately depend on the question being addressed,
although issues of practicality (e.g., availability of equipment and time for train-
ing) will inevitably affect the choice. Guidance regarding the variety of attention to
be assessed should be taken from literature on the disease state or phenomenon of
interest. A first cut might be to decide whether the aspect of attention most likely
to be affected is selective attention, sustained attention, or attentional set-shifting.
Sustained attention is better assessed by the serial reaction time tasks and discrete-
trial signal detection methods described in sections 7.2 and 7.3. Selective attention
is best tapped by serial reaction time tasks with periodic presentation of distracters.
EDS methods provide an index of attentional set-shifting and, indirectly, also a mea-
sure of selective attention; that is, impaired selective attention would be expected to
impair EDSs but slower EDS mastery does not necessarily indicate impaired selec-
tive attention, as discussed above. The serial reaction time tasks also provide an
index of impulsivity or inhibitory control, an area of dysfunction that is prominent
in attention deficit hyperactivity disorder, and therefore is also of interest in many
assessments of attentional function.
Sustained attention can involve both temporal and spatial components. The 5-
CSRTT combines the two, and can be arranged to focus on one or the other of these
dimensions. For example, the spatial distribution of the stimuli can be widened by
using only the peripheral ports, or narrowed by using only the central ones. The
temporal uncertainty and the duration of the cue are routinely varied to manipulate
the “attentional load” placed on the animal. One attractive feature of this task is that
sustained and selective attention as well as inhibitory control can all be indepen-
dently assessed in the same task. A drawback of the standard method is that trials
are initiated by the animal, so that control over the pacing of the session is under the
animal’s control. However, this aspect is easily modified, and has been placed under
the control of the experimenter in both the 5-CSRTT23,24 and the 3-choice variant
(e.g., see6,39,45,48).
The discrete-trial signal detection method removes the spatial component and
focuses only on the temporal dimension. Trial pacing is determined by the program-

ming, so that the experimenter retains control over the entire test session. In addi-
tion, if the intensity of the signal can be manipulated independently of its duration
and timing, then the method can be used in a psychophysical manner to determine
changes in threshold for detecting increments in stimulus intensity as a check for
visual dysfunction (in contrast to attentional problems).
EDS tasks (both operant and sand-digging) provide an index of attentional set-
shifting, an aspect of attention frequently assessed in human neuropsychological
testing batteries, and a classic index of frontal lobe functioning. An attractive feature
of this task series is that virtually identical tasks have been developed for humans,
nonhuman primates, and rodents, facilitating cross-species extrapolation of findings.
An additional advantage of the sand-digging version of this task is that it does not
require expensive equipment and can be administered in a few days. As noted above,
this task series provides an assessment of two types of flexibility, as well as an indi-
rect index of selective attention.
The database of literature should also be considered when selecting a test. The 5-
CSRTT has been more widely used in the study of attention in rodents than any other
task; early work with it focused on the neurobiological pathways mediating behavior
in the test,12 whereas more recent work has included the psychopharmacology of
systemic treatments as well.13 The discrete-trial signal detection task has also been
used to evaluate CNS pathways involved in attention,76 with a focus on questions of
drug addiction and mental disorders.77 It has also been used extensively to study the
acute effects of psychoactive drugs65,101 and volatile organic solvents.78
DISCLAIMER
This manuscript has been reviewed by the National Health and Environmental
Effects Research Laboratory, U.S. Environmental Protection Agency and approved
for publication. Approval does not signify that the contents necessarily reflect the
policies of the Agency nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
REFERENCES
1. Bushnell, P.J., Behavioral approaches to the assessment of attention in animals. Psy-
chopharmacology (Berl), 1998. 138(3-4): p. 231-59.
2. James, W., The Principles of Psychology. 1950, New York: Dover Publications Inc.
3. Carli, M., et al., Effects of lesions to ascending noradrenergic neurones on performance
of a 5-choice serial reaction task in rats; implications for theories of dorsal noradren-
ergic bundle function based on selective attention and arousal. Behav Brain Res, 1983.
9(3): p. 361-80.
4. Wilkinson, R., Interaction of noise with knowledge of results and sleep deprivation.
Journal of Experimental Psychology, 1963. 66: p. 332-337.
5. Sahakian, B., et al., Further analysis of the cognitive effects of tetrahydroaminoacri-
dine (THA) in Alzheimer’s disease: Assessment of attentional and mnemonic function
using CANTAB. Psychopharmacology (Berl), 1993. 110: p. 395-401.

6. Bunsey, M.D. and B.J. Strupp, Specific effects of idazoxan in a distraction task: evi-
dence that endogenous norepinephrine plays a role in selective attention in rats. Behav
Neurosci, 1995. 109(5): p. 903-11.
7. Gendle, M.H., et al., Enduring effects of prenatal cocaine exposure on selective atten-
tion and reactivity to errors: evidence from an animal model. Behav Neurosci, 2004.
118(2): p. 290-7.
8. Stangle, D.E., et al., Succimer chelation improves learning, attention, and arousal regu-
lation in lead-exposed rats but produces lasting cognitive impairment in the absence of
lead exposure. Environ Health Perspect, 2007. 115(2): p. 201-209.
9. Gendle, M.H., et al., Alterations in reactivity to errors and sustained attention in an
animal model of prenatal cocaine exposure (Special Edition on Drugs of Abuse, invited
contribution). Developmental Brain Research 2003. 147: p. 85-96.
10. Morgan, R.E., et al., Early lead exposure produces lasting changes in sustained atten-
tion, response initiation, and reactivity to errors. Neurotoxicol Teratol, 2001. 23(6): p.
519-31.
11. Morgan, R.E., et al., Enduring effects of prenatal cocaine exposure on attention and
reaction to errors. Behav Neurosci, 2002. 116(4): p. 624-33.
12. Robbins, T.W. and B.J. Everitt, Arousal systems and attention, in The Cognitive Neuro-
sciences, M.S. Gazzaniga, Editor. 1995, MIT Press: Cambridge, MA. p. 703-720.
13. Robbins, T.W., The 5-choice serial reaction time task: behavioural pharmacology and
functional neurochemistry. Psychopharmacology (Berl), 2002. 163(3-4): p. 362-80.
14. Puumala, T., et al., Behavioral and pharmacological studies on the validation of a new
animal model for attention deficit hyperactivity disorder. Neurobiol Learn Mem, 1996.
66(2): p. 198-211.
15. Gendle, M.H., et al., Impaired sustained attention and altered reactivity to errors in an
animal model of prenatal cocaine exposure. Brain Res Dev Brain Res, 2003. 147(1-2):
p. 85-96.
16. Humby, T., et al., Visuospatial attentional functioning in mice: interactions between
cholinergic manipulations and genotype. Eur J Neurosci, 1999. 11(8): p. 2813-23.
17. van Gaalen, M.M., et al., Reduced attention in mice overproducing corticotropin-
releasing hormone. Behav Brain Res, 2003. 142(1-2): p. 69-79.
18. Greco, B. and M. Carli, Reduced attention and increased impulsivity in mice lack-
ing NPY Y2 receptors: relation to anxiolytic-like phenotype. Behav Brain Res, 2006.
169(2): p. 325-34.
19. Pattij, T., et al., Strain specificity and cholinergic modulation of visuospatial attention
in three inbred mouse strains. Genes Brain Behav, 2007. 6(6): p. 579-87.
20. Greco, B., R.W. Invernizzi, and M. Carli, Phencyclidine-induced impairment in atten-
tion and response control depends on the background genotype of mice: reversal by the
mGLU(2/3) receptor agonist LY379268. Psychopharmacology (Berl), 2005. 179(1): p.
68-76.
21. Hoyle, E., et al., Impaired performance of alpha7 nicotinic receptor knockout mice in
the five-choice serial reaction time task. Psychopharmacology (Berl), 2006. 189(2): p.
211-23.
22. Davies, W., et al., X-monosomy effects on visuospatial attention in mice: a candidate
gene and implications for Turner syndrome and attention deficit hyperactivity disorder.
Biol Psychiatry, 2007. 61(12): p. 1351-60.
23. Moon, J., et al., Attentional dysfunction, impulsivity, and resistance to change in a
mouse model of fragile X syndrome. Behav Neurosci, 2006. 120(6): p. 1367-79.
24. Driscoll, L.L., et al., Impaired sustained attention and error-induced stereotypy in the
aged Ts65Dn mouse: a mouse model of Down syndrome and Alzheimer’s disease.
Behav Neurosci, 2004. 118(6): p. 1196-205.

25. Domenger, D. and R.K. Schwarting, Sequential behavior in the rat: role of skill and
attention. Exp Brain Res, 2007. 182(2): p. 223-31.
26. Domenger, D. and R.K. Schwarting, The serial reaction time task in the rat: effects of
D1 and D2 dopamine-receptor antagonists. Behav Brain Res, 2006. 175(2): p. 212-22.
27. Domenger, D. and R.K. Schwarting, Sequential behavior in the rat: a new model using
food-reinforced instrumental behavior. Behav Brain Res, 2005. 160(2): p. 197-207.
28. Pearce, J.M. and G. Hall, A model for Pavlovian learning: variations in the effectiveness
of conditioned but not of unconditioned stimuli. Psychol Rev, 1980. 87(6): p. 532-52.
29. Maddux, J.M., et al., Dissociation of attention in learning and action: effects of lesions
of the amygdala central nucleus, medial prefrontal cortex, and posterior parietal cortex.
Behav Neurosci, 2007. 121(1): p. 63-79.
30. Holland, P.C., Disconnection of the amygdala central nucleus and the substantia
innominata/nucleus basalis magnocellularis disrupts performance in a sustained atten-
tion task. Behav Neurosci, 2007. 121(1): p. 80-9.
31. Ator, N.A., Subjects and instrumentation., in Experimental Analysis of Behavior, Part
1, I.H. Iversen and K.A. Lattal, Editors. 1991, Elsevier Science Publishers: Amsterdam.
p. 1-62.
32. Ali, J.S., et al., A LOTUS 1-2-3 - based animal weighing system with a weight mainte-
nance algorithm. Behavior Research Methods, Instrumentation, and Computers, 1992.
24: p. 82-87.
33. Patel, S., et al., Attentional performance of C57BL/6 and DBA/2 mice in the 5-choice
serial reaction time task. Behav Brain Res, 2006. 170(2): p. 197-203.
34. Wrenn, C.C., et al., Performance of galanin transgenic mice in the 5-choice serial reac-
tion time attentional task. Pharmacol Biochem Behav, 2006. 83: p. 428-440.
35. De Bruin, N.M., et al., Attentional performance of (C57BL/6Jx129Sv)F2 mice in the
five-choice serial reaction time task. Physiol Behav, 2006. 89(5): p. 692-703.
36. De Bruin, N.M., et al., Combined uridine and choline administration improves cogni-
tive deficits in spontaneously hypertensive rats. Neurobiol Learn Mem, 2003. 80(1): p.
63-79.
37. Eichenbaum, H., A. Fagan, and N.J. Cohen, Normal olfactory discrimination learning
set and facilitation of reversal learning after medial-temporal damage in rats: implica-
tions for an account of preserved learning abilities in amnesia. J Neurosci, 1986. 6(7):
p. 1876-84.
38. Strupp, B.J. and A. Diamond, Assessing cognitive function in animal models of mental
retardation. Mental Retard Develop Dis Res Rev, 1996. 2(4): p. 216-226.
39. ayer, L.E., et al., Prenatal cocaine exposure increases sensitivity to the attentional
effects of the dopamine D1 agonist SKF81297. J Neurosci, 2000. 20(23): p. 8902-8.
40. Bayer, L.E., et al., Increased sensitivity to idazoxan in rats exposed to cocaine in utero:
Evidence for enduring effects on catecholaminergic systems underlying attention.
Behav Brain Res, 2002. 133(2): p. 185-196.
41. Garavan, H., et al., Prenatal cocaine exposure impairs selective attention: evidence
from serial reversal and extradimensional shift tasks. Behav Neurosci, 2000. 114(4): p.
725-38.
42. Gendle, M.H., et al., Prenatal cocaine exposure does not alter working memory in adult
rats. Neurotoxicol Teratol, 2004. 26(2): p. 319-29.
43. Alber, S.A. and B.J. Strupp, An in-depth analysis of lead effects in a delayed spatial
alternation task: assessment of mnemonic effects, side bias, and proactive interfer-
ence. Neurotoxicol Teratol, 1996. 18(1): p. 3-15.
44. Garavan, H., et al., Enduring effects of early lead exposure: evidence for a specific
deficit in associative ability. Neurotoxicol Teratol, 2000. 22(2): p. 151-64.

45. Hilson, J.A. and B.J. Strupp, Analyses of response patterns clarify lead effects in olfac-
tory reversal and extradimensional shift tasks: assessment of inhibitory control, asso-
ciative ability, and memory. Behav Neurosci, 1997. 111(3): p. 532-42.
46. Morgan, R.E., D.A. Levitsky, and B.J. Strupp, Effects of chronic lead exposure on
learning and reaction time in a visual discrimination task. Neurotoxicol Teratol, 2000.
22(3): p. 337-45.
47. Beaudin, S.A., et al., Succimer chelation normalizes reactivity to reward omission and
errors in lead-exposed rats. Neurotoxicol Teratol, 2007. 29(2): p. 188-202.
48. Strupp, B.J., et al., Deficient cumulative learning: an animal model of retarded cogni-
tive development. Neurotoxicol Teratol, 1994. 16(1): p. 71-9.
49. Strupp, B.J., et al., Cognitive profile of rats exposed to lactational hyperphenylalanin-
emia: correspondence with human mental retardation. Dev Psychobiol, 1990. 23(3): p.
195-214.
50. Muir, J.L., B.J. Everitt, and T.W. Robbins, AMPA-induced excitotoxic lesions of the
basal forebrain: a significant role for the cortical cholinergic system in attentional func-
tion. J Neurosci, 1994. 14(4): p. 2313-26.
51. Muir, J.L., Attention and stimulus processing in the rat. Brain Res Cogn Brain Res,
1996. 3(3-4): p. 215-25.
52. Beaudin, S.A., et al., Succimer chelation normalizes reactivity to reward omission and
errors in lead-exposed rats. Neurotoxicol Teratol, 2006.
53. Parasuraman, R., The psychobiology of sustained attention., in Sustained attention in
human performance, W. JS, Editor. 1984, Wiley: Nw York. p. 61-101.
54. Craig, A. and D. Davies, Vigilance: Sustained visual monitoring and attention., in
Vision and Visual Dysfunction, J. Roufs, Editor. 1991, MacMillan: Basingstoke, UK. p.
83-98.
55. Koelega, H.S., Benzodiazepines and vigilance performance: a review. Psychopharma-
cology (Berl), 1989. 98(2): p. 145-56.
56. Koelega, H.S., Stimulant drugs and vigilance performance: a review. Psychopharma-
cology (Berl), 1993. 111(1): p. 1-16.
57. Koelega, H.S., Alcohol and vigilance performance: a review. Psychopharmacology
(Berl), 1995. 118(3): p. 233-49.
58. Bushnell, P.J., Detection of visual signals by rats:Effects of signal intensity, event rate
and task type. Behavioural Processes 1999. 46: p. 141-150.
59. Bushnell, P.J., K.L. Kelly, and K.M. Crofton, Effects of toluene inhalation on detection
of auditory signals in rats. Neurotoxicol Teratol, 1994. 16(2): p. 149-60.
60. McGaughy, J. and M. Sarter, Behavioral vigilance in rats: task validation and effects of
age, amphetamine, and benzodiazepine receptor ligands. Psychopharmacology (Berl),
1995. 117(3): p. 340-57.
61. Bushnell, P.J., V.A. Benignus, and M.W. Case, Signal detection behavior in humans and
rats: a comparison with matched tasks. Behav Processes, 2003. 64(1): p. 121-129.
62. Rezvani, A.H., D.P. Caldwell, and E.D. Levin, Chronic nicotine interactions with clo-
zapine and risperidone and attentional function in rats. Prog Neuropsychopharmacol
Biol Psychiatry, 2006. 30(2): p. 190-7.
63. Rezvani, A.H., D.P. Caldwell, and E.D. Levin, Nicotinic-serotonergic drug interac-
tions and attentional performance in rats. Psychopharmacology (Berl), 2005. 179(3): p.
521-8.
64. Rezvani, A.H. and E.D. Levin, Nicotine-alcohol interactions and attentional perfor-
mance on an operant visual signal detection task in female rats. Pharmacol Biochem
Behav, 2003. 76(1): p. 75-83.
65. Rezvani, A.H. and E.D. Levin, Nicotinic-glutamatergic interactions and attentional
performance on an operant visual signal detection task in female rats. Eur J Pharma-
col, 2003. 465(1-2): p. 83-90.

66. Bushnell, P.J., W.M. Oshiro, and B.K. Padnos, Detection of visual signals by rats:
effects of chlordiazepoxide and cholinergic and adrenergic drugs on sustained atten-
tion. Psychopharmacology (Berl), 1997. 134(3): p. 230-41.
67. McGaughy, J. and M. Sarter, Effects of ovariectomy, 192 IgG-saporin-induced corti-
cal cholinergic deafferentation, and administration of estradiol on sustained attention
performance in rats. Behav Neurosci, 1999. 113(6): p. 1216-32.
68. urchi, J., L.A. Holley, and M. Sarter, Effects of nicotinic acetylcholine receptor ligands
on behavioral vigilance in rats. Psychopharmacology (Berl), 1995. 118(2): p. 195-205.
69. Rezvani, A.H., P.J. Bushnell, and E.D. Levin, Effects of nicotine and mecamylamine
on choice accuracy in an operant visual signal detection task in female rats. Psycho-
pharmacology (Berl), 2002. 164(4): p. 369-75.
70. McGaughy, J., T. Kaiser, and M. Sarter, Behavioral vigilance following infusions of
192 IgG-saporin into the basal forebrain: selectivity of the behavioral impairment
and relation to cortical AChE-positive fiber density. Behav Neurosci, 1996. 110(2): p.
247-65.
71. McGaughy, J. and M. Sarter, Sustained attention performance in rats with intracortical
infusions of 192 IgG-saporin-induced cortical cholinergic deafferentation: effects of
physostigmine and FG 7142. Behav Neurosci, 1998. 112(6): p. 1519-25.
72. Turchi, J. and M. Sarter, Cortical cholinergic inputs mediate processing capacity:
effects of 192 IgG-saporin-induced lesions on olfactory span performance. Eur J Neu-
rosci, 2000. 12(12): p. 4505-14.
73. Burk, J.A. and M. Sarter, Dissociation between the attentional functions mediated via
basal forebrain cholinergic and GABAergic neurons. Neuroscience, 2001. 105(4): p.
899-909.
74. Turchi, J. and M. Sarter, Antisense oligodeoxynucleotide-induced suppression of basal
forebrain NMDA-NR1 subunits selectively impairs visual attentional performance in
rats. Eur J Neurosci, 2001. 14(1): p. 103-17.
75. Turchi, J. and M. Sarter, Bidirectional modulation of basal forebrain N-methyl-D-
aspartate receptor function differentially affects visual attention but not visual dis-
crimination performance. Neuroscience, 2001. 104(2): p. 407-17.
76. Sarter, M., J.P. Bruno, and B. Givens, Attentional functions of cortical cholinergic
inputs: what does it mean for learning and memory? Neurobiol Learn Mem, 2003.
80(3): p. 245-56.
77. Sarter, M., et al., Unraveling the attentional functions of cortical cholinergic inputs:
interactions between signal-driven and cognitive modulation of signal detection. Brain
Res Brain Res Rev, 2005. 48(1): p. 98-111.
78. Sarter, M., et al., Forebrain dopaminergic-cholinergic interactions, attentional effort,
psychostimulant addiction and schizophrenia. Exs, 2006. 98: p. 65-86.
79. Bushnell, P.J., et al., A dosimetric analysis of the acute behavioral effects of inhaled
toluene in rats. Toxicol Sci, 2007. 99(1): p. 181-9.
80. Oshiro, W.M., Q.T. Krantz, and P.J. Bushnell, Characterizing tolerance to trichloroeth-
ylene (TCE): effects of repeated inhalation of TCE on performance of a signal detec-
tion task in rats. Neurotoxicol Teratol, 2001. 23(6): p. 617-28.
81. Bushnell, P.J., Concentration-time relationships for the effects of inhaled trichloroeth-
ylene on signal detection behavior in rats. Fundam Appl Toxicol, 1997. 36(1): p. 30-8.
82. Samsam, T.E., D.L. Hunter, and P.J. Bushnell, Effects of chronic dietary and repeated
acute exposure to chlorpyrifos on learning and sustained attention in rats. Toxicol Sci,
2005. 87(2): p. 460-8.
83. Bushnell, P.J., et al., Neurobehavioral assessments of rats perinatally exposed to a com-
mercial mixture of polychlorinated biphenyls. Toxicol Sci, 2002. 68(1): p. 109-20.

84. Geller, A.M., et al., Gender-dependent behavioral and sensory effects of a commercial
mixture of polychlorinated biphenyls (Aroclor 1254) in rats. Toxicol Sci, 2001. 59(2): p.
268-77.
85. Burk, J.A., Introduction of a retention interval in a sustained attention task in rats:
effects of a visual distracter and increasing the inter-trial interval. Behav Processes,
2004. 67(3): p. 521-31.
86. Martin, L.A., D. Goldowitz, and G. Mittleman, Sustained attention in the mouse: a study
of the relationship with the cerebellum. Behav Neurosci, 2006. 120(2): p. 477-81.
87. Davenport, J., Combined autoshaping-operant (AO) training: CS-UCS interval effects
in the rat. Bulletin of Psychonomic Sciences 1974. 3: p. 383-385.
88. Bushnell, P.J., Behavioral effects of acute p-xylene inhalation in rats: autoshaping,
motor activity, and reversal learning. Neurotoxicol Teratol, 1988. 10(6): p. 569-77.
89. Fleschler, M. and H. Hoffman, A progression for generating variable-interval sched-
ules. Journal of the Experimental Analysis of Behavior, 1963. 5: p. 529-530.
90. Grier, J.B., Nonparametric indexes for sensitivity and bias: computing formulas. Psy-
chol Bull, 1971. 75(6): p. 424-9.
91. Green, D. and J. Swets, Signal Detection Theory and Psychophysics. 1974, Huntington,
NY: R.E. Krieger Publishing.
92. Frey, P. and J. Colliver, Sensitivity and responsivity measures for discrimination learn-
ing. Learning and Motivation, 1973. 4: p. 327-342.
93. Milner, B., Effects of different brain lesions on card sorting. Archives of Neurology,
1963. 9: p. 90-99.
94. Robbins, T.W., et al., Cambridge Neuropsychological Test Automated Battery (CAN-
TAB): a factor analytic study of a large sample of normal elderly volunteers. Dementia,
1994. 5(5): p. 266-81.
95. Chudasama, Y. and T.W. Robbins, Functions of frontostriatal systems in cognition:
comparative neuropsychopharmacological studies in rats, monkeys and humans. Biol
Psychol, 2006. 73(1): p. 19-38.
96. Roberts, A.C., et al., 6-Hydroxydopamine lesions of the prefrontal cortex in monkeys
enhance performance on an analog of the Wisconsin Card Sort Test: possible interac-
tions with subcortical dopamine. J Neurosci, 1994. 14(5 Pt 1): p. 2531-44.
97. Birrell, J.M. and V.J. Brown, Medial frontal cortex mediates perceptual attentional set
shifting in the rat. J Neurosci, 2000. 20(11): p. 4320-4.
98. Dias, R., T.W. Robbins, and A.C. Roberts, Primate analogue of the Wisconsin Card
Sorting Test: effects of excitotoxic lesions of the prefrontal cortex in the marmoset.
Behav Neurosci, 1996. 110(5): p. 872-86.
99. Roberts, A.C., T.W. Robbins, and B.J. Everitt, The effects of intradimensional and
extradimensional shifts on visual discrimination learning in humans and non-human
primates. Q J Exp Psychol B, 1988. 40(4): p. 321-41.
100. Colacicco, G., et al., Attentional set-shifting in mice: modification of a rat paradigm, and
evidence for strain-dependent variation. Behav Brain Res, 2002. 132(1): p. 95-102.
101. McAlonan, K. and V.J. Brown, Orbital prefrontal cortex mediates reversal learning and
not attentional set shifting in the rat. Behav Brain Res, 2003. 146(1-2): p. 97-103.
102. Rezvani, A.H. and E.D. Levin, Nicotine-antipsychotic drug interactions and attentional
performance in female rats. Eur J Pharmacol, 2004. 486(2): p. 175-82.

APPENDIX: NAMES AND ADDRESSES OF VENDORS DISCUSSED IN

THE TEXT
Behavioral Test Systems

(USA/Canada/Mexico)
Campden Instruments Ltd. TSE Systems, Inc.
http://www.campdeninstruments. 784 S. Poseyville Road
com/home.htm Midland, MI 48640 USA
(Europe) Tel: 989-698-3067
Loughborough (Worldwide)
LE127XT, England TSE Systems GmbH
Tel: 0150-981-14790 Siemensstr. 21
UKsales@campdeninstruments.com 61352 Bad Homburg, Germany
(Worldwide) Tel: +49-(0)6172-789-0
Lafayette, IN USA
Tel: 765-423-1505
USsales@campdeninstruments.com Calibration: Audiometric
Brüel & Kjær Instruments, Inc.

Coulbourn Instruments, LLC 185 Forest St.
7462 Penn Drive Marlborough, MA 10752 USA
Allentown, PA 18106 USA
Tel: 610-395-3771
Calibration: Photometric
www.coulbourninst.com
EG&G Gamma Scientific
Med Associates Inc. 8581 Aero Drive
PO Box 319 San Diego, CA 92123 USA
St. Albans, VT 05478 USA
Tel: 802-527-2343
Food Pellets
www.med-associates.com
Bio-Serv
Panlab, SL One 8th Street, Suite 1
C/ Energia,112 Frenchtown, NJ 08825 USA
08940 Cornellà (Barcelona), Spain www.bio-serv.com
Tel: +34-934-750-697 (Int’l Sales)
http://www.panlab.com/ P.J. Noyes Company, Inc.
PO Box 381
TSE Systems Bridge St.
http://www.tse-systems.com/ Lancaster, NH 03584 USA

8 The Behavioral
Assessment of
Sensorimotor Processes
in the Mouse
Acoustic Startle, Sensory
Gating, Locomotor Activity,
Rotarod, and Beam Walking
Peter Curzon, Min Zhang, Richard
J. Radek, and Gerard B. Fox
CONTENTS
8.1 Introduction................................................................................................. 146
8.2 Acoustic Startle.......................................................................................... 148
8.2.1 Equipment ........................................................................................ 148
8.2.2 Setup and Decibel Confirmation ..................................................... 149
8.2.3 Stimulus Parameters ........................................................................ 149
8.2.4 Testing Location .............................................................................. 149
8.2.5 Subjects ............................................................................................ 149
8.2.6 Acoustic Startle Protocol ................................................................. 150
8.2.7 Startle Habituation........................................................................... 151
8.3 Sensory Gating............................................................................................ 152
8.3.1 Prepulse Inhibition........................................................................... 152
8.3.1.1 Statistical Analysis ............................................................. 152
8.3.1.2 Association of PPI and Startle Responses.......................... 153
8.3.1.3 Sample Prepulse Inhibition Experiment ............................ 153
8.3.2 N-40 Sensory Gating ....................................................................... 155
8.3.2.1 Introduction ........................................................................ 155
8.3.2.2 Method ............................................................................... 158
8.3.2.3 Subjects and Surgery.......................................................... 158
8.3.2.4 Recording of Paired Stimulus Sensory Gating Evoked
Potentials ............................................................................ 159
145

8.3.2.5 Auditory Stimuli................................................................. 160

8.3.2.6 Evoked Potential Analysis.................................................. 161
8.3.2.7 Example Drug Studies with DBA/2 Mice.......................... 161
8.4 Motor Function and Spontaneous Exploration ........................................... 163
8.4.1 Spontaneous Activity....................................................................... 165
8.4.1.1 Open Field (Non-automated).............................................. 165
8.4.1.2 Typical Protocol ................................................................. 165
8.4.1.3 Open Field (Automated)..................................................... 166
8.4.1.5 Variations ........................................................................... 167
8.4.1.6 Sample Experiment ............................................................ 168
8.4.1.7 Motor Function................................................................... 168
8.4.2 Rotarod............................................................................................. 168
8.4.2.2 Variation ............................................................................. 170
8.4.2.3 Sample Experiment ............................................................ 170
8.4.3 Beam Balance/Walking ................................................................... 171
8.4.3.2 Variation ............................................................................. 172
8.4.3.3 Example Experiment.......................................................... 172
References.............................................................................................................. 174
Appendix: Equipment Suppliers ............................................................................ 177
8.1 INTRODUCTION
Assessment of sensorimotor competence is an important part of the evaluation of ani-
mal behavior. Measurement of sensorimotor performance is of obvious importance
in investigations of sensory or motor processes; however, the effects of experimental
manipulations on sensorimotor performance have broader implications for behavioral
neuroscience because behavioral experiments typically measure motor responses to
sensory information. Thus, the results of behavioral experiments designed to assess
other neurobiological processes often cannot be properly interpreted without con-
sidering concomitant effects on sensorimotor function. For example, if a lesion or
genetic manipulation impairs performance on a spatial memory test, such as the
radial arm maze, this impairment cannot be interpreted as evidence of cognitive
dysfunction unless it is first established that it is not the result of sensorimotor defi-
cits. Moreover, sensorimotor effects of manipulations can often be used in animal
models as surrogates for effects that are more difficult to measure, and relatively
simple variations of sensorimotor measures can be used as indices of performance in
other behavioral domains, including cognition and emotion. A number of behavioral
tasks have been designed to assess sensorimotor performance in rodents, and this
chapter focuses on five general procedures—acoustic startle, sensory gating, open
field exploration, rotarod, and beam walking.
The startle reflex is a stereotyped motor response to a sudden, intense stimulus
that has been assessed experimentally in a variety of species, including rats, mice,
cats, monkeys, and humans.1,2 In rodents, the startle response is typically evoked

The Behavioral Assessment of Sensorimotor Processes in the Mouse 147
using either acoustic or tactile stimuli and is characterized by contractions of the

major muscles of the body, generally leading to extension of the forepaws and hind
paws followed by muscle flexion into a hunched position. Mapping studies have
demonstrated that the acoustic startle reflex is mediated by a specific neural path-
way with acoustic information entering the CNS through auditory nerve input to the
cochlear nucleus, which projects to the reticular pontine nucleus via the lateral lem-
nisus. Motor outputs are generated in the reticular pontine nucleus, which projects
to the ventral spinal horn through the reticulospinal tract. Although the basic reflex
pathway appears to be relatively simple, the reflex is subject to modulatory influ-
ences from higher brain structures.3
Measurement of acoustic startle responses can provide general information
regarding sensorimotor processing, but measurement of the reflex under conditions
that engage the influence of higher brain centers provides an even richer source of
data. For example, presentation of lower intensity acoustic stimuli immediately
prior to the acoustic startle stimulus attenuates the response to the startle stimulus.
This phenomenon, called prepulse inhibition (PPI) of startle reflex, is regulated by
forebrain neural circuits and is considered an operational measure of sensorimotor
gating, a filtering mechanism to prevent information flooding in the brain so that
attentional recourses can be selectively allocated to salient stimuli.4,5 The normally
functioning brain has endogenous mechanisms that filter out the multitude of irrel-
evant sensory stimuli from those of importance.
Sensory gating is a term given to this filtering mechanism and can be mea-
sured. A sensory gating deficit, or the lack of an ability to inhibit responding to the
test stimulus, is used as a clinical measure of schizophrenia.6–9 The two methods
of sensory gating described in this chapter, PPI and N-40 gating, both measure the
transmission of auditory sensory information to the nervous system. For PPI the
observable and measured response is based on the motor output (startle) following a
loud acoustic stimulus. Inhibition of the response to this stimulus is observed if the
stimulus is preceded very shortly (10–500 msec) by a prepulse stimulus to which the
organism normally does not respond. PPI is defined as reduction of a response to a
stimulus. In contrast, the N-40 response is a measure of electrical brain activity to
repeated auditory stimuli. Although more involved, there are several advantages to
measuring sensory gating using electrophysiological recordings. First, afferent neu-
ronal activity largely contributes to sensory gating evoked potential (EP) response,
while afferent, efferent, and neuromuscular activity contribute to the PPI response.
Therefore, recording EPs allows for the study of sensory information processing
without the influence of descending motor responding as in PPI. Secondly, although
the hippocampus and cortex are most frequently recorded, electrophysiological tech-
niques allow the study of the inhibitory processes in different brain regions.10
Pharmacological manipulations and strains of mice yield sensory deficits remi-
niscent of those seen in schizophrenic patients. For example, the pro-psychotic
amphetamine produces N-40 sensory gating deficits in rodents and P-50 deficits in
humans.11,12 As another example, the F7 nicotinic receptor subtype, is thought to be
deficient in schizophrenia, and decreased hippocampal F7 receptor levels in DBA/2
mice are thought to account for the N-40 sensory gating deficits expressed in this
strain.13 Also, as in schizophrenic patients and PPI procedures in rodents, atypi-

cal antipsychotics reverse sensory gating deficits in both PPI and N-40 in DBA/2
mice.14,15 Thus, sensory gating measures show considerable parallels across species
and have good rodent–human translation. Other sensory gating deficit models have
been reported, but the use of the DBA/2 mouse is described herein as representative
of an approach to study the nicotinic-acetylcholine system and sensory gating.
8.2 ACOUSTIC STARTLE

8.2.1 EQUIPMENT
The equipment used to measure the startle response has varied from simple lab-
made devices16,17 to more sophisticated units available from various commercial sup-
pliers. We have used the SR-LAB Startle Response System (San Diego Instruments,
San Diego, California, USA) and Kinder Scientific Startle Monitor (SM 100 ver-
sion 4.1, Poway, California, USA). In addition, acoustic startle equipment can also
be obtained from Coulbourn Instruments. The ability to deliver stimuli for 5–1000
msec with consistent intensity is important. Each device is typically enclosed in a
larger soundproof cubicle that isolates the animal in the presence of background
noise. This also serves to protect the animals in the immediate vicinity from being
exposed to the acoustic startle stimulus. A simpler, cheaper SR-LAB screening sys-
tem is also available from San Diego Instruments, but it has no enclosure and the
animals must obviously be isolated by location from other test animals. The magni-
tude of the response of an animal depends on the size of the animal, which means
that the assessment of the acoustic startle reflex in the mouse requires more sensitive
equipment than the assessment of the reflex in the rat.
The SR-LAB system and Kinder Scientific Startle Monitor include a separate
isolation chamber for each individual startle unit. The outer sound-attenuating
chamber is illuminated and ventilated with a small fan that also provides some level
of background noise. An acoustic sound source is located in the upper part of this
chamber and consists of a loudspeaker that delivers a full spectrum white noise that
is computer controlled for duration and decibel level. The startle unit is available in
assorted sizes to accommodate mice or rats. With the SR-LAB system, the cylin-
drical animal enclosure and the enclosure base are installed within the SR-LAB
chamber. Each animal enclosure has an attached piezoelectric motion sensor that
detects the movement of the animal. The signal from the motion sensor is sent to
a computer for digital transformation. The motion sensor supplied by San Diego
Instruments has an adjustable potentiometer on the underside. Compared to the SR-
LAB system, the Kinder Scientific Startle Monitor has some unique features that
we liked. The Kinder Scientific equipment animal enclosure is separable from the
sensing plate that serves as a motion sensor, allowing for easy cleaning of the animal
enclosure without damaging the motion sensor. Also, the SR-LAB system requires
calibration to standardize test chambers before each test, whereas calibration is not
necessary for each test with the Kinder Scientific Startle Monitor. Furthermore,
Kinder Scientific uses direct readings for the magnitude of acoustic stimuli rather
than analog levels used by the SR-LAB system that require the use of a sound meter
to calculate decibel levels.

8.2.2 SETUP AND DECIBEL CONFIRMATION

The initial setup of the equipment is fairly simple and the experimental program-
ming is easily accomplished by following the manual provided. Measuring the deci-
bel levels in each enclosure prior to commencing experiments is recommended. A
RadioShack (Allied Electronics) Sound Level Meter #33-2050 set on slow response
(“A” weighting) is a simple and inexpensive device for measuring the intensity of the
acoustic stimuli. Kinder Scientific sells a device with a sensor connected via a cord
to the decibel meter, making it convenient to read the sound level outside the closed
chamber. The startle unit sound duration must be set long enough (e.g., 6000 msec)
to accurately measure sound level, and the sound meter should be placed in the posi-
tion normally occupied by the animal holder. If isolation chambers are used, calibra-
tion should be done with all of the chamber doors closed to reduce the likelihood
that sound from the other chambers will influence the reading made in the chamber
being calibrated.
8.2.3 STIMULUS PARAMETERS

Startle responses can be measured over a period of time up to 1 sec after the pre-
sentation of the startle stimulus. However, since the startle response is typically over
within 100 msec of stimulus presentation, the window should be set to include only
movements generated within the first 200 msec following the startle stimulus. When
analyzing data, use the maximal voltage generated during this 200-msec period;
however, it is also possible to use average voltage across the entire response win-
dow if desired. For mice, stimuli with intensities of 90 dB or higher will typically
produce startle responses, although there is some strain-dependent variability.18 The
magnitude of the startle response varies as a function of the intensity of the startle
stimulus, so more reliable responses are often obtained at higher intensities. Acous-
tic stimuli intensities should not be set higher than 120 dB to avoid producing dam-
age to the ear and the loudspeakers.
8.2.4 TESTING LOCATION

Animals are brought to a convenient holding area near the room containing the star-
tle chambers for acclimation. Thus, the startle equipment is best located within an
inner room with a heavy door. This provides additional sound attenuation and keeps
animals held in the vicinity of the testing room from being exposed to the startle
stimulus.
8.2.5 SUBJECTS
Among the mouse strains used, the DBA/2J (Jackson Laboratories, USA) mice,
11–17 g, exhibit a naturally occurring low PPI and thus provide a window to detect a
PPI enhancing effect (See Figure 8.1).19,20 One caveat concerning the DBA/2J mice is
mice older than 8 wk have hearing loss and thus only young DBA/2J mice are used.
Also used are CD1 and C57BL/6 mice, 28–40 g, for PPI or startle habituation stud-
ies. The animals are housed eight per cage (reflecting the number of test chambers

70
CD1 mice
60
C57BL/6
50 DBA/2J
40
% PPI
30
20
10
0
70 dB 75 dB 80 dB
Prepulses
FIGURE 8.1 The effect of 70, 75, and 80 dB prepulse on prepulse inhibition (PPI) displayed
by CD1, C57BL/6, and DBA/2J mice. Shown are mean ± SEM. Note that increasing the
intensity of the prepulse levels increases PPI. Kinder Scientific Startle Monitor was used. N
= 8–12 per group. Source: Author’s unpublished data.
employed) with water and food available ad libitum. Aggressive dominant males
should be removed from holding cages before commencement of any experiment.
Best results are often obtained from animals that have been protected from stressors
and have been habituated to the laboratory/animal quarters for at least 7 days.
8.2.6 ACOUSTIC STARTLE PROTOCOL

After placing the animal in the test chambers allow a 5-min adaptation period before
the start of the session. Background white noise (65 dB) is present during this adap-
tation period and throughout the session. The session starts with four 120 dB, 40
msec sound bursts. These are not included in the analysis because the responses to
the first few startle bursts generally differ in magnitude from the rest of the trials.
Thus, exposure to these initial bursts allows for the establishment of a stable base-
line. Following this, acoustic startle trials are initiated. For simple assessment of
acoustic startle responses, use two or three stimulus intensities (90 and 105 dB or 90,
105, and 120 dB). The stimuli are 40 msec in duration and are presented in a quasi
random order so that an equal number of presentations of each stimulus intensity is
included in each half of the session, and no single intensity is presented more than
two times in succession. The time between stimuli averages 15 sec but this interval
should be varied within a range of 5–30 sec so that the animals do not anticipate the
stimulus. At least 10 trials at each stimulus intensity should be used to obtain reli-
able results. An alternative to evaluating the magnitude of the startle response is to
measure the startle threshold. Here, a wider range of stimulus intensities, e.g., from
70 to 120 dB, is used. Stimuli are presented in quasi-random order, and the lowest
intensity producing a reliable response is determined. Since some habituation of the

response can occur both within a session and between sessions (see below), trials at
each intensity should always be evenly distributed within a session (but they should
not, of course, be presented in a predictable sequence).
8.2.7 STARTLE HABITUATION

This measure is of interest in the area of schizophrenia research, since schizo-
phrenics show impaired startle habituation, an impairment that may be related
to the hypervigilance characteristic of this condition.21,22 For startle habituation,
a single stimulus intensity is repeatedly presented using either a fixed or variable
interval. Responses normally decline (habituate) over trials. An example of the
data generated in such an experiment is shown in Figure 8.2. In this case, each
session is initiated with a 10-min acclimation period followed by 121 successive
120 dB, 40 msec trials separated by a fixed inter-trial interval (ITI) of 10 sec. The
first trial is excluded from analysis due to variability. Each block has 20 trials.
Two groups of CD1 mice received ip injection of water and MK-801 at 0.3 mg/kg,
respectively, 15 min before the test. As shown in the figure, startle habituation is
displayed by the water-treated group but not by the MK-801-treated group. Long-
term habituation of the startle response can also be used as a memory index by
conducting a second session at a later time (e.g., 24 hr later) and assessing long-
term retention of habituation.
Startle Response to s120

1.6
1.5
VEH
1.4
MK-801 0.3
1.3
1.2
Newtons (± SE M)
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0 1 2 3 4 5 6
Blocks
FIGURE 8.2 An example of startle habituation in two groups of CD1 mice treated with water
and MK-801 at 0.3 mg/kg, respectively. The study was carried out in Kinder Scientific Star-
tle Monitor equipment. Shown are mean ± SEM. The study showed that MK-801 treatment
impaired startle habituation. N = 9–12 per group. Source: Author’s unpublished data.

8.3 SENSORY GATING

8.3.1 PREPULSE INHIBITION
In this version of the test, the attenuation produced by a low intensity stimulus pre-
sented just before the startle stimulus is assessed. We sometimes assess PPI on the
day following standard startle testing. This serves to habituate the animals to the
basic handling procedures and, in the case of pharmacological studies of PPI, allows
groups to be matched for baseline startle, as well as elimination of animals that
either startle excessively or do not respond. The animals are given a 5-min accli-
mation period in the startle chambers during which a 65-dB background noise is
presented. This background noise remains throughout the entire test. Following
the 5-min acclimation period, four successive trials of 40-msec noise bursts at 120
dB are presented. These trials are not included in data analysis. Subjects are then
exposed to five different types of acoustic stimuli in a randomized order: pulse alone
(120-dB noise for 40 msec), no stimulus (no stimulus is presented), and three sepa-
rate prepulse + pulse combinations, with prepulse set at three sound levels of 70, 75,
and 80 dB for 20 msec followed by a 40-msec pulse at 120 dB. There are 100 msec
between the prepulse and the pulse. A total number of 12 trials under each acous-
tic stimulus condition are presented with average 20-sec variable intervals ranging
from 5 sec to 25 sec. The inclusion of four pulse-alone trials in the beginning of the
experiments is to help normalize the response of the mice, as there is rapid habitua-
tion to the startle responses seen within the first few trials.
More than one level of prepulse intensity is used, whereas a “no stimulus” trial is
used to assess the influence of background movement on startle measures. To reduce
variability, at least 12 trials of each type should be conducted within a single session.
As with standard startle testing, trial types and ITIs should be presented in a quasi-
random, balanced manner with equal representations of trial types and intervals
in each half of the session. This allows the session to be analyzed in two blocks to
assess any changes over time.
8.3.1.1 Statistical Analysis

The results can either be printed out or, more practically, assembled in a computer
text file. Microsoft Excel or similar spreadsheet software easily assemble the data
into the various trial types. Percent PPI is calculated as follows:
[1 − (startle response to prepulse + pulse) ÷ (startle response to pulse alone)] × 100.
If there is a drug treatment involved using naïve rodents, data are typically ana-
lyzed with two-way analysis of variance (ANOVA), with the drug treatment as a
between-subjects variable and prepulse level as a repeated measure. If a significant
treatment effect and interaction of treatment and prepulse are identified, percent PPI
is then analyzed by post-hoc comparisons to compare group means at each prepulse
level. If a significant treatment effect is identified without the presence of significant
interaction of treatment and prepulse, percent PPI can be collapsed across prepulse
levels and analyzed by post-hoc comparisons for treatment difference. Use of a post-

hoc test such as Fisher’s post-hoc PLSD and Dunnett’s test in order to compare the
groups treated with vehicle and the groups treated with drug should be decided a
priori. To evaluate drug effects on the response to the startle (s120) alone, a separate
one-way ANOVA, followed by post-hoc comparisons if there is a significant treat-
ment effect, is calculated. Programs such as JMP statistics software package (SAS
Institute, Cary, North Carolina, USA) or Graph Pad Prism (Graph Pad Software, San
Diego, California, USA) are useful for statistical analysis.
8.3.1.2 Association of PPI and Startle Responses

Substantial effects of an experimental treatment, resulting in an increase or decrease
of the startle response (s120), may confound the effects on PPI. However, there is
no agreement upon how PPI and startle responses are correlated. A reduced startle
response could be accompanied by a PPI reduction, PPI enhancement, or no change
of PPI dependent upon the drug treatment. Our previous studies show that antipsy-
chotics enhance PPI responding while substantially decreasing startle responses
(Figure 8.3).20 To address the question whether the PPI-enhancing effects observed
with antipsychotics is secondary to their effects on startle response,20 we pooled data
of vehicle and 1 mg/kg risperidone-treated mice from studies using risperidone as
positive control. We then sorted the data by startle responses in ascending values.
In order to obtain equal startle magnitude in the vehicle and risperidone groups,
the pairs of individuals with risperidone and vehicle treatment listed next to each
other were kept for further analysis. In cases where there were two possible pairs
(e.g., if there was a sequence of vehicle-risperidone-vehicle, risperidone could be
paired with the first vehicle and the second vehicle), the pair that had the smaller
startle difference between vehicle- and risperidone-treated mice was chosen. This
selection resulted in n = 19 in water- and risperidone-treated groups, respectively.
The remaining mice were excluded. One-way ANOVA revealed (Figure 8.4) that
there was a significant difference (p < 0.01) in percent PPI between risperidone- and
water-treated mice when their mean values of startle magnitude were equal, suggest-
ing that the effects on PPI responding can be independent of the effects on startle
magnitude. A study investigating basal acoustic startle responses and PPI among
several inbred mouse strains showed there was no correlation between the magni-
tude of basal acoustic startle responses and PPI,23, suggesting that different physi-
ological processes are involved in basal acoustic startle and PPI of startle reflex.
These observations support the notion that different neurobiological processes may
underlie gating processes and the startle reflex.
8.3.1.3 Sample Prepulse Inhibition Experiment

An experiment was conducted to evaluate the effects of BP 897, a preferential dopa-
mine D3 receptor antagonist, on PPI.20 In pharmacological experiments, it is often
important to conduct the maximum number of trials in the shortest possible time
because of pharmacokinetic considerations with many test compounds. Session
duration will obviously increase as more trial types/numbers are added, and this
was taken into account when we set up our standard rodent testing paradigm. Please
see section 8.3.1, “Prepulse Inhibition,” for details of the paradigm. The sequence of

Haloperidol
55 0.6
50
45 ** ** 0.5
*
Startle to s120
40
35 0.4
% PPI
30 0.3
25
20 0.2
15
10 0.1
5
0 0.0
0 0.3 1 3 0 0.3 1 3
(Treatment mg/kg IP) (Treatment mg/kg IP)
(a)
Risperidone
55 ** 0.6
50
45 ** 0.5
Startle to s120
40
35 0.4
% PPI
30
0.3
25
20 0.2
15
10 0.1
5
0 0.0
0 0.1 0.3 1 0 0.1 0.3 1
(b)
FIGURE 8.3 Effects of the antipsychotics, haloperidol and risperidone on prepulse inhibi-
tion (PPI) in DBA/2J mice. The effects on PPI and startle responses are presented on the left
and the right side of the panels, respectively. Haloperidol (a) and respiridone (b), N = 9–10
per group, significantly increased PPI at all of the doses tested while eliciting a nonsignifi-
cant reduction of startle response to pulse alone. Shown are mean ± SEM. *p < 0.05 and **p
< 0.01, compared to vehicle-treated alone group. Source: Data reproduced with permission
from Zhang, M., et al. 2006. Effect of dopamine D3 antagonists on PPI in DBA/2J mice
or PPI deficit induced by neonatal hippocampal lesions in rats. Neuropsychopharmacology
31(7):1382–92. 2006.
trials is illustrated in Table 8.1. The 68-trial session used ran for approximately 25
min. In this experiment, DBA/2J mice (11–13 per group) were used. Included in the
study was a positive control group in order to be certain that the experiment ran as
expected. In this study, risperidone at 1 mg/kg was used as a positive control. Both
BP 897 and risperidone were dissolved in 1 N HCL and then titrated to a final pH of
5 with 1 N NaOH. Both of the compounds were given ip in a volume of 10 mL/kg
30 min before the test. The results, which were analyzed using a two-way ANOVA
with treatment as a between-subjects variable and prepulse as a repeated measure,
revealed a significant main effect of treatment on percent PPI [F(5, 66) = 2.936, p <
0.05] in the absence of significant interaction of treatment and prepulse; the data was
collapsed across the three prepulses and the average percent PPI values of the three

Risperidone (re-analysis of the pooled data)

55 0.6
50 **
45 0.5
40
0.4
Startle to s120
35
% PPI
30
0.3
25
20 0.2
15
10 0.1
5
0 0.0
0 1 0 1
FIGURE 8.4 All of the data of vehicle- and 1-mg/kg-risperidone-treated DBA mice from
several mouse studies were pooled and then the startle magnitude of vehicle- and risperi-
done-treated individuals were matched (for details, see section 8.3.1, “Prepulse Inhibition”)
to compare the effect of vehicle and risperidone on percent prepulse inhibition (PPI) when
the startle magnitude was equal. This selection resulted in N = 19 in water- and risperidone-
treated groups, respectively. One-way ANOVA revealed a significant difference (p < 0.01)
between risperidone- and water-treated mice when their mean values of startle magnitude
were equal, suggesting that the effects on PPI responding can be independent of the effects
on startle magnitude. Shown are mean ± SEM. Hamilton Kinder Startle Monitor was used.
Source: Data reproduced with permission from Zhang, M., et al. 2006. Effect of dopamine D3
antagonists on PPI in DBA/2J mice or PPI deficit induced by neonatal hippocampal lesions in
rats. Neuropsychopharmacology 31(7):1382–92. 2006.
prepulses were analyzed with Fisher’s PLSD to compare the vehicle-treated group to
each of the drug-treated groups. As shown in Figure 8.5, the positive control, risperi-
done, significantly increased PPI (p < 0.05), suggesting the study was valid. The test
compound, BP 897, also significantly enhanced PPI at 8 mg/kg (p < 0.05). Startle
response to s120 was analyzed with a one-way ANOVA with treatment as a between-
subject variable. A significant main effect of treatment on startle was identified [F(5,
66) = 5.575, p < 0.01]. The follow-up Fisher’s PLSD showed that risperidone, but not
BP 897, significantly reduced startle response to s120.
As previously mentioned, DBA/2J mice have a lower PPI response compared
to other mouse strains and thus would provide enough of a window for seeing a
PPI-enhancing effect following a drug treatment. We also use other mouse strains
such as CD1 mice in studies to investigate PPI deficits. Pharmacological disruption
of PPI in mice can also be obtained with compounds that influence dopaminergic
(e.g., apomorphine, amphetamine), glutaminergic (e.g., phencyclidine, MK-801),
muscarinic cholinergic (e.g., scopolamine), and serotoninergic (e.g., 2,5. dimethoxy-
4-iodoamphetamine) neurotransmission.15
8.3.2 N-40 SENSORY GATING

8.3.2.1 Introduction
Evoked potentials are synchronous discharges of neuronal circuits or populations
that are time locked to a sensory stimulus. Sensorimotor gating can be measured by

TABLE 8.1
Session Protocol for Prepulse Inhibition Experiment
Trial # Trial Type (dB) ITI (msec) Trial # Trial Type ITI (msec)
1 Pulse alone 120 10 35 Prepulse 70 20
3 Pulse alone 120 20 37 No-stimulus 15
4 Pulse alone 120 15 38 Pulse alone 120 25
5 Prepulse 70 5 39 Prepulse 80 15
7 No-stimulus 15 41 Prepulse 70 20
11 Prepulse 70 10 45 No-stimulus 20
15 No-stimulus 20 49 No-stimulus 10
19 No-stimulus 20 53 Pulse alone 120 10
25 Prepulse 70 20 59 Pulse alone 120 25
28 No-stimulus 15 62 No-stimulus 15
30 No-stimulus 15 64 Prepulse 75 5
32 No-stimulus 5 66 Pulse alone 120 10
aITI, inter-trial interval, the delay (in sec) prior to the initiation of the trial
bEach prepulse trial consists of a 20 msec prepulse stimulus and a 40 msec, 120 dB startle stimulus
(separated by 100 msec); each pulse alone trial consists of a 40 msec presentation of the 120 dB
startle stimulus by itself.

BP 897
60 0.8
Startle Response to Pulse Alone

0.7
50
0.6
*
40 *
0.5
% PPI
30 0.4
0.3
20
0.2 **
10 0.1
0 0
0 1 2 4 8 RISP 1 0 1 2 4 8 RISP 1
Treatment (mg/kg IP)
FIGURE 8.5 Prepulse inhibition (PPI)-enhancing effects of BP 897 in DBA/2J mice. BP

897 at 8 mg/kg significantly increased percent PPI without affecting startle response to pulse
alone, while the positive control, risperidone, significantly increased PPI and decreased star-
tle response to pulse alone. *p < 0.05 and **p < 0.01, compared to the vehicle group. Shown
are mean ± SEM. Hamilton Kinder Startle Monitor was used. N = 11–13 per group. Source:
Data reproduced with permission from Zhang, M., et al. 2006. Effect of dopamine D3 antago-
nists on PPI in DBA/2J mice or PPI deficit induced by neonatal hippocampal lesions in rats.
Neuropsychopharmacology 31(7):1382–92. 2006.
recording electroencephalographic (EEG) EP responses to pairs of identical audi-

tory stimuli.9 Each stimulus of a pair in a sensory gating paradigm is given a specific
name; the first stimulus of a pair is typically called the conditioning stimulus, or S1,
while the second is referred to as the test stimulus, or S2. The terminology used for
stimuli in this section will be “condition” and “test.” Ordinarily, the EP elicited by
the test stimulus, typically an auditory click or tone, is smaller in amplitude than that
evoked by the conditioning stimulus (Figure 8.6). A lowered test response amplitude
is widely considered to be indicative of sensory gating or the filtering function of
the brain.6,7 A gated EP response to visual stimuli, such as to paired strobe flashes,
is not observed in humans; however, cross-modal visual to auditory sensory gating
phenomenon have been reported.24–26 The majority of animal sensory gating studies
are reported to use auditory stimuli.
The component of the human auditory EP that shows this gating response is
the P-50 wave, a major positive deflection in the ongoing EEG with a latency of
about 50 msec after the stimulus. It is thought that some form of active inhibition of
neuronal activity is initiated by the first stimulus, and suppresses neuronal activity
thereafter.27,28 Although somewhat controversial, in rodents the major EP component
analogous to the human P-50 is thought by some to be the N-40 (negative polarity,
40 msec latency), and is typically suppressed in many strains after presentation of
the test stimulus.29,30

Condition Test
N-40
N-40
200 ms
FIGURE 8.6 Example of an auditory evoked potential sensory (inhibitory) gating response
from a C3H strain mouse. These electrographic potentials were recorded from the CA3 region
of the hippocampus. The condition and test stimuli are separated by a delay of 500 msec. The
major negative deflection of the potential is approximately 40 msec after the stimulus, and
thus is termed the N-40 wave. The N-40 amplitude in response to the conditioning stimulus is
much larger than the test stimulus amplitude, and is thought to reflect the inhibitory processes
of sensory filtering.
8.3.2.2 Method
The techniques for recording sensory EPs in anesthetized and unanesthetized rodents
are well established. Recording freely moving unanesthetized versus anesthetized
rats and mice is somewhat more difficult because of movement artifact; however,
this approach avoids any possible confounds of the anesthetic in pharmacological
studies. The recording techniques described herein are of auditory EPs recorded in
the hippocampus from unanesthetized DBA/2 mice, but many of these methods are
applicable to recordings in rats and recordings in other areas of the brain.
8.3.2.3 Subjects and Surgery

For implantation of EEG recording electrodes, DBA/2 mice (16–20 g, 6–7 wk,
Harlan) are anesthetized with a solution of 2.8% ketamine, 0.28% xylazine, 0.05%
acepromazine (Sigma Chem. Co.) at 140 mg/kg of ketamine. Other anesthetics pro-
duce less satisfactory results in mice in terms of survival. After achieving a stable
plane of anesthesia, scalp hair is removed and the skin is cleaned with a standard vet-
erinary disinfectant solution (e.g., povidone iodine). The mouse is placed in a Kopf
student stereotaxic frame and a sagittal incision approximately 6 mm long is made
along the centerline to expose the bone between lambda and bregma. The bone is
dried with a 30% hydrogen peroxide solution, which makes suture landmarks easier
to see. Three drill holes (#68 drill bit) are made at medial lateral (ML) 1.0, 1.8, and
2.6 mm from the central suture. All three are located at anterior posterior (AP) −1.8

mm from bregma, thus, they are in a plane perpendicular to the central suture. The
hole at ML 2.6 and AP −1.8 mm is for the electrode directed at the CA3 region of
the hippocampus. The holes at ML 1.0, AP −1.8 mm, and ML 1.8, AP −1.8 mm, are
for two electrodes that lie on the surface of the cortex. The depth of the hippocampal
electrode tip is dorsal ventral (DV) 1.65–1.70 mm below the surface of the cortex.
The depth of the cortical electrodes are dorsal ventral (DV) 0.5 mm from the surface
of the skull; a distance that results in the electrode tip being in contact with, but not
penetrating, the cortical tissue. The electrodes we use for recording mouse EPs are
from Plastics One, Inc., Roanoke, Virginia, USA. They are tripolar stainless steel
wires that have an integral mounting pedestal for connecting a tether when record-
ing the EEG. These electrodes are cut to a length that is appropriate for the cortical
or subcortical target. After experiments are completed, the accuracy of the electrode
placement should be verified histologically (e.g., crystal violet staining) and, if inac-
curate, coordinates and electrode length should be adjusted accordingly. Two addi-
tional holes are drilled in the contralateral skull for placement of anchoring screws
(#00-90, 1/16 in). After the screws are driven into the skull, the tripolar electrode is
lowered into the brain with a stereotaxic electrode holder. Before completely insert-
ing the electrodes, a drop of cyanoacrylic glue is placed on the skull underneath the
electrode pedestal. The electrodes and pedestal are then completely lowered and the
glue is allowed to dry for several minutes. The pedestal is permanently affixed to the
skull with dental acrylic. The mice should be allowed to recover for at least four days
before conducting experiments.
8.3.2.4 Recording of Paired Stimulus Sensory Gating Evoked Potentials

Routine use of this technique requires a dedicated lab space. To minimize the influ-
ence of ambient noise, the animals are recorded in plastic shoebox cages within
acoustically isolated chambers (Med Associates) that are internally lined with
sound-absorbent foam. A mouse previously implanted with indwelling electrodes is
connected to a flexible electrical tether and swivel device, also called a commutator,
which is mounted to the ceiling of the isolation chamber. The commutator allows the
animal to rotate freely without twisting the tether. Commercially available tether and
commutator systems from Plastics One, Inc. have at least two channels, and some
models have up to 16 channels or more to record multiple brain areas simultane-
ously. For recording EEGs from one location, two channels will be required; one
is the active channel, in our example the hippocampal electrode, and the other is a
reference channel, in our example one of the cortical electrodes. We do not use the
third electrode of the tripolar configuration. The small microvolt EEG biosignals
must be amplified and filtered, thus the commutator is electrically cabled to differ-
ential AC amplifiers (Grass Instrument Division, Astro-Med, Inc., West Warwick,
Rhode Island, USA). Cortical and subcortical EEGs are typically amplified by a fac-
tor of 1000, and band pass filters are set at 1 and 300 Hz. A standard PC with large
storage capacity is used in conjunction with acquisition software (e.g., Datawave,
Inc., Berthoud, Colorado, USA) that digitizes the EEG signals at 1000 Hz, a sample
rate that captures the rapid rise and fall of auditory EP waveforms. Currently, our lab
has eight recording chambers that can record two channels of EEG per mouse, thus

EEGs from eight mice can be recorded simultaneously. This shortens the amount of
time required to record animals in a given day without reducing N size.
8.3.2.5 Auditory Stimuli

Sensory gating auditory EPs are generated by presentation of paired white noise
bursts (5 msec duration) from a wide-dynamic-range speaker mounted within the
recording chamber at a distance of approximately 15–20 cm from the mouse. Tone
bursts of 2–3 KHz have also been successfully used. The auditory stimulus can be
generated by a computer, or by an external audio generator (e.g., from Med Associ-
ates). The first conditioning auditory stimulus is followed 0.5 sec later by an identical
auditory test stimulus. The latency between the two stimuli of the pair, or inter-stim-
ulus interval (ISI), is critical. Generally, ISIs of less than 1 sec result in lowered test
stimulus EP amplitudes. While an ISI of 0.5 sec produces a consistent suppression of
the test EP response in humans and animals, ISIs of greater than 2–3 sec yield simi-
lar condition and test stimulus EP amplitudes (Figure 8.7). Thus, gating function,
initiated by sensory input, seems to gradually extinguish over time to a point where
the brain is reset to a ready-state level of maximal sensitivity to sensory input. The
length of time between stimulus pairs, or trials, is 15 sec. ITIs are generally reported
to be 10–15 sec, a range that minimizes a potential influence from the previous
stimulus pair. Clear and measurable auditory EP responses cannot be obtained from
a single stimulus because of variability of the background EEG. We present 120
paired auditory stimuli to subjects and the data acquisition computer averages each
EEG trace starting 100 msec before and 900 msec after every conditioning stimulus.
This averaging produces prototypical-evoked waveforms (potentials) with peaks at
relatively stable latencies after the auditory stimulus. Thirty minutes are required
to obtain an average EP using an ITI of 15 sec. Other labs using anesthetized mice
report using fewer stimulus repetitions to obtain auditory EPs, probably because of a
more stable background EEG under anesthesia. The volume of auditory stimuli is 65
dB, which is about 5 dB above the constant 60 dB background noise of the recording
A. 500 msec ISI, 15 sec ITI B. 2000 msec ISI, 15 sec ITI
Stim Stim
50 µV
Condition Condition
Test Test
0 100 200 300 400 500 0 100 200 300 400 500
msec msec
FIGURE 8.7 Condition and test stimulus evoked potential (EP) responses are superimposed.
The recording is from the hippocampus of a CD-1 mouse. In A, the 500 msec inter-stimulus
interval (ISI) results in an attenuated test stimulus EP amplitude. In B, no attenuation of the
test stimulus EP amplitude is seen when the ISI is increased to 2000 msec.

chambers. The use of louder intensities has been reported, but this relatively low
level yields clear averaged EPs for us without producing a startle response. The
acquisition software is set to acquire 1 sec of data starting 100 msec before and
ending 900 msec after the initial conditioning stimulus. In addition to recording the
EEG EPs, the software triggers the audio generator to synchronize the EP recording
with the stimulus.
8.3.2.6 Evoked Potential Analysis

The hippocampal auditory EP response is identified as a peak in the ongoing EEG at
a latency of 15–25 msec after the stimulus, followed by the peak of opposite polarity
at 30–50 msec after the stimulus. The difference in amplitude between these peaks
is defined as the N-40 amplitude in microvolts (μV). N-40 amplitude is determined
for both the averaged conditioning (CAMP) and test (TAMP) EPs. A sensory gating
ratio is calculated by dividing the test amplitude by the conditioning amplitude. This
calculation, termed the T:C ratio, is the index by which sensory gating is assessed
(Figure 8.8). T:C ratios in schizophrenic patients are often well above 0.4; in other
words, TAMP is greater than 40% of CAMP. In anesthetized and unanesthetized
DBA/2 mice, T:C ratios usually exceed 0.5, while in other strains such as C3H, T:
C ratios often are below 0.4 (Figure 8.8). These findings have led to a considerable
use of the DBA/2 mouse strain as a model of sensory gating deficits. On average,
we have found that about 10% of DBA/2 mice have control T:C ratios below 0.4.
Mice with low T:C ratios are considered to be clinically normal and show little or no
response to drugs that improve gating. Thus, animals with control T:C ratios below
0.4 are retrospectively dropped from studies and data analysis.
8.3.2.7 Example Drug Studies with DBA/2 Mice

In a typical study, drugs are administered immediately before mice are placed into
the isolation chambers and initiation of auditory EP recording. The duration of the
recordings can vary, but they last for 30 min in many of our studies. Multiple treat-
ments (e.g., doses), including vehicle control, are administered to each mouse on sep-
arate days with at least 48 hr between treatments. So, for example, a three-point dose
response with vehicle would require 4 recording days. With a 48-hr washout between
days, this study would take 7 days. This within-subjects design allows each mouse
to serve as its own control. For drugs with robust effects on sensory gating, such
as nicotine, an N size of eight is sufficient to demonstrate a statistically significant
effect. However, we routinely use an N size of 12 or more to detect low dose effects,
or the effects of weaker drugs. Paired t-tests for two–group or repeated measures
ANOVA for multiple-group statistical evaluation are used. For a drug study where
only the dose is varied, a one-way ANOVA is used, while for time course or multiple
drug studies, a two-way ANOVA is employed. Post-hoc analyses for between-group
comparisons are Newman-Keuls for the one-way ANOVA and Boferroni for the two-
way ANOVA.
As mentioned before, the atypical antipsychotics clozapine and olanzapine
improve sensory gating in DBA/2 mice, that is, they lower T:C ratios (Figure 8.9).15
Sensory gating deficits in schizophrenia are hypothesized to be, in part, mediated

DBA mice C3H mice
Condition Test Condition Test
N40
N40
N40 N40
Condition Test Condition Test
= 77 µV = 72 µV = 67 µV = 31 µV
Amplitude Amplitude Amplitude Amplitude
72/77 = T:C ratio 0.93 29/69 = T:C ratio 0.42

(a)
0.8
0.7
0.6
0.5
**
T:C Ratio
0.4
0.3
0.2
0.1
0.0
DBA/2 C3H
Strain
(b)
Figure 8.8 Examples of sensory evoked potentials from either DBA/2 or C3H mice. The
N-40 amplitude is measured from the positive peak about 20 msec after the stimulus (P20) to
the trough of the N-40 wave. In (a), the DBA/2 mouse has a calculated test amplitude:condi-
tioning amplitude (T:C) ratio of 0.93, while the C3H mouse has a T:C ratio of 0.42. The high
T:C ratio of the DBA/2 mouse is indicative of a sensory gating deficit. (b) shows average T:
C ratios for groups of DBA/2 and C3H mice. The DBA/2 group had significantly higher T:C
ratios. **p = 0.0022, t51 = 3.232, unpaired, two-tailed t-test.
by α-7 receptor hypofunction.31 The selective α-7 agonist GTS-21 and the nonse-
lective nicotinic agonist nicotine improve sensory gating in both DBA/2 mice and
schizophrenic patients.32–34 Thus, one of the uses of the DBA/2 mouse is to study
compounds that may have potential therapeutic effects in schizophrenia. Figure 8.10
shows the effects of the selective α-7 agonist A-582941 on sensory gating T:C ratios

0.8
0.6
**
T:C Ratio
0.4
0.2
0.0
0.0 0.3
Clozapine mg/kg IP
FIGURE 8.9 Clozapine significantly lowers T:C ratios in DBA/2 mice. **p = 0.0056, t 9 =
3.622, paired, two-tailed t-test. Source: Author’s unpublished data.
in DBA/2 mice.35 As with GTS-21, A-582941 lowers T:C ratios, an effect consistent
with improved sensory gating. Substantial literature exists on the pharmacology of
sensory gating using in vivo electrophysiological recordings of auditory EPs. It has
been particularly useful in advancing the understanding of the role of brain cholin-
ergic systems in information processing, and the role of cholinergic deficiencies in
disease states such as schizophrenia.
8.4 MOTOR FUNCTION AND SPONTANEOUS EXPLORATION

Although many experimenters view locomotor activity as an overly simplistic mea-
sure that provides only limited information, alterations in this behavior can reveal
important information on potential mechanisms of drug action. Moreover, locomo-
tor activity may influence functional outcome in animal models of CNS injury or
disease. For example, many different psychoactive drugs can act at neuronal receptor
sites and directly affect motor function. Similarly, brain injury models employed
by many researchers can produce subtle or sometimes pronounced alterations in
motor behavior. Furthermore, genetically altered animals have become popular in
an attempt at unmasking the molecular and cellular correlates of such behaviors as
learning and memory in addition to numerous disease states. These animal mod-
els, however, are not without their drawbacks, including profound changes in motor
function that can confound the interpretation of behavioral results. Therefore, it is
important for the neuroscientist to be aware of, and to characterize, these changes
carefully. The following examines several methods for assessing motor and explor-
atory activity in the adult mouse.
Locomotor and exploratory behavior may also be influenced by several other
factors such as time of day (rodents are nocturnal animals and are therefore signifi-
cantly more active during periods of darkness); anxiety (animals may be more or less
active depending on the situation to which they are exposed); state of wakefulness
or arousal (stimulants will tend to increase activity, whereas sedatives will tend to

T:C Ratio
1.0
0.8
0.6 *
0.4
0.2
0.0
0.0 1.0 3.0
A-582941 µmol/kg IP
(a)
CAMP
100
80
60
µV
40
20
0
0.0 1.0 3.0
A-582941 µmol/kg
(b)
TAMP
60
*
40
µV
20
0
0. 0 1. 0 3. 0
A-582941 µmol/kg
(c)
FIGURE 8.10 Example of the effects of a selective B7 agonist. In panel (a), A-582941 (3.0
μmol/kg ip) significantly lowers test amplitude:conditioning amplitude (T:C) ratios in DBA/2
mice, one-way repeated measures ANOVA p = 0.0158 F(2,38) = 4.955, *p < 0.05 Newman-
Keuls post-hoc test vs. vehicle. Panel (c) shows that the T:C ratio in this instance was lowered
by significantly decreasing the amplitude of the test stimulus evoked potential (TAMP), one-
way repeated measures ANOVA p = 0.0427 F(2,38) = 3.608, *p < 0.05, Newman-Keuls post-
hoc test vs. vehicle. The condition stimulus amplitude in panel (b) (CAMP) was unchanged,
one-way repeated measures ANOVA p = 0.7516 F(2,38) = 0.289. Source: Data reproduced
with permission from Bitner, R. S., Bunnelle, W. H., Anderson, D. J., et al. 2007. Broad-spec-
trum efficacy across cognitive domains by alpha-7 nicotinic acetylcholine receptor agonism
correlates with activation of ERK1/2 and CREB phosphorylation pathways. Journal of Neu-
roscience, 27:10578–87, 2007.

decrease activity, although the magnitude of these effects can often depend upon the
strain of animal used); environmental novelty (mice tend to exhibit increased explor-
atory behavior when exposed to a novel environment and decreased activity upon
reexposure to the same environment, i.e., “habituation”); motivation (food-deprived
mice may show increased activity); age (younger rodents are more active than aged
animals); general health; and genetic strain (C57BL/6 mice are more active than
129-derived animals). All of these factors necessitate careful experimental design
and it is therefore prudent to control for these and maintain consistency from the
outset. However, natural variations in activity and stress levels often exist between
mice of the same strain despite controlling for all of the factors described above,
hence the need for adequate group sizes that will accommodate appropriate statisti-
cal analyses.
8.4.1 SPONTANEOUS ACTIVITY

8.4.1.1 Open Field (Non-automated)
Perhaps the simplest and most economical method for assessing both exploratory
and locomotor activity is the open-field apparatus. As the name suggests, this gen-
erally consists of a square (or circular, depending on personal preference) arena of
adequate size (e.g., 50 × 50 cm or 50 cm diameter for mice) surrounded by walls to
prevent the animal from escaping. The box itself may be composed of either wood
or plastic, although the latter is preferred to reduce olfactory issues and for ease
in cleaning. In its simplest form, the floor is divided into equally spaced regions
by marker pen which has been allowed sufficient time to dry so as not to produce
unwanted olfactory effects. Alternatively, the box can be video monitored and lines
can be drawn on the screen to divide the arena.
8.4.1.2 Typical Protocol

1. The open field should be located in a quiet room with controlled tempera-
ture and ventilation. A low-level illumination is preferred to reduce anxiety
and thus lessen freezing behaviors, unless this is a component of the task
that you wish to study. The observer should be seated comfortably at a
distance from the apparatus, or ideally watching a monitor fed by a video
camera positioned above the open field. If visible to the behaving rodent,
the investigator should be consistent with seating position, clothing, and
potential olfactory cues.
2. If stimulant activity of a drug is to be examined, the rodent should first be
habituated to the apparatus for three or more 5-min sessions to reduce base-
line activity. Habituation should be omitted if anxiety or response to novelty
is being studied. For brain lesion or injury studies, you may wish to examine
performance at discreet times before and after surgery. Bear in mind, how-
ever, that in general, activity and/or exploration following repeated expo-
sure to the open field will decrease with habituation and cognition.
3. On the test day, administer the test drug, if required, at an appropriate time
prior to placing the rodent into the center of the open field. The investiga-

tor records the following specific behaviors using prepared data sheets and
appropriate counters over a specified period of time, usually 5–10 min.
Parameters to record: locomotion (number of square crossings within the specified

time); rearing; grooming; and stereotypical behaviors such as licking, biting, and
head weaving. These activities may be recorded separately for peripheral regions
or in the center of the arena, the latter being thought to reflect the degree of anxiety
experienced by the rodent (i.e., animals with higher levels of activity in the center
of the arena are less anxious). Defecation frequency may also be recorded as a mea-
sure of fear, but this tends to be more variable due to the relatively small numbers
involved.
8.4.1.3 Open Field (Automated)

For large studies it is impractical to directly observe each animal individually. This
can be resolved by making use of a set of automated activity boxes consisting of
arenas similar to those described above, but with regions demarcated by infrared
beams instead of marker pen. Each box is connected individually to a computer that
collates all data from up to 30 or more boxes at a time. Equipment available from
AccuScan Instruments (formerly Omnitech, Columbus, Ohio, USA; www.accuscan-
usa.com) offers flexibility in experimental design and data analysis. Although the
cost can be somewhat prohibitive for smaller laboratories (upwards of $80,000 for a
set of 16 boxes), the major advantage of such systems is that they allow the collection
of both vertical (rearing) and horizontal activity from the periphery and center of
the apparatus over time periods that could not be accurately completed by a manual
observer. For example, the computer can be programmed to accept data every 15
sec and to calculate a mean for each 1 min time bin for up 2 hr or more. Clearly the
amount of data collected rises considerably with increased time. In our laboratories,
a system of 16 boxes is employed (AccuScan Instruments) in a dedicated quiet room
with dimmed lighting. Each arena is 40 × 40 cm in size with removable clear Plexi-
glas chambers for ease of cleaning. Two sets of infrared photocells (one for detecting
rearing, the other for locomotion), are fixed to a rack that surrounds the Plexiglas and
that can be adjusted in the vertical plane to allow measurements from rats or mice.

1. If a stimulant drug-induced increase in activity is expected, habituate the
mice to the apparatus for several 1-hr sessions. Do not habituate for drugs
expected to decrease activity.
2. Program the computer to record activity as desired. An example would be
to bin data every 5 min for up to 1 hr and to distinguish horizontal activity
from vertical in both the peripheral and central regions of each arena.
3. Administer test drug at the appropriate time. If decreased activity is
expected, inject before placing mice into the center of the arena. Con-
versely, if increased activity is expected, place mice into the center of the
arenas and allow for habituation to the novel environment for at least 30
min before drug administration. If animals are subjected to brain injury or

other surgery, allow sufficient time for recovery (at least 24 hr, if not more,
depending on severity) before placement into the arenas.
4. Record data for a predetermined period of time, usually 30–120 min. Print
out all raw data as a hard copy backup and convert the data file produced into
a form suitable for analysis using software programs such as Microsoft Excel
(Microsoft Corporation, Seattle, Washington, USA; www.microsoft.com)
and JUMP (SAS Institute, Cary, North Carolina, USA; www.sas.com).
Arenas employed by more advanced systems (AccuScan Instruments; www.accus-

can-usa.com) can be subdivided into zones both physically by using a Plexiglas
insert and virtually by specifying different software parameters. This allows many
additional animals to be assessed simultaneously. In addition, this system can allow
measurements of distance traveled, movement time, rearing duration, etc. If the
additional cost of such systems is prohibitive, but something more sophisticated than
the manual version of the open field apparatus is desired, other equipment options
are available. Software for video tracking systems that identify images in contrast
to the background and track a center point for movement, and which have been used
for tracking in water mazes, are available for multiple chambers for analysis. These
can produce measures of distance and orientation over time, although rearing activ-
ity may have to be recorded separately. In addition, there are more sophisticated
video systems that track the directionality of the animal and other behaviors such
as weaving and stereotypy. Therefore, if cost is an important factor, the investigator
should determine whether existing equipment could be easily modified to measure
motor activity. In addition, some photocell systems have been designed to measure
activity in the home cage, where response to a novel environment is undesirable,
or for monitoring over the light-dark cycle. Such systems tend to come in specially
designed cages, and racks and can be expensive. However, there is now a very com-
petitive market for analyzing locomotor activity, and this has led to a reduction in the
price of systems, allowing a small lab to tailor equipment according to need. A list of
vendors is included at the end of the chapter.
8.4.1.5 Variations
Locomotor activity can also provide indices of learning and memory and anxiety.
Habituation of locomotor activity in a novel environment can be used to assess
memory in mice.36 For this procedure, the mouse is briefly exposed (e.g., 5 min) to
a novel open field and locomotor activity is assessed. Memory for the novel experi-
ence is then tested at a later time by reexposing the mouse to the same open field.
Activity during the second exposure is used as an index for assessing memory, with
lower activity being indicative of better memory for the open field. Of course, it is
important that the treatments evaluated with this method do not have direct effects
on locomotor behavior. One way to minimize the effects of the drug is to treat imme-
diately following the first activity session.
The pattern of exploration can also be an important index of anxiety. Informal
assessment of anxiety can be derived by comparing time spent in the periphery of
the arena relative to time spent in the center. Anxious animals tend to spend more

time in the periphery. In addition, initial freezing in an open field is an index of

anxiety, so the latency to move a given distance (or to move through a given number
of squares) can also be used to assess fear and anxiety. A more formal assessment of
anxiety can be made using a modified open-field apparatus. The open field is sepa-
rated into a well-lit area and a dark area and the relative time and activity in these
two zones is compared. Anxiolytics increase the time spent in the well-lit zone in
this light–dark test.37
8.4.1.6 Sample Experiment

Data from a typical automated experiment are presented in Figure 8.11. In this
study, vertical (or rearing) (Figure 8.11A) and horizontal (Figure 8.11B) activity was
assessed for three mouse strains, with data collected in 5-min bins for 30 min.38 Note
that BALB/c mice appear significantly less active than animals from the other strains.
Note also the habituation response indicated by decreased activity over time for most
groups. It is important that appropriate statistical methods be used for analysis of
behavioral data. Locomotor data are generally normally distributed so parametric
ANOVAs are used routinely. A repeated measures ANOVA should be considered in
most cases when a time course is employed. Post-hoc tests that examine the mean
square error relative to the overall analysis (e.g., Tukey’s) can then be used for mul-
tiple comparisons between groups. Individual t-tests should not be used unless they
are corrected for multiple comparisons. For the data presented in Figure 8.11B, a
repeated measures ANOVA yielded a significant group effect [F(2,26) = 4.382, p <
0.0229], indicating overall differences between the strains in the study; time effect
[F(5,130) = 126.103, p < 0.0001], reflecting the decreased activity with time (habitua-
tion) overall; and group × time interaction [F(10,130) = 3.049, p < 0.0017], indicating
significant differences between groups over time. Post-hoc analysis with Tukey’s
pairwise comparisons detected significant differences for the 5-, 10-, 15-, and 30-
min time points between C57BL/6 and BALB/c mice (p < 0.01).
8.4.1.7 Motor Function

Motor function can be differentially affected depending on experimental parameters.
For example, unilateral brain injury models often produce hemiparesis-like effects,
which may be reflected by deficits in grip strength, balance, and turning behavior, or
may induce forepaw flexion. Many drugs can have either sedative or stimulant prop-
erties. Consequently, several models have been developed to examine specific motor
deficits such as these. Two commonly used procedures are thus described.
8.4.2 ROTAROD
The ability of a rodent to maintain balance and keep pace with a rotating rod has
been used with varying degrees of success over the years to assess motor function.
Several versions of this test (commonly referred to as the rotarod test) have been
described over the years. Most require the mouse to walk on a rotating rod of fixed
diameter (3.5 cm for the apparatus we use) that increases in speed over a prede-
termined period of time until the animal can no longer maintain its position. The

110
100 C57BL/6
Mean no. Beam Breaks

90 CD-1
BALB/c
80
70
60
50
40
30
5 10 15 20 25 30
Time (min)
(a)
800
700 C57BL/6
Mean no. Beam Breaks
CD-1
600
BALB/c
500
400
300
200
100
5 10 15 20 25 30
Time (min)
(b)
FIGURE 8.11 Automated measure of vertical (a) and horizontal (b) activity in the same
animals from three different mouse strains. Data were collected over 5-min intervals for a
total of 30 min. Most mice habituated to the test environment, as evidenced by the decline
in activity over the duration of the experiment. Note that BALB/c mice were less active than
mice from the other two strains. (Statistical significance described in detail in main text.)
Source: Author’s unpublished data.
rotarod apparatus employed in our laboratories consists of a central drive rod con-
nected to a stepper motor (AccuScan Instruments) that is divided into four separate
testing stations. The speed at which the rod rotates can be accelerated up from 0 rpm
to over 100 rpm over a set time period. Other rotarod models are available and can
be found in the vendor list at the end of the chapter.

1. Administer drug, as appropriate. For lesioned or injured animals, wait at
least 24 hr following the surgical procedure.
2. Set the apparatus to accelerate from 0–40 rpm over 60 sec. This is a good
standard for young adult mice, although it should be noted that juvenile and
older animals perform poorly at this task.

3. Place four mice on the rotarod, one per testing station, and then start the
stepper motor and timer. Many models come equipped with a timer that
begins automatically when the motor is switched on and stops when the
animal falls down to the floor of the apparatus, as detected by interruption
of an infrared beam.
4. As the speed increases, the mouse is required to walk faster to remain in a
stationary position. The latency to fall from the rotating rod is determined
and taken as a measure of motor function. It is generally a good idea to take
the mean of at least two to three measures from each animal.
8.4.2.2 Variation
Some investigators39,40 modify the rod itself by enclosing the core of the rod with a
series of stainless steel bars of a specific diameter (Figure 8.12A). In this instance the
time either to fall (Figure 8.12B) or to cling and make two full rotations is recorded
as the outcome measure. This design may offer some advantages over the more tra-
ditional, relatively smooth rod in that data, particularly in brain injury studies, may
be more consistent within groups. With rodent strains that exhibit a poor baseline
performance in this task, it is usually beneficial to pre-train these animals at least
two to three times before commencing the study proper.
8.4.2.3 Sample Experiment

Data from a typical experiment are presented in Figure 8.12C, where the effect of
sham surgery and controlled cortical impact (CCI) brain injury on time spent on
the rotating rod is shown for three mouse strains. Sham operated controls exhib-
ited a stable performance over the 4 wk of testing, whereas a decrease in time
spent on the rotarod device was observed in injured mice from all strains for up to
7 days following injury. Because of these pronounced deficits, it would be unwise
to conduct cognitive experiments with a significant motor component (e.g., Morris
water maze) during this time period. Once again a repeated measures ANOVA is
appropriate for comparing groups over time as rotarod data tends to be normally
distributed and this test was conducted repeatedly over a 4-wk period in this study.
A significant group effect [F(5,57) = 16.601, p < 0.0001], indicating overall differ-
ences between the strains in the study; time effect [F(7,399) = 47.183, p < 0.0001],
reflecting the attenuation of the deficits with time in the injured groups; and group
× time interaction [F(35,399) = 6.480, p < 0.0001], indicating significant differ-
ences between groups over time, were observed. Using a post-hoc test (Tukey’s
pairwise comparison), a significant impairment was detected among CCI injured
mice from all three strains when compared with their respective surgery controls
on days 1, 2, and 3 (p < 0.05) following surgery. A one-way ANOVA would be
suitable for comparing these groups if no time component was involved. A t-test
may also be appropriate in such instances. No significant difference was observed
between strains for either treatment group.

(a) (b)
42
40
38
Time on Rotarod (sec)
36
*
34 Sham 129/SvEMS
* Sham C57
32 Sham FVB/N
*
30 CCI 129/SvEMS
28 CCI C57
CCI FVB/N
26
0 1 2 3 7 14 21 28
Days Post-Surgery
(c)
FIGURE 8.12. The effect of moderate controlled cortical impact (CCI) brain injury on
rotarod performance. As the device gradually ramped up to speed (35 rpm), the mouse was
required to walk faster to maintain a stationary position on the rod, which has been modified
here to include a series of stainless steel bars (a). When the mouse can no longer keep up with
the speed of rotation, it either falls from the bars (b) or clings tight and begins to rotate with
the rod (not shown). Uninjured or sham-operated mice are generally adept at this task. How-
ever, a significant deficit can be seen for approximately 7 days following CCI brain injury,
shown for three different mouse strains in (c). Photos depict an adult male C57BL/6 mouse.
8.4.3 BEAM BALANCE/WALKING

While the rotarod is useful for determining gross motor deficits in the rodent, the
detection of more subtle motor effects requires a different approach. Fine motor

coordination, for example, can be assessed using a beam walking or balance task.
This test essentially examines the ability of the animal to remain upright and to
walk on an elevated and relatively narrow beam (Figure 8.13A) without falling to the
cushioned pads below or slipping to one side of the beam. Again, unilateral brain
injury models tend to induce a hemiparesis-like effect, which can cause the rodent to
slip to one side, usually contralateral to the injury site (Figure 8.13B).

1. For mice, set up a beam approximately 0.6 cm wide and 120 cm in length,
suspended about 60 cm above some foam pads. (A larger beam, approxi-
mately 1.8 cm wide and 240 cm in length, in addition to a flat platform at
one end to rest between trials is required for rats.)
2. Place the animal on one end of the beam (for the rat this would be farthest
from the platform). Animals from active strains such as the C57BL/6 mouse
or the Long-Evans rat will instinctively walk along the beam to reach the
opposite end. Once at this point they will generally turn 180° and continue
to walk on to the opposite end. Establish a basal level of performance before
surgery or treatment, and allow sufficient time for recovery (at least 24 hr)
before retesting.
3. Count the number of foot faults, defined as the number of times the fore-
paws and/or hindpaws slip from the horizontal surface of the beam over
a predetermined number of steps (50 is usually adequate). Allow the per-
forming animal sufficient time (approximately 5 min) to complete this task.
(It is useful to use a mirror on the side of the beam opposite the observer
and to videotape the performance for scoring.)
4. Remove to home cage and retest as appropriate. It should be noted that
rodents, especially rats, tire and are reluctant to move if exposed to this test
repeatedly over a short period on the same day.
8.4.3.2 Variation
This task works well for active rodent strains and may not be suitable for less active
animals. Another variation partly designed to address this issue in the rat involves
training animals to walk across the beam to a “safe” dark box; the cognitive require-
ments for this version, however, may influence motor outcome to some degree so care
should be taken here. A simpler approach measures the time taken to fall down onto
the foam pads. In this instance, the investigator should vary the beam width until
an acceptable latency is found for the particular strain to be used. Attention should
also be paid to the body weight of the animal, as the suitable width of the beam may
change according to the mouse’s ability to grip the edge of the beam, for example,
mice heavier than 35 g generally require a beam approximately 0.9-cm thick.
8.4.3.3 Example Experiment

Data from a typical experiment are reproduced in Figure 8.13C. In this experiment,
adult C57BL/6 mice were subjected to mild (4.5 m/s) or moderate (6.0 m/s) unilateral

(a)
(b)
50
CCI 6.0 m/s
Mean Number Footfaults
40
30
CCI 4.5 m/s
20
10
Sham
0
0 1 2 3 7 14 21 28
Days Post-surgery
(c)
FIGURE 8.13 The effect of moderate controlled cortical impact (CCI) brain injury on
beam walking performance for the C57BL/6 mouse. Surgery-naive or sham-operated mice
perform well on this task, traversing the beam several times, gripping its horizontal edge with
the innermost digits (arrow in (a) illustrates this point). Foot faults, defined as forelimb and/or
hind limb slipping from the horizontal surface of the beam, (arrow in (b) shows contralateral
hind limb slipping down the side of the beam) are generally counted over a total of 50 steps;
a foot fault frequency of 15% or less is normal for control mice from this strain (c). However,
mice subjected to mild (low velocity, 4.5 m/s) or moderate (higher velocity, 6.0 m/s) unilateral
CCI brain injury exhibit a highly significant deficit (statistical significance described in detail
in main text) in this task, which is dependent on injury severity and persists for an extended
period (c). Source: Data reproduced with permission from Fox, G. B., et al. 1998. Sustained
sensory/motor and cognitive deficits with neuronal apoptosis following controlled cortical
impact brain injury in the mouse. J. Neurotrauma 15(8):599–614.

CCI brain injury, and the number of contralateral hind limb foot faults were recorded
over a 4-wk period. An obvious deficit, dependent on injury severity, was observed
when compared with sham-operated controls. For statistical analysis, beam-walk-
ing data are generally normally distributed so parametric ANOVAs are advised. A
repeated measures ANOVA should be considered in most cases when a time course
such as that presented above is employed. Post-hoc tests that examine the mean
square error relative to the overall analysis (e.g., Tukey’s) can then be used for mul-
tiple comparisons between groups. Individual t-tests should not be used. For the data
presented in Figure 8.13C, a repeated measures ANOVA yielded a significant group
effect [F(2,33) = 94.265, p < 0.0001], indicating overall differences between the dif-
ferent treatment groups in the study; time effect [F(7,231) = 89.383, p < 0.0001],
indicating significant overall changes in performance over the duration of the study;
and group × day interaction [F(14,231) = 20.995, p < 0.0001], indicating significant
performance differences between groups over time. Post-hoc analysis with Tukey’s
pairwise comparisons detected significant differences for days 1–28 between sham
controls and CCI-injured mice from both groups (p < 0.001). There were no signifi-
cant differences between groups before injury (day 0; p > 0.05).
REFERENCES
1. Davis, M. 1980. Neurochemical modulation of sensory-motor reactivity: Acoustic and
tactile startle reflexes. Neurosci. Biobehav. Rev. 4(2):241–63.
2. Davis, M. 1984. The mamalian startle response. In Neural mechanisms of startle
behavior, ed. R.C. Eaton, 287–351. New York: Plenum Press.
3. Koch, M., and Schnitzler, H. U. 1997. The acoustic startle response in rats—circuits
mediating evocation, inhibition and potentiation. Behav. Brain Res. 89(1–2):35–49.
4. Braff, D. L., et al., 2001. Impact of prepulse characteristics on the detection of senso-
rimotor gating deficits in schizophrenia. Schizophr. Res. 49(1–2):71–8.
5. Swerdlow, N. R. 1996. Cortico-striatal substrates of cognitive, motor and sensory
gating: Speculations and implications for psychological function and dysfunction. In
Advances in biological psychiatry, Vol. 2, ed. Panksepp, J. Greenwich, CT: JAI Press.
6. Alder, L. E., et al. 1982. Neurophysiological evidence for a defect in neuronal mechanisms
involved in sensory gating in schizophrenia. Biological Psychiatry 17(6):639–54.
7. Boutros, N. N., Zouridakis, G., and Overall, J. 1991. Replication and extension of P50
findings in schizophrenia. Clinical EEG (electroencephalography) 22(1):40–45.
8. Freedman, R., Alder, L. E., Myles-Worsley, M., et al., 1996. Inhibitory gating of an
evoked response to repeated auditory stimuli in schizophrenic and normal subjects:
Human recordings, computer simulation, and an animal model. Archives of General
Psychiatry 53(12):1114–1121.
9. Freedman, R., Alder, L. E., Waldo, M. C., Pachtman, E., and Franks, R. D., 1983. Neu-
rophysiological evidence for a defect in inhibitory pathways in schizophrenia: Com-
parison of medicated and drug-free patients. Biological Psychiatry 18(5):537–551.
10. Bickford, P. C., Luntz, L. V., and Freedman, R. 1993. Auditory sensory gating in the rat
hippocampus: Modulation by brainstem activity. Brain Research 607(1-2):33–38.
11. Adler, L. E., Rose, G., and Freedman, R. 1986. Neurophysiological studies of sensory
gating in rats: Effects of amphetamine, phencyclidine, and haloperidol. Biological Psy-
chiatry 21(8–9):787–798.
12. Light, G. A., Malaspina, D., Geyer, M. A., et al., 1999. Amphetamine disrupts P50 sup-
pression in normal subjects. Biological Psychiatry 46(7):990–996.

13. Stevens, K. E., Freedman, R., Collins, A. C., et al. 1996. Genetic correlation of inhibi-
tory gating of hippocampal auditory evoked response and alpha-bungarotoxin-binding
nicotinic cholinergic receptors in inbred mouse strains. Neuropsychopharmacology:
15(2):152–162.
14. Light, G. A., Geyer, M. A., Clementz, B. A., Cadenhead, K. S., and Bratt, D. L. 2000.
Normal P50 suppression in schizophrenia patients treated with atypical antipsychotic
medications. The American Journal of Psychiatry 157(5):767–771.
15. Simosky, J. K., Stevens, K.E., and Freedman, R. 2002. Nicotinic agonists and psycho-
sis: Current drug targets. CNS and Neurological Disorders 1(2):149–162.
16. Hunter, K. P., and Willott, J. F. 1993. Effects of bilateral lesions of auditory cortex in
mice on the acoustic startle response. Physiology & Behavior 54(6):1133–39.
17. Weiss, G. T., and Davis, M. 1976. Automated system for acquisition and reduction of
startle response data. Pharmacol. Biochem. Behav. 4(6):713–20.
18. Crawley, J. N., et al., 1997. Behavioral phenotypes of inbred mouse strains: Impli-
cations and recommendations for molecular studies. Psychopharmacology (Berl.)
132(2):107–24.
19. Olivier, B., et al., 2001. The DBA/2J strain and prepulse inhibition of startle: A model
system to test antipsychotics? Psychopharmacology (Berl.) 156(2–3):284–90.
20. Zhang, M., et al., 2006. Effect of dopamine D3 antagonists on PPI in DBA/2J mice or
PPI deficit induced by neonatal ventral hippocampal lesions in rats. Neuropsychophar-
macology 31(7):1382–92.
21. Geyer, M. A., and Braff, D. L. 1987. Startle habituation and sensorimotor gating in
schizophrenia and related animal models. Schizophr. Bull. 13(4):643–68.
22. Geyer, M. A., et al., 1990. Startle response models of sensorimotor gating and habitua-
tion deficits in schizophrenia. Brain Res. Bull. 25(3):48598.
23. Paylor, R., and Crawley, J. N. 1997. Inbred strain differences in prepulse inhibition of
the mouse startle response. Psychopharmacology (Berl.) 132(2):169–80.
24. Adler, L. E., Waldo, M. C., and Freedman, R. 1985. Neurophysiologic studies of sen-
sory gating in schizophrenia: Comparison of auditory and visual responses. Biological
Psychiatry 20(12):1284–1296.
25. Jin, Y., and Potkin, S. G. 1996. P50 changes with visual interference in normal sub-
jects: A sensory distraction model for schizophrenia. Clinical EEG (electroencepha-
lography) 27(3):151–154.
26. Lebib, R., Papo, D., de Bodes, S., and Bandonnière, P. M. 2003. Evidence of a visual-
to-auditory cross-modal sensory gating phenomenon as reflected by the human P50
event-related brain potential modulation. Neuroscience Letters 341(3):185–188.
27. Moxon, K. A., Gerhart, G. A., Brickford, P. C., et al. 1999. Multiple single units and
population responses during inhibitory gating of hippocampal auditory response in
freely-moving rats. Brain Research 825(1–2):75–85.
28. Moxon, K. A., Gerhart, G. A., Gwinetto, M., and Alder, L. E. 2003. Inhibitory control
of sensory gating in a computer model of the CA3 region of the hippocampus. Biologi-
cal Cybernetics 88(4):247.
29. Ellenbroek, B. A. 2004. Pre-attentive processing and schizophrenia: Animal studies.
Psychopharmacology 174(1):65–74.
30. Miyazato, H., et al., 1995. A middle-latency auditory-evoked potential in the rat. Brain
Research Bulletin 37(3):247–255.
31. Martin, L. F., Freedman, R., and Anissa AbiDargham, G., and Olivier. 2007. Schizo-
phrenia and the 7 Nicotinic Acetylcholine Receptor. International review of neurobiol-
ogy. ed. 78:225–246.
32. Alder, L. E., et al., 1993. Normalization of auditory physiology by cigarette smoking in
schizophrenic patients. American Journal of Psychiatry 150:185–61.

33. Olincy, A., Harris, J. G., Johnson, L. L. et al., 2006. Proof-of-concept trial of and 7
nicotinic agonist in schizophrenia. Arch. Gen. Psychiatry 63(6):630–638.
34. Stevens, K. E., and Wear, K. D. 1997. Normalizing effects of nicotine and a novel nico-
tinic agonist on hippocampal auditory gating in two animal models. Pharmacology
Biochemistry and Behavior 57(4):869–874.
35. Bitner, R. S., Bunnelle, W. H., Anderson, D. J., et al. 2007. Broad-spectrum efficacy
across cognitive domains by alpha-7 nicotinic acetylcholine receptor agonism corre-
lates with activation of ERK1/2 and CREB phosphorylation pathways. Journal of Neu-
roscience 27(39):10578–10587.
36. Platel, A., and Porsolt, R. D. 1982. Habituation of exploratory activity in mice: A screen-
ing test for memory enhancing drugs. Psychopharmacology (Berl.) 78(4): 346–52.
37. Costall, B., et al., 1988. Actions of buspirone in a putative model of anxiety in the
mouse. J. Pharm. Pharmacol. 40(7):494–500.
38. Fox, G. B., LeVasseur, R. A., and Faden, A. I. 1999. Behavioral responses of C57BL/6,
FVB/N, and 129/SvEMS mouse strains to traumatic brain injury: Implications for gene
targeting approaches to neurotrauma. J. Neurotrauma 16(5): 377–89.
39. Fox, G. B., et al. 1998. Sustained sensory/motor and cognitive deficits with neuro-
nal apoptosis following controlled cortical impact brain injury in the mouse. J. Neu-
rotrauma 15(8):599–614.
40. Hamm, R. J., et al. 1994. The rotarod test: An evaluation of its effectiveness in assess-
ing motor deficits following traumatic brain injury. J. Neurotrauma 11(2):187–96.

APPENDIX: EQUIPMENT SUPPLIERS
AccuScan Instruments, Inc. Columbus Instruments

5098 Trabue Road 950 N. Hague Avenue
Columbus, OH 43228 USA Columbus, OH 43204 USA
Tel: 614-878-6644; 800-822-1344 Tel: 614-276-0861; 800-669-5011
(USA/Canada) Fax: 614-276-0529
Fax: 866-650-8265 Email: sales@colinst.com
E-mail: sales@accuscan-usa.com
Kinder Scientific
Coulbourn Instruments Michael Kinder, President
7462 Penn Drive 12655 Danielson Court, Suite 308
Allentown, PA 18106 USA Poway, CA 92064 USA
Tel: 610-395-3771 Tel: 858-679-1515
E-mail: sales@coulbourn.com Fax: 858-679-4811
E-mail: mkinder@hamiltonkinder.com
Clever Sys., Inc. Noldus Information Technology, Inc.

11425 Isaac Newton Square, Suite 202 1503 Edwards Ferry Road, Suite 201
Reston, VA 20190 USA Leesburg, VA 20176 USA
Tel: 703-787-6946 Tel: 703-771-0440; 800-355-9541
Fax: 703-757-7467 Fax: 703-771-0441
www.cleversysinc.com E-mail: info@noldus.com
Source: Data reproduced with permission from Fox, G. B., LeVasseur, R. A., and Faden, A. I. 1999.
Behavioral responses of C57BL/6, FVB/N, and 129/SvEMS mouse strains to traumatic brain
injury: Implications for gene targeting approaches to neurotrauma. J. Neurotrauma 16(5):
377–89.

9 Intravenous Drug
Self-Administration in
Nonhuman Primates
Leonard L. Howell and William E. Fantegrossi
CONTENTS
9.1 Introduction................................................................................................. 179
9.2 Surgical Procedures .................................................................................... 180
9.3 Schedules of Reinforcement........................................................................ 181
9.3.1 Initial Training................................................................................. 181
9.3.2 Fixed-Ratio Schedules ..................................................................... 182
9.3.3 Fixed-Interval Schedules ................................................................. 182
9.3.4 Second-Order Schedules.................................................................. 183
9.3.5 Progressive-Ratio Schedules............................................................ 183
9.4 Research Application and Data Interpretation............................................ 184
9.5 Discussion ................................................................................................... 191
References.............................................................................................................. 194
9.1 INTRODUCTION
The abuse of psychoactive drugs such as cocaine and heroin has spanned several
decades and continues to be widespread in the United States. Currently, research
efforts have focused on the development of therapeutics to treat drug abuse. Drug
self-administration studies have done much to help us understand the behavioral
and pharmacological mechanisms underlying drug abuse. An understanding of these
mechanisms will in turn aid in the development of effective therapeutic agents.
Important to the study of drug effects on behavior is the understanding that
drugs can function as stimuli to control behavior.1 Based on the principles of oper-
ant conditioning, presentation of a stimulus as a consequence of behavior may either
increase or decrease the probability that a behavior will occur again.2 If the presen-
tation of a stimulus increases the probability that a behavior will recur, then that
stimulus is defined as a positive reinforcer.2 Stimuli such as food and water function
as positive reinforcers, and data from self-administration studies indicate that most
drugs of abuse, most notably psychostimulants and opioids, can also function as pos-
itive reinforcers under the appropriate schedule contingencies. Drug self-administra-
tion procedures in animals have been used extensively to evaluate the reinforcing
effects of drugs. The first studies examining the reinforcing effects of drugs in the
179

1950s and 1960s focused on morphine self-administration in morphine-dependent

animals.3,4 Later studies demonstrated that dependence was not necessary to initiate
self-administration.5,6 Since these early studies, self-administration procedures have
been used as a model of drug taking that can be studied under controlled laboratory
conditions and applied to human drug use.
Drug self-administration studies in animals have contributed substantially to our
knowledge of the neuropharmacological mechanisms controlling drug abuse. For
example, studies with opioids have shown that drugs with high affinity for μ-opioid
receptors function as positive reinforcers,7,8 whereas opioids with high affinity for
P-opioid receptors generally do not.8,9 Additionally, the self-administration of psy-
chomotor stimulants and opioids has been found to be affected by the administra-
tion of antagonists either systemically or centrally (cf.10,11). If administration of an
antagonist shifts the dose-response function for a self-administered drug to the right,
then it can be assumed that the site of action of the antagonist is important for the
reinforcing effects of the drug.12–15 In addition to the administration of antagonists,
self-administration of drugs has been affected by lesions of certain brain neurotrans-
mitter systems (cf.10,11). Studies have also found that animals will self-administer
drugs directly into certain brain areas, suggesting a neuropharmacological mecha-
nism for their reinforcing effects (cf.10).
Over the years, drug self-administration procedures in animals have been found
to be valid and reliable for determining the abuse liability of drugs in humans. It
is well established that animals will self-administer most drugs that are abused by
humans.16,17 In particular, studies in nonhuman primates have made a significant
contribution to the field of drug abuse research. Nonhuman primates are ideal sub-
jects since they are phylogenetically more closely related to humans than are other
species.18 Potential species differences in drug metabolism also illustrate the impor-
tance of nonhuman primate models in substance abuse research. Thus, we can apply
information obtained from nonhuman primates to problems of human drug abuse
with greater accuracy. This chapter discusses methods of self-administration in non-
human primates, including different preparations and schedules of drug reinforce-
ment. While the focus is primarily on self-administration of psychoactive stimulants
such as cocaine, the methodology and general principles apply to other pharmaco-
logical classes including opiates, benzodiazepines, and alcohol.
9.2 SURGICAL PROCEDURES

Protocols for intravenous drug self-administration require the surgical implantation
of a chronic intravenous catheter to permit infusion of the drug solution. Typically,
a superficial vessel, such as the external jugular or femoral vein, is accessed via a
surgical cut-down procedure.19 Using appropriate anesthesia, either inhaled anes-
thetics (e.g., isoflurane) or injected ketamine in combination with a benzodiazepine,
and under aseptic conditions, one end of a catheter is implanted into the vessel while
the other end is routed subcutaneously to a point of access. If the distal end of the
catheter is externalized, an appropriate jacket is used to prevent the animal from
damaging the preparation,20 and the catheter is sealed with a stainless-steel obtura-
tor when not in use. An alternative means of access involves the attachment of the

Intravenous Drug Self-Administration in Nonhuman Primates 181
distal end of the catheter subcutaneously to a vascular access port.21 A Huber needle
designed to minimize insult to the skin or port membrane is inserted perpendicular
to the port to allow for injection of drug solution. Lastly, a tethering system can be
used to protect the catheter while providing convenient access.5,22 The preparation
requires continuous housing in an experimental chamber, and restraint by a harness
and a spring arm attached to the top or back of the chamber. However, movement of
the animal within the chamber is not restricted by the tethering system. The distal
end of the catheter is routed subcutaneously to exit between the monkey’s scapulae,
and is threaded through the spring arm. For each of the preparations, the catheter is
connected via plastic tubing to a motor-driven syringe located outside the test cham-
ber during experimental sessions. At least twice weekly, catheters are flushed with
sterile saline or water, and filled with heparinized saline (100 units of heparin per
mL of saline). All solutions that come in contact with the catheter are prepared with
sterile components and stored in sterile glassware.
9.3 SCHEDULES OF REINFORCEMENT

9.3.1 INITIAL TRAINING
Drug self-administration involves operant behavior that is reinforced and maintained
by drug delivery. Animals acquire drug injections by emitting a discrete response,
such as pressing a lever or key. The number and pattern of responses required for
each injection are defined by the schedule of reinforcement. Availability of drug
under a given schedule typically is signaled by an environmental stimulus, such as
the illumination of a stimulus light located proximal to the response lever. Schedule
parameters and stimulus conditions are controlled by computers, while responses
emitted by the animal are recorded simultaneously. The primary dependent mea-
sures are number of drug injections and rate of responding during each session. As
with behavior maintained by nondrug reinforcers such as food, responding main-
tained by drug injections is determined by the schedule parameters and the behav-
ioral history of the animal.
Daily experimental sessions are conducted in the home cage or in a standard
primate chair either custom designed22 or commercially available.24,25 If a primate
chair is used, location of the chair within a ventilated, sound-attenuating chamber
will minimize distractions and interference from daily laboratory activities. Typi-
cally, responding is initiated using a 1-response, fixed-ratio schedule so that each
response in the presence of a stimulus light will result in the intravenous injection
of a drug solution. The drug dose is determined by the concentration and volume
of solution, and should be sufficient to maintain reliable drug self-administration in
a well-trained animal. The saliency of the drug injection is enhanced by a change
in the stimulus lights during the injection period. It is critical to avoid excessive
drug intake and toxicity during training sessions. Drug intake can be limited by
scheduled timeout periods following each injection, during which stimulus lights
are extinguished and responding has no scheduled consequence. Defined limits
on the number of injections per session are also recommended. Once the animal
has acquired the lever press response and behavior is reliably maintained by drug

injections, the response requirement can be gradually increased under a variety of

intermittent schedules of reinforcement.
9.3.2 FIXED-RATIO SCHEDULES

The most basic schedule of reinforcement is the fixed-ratio (FR) schedule, which
defines the number of responses required per drug injection. Once responding on the
lever is engendered, the response requirement is gradually increased to the terminal
value. Nonhuman primates rapidly acquire drug self-administration behavior under
FR schedules, and stable daily performances can be obtained in several weeks. FR
schedules typically generate high response rates and a “break and run” pattern of
responding characterized by a brief pause in responding after each drug injection,
followed by an abrupt change to a steady high rate of responding until the next FR
is completed. It is important to note that total session intake of drug is a direct func-
tion of response rate under FR schedules. Typically, higher unit doses are required
to maintain behavior at higher FR values. Drug intake can be limited by scheduling
timeouts following each injection and by restricting the total number of drug injec-
tions per session. The duration of the timeout value following each injection can also
have significant effects on behavior.
9.3.3 FIXED-INTERVAL SCHEDULES

In contrast to FR schedules, fixed-interval (FI) schedules are time based and specify
a minimal inter-injection interval. They represent a suitable alternative to FR sched-
ules because they engender high levels of behavioral output. Typically, a stimulus
light is illuminated in the test chamber during the FI to serve as a discriminative
stimulus. Once the FI has elapsed, a single response is required for drug delivery.
A limited hold can be imposed following the FI to restrict the time period dur-
ing which a response is reinforced, resulting in higher rates of responding. Tem-
poral control over behavior is enhanced if a timeout is scheduled following each
injection. Once responding on the lever is engendered at a very short FI (1–5 sec),
the interval is gradually increased to the terminal value. In contrast to the “break
and run” pattern engendered by FR schedules, FI schedules engender a “scalloped”
pattern of responding characterized by little or no responding early in the interval
and increased rates of responding as the interval elapses. Nonhuman primates often
require several months of training before a stable pattern of responding develops.
Also, response rate can vary markedly with little or no change in the total number of
injections per session. It is important to emphasize that quantitative aspects of self-
administration are dictated by the schedule of reinforcement. For example, nicotine
maintained i.v. self-administration under an FR schedule in rhesus monkeys, but at
response rates much lower than those maintained by cocaine.26 In contrast, nicotine
maintained i.v. self-administration in squirrel monkeys under an FI schedule with
peak response rates much more similar to those maintained by cocaine. Hence, the
apparent strength of nicotine to maintain behavior was markedly influenced by the
schedule of reinforcement.

9.3.4 SECOND-ORDER SCHEDULES

Environmental stimuli that are paired with reinforcement can substitute for the
reinforcer itself through Pavlovian conditioning. Moreover, these conditioned
stimuli can reinforce behavior that results in their presentation through the process
of second-order conditioning.27 When a stimulus light that has been paired with
drug injection is presented following an operant response, the frequency of that
response increases.28,29 Second-order schedules involve a contingency arrangement
under which a series of responses under a schedule of conditioned reinforcement is
treated as a unit response under a second schedule that is simultaneously in effect.
Second-order schedules of drug self-administration can generate very high behav-
ioral outputs for a single injection of drug. If the stimuli are presented early in the
drug-taking history of the animal, they can also enhance the acquisition of drug
self-administration. In a typical example of a second-order schedule, responding
is initiated using a 1-response FR schedule so that each response in the presence of
a stimulus light (e.g., red) will produce an intravenous drug injection and the brief
illumination of a different stimulus light (e.g., white), followed by a timeout. The
ratio value is gradually increased as responding increases. When the schedule value
reaches a terminal value, drug injection no longer follows completion of each FR
and, instead, is arranged to follow an increasing number of FR components during
a predetermined interval of time. As the interval duration is extended during train-
ing, a greater number of FR components will be completed per drug injection. Ulti-
mately, the terminal schedule will arrange for drug injection following the first FR
component completed after the FI has elapsed. Drug administration is accompanied
by a change in the stimulus light (e.g., from red to white), followed by a timeout.
The drug-paired stimulus light is also presented briefly upon completion of each
FR component. Daily sessions can consist of several consecutive FIs depending on
the interval and session duration. By using this second-order procedure and limit-
ing the daily session to approximately 1 hr, any direct effects the self-administered
drug might have on rate and pattern of responding will be absent during the first
component and minimized during the experimental session. Hence, performance
measures can be related directly to the reinforcing effects of the drug. Cumulative
records typical of performance engendered by a second-order FI schedule with FR
components nicely illustrate that introducing an imbedded schedule of conditioned
reinforcement results in much higher and persistent rates of self-administration with
the same dose of cocaine and the same FI value.30
9.3.5 PROGRESSIVE-RATIO SCHEDULES

Progressive-ratio self-administration procedures are designed to quantify the
reinforcing effects of drugs and to determine their reinforcing strength. Reinforc-
ing strength is often referred to as the maximum reinforcing effect of a drug or
other reinforcer.31 Generally, it has been inferred from the strength of the behavior
maintained by the drug. In a progressive-ratio procedure, the number of responses
required to obtain a reinforcer progressively increases over the duration of the ses-
sion. Eventually, responding for the reinforcer will cease when the response require-
ment becomes too great. This point, termed a break point, is a measure of a drug’s

500 Responses
15 Minutes
FIGURE 9.1 A cumulative record of lever pressing maintained in a rhesus monkey under a
progressive-ratio schedule of cocaine (0.1 mg/kg/injection) self-administration over a daily
session. The daily session consisted of five components, each made up of four trials at a par-
ticular response requirement. The response requirement began at a fixed ratio (FR) of 120 and
doubled in subsequent components (i.e., 120, 240, 480, 960, 1920). A trial ended with a drug
injection or the expiration of a limited hold. The session ended if the limited hold expired two
consecutive times. Abscissa: time. Ordinate: cumulative number of responses. The response
pen reset vertically upon completion of the FR or when the pen reached the top of the paper.
Injections are indicated by a deflection of the response pen.
reinforcing efficacy. Under a progressive-ratio schedule, the response requirement

can increase either following the delivery of the drug32 or at the beginning of each
daily session.33,34 If the response requirement increases following the delivery of the
drug, it will be incremented within a daily session, and completion of the response
requirement one time will result in the delivery of drug. If the response require-
ment increases at the beginning of a daily session, it will be fixed over the dura-
tion of a daily session. The same response requirement will be in effect for several
days (i.e., until stability criteria are met) before progressing to the next response
requirement, allowing multiple determinations of self-administration at a particular
response requirement. More recent progressive-ratio procedures combine both of
these approaches and require that the animal respond a given number of times at a
particular response requirement before proceeding to the next response requirement
within a daily session.35,36 Thus, these procedures have the advantage of collecting
multiple determinations of self-administration at a particular response requirement
within a daily session while still allowing the response requirement to be increased
within the session. A cumulative record typical of performance engendered by a
progressive-ratio schedule is shown in Figure 9.1.
9.4 RESEARCH APPLICATION AND DATA INTERPRETATION

The primary focus of drug self-administration research in nonhuman primates has
been to establish the reinforcing properties of drugs of abuse and to identify neu-

rochemical mechanisms underlying drug use. A better understanding of the neu-

rochemical basis of drug self-administration is essential for the development of
treatment medications for human drug abusers. The key feature of drug-reinforced
behavior is control of behavior by response-contingent drug delivery.37 Hence,
drug-reinforced behavior should be distinguished from drug self-administration. A
number of control procedures have been described to demonstrate that increases
in behavior that result in drug delivery are caused specifically by the reinforcing
effects of the drug.6 The most commonly used procedure is to substitute saline for
the drug solution and determine whether the behavior undergoes extinction. The rate
and pattern of responding maintained by drug delivery depends on a number of vari-
ables including the schedule of reinforcement, drug dose, the volume and duration
of injection, and the duration of drug self-administration sessions. Drug self-admin-
istration studies have consistently obtained an inverted U-shaped dose-effect curve
relating the unit dose of drug delivered per injection and response rate or number of
injections delivered. The dose-effect function reflects a combination of reinforcing
effects and unconditioned stimulus effects such as sedation or marked hyperactiv-
ity. Typically, the ascending limb of the dose-effect curve reflects the reinforcing
effects, and response rate increases with drug dose. In contrast, the descending limb
of the curve reflects a nonspecific disruption of operant behavior as excessive drug
accumulates over the session, and response rate decreases with drug dose. It should
be noted that the dose-effect curve relating the unit dose of drug delivered per injec-
tion and drug intake in mg/kg is typically a monotonic increasing function. Lastly,
an inverse relationship has been obtained between infusion duration and reinforcing
effects.38,39 The longer the infusion time required to deliver a constant volume of
drug solution, the less effectively the drug functions as a reinforcer. However, the
latter relationship is typically not observed until the infusion duration is extended to
a minute or more.
A study by Glowa et al.40 illustrates the use of an FR schedule of drug self-
administration to characterize the effectiveness of a dopamine reuptake inhibitor to
alter the reinforcing effects of cocaine in rhesus monkeys. A standard tethered-cath-
eter, home-cage system was used for i.v. drug delivery.5 Animals were trained to self-
administer cocaine under an FR 10 schedule of drug delivery during 90-min daily
sessions. Pretreatment with the high-affinity dopamine reuptake inhibitor, GBR
12909, dose-dependently decreased rates of cocaine-maintained responding, and the
effect was larger when lower doses of cocaine were used to maintain responding.
Moreover, the rate-decreasing effects of GBR 12909 were greater on cocaine-main-
tained responding than on food-maintained responding under a multiple schedule of
drug and food delivery. The results obtained were consistent with previous reports
demonstrating that drugs with dopamine agonist effects can decrease cocaine-main-
tained responding.41,42 This type of drug interaction has been attributed to a satiation
of cocaine-maintained responding by pretreatment with a drug having dopaminergic
effects. The latter approach to cocaine medication development has been referred to
as substitute agonist pharmacotherapy.43 Hence, response-independent delivery of
a dopamine reuptake inhibitor may have decreased cocaine self-administration by
substituting for the reinforcing effects of response-produced cocaine. This interpre-
tation is supported by studies showing that GBR 12909 will substitute for cocaine as
a reinforcer in squirrel monkeys.44–46

Note that Glowa et al.40 incorporated several important design features in their
self-administration study. First, multiple unit doses of cocaine were self-adminis-
tered on separate occasions in order to establish a complete dose-effect curve for
cocaine. Hence, pretreatment effects of GBR 12909 could be assessed over a broad
range of cocaine doses. Second, multiple pretreatment doses of GBR 12909 were
administered in order to establish dose-dependency of pretreatment effects and to
identify the optimal pretreatment dose that lacked overt behavioral toxicity. Lastly,
the specificity of pretreatment effects on cocaine-maintained behavior was assessed
by comparing drug effects on food-maintained behavior. The multiple schedule that
alternated cocaine and food as maintaining events during separate components was
well suited for this application. Moreover, cocaine dose was manipulated to match
response rate to that obtained during the food component of the multiple schedule.
The finding that GBR 12909 suppressed cocaine-maintained responding at doses
that had little or no effect on food-maintained responding under identical schedules
and comparable response rates provides convincing evidence that GBR 12909 selec-
tively attenuated the reinforcing effects of cocaine.
A study by Woolverton47 provides another example of cocaine self-administra-
tion under an FR schedule in rhesus monkeys. The objective was to characterize the
effectiveness of dopamine antagonists to alter the reinforcing effects of cocaine.
A standard tethered-catheter, home-cage system was used for i.v. drug delivery.
Animals were trained to self-administer cocaine under an FR 10 schedule of drug
delivery during 2-hr daily sessions. When responding was stable, the animals were
pretreated with the D1 antagonist SCH 23390, or the D2 antagonist pimozide. Inter-
mediate doses of pimozide generally increased cocaine self-administration, whereas
SCH 23390 either had no effect or decreased cocaine self-administration. High
doses of both antagonists decreased the rate of cocaine self-administration, but also
produced pronounced catalepsy. Hence, the latter effects could not be attributed to
a selective interaction with the reinforcing effects of cocaine. The author concluded
that the selective increase in responding maintained by cocaine following pimozide
pretreatment suggested a role for the D2-receptor in cocaine self-administration.
Strengths of the Woolverton47 study design included multiple unit doses of
cocaine and multiple pretreatment doses of both dopamine antagonists. Extinction of
cocaine self-administration when saline was substituted for cocaine was also charac-
terized. Note that response rate for cocaine increased following pretreatment with the
D2-selective antagonist. The latter effect is interpreted as a behavioral compensation
to overcome the attenuation of the reinforcing effects of cocaine by pimozide. Since
drug intake is a direct function of response rate under FR schedules, an increase
in rate will result in greater session intake of cocaine, which may effectively sur-
mount the dopamine antagonist effects of pimozide. The finding that the pattern of
responding following pimozide was virtually identical to that seen in the first session
of extinction supports the view that pimozide was attenuating the reinforcing effects
of cocaine. However, alternative interpretations were acknowledged, largely because
specificity of pretreatment effects on cocaine-maintained behavior was not assessed
by comparing drug effects on behavior maintained by nondrug reinforcers.
Nader et al.48 used an FI schedule of drug self-administration to characterize
the effectiveness of a novel cocaine analog to alter the reinforcing effects of cocaine

in rhesus monkeys. A standard tethered-catheter, home-cage system was used for

intravenous drug delivery. Animals were trained to self-administer cocaine under an
FI 5-min schedule during 4-hr daily sessions. Under this schedule, the first response
after 5 min produced a 10-sec cocaine injection. Pretreatment with the high-affinity
dopamine reuptake inhibitor 2G-propanoyl-3G-(4-tolyl)-tropane (PTT) dose-depend-
ently decreased rates of cocaine-maintained responding and total session intake of
cocaine. The reinforcing effects of PTT were also evaluated in a separate group of
animals. When substituted for cocaine, PTT maintained response rates that were
similar to those maintained by saline and significantly lower than rates maintained
by cocaine. The results demonstrated that a long-acting dopamine reuptake inhibitor
could effectively decrease cocaine self-administration in nonhuman primates. Also,
failure of PTT to maintain rates of self-administration greater than those obtained
during extinction conditions suggested that PTT may have limited abuse liability.
The study of Nader et al.48 was consistent with previous findings using dopa-
mine reuptake inhibitors to decrease cocaine self-administration under FR sched-
ules of drug self-administration.37 Hence, a generality of pretreatment effects has
been demonstrated across experimental conditions, further supporting a substitute
agonist approach to cocaine medication development. Both studies included multiple
unit doses of cocaine and multiple pretreatment doses of the dopamine reuptake
inhibitors. In addition, assessment of the reinforcing properties of PTT provided
critical information concerning the abuse liability of the candidate medication. The
fact that PTT did not reliably maintain self-administration behavior, whereas GBR
12909 has been shown to function effectively as a reinforcer in monkeys,44–46 illus-
trates the potential importance of pharmacokinetic factors in drug self-administra-
tion studies. It is possible that low rates of PTT self-administration were a result of
its relatively long duration of action at inhibiting dopamine uptake. Hence, its long
duration of action compared with cocaine may have required lower session intake to
produce cocaine-like reinforcing effects. It should be noted, however, that the rate
of onset appears to play a more prominent role in the reinforcing effects of psycho-
stimulants than does duration of action.49–51
Human drug use often involves a ritualized sequence of behaviors that occurs
in a specific environment. The environmental stimuli associated with drug use are
believed to play a major role in the maintenance of drug-seeking behavior.29 Second-
order schedules of drug self-administration have been used in nonhuman primates to
maintain extended sequences of responding between drug injections20,45,46,52 analo-
gous to patterns of drug use in humans. The second-order schedule is well suited for
drug-interaction and drug-substitution experiments because response rate increases
as a direct function of the unit dose administered at low and intermediate doses (Fig-
ure 9.2). Note that high doses of drug can disrupt performance during the latter com-
ponents of a session as multiple doses accumulate. In drug-interaction experiments,
changes in the positioning of the cocaine dose-effect curve leftward or rightward
will indicate altered potency of cocaine to function as a reinforcer. A downward
shift in the cocaine dose-effect curve will indicate an insurmountable attenuation
of cocaine self-administration. In drug-substitution experiments, maximum rates of
responding maintained over a range of drug doses can be used to compare reinforc-
ing effectiveness.

1.00
N
= 3
0.75
Responses/S econd
0.50
0.25
0.00
0.03 0.1 0.3 1.0
Cocaine (mg/inj ection)
FIGURE 9.2 Mean (± SEM) rate of responding maintained in a group of three squirrel
monkeys under a second-order fixed interval 900-sec schedule of cocaine intravenous self-
administration with fixed ratio 20 components. Data for each dose of cocaine were derived
from at least five consecutive sessions on two separate occasions. Abscissae: dose, log scale.
Ordinate: mean response rate expressed as responses per second.
A study by Howell et al.53 provides an example of cocaine self-administration

under a second-order FI schedule in squirrel monkeys. The objective was to char-
acterize the effectiveness of a phenyltropane analog of cocaine to alter the reinforc-
ing effects of cocaine. The distal end of the catheter was externalized and exited
between the scapulae, and a nylon-mesh jacket protected the catheter when not in
use. Animals were trained to self-administer cocaine under a second-order FI 15-
min schedule with FR 20 components during 1-hr daily sessions. Pretreatment with
the dopamine reuptake inhibitor RTI-113 significantly decreased rates of cocaine
self-administration, and the effect was not surmounted by increasing the unit dose
of cocaine (Figure 9.3). The latter findings are consistent with previous studies using
dopamine reuptake inhibitors to decrease cocaine self-administration under FR40
and FI48 schedules of drug self-administration. Hence, the generality of pretreatment
effects has been demonstrated over a range of experimental conditions and in two
different primate species. Note that low doses of RTI-113 actually increased rates
of cocaine self-administration at the low unit dose of cocaine, providing evidence
of additivity of effects. The latter finding is an important consideration when con-
ducting drug interaction studies with two drugs that have a similar mechanism of
action.

350 350
0.1 mg/injection Cocaine 0.3 mg/injection Cocaine
300 300
250 250
Percent Control
200 200
150 150
100 100
50 50
*
*
0 0
0.03 0.1 0.3 0.03 0.1 0.3
RTI 113 (mg/kg) RTI 113 (mg/kg)
FIGURE 9.3 Mean (± SEM) rate of responding in a group of three squirrel monkeys
maintained under a second-order fixed interval 900-sec schedule of cocaine (0.1 and 0.3
mg/injection) intravenous self-administration with fixed ratio 20 components. Cocaine was
self-administered alone (dashed lines) or following pretreatment with RTI-113 (closed sym-
bols). Subjects were pretreated with each dose of RTI-113 for three consecutive sessions, and
each subject received all drug combinations on two separate occasions. Abscissae: dose, log
scale. Ordinates: mean response rate expressed as a percentage of control rate obtained when
subjects were pretreated with saline. Asterisks indicate a significant (p < 0.05) effect of RTI-
113 pretreatment.
Similar to other self-administration schedules, the reinforcing potencies of drugs

can be determined under progressive-ratio schedules.33,36 For example, cocaine is
10-fold more potent than the local anesthetic procaine under a progressive-ratio pro-
cedure.36 However, progressive-ratio procedures are most useful for determining the
reinforcing strength of self-administered drugs. Drugs can be rank-ordered based
on their relative reinforcing effects as determined by break point in the progres-
sive ratio.33,36 For example, cocaine has been found to maintain higher break-point
values than diethylpropion, chlorphentermine, and fenfluramine in baboons.33 More
recently, cocaine has been found to maintain higher break-point values than pro-
caine in rhesus monkeys (Figure 9.4).36

Cocaine Procaine
20
FR 1920
16
(5) (5) (2)
FR 960
Mean inj/Session
12 (5)
(5) (3)
(5) (5) FR 480
(4)
8
(3)
(3) FR 240
(3)
4
(5) FR 120
(5)
0
Sal 0.01 0.1 1.0 10
Dose (mg/kg/inj)
FIGURE 9.4 Break-point values and injections/session maintained by cocaine (closed sym-
bols) and procaine (open symbols) under a progressive-ratio schedule in rhesus monkeys.
Data are the mean (± SEM) for the number of monkeys indicated in parentheses above each
dose. The daily session consisted of five components, each made up of four trials at a particu-
lar response requirement. The response requirement began at a fixed ratio of 120 and doubled
in subsequent components (i.e., 120, 240, 480, 960, 1920). Absicssae: dose, log scale. Left
ordinate: mean injections/session. Right ordinate: break-point values. Dashed lines represent
the mean number of injections/session taken at a particular break-point value.
As mentioned above, break point typically is used as the dependent measure

to assess reinforcing effectiveness under a progressive-ratio procedure. However,
break-point data violate the assumption of homogeneity of variance necessary for
reliable statistical analysis. Variability is greater at high break-point values than at
low break-point values, making it difficult to determine effects based on drug dose
at high break points.54,55 Therefore, some researchers have applied a data transforma-
tion to break-point data before analysis (see Rowlett et al.55). In addition, the num-
ber of injections per session has been found to be a reliable measure of reinforcing
strength and does not violate the assumption of homogeneity of variance.36,55 With
intravenous self-administration paradigms, it is important to consider that effects
other than reinforcing effects may influence responding for drug injections when
high doses of a drug are available. Downward turns in dose-response curves have
been explained by drug accumulation. To address this issue in the progressive ratio,
researchers have used a timeout after each injection. The idea is that the timeout will
allow the effects of the drug to dissipate before another injection can be obtained.
A timeout length of 30 min is effective for studying cocaine in the progressive-ratio
procedure.36,55
In addition to progressive-ratio procedures, choice procedures are used to study
the reinforcing strength of drugs. The choice paradigm allows animals access to two
reinforcers and evaluates the preference of one reinforcer over the other. Typically,

animals choose between a food and a drug reinforcer or two drug reinforcers.56,57
Reinforcing strength can be determined based on the preference of one reinforcer
over the other. For example, rhesus monkeys given a choice between a high and a
low dose of cocaine will prefer the higher cocaine dose.57 A study by Johanson and
Aigner58 suggested a difference in the maximum reinforcing effects of cocaine and
procaine using a choice procedure. They evaluated the preference for an i.v. injec-
tion of cocaine versus an i.v. injection of procaine in rhesus monkeys. At equipotent
doses for reinforcing effects, monkeys chose i.v. injections of cocaine more than 80%
of the time.52 These results are consistent with those of the Woolverton33 progres-
sive-ratio study mentioned above. Thus, choice paradigms are reliable for studying
reinforcing efficacy.
Lastly, behavioral economics provides a means to quantify the reinforcing effects
of drugs independent of dose.59 Such studies apply microeconomic concepts includ-
ing consumer demand and labor supply theories to help understand how behavior is
maintained by various reinforcers, referred to as “commodities” in economic par-
lance. Behavioral economic studies use total daily consumption of a commodity,
rather than response rate, as the primary indicator of demand for that commod-
ity. In drug self-administration experiments, subjects regulate their consumption
by responding to obtain multiple presentations of the commodities of interest. The
function generated by assessing consumption across increasing “cost” of a com-
modity is known as a “demand curve,” and these functions generally reveal that
consumption decreases as the cost of a commodity increases. Cost is manipulated
by increasing the work requirement—in the simplest case, increasing the FR value
required to receive an injection. As the FR value is increased, consumption levels
decrease, reflecting the behavioral sensitivity to price. By comparing consumption
at a given price, relative to the level of consumption at the lowest price (i.e., at FR
1), one can gauge the “elasticity of demand” of a given commodity.60 Demand is
“inelastic” when consumption is defended across large increases in price. In con-
trast, demand is “elastic” when consumption declines rapidly with increasing price.
An example of the relationship between demand and onset of drug action was dem-
onstrated with the drugs fentanil, alfentanil, and remifentanil. All three compounds
are full agonists at μ opioid receptors and have immediate onsets of action following
i.v. administration, but the durations of action for these compounds differ mark-
edly, with remifentanil having the shortest duration of action and fentanil having the
longest duration of action. Despite differences in durations of action, and apparent
differences in the absolute rates of responding maintained by these three compounds
in self-administration experiments, demand curve analysis suggests that these drugs
do not differ in their reinforcing effectiveness.61 Thus, duration of action does not
seem to contribute to the reinforcing effectiveness of opioids, or, perhaps, for other
drug classes as well.
9.5 DISCUSSION
Nonhuman primate models of drug self-administration provide a rigorous, system-
atic approach to characterize the reinforcing effects of psychoactive drugs. The lon-
gevity of nonhuman primates is an important consideration, allowing for long-term

studies to be conducted and repeated-measures designs to be employed. A single

venous catheter can be readily maintained for over a year, and multiple implants
permit the conduct of self-administration experiments for several years in individual
subjects. Long-term studies with repeated measures are well suited for comprehen-
sive drug-interaction experiments. While rodent models of drug self-administration
have substantially contributed toward an understanding of neuropharmacology, the
nonhuman primate represents an animal model with unique relevance to under-
standing the neurochemical basis of substance abuse in humans. For example, the
complexity of the topographical organization of the striatum and its connections
with surrounding areas in primates62–64 complicates extrapolations from rodents to
primates. Moreover, a large number of brain regions respond differently to acute
drug administration in monkeys65 compared to rodents.66,67 Both the topography
of altered brain metabolism and the direction of metabolic responses differ mark-
edly.65,67 The metabolic effects reported in monkeys are more consistent with data on
functional activity in humans.68,69 These findings, in conjunction with documented
species differences in drug metabolism,18 illustrate the importance of nonhuman
primate models in substance abuse research.
Research efforts that have used nonhuman primate models of drug self-admin-
istration have focused primarily on the identification of neurochemical mechanisms
that underly drug reinforcement, and the development of pharmacotherapies to
treat drug addiction. In clinical evaluations of new medications, a decrease in drug
self-administration is the goal of treatment.70–72 Preclinical evaluations of pharma-
cotherapies require the establishment of stable baseline patterns of drug self-admin-
istration prior to drug-interaction studies. Subsequently, the treatment medication is
administered as a pretreatment before the conduct of self-administration sessions. It
is critical to study several doses of the treatment medication to determine an effec-
tive dose range and a maximally effective dose that lacks overt behavioral toxicity.
The effects of the treatment medication typically are evaluated first in combina-
tion with a dose of the self-administered drug on the ascending limb of the dose-
effect curve that maintains high rates of self-administration. However, a complete
dose-effect curve should be characterized for the self-administered drug because
pretreatment effects can differ depending on the unit dose of the drug self-admin-
istered. A rightward shift in the dose-effect curve suggests that drug pretreatment
is antagonizing the reinforcing effects of the self-administered drug. A downward
displacement of the dose-effect curve indicates an insurmountable attenuation of the
reinforcing effects. Alternatively, a leftward shift is consistent with an enhancement
of the reinforcing effects. Medications that shift the dose-effect curve downward
and decrease self-administration over a broad range of unit doses are most likely to
have therapeutic utility. Medications that shift the dose-effect curve to the right and
simply alter the potency of the self-administered drug may prove to be ineffective at
higher unit doses. Clinically, most medications are administered on a chronic basis
and may require long-term exposure before therapeutic effects are noted.73,74 Accord-
ingly, preclinical studies should include repeated daily exposure to the medication to
characterize peak effectiveness and to document continued effectiveness over multi-
ple sessions.75 Figure 9.5 illustrates the effects of chronic amphetamine treatment on
cocaine self-administration in rhesus monkeys. Note that amphetamine maintained

80 100
Injections/Day
60 75
Pellets/Day
40 50
20 25
0 0
c 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 P1 P2 P3 P4 P5 P6 P7
Days of Treatment
FIGURE 9.5 Time course of effects of saline or d-amphetamine (0.01–0.1 mg/kg per hr)
on responding for 0.01 mg/kg per injection cocaine and food pellets. Abscissae: consecutive
days of treatment. Left ordinate: number of cocaine injections (0.01 mg/kg per injection)
delivered on each day of treatment (filled triangles, maximum = 80). Right ordinate: number
of food pellets delivered on each day of treatment (open circles, maximum = 100). Each point
shows mean data from four monkeys, and error bars show the SEM.
its effectiveness to reduce cocaine self-administration over multiple weeks. Also,

there was an initial disruption of food-maintained behavior, but responding for food
returned to baseline values over the first week of amphetamine treatment. It is also
critical to reestablish baseline levels of drug self-administration between successive
exposures to the medication to ensure that the catheter preparation is functional and
that persistent effects of the pretreatment drug do not interfere with the interpreta-
tion of drug interactions obtained.
The primary treatment outcome measures in drug self-administration studies
are rate of responding and the number of drug injections delivered per session. Both
measures are influenced by the schedule of reinforcement, drug dose, the volume
and duration of injection, and the duration of the self-administration session. More-
over, most drugs that are self-administered have direct effects on rate of responding
that may be distinct from their reinforcing effects. For example, cocaine injections
may increase rate of responding early in the session, but suppress behavior later in
the session as total drug intake accumulates. Another important consideration in
evaluating medication effectiveness is the selectivity of effects on drug self-admin-
istration. If the drug pretreatment decreases drug self-administration at lower doses
or to a greater extent than behavior maintained by a nondrug reinforcer such as food,
the outcome is indicative of selective interactions with the reinforcing properties of
the self-administered drug. In contrast, a nonspecific disruptive effect on operant
behavior will likely suppress drug- and food-maintained responding to a compa-
rable extent. Lastly, the reinforcing properties and abuse potential of the medica-
tion should be evaluated by substituting a range of doses of the medication for the
self-administered drug. Since reinforcing effects in preclinical studies are correlated
with abuse liability in humans, reliable self-administration of the medication is usu-
ally considered undesirable.
While this chapter focuses on the i.v. route of drug self-administration, the rein-
forcing properties of drugs have been studied effectively in nonhuman primates via
the oral and inhalation routes. For example, orally delivered cocaine can function
as a reinforcer in rhesus monkeys, and persistent and orderly responding is obtained
when dose and FR size are varied.76 Orally delivered phencyclidine and ethanol also

maintain self-administration behavior in rhesus monkeys under progressive-ratio

schedules.77 However, establishing drugs as reinforcers via oral administration can
be difficult because of metabolic effects associated with the gastrointestinal system
and delayed onset of CNS activity associated with slow absorption and distribution.
In addition, drug solutions often have a bitter taste that may be aversive to nonhu-
man primates. Accordingly, complex induction procedures are frequently used to
establish oral self-administration of drug solutions. Studies that have demonstrated
cocaine’s ability to function as a reinforcer have used a fading procedure from an
initial baseline of ethanol-maintained responding,76 although concurrent access
to cocaine and vehicle solution is sufficient to establish oral self-administration.78
Lastly, cocaine and heroin are self-administered by rhesus monkeys via smoke inha-
lation under FR and progressive-ratio schedules.79–81 Although initial training is dif-
ficult because of the aversive characteristics of smoke, and drug dose is difficult
to quantify, rhesus monkeys can rapidly learn to self-administer the drugs via the
inhalation route. Given the above considerations, the advantages of intravenous self-
administration procedures are clearly evident. Drug dose is easily manipulated and
quantified, metabolic effects in the gastrointestinal system and slow absorption are
avoided, and onset of CNS activity is rapid. Importantly, orderly and reliable dose-
effect curves are obtained that are sensitive to pharmacological manipulation.
REFERENCES
1. Thompson, T. and Pickens, R., eds. 1971. Stimulus properties of drugs. New York:
Appleton-Century-Crofts.
2. Skinner, B. F. 1938. The behavior of organisms. New York: Appleton-Century-Crofts.
3. Headlee, C. P., Coppock, H. W., and Nichols, J. R. 1955. Apparatus and technique
involved in laboratory method of detecting the addictiveness of drugs. J. Am. Pharm.
Assoc., Sci. Ed., 44:229.
4. Thompson, T., and Schuster, C. R. 1964. Morphine self-administration, food-rein-
forced, and avoidance behaviors in rhesus monkeys. Psychopharmacologia 5:87.
5. Deneau, G., Yanagita, T., and Seevers, M. H. 1969. Self-administration of psychoactive
substances by the monkey: A measure of psychological dependence. Psychopharmaco-
logia 16:30.
6. Pickens, R., and Thompson T. 1968. Cocaine-reinforced behavior in rats: Effects of
reinforcement magnitude and fixed-ratio size J. Pharmacol. Exp. Ther. 161:122.
7. France, C. P., Winger, G. D., Medzihradsky, F., Seggel, M. R., Rice, K. C., and Woods,
J. H. 1991. Mirfentanil: Pharmacological profile of a novel fentanyl derivative with
opioid and nonopioid effects. J. Pharmcol. Exp. Ther. 258:502.
8. Young, A. M., Stephens, K. R., Hein, D. W., and Woods, J. H. 1984. Reinforcing and
discriminative stimulus properties of mixed agonist-antagonist opioids. J. Pharmacol.
Exp. Ther. 229:118.
9. Tang, A. H., and Collins, R. J. 1985. Behavioral effects of a novel kappa-opioid analge-
sic, U-50488, in rats and rhesus monkeys. Psychopharmacology 85:309.
10. Koob, G. F., and Bloom, F. E. 1988. Cellular mechanisms of drug dependence. Sci-
ence 242:715.
11. Koob, G. F., and Weiss, F. 1990. Pharmacology of drug self-administration. Alcohol
7:193.

12. Bergman, J., Kamien, J. B., and Spealman, R. D. 1990. Antagonism of cocaine self-
administration by selective dopamine D1 and D2 antagonists. Behav. Pharmacol.
1:355.
13. Bertalmio, A. J., and Woods, J. H. 1989. Reinforcing effects of alfentanil is mediated
by mu opioid receptors: Apparent pA2 analysis. J. Pharmacol. Exp. Ther. 251:455.
14. Dewit, H., and Wise, R. A. 1977. Blockade of cocaine reinforcement in rats with the
dopamine receptor blocker pimozide, but not with the noradrenergic blockers phentol-
amine or phenoxybenzamine. Can. J. Psychol. 31:195.
15. Wilson, M. C., and Schuster, C. R. 1972. The effects of chlorpromazine on psychomo-
tor stimulant self-administration in the rhesus monkey. Psychopharmacologia 26:115.
16. Griffiths, R. R., Bigelow, G. E., and Henningfield, J. E. 1980. Similarities in animal and
human drug-taking behavior. In Advances in substance abuse, Vol. 1, ed. N. K. Mello,
1. Greenwich, CT: JAI Press.
17. Mello, N. K. 1979. Behavioral pharmacology of narcotic antagonists. In The interna-
tional challenge of drug abuse, NIDA research monograph series 19, ed. R. C. Peter-
son, 126. Washington, DC: U.S. Government Printing Office.
18. Weerts, E. M., Fantegrossi, W. E., and Goodwin, A. K. 2007. The value of nonhuman
primates in drug abuse research. Exp. Clin. Psychopharmacol. 15:309.
19. Herd, J. A., Morse, W. H., Kelleher, R. T., and Jones, L. G. 1969. Arterial hypertension
in the squirrel monkey during behavioral experiments. Am. J. Physiol. 217:24.
20. Howell, L. L., and Byrd, L. D. 1995. Serotonergic modulation of the behavioral effects
of cocaine in the squirrel monkey. J. Pharmacol. Exp. Ther. 275:1551.
21. Wojnicki, F. H. E., Rothman, R. B., Rice, K. C., and Glowa, J. R. 1999. Effects of phen-
termine on responding maintained under multiple fixed-ratio schedules of food and
cocaine presentation in the rhesus monkey. J. Pharmacol. Exp. Ther. 288:550.
22. Byrd, L. D. 1979. A tethering system for direct measurement of cardiovascular function
in the caged baboon. Am. J. Physiol.: Heart Circ. Physiol. 5:H775.
23. Byrd, L. D. 1979. The behavioral effects of cocaine: Rate dependency or rate constancy.
Eur. J. Pharmacol. 56:355.
24. Howell, L. L., and Landrum, A. M. 1994. Behavioral and pharmacological modulation
of respiration in rhesus monkeys. J. Exp. Anal. Behav. 62:57.
25. Howell, L. L., and Landrum, A. M. 1997. Effects of chronic caffeine administration on
respiration and schedule-controlled behavior in rhesus monkeys. J. Pharmacol. Exp.
Ther. 283:190.
26. Slifer, B. L., and Balster, R. L. 1985. Intravenous self-administration of nicotine: With
and without schedule-induction. Pharmacol. Biochem. Behav. 22:61.
27. Rescorla, R. A. 1980. Pavlovian second-order conditioning: Studies in associative
learning, 120. Hillsdale, NJ: Lawrence Erlbaum.
28. Schindler, C. W., Katz, J. L., and Goldberg, S. R. 1988. The use of second-order sched-
ules to study the influence of environmental stimuli on drug-seeking behavior. In
Learning factors in substance abuse, NIDA research monograph series 84, ed. B. A.
Ray, 180. Washington, DC: U.S. Government Printing Office.
29. Katz, J. L., and Goldberg, S. R. 1975. Second-order schedules of drug injection: Impli-
cations for understanding reinforcing effects of abused drugs. Adv. Subst. Abuse
(4)205:1991.
30. Kelleher, R. T. 1975. Characteristics of behavior controlled by scheduled injections of
drugs. Pharmacol. Rev. 27:307.
31. Hodos, W. 1961. Progressive ratio as a measure of reward strength. Science 134:943.
32. Bedford, J. A., Baily, L. P., and Wilson, M. C. 1978. Cocaine reinforced progressive
ratio performance in the rhesus monkey. Pharmacol. Biochem. Behav. 9:631.

33. Griffiths, R. R., Brady, J. V., and Snell, J. D. 1978. Progressive-ratio performance main-
tained by drug infusions: Comparison of cocaine, diethylproprion, chlorphentermine,
and fenfluramine. Psychopharmacology 56:5.
34. Griffiths, R. R., Bradford, L. D., and Brady, J. V. 1979. Progressive-ratio and fixed-ratio
schedules of cocaine-maintained responding in baboons. Psychopharmacology 65:125.
35. Rowlett, J. K., and Woolverton, W. L. 1997. Self-administration of cocaine and heroin
combinations by rhesus monkeys responding under a progressive-ratio schedule. Psy-
chopharmacology 133:363.
36. Woolverton W. L. 1995. Comparison of the reinforcing efficacy of cocaine and pro-
caine in rhesus monkeys responding under a progressive-ratio schedule. Psychophar-
macology 120:296.
37. Young, A. M., and Herling, S. 1986. Drugs as reinforcers: Studies in laboratory ani-
mals. In Behavioral analysis of drug dependence, eds. S. R. Goldberg and I. P. Stoler-
man, I. P., 9. San Diego: Academic Press.
38. Balster, R. L., and Schuster, C. R. 1973. Fixed-interval schedule of cocaine-reinforce-
ment: Effect of dose and infusion duration. J. Exp. Anal. Behav. 20:119.
39. Kato, S., Wakasa, Y., and Yanagita, T. 1987. Relationship between minimum reinforc-
ing doses and injection speed in cocaine and pentobarbital self-administration in crab-
eating monkeys. Pharmacol. Biochem. Behav. 28:407.
40. Glowa, J. R., Wojnicki, F. H. E., Matecka, D., et al. 1995. Effects of dopamine reuptake
inhibitors on food- and cocaine-maintained responding. I: Dependence on unit dose of
cocaine. Exp. Clin. Psychopharmacol. 3:219.
41. Caine, S. B., and Koob, G. F. 1993. Modulation of cocaine self-administration in the rat
through D3 dopamine receptors. Science 260:1814.
42. Skjoldager, P., Winger, G., and Woods, J. H. 1993. Effects of GBR 12909 and cocaine
on cocaine-maintained behavior in rhesus monkeys. Drug Alcohol Depend. 33:31.
43. Carroll, F. I., Howell, L. L., and Kuhar, M. J. 1999. Pharmacotherapies for treatment of
cocaine abuse: Preclinical aspects. J. Med. Chem. 42:2721.
44. Bergman, J., Madras, B. K., Johnson, S. E., and Spealman, R. D. 1989. Effects of
cocaine and related drugs in nonhuman primates. III. Self-administration by squirrel
monkeys. J. Pharmacol. Exp. Ther. 251:150.
45. Howell, L. L., and Byrd, L. D. 1991. Characterization of the effects of cocaine and GBR
12909, a dopamine uptake inhibitor, on behavior in the squirrel monkey. J. Pharmacol.
Exp. Ther. 258:178.
46. Howell, L. L., Czoty, P. W., and Byrd, L. D. 1997. Pharmacological interactions between
serotonin and dopamine on behavior in the squirrel monkey. Psychopharmacology
131:40.
47. Woolverton, W. L. 1986. Effects of a D1 and D2 dopamine antagonist in the self-admin-
istration of cocaine and piribedil by rhesus monkeys. Pharmacol. Biochem. Behav.
24:351.
48. Nader, M. A., Grant, K. A., Davies, H. M. L., Mach R. H., and Childers, S. R. 1997.
The reinforcing and discriminative stimulus effects of the novel cocaine analog 2G-
propanoyl-3G-(4-tolyl)-tropane in rhesus monkeys. J. Pharmacol. Exp. Ther. 280:541.
49. Kimmel, H. L., O’Connor, J. A., Carroll, F. I., and Howell, L. L. 2007. Faster onset and
dopamine transporter selectivity predict stimulant and reinforcing effects of cocaine
analogs in squirrel monkeys. Pharmacol. Biochem. Behav. 86:45.
50. Lile, J. A., Wang, Z., Woolverton, W. L., et al. 2003. The reinforcing efficacy of psycho-
stimulants in rhesus monkeys: The role of pharmacokinetics and pharmacodynamics.
J. Pharmacol. Exp. Ther. 307:356.
51. Woolverton, W. L., Ranaldi, R., Wang, Z., et al. 2002. Reinforcing strength of a novel
dopamine transporter ligand: Pharmacodynamic and pharmacokinetic mechanisms. J.
Pharmacol. Exp. Ther. 303:211.

52. Kelleher, R. T., and Goldberg, S. R. 1977. Fixed-interval responding under second-
order schedules of food presentation or cocaine injection. J. Exp. Anal. Behav. 28:221.
53. Howell, L. L., Czoty, P. W., Kuhar, M. J., and Carroll, F. I. Comparative behavioral
pharmacology of cocaine and the selective dopamine uptake inhibitor, RTI-113, in the
squirrel monkey. J. Pharmacol. Exp. Ther., in press.
54. Depoortere, R. Y., Li, D. H., Lane, J. D., and Emmett-Oglesby, M. W. 1993. Parameters
of self-administration of cocaine in rats under a progressive-ratio schedule. Pharma-
col. Biochem. Behav. 45:539.
55. Rowlett, J. K., Massey, B. W., Kleven, M. S., and Woolverton, W. L. 1996. Parametric
analysis of cocaine self-administration under a progressive-ratio schedule in rhesus
monkeys. Psychopharmacology 125:361.
56. Iglauer, C., and Woods, J. H. 1974. Concurrent performances: Reinforcement by differ-
ent doses of intravenous cocaine in rhesus monkeys. J. Exp. Anal. Behav. 22:79.
57. Johanson, C. E. 1976. Pharmacological and environmental variables affecting drug
preference in rhesus monkeys. Pharmacol. Rev. 27:343.
58. Johanson, C.-E., and Aigner, T. 1981. Comparison of the reinforcing properties of
cocaine and procaine in rhesus monkeys. Pharmacol. Biochem. Behav. 15:49.
59. Hursh, S. R., Galska, C. M., Winger, G., and Woods, J. H. 2005. The economics of drug
abuse: A quantitative assessment of drug demand. Mol. Interv. 5:20.
60. Bickel, W. K., Marsch, L. A., and Carroll, M. E. 2000. Deconstructing relative reinforc-
ing efficacy and situating the measures of pharmacological reinforcement with behav-
ioral economics: A theoretical proposal. Psychopharmacology 153:44.
61. Ko, M. C., Terner, J., Hursh, S., Woods, J. H., and Winger, G. 2002. Relative reinforc-
ing effects of three opioids with different durations of action. J. Pharmacol. Exp. Ther.
301:698.
62. Haber, S. N., Kunishio, K., Mizobuchi, M., and Lynd-Balta, E. 1995. The orbital and
medial prefrontal circuit through the primate basal ganglia. J. Neurosci. 15:4851.
63. Lynd-Balta, E., and Haber, S. N. 1994. The organization of midbrain projections to the
ventral striatum in the primate. Neuroscience 59:609.
64. Lynd-Balta, E., and Haber, S. N. 1994. Primate striatonigral projections: A compari-
son of the sensorimotor-related striatum and the ventral striatum. J. Comp. Neurol.
3345:562.
65. Lyons, D., Friedman, D. P., Nader, M. A., and Porrino, L. J. 1996. Cocaine alters cere-
bral metabolism within the ventral striatum and limbic cortex of monkeys. J. Neurosci.
16:1230.
66. Porrino, L. J. 1993. Functional effects of cocaine depend on route of administration.
Psychopharmacology 112:343.
67. Porrino, L. J., Domer, F. R., Crane, A. M., and Sokoloff, L. 1988. Selective alterations
in cerebral metabolism within the mesocorticolimbic dopaminergic system produced
by acute cocaine administration in rats. Neuropsychopharmacoalogy,1:109.
68. London, E. D., Cascella, N. G., Wong, D. F., et al. 1990. Cocaine-induced reduction of
glucose utilization in human brain. Arch. Gen. Psychiat. 47:567.
69. Pearlson, G. D., Jeffery, P. J., Harris, G. J., Ross, C. A., Fischman, M. W., and Camargo,
E. E. 1993. Correlation of acute cocaine-induced changes in local cerebral blood flow
with subjective effects. Am. J. Psychiat. 150:495.
70. Mello, N. K., and Mendelson, J. H., 1980. Buprenorphine suppresses heroin use by
heroin addicts. Science 27:657.
71. Mello, N. K., Mendelson, J. H., and Bree, M. P. 1981. Naltrexone effects on morphine and
food self-administration in morphine-dependent rhesus monkeys. J. Pharmacol. Exp. Ther.
218:550.
72. Mello, N. K., Mendelson, J. H., and Kuehnle, J. C. 1982. Buprenorphine effects on
human heroin self-administration. J. Pharmacol. Exp. Ther. 230:30.

73. Gawin, F. H. 1991. Cocaine addiction: Psychology and neurophysiology. Science

251:1580.
74. Gawin, F. H., and Ellinwood, E. H. 1988. Cocaine and other stimulants: Actions, abuse
and treatment. N. Engl. J. Med. 318:1173.
75. Negus, S. S., and Mello, N. K. 2003. Effects of chronic d-amphetamine treatment on
cocaine- and food-maintained responding under a second-order schedule in rhesus
monkeys. Drug Alcohol Depend. 70:39.
76. Meisch, R. A., Bell, S. M., and Lemaire, G. A. 1993. Orally self-administered cocaine
in rhesus monkeys: Transition from negative or neutral behavioral effects to positive
reinforcing effects. Drug Alcohol Depend. 32:143.
77. Rodefer, J. S., and Carroll, M. E. 1996. Progressive ratio and behavioral economic
evaluation of the reinforcing efficacy of orally delivered phencyclidine and ethanol in
monkeys: Effects of feeding conditions. Psychopharmacology 128:265.
78. Macenski, M. J., and Meisch, R. A. 1995. Oral cocaine self-administration in rhesus mon-
keys: Strategies for engendering reinforcing effects. Exp. Clin. Psychopharmacol. 3:129.
79. Mattox, A. J., Thompson, S. S., and Carroll, M. E. 1997. Smoked heroin and cocaine
base (speedball) combinations in rhesus monkeys. Exp. Clin. Psychopharmacol.
5:113.
80. Carroll, M. E., Krattiger, K. L., Gieske, D., and Sadoff, D. A. 1990. Cocaine-base smok-
ing in rhesus monkeys: Reinforcing and physiological effects. Psychopharmacology
102:443.
81. Siegel, R. K., Johnson, C. A., Brewster, J. A., and Jarvik, M. E. 1976. Cocaine self-
administration in monkeys by chewing and smoking. Pharmacol. Biochem. Behav.
4:461.

10 Contextually
Drug Seeking
Induced
During Protracted
Abstinence in Rats
Jerry J. Buccafusco and Laura Shuster
CONTENTS
10.1 Introduction................................................................................................. 199

10.2 General Methods.........................................................................................200
10.2.1 Animal Subjects...............................................................................200
10.2.2 Use of Food to Shape Responding...................................................200
10.2.3 Implantation of the i.v. Infusion Line ............................................. 201
10.2.4 The Morphine Regimen................................................................... 201
10.2.5 The Cocaine Regimen .....................................................................202
10.2.6 Spontaneous Morphine Withdrawal Syndrome...............................202
10.2.7 Reinstatement of Lever Pressing .....................................................202
10.3 Methodological Details...............................................................................204
10.3.1 Things to Prepare Before Surgery ...................................................204
10.3.2 Installing a Permanent Intravenous Line.........................................205
10.3.3 Self-Administration Dose Calculations ...........................................206
10.3.4 Ordering Information.......................................................................207
10.3.5 Coulbourn Instruments—Required Equipment ..............................208
10.4 Actual Data .................................................................................................208
10.4.1 Morphine Self-Administration ........................................................208
10.4.2 Context-Induced Post-Withdrawal Lever Responding..................... 210
10.5 Discussion ................................................................................................... 210
10.5.1 The Drug Abuse Cycle and Pharmacological Intervention ............. 210
References.............................................................................................................. 212
10.1 INTRODUCTION
Over the past half century great strides have been made in the development of useful
animal models for the drug abuse triad—self-administration, physical or psychologi-
199

cal dependence, and withdrawal. In fact, the compulsion to self-administer cocaine

even in the face of adverse consequences is not limited to human beings.1,2 Practi-
cally, it is not that difficult to detoxify a drug addict, but the problem lies in the pro-
pensity for former addicts to relapse to drug-seeking behavior, a risk factor that does
not appear to decrease in potency over time. Recently there has been an increasing
focus on the issue of protracted withdrawal. This feature of drug addiction mir-
rors classical conditioning in that certain contextual cues or environmental stimuli
associated with drug taking can readily initiate a form of withdrawal or craving in
addicts that often leads to renewed drug seeking and relapse (for review, see3–5).
Indeed, both the rat and human share common triggers of relapse, including the
drug of abuse itself, stress, and stimuli or the environment conditioned to the drug of
abuse.6 Rodent models of human drug craving and relapse have used paradigms of
extinction and reinstatement. Such models have shown predictive validity by dem-
onstrating that clinically effective anti-craving drugs reduce drug-seeking behavior
as a component of the model.7
10.2 GENERAL METHODS

Male Wistar rats (Harlan, Indianapolis, Indiana, USA), weighing 300 to 380 g, are
housed and tested in environmentally controlled rooms on a 12/12-hr day/night
cycle, and they are maintained on standard rat chow and tap water (unlimited). All
current animal protocols are approved by the Augusta Veterans Administration Ani-
mal Care and Use Committee. Initially each rat is maintained at about 85% of free-
feeding levels for about 5 days. During this time the animals are acclimated to the
operant chamber and trained for lever pressing. Once a rat is assigned to a particular
operant chamber they are maintained there on a 24-hr basis except during the with-
drawal phase (see below).
10.2.2 USE OF FOOD TO SHAPE RESPONDING

Some investigators use food reinforcement to shape animals for lever responding.
The advantage is that animals that are poor responders can be weeded out before
they are prepared for i.v. self-administration of the drug of interest. The disadvantage
is that the investigator must take care to insure that subsequent lever responding,
whether during the i.v. self-administration phase or the post-withdrawal phase, is
not reflecting the expectation or the habit of obtaining food rewards. Presently we
prefer not to shape animals using food reinforcement, but for those that prefer this
approach, and for those that plan to study food reinforcement as the sole reinforcing
agent, we describe the method below.
We generally train animals on an operant food-reinforcement fixed ratio-1 (FR1)
schedule during 2-hr daily sessions. Lever presses are reinforced by the automated
delivery of a 45-mg food pellet. The lever is signaled active by the illumination

Contextually Induced Drug Seeking During Protracted Abstinence in Rats 201
of a stimulus light mounted above the lever. The only time the stimulus light is
extinguished is during a post-reward 50-sec timeout period. Training and testing are
accomplished in a Coulbourn Instruments (Allentown, Pennsylvania, USA) com-
puter-controlled operant system that includes 16 operant chambers (represents an
upgrade from current six-station system) with light cues and retractable levers. Each
operant chamber is housed in a sound-attenuated and fan-cooled environmental
compartment. Rats that maintain at least 100 responses for three consecutive ses-
sions are surgically prepared for i.v. self-administration sessions as described below.
The pellet feeder is removed from the operant chamber (to be replaced by i.v. infu-
sion of morphine), but throughout the remainder of the study, the rats have unlimited
access to standard rat chow and water.
10.2.3 IMPLANTATION OF THE I.V. INFUSION LINE

The trained rats are anesthetized with sodium methohexital (65 mg/kg, i.p.) and
under aseptic conditions a midline incision is made ventrally over the neck region.
The jugular vein is exposed by blunt dissection and a small nick made to allow intro-
duction of a non-thromogenic, softening, vascular implant tubing (Data Sciences, St.
Paul, Minnesota, USA), which is filled with dilute heparinized (20 units/mL) sterile
saline. The tubing is advanced about 2.5 cm and then it is tied off and secured to
surrounding fascia. The tubing is tunneled under the skin to emerge at the nape of
the neck. There the tubing is stabilized to a subcutaneous plastic anchor button and
fixed to a water-tight swivel cannula mounted at the top of the operant chamber. The
swivel is connected to the computer-controlled infusion pump. Two days later the
patency of the venous infusion system is tested by the rapid i.v. administration of the
short-acting anesthetic agent sodium methohexital (2.0 mg). Intravenous administra-
tion of methohexital leads to immediate loss of muscle tone and righting reflex in
patent animals. The heparin-saline solution is maintained in the catheter until the
start of morphine/cocaine self-administration (third day after surgery). It is impor-
tant to maintain each rat in its previously assigned operant chamber.
10.2.4 THE MORPHINE REGIMEN

Rats are permitted to self-administer morphine sulfate according to an FR1 sched-
ule with a 50-sec timeout instituted after each infusion. An illuminated stimulus
light signifies the beginning of the session and indicates that the lever is active (one
press, one infusion). The light is extinguished and the lever is made inactive (presses
elicit no infusion) for 50 sec after a reward is delivered. The pump is set to deliver
morphine sulfate over a 5-sec period in 0.165 mL of saline. An experimental cohort
usually consists of eight rats (usually provides adequate statistical power). The start-
ing dose of morphine in a single infusion is 0.25 mg/kg. Animals are permitted
to self-administer 0.25 mg/kg/infusion over the first three days. During the next
three days animals self-administer 0.5 mg/kg/infusion, and during the next four days
they self-administer 1.0 mg/kg/infusion. The only interruptions in the 24-hr access
schedule are the brief periods needed to replace empty infusion syringes between the

changes in drug concentrations. Over the final five days, the morphine levels in the
infusion solution are tapered to 0.25 mg/kg/infusion (on days 10, 11, 12, 13, and 14
the morphine concentration is adjusted accordingly to 0.75, 0.75, 0.5, 0.5, and 0.25
mg/kg/infusion, respectively).
10.2.5 THE COCAINE REGIMEN

The procedure for cocaine i.v. self-administration is similar to that for morphine except
that the first dose of cocaine to be self-administered is 0.5 mg/kg/infusion, and the
timeout period after each infusion is 280 sec. The timeout period of 280 sec is insti-
tuted so that the animal cannot self-administer more than 300 doses per day. During
the next three days animals will self-administer 1.0 mg/kg/infusion, and during the
next four days they will self-administer 2.5 mg/kg/infusion. In practice, the animals
will maximally self-administer no more than about 150 doses in 24 hr (Figure 10.1).
Over the final five days, the cocaine levels in the infusion solution are tapered to 0.5
mg/kg/infusion (on days 10, 11, 12, 13, and 14 the cocaine concentration is adjusted
accordingly to 2.0, 2.0, 0.5, 1.0, and 0.5 mg/kg/infusion, respectively).
10.2.6 SPONTANEOUS MORPHINE WITHDRAWAL SYNDROME

On the final day of self-administration rats continue to have access to the morphine
infusion lever until 1700 hr, at which time the i.v. infusion line is disconnected and
plugged. Body weight and core temperature (rectal temperature measured by using
a thermistor probe) are recorded (0 hr post-withdrawal) and the animals are placed
back into their individual home cages. At 0900 hr on the following day (16 hr post-
withdrawal) body weight and core temperature are again measured. Next the rats
are placed in a standard open-field environment to assess abstinence signs associ-
ated with spontaneous morphine withdrawal by using a standardized withdrawal
checklist.8,9 These symptoms are scored by a “blinded” rater during 30-min obser-
vation periods. Withdrawal symptoms include withdrawal body (wet-dog) shakes,
escape attempts (attempting to leap out of the cage), writhing, defecation/diarrhea,
and chromodacryorrhea (reddish tears). This procedure is initiated three more times
at consecutive 2-hr intervals, and the incidences of the scored symptoms over the
four observation periods are totaled. In our previous studies, morphine withdrawal
symptoms were shown to peak between 12 and 20 hr after withdrawal and they
ended by about 100 hr.10,11 Body weight and core (rectal) temperature are measured
after the conclusion of the last observation period (1700 hr). Thereafter, body weight
and temperature are measured at 900 hr and at 1700 hr each of the next two days
(withdrawal symptoms will no longer need to be recorded), and body weight (which
requires the longest period to recover, Figure 10.2) is recorded once daily for the next
three days. After the last measurements, the animals are left undisturbed in their
individual home cages for the remainder of the 6 wk post-withdrawal period.
10.2.7 REINSTATEMENT OF LEVER PRESSING

After the 6-wk post-withdrawal period the subjects are returned to the operant cham-
bers. The condition of the chambers is identical to that during the morphine self-

N: 21 21 21 21 21 21 15 15 15 3 3 3 2 1
4
mg/kg/Infusion
3
2
1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
(a)
400
350
Lever Responses
300
250
200
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
(b)
200
180
Number of Infusions
160
140
120
100
80
60
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
(c)
400
Dose SA (mg/kg/24 hr)
200
120
100
80
60
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Days of Morphine Self-administration
(d)
FIGURE 10.1 The self-administration of an escalating dose regimen of i.v. morphine infu-
sion by 21 rats under a contingent FR1 schedule of reinforcement with 50-sec timeouts.
Access to the reinforcement response lever was available 24 hr per day. (a) The dose of
morphine self-administered with each infusion. The integers above the panel indicate the
number of animals completing the regimen to that point. (b) The number of reinforcement
lever responses per 24 hr. (c) The number of morphine infusions as a consequence of contin-
gent lever responses per 24 hr. (d) The dose of morphine self-administered per 24 hr. Error
bars refer to the SEM.

360
340
Body Weight (g)
320
300
280
260
240
0 20 40 60 80 100 120 140
Hr Post-withdrawal
FIGURE 10.2 The change in body weight in morphine-dependent rats after discontinua-
tion of self-administration.
administration phase, and each rat will always be returned to its original chamber.
Again rats are allowed to lever press according to the 24-hr access schedule used
during morphine or cocaine self-administration, including the (contingent) cue light.
The animals have unlimited access to standard rat chow and water. In this case the
i.v. line is not reconnected, and lever pressing will not result in a reward. Rats are
maintained in the operant chamber for at least seven days.
10.3 METHODOLOGICAL DETAILS
10.3.1 THINGS TO PREPARE BEFORE SURGERY

1. Order male Wistar rats, 300–324 g, from Harland, Indianapolis, Indiana,
USA. Allow the animals at least 1 wk to acclimate after shipping.
&RQQHFWWKH3(WXELQJWRWKHERWWRPRIWKHVZLYHO7KLVFDQEHGLI¿FXOW
,QVHUWDVPDOOJDXJHQHHGOHLQLW¿UVWWRVWUHWFKLWDOLWWOH7KHQLQVHUWD
piece of 22-gauge wire. This can help open the tubing. Attach the swivel to
the tubing. By doing this before surgery it can be reattached easily to save
time.
3. The surgical instruments are sterilized in a glass bead sterilizer (Fine Sci-
HQWL¿F7RROV)RVWHU&LW\&DOLIRUQLD86$ %HIRUHSODFLQJWKHLQVWUXPHQWV
in the sterilizer they should be cleaned and scrubbed in rubbing alcohol.
Insert the dry instruments into the beads at 250°C (it requires about 20 min
for the sterilizer to reach 250°& IRUDWOHDVWPLQ$OORZWKHLQVWUXPHQWV
to cool before starting surgery. The sterilizer does not need to be turned off
EHWZHHQVXUJLFDOSURFHGXUHV*ORYHVVXUJLFDOJRZQDQGPDVNVKRXOGEH
worn during all surgical procedures.
4. Dilute hospital-grade sodium heparin in sterile normal saline to 20
units/mL.

10.3.2 INSTALLING A PERMANENT INTRAVENOUS LINE

One hour before surgery rats are administered carprofen sterile injectable solution
purchased in 50 mg/mL vials (Pfizer Animal Health, Exton, Pennsylvania, USA).
Carprofen is an analgesic agent administered by subcutaneous injection at a dose of
5 mg/kg. Rats are anesthetized by i.p. injection with methohexital sodium (Brevital)
60 mg/kg. Brevital (purchased as 500 mg of the dry powder in 50 mL vial to provide
a working solution of 10 mg/mL) is a short-acting anesthetic and in most cases the
entire surgical procedure can be accomplished after a single injection. However,
if the animal begins to recover (assess response to toe pinch), an additional 5 mg
of Brevital should be administered. The fur should be clipped ventrally over the
neck and on the area dorsally over the neck just behind the head after the animal is
anesthetized. Next, the exposed skin in the surgical fields should be swabbed with
Betadine antiseptic solution. A midline incision is then made in the skin over the
ventrolateral aspect of the neck. The incision is about 2-cm long. A small pair of
round-tipped scissors can be used to bluntly dissect the fascia without cutting. The
right exterior jugular vein will be exposed. Care should be taken not to overly stretch
the vein as it will become narrower and more difficult to cannulate. The vein can be
isolated from the surrounding fascia by using curved forceps, but care must be taken
not to puncture the vein. Place a small forceps under the vein to secure it. Place two
pieces of 3-0 black braided suture at the top and bottom of the cleared vessel. Tie
a knot on the cranial aspect of the vein to seal it off. Make the venotomy (incision)
approximately 5 mm cranial to the site of crossover of the pectoralis major muscle.
Before inserting the small animal vascular catheter into the vein, cut a bevel on the
end to be inserted. Make sure that the tip is not too sharp. The catheter should be
about 20-cm long. Fill the cannula with heparinized saline (20 units/mL). Insert a #7
Dumont tweezer into the vein and lift slightly, i.e., insert the right tip of the forceps
into the vein and then grasp the vein. Slowly insert the small animal tubing into
the vein for about 1.5 cm. Hold the vein and tubing with the Dumont tweezer until
insertion is complete. (If nick in the vein is difficult to see, the vein can be “milked”
by gentle rubbing and getting the blood to flow again. Application of saline to the
cut end could also help.) Tie the vein and tubing in with the 3-0 suture that was
placed there earlier. Both sutures should be tied to the vein to secure the catheter. A
subcutaneous tunnel should be created from the neck area to the dorsum. A small
incision (about 1.5 mm) is made in the skin just behind the head. Make it a pocket
by inserting the scissors and opening them to clear the fascia away. Pass a trochar
(point-sharpened 14-gauge stainless steel tubing) between the skin and muscle lay-
ers. (Note: Do not pass through the muscle.) around to the back of the neck behind
the head and out the small incision. Pass the catheter through the trochar. Remove
the stainless steel tubing. Pull the catheter through the anchor button and place the
button under the skin. Suture the skin to the button with 3-0 black braided suture.
Attach a connector made from 26-gauge stainless steel tubing to the catheter. Next,
place a piece of PE 20 tubing filled with heparin saline to the connector. Place a
piece of Vardex tubing about 4-in long over the spring support. (This is optional
but keeps the rat from chewing into the spring support.) Feed the tubing through

the spring support and attach the spring to the button. The tubing from the bottom
of the swivel to the top of the animal should be long enough to include a loop in the
top before attaching the catheter to the swivel (approximately 13-in long—you can
always shorten it by cutting some from the end that joins to the connector). The loop
should be small, about 2.5 cm in diameter. The loop absorbs tension on the catheter
as the animal moves. If the loop is too large it can get caught on the swivel holder.
Note: Coulbourn tethers have attachments at the bottom to fit their rodent harnesses.
We do not use the harness and so the attachments are clipped off with a wire cutter.
Tethers are cut to 32 cm. With a dull pair of scissors, squeeze between turns on the
spring to create an offset. The output end of the swivel is twisted tight into the entry
of the offset and the looped PE 20 tubing from the rat connects to the swivel output.
This offset allows the rat to move more freely in the Habitest cage. Place a 5-in
piece of Tygon micro-bore tubing, 0.2 in inner diameter (id) × 0.6 in outer diameter
(od), on the input of the swivel. Fill the swivel and tubing with heparinized saline.
Attach the swivel to the spring and then attach the looped tubing to the bottom of
the swivel. Fill a 60-mL syringe with 30–40 mL of heparin saline. Insert a blunt 22-
gauge needle into the end of the syringe and attach it to 0.2 in id × 0.6 in od Tygon
tubing long enough to reach the top of the swivel. Place a 22-gauge stainless steel
connector on the end of the tubing. Fill the tubing with heparin saline or drug. Con-
nect it to the tubing from the swivel input. Infuse heparin saline at a rate of about 7.5
mL/day. Place the rat in the Habitest cage. Hook the swivel to the swivel holder. A
weight at the back of the balance arm can be moved to make the line taut. Place the
water bottle on the cage and place food in bottom of cage. (Note: Place 35 g of rodent
chow per day.) Food consumption should be recorded on a daily basis. The Habitest
Linc program for controlling the operant aspects of the task and for recording lever
responses can be initiated at any time.
10.3.3 SELF-ADMINISTRATION DOSE CALCULATIONS

Morphine Self-Administration
Day 1–3 0.25 mg/kg/day

Cocaine Self-Administration

Sample Calculations for Preparing Stock Solution of Drugs for Self-

Administration: For a rat with average body weight of 300 g and a required dose of
0.25 mg/kg/infusion:

Infusion pump is set to 0.165 mL/infusion

Dose (mg/kg/infusion)/0.165 mL/infusion × weight in kg = mg/mL
0.25 mg/kg/infusion × 0.3 kg/0.165 mL/infusion = 0.455 mg/mL (concen-
tration of stock solution)
Prepare enough drug solution to fill all infusion syringes.
To calculate daily (24 hr) dose self-administered, divide the dose infused (mg/
kg/infusion) by the number of infusions self-administered per 24 hr.
10.3.4 ORDERING INFORMATION

Small animal vascular catheter Data Sciences #277-0011-002
St. Paul, MN 55126-6164 USA
Button for DC95 tethers Instech #DC95BS
www.instechlabs.com
Tygon tubing 0.02 w 0.06 VWR 63018-044 or
Fisher 14-170-15B
Stainless steel tubing and wire
14-gauge stainless steel tubing Small Parts HTX-14
www.smallparts.com
22-gauge stainless steel wire Small Parts J-SWX-022
PE20 tubing 0.015 w 0.043 VWR 63019-025
Male Wistar rats 300–324 g Harlan, Inc. Indianapolis, IN, USA
Glass bead sterilizer Fine Scientific Tools
FST 250 sterilizer
info@finescience.com
Rodent 22 g swivel Braintree Scientific RS-22G
(Need this or the ones from Coulbourn) www.braintreesci.com
Optional
Vardex interbraided tubing 0.25 id w 0.453 od Newage Industries
#140-0070-100
Southampton, PA 18966 USA
www.Newageindustries.com
Utility cutter short-angled blade scissors Fisher Scientific 14-277
SuperCut iris scissors, straight, 4.5 in Braintree Scientific SC528
www.braintreesci.com

10.3.5 COULBOURN INSTRUMENTS—REQUIRED EQUIPMENT

Description Part Number
Cue, single high bright light—rat H11-03R
House light H11-01R
Balance arm H-29-01
Lever H21-03R
Swivel A73-51
Panel, blank metal assortment H-90-00R-M-KT01
Pump, infusion—programmable E73-02
Habitest power base w/lid H01-01
Habitest Linc H02-08 (for two rats)
Environmental control board H03-04
Test cage H10-11R-TC
Non-shock floor H10-11R-TC-NSF
Catheter protection and tethers A73-59R
Rat harness to fit 150–500 g rat A71-21R-350/500
Graphic State software, version 3.03 GS3.03
PCI-3 computer interface card PCI-3-Kit
Computer controller requirements Sys Control 1F P4, 3.2 GHz , 1GB RAM,
250 GB HD, 17-in LCD, DVD-RW
(May use Windows XP with PC1 slot and
turn-key package)
Note:7RVHOIDGPLQLVWHUIRRGSHOOHWVDIHHGHUDQGWKHZLUHVWRFRQQHFWWKHPWRWKH
board are required. There may be other cables and accessories required depending
RQWKHFKRVHQV\VWHPFRQ¿JXUDWLRQ7KLVLQIRFDQEHREWDLQHGIURP&RXOERXUQ
,QVWUXPHQWV ZZZFRXOERXUQFRP
10.4 ACTUAL DATA
10.4.1 MORPHINE SELF-ADMINISTRATION

These data reflect the trial of different morphine self-administration regimens.
Twenty-one rats initially participated in the first experimental series, and each animal
was trained to self-administer 0.25 mg/kg doses of morphine. These data are pre-
sented in Figure 10.1. Animals generally maintained their level of lever responding
(about 100 responses/24 hr) that carried over from the earlier lever training experi-
ments where food pellets provided the response motivation. That the animals trans-
ferred this behavior to morphine-reinforced responding was insured by the removal
of the food hopper (the space was covered by a flat insert indistinguishable from the
surrounding wall), and by the continuous availability of food in the operant chamber.

Also, in separate studies in which rats shaped on food rewards were transferred to
saline self-administration, the lever responding that carried over extinguished over
the same time period.12 Note that it is possible that rats can learn to self-administer
without the use of prior shaping with food reinforcement. This is because with 24-hr
access, inadvertent lever responses occur with sufficient frequency to help encourage
the behavior. Should you use this approach, you should be prepared to encounter some
animals that fail to lever respond sufficiently during the first few days to the extent
that they are removed from the study. As indicated in Figure 10.1A, lever responding
and the daily dose self-administered increased over the first 3 days. Lever respond-
ing became relatively constant over the next 6 days during the self-administration of
the 0.5 mg/kg and the 1 mg/kg doses. Though the numbers of animals remaining in
the study decreased dramatically over the last 4 days (the self-administration phase
was terminated at different times to enhance the variability in the total amount of
morphine consumed prior to withdrawal), there was a dramatic decrease in respond-
ing when the infusion dose was increased to 2 mg/kg. The level of responding recov-
ered even for the two rats that self-administered 4 mg/kg/infusion. The daily dose
of morphine was maintained fairly constantly during the self-administration of the
0.5 mg/kg and 1 mg/kg doses per infusion (Figure 10.1D). This was evident in the
observation that after day 6 (the last day of 0.5 mg/kg/infusion), when the dose of
morphine was increased to 1 mg/kg/infusion, responding decreased slightly so that
the average total dose self-administered could be maintained at about 70 mg/kg/day.
This profile of responses, and the amount of morphine self-administered, was again
apparent in transitioning from the 1 mg/kg to the 2 mg/kg doses per infusion. Two
animals continued to self-administer 4 mg/kg morphine per infusion with additional
fall-off in the number of lever responses. By varying the number of days of self-
administration, and allowing some animals to self-administer high concentrations of
morphine, we were able to obtain a broad spectrum of total morphine doses as well
as the associated total number of lever responses. For morphine self-administration,
we allow rats to self-administer only up to 1 mg/kg/infusion. If the animals maintain
a good level of responding (> 70 responses per day) they will become dependent on
the drug after about 4–5 days self-administering this dose.
At the completion of the self-administration phase of the study, animals are
returned to their home cages and allowed to undergo withdrawal. Figure 10.2 shows
the change in body weight as a function of time after withdrawal. There was a char-
acteristic sharp decrease in body weights averaging 16.2 g measured 36 hr after
withdrawal. Thereafter, body weight gradually increased to near control levels by 84
hr post-withdrawal. The animals continued to gain weight through the last observa-
tion period at 6 days after withdrawal. More importantly, the withdrawal-associ-
ated decrease in body weight was shown to be linearly related to the total dose of
morphine self-administered.12 These data therefore relate the quantity of morphine
consumed during the dependence phase to magnitude of the expression of this with-
drawal symptom. In general, other withdrawal symptoms were not as dramatic or
intense as with opiate antagonist-precipitated withdrawal, and the most prevalent of
the visually observed symptoms was withdrawal body shakes. Of the other symp-

700
600
Lever Responses
500
400
300
200
100
0
1 2 3 4 5 6 7
Days Post Lever Reinstatement
FIGURE 10.3 &RQWH[WLQGXFHGUHLQVWDWHPHQWRIOHYHUUHVSRQGLQJZNDIWHUGLVFRQWLQXD-

tion of morphine self-administration in dependent rats. The number of lever responses (per
KU GXULQJWKHGD\OHYHUUHLQVWDWHPHQWSHULRGZLWKOHYHUUHVSRQGLQJRQWKHODVWGD\RI
the morphine self-administration regimen.
toms, defecation and diarrhea were noted most often, though these are not necessar-
ily characteristic withdrawal symptoms.
10.4.2 CONTEXT-INDUCED POST-WITHDRAWAL LEVER RESPONDING

After completing the 6-wk withdrawal period, rats were placed back into their origi-
nal operant environment and allowed to make lever responses with no reward con-
sequences (even though the illuminated stimulus light continued to signal an active
lever). These data are presented in Figure 10.3. Returning the animal to the operant
chamber resulted in a doubling of the average 24-hr lever response rate that was
measured on the last day of the self-administration schedule. Average responding
decreased over the next two days, but it rebounded during test days 4 and 5. By day
7 the response rate extinguished to that measured during the last day of self-admin-
istration. The differences among the means for each day of testing relative to the
pre-withdrawal level of responding was statistically significant (P < 0.01).
10.5 DISCUSSION
10.5.1 THE DRUG ABUSE CYCLE AND PHARMACOLOGICAL INTERVENTION

Animals that self-administered morphine according to our standard protocol were
shown to be physically dependent on morphine. This was evidenced by the appear-
ance of characteristic symptoms of abstinence, and more specifically by the precipi-
tous decrease in body weight that was maximal on the second day after morphine
withdrawal. After monitoring body weight for 6 days, the rats were returned to their
home cages to complete the 6-wk protracted abstinence period. Harris and Aston-
Jones13 reported that the preference for a morphine-paired environment in formerly

dependent rats was maintained from 2–5 wk post-withdrawal. In fact, increased drug-
seeking behavior after a protracted term of abstinence was noted for several drugs
of abuse.14 Likewise, in the present study, returning formerly dependent animals
to the self-administration environment resulted in an initial doubling of the 24-hr
response rate as compared with the last few days of self-administration. Extinction
of enhanced lever-pressing activity came slowly, but self-administration levels were
attained by day 7.
The 24-hr access model described here provides a step closer in relevance to real-
world conditions than studies that rely on one- or two-hour daily test sessions that
often occur during the animals’ sleep cycle. Another important feature of the para-
digm is that treatment interventions can be studied within the same model for each
component of the drug abuse cycle: self-administration, dependence, withdrawal,
protracted withdrawal, and renewed drug-seeking behavior. Figure 10.4 illustrates
the days of administration for each treatment intervention (arrowheads) across the
three regimens. The regimen outlined in the uppermost graph of Figure 10.4 permits
the evaluation of treatment intervention administered during the self-administra-
tion phase on subsequent acute and protracted withdrawal behaviors. Note that the
first treatment intervention occurs on day 9—the last day of the highest concentra-
tion self-administered. This paradigm is designed to examine the direct effect of
treatment intervention on self-administration behavior. At the conclusion of day 10,
self-administration is maintained over the following 5 days. Treatment intervention
can continue to be administered up until the start of withdrawal at the end of day 14.
This paradigm of interdiction after the development of physical dependence is more
clinically relevant than beginning the treatment intervention simultaneously with
the start of self-administration, though the latter paradigm can be used to assess the
effects of treatment interventions designed to inhibit self-administration behavior.
When treatment intervention is initiated at the start of the self-administration period,
if lever responding is decreased, the expression of downstream withdrawal symp-
toms will automatically be reduced. Thus the regimen outlined in the upper graph of
Figure 10.4 circumvents this limitation.
The regimen outlined in the middle graph of Figure 10.4 permits the evaluation
of treatment intervention during the acute withdrawal period on acute and protracted
withdrawal behaviors. The regimen outlined in lowermost graph of Figure 10.4 per-
mits the evaluation of treatment intervention during the protracted withdrawal period
on protracted withdrawal behavior (contextually induced lever responding). There-
fore, insight can be gained into the specificity of each treatment intervention on the
component of the drug abuse cycle, and information can be obtained regarding the
role of each preceding component on the expression of the subsequent components
of the cycle.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 40 41 42 43 44 45 46 47 48 49
Self-administration
(Dependence) Acute
Protracted
Withdrawal Contextual Reinstatement
Withdrawal
(drug seeking)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 40 41 42 43 44 45 46 47 48 49
Self-administration
(Dependence) Acute
Protracted
Withdrawal
(drug seeking)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 40 41 42 43 44 45 46 47 48 49
Self-administration
(Dependence) Acute
Protracted
Withdrawal
(drug seeking)
FIGURE 10.4 The phases of the drug-abuse cycle as provided by the 24-hr access model
for self-administration. Arrows indicate the days in which a treatment intervention is
administered.
REFERENCES
1. Deroche-Gamonet, V., Belin, D., and Piazza, P. V. 2004. Evidence for addiction-like
behavior in the rat. Science 305:1014–17.
2. Vanderschuren, L. J. M. J., and Everitt, B. J. 2004. Drug seeking becomes compulsive
after prolonged cocaine self-administration. Science 305:1017–19.
3. Koob, G. F., and Le Moal, M. 2001. Drug addiction, dysregulation of reward, and allo-
stasis. Neuropsychopharmacology 24:97–129.
4. Shalev, U., Grimm, J. W., and Shaham, Y. 2002. Neurobiology of relapse to heroin and
cocaine seeking: A review. Pharmacological Rev. 54:1–42.
5. Aston-Jones, G., and Harris, G. C. 2004. Brain substrates for increased drug seeking
during protracted withdrawal. Neuropharmacology 47(suppl 1):167–79.
6. Vorel, S. R., Liu, X., Hayes, R. J., Spector, J. A., and Gardner, E. L. 2001. Relapse to
cocaine-seeking after hippocampal theta burst stimulation. Science 292:1175–78.
7. Fuchs, R. A., Tran-Nguyen, L. T. L., Specio, S. E., Groff, R. S., and Neisewander, J. L.
1998. Predictive validity of the extinction/reinstatement model of drug craving. Psy-

8. Buccafusco, J. J., Marshall, D. C., and Turner, R. M. 1984. A comparison of the inhibi-
tory effects of clonidine and guanfacine on the behavioral and autonomic components
of morphine withdrawal in rats. Life Sci. 35:1401–08.
9. Marshall, D. C., and Buccafusco, J. J. 1985. A comparison of the cardiovascular and the
behavioral changes following naloxone as measures of the degree of physical depen-
dence on morphine in rats. Drug Devel. Res. 5:271–80.
10. Buccafusco, J. J. 1983. Cardiovascular changes during morphine withdrawal in the rat:
Effects of clonidine. Pharmacol. Biochem. Behav. 18:209–15.
11. Zhang, L. C., and Buccafusco, J. J. 1998. Prevention of morphine-induced muscarinic
(M2) receptor adaptation suppresses the expression of withdrawal symptoms. Brain
Res. 803:114–21.
12. Buccafusco, J .J., and Bain, J. N. 2007. A 24-hour access i.v. self-administration sched-
ule of morphine reinforcement and the estimation of recidivism: Pharmacological
modification by arecoline. Neuroscience 149:487–89.
13. Harris, G., and Aston-Jones, G. 2003. Enhanced morphine preference following pro-
longed abstinence: Association with increased Fos expression in the extended amyg-
dale. Neuropsychopharmacology 28:292–99.
14. Harris, G., and Aston-Jones, G. 2001. Augmented accumbal serotonin levels decrease
the preference for a morphine associated environment during withdrawal. Neuropsy-

11 Operant Analysis
of Fronto-striatal
Function in Rodents
Máté D. Döbrössy, Simon Brooks, Rebecca
Trueman, Peter J. Brasted, and Stephen B. Dunnett
CONTENTS
11.1 Introduction................................................................................................. 215
11.2 The Neuropathological and Behavioral Profile of HD ............................... 216
11.2.1 HD Pathology................................................................................... 216
11.2.2 HD Symptomatology ....................................................................... 217
11.3 Animal Models of HD ................................................................................ 218
11.4 Operant Conditioning and Operant Chambers ........................................... 220
11.4.1 Operant Chambers ...........................................................................220
11.4.2 Operant Tasks to Assess Striatal Function in Rodents .................... 222
11.5 Operant Analysis of Motor Responding: Striatal Lesioned and HD
Transgenic Animals .................................................................................... 223
11.5.1 Operant Analysis of the Sensory and Motor Aspects of
“Sensorimotor” Striatal Neglect ...................................................... 223
11.5.2 Operant Tasks to Delimit the Specificity of Striatal Neglect .......... 225
11.6 Operant Analysis of Cognitive Tasks: Striatal Lesioned and
Genetically Modified HD Models............................................................... 228
11.6.1 Delayed Matching Tasks.................................................................. 229
11.6.2 Delayed Alternation Tasks............................................................... 230
11.6.3 5-Choice Reaction Time Task.......................................................... 235
11.6.4 Serial Implicit Learning Task .......................................................... 236
11.7 Operant Analysis of Striatal Lesions: Deficits in Motivational State......... 238
11.8 Conclusion...................................................................................................240
References.............................................................................................................. 241
11.1 INTRODUCTION
The basal ganglia were once believed to function as part of an “extrapyramidal”
motor system, operating separately from the pyramidal tract.1,2 However, this con-
cept has been discarded for two fundamental reasons. First, the basal ganglia have
been shown to be an intrinsic part of well-defined anatomical circuits that not only
receive cortical input but also send projections, via the thalamus, back to those
215

cortical areas that control motor output. Second, a wealth of experimental work
has shown that the striatum, the main input structure for the basal ganglia, can no
longer be regarded purely as a “motor” structure. The observation that striatal dam-
age could induce deficits in cognitive function led researchers such as H. Enger Ros-
vold and Ivan Divac to state that the striatum may reflect the function of those areas
of neocortex that project to it.3 These pioneers investigating striatal function laid
the foundation for the “functional loop” concept that proposes multiple, topographi-
cally arranged basal ganglia circuits that serve as substrates for motor, oculomotor,
prefrontal, and limbic functions.4 The theory that the striatum may mediate a wide
variety of functions reflecting its diverse cortical innervation has become evident
in studies of patients with basal ganglia disorders. Thus, impairments in cognitive
function are now well documented in patients with neurodegenerative diseases such
as Huntington’s disease (HD) or Parkinson’s disease (PD), disorders which were
once regarded as entirely “movement” related.
Attempts to examine disease states such as HD in experimental animals can
provide both insight into normal brain function and a means by which to assess
potential therapeutic strategies. In either scenario, an operant analysis of behavior
can prove particularly powerful. The detailed functional analyses that are permitted
by operant paradigms not only allow more specific questions to be asked of normal
brain function, but can also provide experimental paradigms that are extremely sen-
sitive to brain insults and subsequent recovery.
11.2 THE NEUROPATHOLOGICAL AND

BEHAVIORAL PROFILE OF HD
11.2.1 HD PATHOLOGY
Originally reported by George Huntington in 1872,5 HD is a fatal inherited neu-
rodegenerative disorder, the genetic basis of which has recently been identified.6
The predominant pathological signature of the disease is the early and progressive
loss of GABAergic medium spiny projection neurons from within the striatum (cau-
date nucleus and putamen); however, anatomical changes in other regions have been
described in preclinical stages of the disease.7 The degeneration begins in the caudate
nucleus and progresses through the entire striatum in a medial-to-lateral and dorsal-
to-ventral fashion.8,9 Striatal degeneration involves, in particular, loss of the medium
spiny projection neurons, with relative sparing of the large aspiny interneurons.
There is post mortem evidence that the earliest striatal neurons affected are those in
the striosomes, projecting to the substantia nigra pars compacta, and Hedreen and
Folstein have proposed that more diverse striatal projections via this nigral feedback
are an essential component in the spread of the disease.10 As the disease progresses,
striatal atrophy, gliosis, and cell loss becomes progressively more marked, which
has been characterized by Vonsattel in a widely used five-point grading system.8 In
advanced disease, not just the striatum but widespread areas of the forebrain—in
particular areas such as the neocortex or substantia nigra pars reticulata that are
sites of afferent and efferent connections with the striatum—also undergo atrophy

Operant Analysis of Fronto-striatal Function in Rodents 217
and cell loss. Other changes accompanying the disease include loss of overall brain
weight and ventricle enlargements.
One of the pathological hallmarks of HD is the appearance of nuclear aggregates
containing a truncated fragment of the expanded mutant polyglutamine repeat of
the huntingtin protein. The gene and its protein product have been the subject of a
great wealth of research since its discovery in 1993; nevertheless, the mechanism
of its toxicity remains yet unresolved.6 Transgenic models of the disease replicate
the formation of abnormal nuclear inclusions of N-terminal truncated fragments of
huntingtin,11–14 which have now been confirmed as being widely distributed in the
brains of affected individuals.15 However, previously thought of as a predictor of cel-
lular dysfunction, the balance of evidence currently points to the truncated hunting-
tin fragment containing aggregates as a consequence of the disease process, rather
than a cause, and some even argue that they might be neuroprotective.16 Evidence
is growing that the mutant form of huntingtin is directly responsible for the disrup-
tion of a wide range of essential cellular processes, including transcription and the
transport of trophic support to the striatum, protease cascades, and mitochondrial
energy metabolism.17
11.2.2 HD SYMPTOMATOLOGY
The uncontrollable movements (or “chorea”) that characterize HD are now recog-
nized to be only one part of the disease’s behavioral profile. In fact, HD presents
with a triad of motor, cognitive, and affective symptoms, all of which worsen as
the disease progresses inexorably, in parallel with the progress of the underlying
degeneration. Indeed, the introduction of genetic screening in conjunction with more
sensitive imaging and clinical tests has led to the suggestion that subtle cognitive and
psychiatric aspects of the disease are apparent before the onset of chorea.7,18 A large
multi-center longitudinal study, PREDICT-HD, is now in progress with the aim of
characterizing early, preclinical, and presymptomatic anatomical, motor, and cogni-
tive changes in patients that carry the expanded mutant form of the huntingtin gene.
The aim of the study is to promote better design and outcome measures for preven-
tive clinical trials in HD.19
The most striking aspect of HD is the chorea that originally gave its name to the
disorder, until the diverse nature of impairments in HD was acknowledged. However,
these unwanted choreic actions often mask an underlying bradykinesia, and deficits
in initiating responses in reaction time paradigms20,21 are indicative of impairments
in initiating and selecting motor programs. Consequently, it is now recognized that
both hypokinetic and hyperkinetic symptoms coexist in HD.22–24
Almost four decades ago, Divac proposed that the striatum has a role in both
motor and higher order functions,25,26 and it has now become widely accepted that
HD patients express a profile of cognitive and neuropsychological deficits similar to
that seen in patients with prefrontal cortical damage.27,28 This includes impairments
in learning29,30 and working memory,30,31 as well as deficiencies in “executive” tasks
that assess planning and attentional control.31,32 A number of psychiatric symptoms
are also present in HD, such as depression and anxiety.33

11.3 ANIMAL MODELS OF HD

HD is a uniquely human predicament and the experimental exploration of the dis-
ease, and of potential therapies, necessitates access to animal models. Numerous
models are available. However, the advantages and disadvantages of each needs to
be understood so that the appropriate model is used to shed light on particular ques-
tions.34 By recreating a milieu in experimental animals that resembles the pathology
of diseases such as HD, we can attempt to define the specific functional role of par-
ticular neural mechanisms more precisely than can be achieved in the human condi-
tion. These animal “models” of human disease can also provide a basis for assessing
potential strategies for repair (e.g., neural transplantation, neuroprotective agents,
or gene therapy) by giving rise to measurable changes against which the functional
efficacy of any particular strategy can be evaluated. The behavioral impact of striatal
damage on motor and cognitive function, as well as the motivational state, can all be
examined experimentally in the rat and the mouse. The two main classes of rodent
models currently being used to investigate fronto-striatal function in HD, induced
lesion (excitotoxic and metabolic toxins) and genetically modified models, are the
subject of a brief analysis below.
Striatal function was initially studied in experimental animals with the use of
basal ganglia lesions.2,35 However, there were major difficulties in interpreting the
consequences of lesions made by electrolysis, radiofrequency, or direct surgical
excision because of the inevitable damage to the immediately adjacent afferent and
efferent fibers of the internal capsule connecting the cortex to subcortical structures
including the thalamus. However, this changed dramatically with the introduction
of excitotoxic lesions in the mid 1970s, opening the way for the modern era of basal
ganglia research. The primary excitotoxins are amino acids (such as monosodium
glutamate, N-methyl-D-aspartic acid and kainic acid), which are glutamate agonists
that can be toxic above physiological doses by acting on glutamate-receptor-bear-
ing neurons, a feature of most neurons of the nervous system. When administered
directly into the striatum, excitotoxins specifically target and kill neurons within the
striatum without damaging the axons of the corticofugal and corticopetal pathways
passing through and adjacent to the striatum. Furthermore, injections of excitotoxins
into the striatum produce neurochemical and pathological changes similar to that
seen in HD. Initially kainic and ibotenic acid were used for this purpose.36–38 How-
ever, quinolinic acid has become the toxin of choice on account of numerous neuro-
chemical studies that demonstrate a more selective neuronal loss within the striatum,
with the medium spiny GABA neurons being particularly vulnerable, and the large
aspiny cholinergic, neuropeptide Y and NADPH-diaphorase–positive interneurons
being relatively resisitant to this toxin, corresponding to the profile of degeneration
observed in HD.39–41
While a single neurotoxic insult is able to mimic the neuropathology of HD, it
cannot reproduce the slow and progressive degeneration that is a characteristic fea-
ture of the human disease. These features can be better mimicked by metabolic tox-
ins, such as 3-nitropropionic acid, which target the striatal neurons selectively, even
when administered peripherally.42,43 Nevertheless, excitotoxins continue to be widely

used because of the fact that they typically produce more convenient, consistent, and
reproducible lesions than appear achievable with the metabolic toxins.44
The identification of the mutant huntingtin gene has permitted the production
of genetically modified animal models, predominantly, but not exclusively, based on
mice, and include transgenic, knock-in, knockout, and virally inserted polyQ tract
models.34 These models have the potential to be more authentic to HD on the grounds
that any neuropathology and behavioral deficits are caused by the expression of the
mutant huntingtin gene, and the onset of the “disease state” is progressive. However,
there are significant differences between models of genetically modified animals
regarding both behavioral and cellular consequences of the expression of the mutant
gene, the rate of onset of the pathology, the distribution pattern of the inclusions, the
degree of neuronal death, and the expected life span of the animals. The variability
between models is attributed to several factors, including the diverse length of the
CAG repeat present on the mutant huntingtin gene, whether an exon 1 or the full
length insert is used, and the technical method used for the insertion of the repeat, as
well as the characteristics of the specific background strain used to generate the ani-
mal model. As a consequence, the suitability of the genetically modified model will
depend on the issues on which the investigator is focusing. For example, studying
the process of formation and modification of the nuclear inclusions, or the changes
in energy metabolism induced by the mutant huntingtin protein, or the subsequent
behavioral consequences, might require a different HD animal model. The best
characterized transgenic mouse models are the R6/2 lines generated by Gill Bates
and colleagues.45 These mice show a clear profile of motor and cognitive deficits.46,47
However, the broad extra-striatal profile of the pathology and very rapid progression
of the disease48 make detailed analysis of the behavioral impairment difficult. By
contrast, the knock-in models show a slower, more progressive impairment, often
with focal striatal pathology, making them more suitable for detailed analysis of
specific cognitive deficits. The best characterized knock-in mouse in our hands for
behavioral studies is the Hdh Q92/Q92 (Q92) mouse. These knock-in mice have 90 CAG
repeats inserted into their endogenous huntingtin gene sequence resulting, on aver-
age, in 92 glutamine repeats.49 The animals develop nuclear inclusions and behavior
impairments, which are the subject of further discussion later on in the chapter.
The destruction of striatal cells in genetically modified models or with excito-
toxins not only produces some of the pathological hallmarks of HD, but can also
give rise to behavioral sequelae, which reflect many of the symptoms seen clini-
cally. Both approaches have advantages, and the use of one should not exclude the
other. The first use of excitotoxins to model the pathology of HD showed unilateral
striatal lesions to induce a marked rotation towards the ipsilateral side 48 hr after
surgery.38 This motor asymmetry reflected the inability of the lesioned striatum to
mediate contralateral movement, and this biasing of motor output to the side ipsilat-
eral to the lesioned striatum is evident in many indices of striatal dysfunction such as
amphetamine-induced rotation50,51 or the elevated body swing test.52 An additional
advantage of unilateral striatal lesions is that they can allow a within-subject analy-
sis of dysfunction and recovery. Paw-reaching,53,54 as well as several of the oper-
ant tasks elaborated below, typically take advantage of this asymmetry and allow
performance mediated by the intact striatum to be compared with performance that

is under the control of the lesioned striatum. However, this laterality of motor func-
tion is less applicable in tests of cognitive function or motivation, and paradigms
that assess these aspects of behavior typically employ bilateral lesions. On the other
hand, the main advantage of some of the genetically modified models is that they
offer a more mechanistic understanding of the abnormalities of HD by reproducing
specific aspects of the pathogenetic process of the human disease. Studying the onset,
the progression, and the severity of the behavioral consequences of the disease, in
conjunction with the cellular and molecular events in the genetically modified HD
models will promote understanding of the relationship between mechanism of cell
death, the expanded polyglutamine repeat, the formation of intranuclear inclusions,
and the behavioral symptoms.
11.4 OPERANT CONDITIONING AND OPERANT CHAMBERS

Although many standard tests, such as rotation, can provide a useful index of func-
tional capacity, it is often desirable to examine a subject’s ability to perform an
action, or series of actions, that is more purposeful, or goal directed. Thus, rather
than assess a general and poorly defined motor capacity, one can examine the capa-
bility of an animal to produce specific motor responses that are required to achieve
a particular outcome. Operant conditioning refers to this ability to train animals to
perform purposeful actions, and the underlying learning processes involved.
In an operant task, a subject learns to respond in a particular way to gain reward
(or to avoid punishment) when a discrete cue, or stimulus, signals that this response
will be positively reinforced. The principal advantage of modern operant testing
is the absence of interference from the investigator, as the tests are conducted in
isolated, soundproof boxes controlled by a computer. As it is the investigator who
designs the computer program, and hence the task, the variety of paradigms one can
use are almost without limit. Furthermore, operant tasks also permit the collection
and study of a large number of response parameters, enhancing the analysis of the
behavioral performance.
11.4.1 OPERANT CHAMBERS

A standard operant chamber will typically contain:
• The operandum, such as a lever, on which an animal can “operate,” i.e., act
on or respond to.
• Reinforcers, stimuli that increase the likelihood of responding. Animals
may either act to obtain positive reinforcers (e.g., food or water), or alter-
natively they may act to terminate or postpone negative reinforcers (e.g.,
electric shock).
• Discriminative stimuli, typically lights, sounds, or olfactory cues, the prop-
erties of which signal the timing and location of reinforcers and thereby
control the animal’s responding.

FIGURE 11.1 Schematic illustration of a two-retractable lever operant chamber (Skinner box).
The operant chamber is an automated apparatus designed to present discriminative

stimuli (e.g., lights that differ in color, location, or intensity) to record and measure
responses to the operanda (e.g., presses of the lever, presses of the panel covering the
food well, licks at a drinking tube) and deliver reinforcements (e.g., by operating a
dispenser to deliver food pellets). The contingencies for particular experiments can
be specifically included in the software running the task.
The classic test apparatus for evaluating operant behaviors is the Skinner box,
an automated test apparatus first devised and developed by B. F. Skinner when ana-
lyzing the behavior of rats responding to obtain food reward.55 As illustrated in Fig-
ure 11.1, a typical Skinner box provides one or two levers as operanda to which the
rat may respond. The timing of responding into discrete trials can be achieved by
making the levers retractable and only available at discrete points within the trial.
Discriminative stimuli are provided by a variety of different lights located above
the levers, above and within the food hopper, and in the roof of the test chamber. A
loudspeaker in the chamber can also present auditory stimuli either as white noise or
discrete tones of controlled frequency and intensity. A variety of different reinforc-
ers can be built in. This is most typically either a dispenser to deliver food pellets or
a liquid dipper to present water into the reward chamber. These will only be effective
if the animal is suitably motivated by hunger or thirst, achieved by some hours of
food or water deprivation, respectively, prior to the training session. However, other
reinforcers are also possible, such as presentation of a receptive female rat to a male
rat in studies of hormonal control of sexual motivation.56
An alternative type of operant chamber that has proved highly effective, in par-
ticular for mice, is the nine-hole box (9-HB) (Figure 11.2). The 9-HB is conceptually

Stimulus
Lights
Photocell Well
Detectors Blanks
Food
Well
FIGURE 11.2 Schematic illustration of the nine-hole box operant chamber.
similar to the standard Skinner box except that instead of levers, the box is supplied
with an arc of nine square-shaped holes. Discriminative stimuli are provided by
lights at the rear of each hole, responses (nose pokes) are monitored by infrared
beams at the entrance to each hole, and food pellets or liquid reinforcements are
delivered to a well positioned at the rear of the chamber. Rats typically investigate
stimulus lights associated with food reward by producing a nose poke into individual
holes, which is arguably a more ethologically natural action for a rat than press-
ing or releasing a lever. Although originally designed to assess attentional function,
the physical configuration of the apparatus in the 9-HB allows great specificity in
defining both stimuli and responses, and this has allowed for the laterality of motor
function evident in structures such as the striatum to be analyzed with great preci-
sion (see below).
Operant chamber paradigms enable far more precise control of the factors deter-
mining behavior than can be achieved by conventional observational and hand test-
ing methods. Using different stimuli to signal the class of responses that will be
reinforced, it is possible determine the nature of the sensory discriminations that an
animal can make, and subsequently its performance on cognitive tasks, as well as the
effects of changes in reward value or magnitude.
11.4.2 OPERANT TASKS TO ASSESS STRIATAL FUNCTION IN RODENTS

Tasks that are administered in operant chambers usually require animals to perform
a number of intermediate training sessions before they are able to perform the task
in full. Since it is important to design and use tasks correctly to assess different
specified functions, the rationale for a variety of tasks that assess striatal function

is outlined below. Reference is made to those sources that contain more detailed
descriptions of how to accomplish individual behavioral tasks.
Behavioral testing in operant chambers allows both a high degree of experimen-
tal control and a detailed and thorough behavioral analysis. Consequently, operant
conditioning has proved to be a valuable way in which to assess cognition, motor
function, and motivation in the rat. Simple operant responses, such as pressing a lever
to gain food reward, are unaffected by striatal lesions or in transgenic animals.57,58 In
addition, due to the capacity to precisely control spatial and visual cues, more elabo-
rate schedules requiring conditional responses are also possible.59,60 However, more
complex operant tasks turn out to reveal subtle but specific and robust behavioral
impairments in many functional domains in animals with striatal damage. Indeed,
it is often not possible to even detect, let alone to precisely define, the nature of such
impairments with alternative observational or manual testing methods.
11.5 OPERANT ANALYSIS OF MOTOR RESPONDING:

STRIATAL LESIONED AND HD TRANSGENIC ANIMALS
The unilateral disruption of striatal function, either through targeting the striatal
projecting neurons or through depletion of striatal dopamine by central injection
of 6-OHDA, is known to cause a general sensorimotor impairment on the side con-
tralateral to the lesion. This impairment takes the form of an ipsilateral rotation,61
clumsy and inefficient use of the contralateral paw,62–64 and also a failure to respond
to stimuli that are presented to the side of the rat contralateral to the lesion,65 a phe-
nomenon that has been named “striatal neglect.” This syndrome of impairments has
many similarities to the “sensorimotor” syndrome produced by electrolytic lesions
of the lateral hypothalamus.66,67 Marshall and colleagues suggested in the 1970s
that the neglect that resulted from electrolytic lesions of the lateral hypothalamus
might in fact be the product of damage to the nigrostriatal projection.64 Thus, tasks
that used lateralized stimuli and responses that were first introduced to define the
precise nature of deficits resulting from hypothalamic lesions66,68 came to influence
subsequent behavioral paradigms that sought to analyze nigrostriatal and striatal
function. Assessing lateralized behavioral deficits in operant tasks can provide the
most sensitive within-subjects dependent variables in the context of unilateral lesion
models. However, transgenic animals show bilateral striatal pathology and symp-
toms where the overall level of performance compared to the controls provides the
main between-subjects dependent variables.
11.5.1 OPERANT ANALYSIS OF THE SENSORY AND MOTOR

ASPECTS OF “SENSORIMOTOR” STRIATAL NEGLECT
Carli and others sought to distinguish the sensory and motor aspects that constituted
the sensorimotor components of striatal neglect by designing a visual choice reac-
tion time task in the 9-HB apparatus that allowed the spatial location of the discrimi-
native stimuli and conditioned responses to be separated.69,70 In this task, only the
central three holes were exposed, the remaining holes being covered with a masking
plate (Figure 11.3). Two groups of rats were trained to make a sustained nose poke

Hold
Detect
Withdraw
(RT)
Move (MT)
Correct SAME Error
Error OPPOSITE Correct
FIGURE 11.3 Schematic illustration of the SAME and OPPOSITE versions of the Carli
choice reaction time task. In addition to measures of accuracy and response bias, the speed of
initiating (reaction time) and executing (movement time) of correct responses to the two sides
are also recorded. Source: Carli, M., Evenden, J. L., and Robbins, T. W. 1985. Depletion of
unilateral striatal dopamine impairs initiation of contralateral actions and not sensory atten-
tion. Nature 313:679–82.
into the middle of the three holes for a variable period until the presentation of a brief
light cue, which flashed unpredictably in one of the side holes, either to the left or
the right of the center hole. Rats had to make a rapid lateralized nose-poke response
in order to gain food reward, but the rule defining a correct response was different
for the two groups. The rats of one group were required to make a nose poke into the
same response hole at which the stimulus was presented (the “SAME” condition),
whereas the rats of the second group had to respond on the side that was not signaled
(the “OPPOSITE” condition).
For animals trained in the SAME condition, because of the crossover of connec-
tions between the brain and the periphery, we would predict that unilateral lesions—
whether of the dopamine system or the striatum itself—would produce deficits on
the contralateral side of the body. This is equally true whether the deficit is sensory,
motor, or associative in nature. However, the dissociation between the location of
the stimuli and the response holes in the OPPOSITE condition allows differential
predictions of the outcome depending on the nature of the underlying deficit. Thus, if
the animals have a sensory impairment in the detection of the stimuli, then we would
expect the rats with unilateral lesions to be impaired making an ipsilateral response
to a contralateral stimulus, whereas a response to an ipsilateral stimulus would be
unaffected. Conversely, if the deficit was primarily in the selection or initiation of
motor response, then we would expect the animal to be impaired making a contra-
lateral response to an ipsilateral stimulus, but be unimpaired in responding on the

ipsilateral side even though the discriminative stimuli are presented in contralateral
space.
Carli and colleagues70 reported that in both the SAME and the OPPOSITE condi-
tion, rats were impaired in effecting responses to the side contralateral to the lesion,
while neither group was impaired in producing ipsilateral responses (Figure 11.4).
This pattern of impairments was not consistent with either a sensory impairment
or a sensorimotor integration deficit; rather, the ipsilateral bias induced by unilat-
eral striatal dopamine depletion was interpreted as a bias in responding. The extent
of this ipsilateral bias has subsequently been shown to correlate with the extent of
dopamine depletion within the striatum.71 Furthermore, the pattern of results sug-
gested that the general motor deficits reported following striatal damage may be
caused by an increase in the latency to execute responses to the contralateral side.
However, a detailed analysis of this action showed that contralateral movement was
not uniformly impaired. The time taken to execute a response is considered to have
two components: “reaction time,” defined as the latency to initiate a response by
withdrawing the nose from the central hole (which therefore contains no lateral-
ized component), and the “movement time,” defined as the subsequent latency to
complete the lateralized nose poke response into the response hole. Animals with
unilateral striatal damage, including nigrostriatal lesions, are particularly impaired
in the reaction time.69–71 This indicates that the deficit is attributable to an impair-
ment in the planning, selection, or initiation of the lateralized response rather than in
its execution, suggesting an “executive” impairment, similar to that seen after frontal
lesions, rather than a pure motor deficit per se.
11.5.2 OPERANT TASKS TO DELIMIT THE SPECIFICITY OF STRIATAL NEGLECT

The study of Carli et al.70 demonstrated that unilateral striatal damage impaired
animals in responding toward their contralateral side, while at the same time show-
ing that, once initiated, movements made in contralateral space were not impaired.
This suggested that if the motor actions that comprised the response itself were not
themselves impaired, perhaps the striatum was more involved in some more abstract
elements of responding that influence motor output at an early stage of response prep-
aration. For example, it was possible that the coordinate frame in which responses
are organized were disrupted by striatal damage.72 We therefore sought to modify
the experimental design to define the basis of this “response space”; what is the
“contralateral” deficit seen in striatal neglect actually contralateral to? The nature
of this neglect has been quantified using both unilateral lesions of the nigrostriatal
dopamine neurons and excitotoxic lesions of intrinsic striatal neurons.
Animals were trained to perform two discriminations, independently, on alter-
nate days. As in the Carli et al.70 study, the task comprised a central hole and two
response holes. However, unlike the Carli et al.70 study, both response holes were
on the same side. So, on one day, animals were required to respond to the two holes
on the left, and on the next day to the two holes on the right (Figure 11.5).73 All
responses required rats to detect a stimulus light in one of the two response holes,
and to make a nose poke response in the same hole. Once trained, animals received
unilateral striatal lesions with central injections of quinolinic acid.

Ipsilateral Response Pre-op

100 Contralateral Response 100 Post-op
75 75
% Contralateral Bias
% Correct Responses
50 50
25 25
0 0
Same Opposite Same Opposite
Accuracy Contralateral Response Bias
(a) (b)
Ipsilateral Response
1.00
Contralateral Response
0.75
Reaction Time (sec)
0.50
0.25
0.00
Same Opposite
Reaction Time (RT)
(c)
FIGURE 11.4 Unilateral 6-hydroxydopamine lesions of the nigrostriatal pathway produces

marked deficits in initiating action (reaction time) specifically for contralateral responses in
the Carli task. (a) Postoperative accuracy of responding to the ipsilateral and contralateral
sides in the SAME and OPPOSITE task contingencies; (b) Degree of contralateral response
bias preop and post-op; (c) Postoperative reaction times to initiate ipsilateral and contralateral
responses in the SAME and OPPOSITE task contingencies. Source: Data from Carli, M.,
Evenden, J. L., and Robbins, T. W. 1985. Depletion of unilateral striatal dopamine impairs
initiation of contralateral actions and not sensory attention. Nature 313:679–82.

Hold
Detect
Withdraw
(RT)
Respond (MT) Error !!
FIGURE 11.5 Schematic illustration of the Brasted lateralized choice reaction time task
in the nine-hole box. Source: Brasted, P. J., Humby, T., Dunnett, S. B., and Robbins, T. W.
1997. Unilateral lesions of the dorsal striatum in rats disrupt responding in egocentric space.
J. Neurosci. 17: 8919–26.
When testing resumed a week later, the lesioned rats showed a severe impairment
responding on the contralateral side. This impairment took the form of a marked bias
toward the “near” hole, i.e., the hole closer to the center, when the holes were on the
side contralateral to the lesions. This “response bias” was so severe that lesioned ani-
mals were rarely able to produce responses to the “far” hole on the contralateral side.
In stark contrast to this impairment, lesioned rats were able to respond efficiently
and correctly when the holes were on the side ipsilateral to the lesion.73
The design of this operant task allowed a specific comparison to be made between
two specific hypotheses concerning the nature of a striatally mediated “response
space.” If responses were coded relative to an external reference within the animals’
environment (allocentric coding) then animals would be expected to always neglect
the relatively contralateral hole, regardless of which side of the space the holes were
presented. Allocentric coding is often seen in perceptual neglect, when patients with
cortical lesions neglect the contralateral side of an object, regardless of where the
object is located in space.74,75 In this task, an allocentric-based deficit would manifest
itself as a bias toward the far hole when the task is performed to the ipsilateral side,
and a bias toward the near hole when the task is performed to the contralateral side.
Alternatively, if responses were coded with respect to the subject’s body (egocentric
coding), then one would predict responding to be disrupted only on the contralateral
side.
The data clearly show that striatal neglect is not seen uniformly in all parts of
space, but is restricted to the contralateral side and thus consistent with the latter,
egocentric hypothesis. When responding to the ipsilateral holes, animals showed no
evidence of biasing their responding toward the far (i.e., relatively contralateral) hole.

Control rats
Unilateral striatal lesions
100
Near Hole Bias (%)
80
70
60
50
40
ipsi contra ipsi contra

Pre-lesion Post-lesion
FIGURE 11.6 Unilateral striatal lesions produce a marked postoperative ipsilateral response
bias, which is more marked for discriminations on the contralateral than on the ipsilateral
side. Source: Data from Brasted, P. J., Humby, T., Dunnett, S. B., and Robbins, T. W. 1997.
Unilateral lesions of the dorsal striatum in rats disrupt responding in egocentric space. J.
Neurosci. 17: 8919–26.
In contrast, animals were markedly impaired when performing on the contralateral

side and were completely unable to select responses to the far (i.e., relatively contra-
lateral) hole (Figure 11.6). A similar impairment was seen in studies that examined
unilateral striatal dopamine lesions using a between-subject design.72
11.6 OPERANT ANALYSIS OF COGNITIVE TASKS: STRIATAL

LESIONED AND GENETICALLY MODIFIED HD MODELS
Rosvold’s experimental work in the 1950s seriously challenged the concept of the
striatum as a structure involved purely in motor function. Based on the topographi-
cal nature of projections from the prefrontal cortex to the caudate nucelus, it was
shown that caudate lesions in monkeys produced impairments on the same tasks that
were known to be sensitive to frontal lesions.3 In particular, bilateral lesions of the
caudate nucelus were seen to impair accuracy on Jacobsen’s classic tasks of frontal
function designed to assess spatial working memory, such as spatial alternation and
delayed response.76,77
A corresponding functional organization was demonstrated in the rat only
when the prefrontal cortex came to be defined in terms of the projection areas of
the mediodorsal nucleus of the thalamus—namely the medial and orbital walls of
the frontal cortex rather than the pole of the frontal lobe, as had previously been
assumed.78,79 Lesions in the rostral medial striatal areas that were innervated by the
medial wall prefrontal cortex again produced deficits in delayed alternation and spa-
tial navigation tasks.80–82 It is these advances in functional neuroanatomy, as well as
the neuropsychological deficits reported in basal ganglia disorders such as HD and

PD, that provide the theoretical foundation for operant tasks designed specifically to
assess the impact of striatal lesions on cognitive function.
11.6.1 DELAYED MATCHING TASKS

The automation of delayed matching procedures represented one such attempt to
assess the impact of striatal lesions on working memory. This requires an animal to
be presented initially with a sample stimulus, and then after a variable delay, to be
presented with a choice between two stimuli, one being the earlier sample and the
other novel. Memory of the earlier stimulus presentation is then tested by requiring
the animal to choose either the initial sample (delayed-matching-to-sample) or the
novel stimulus (delayed-non-matching-to-sample). A correct response results in the
delivery of food reward, whereas an incorrect response is signalled by a timeout or
some other error signal. The introduction of a variable delay between sample pre-
sentation and choice response allows the rate of forgetting to be determined from
the plot of decline in choice as the length of the delay interval increases. DMTS
was first developed for assessing memory in primates and required monkeys to dis-
criminate between objects primarily on the basis of the visual properties (e.g., shape,
color) of objects.83 Procedures for rats, however, can require animals to discriminate
between objects in a maze84 or between different odors. Alternatively, choice dis-
criminations can be made in the spatial modality, and this has led us to develop a
delayed-matching-to-position (DMTP) task for use in rats.85,86 Animals are trained
in standard two-retractable-lever Skinner boxes (Figure 11.1). On each trial, one of
the two retractable levers is inserted into the chamber as the sample, which the rat
must press to register. The lever is then retracted and the rat must press the central
panel until, after a random variable delay period, the next panel press causes both
levers to be extended back into the chamber. A correct matching response on the
same lever as that produced in the sample stage is rewarded with the delivery of a
food pellet, whereas an incorrect response to the other lever results in a timeout (all
lights are turned off for a short period to signal an error). After a further short inter-
val, the next trial commences.
Delayed-matching-to-sample is sensitive to prefrontal cortex (PFC) damage in
primates.87–89 Consequently, we have developed DMTP to investigate the compara-
tive contributions of the medial PFC and medial striatum in a working memory task
in rats. Cortical lesions that are restricted to the anterior medial PFC result in a pro-
gressive deficit such that accuracy is increasingly impaired at progressively longer
delays.90 This delay-dependent pattern of impairment suggests a rather specific dis-
turbance in short-term memory. However, larger lesions that extend into the anterior
cingulate cortex produce a broader impairment at all delays, reflecting a more gen-
eralized deficit in the animal’s ability to perform the matching rule (Figure 11.7).90
Similarly, when we looked at impairments following striatal lesions, lesions in the
ventral striatum (which receives restricted prefrontal inputs) exhibited a clear delay-
dependent deficit, whereas lesions in the dorsal striatum (which receives cingulate as
well as frontal inputs) disrupted animals’ performance at all delays.90–92
These studies illustrate, first, the way in which an operant task may be designed
to tap discrete aspects of cognitive function—in this case short-term working mem-

Controls
100 Prefrontal cortex lesion
Neostriatal lesion
Ventral striatal lesion
90
% Correct
80
70
60
50
0 2 4 8 12 18 24
Delay Interval (sec)
FIGURE 11.7 Prefrontal cortex and dorsal and ventral striatal lesions produce marked defi-
cits on the DMTP task. Note that while prefrontal and ventral striatal lesions produce a delay-
dependent deficit, the neostriatal lesions induce a more global deficit at immediate as well
as long delays. Source: Data from Dunnett, S. B. 1990. Role of prefrontal cortex and striatal
output systems in short-term memory deficits associated with ageing, basal forebrain lesions,
and cholinergic-rich grafts. Can. J. Psychol. 44:210–32.
ory—and second, that selective lesions within the fronto-striatal circuitry yield dis-
tinctive patterns of cognitive deficit associated with the particular cortical loop(s)
disturbed.
11.6.2 DELAYED ALTERNATION TASKS

The spatial delayed alternation task (DA) is another task that involves short-term
working memory and requires animals to alternate their responding between two
spatially distinct locations. Like the DMTP task, it is very sensitive to damage of the
medial PFC. Indeed, deficits in delayed alternation in Wisconsin boxes (for monkeys)
and in T-mazes (for rats) was one of the defining features of the prefrontal deficit
described by Jacobsen in his classic primate studies in the 1930s93,94 and replicated
since.30 Rats similarly exhibit clear deficits in delayed alternation after prefrontal
lesions when tested in a T-maze.92,95,96
There have been several attempts to adapt this classic task to the operant box,97–
100 although with varying degrees of success. In our adaptation,101 rats are trained to
press the centrally located food panel until the end of a variable delay period (5–20
sec). A panel press subsequent to the end of the delay period results in the extension
of both the left and the right levers. On the first trial of the day, pressing either lever
produces a food pellet reward. On all subsequent trials, the rat is required to press
the lever that was not pressed on the previous trial (Figure 11.8). A correct press
is rewarded with a food pellet, whereas an incorrect press (repetition of the same

$)')' " $)'+"(

$%"+'(%'" )(
'())' "&'($)%)"+'(
')'(&%$(")%'','
%'')'(&%$( " +'%%&"")

&$"" )%$ , )*$) "$%(&%!
$)')' " $)'+"(

$%"+'(%'" )(
$"" )%$ &)$%(&%!(

(*"%+'/(
&(
&$"" )%.
"+'(-)$
&&%( )'(&%$( #'(&%$(

&$"" )%$" +'',' """ )(%.) #%*) &$(%$%'')
%'''%''(&%$(
-) $)')' " $)'+"

$%"+'(%'" )(
FIGURE 11.8 Schematic illustration of the delayed operant spatial alternation test in a two-
retractable-lever Skinner box.
lever as on the previous trial) has no consequence. In either case, after pressing one
lever, both levers are withdrawn and the timer for the next variable delay interval is
started. The distinctive feature of our variant of operant-delayed alternation, as in
the DMTP/DNMTP task described above, is that the animal is required to nose poke
at the central panel during the delay in order to trigger presentation of the two choice
levers. This serves to keep the rat centralized between the two response locations
and reduces the opportunity for it to adopt a simple mediating response strategy
during the delay (i.e., simply waiting at the location where the correct lever will next
appear). We vary the delay interval on each trial and thereby accumulate information
about the animal’s level of accuracy over different lengths of time that the last trial
response must be held in memory.
In recent studies we have begun to investigate differential roles of the PFC and
the medial aspect of the striatum in this task. We find that lesions both of the medial
prefrontal cortex (mPFC) and of the medial striatum disrupt performance on this
task (Figure 11.9A). However, detailed analysis of the impairments suggest slightly
different reasons for the deficit after each lesion.101 In particular, the mPFC animals
exhibit a relatively straightforward impairment in task accuracy that may be related
to an executive deficit in determining the correct response based on short-term

Control
A. Previous Trial Correct Prefrontal lesions
90 Striatal lesions
Chance (50%)
80
Correct (%)
70
60
50
40
B. Previous Trial Incorrect
90
80
Correct (%)
70
60
50
40
Pre-lesion Post-lesion Probe Trials
FIGURE 11.9 Prefrontal and striatal lesions both produce a marked decline in response
accuracy in the operant delayed alternation test. The upper and lower panels illustrate choice
accuracy depending on whether the previous trial was correct or incorrect. While the pre-
frontal lesions induced a similar deficit on all trials irrespective of performance on previous
trials, the striatal lesions induced a deficit whereby a deficit on the previous trial increased the
chance of a deficit on the present trial. This is directly against an interference effect between
trials and suggests a perseverative tendency of the rats with striatal lesions. Source: Data
from Dunnett, S. B., Nathwani, F., and Brasted, P. J. 1999. Medial prefrontal and neostriatal
lesions disrupt performance in an operant delayed alternation task in rats. Behav. Brain Res.
106:13–28.
memory for the last response. By contrast, the rats with the striatal lesions exhibit a
tendency to preseverate their responses. Thus, for example, if we analyze the errors
from trial to trial, whereas the conditional probability of an error on a particular
trial is, after prefrontal lesions, independent of how the animal performed on the
previous trial, the chance of an error (involving repetition rather than alternation of
the previous response) after a striatal lesion increases as the animal makes a run of
errors in a row (Figure 11.9B). It is worth noting that this pattern of errors is quite
different from that which would be expected if the animals’ deficits were caused by a
memory failure, for example by an increased susceptibility to proactive interference.
Furthermore, analyses of other behavioral measures indicate that the time taken to
collect food reward is unaffected, suggesting that neither a general motor deficit nor
a motivational deficit are the basis for these impairments.

Midline Unilateral Ipsi/contra

knife-cut striatal prefrontal
90 lesions lesions lesions
80
Accuracy (%)
70
60
Controls Lpsilat double lesions

50
Single striatal lesions Crossed double lesions
4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80
Test Days (in 2-day blocks)
FIGURE 11.10 Delayed alternation performance accuracy over the period of initial train-
ing and recovery after midline knife cut lesions in all rats, after additional unilateral stria-
tal lesions, and after further lesions of the ipsilateral or contralateral prefrontal cortex. All
animals rapidly recovered after each of the first two surgeries. The effects of double lesions
depended on the side of the second, prefrontal cortical lesions: those rats receiving lesions
in the hemisphere contralateral to the earlier striatal lesions (crossed lesions) exhibited last-
ing impairments, whereas lesions in the same hemisphere (ipsilateral lesions) produced no
impairment. Source: Data from White, A., and Dunnett, S. B. 2006. Fronto-striatal discon-
nection disrupts operant delayed alternation performance in the rat. Neuroreport 17:435–41.
Thus, the range of measures of different aspects of task performance allow not
only a dissociation between different lesions—even though they all disrupt task
performance—but also the beginnings of an analysis of the precise nature of the
functional impairment that results following disturbance of each neuroanatomical
component in a connected fronto-striatal circuit.
The role of the fronto-striatal/cortico-striatal system in mediating the perfor-
mance in delayed alternation was recently further tested by using a crossover lesion
paradigm.102 A midline transection of the corpus callosum was made to separate the
hemispheres, followed by crossed lesions of the striatum on one side and the PFC
on the other side. This crossover lesion produced a significant and long-term deficit
in delayed alternation. Detailed analysis showed that the double-lesion affected the
working-memory aspects of cortico-striatal function in a delay-dependent fashion,
without effecting response bias. Interestingly, in a control group in which similar
striatal and cortical lesions were made but on the same side, the intervention had
little detectable effect. The crossover lesion disrupted the cortico-striatal system
completely, while the same side double-lesion retained the system intact on one side.
This disconnection paradigm elegantly demonstrated that efficient performance on

A Trial Numbers B Accuracy

100
250 Wild-type mice
Q92 Transgenics 80
Percentage Correct (%)

200
Trials Initiated (n)
60
150
40
100
50 20
0 0
FIGURE 11.11 Delayed alternation testing of the Q92 mouse model. When tested on the
delayed alternation task, with no delays present, the knock-in Q92 (Hdh Q92/Q92) mice initiated
fewer trials compared to their wild-type littermates (A) and exhibited a severe deficit in per-
formance accuracy (B). Source: Data produced by Rebecca Trueman.
the classical prefrontal delayed alternation task is dependent upon an intact cortico-
striatal system (Figure 11.10).
The delayed alternation paradigm has also been used to investigate the Q92
mouse model of HD, in this instance using mouse nine-hole operant chambers. To
adapt the task for these boxes the mouse is required to respond to the illuminated
center hole (hole 5), until this hole is extinguished and two lateral lights are illumi-
nated (hole 3 and 7; the remaining 6 holes are blocked). The center hole therefore
acts like the panel in the Skinner boxes to try to reduce any mediating response strat-
egies that the mouse may make, and the lateral lights act as levers with the mouse
being required to alternate responding between the two locations.
We have found a severe disruption in responding, evident from both accuracy
and total trials measures when testing naive 12-mo-old Q92 mice on this task. This
was before any delay was introduced into the task. However, unlike striatal lesioned
rats, there was no increase in perseveration to the response holes (when adjusted for
the number of correct trials) and panel latencies were significantly longer, indicating
possible motor and/or motivational deficits and that the dysfunction in these mice
may not be wholly attributed to striatal dysfunction.
When we examined a second group of Q92 mice trained at 5 mo of age and then
retested at 12 mo, the pretraining had improved the performance of these mice in
such a way that it was possible to introduce delays into the task. With this group of
Q92 mice, increasing delays had a deleterious effect on accuracy of both the wild-
type and knock-in groups; however, overall the knock-in mice were less accurate
than their wild-type littermates. As in the previous group of Q92 mice there was no
evidence for increased perseveration and panel latencies were longer (Figure 11.11).

11.6.3 5-CHOICE REACTION TIME TASK

The 5-Choice Reaction Time Task (5-CRTT), initially developed by Robbins,103 is
a task that measures the sustained visual attention of an animal, and the animal’s
subsequent ability to respond to the identified stimuli. These integrated attentional/
motor processes rely on intact cortical (especially anterior regions) and striatal brain
areas, with damage to either producing profound behavioral deficits on the task.
Consequently, this task is an ideal probe of fronto-striatal dysfunction in animal
models that could have discrete frontal cortex or striatal neuropathology, as would be
expected in HD, for example. The integrity of the underlying attentional/motor pro-
cesses are assessed through the ability of the animal to attend and respond to visual
stimuli randomly presented across five of the nine holes of the 9-HB. In the 5-CRTT,
five holes (1, 3, 5, 7, and 9) are available, while holes 2, 4, 6, and 8 are masked. A
stimulus light is presented randomly in one of the five holes, and a correct response
to the light extinguishes it and produces a reward delivery signaled by the illumi-
nation of the food magazine. On collection of the reward and subsequent removal
of the head from the magazine, a short inter-trial interval (ITI) (2 sec) is followed
by the next trial. By presenting the stimulus light randomly across the array, the
animal must attend to the whole of the array to make a correct nose-poke response.
Response accuracy, presented as the percentage of correct trials, and reaction time,
the time between the light stimulus being presented and the correct response, are the
two main dependent variables on which behavior is assessed. But several error types
are also measured to provide information as to why one group of animals performs
better/worse than the other. Of particular relevance to HD would be perseverative
errors, akin to compulsive responding, where the individual is unable to refrain from
making a response, despite knowing that that particular response is inappropriate,
which on this task is measured as a nose poke in the correct location, but after the
stimulus light has been extinguished.
As with most operant tasks, the task parameters can be manipulated to increase/
decrease task difficulty. The following example using the Q92 HD mouse demon-
strates the flexibility of the apparatus and task in teasing out subtle performance
deficits in this knock-in mouse line.
The Q92 mouse has a normal life span and to the trained observer looks normal.
When introduced to the 9-HB with the parameters set to their easiest settings on
the 5-CRTT (2-sec stimulus light duration), no differences in accuracy are found
between Q92 mice and the wild-type controls, but a small difference in response
latencies were demonstrated in the Q92 mice. By decreasing the duration of the stim-
ulus light from 2 sec to 0.5 sec, differences between the groups on both reaction time
and accuracy measures become more apparent because of the increased difficulty of
the task. Running a further experimental session where the duration of the stimulus
light varies between 0.5 and 2 sec (presented in a pseudo-random fashion), response
latencies between the two groups diverged with the Q92 mice becoming slower at
the task, while the wild-type animals reduced their response latencies in response
to the new conditions. The variable stimulus light durations also uncovered differ-
ences in the overall levels of accuracy between the groups, with the wild-type ani-
mals responding with greater accuracy for each of the stimulus light durations. The

introduction of the variable stimulus light duration increased the overall difficulty of
the task and provided a more sensitive measure of behavioral deficit in the Q92 mice
than the single-light duration condition, and clearly demonstrates the breakdown in
the integration of attentional and motor aspects of behavior in the response pattern.
This effect also demonstrates the sensitivity of the task and the utility of the task for
use with mice, which has important implications for the use of transgenic and knock-
in/knockout animals.
11.6.4 SERIAL IMPLICIT LEARNING TASK

The impetus for designing the Serial Implicit Learning Task (SILT) was the wish
to identify a selective HD deficit that could then be probed for in animal models of
HD, with the goal being the reversal of the deficit with a novel therapeutic approach.
Implicit learning is the ability to learn information without explicit knowledge of
the learning experience. The classic example of implicit learning is the ability of a
young child to acquire the rules of grammar and syntax of language, which are gen-
erally not taught formally until the child is around 5 yr of age, by which time some
language has developed. People with HD have been found to be deficient in learning
tasks with an implicit element,104 and this learning deficit has been demonstrated to
be independent of any motor abnormalities the person may have. Typically, implicit
learning tasks are procedural learning tasks, whereby strings of stimuli are repeated
in blocks and reaction times to respond to the stimuli are recorded, with the infer-
ence being that familiar sequences of stimuli would elicit shorter response latencies
than randomly presented stimuli. This is, in fact, the case. Repeated sequences of
stimuli do produce shorter response latencies than randomly presented stimuli, and
importantly, the subjects must report no awareness of the sequential nature of the
predictable stimuli, making the sequences implicitly learned. The SILT uses the 9-
HB for presenting random two-phase sequences of light stimuli to the animals, with
an embedded, predictable, two-step sequence, which is our implicit learning probe.
Within the 9-HB, five of the nine stimulus lights are available, designated A, B,
C, D, and E (corresponding to holes 1, 3, 5, 7, and 9 in the 9-HB, if counting from the
left) with the remaining four holes masked. The SILT is a two-phase task, whereby
the animal has to respond correctly to both the S1 and S2 stimuli to receive a reward.
The S1 stimulus is presented randomly in any of the five available holes, and a cor-
rect response to the S1 stimulus results in this light being extinguished and the ran-
domly chosen S2 stimulus light being illuminated. A correct nose poke to this light
results in the reward presentation in the food magazine, which is signaled by the
magazine light being lit. Removal of the head from the magazine initiates the ITI (2
sec) prior to the onset of the next trial and the illumination of the next S1. In order to
probe for implicit learning, a predictable two-phase sequence is embedded into the
random stimuli presentation. The predictable sequence that is generally used in our
experiments is hole B (S1) to hole D (S2). This particular sequence has the advantage
of having a direct nonpredictable comparison of hole D (S1) to hole B (S2), so that
performance of the predictable option can be directly compared with performance
levels of the comparable nonpredictable option. As with other 9-HB tasks, task dif-

S1 Accuracy S1 Reaction Time

100 2.5
Control mice Intact Control mice Intact
90 Striatal lesions Lesion Striatal lesions Lesion
* 2.0
* *
*
Latency (secs)
80
% Correct
* 1.5
70
1.0
60
50 0.5
40 0.0
A B C D E A B D C E
Holes Holes
(a) (b)
Predicted S2 Reaction Time Predicted
S2 Accuracy
100 * Unpredicted 2.0 Unpredicted
* *
* *
90 *
* 1.5 *
Latency (secs)
80
% Correct
70 1.0
60
0.5 Control mice
50 Striatal lesions
Control mice
Striatal lesions
40 0.0
1 2 3 4 Intact Lesion 1 2 3 4 Intact Lesion
Steps Steps
(c) (d)
FIGURE 11.12 Performance on the serial choice discrimination task by sham-operated

(control) and striatal lesioned mice, collapsed over the 20 days of postoperative testing. (a
and b) Response accuracy and reaction times, respectively, to the first stimulus are plot-
ted separately for trials in which S1 was presented in the different response locations A-E,
with average performance on trials shown as vertical bars for the intact and lesion groups.
Significant differences in accuracy between control and lesioned animals at each hole were
demonstrated. (c and d) Response accuracy and reaction times, respectively, to the second
are plotted separately for the number of steps 1–4, between S1 and S2. Control mice demon-
strated significantly greater accuracy and significantly quicker reaction times than lesioned
mice for each of the step distances (d). The two-step trials are then subdivided into predict-
able and unpredictable trials, shown as vertical bars (within group analysis). Source: Adapted
from Trueman, R. C., Brooks, S. P., and Dunnett, S. B. 2005. Implicit learning in a serial
choice visual discrimination task in the operant 9-hole box by intact and striatal lesioned
mice. Behav. Brain Res. 159:313–22.
ficulty can be manipulated through the shortening of the stimulus duration of the S2
stimuli.
Two measures of predictability are recorded: accuracy for, and reaction time to,
the predictable sequence. In addition, the SILT measures several other performance
parameters. Responding to S1 stimuli is assessed by the accuracy and reaction time

measures for each individual hole, as this S1 responding is essentially the 5-CRTT.
The typical response pattern for S1 accuracy and reaction times is illustrated by the
control animals in Figure 11.12A and B, respectively, which show greater accuracy
and reduced response latencies for the center holes. Responding to the S2 stimuli
can be broken down into accuracy of response and reaction times for the two-phase
movements from S1 to S2, which are of varying lengths. For example, if S1 = hole
A and S2 = hole E, this is a four-step movement (B > C > D > E) to make a correct
response, whereas if S1 = hole D and S2 = hole C, this is a one-step movement.
Typically, animals are more accurate and quicker to respond to shorter movement
steps than longer (Figure 11.12C and D, respectively). Finally, in our experiments to
date, animals with striatal lesions and genetically modified animals have been able
to use the predictable information embedded in the sequences, resulting in greater
accuracy and shorter response latencies for the predictable trials (see bars in Fig-
ure 11.12C and D).105
11.7 OPERANT ANALYSIS OF STRIATAL LESIONS:

DEFICITS IN MOTIVATIONAL STATE
One of the hallmarks of HD is that patients present with psychiatric as well as motor
and cognitive disturbances. A true psychiatric impairment is, of course, difficult to
evaluate in primates, let alone in the laboratory rat. However, as an alternative to
trying to determine the animal’s emotional state, it is certainly possible to evaluate
motivational state, in particular the effects of striatal manipulations on changes in
responding to stimuli with motivational significance. For example, the hyperactivity
that is apparent following lesions of the medial striatum is not independent of the ani-
mals’ motivational state, but is particularly associated with food deprivation.106–108
Traditionally, it is the ventral striatum that is believed to play a central role in
reward-related processes.109,110 More recently, research interest has turned to whether
the ventral striatum not only mediates the evaluation of reward, but also the “effort”
expended in obtaining it.111,112 One way in which to assess this “cost–benefit” analy-
sis is to employ a progressive ratio (PR) schedule of operant lever pressing.113 Moti-
vational deficits are a well-established symptom in HD and often present at an early
stage in the disease, and as such these deficits are more associated with damage in
the dorsal striatum. It is primarily this clinically based rationale that has led us to
investigate reward-related processes in the dorsal striatum using a PR schedule.114
PR schedules require the rat to simply press a single lever a number of times to
receive a food pellet reward. At first, each lever press results in a reward, then after
five rewards three or five presses are required for each reward, then 10 presses per
reward, and so on. As the session progresses, a progressively greater amount of work
(e.g., number of lever presses) is required in order to gain an unchanging amount
of reward (e.g., a single food pellet each time). Such a schedule makes it possible to
examine a number of motivational measures that reflect the willingness of a rat to
work for reward. These measures typically include the breakpoint of each animal
(e.g., the greatest number of lever presses that an animal is prepared to make in order
to obtain a single food pellet). A reduction in breakpoint is regarded as a general
indication of lower levels of motivation.113,115 A second useful measure is the “post-

25
Wild-type mice
Q92 transgenics
20
Last Ratio Completed

15
10
FIGURE 11.13 Progressive ratio testing of the Q92 mouse model. The last ratio completed
by the knock-in mice (HdhQ92/Q92) was significantly lower than that attained by the wild-type
mice (Hdh+/+), demonstrating a motivational deficit in the Q92 knock-in mouse line. Source:
Data produced by Rebecca Trueman.
reinforcement pause,” which describes the latency to resume lever pressing follow-
ing the presentation of food. This is believed to reflect the reluctance to resume lever
pressing as the “cost” of reward increases.
In contrast to lesions of the ventral striatum, we have found that focal lesions of
the dorsal striatum do not induce deficits as revealed by these primary motivational
measures.114 Neither lesions of the dorsomedial striatum nor lesions of the dorsolat-
eral striatum produced any changes in either breakpoint or the post-reinforcement
pause. Nevertheless, these restricted striatal lesions did induce some specific perfor-
mance deficits. Animals in both lesion groups took significantly longer to collect food
rewards once a food pellet was delivered at the completion of a schedule. Moreover,
rats with dorsolateral striatal lesions also continued to press the lever once a sched-
ule had been completed and a food reward was available for consumption.114 Thus,
although striatal lesions did not cause any overall deficit in motivation, rats with
striatal lesions again demonstrated a perseverative deficit. This suggested that while
striatal lesions did not affect the ability of rats to regulate their rates of responding
with changing reward cost, there was evidence that striatal lesions compromised the
ability of rats to sequence and switch responding efficiently and appropriately.
Recently we have also tested the knock-in Q92 mice on the PR task, using
the mouse nine-hole operant chambers. In this version of the task the mouse was
required to respond to the illuminated center hole with one nose poke for three trials;
for the proceeding three trials three nose pokes were necessary in order to receive
the reward; and for the next three trials, six nose pokes were required, and so on.
Unlike the striatal lesioned rats, the Q92 mice had a severe decrease in breakpoint
attained compared to wild-type littermates and an increase in post-reinforcement
pause. However, no increase in perseveration was evident, showing that these mice
have a motivational deficit and suggesting that the dysfunctions seen in this HD
model are not classical of striatal lesions (Figure 11.13).

11.8 CONCLUSION
By reviewing a handful of recent studies examining striatal function in rats and
mice, this chapter aims to demonstrate the value of operant tasks in detecting spe-
cific deficits in the fronto-striatal system. However, new tasks are constantly being
developed with the objective of shedding light on striatal function in both intact and
pathological states, particularly trying to view the impact of the rodent studies in the
context of ongoing primate and clinical research.
There are many advantages to using operant tasks to assess function. The impres-
sive degree of experimental control afforded by such paradigms provides not only an
extensive stimulus control over an animal’s behavior, but also permits the quantifica-
tion of the varied response options that an animal can choose to make. In addition,
the automation of such paradigms overcomes the inherent biases in the observational
recording of behavior, and also allows for greater efficiency in collecting and pro-
cessing data. In addition to allowing detailed functional studies, operant tests allow a
functional assessment of a wide variety of those surgical, cellular, or pharmacologi-
cal interventions with potential clinical relevance. While hand testing and observa-
tional techniques may be more appropriate if the evaluation of a novel therapy is at
an early stage, operant tasks allow for a fuller evaluation of the functional efficacy of
treatments. Moreover, an understanding of how particular neural structures mediate
function is crucial to the design of such interventions.116
The use of genetically modified animals has posed a challenge on several
grounds. The development of transgenic, knock-in and knockout animal lines to
recapitulate diseases has produced numerous mouse lines (less so in rats) for most
major diseases. In the case of HD, the mouse lines, of which there are several, exhibit
neuropathology that varies between lines in severity, anatomical location, and pace
of development, resulting in a wide range of behavioral deficits that also vary in
severity and age of onset. As a general rule, transgenic animals demonstrate a greater
degree of behavioral and neuronal pathology than the knock-in models of HD. The
severity and age of onset of the disease in any given mouse line has important impli-
cations for operant testing generally, and should be taken into account prior to the
onset of experiments. Some mouse lines (for example, the R6/2) have an especially
aggressive phenotype development that would make operant testing and subsequent
therapeutic intervention extremely difficult because of the operant training times
required.
A further complication is the background strain of the mice that carry the
genetic modification. Typically, transgenic animals are bred from strains where the
mouse strains chosen to create the new mouse line are chosen for reasons other than
behavioral performance. Consequently, in several genetically modified mouse lines
it would be very difficult to produce good behavioral readouts, as they have some
characteristic that confounds testing, an example being the HD YAC mouse lines117
that were created on an FVB background strain and consequently have severe retinal
degeneration. Other mouse lines display normal behavioral profiles generally, but
particular features of their behavior are at the extreme limit of what would be consid-
ered normal. In terms of spontaneous locomotor activity, DBA mice are very active
(possibly hyperactive), whereas the 129/S2 mouse exhibits relatively little sponta-

neous motor activity. Both of these mouse lines are used as background strains.
However, several mouse lines are very suitable for behavioral testing, and our recent
experiences with Q92 mice has clearly demonstrated the utility of using genetically
modified animals in an operant setting.
Genetically modified animals in HD, and lesion models to a lesser degree, have
the potential to recapitulate the disease process with biological authenticity, permit-
ting the testing and validation of novel therapeutic interventions. Operant testing of
fronto-striatal function is an essential part of analyzing the behavioral sequelae of
basal ganglia pathology and must be continually addressed.
REFERENCES
1. Wilson, S. A. K. 1914. An experimental research into the anatomy and physiology of
the corpus striatum. Brain 492.
2. Jung, R., and Hassler, R. 1960. The handbook of physiology. I. Neurophysiology, Vol.
2. 863. Washington, DC: The American Physiological Society.
3. Divac, I., Rosvold, H. E., and Szwarcbart, M. K. 1967. Behavioral effects of selective
ablation of the caudate nucleus. J. Comp. Physiol. Psychol. 63:184–90.
4. Alexander, G. E., Crutcher, M. D., and DeLong, M. R. 1990. Basal ganglia-thalamo-
cortical circuits: Parallel substrates for motor, oculomotor, “prefrontal” and “limbic”
functions. Prog. Brain Res. 85:119–46.
5. Huntington, G. 1872. On chorea. Adv. Neurol. 1.
6. Huntington’s Disease Collaborative Research Group. 1993. A novel gene contain-
ing a trinucleotide repeat that is expended and unstable on Huntington’s disease. Cell
72:971.
7. Paulsen, J. S., et al. 2006. Brain structure in preclinical Huntington’s disease. Biol.
8. Vonsattel, J. P., et al. 1985. Neuropathological classification of Huntington’s disease. J.
Neuropathol. Exp. Neurol. 44:559–77.
9. Myers, R. H., et al. 1991. Decreased neuronal and increased oligodendroglial densities
in Huntington’s disease caudate nucleus. J. Neuropathol. Exp. Neurol. 50:729–42.
10. Hedreen, J. C., and Folstein, S. E. 1995. Early loss of neostriatal striosome neurons in
Huntington’s disease. J. Neuropathol. Exp. Neurol. 54:105–20.
11. Hackam, A. S., Singaraja, R., Zhang, T., Gan, L., and Hayden, M. R. 1999. In vitro
evidence for both the nucleus and cytoplasm as subcellular sites of pathogenesis in
Huntington‘s disease. Hum. Mol. Genet. 8:25–33.
12. Reddy, P. H., et al. 1999. Transgenic mice expressing mutated full-length HD cDNA:
A paradigm for locomotor changes and selective neuronal loss in Huntington’s disease.
Philos. Trans. R. Soc. Lond B Biol. Sci. 354:1035–45.
13. Ordway, J. M., et al. 1997. Ectopically expressed CAG repeats cause intranuclear
inclusions and a progressive late onset neurological phenotype in the mouse. Cell
91:753–63.
14. Davies, S. W., et al. 1997. Formation of neuronal intranuclear inclusions underlies the
neurological dysfunction in mice transgenic for the HD mutation. Cell 90:537–48.
15. DiFiglia, M. et al. 1997. Aggregation of huntingtin in neuronal intranuclear inclusions
and dystrophic neurites in brain. Science 277:1990–93.
16. Van Raamsdonk, J. M., Pearson, J., Murphy, Z., Hayden, M. R., and Leavitt, B. R.
2006. Wild-type huntingtin ameliorates striatal neuronal atrophy but does not prevent
other abnormalities in the YAC128 mouse model of Huntington disease. BMC. Neuro-
sci. 7:80.

17. Li, S., and Li, X. J. 2006. Multiple pathways contribute to the pathogenesis of Hunting-
ton disease. Mol. Neurodegener. 1:19.
18. Paulsen, J. S., et al. 2001. Clinical markers of early disease in persons near onset of
Huntington’s disease. Neurology 57: 658–62.
19. Paulsen, J. S., et al. 2006. Preparing for preventive clinical trials: The Predict-HD
study. Arch. Neurol. 63:883–90.
20. Girotti, F., Marano, R., Soliveri, P., Geminiani, G., and Scigliano, G. 1988. Rela-
tionship between motor and cognitive disorders in Huntington’s disease. J. Neurol.
235:454–57.
21. Jahanshahi, M., Brown, R. G., and Marsden, C. D. 1993. A comparative study of simple
and choice reaction time in Parkinson’s, Huntington’s and cerebellar disease. J. Neurol.
Neurosurg. Psychiatry 56:1169–77.
22. Bradshaw, J. L., et al. 1992. Initiation and execution of movement sequences in those
suffering from and at-risk of developing Huntington’s disease. J. Clin. Exp. Neuropsy-
chol. 14:179–92.
23. Thompson, P. D., et al. 1988. The coexistence of bradykinesia and chorea in Hunting-
ton’s disease and its implications for theories of basal ganglia control of movement.
Brain 111(Pt 2):223–44.
24. Penney, J. B. Jr., et al. 1990. Huntington’s disease in Venezuela: 7 years of follow-up on
symptomatic and asymptomatic individuals. Mov. Disord. 5:93–9.
25. Divac, I. 1968. Effects of prefrontal and caudate lesions on delayed response in cats.
Acta. Biol. Exp. (Warsz. ) 28:149–67.
26. Divac, I. 1972. Neostriatum and functions of prefrontal cortex. Acta Neurobiol. Exp.
(Warsz. ) 32:461–77.
27. Stout, J. C., and Johnson, S. A. 2005. Cognitive impairment and dementia in basal
ganglia disorders. Curr. Neurol. Neurosci. Rep. 5:355–63.
28. Hoth, K. F., et al. 2007. Patients with Huntington’s disease have impaired awareness of
cognitive, emotional, and functional abilities. J. Clin. Exp. Neuropsychol. 29:365–76.
29. Knopman, D., and Nissen, M. J. 1991. Procedural learning is impaired in Huntington‘s
disease: Evidence from the serial reaction time task. Neuropsychologia 29:245–54.
30. Butters, N., Sax, D., Montgomery, K., and Tarlow, S. 1978. Comparison of the neuro-
psychological deficits associated with early and advanced Huntington’s disease. Arch.
Neurol. 35:585–89.
31. Lange, K. W., Sahakian, B. J., Quinn, N. P., Marsden, C. D., and Robbins, T. W. 1995.
Comparison of executive and visuospatial memory function in Huntington’s disease
and dementia of Alzheimer type matched for degree of dementia. J. Neurol. Neurosurg.
32. Lawrence, A. D., et al. 1996. Executive and mnemonic functions in early Huntington’s
disease. Brain 119(Pt 5):1633–45.
33. Harper, P. S. 1996. Huntington’s Disease. London: W. B. Saunders.
34. Wang, L. H., and Qin, Z. H. 2006. Animal models of Huntington‘s disease: Implica-
tions in uncovering pathogenic mechanisms and developing therapies. Acta Pharma-
col. Sin. 27:1287–1302.
35. Laursen, A. M. Corpus Striatum. 1963. Acta Physiol. Scand. Suppl. (suppl)211–106.
36. McGeer, E. G., and McGeer, P. L. 1976. Duplication of biochemical changes of
Huntington’s chorea by intrastriatal injections of glutamic and kainic acids. Nature
263:517–19.
37. Schwarcz, R., and Coyle, J. T. 1977. Striatal lesions with kainic acid: Neurochemical
characteristics. Brain Res. 127:235–49.
38. Coyle, J. T., and Schwarcz, R. 1976. Lesion of striatal neurones with kainic acid pro-
vides a model for Huntington‘s chorea. Nature 263:244–46.

39. Beal, M. F., Ferrante, R. J., Swartz, K. J., and Kowall, N. W. 1991. Chronic quinolinic
acid lesions in rats closely resemble Huntington’s disease. J. Neurosci. 11:1649–59.
40. Schwarcz, R., et al. 1979. Ibotenic acid-induced neuronal degeneration: A morphologi-
cal and neurochemical study. Exp. Brain Res. 37:199–216.
41. Schwarcz, R., Whetsell, W. O. Jr., and Mangano, R. M. 1983. Quinolinic acid: An endog-
enous metabolite that produces axon-sparing lesions in rat brain. Science 219:316–18.
42. Beal, M. F. et al. 1993. Neurochemical and histologic characterization of striatal
excitotoxic lesions produced by the mitochondrial toxin 3-nitropropionic acid. J.
Neurosci.13:4181–92.
43. Borlongan, C. V., Koutouzis, T. K., and Sanberg, P. R. 1997. 3-Nitropropionic acid
animal model and Huntington’s disease. Neurosci. Biobehav. Rev. 21:289–93.
44. Meldrum, A., Page, K. J., Everitt, B. J., and Dunnett, S. 1999. Mitochondrial inhibitors
as tools for neurobiology. In Mitochondrial Inhibitors and Neurodegenerative Disor-
ders, eds. P. R. Sanberg, H. Nishino,H, and C. Borlongan. Totowa, NJ: Human Press,
201–208.
45. Mangiarini, L., et al. 1996. Exon 1 of the HD gene with an expanded CAG repeat
is sufficient to cause a progressive neurological phenotype in transgenic mice. Cell
87:493–506.
46. Lione, L. A., et al. 1999. Selective discrimination learning impairments in mice express-
ing the human Huntington’s disease mutation. J. Neurosci. 19:10428–37.
47. Carter, R. J., et al. 1999. Characterization of progressive motor deficits in mice trans-
genic for the human Huntington’s disease mutation. J. Neurosci. 19:3248–57
48. Morton, A. J., Lagan, M. A., Skepper, J. N., and Dunnett, S. B. 2000. Progressive for-
mation of inclusions in the striatum and hippocampus of mice transgenic for the human
Huntington’s disease mutation. J. Neurocytol. 29:679–702.
49. Wheeler, V. C., et al. 1999. Length-dependent gametic CAG repeat instability in the
Huntington’s disease knock-in mouse. Hum. Mol. Genet. 8:115–22.
50. Schwarcz, R., Fuxe, K., Agnati, L. F., Hokfelt, T., and Coyle, J. T. 1979. Rotational
behaviour in rats with unilateral striatal kainic acid lesions: A behavioural model for
studies on intact dopamine receptors. Brain Res. 170:485–95.
51. Dunnett, S. B., Isacson, O., Sirinathsinghji, D. J., Clarke, D. J., and Bjorklund, A. 1988.
Striatal grafts in rats with unilateral neostriatal lesions--III. Recovery from dopa-
mine-dependent motor asymmetry and deficits in skilled paw reaching. Neuroscience
24:813–20.
52. Borlongan, C. V., Randall, T. S., Cahill, D. W., and Sanberg, P. R. 1995. Asymmetrical
motor behavior in rats with unilateral striatal excitotoxic lesions as revealed by the
elevated body swing test. Brain Res. 676:231–34.
53. Montoya, C. P., Astell, S., and Dunnett, S. B. 1990. Effects of nigral and striatal grafts
on skilled forelimb use in the rat. Prog. Brain Res. 82:459–66.
54. Montoya, C. P., Campbell-Hope, L. J., Pemberton, K. D., and Dunnett, S. B. 1991. The
“staircase test”: A measure of independent forelimb reaching and grasping abilities in
rats. J. Neurosci. Methods 36:219–28.
55. Skinner, B. F. 1938. The behavior of organsims. New York: Appleton-Century-Crofts.
56. Everitt, B. J., Fray, P., Kostarczyk, E., Taylor, S., and Stacey, P. 1987. Studies of
instrumental behavior with sexual reinforcement in male rats (Rattus norvegicus):
I. Control by brief visual stimuli paired with a receptive female. J. Comp. Psychol.
101:395–406.
57. Dunnett, S. B., and Iversen, S. D. 1982. Neurotoxic lesions of ventrolateral but not
anteromedial neostriatum in rats impair differential reinforcement of low rates (DRL)
performance. Behav. Brain Res. 6:213–26.

58. Sanberg, P. R., Pisa, M., and Fibiger, H. C. 1979. Avoidance, operant and locomotor
behavior in rats with neostriatal injections of kainic acid. Pharmacol. Biochem. Behav.
10:137–44.
59. Dobrossy, M. D., Svendsen, C. N., and Dunnett, S. B. 1995. The effects of bilateral
striatal lesions on the acquisition of an operant test of short term memory. Neuroreport
6:2049–53.
60. Reading, P. J., Dunnett, S. B., and Robbins, T. W. 1991. Dissociable roles of the ventral,
medial and lateral striatum on the acquisition and performance of a complex visual
stimulus-response habit. Behav. Brain Res. 45:147–61.
61. Ungerstedt, U., and Arbuthnott, G. W. 1970. Quantitative recording of rotational behav-
ior in rats after 6-hydroxydopamine lesions of the nigrostriatal dopamine system.
Brain Res. 24:485–93.
62. Whishaw, I. Q., O’Connor, W. T., and Dunnett, S. B. 1986. The contributions of motor
cortex, nigrostriatal dopamine and caudate- putamen to skilled forelimb use in the rat.
Brain 109(Pt 5):805–43.
63. Evenden, J. L., and Robbins, T. W. 1984. Effects of unilateral 6-hydroxydopamine
lesions of the caudate-putamen on skilled forepaw use in the rat. Behav. Brain Res.
14:61–68.
64. Marshall, J. F., Richardson, J. S., and Teitelbaum, P. 1974. Nigrostriatal bundle damage
and the lateral hypothalamic syndrome. J. Comp. Physiol. Psychol. 87:808–30.
65. Ljungberg, T., and Ungerstedt, U. 1976. Sensory inattention produced by 6-hydroxydo-
pamine-induced degeneration of ascending dopamine neurons in the brain. Exp. Neu-
rol. 53:585–600.
66. Marshall, J. F., Turner, B. H., and Teitelbaum, P. 1971. Sensory neglect produced by
lateral hypothalamic damage. Science 174:523–25.
67. Marshall, J. F., and Teitelbaum, P. 1974. Further analysis of sensory inattention follow-
ing lateral hypothalamic damage in rats. J. Comp. Physiol. Psychol. 86:375–95.
68. Turner,B.H. 1973. Sensorimotor syndrome produced by lesions of the amygdala and
lateral hypothalamus. J. Comp Physiol Psychol. 82: 37-47.
69. Robbins, T. W., Muir, J., Killcross, A. S., and Price, B. H. 1993. In Behavioural neuro-
science, Vol. I, ed. A. Sahgal. Oxford: IRL Press.
70. Carli, M., Evenden, J. L., and Robbins, T. W. 1985. Depletion of unilateral striatal
dopamine impairs initiation of contralateral actions and not sensory attention. Nature
313:679–82.
71. Carli, M., Jones,G. H., and Robbins, T. W. 1989. Effects of unilateral dorsal and ven-
tral striatal dopamine depletion on visual neglect in the rat: A neural and behavioural
analysis. Neuroscience 29:309–27.
72. Brown, V. J., and Robbins, T. W. 1989. Deficits in response space following unilateral
striatal dopamine depletion in the rat. J. Neurosci. 9:983–89.
73. Brasted, P. J., Humby, T., Dunnett, S. B., and Robbins, T. W. 1997. Unilateral lesions
of the dorsal striatum in rats disrupt responding in egocentric space. J. Neurosci. 17:
8919–26.
74. Heilman, K. M. 1979. In Clinical neuropsychology, ed. E. S. Valenstein. Oxford:
Oxford University Press.
75. Bisiach, E., and Luzzatti, C. 1978. Unilateral neglect of representational space. Cortex
14:129–33.
76. Rosvold, H. E. 1972. The frontal lobe system: Cortical-subcortical interrelationships.
Acta Neurobiol. Exp. (Warsz. ) 32:439–60.
77. Rosvold, H. E., and Delgado, J. M. 1956. The effect on delayed-alternation test perfor-
mance of stimulating or destroying electrically structures within the frontal lobes of
the monkey’s brain. J. Comp. Physiol. Psychol. 49:365–72.

78. Krettek, J. E., and Price, J. L. 1977. The cortical projections of the mediodorsal nucleus
and adjacent thalamic nuclei in the rat. J. Comp. Neurol. 171:157–91.
79. Leonard, C. M. 1969. The prefrontal cortex of the rat. I. Cortical projection of the
mediodorsal nucleus. II. Efferent connections. Brain Res. 12:321–43.
80. Dunnett, S. B., and Iversen, S. D. 1981. Learning impairments following selective kai-
nic acid-induced lesions within the neostriatum of rats. Behav. Brain Res. 2:189–209.
81. Divac, I., Markowitsch, H. J., and Pritzel, M. 1978. Behavioral and anatomical
consequences of small intrastriatal injections of kainic acid in the rat. Brain Res.
151:523–32.
82. Sanberg, P. R., Lehmann, J., and Fibiger, H. C. 1978. Impaired learning and memory
after kainic acid lesions of the striatum: A behavioral model of Huntington‘s disease.
Brain Res. 149:546–51.
83. D’Amato, M. R., and O’Neill, W. 1971. Effect of delay-interval illumination on match-
ing behavior in the capuchin monkey. J. Exp. Anal. Behav. 15:327–33.
84. Aggleton, J. P. 1985. One-trial object recognition by rats. Quart. J. Exp. Psychol. B.
37:279.
85. Dunnett, S. B. 1993. In Behavioural neuroscience: A technical approach, ed. A Sahgal,
123. Oxford: IRL Press.
86. Dunnett, S. B. 1985. Comparative effects of cholinergic drugs and lesions of nucleus
basalis or fimbria-fornix on delayed matching in rats. Psychopharmacology (Berl.)
87:357–63.
87. Goldman-Rakic, P. S. 1989. Handbook of physiology: The nervous system V, 373. Bal-
timore: American Physiological Association.
88. Kowalska, D. M., Bachevalier, J., and Mishkin, M. 1991. The role of the inferior pre-
frontal convexity in performance of delayed nonmatching-to-sample. Neuropsycholo-
gia 29:583–600.
89. Mishkin, M., and Manning, F. J. 1978. Non-spatial memory after selective prefrontal
lesions in monkeys. Brain Res. 143:313–23.
90. Dunnett, S. B. 1990. Role of prefrontal cortex and striatal output systems in short-term
memory deficits associated with ageing, basal forebrain lesions, and cholinergic-rich
grafts. Can. J. Psychol. 44:210–32.
91. Dobrossy, M. D., Svendsen, C. N., and Dunnett, S. B. 1996. Bilateral striatal lesions
impair retention of an operant test of short- term memory. Brain Res. Bull. 41:159–65.
92. Dunnett, S. B. 1990. Is it possible to repair the damaged prefrontal cortex by neural
tissue transplantation? Prog. Brain Res. 85:285–96.
93. Jacobsen, C. F. 1936. Studies of cerebral function in primates. I. The functions of the
frontal association areas in monkeys. Comp. Psychol. Monogr. 13:3.
94. Jacobsen, C. F. 1937. Studies of cerebral function in primates. IV. The effect of frontal
lobe lesions on the delayed alternation habit in monkeys. J. Comp. Psychol. 23:101.
95. Brandt, J., et al. 1984. Clinical correlates of dementia and disability in Huntington’s
disease. J. Clin. Neuropsychol. 6:401–12.
96. Georgiou, N., Bradshaw, J. L., Phillips, J. G., and Chiu, E. 1997. Effect of directed
attention in Huntington’s disease. J. Clin. Exp. Neuropsychol. 19:367–77.
97. van Haaren, F., de Bruin, J. P., Heinsbroek, R. P., and van de Poll, N. E. 1985. Delayed
spatial response alternation: Effects of delay-interval duration and lesions of the medial
prefrontal cortex on response accuracy of male and female Wistar rats. Behav. Brain
Res. 18:41–49.
98. Numan, R., and Quaranta, J. R. Jr. 1990. Effects of medial septal lesions on operant
delayed alternation in rats. Brain Res. 531:232–41.
99. Heise, G. A., Conner, R., and Martin, R. A. 1976. Effects of scopolamine on vari-
able intertrial interval spatial alternation and memory in the rat. Psychopharmacology
(Berl.) 49:131–37.

100. Mogensen, J., Iversen, I. H., and Divac, I. 1987. Neostriatal lesions impaired rats‘
delayed alternation performance in a T-maze but not in a two-key operant chamber.
Acta Neurobiol. Exp. (Wars. ) 47:45–54.
101. Dunnett, S. B., Nathwani, F., and Brasted, P. J. 1999. Medial prefrontal and neostriatal
lesions disrupt performance in an operant delayed alternation task in rats. Behav. Brain
Res. 106:13–28.
102. White, A., and Dunnett, S. B. 2006. Fronto-striatal disconnection disrupts operant
delayed alternation performance in the rat. Neuroreport 17:435–41.
103. Robbins, T. W. 2002. The 5-choice serial reaction time task: Behavioural pharmacol-
ogy and functional neurochemistry. Psychopharmacology (Berl.)163:362–80.
104. Knopman, D., and Nissen, M. J. 1991. Procedural learning is impaired in Huntington‘s
disease: Evidence from the serial reaction time task. Neuropsychologia 29:245–54.
105. Trueman, R. C., Brooks, S. P., and Dunnett, S. B. 2005. Implicit learning in a serial
choice visual discrimination task in the operant 9-hole box by intact and striatal
lesioned mice. Behav. Brain Res. 159:313–22.
106. Isacson, O., Dunnett, S. B., and Bjorklund, A. 1986. Graft-induced behavioral recovery
in an animal model of Huntington disease. Proc. Natl. Acad. Sci. USA 83:2728–32.
107. Jacobsen, C. F. 1936. Studies of cerebral function in primates. III. A note on the effect
of motor and premotor lesions on delayed response in monkeys. Comp. Psychol. Monogr.
13:66.
108. Sanberg, P. R., and Coyle, J. T. 1984. Scientific approaches to Huntington’s disease.
CRC Crit. Rev. Clin. Neurobiol. 1:1–44.
109. Berridge, K. C. 1996. Food reward: Brain substrates of wanting and liking. Neurosci.
Biobehav. Rev. 20:1–25.
110. Mogenson, G. J., Jones, D. L., and Yim, C. Y. 1980. From motivation to action: Func-
tional interface between the limbic system and the motor system. Prog. Neurobiol.
14:69–97.
111. Salamone, J. D., Cousins, M. S., and Snyder, B. J. 1997. Behavioral functions of nucleus
accumbens dopamine: Empirical and conceptual problems with the anhedonia hypoth-
esis. Neurosci. Biobehav. Rev. 21:341–59.
112. Salamone, J. D., Kurth, P., McCullough, L. D., and Sokolowski, J. D. 1995. The effects of
nucleus accumbens dopamine depletions on continuously reinforced operant respond-
ing: Contrasts with the effects of extinction. Pharmacol. Biochem. Behav. 50:437–43.
113. Hodos, W., and Kalman, G. 1963. Effects of increment size and reinforcer volume on
progressive ratio performance. J. Exp. Anal. Behav. 6:387–92.
114. Eagle, D. M., Humby, T., Dunnett, S. B., and Robbins, T. W. 1999. Effects of regional
striatal lesions on motor, motivational, and executive aspects of progressive-ratio per-
formance in rats. Behav. Neurosci. 113:718–31.
115. Skjoldager, P., Pierre, P. J., and Mittleman, G. 1993. Reinforcer magnitude and progres-
sive ratio responding in the rat: Effects of increased effort, prefeeding and extinction.
Learn. Motiv. 24:303
116. Dunnett, S. B., and Everitt, B. J. 1998. Cell transplantation for neurological disorders,
eds. T. B. Freeman and J. H. Kordower, 135. Totowa, NJ: Humana Press.
117. Hodgson, J. G., et al. 1999. A YAC mouse model for Huntington’s disease with full-
length mutant huntingtin, cytoplasmic toxicity, and selective striatal neurodegenera-
tion. Neuron 23:181–92.

12 Working Memory
Delayed Response
Tasks in Monkeys
Jesse S. Rodriguez and Merle G. Paule
CONTENTS
12.1 Introduction................................................................................................. 247
12.2 Methods.......................................................................................................248
12.2.1 Animal Subjects...............................................................................248
12.2.2 Equipment ........................................................................................ 249
12.2.3 Delayed Response Tasks.................................................................. 250
12.2.4 Delayed Alternation Tasks............................................................... 251
12.2.5 Delayed Matching-to-Sample Tasks ................................................ 251
12.2.6 Delayed Non-Matching-to-Sample Tasks ........................................ 252
12.2.6.1 Trial-Unique Versus Repetitive Stimuli ............................ 252
12.3 Data Analysis and Interpretation ................................................................ 252
12.6 Limitations and Conclusions....................................................................... 258
References.............................................................................................................. 262
12.1 INTRODUCTION
Accepted taxonomies of memory typically distinguish among different kinds of
remembering depending upon the information that must be remembered. A basic
dichotomy distinguishes between the retention of factual or experiential information
on the one hand, and the retention of habits and motor skills on the other. Factual
memory can be further differentiated into working and reference memory. As first
described by Werner Honig’s group,1 working memory is required when “different
stimuli govern the criterion response on different trials, so that the cue that the ani-
mal must remember varies from trial to trial.” Thus, working memory is required
for remembering information that varies unpredictably in time and/or in content: it
is this type of memory that is decimated in Alzheimer’s disease and other demen-
tias.2 In contrast, reference memory is used to retain information that remains con-
stant over time (e.g., removing the cup from a baited well provides access to food).
The study of working memory processes is generally accomplished using delayed
response (DR) tasks and, within the confines of even relatively short test sessions,
one can readily assess processes associated with short-term memory using these
247

procedures. In such tasks, one presents a bit of information (a sample) to a subject,

withdraws that information, waits for a period of time (recall delay), then presents
that same bit of information along with a comparison bit and asks the subject to
identify (choose) which bit of information was presented previously. This process is
then repeated across a number of trials.
DR tasks using nonhuman primates as surrogates are used primarily to deter-
mine the biological underpinnings of human learning and memory processes and
their associated mechanisms. The complex brain functions associated with the per-
formance of these tasks can be assessed using a variety of approaches, and the care-
ful monitoring of behavioral outputs provides important experimental advantages.
Many of the different types of assessments of memory processes have evolved from
studies ranging from those employing neuroanatomical lesions to electrophysiologi-
cal recordings and, more recently, neuroimaging techniques. Goldman-Rakic et al.
have shown the importance of the prefrontal cortex in the performance of DR tasks
in nonhuman primates by employing lesions and electrophysiological and neuroim-
aging techniques.3–6 Brain lesion experiments in monkeys have also demonstrated
the importance of the hippocampus and its relevance to performance in DR tasks.7
Through the use of lesions, Zola et al. (2000) and Squire (2004) have also demon-
strated the importance of the medial temporal lobe in performance of these types
of tasks in monkeys.8,9 Taken together these and other studies have elucidated the
importance of the prefrontal cortex in visual-spatial functions, inhibition of behavior
(temporal mechanisms), decision making, working memory, problem solving, plan-
ning, and organizing. Likewise, the findings that the hippocampal formation (medial
temporal lobe) is important in memory consolidation, associative memory formation,
declarative memory, and recognition memory stem from these and similar studies.
The prefrontal cortex and the hippocampal formation are two brain areas intricately
involved in the performance of DR tasks. The understanding of short-term memory
and, more specifically, working memory has benefited greatly from this work.
12.2 METHODS
Animal models are essential components of basic science and preclinical research,
and their use in behavioral experiments is often more practical than is the use of
human subjects. While the most commonly used research animals are rodents, non-
human primates are phylogenetically the most similar to humans. Goldman-Rakic
et al. have observed that “the organization of the cortical dopamine system is essen-
tially the same in macaque monkey and human and that the nonhuman primate is
a suitable animal model for analysis of dopamine function in prefrontal cortex.”10
Nonhuman primates are excellent models for studying behavior during different
periods of development and for use in pharmacological and toxicological studies,
and the ethics of using these animals in research has been eloquently addressed.11
Since DR tasks often use food reinforcement, food restriction is commonly
used as a means of increasing motivation and positive behavioral output.12,13 Caloric
intake is usually reduced by about 15%–20% from that of free-feeding animals and

Working Memory 249
can result in many positive effects such as enhanced quality and length of life.14–17
Studies performed in nonhuman primates are easily relatable to studies conducted in
lower species, such as rodents, and in humans.18
12.2.2 EQUIPMENT
Most of the early DR tasks were conducted using the Wisconsin General Test Appa-
ratus or WGTA (Figure 12.1). In this apparatus a monkey sits in a cage or a restraint
chair in front of a tray that contains recessed food wells. The experimenter baits a
well with food, covers it, and then lowers a screen to block the experimental tray
from the monkey’s view. Subjects can retrieve the food reinforcer after the screen
is removed at some later time. This approach, while very productive, is labor inten-
sive and prone to the vagaries of experimenter–subject interactions. More recently,
however, automated behavioral systems have become increasing popular. In a typical
primate behavioral test system, the subject is either placed in a behavior chamber
in which it interacts with a response panel, or a test panel is positioned so that the
animal can interact with it; for example, a panel can be temporarily affixed to its
home cage during test sessions. The response panels can be outfitted with a vari-
ety of manipulanda such as response levers or bars or press-plates on which visual
stimuli can be presented. In these configurations, there is generally a food cup or
trough into which food reinforcers are delivered when correct responses are made.
An example of the behavioral panel used in one such system, in this case the National
Center for Toxicological Research (NCTR) Operant Test Battery or OTB,19 is shown
in Figure 12.2.
Recently, a touch-screen-based system, the CANTAB (Cambridge Neuropsy-
chological Test Automated Battery) system was developed for the assessment of cog-
nitive deficits in humans with neurodegenerative diseases or brain damage. It is now
One-way
vision screen
Forward opaque screen
Transport cage Stimulus

tray
FIGURE 12.1 Wisconsin General Test Apparatus. Source: Buccafusco, J. J. 2001. Methods
of behavior analysis in neuroscience, First Edition, Boca Raton, FL: Taylor and Francis.,
with permission.

FIGURE 12.2 Operant Test Battery panel used for nonhuman primate and human experiments.
being used by some laboratories to test nonhuman primates.20,21 The battery consists
of computerized tests of memory, attention, and executive function, one of which is
a delayed matching-to-sample (DMTS) task for the assessment of working memory.
The CANTAB apparatus, like all of the other systems discussed here, is nonver-
bal in nature, making it language independent and largely culture-free. CANTAB
performance has been standardized in an elderly human population and validated
in neurosurgical patients and in patients with basal ganglia disorders, Alzheimer’s
disease, depression, and schizophrenia.
12.2.3 DELAYED RESPONSE TASKS

As mentioned earlier, DR tasks have historically been performed using the WGTA.
Initially, in full view of the subject, the experimenter baits one of two or three rein-
forcer wells with food, covers all food wells with identical items (e.g., cardboard
plaques), then lowers an opaque screen to block the experimental tray from the mon-
key’s view. After a delay period the screen is lifted, allowing the monkey to recall
from memory and choose the baited food well. This DR task measures mnemonic
processes associated with working memory.22,23 Reinforcers are randomly distrib-
uted between the food wells over a number of trials (e.g., 30) that make up a daily test
session. During the initial training phase, delays are held constant at short values and
are gradually increased according to a stepwise procedure as animals demonstrate
mastery of the task.24 DR task performance has been used to examine the involve-

Working Memory 251
ment of specific neurotransmitters with aspects of working memory. For example, it

has been shown that guanfacine, a norepinephrine agonist, improves performance in
young adult rhesus monkeys,25 suggesting that the norepinephrine system is impor-
tant in the modulation of working memory.
12.2.4 DELAYED ALTERNATION TASKS

There are several variations of delayed alternation tasks, for example, delayed spatial
alternation and delayed object alternation (nonspatial). For both of these tasks, the
trials are temporally related to one another: a correct response is dependent upon the
previous response. The delayed spatial alternation task is a two-choice task in which
the monkey has to alternately displace a left or right plaque to retrieve a reward, i.e.,
if the right food well was baited on one trial, then the left food well will be baited
on the next trial. This task is typically used with a delay or recall duration of at least
5 sec. The delayed object alternation task is also a two-choice task in which the
monkey needs to correctly select an object that alternates between trials. Levy et
al.26 have reported that delayed spatial alternation tasks increase neural metabolic
activity, as measured by local cerebral glucose utilization, in the head of the caudate
nucleus where efferents from the dorsolateral prefrontal cortex project most densely.
Performance of a delayed object alternation task increase neural metabolic activity
in the body of the caudate nucleus, which is innervated by the temporal cortex. These
findings show that spatial and nonspatial cognitive operations are subserved by dis-
tinct cortico-striatal circuits.
12.2.5 DELAYED MATCHING-TO-SAMPLE TASKS

For DMTS tasks, as with other DR tasks, each trial begins with the presentation of a
sample stimulus. As used in the NCTR OTB, this involves illumination of the center
one of three horizontally aligned press-plate manipulanda (Figure 12.2) with one
of seven white-on-black geometric shapes. Subjects make an observing response to
the initial shape (sample stimulus) by pushing the illuminated plate, after which it is
immediately extinguished. Following an interval (recall delay) that varies randomly
(from 2 to 32 sec, for example), all three press-plates are illuminated, each with a
different stimulus shape, but one of which matches the sample stimulus. A choice
response to the plate matching the sample stimulus results in food reinforcer deliv-
ery. Incorrect choices are followed by a 10-sec timeout, then initiation of a new trial.
As indicated by lesion studies,27 this task28 and other DMTS procedures are thought
to depend upon the prefrontal and inferior temporal cortices. Damage to the higher
visual area inferior temporal cortex interferes with the performance of the task for
visual stimuli,27 while lesions of the prefrontal cortex still allow successful comple-
tion of the task but only for short recall delay periods.29
The use of distractors in DMTS tasks has been employed to increase task dif-
ficulty (for review, see30). Distractors are thought to inhibit the subject’s ability to
sustain attention and serve to impair working memory. Distractors are usually either
auditory or visual stimuli presented during the delay or recall interval to distract
the monkey from relevant test stimuli. It is believed that the presentation of distrac-
tors during the delay or recall interval, a time during which selective attention and

rehearsing are thought to be important, disrupts the cognitive processes subserving

working memory.
12.2.6 DELAYED NON-MATCHING-TO-SAMPLE TASKS

These procedures are conceptually similar to DMTS paradigms, however, in delayed
non-matching-to-sample (DNMTS) tasks, the subject is required to choose the stimu-
lus that does not match the test (sample) stimulus that was presented before the recall
delay. This task has been used extensively in monkeys to identify the brain regions
involved in recognizing previously presented stimuli.8,31 In a typical DNMTS task,
the subject is presented with a trial-unique (unfamiliar) sample stimulus followed by
a recall delay during which the sample is removed from sight, and then presentation
of the sample and a novel stimulus item. Choosing the novel stimulus is reinforced
as the correct response. It is thought that monkeys learn to perform DNMTS tasks
much more rapidly than they learn to perform DMTS tasks because of their natural
tendency to preferentially attend to novel stimuli, in this case, the non-matching
stimulus. Lesion studies in monkeys have revealed a number of interconnected areas
involved in the encoding, retention, and retrieval of stimulus representations within
DNMTS trials. This memory circuit consists principally of the ventromedial and
ventrolateral prefrontal cortices, rhinal (entorhinal and perirhinal) cortex, thalamus,
and the inferotemporal cortex.31–34 The hippocampus, located in the medial temporal
cortex, might also be a component of this circuit, but its precise role is currently
the subject of debate.8,31 Recent findings have suggested that an intact hippocampal
formation is critical for acquisition of DNMTS task performance, and specifically
for the delay component.35
12.2.6.1 Trial-Unique Versus Repetitive Stimuli

Trial-unique stimuli are stimuli that do not appear in more than one trial during a
test session or during many test sessions (i.e., they are unfamiliar). Repetitive stim-
uli, on the other hand, may repeat in multiple trials during a given test session. Tasks
that employ trial-unique stimuli activate the medial temporal lobe more than tasks
that employ repetitive stimuli.36 This suggests a stronger role for the medial tem-
poral lobe in working memory for complex, novel, trial-unique stimuli, relative to
working memory for familiar stimuli. However, repetitive stimuli tasks activate the
prefrontal and parietal cortices more than tasks that employ trial-unique stimuli.36
This suggests that the prefrontal and parietal cortices play a stronger role in working
memory for familiar stimuli. These differences are thought to reflect the engagement
of medial temporal lobe regions in tasks that require the formation and maintenance
of new short-term representations; whereas the prefrontal cortex is preferentially
engaged when prior representations already exist in the brain but must be selectively
updated and monitored to avoid interference effects.36
12.3 DATA ANALYSIS AND INTERPRETATION

The analyses of data obtained using the working memory tasks described here some-
times vary between laboratories, but, generally, common endpoints are used. Ideal-

Working Memory 253
100
Percent Accuracy
75
50
25 A
B
0
D C
0 5 10 15 20 25 30 35
Delay Intervals (seconds)
FIGURE 12.3 Idealized delayed response task data
ized data for typical DR tasks are shown in Figure 12.3, where response accuracy is
plotted on the ordinate, and length of recall delay is plotted on the abscissa. Typical
data from normal subjects might be represented by line A, where it can be seen that
at zero—or very short—delays recall accuracies are very high. The point where
these data intercept the y axis is thought to be related to the ability of subjects to
attend to the task, discriminate the stimuli, and encode the information relevant
to problem solution. Thus, if aspects of attention, discriminability, or encoding are
degraded, accuracy at zero or very short delays will decrease. The slope of this line
indicates the normal rate of short-term/working memory decay. Line B represents
data showing a decrease in initial attention and/or encoding, with a normal rate of
forgetting; similar observations have been noted after the administration of the hal-
lucinogen lysergic acid diethylamide (LSD) (see also Figure 12.7).37 The data in line
C represent a circumstance under which the initial level of attention and/or encoding
is no different from that of the normal condition (line A), but under which the rate of
forgetting has been increased (slope is steeper). This kind of effect can be seen after
surgical ablations of the hippocampus and amygdala (Figure 12.4),8 after treatment
with certain drugs, such as tetrahydrocannabinol (THC),38 and when distractors are
used.39 Line D represents data showing a decrease in both initial attention and/or
encoding and an increase in the rate of forgetting. These data examples are rel-
evant to DR tasks whether or not they have time limitations (predetermined session
lengths), and whether or not they include the measurement of response speed (laten-
cies). By setting maximum session lengths and measuring response speed, however,
several additional important metrics of task performance can be had.
The typical measure used by our lab for DNMTS and DMTS tasks is the percent
task completed (PTC) measure. This measure is a function of both response accuracy
and speed, so changes in either of those measures can affect the PTC. Conceivably,
an experimental treatment (drug or toxicant exposure, brain lesion, other stressor,
etc.) could increase response rate while decreasing accuracy and, thus, have no effect
on PTC. Conversely, response rate could be decreased and accuracy increased, again
resulting in no effect on PTC. Practically speaking, these effect combinations have
not been seen in our lab. Typically, if a treatment has an effect, it will manifest in

Delayed Nonmatching to Sample

100
*
*
90
Percent Correct
80 *
*
N (10)
70
H (18a)
60
50
8s 15 s 1 m 10 m 40 m
Delay
FIGURE 12.4 Data are from hippocampal lesioned monkeys (closed squares) as compared
to nonlesioned animals (open squares) for a delayed non-matching-to-sample task. Lesioned
animals performed significantly lower in accuracy for most delay intervals as compared to
the control group. Source: From Zola, S. M., Squire, L. R., Teng, E., Stefanacci, L., Buffalo,
E. A., and Clark, R. E. 2000. Impaired recognition memory in monkeys after damage limited
to the hippocampal region. J. Neurosci 20:451–63, with permission.
the PTC metric, alerting one to look further for more information as to the spe-
cific nature of the effect. Thus, overall accuracy (collapsed across all recall delays)
and overall response latencies (both observing and choice) are examined next. Since
latencies are inverses of response rates, they provide metrics of speed of responding.
Observing response latency is defined as the time elapsed before the subject initiates
a trial by responding to or “observing” the initial sample stimulus. Choice response
latency is defined as the time to respond to a choice stimulus following the presen-
tation of choices. Choice response latency can be determined for each recall delay,
whereas observing response latency is generally independent of recall delay.
An increase or decrease in overall accuracy is indicative of short-term or work-
ing memory effects, particularly if this effect is seen in the absence of any effect on
response latencies. This effect can manifest during a particular delay interval and
may or may not be delay dependent. Delay-dependent effects would be those that
change as a function of recall delay. For example, a treatment may have no effect on
response accuracy at short delays (no or little effect on attention and/or encoding),
but have significant effects on accuracy at the longer delays (increased rate of forget-
ting or distractibility; see also hypothetical line C in Figure 12.3).
An increase or decrease in observing response latency following a pharmaco-
logical challenge or other experimental manipulation can be indicative of effects
on reaction time (psychomotor speed), motivation, and/or motoric capabilities. An
increase or decrease in overall choice response latency can be indicative of an effect
on attention, reaction time, motivation, or motoric capabilities. If a treatment effect
on choice response latency increases with increasing length of recall delay, then that

Working Memory 255
effect is likely related to attentional processes/distractibility. This stems from the

fact that there is increasing opportunity for distraction with increasing recall delay.

There is a wide range of applications for which DR tasks are useful. Many stud-
ies have investigated the effects of pharmacological challenges on DR tasks in the
developing monkey (for review, see40). Other studies have observed the effects of
brain lesions,41 environmental factors,42,43 and human diseases such as Alzheimer’s
disease and attention deficit hyperactivity disorder (ADHD).44–46
Old monkeys can sometimes serve as good models for age-related cognitive
decline and Alzheimer’s disease because they show impaired performance in the
DMTS task as compared with younger monkeys.47 Some of the earliest findings of
drug effects on DMTS task performance demonstrated the importance of the nor-
epinephrine system in performance of DR tasks. Jackson and Buccafusco48 showed
that the norepinephrine agonist, clonidine, when administered to either young or
aged monkeys, resulted in a significant improvement across all trials in DMTS task
performance. Similarly, Franowicz and Arnsten25 reported that guanfacine, another
norepinephrine agonist, improved DR task performance in young adult rhesus
monkeys.
The function of the acetylcholine system has, however, remained the focus of
memory and Alzheimer’s disease research because of its suspected role in learn-
ing and memory processes and in the etiology of Alzheimer’s disease. Applica-
bly, a study by Decker et al.49 showed that administration of ABT-089 (a nicotinic
acetylcholine receptor agonist) to aged monkeys produced a robust improvement
in DMTS accuracy at all delay intervals. This effect was not age-specific because
task improvement was also seen in younger adult monkeys. Similarly, Prendergast
et al.50 reported improvement in accuracy in a delayed recall task in both aged and
younger adult monkeys after administration of the nicotinic receptor agonist, ABT-
418. Additionally, Jackson et al.51 found that the acetylcholinesterase inhibitor, veln-
acrine, improved DMTS performance in aged monkeys. These findings, relevant to
the study and treatment of Alzheimer’s disease, clearly demonstrate the importance
of the acetylcholine system in working memory processes and age-associated mem-
ory decline.
Monkey subjects have also been used to evaluate aspects of attention and focus
using DR tasks. ADHD patients exhibit difficulties in sustaining focus/attention.
Therefore, the DMTS task has been modified to include a component in which dis-
tractors (flashing sample and choice stimuli) are presented during the recall delay
intervals. Distractors inhibit the subject’s ability to sustain attention and impair
working memory during DR tasks. Prendergast et al.52 found that nicotine, ABT-
089, and ABT-418 (all acetylcholine agonists) attenuated the effects of distractors
in the DMTS task by preventing distractor-induced declines in recall accuracy in
adult monkeys. Other potentially beneficial drugs derive from a class of serotonin
(5-HT) receptor antagonists. Terry et al.53 reported that the 5-HT3 receptor antago-
nist RS56812 enhances DMTS accuracy in monkeys at long recall delays. This com-
pound presumably acts through 5-HT3 receptors on acetylcholine axon terminals

where 5-HT receptor inactivation is thought to lead to increases in acetylcholine

release. Additionally, Terry et al.54 found that the 5-HT4 receptor antagonist RS17017
enhances DMTS accuracy in both young and old macaque monkeys. These findings
suggest the potential development of new treatments for ADHD and other memory
disorders that would be alternatives to the psychostimulant medications currently in
common use.
In addition to the cholinergic agonists and serotonin antagonists that enhance
performance on DR tasks, there are, not unexpectedly, also a host of drugs that
degrade such behavior in nonhuman primates. Schulze et al.55,56 investigated the
acute affects of delta-9-THC (the active ingredient in marijuana smoke) and mari-
juana smoke itself in rhesus monkeys as measured by performance in the NCTR
OTB. The results for the DMTS task showed that THC decreased both the num-
ber of reinforcers earned and the percent task completed, primarily by increasing
the time it took subjects to begin each trial (increased mean observing response
latencies). Marijuana smoke administration, at exposure levels that produced plasma
THC levels similar to those seen in humans after marijuana smoke exposure, also
significantly increased response latencies (increased time to initiate trials and to
make recall choices). Similarly, this lab reported that diazepam, a gamma-amino
butyric acid (GABA) agonist, significantly decreased DMTS task accuracy in rhesus
monkeys across several different time delays.56 Additionally, the μ opiate receptor
agonist, morphine, significantly decreased the number of reinforcers earned and
percent task completed and increased response latencies, while having little effect
on recall accuracy.57 Finally, the psychostimulant amphetamine, while having no
effects on DMTS task performance at low doses, significantly decreased percent
task completed, increased observing response latencies, and decreased accuracy at
the longer recall delays.58

In this section representative nonhuman primate data for all of the tasks described
in this chapter are briefly discussed. Figure 12.5A and B show data for a DR task
in which performance has been enhanced: overall accuracy (Figure 12.5A) was
increased and rate of forgetting (Figure 12.5B) was decreased following pretreat-
ment with bextaxol, a G-1 adrenergic receptor antagonist.59 These data reveal a role
of the adrenergic system in DR performance. Figure 12.6 shows representative data
for monkey performance in a delayed alternation task,60 showing an impairment
in recall accuracy after the administration of the anxiogenic drug FG 7142 (a G-
carboline, GABAA receptor partial inverse agonist), and the amelioration of that
effect by pretreatment with three different dopamine receptor antagonists (haloperi-
dol, SCH23390, and clozapine). Additionally, a benzodiazepine receptor antagonist
was administered along with FG 7142 to illustrate the GABA agonist’s receptor
specificity. These data demonstrate the involvement of the GABA and dopamine
systems in the performance of delayed alternation tasks.
Figure 12.7 shows a decrease in delayed matching-to-sample task accuracy
following treatment with LSD, a serotonin receptor partial agonist.37 These data
illustrate how this compound preferentially affects discriminability, encoding, or

Working Memory 257
# of Correct Choices (out of 6)

80 6
* *
Percent Correct
5 *
70
4
60
3
50
Saline .00011 .0011 .011 A B C D E
Delay Length
Betax (mg/kg)
(a) (b)
FIGURE 12.5 (a) The effect of betaxolol (G-1 adrenergic receptor antagonist) enhanced
delayed response performance in monkeys (n = 10). Data are shown as mean percent correct ±
SEM (out of 30 trials) following systemic administration of saline, .00011, .0011, or .011 mg/
kg of betaxolol. *Significant difference from saline (p = .008). (b) The effect of betaxolol on
the delayed response performance of monkeys at different recall delays. The most effective
dose was selected for each monkey (n = 8). Data are shown as mean number correct ± SEM
(out of six trials/delay) for five different delays (A, B, C, D, and E) following betaxolol treat-
ment (closed squares). *Denotes significant difference from saline (open squares). Source:
From Ramos, B. P., Colgan, L., Nou, E., Ovadia, S., Wilson, S. R., and Arnsten, A. F. 2005.
The beta-1 adrenergic antagonist, betaxolol, improves working memory performance in rats
and monkeys. Biol. Psychiatry 58:894–900, with permission.
attention, while not affecting rate of memory decay. Additionally, a recent report has
demonstrated an enhanced performance accuracy in the DMTS task following treat-
ment with a serotonin receptor antagonist.61 In that report, the length of recall delays
were adjusted subject to attain similar levels of performance accuracy for all subjects:
(1) a least difficult zero delay (85%–100% accuracy), (2) a short delay (75%–84%
accuracy), (3) a medium delay (65%–74% accuracy), and (4) a long delay represent-
ing chance performance (55%–64% accuracy). Figure 12.8A shows representative
data of drug-enhanced DMTS task accuracy in young monkeys at a medium-dif-
ficulty recall delay. This increase in accuracy resulted from treatment with EMD
281014, a serotonin 2A receptor antagonist.61 Figure 12.8B shows data for this same
compound where it is shown to improve delayed matching performance in aged rhe-
sus monkeys at both medium- and long-recall delays.61 Figure 12.9 shows data from
animals performing a DNMTS task using the CANTAB system. Here a decrease
in accuracy of task performance was clear following treatment with the muscarinic
cholinergic receptor antagonist scopolamine.62 Figure 12.10, on the other hand, dem-
onstrates that the cholinesterase inhibitor physostigmine can enhance accuracy of
DNMTS task performance.63 These examples of representative data serve to show
how different drugs from a variety of pharmacological classes can either degrade or
enhance performance in different DR tasks. These data also suggest that numerous
neurotransmitter systems are involved in performance of DR tasks.

100
Delayed Response, Percent Correct
90 **
**
**
80 * *
**
70 *
60
50
VEH VEH RO RO HAL HAL CLZ CLZ SCH SCH
+ + + + + + + + + +
VEH FG VEH FG VEH FG VEH FG VEH FG
(n = 4) (n = 4) (n = 4) (n = 4) (n = 4) (n = 4) (n = 4) (n = 4) (n = 3) (n = 3)
FIGURE 12.6 Effects of FG 7142 (a G-carotine, partial inverse agonist of the benzodiaz-
epine receptor) and dopamine receptor antagonists on delayed alternation performance in the
monkey. FG 7142 produced a significant impairment in response accuracy when compared
with vehicle performance. Although there was no change in performance with R015-1788
(a benzodiazepine receptor antagonist) when given alone, pretreatment with R015-1788 pre-
vented the FG 7142-induced impairment. Haloperidol and SCH23390 (both dopamine recep-
tor antagonists) significantly impaired cognitive performance when given alone. Clozapine
(another dopamine receptor antagonist) produced a small nonsignificant impairment in per-
formance when given alone. Haloperidol, clozapine, and SCH23390 all ameliorated the FG
7142–associated cognitive impairment when given as a pretreatment. Source: From Murphy,
B. L., Arnsten, A. F., Goldman-Rakic, P. S., and Roth, R. H. 1996. Increased dopamine
turnover in the prefrontal cortex impairs spatial working memory performance in rats and
monkeys. Proc. Natl. Acad. Sci. USA 93:1325–9, with permission.
12.6 LIMITATIONS AND CONCLUSIONS

DR tasks fulfill many of the criteria needed to measure short-term or working mem-
ory. There are, however, some aspects of these paradigms that pose problems if
not considered. First, in tasks employing few response locations, subjects frequently
develop a position bias, which is evidenced by a majority of their choice responses
being restricted to one of the locations, for example, to either the right or left response
position in a two-choice paradigm. As a result, on approximately 50% of all trials
the subject’s response to the right-hand location can result in a correct response and
reinforcement.64 In addition, even when such biases are not observed under control
conditions, they can be seen following drug administration, making interpretation of
dose-response curves and drug effects difficult.65 DR tasks can also suffer from the
occurrence of prominent ceiling and/or floor effects. If the task is not challenging
enough, response accuracies even at long delays can be very high, making detection

Working Memory 259
Short-Term Memory Task

120
100
80
Accuracy
60
0
40 0.0003
0.001
0.003
20
0.01
0.03
0
2 8 16 32 48 64
Time Delay (seconds)
FIGURE 12.7 Data from a delayed-matching-to-sample task comparing vehicle (shaded

area represents 95% confidence interval after saline administration) to lysergic acid diethyl-
amide (LSD, serotonin receptor partial agonist) in monkeys. The highest dose of LSD shows
a significant effect on accuracy of performance indicative of a decrease in attention and/or
encoding with no effect on rate of forgetting (slope of decay is unaffected). Source: From
Frederick, D. L., Gillam, M. P., Lensing, S., and Paule, M. G. 1997. Acute effects of LSD on
rhesus monkey operant test battery performance. Pharmacol. Biochem. Behav. 57:633–41,
with permission.
of treatment-induced increases in memory impossible. This problem can usually be

overcome by lengthening the delays and/or increasing the complexity of the stimuli
so that discrimination between sample and choice stimuli becomes more difficult.
Conversely, if the task is too difficult, a floor effect can be apparent, in which case
delays can be shortened or the difficulty of the discrimination can be made less
demanding.
In summary, DR tasks in nonhuman primates are very useful for assessing and
elucidating the biological underpinnings of working memory that are relevant to
humans. The clinical relevance of these types of tasks is further supported by the
observation that DMTS performance in humans correlates significantly with IQ.66
By inference, then, use of these tasks in animals would seem to provide direct access
to important aspects of animal intelligence.

Young Monkeys
100 Zero Delays 100 Short Delays
90 90
Matching Performance (Percent Correct)
80 80
70 70
60 60
50 50
80 80
Medium Delays Long Delays
*
Vehicle
70 70 EMD Compound
60 60
50 50
Veh 0.1 1.0 3.0 10.0 Veh 0.1 1.0 3.0 10.0
Dose (mg/kg)
(a)
Aged Monkeys
100 Zero Delays 100 Short Delays
90 90
Matching Performance (Percent Correct)
80 80
70 70
60 60
50 50
80 80
Medium Delays Long Delays
Vehicle
* EMD 281014 *
70 70
60 60
50 50
Veh 0.1 1.0 3.0 10.0 Veh 0.1 1.0 3.0 10.0
Dose (mg/kg)
(b)
FIGURE 12.8 (a) Representative data of drug-treatment-enhanced accuracy of delayed-

matching-to-sample (DMTS) task performance in young monkeys. Increased accuracy is
observed for medium recall delays after treatment with the selective serotonin 5-HT 2A recep-
tor antagonist EMD 281014. (b) The effect of EMD 281014 to improve DMTS performance
in aged rhesus monkeys at both medium and long recall delays. Delay durations were titrated
for each subject in order to equate task performance among subjects. Source: From Terry,
A. V. Jr., Buccafusco, J. J., and Bartoszyk, G. D. 2005. Selective serotonin 5-HT2A receptor
antagonist EMD 281014 improves delayed matching performance in young and aged rhesus
monkeys. Psychopharmacology (Berl.) 179:725–32, with permission.

Working Memory
100%
Simultaneous
0s 16s
90% 32s 64s
Percent Correct Choices
80%
70%
60%
50%
Baseline Vehicle 3 10 14 17
Scopolamine (µg/kg, i.m.)
FIgure . Effect of scopolamine on mean (n = 6, except for the 14 μg/kg dose; ± SEM)
choice accuracy in a delayed-non-matching-to-sample task expressed as the percent of trials
completed on which a correct choice was made showed impairment. Data are shown for the
simultaneous condition as well as for 0-, 16-, 32-, and 64-sec retention intervals. Random
responding corresponds to a choice accuracy of 50%. There was a significant main effect
of both retention interval and drug condition (p < 0.05). These data were generated using a
touch screen apparatus, i.e., CANTAB. Source: From Taffe, M. A., Weed, M. R., and Gold,
L. H. 1999. Scopolamine alters rhesus monkey performance on a novel neuropsychological
test battery. Brain Res. Cogn. Brain Res. 8:203–12, with permission.
90
*
80
Percent Correct
70
60
50
S 3.2 10 32
FIgure .0 The effect of physostigmine, a cholinesterase inhibitor, was to enhance

monkey accuracy on a delayed non-matching-to-sample task but only at a specific dose (s =
saline, 3.2, 10, and 32 μg/kg). Scores are group means (± SEM) as a function of dose. Mon-
keys received saline or the indicated dose of physostigmine i.m. 20 min before testing. *p <
0.05 vs. saline treatment. Source: From Ogura, H., and Aigner, T. G. 1993. MK-801 impairs
recognition memory in rhesus monkeys: comparison with cholinergic drugs. J. Pharmacol.
Exp. Ther. 266:60–4, with permission.

REFERENCES
1. Hulse, S. H. and Honig, W. K. 1978. Studies in working memory in the pigeon. In
Cognitive processes in animal behavior, eds. S. H. Hulse, H. Fowler, and W. K. Honig,
211–248. Hillsdale, NJ: Lawrence Erlbaum Associates.
2. Sullivan, E. V., Sagar, H. J., Gabrieli, J. D., Corkin, S., and Growdon, J. H. 1989. Differ-
ent cognitive profiles on standard behavioral tests in Parkinson’s disease and Alzheim-
er’s disease. J. Clin. Exp. Neuropsychol. 11:799–820.
3. Davachi, L., and Goldman-Rakic, P. S. 2001. Primate rhinal cortex participates in both
visual recognition and working memory tasks: Functional mapping with 2-DG. J. Neu-
rophysiol. 85:2590–601.
4. Constantinidis, C., Franowicz, M. N., and Goldman-Rakic, P. S. 2001. The sensory
nature of mnemonic representation in the primate prefrontal cortex. Nat. Neurosci.
4:311–6.
5. Goldman-Rakic, P. S. 1990. Cellular and circuit basis of working memory in prefrontal
cortex of nonhuman primates. Prog. Brain Res. 85:325–35; discussion 335–6.
6. Goldman-Rakic, P. S. 2002. The “psychic cell” of Ramon y Cajal. Prog. Brain Res.
136:427–34.
7. Murray, E. A., and Mishkin, M. 1998. Object recognition and location memory in
monkeys with excitotoxic lesions of the amygdala and hippocampus. J. Neurosci.
18:6568–82.
8. Zola, S. M., Squire, L. R., Teng, E., Stefanacci, L., Buffalo, E. A., and Clark, R. E.
2000. Impaired recognition memory in monkeys after damage limited to the hippo-
campal region. J. Neurosci. 20:451–63.
9. Squire, L. R. 2004. Memory systems of the brain: A brief history and current perspec-
tive. Neurobiol. Learn. Mem. 82:171–7.
10. Goldman-Rakic, P. S., Lidow, M. S., Smiley, J. F., and Williams, M. S. 1992. The
anatomy of dopamine in monkey and human prefrontal cortex. J. Neural. Transm.
Suppl. 36:163–77.
11. Evans, H. L. 1990. Nonhuman primates in behavioral toxicology: Issues of validity,
ethics and public health. Neurotoxicol. Teratol. 12:531–6.
12. Weed, J. L., Lane, M. A., Roth, G. S., Speer, D. L., and Ingram, D. K. 1997. Activ-
ity measures in rhesus monkeys on long-term calorie restriction. Physiol. Behav.
62:97–103.
13. Taffe, M. A. 2004. Effects of parametric feeding manipulations on behavioral perfor-
mance in macaques. Physiol. Behav. 81:59–70.
14. Pugh, T. D., Klopp, R. G., and Weindruch, R. 1999. Controlling caloric consumption:
Protocols for rodents and rhesus monkeys. Neurobiol. Aging 20:157–65.
15. Lane, M. A., Black, A., Handy, A., Tilmont, E. M., Ingram, D. K., and Roth, G. S. 2001.
Caloric restriction in primates. Ann. NY Acad. Sci. 928:287–95.
16. Roth, G. S., Ingram, D. K., and Lane, M. A. 2001. Caloric restriction in primates and
relevance to humans. Ann. NY Acad. Sci. 928:305–15.
17. Mattison, J. A., Black, A., Huck, J., et al. 2005. Age-related decline in caloric intake and
motivation for food in rhesus monkeys. Neurobiol. Aging 26:1117–27.
18. Dellinger, J. A. 1991. Pharmacologic challenges for establishing interspecies extrapola-
tion models in neurotoxicology. Neurosci. Biobehav. Rev. 15:21–3.
19. Paule, M. G. 1990. Use of the NCTR Operant Test Battery in nonhuman primates.
Neurotoxicol. Teratol. 12:413–8.
20. Fray, P. J., and Robbins, T. W. 1996. CANTAB battery: Proposed utility in neurotoxi-
cology. Neurotoxicol. Teratol. 18:499–504.

Working Memory 263
21. Weed, M. R., Taffe, M. A., Polis, I., et al. 1999. Performance norms for a rhesus monkey
neuropsychological testing battery: Acquisition and long-term performance. Brain Res.
Cogn. Brain Res. 8:185–201.
22. Friedman, H. R., and Goldman-Rakic, P. S. 1988. Activation of the hippocampus and
dentate gyrus by working-memory: A 2-deoxyglucose study of behaving rhesus mon-
keys. J. Neurosci. 8:4693–706.
23. Goldman-Rakic, P. S. 1987. Development of cortical circuitry and cognitive function.
Child Dev. 58:601–22.
24. Arnsten, A. F., Cai, J. X., and Goldman-Rakic, P. S. 1988. The alpha-2 adrenergic ago-
nist guanfacine improves memory in aged monkeys without sedative or hypotensive
side effects: Evidence for alpha-2 receptor subtypes. J. Neurosci. 8:4287–98.
25. Franowicz, J. S., and Arnsten, A. F. 1998. The alpha-2a noradrenergic agonist, guanfa-
cine, improves delayed response performance in young adult rhesus monkeys. Psycho-
pharmacology (Berl.) 136:8–14.
26. Levy, R., Friedman, H. R., Davachi, L., and Goldman-Rakic, P. S. 1997. Differential
activation of the caudate nucleus in primates performing spatial and nonspatial work-
ing memory tasks. J. Neurosci. 17:3870–82.
27. Gaffan, D., and Weiskrantz, L. 1980. Recency effects and lesion effects in delayed non-
matching to randomly baited samples by monkeys. Brain Res. 196:373–86.
28. Paule, M. G., Bushnell, P. J., Maurissen, J. P., et al. 1998. Symposium overview: The use
of delayed matching-to-sample procedures in studies of short-term memory in animals
and humans. Neurotoxicol. Teratol. 20:493–502.
29. Mishkin, M., and Manning, F. J. 1978. Non-spatial memory after selective prefrontal
lesions in monkeys. Brain Res. 143:313–23.
30. Buccafusco, J. J. 2001. Methods of Behavior Analysis in Neuroscience. Boca Raton,
FL: Taylor and Francis.
31. Murray, E. A., Bussey, T. J., Hampton, R. R., and Saksida, L. M. 2000. The parahip-
pocampal region and object identification. Ann. NY Acad. Sci. 911:166–74.
32. Meunier, M., Bachevalier, J., and Mishkin, M. 1997. Effects of orbital frontal and ante-
rior cingulate lesions on object and spatial memory in rhesus monkeys. Neuropsycho-
logia 35:999–1015.
33. Kowalska, D. M., Bachevalier, J., and Mishkin, M. 1991. The role of the inferior pre-
frontal convexity in performance of delayed nonmatching-to-sample. Neuropsycholo-
gia 29:583–600.
34. Buffalo, E. A., Ramus, S. J., Clark, R. E., Teng, E., Squire, L. R., and Zola, S. M. 1999.
Dissociation between the effects of damage to perirhinal cortex and area TE. Learn.
Mem. 6:572–99.
35. Beason-Held, L. L., Rosene, D. L., Killiany, R. J., and Moss, M. B. 1999. Hippocampal
formation lesions produce memory impairment in the rhesus monkey. Hippocampus
9:562–74.
36. Stern, C. E., Sherman, S. J., Kirchhoff, B. A., and Hasselmo, M. E. 2001. Medial tem-
poral and prefrontal contributions to working memory tasks with novel and familiar
stimuli. Hippocampus 11:337–46.
37. Frederick, D. L., Gillam, M. P., Lensing, S., and Paule, M. G. 1997. Acute effects of
LSD on rhesus monkey operant test battery performance. Pharmacol. Biochem. Behav.
57:633–41.
38. Hampson, R. E., and Deadwyler, S. A. 1999. Cannabinoids, hippocampal function and
memory. Life Sci. 65:715–23.
39. Buccafusco, J. J., Terry, A. V. Jr., Decker, M. W., and Gopalakrishnan, M. 2007. Profile
of nicotinic acetylcholine receptor agonists ABT-594 and A-582941, with differential
subtype selectivity, on delayed matching accuracy by young monkeys. Biochem. Phar-
macol. 74:1202–11.

40. Paule, M. G. 2005. Chronic drug exposures during development in nonhuman pri-
mates: Models of brain dysfunction in humans. Front. Biosci. 10:2240–9.
41. Zola, S. M., and Squire, L. R. 2001. Relationship between magnitude of damage to the
hippocampus and impaired recognition memory in monkeys. Hippocampus 11:92–8.
42. Burbacher, T. M., and Grant, K. S. 2000. Methods for studying nonhuman primates in
neurobehavioral toxicology and teratology. Neurotoxicol. Teratol. 22:475–86.
43. Rice, D. C. 2000. Parallels between attention deficit hyperactivity disorder and behav-
ioral deficits produced by neurotoxic exposure in monkeys. Environ. Health Perspect.
108,suppl 3:405–8.
44. Coghill, D. R., Rhodes, S. M., and Matthews, K. 2007. The neuropsychological effects
of chronic methylphenidate on drug-naive boys with attention-deficit/hyperactivity dis-
order. Biol. Psychiatry 62(9):954–62.
45. Chelonis, J. J., Edwards, M. C., Schulz, E. G., et al. 2002. Stimulant medication
improves recognition memory in children diagnosed with attention-deficit/hyperactiv-
ity disorder. Exp. Clin. Psychopharmacol. 10:400–7.
46. McCarten, J. R., Kovera, C., Maddox, M. K., and Cleary, J. P. 1995. Triazolam in
Alzheimer’s disease: Pilot study on sleep and memory effects. Pharmacol. Biochem.
Behav. 52:447–52.
47. Arnsten, A. F., and Goldman-Rakic, P. S. 1985. Catecholamines and cognitive decline
in aged nonhuman primates. Ann. NY Acad. Sci. 444:218–34.
48. Jackson, W. J., and Buccafusco, J. J. 1991. Clonidine enhances delayed matching-
to-sample performance by young and aged monkeys. Pharmacol. Biochem. Behav.
39:79–84.
49. Decker, M. W., Bannon, A. W., Curzon, P., et al. 1997. ABT-089 [2-methyl-3-(2-(S)-pyr
rolidinylmethoxy)pyridine dihydrochloride]: II. A novel cholinergic channel modulator
with effects on cognitive performance in rats and monkeys. J. Pharmacol. Exp. Ther.
283:247–58.
50. Prendergast, M. A., Terry, A. V. Jr., Jackson, W. J., et al. 1997. Improvement in accuracy
of delayed recall in aged and non-aged, mature monkeys after intramuscular or trans-
dermal administration of the CNS nicotinic receptor agonist ABT-418. Psychopharma-
cology (Berl.) 130:276–84.
51. Jackson, W. J., Buccafusco, J. J., Terry, A. V., Turk, D. J., and Rush, D. K. 1995. Velna-
crine maleate improves delayed matching performance by aged monkeys. Psychophar-
macology (Berl.) 119:391–8.
52. Prendergast, M. A., Jackson, W. J., Terry, A. V. Jr., Decker, M. W., Arneric, S. P.,
and Buccafusco, J. J. 1998. Central nicotinic receptor agonists ABT-418, ABT-089,
and (-)-nicotine reduce distractibility in adult monkeys. Psychopharmacology (Berl.)
136:50–8.
53. Terry, A. V. Jr., Buccafusco, J. J., Prendergast, M. A., et al. 1996. The 5-HT3 receptor
antagonist, RS-56812, enhances delayed matching performance in monkeys. Neurore-
port 8:49–54.
54. Terry, A. V. Jr., Buccafusco, J. J., Jackson, W. J., et al. 1998. Enhanced delayed match-
ing performance in younger and older macaques administered the 5-HT4 receptor ago-
nist, RS 17017. Psychopharmacology (Berl.) 135:407–15.
55. Schulze, G. E., McMillan, D. E., Bailey, J. R., et al. 1988. Acute effects of delta-9-
tetrahydrocannabinol in rhesus monkeys as measured by performance in a battery of
complex operant tests. J. Pharmacol. Exp. Ther. 245:178–86.
56. Schulze, G. E., McMillan, D. E., Bailey, J. R., et al. Acute effects of marijuana smoke
on complex operant behavior in rhesus monkeys. Life Sci. 45:465–75.
57. Schulze, G. E., and Paule, M. G. 1991. Effects of morphine sulfate on operant behavior
in rhesus monkeys. Pharmacol. Biochem. Behav. 38:77–83.

Working Memory 265
58. Schulze, G. E., and Paule, M. G. 1990. Acute effects of d-amphetamine in a monkey
operant behavioral test battery. Pharmacol. Biochem. Behav. 35:759–65.
59. Ramos, B. P., Colgan, L., Nou, E., Ovadia, S., Wilson, S. R., and Arnsten, A. F. 2005.
The beta-1 adrenergic antagonist, betaxolol, improves working memory performance
in rats and monkeys. Biol. Psychiatry 58:894–900.
60. Murphy, B. L., Arnsten, A. F., Goldman-Rakic, P. S., and Roth, R. H. 1996. Increased
dopamine turnover in the prefrontal cortex impairs spatial working memory perfor-
mance in rats and monkeys. Proc. Natl. Acad. Sci. USA 93:1325–9.
61. Terry, A. V. Jr., Buccafusco, J. J., and Bartoszyk, G. D. 2005. Selective serotonin 5-
HT2A receptor antagonist EMD 281014 improves delayed matching performance in
young and aged rhesus monkeys. Psychopharmacology (Berl.) 179:725–32.
62. Taffe, M. A., Weed, M. R., and Gold, L. H. 1999. Scopolamine alters rhesus monkey
performance on a novel neuropsychological test battery. Cogn. Brain Res. 8:203–12.
63. Ogura, H., and Aigner, T. G. 1993. MK-801 impairs recognition memory in rhesus
monkeys: Comparison with cholinergic drugs. J. Pharmacol. Exp. Ther. 266:60–4.
64. Angeli, S. J., Murray, E. A., and Mishkin, M. 1993. Hippocampectomized monkeys can
remember one place but not two. Neuropsychologia 31:1021–30.
65. Baron, S. P., and Wenger, G. R. 2001. Effects of drugs of abuse on response accuracy
and bias under a delayed matching-to-sample procedure in squirrel monkeys. Behav.
66. Paule, M. G., Chelonis, J. J., Buffalo, E. A., Blake, D. J., and Casey, P. H. 1999. Oper-
ant test battery performance in children: Correlation with IQ. Neurotoxicol. Teratol.
21:223–30.

13 Spatial Navigation
(Water Maze) Tasks
Alvin V. Terry Jr.
CONTENTS
13.1 Introduction................................................................................................. 267
13.2 Standard Procedures ................................................................................... 269
13.2.1 Methodology .................................................................................... 270
13.2.1.1 Testing Apparatus............................................................... 270
13.2.1.2 Hidden Platform Test.......................................................... 271
13.2.1.3 Transfer Test (Probe Trials)................................................ 272
13.2.1.4 Visible Platform Tests ........................................................ 272
13.2.1.5 Relearning Phases .............................................................. 273
13.2.2 Statistical Analyses.......................................................................... 273
13.2.2.1 Hidden Platform Test.......................................................... 273
13.2.2.2 Transfer Test (Probe Trials)................................................ 273
13.2.2.3 Visible Platform Test .......................................................... 274
13.2.2.4 Relearning Phases .............................................................. 274
13.2.3 Representative Data ......................................................................... 274
13.3 Alternative Procedures................................................................................ 277
13.3.1 Place Recall Test.............................................................................. 277
13.3.2 Platform Discrimination Procedures ............................................... 277
13.3.3 Working Memory Procedures.......................................................... 278
13.3.4 Extinction......................................................................................... 278
13.4 Summary and Conclusions ......................................................................... 278
References.............................................................................................................. 279
13.1 INTRODUCTION
Since the early part of the 20th century, a variety of experimental procedures have
been developed for animals that employ the escape from water as a means to motivate
learning and memory processes.1–4 Water maze tasks primarily designed to measure
spatial learning and recall have become quite useful for evaluating the effects of
aging, experimental lesions, and drug effects, especially in rodents. For more than
25 years the Morris water maze (MWM)5 has been the task most extensively used
and accepted by behavioral physiologists and pharmacologists. A cursory literature
search revealed that well over 2500 journal articles have been published since 1982
in which this model (or variations of the model) was used to assess and compare
spatial learning and memory in rodents.
267

The MWM, while simple at first glance, is a challenging task for rodents that
employs a variety of sophisticated mnemonic processes. These processes encompass
the acquisition and spatial localization of relevant visual cues that are subsequently
processed, consolidated, retained, and then retrieved in order to successfully navi-
gate and thereby locate a hidden platform to escape the water5 (see also review6).
The general processes used for “visuospatial navigation” in rats also contribute
considerably to human day-to-day cognitive processes. Importantly, several lines of
evidence confirm the utility of the model for investigations relevant to the study of
neurodegenerative and neuropsychiatric illnesses where cognition is impaired (e.g.,
Alzheimer’s disease, Parkinson’s disease, schizophrenia). While one would readily
acknowledge the differences in complexity between human and rodent behaviors,
several salient observations regarding the utility of the MWM are notable: (1) The
functional integrity of forebrain cholinergic systems, which are essential for efficient
performance of the MWM, appears to be consistently disrupted in patients who
suffer from AD. This disruption correlates well with the degree of dementia (see
reviews7,8) and is also present in many PD patients who suffer cognitive decline.9,10
(2) Cortical and hippocampal projections from the nucleus basalis magnocellularis
(NBM) and medial septum (MS), respectively, are reproducibly devastated in AD
(reviewed7) and accordingly, reductions in central cholinergic activity in rodents
resulting from brain lesions (e.g., NBM, MS, etc.) and age reproducibly impair spa-
tial learning in the MWM (reviewed6). (3) Other data implicate the hippocampus as
an essential structure for place learning,11which, incidentally, is commonly atrophic
in patients with AD.12,13 It is interesting to note that the hippocampal formation (in
particular the hippocampal-dentate complex and the adjacent entorhinal cortex),
which undergoes significant degeneration with age (and particularly so in the set-
ting of dementia), is believed to be intimately involved in cognitive mapping and
the facilitation of context-dependent behavior in a changing spatio-temporal setting
(reviewed14). Evidence to support this premise is now available from living humans
where computerized “virtual water maze tasks” have been shown to be highly sensi-
tive to hippocampal dysfunction. For example, in a virtual analogue of the classic
MWM hidden platform task (with a three-dimensional pool), patients with unilateral
hippocampal resections were severely impaired in their performance relative to age-
matched controls and age-matched patients who had extra-hippocampal resections.15
(4) Anticholinergic agents (e.g., scopolamine), which are routinely used to impair
performance in the MWM, also impair memory in humans and worsen the dementia
in those with AD16 (see also review17). (5) Finally, it is also important to note that
spatial orientation, navigation, learning, and recall (which are used extensively in the
MWM) are quite commonly disrupted in patients with dementia. Visuospatial and
visuoperceptual deficits and topographic disorientation are detectable very early in
the course of AD and become more pronounced as the disease progresses.18–20 The
common observations of spatial and visual agnosia in AD patients also indicate the
disruption of complex processes that involve both visual pathways and mnemonic
processing.21,22
The MWM procedure offers a number of advantages as a means of assessing
cognitive function in rodents when compared to others methods: (1) It requires no
pretraining period and can be accomplished in a short period of time with a relatively

Spatial Navigation (Water Maze) Tasks 269
modest number of animals. For example, young adult, unimpaired (control) rats can
accomplish the most commonly employed versions of the task with asymptotic lev-
els of performance achieved in 10–20 trials, generally requiring no more than a
few days of testing. (2) Through the use of “training” as well as “probe” or “trans-
fer” trials, learning as well as retrieval processes (spatial bias)5 can be analyzed and
compared between groups. (3) The confounding nature of olfactory trails or cues is
eliminated. (4) Through the use of video tracking devices and the measure of swim
speeds, non-mnemonic behaviors or strategies (i.e., taxon, praxis, thygmotaxis, etc.)
can be delineated and motoric or motivational deficits can be identified. (5) Visible
platform tests can identify gross visual deficits that might confound interpretation of
results obtained from standard MWM testing. (6) By changing the platform location,
both learning and relearning experiments can be accomplished. Accordingly, several
doses of experimental drugs can be tested in the same group of animals. (7) While
immersion into water may be somewhat unpleasant, more aversive procedures such
as food deprivation or exposure to electric shock are circumvented. (8) Through the
use of curtains, partitions, etc., operation of the video tracking system by the experi-
menter out of site of the test subjects also reduces distraction. (9) Finally, the MWM
is quite easy to set up in a relatively small laboratory, is comparatively less expensive
to operate than many types of behavioral tasks, and is easy to master by research
and technical personnel. We have found the method quite useful in drug discovery
and development studies for screening compounds for potential cognitive enhancing
effects,23 as well as delineating deleterious effects of neurotoxicants on cognition.24
For a more extensive discussion of the various MWM procedures and their advan-
tages, see Morris5 and reviews.6,25,26
13.2 STANDARD PROCEDURES

The MWM generally consists of a large circular pool of water maintained at room
temperature (or slightly above) with a fixed platform hidden just below (i.e., ~ 1.0 cm)
the surface of the water. The platform is rendered invisible by one of several means:
(1) adding an agent (i.e., powdered milk or a nontoxic dye or food coloring agent)
to render the water opaque; (2) having a clear Plexiglas platform in clear water; (3)
or having the platform painted the same color as the pool wall and floor (e.g., black
on black). Rats are tested individually and placed into various quadrants of the pool
and the time elapsed and/or the distance traversed to reach the hidden platform is
recorded. Various objects or geometric images (e.g., circles, squares, triangles) are
often placed in the testing room or hung on the wall so that the rats can use these
visual cues as a means of navigating in the maze. With each subsequent entry into
the maze the rats progressively become more efficient at locating the platform, thus
escaping the water by learning the location of the platform relative to the distal
visual cues. The learning curves are thus compared between groups. An illustration
of a typical MWM setup (as used in our laboratory) appears in Figure 13.1. The inset
at the top right illustrates typical learning behavior (under vehicle control conditions)
on day 1 of a two trial per day × 6 day hidden platform task when test subjects search
throughout the pool before locating the escape platform. The inset at the middle right
illustrates the search behavior on day 6 when the subject has learned the task and

Day 1
Water Maze Testing
Camera Light
Visual Day 6
Cue
Morris Water Maze

Hidden
Platform
Day 7
Probe Trials
FIGURE 13.1 Diagrammatic illustration of the Morris water maze testing room and apparatus.
can easily locate the platform in a matter of seconds. Finally, the inset on the bottom
right illustrates the clear bias for the previous platform location during the probe trial
on day 7 of testing after the escape platform has been removed.
13.2.1 METHODOLOGY
13.2.1.1 Testing Apparatus
1. Maze testing should be conducted in a large circular pool (e.g., rats, diameter:
180 cm, height: 76 cm; mice, diameter: 100–120 cm, height: 76 cm) made of
plastic (e.g., Bonar Plastics, Noonan, Georgia, USA) and painted black.
2. Fill the pool to a depth of 35 cm of water (maintained at 25°C + 1.0°C) to
cover an invisible (black) 10-cm square platform. The platform should be
submerged approximately 1.0 cm below the surface of the water and placed
in the center of the northeast quadrant.
Note: We have found that using a black platform in a pool with the sides
and floor painted black obviates the need for addition of agents to render the
water opaque. If the experimenter is unsure whether or not the platform is
still visible, closing the curtains to eliminate spatial cues and subsequently
testing a few rats will resolve this question. While rats can use egocentric
cues to eventually acquire the location of the platform, they will not rapidly
(or efficiently) become more successful with each entry into the pool if the

platform is invisible and room lighting is diffuse (i.e., when most of the
allocentric cues are eliminated).
3. The pool should be located in a large room with a number of extramaze
visual cues, including highly visible (reflective) geometric images (squares,
triangles, circles, etc.) hung on the wall, diffuse lighting, and black curtains
to hide the experimenter and the awaiting rats. Swimming activity of each
rat may be monitored via a television camera mounted overhead, which
relays information including latency to find the platform, total distance trav-
eled, time and distance spent in each quadrant, etc., to a video tracking sys-
tem. Tracking may be accomplished via a white rat on a black background.
Note: We have found the Noldus EthoVision¥ system (Leesburg, Virginia,
USA) to be a very reliable system that is also easy to set up. Several other
vendors market similar systems (e.g., San Diego Instruments, Columbus
Instruments).
13.2.1.2 Hidden Platform Test

We commonly employ a method in which each rat is given two trials per day for six
consecutive days.
1. Each day, a trial is initiated by placing each rat in the water facing the pool
wall in one of the four quadrants (designated NE, NW, SE, SW), which are
set up on the computer software so that each quadrant is equal in area. The
daily order of entry into individual quadrants is randomized so that all four
quadrants are used once every two days.
Note: Do not place the rat in adjacent quadrants sequentially since the rat
may adopt a positional or other non-mnemonic strategy (e.g., all right turns)
to locate the platform. Further, the order should be changed on each subse-
quent day of testing.
2. For each trial, the rat is allowed to swim a maximum of 90 sec to find the
hidden platform. When successful, the rat is allowed a 30-sec rest period on
the platform (timed manually with a stopwatch). If unsuccessful within the
allotted time period, the rat is given a score of 90 sec and is then physically
placed on the platform and allowed the 30-sec rest period. In either case the
rat is immediately given the next trial (inter-trial interval = 30 sec) after the
rest period.
Note: In some cases the rat may fall or jump off of the platform and resume
swimming before the elapsed 30-sec interval. When this occurs, the stop-
watch should be immediately stopped and the rat retrieved and placed on
the platform again. The stopwatch should be reactivated so that the remain-
der of the time interval (30 sec) is enforced. This assures that each rat has
equal time to observe spatial cues after each trial.

13.2.1.3 Transfer Test (Probe Trials)

On day 7 (i.e., 24 hr following the last hidden platform trial) a probe trial is con-
ducted in which the platform is removed from the pool to measure spatial bias for
the previous platform location.5 This is accomplished by measuring the percentage
of time spent (and distance swam) in the previous target quadrant as well as the
number of crossings over the previous platform location. These assessments provide
a second estimate of the strength and accuracy of the memory of the previous plat-
form location.
1. Place each rat in the pool and track the animal for 90 sec. This may be repeated
one time (if necessary), since in some cases an unusual level of variance in
performance will be observed in this first trial. It is assumed that some of the
rats are in some way disoriented after the change in testing conditions.
Note: More than two trials may result in “extinction” effects (see section
“Alternative Procedures” below) with less time spent in the target quadrant,
and is thus undesirable for a measure of spatial bias.
2. The time elapsed and distance swam in the previous target quadrant is
recorded. An annulus ring can be circumscribed around the previous target
location (on the computer screen) to localize it more closely. The number of
crossings through this region may be recorded. Alternatively, crossings of
the actual 10-cm square platform target outlined in the previous trials can
be recorded and compared between groups.
13.2.1.4 Visible Platform Tests

A visible platform test may be performed to determine if a drug or other experimen-
tal manipulation results in crude alterations in visual acuity that might confound
the analyses of data that depend on the use of distal visual cues for task perfor-
mance. One must be aware, however, of certain behaviors that might be interpreted
as impaired visual acuity. For example, the absence of search behaviors or thymgot-
axis (swimming constantly along the perimeter of the pool) might be misinterpreted
as visual deficits since the animal does not locate the platform in a reasonable period
of time. Thus, animals must make attempts to cross the pool and then be impaired at
locating the platform in order for an interpretation of visual deficits to be made.
1. Immediately following the probe trial on day 7, place the platform into the
pool in the quadrant located diametrically opposite the original position
(SW quadrant).
2. A cover (available from San Diego Instruments, Inc. and other vendors),
which is rendered highly visible (i.e., with light-reflective glossy or neon
paint), is attached to the platform to raise the surface above the water level
(approximately 1.5 cm).
3. Room lighting may be changed so that the extramaze cues are no longer
available and the visible platform is more highly illuminated.

Note: The video tracker is not necessary for this procedure and only a
stopwatch is needed.
4. Allow each rat one trial in order to acclimate to the new set of conditions
and locate the platform visually. This is accomplished by lowering the rat
into the water in the NE quadrant and allowing the rat to locate the plat-
form. No time limit is placed on this first trial. Once the platform is located,
allow the rat 30 sec on the platform. The rat should then immediately be
given a second trial in the same manner and the latency to find the platform
measured as a comparison of visual acuity.
Note: This procedure may be repeated additional times; however, the plat-
form location should be changed on each subsequent trial to ensure that
visual location of the platform is actually made from a distance and the rat
is not first using nearby stationary visual cues.
13.2.1.5 Relearning Phases

After completion of the first seven days of water maze testing and a rest period
(generally at least 1 wk and often longer), a second series of trials (phase 2) may be
conducted as described above (hidden platform test section), except that the location
of the platform is changed to a different quadrant. Daily performances (average of
two trials/day/rat) are then compared as described above. This method may be used
in order to compare different drug doses or other additional manipulations with the
same groups of animals.
Note: It must be realized that learning curves will generally be steeper than in the
first phase of testing since a number of factors not associated with the actual plat-
form location will have been previously learned (e.g., use of visual cues to navigate,
the fact that escape is not associated with the pool wall, etc.).
13.2.2 STATISTICAL ANALYSES

13.2.2.1 Hidden Platform Test
For the hidden platform test, we generally average the latencies and the distances
swam across the two trials for each rat each day. These means are then analyzed
across the six days of testing. A two-way repeated measures analysis of variance
(ANOVA) is used for main effects (i.e., group or treatment comparisons) with day as
the repeated measure and latency or distance swam as the dependent variable. The
Student Newman-Keuls test is used for post-hoc analyses.
13.2.2.2 Transfer Test (Probe Trials)

For probe trials the means are compared between groups via a one-way ANOVA and
again, the Student Newman-Keuls test is used for post-hoc analyses.

13.2.2.3 Visible Platform Test

For the visible platform test, the means are also compared between groups via a one-
way ANOVA and the Student Newman-Keuls test is used for post-hoc analyses.
13.2.2.4 Relearning Phases

The relearning phase is analyzed identically to the hidden platform test described
above.
13.2.3 REPRESENTATIVE DATA

Several representative MWM hidden platform studies under vehicle control condi-
tions from our laboratory appear in Figure 13.2. The acquisition curves for rats given
one trial per day for 14 consecutive days, two trials per day for six consecutive days,
or four trials per day for four consecutive days are illustrated. We have used each of
these methods in our laboratory in previous studies. Using each of these approaches,
the rats learned to locate the hidden platform with progressively shorter latencies
over the course of the study. We have found that young vehicle-treated test subjects
(i.e., not age impaired or impaired by an amnestic drugs) given only one trial per
day are somewhat less efficient at learning the task, but more sensitive to pro-cog-
nitive agents (e.g., nicotine, see reference27) than subjects given multiple trials per
day. Further, we have found the two- and four-trial-per-day methods to be useful for
amnestic-reversal studies (see references28,29).
An MWM study conduced in our laboratory in young (3–4 mo) and aged (22–24
mo) male Fisher 344 rats is presented in Figure 13.3. Figure 13.3A illustrates the
efficiency of each experimental group to locate a hidden platform in a water maze
task on 10 consecutive days of testing (two trials per day). As expected, under saline
conditions the young rats learned to locate the hidden platform with progressively
shorter latencies until the end of the study, while the aged rats administered saline
were less efficient. For the latency comparisons, there was a highly significant main
effect (p < 0.001), a significant trial effect (p < 0.001), and a significant group × trial
interaction (p < 0.01). Post-hoc analyses indicated that performance by the young
animals was superior to that of the aged animals across a number of days of testing.
Further, the acetylcholinersterase inhibitor (and commonly prescribed AD therapeu-
tic agent) donepezil (0.75 mg/kg) (in aged rats) was associated with superior perfor-
mance over aged saline controls across several days of testing. In addition, all rats
treated with donepezil reached a near-asymptotic level of performance (i.e., latencies
less that 20 sec) by day 10 of testing, whereas this was not the case for aged rats
administered saline.
Figures 13.3B and 13.3C illustrate the performance of probe trials by the various
test groups. As noted above, these experiments are performed after hidden platform
tests (in this case 48 hr later) and reflect a “spatial bias” of animals toward the pre-
vious location of the hidden platform. The results are analyzed separately from the
hidden platform tests and offer a second, easily performed method of estimating
the strength and accuracy of the original learning process.6 It is important to note
that since the pool is divided into four quadrants of equal area, a chance level of

100 2500
Distance to Platform (sec)

Latency to Platform (sec)
80 2000
60 1500
40 1000
20 500
0 0
2 4 6 8 10 12 14 2 4 6 8 10 12 14
Session Session
(a)
60 1400
50 1200
1000
40
800
30
600
20
400
10 200
0 0
1 2 3 4 5 6 1 2 3 4 5 6
Session Session
(b)
60 1400
1200
50
1000
40
800
30
600
20
400
10 200
0 0
1 2 3 4 1 2 3 4
Session Session
(c)
FIGURE 13.2 Illustration of acquisition curves for several versions of the Morris water
maze hidden platform test in young adult Wistar rats. The latency in seconds (left) and the
distance swam in centimeters (right) to find the hidden platform are presented for each study.
(a) A one-trial-per-day procedure conducted for 14 consecutive days; (b) A two-trial-per-day
procedure conducted for six consecutive days; (c) A four-trial-per-day procedure conducted
for four consecutive days. Each point represents the mean ± SEM, N = 10–12 rats per group.

60
% Time in Target Quadrant

70 50 +
40
*
60 30
20
50
Latency Platform (sec)
10
40 0
Young Aged Aged
Vehicle Vehicle Donepezil
30
(b)
20 6
Platform Area Crossings

Aged Vehicle 5
10
Aged Donepezil
Young Vehicle 4
+
0 3
1 2 3 4 5 6 7
*
Day of Testing 2
(a) 1
0
Young Aged Aged
Vehicle Vehicle Donepezil
(c)
FIGURE 13.3 Effects of age and the Alzheimer’s disease therapeutic agent, donepezil, on
performance of the MWM. (a) Hidden platform test. Each point represents the mean latency
in sec ± SEM of two trials per day for 10 consecutive days of testing. (b and c) Probe trials.
The percent of total time spent (in sec) in the previous target quadrant and the number of
crossings over the previous platform area, respectively, 48 hr after the last hidden platform
trial. Each bar represents the mean ± SEM. +=*=significantly different from vehicle-treated
young rats significantly different from vehicle-treated aged rats (one-way ANOVA, p < 0.05,
Student Neuman-Keuls post-hoc test), N = 14–22 rats per group.
performance would mean that the percent of time or distance swam (of the total) in
the previous target quadrant would generally approximate 25%. As indicated in both
Figures 13.3B and 13.3C, there were statistically significant (group-related) effects
on performance (i.e., percent of time spent in the previous target quadrant and cross-
ings over the previous platform area, respectively). Namely, aged vehicle-treated rats
demonstrated less spatial bias than young vehicle-treated subjects, and donepezil
partially reversed this impairment in aged subjects.
In summary, these data support the argument that the MWM (as conducted in
our laboratory) is sensitive to the impairing effects of aging on spatial navigation,
learning, and recall, and further, that the procedure is sensitive to the positive effects
of a well-known pro-cognitive agent (i.e., a positive control).

13.3 ALTERNATIVE PROCEDURES

A number of variations of the water maze tasks described above have been employed
for the study of memory processes in rats and a full review of these procedures in
beyond the scope of this chapter. A short summary of a few of these procedures is
outlined below, however. For a more detailed overview of these and additional water
maze procedures see Morris5 and reviews.6,25
13.3.1 PLACE RECALL TEST

In this procedure, hidden platform tests are first performed as described above in
intact animals so that the location of the platform is well learned. Subsequently,
the rats are experimentally manipulated (i.e., given brain lesions, drugs, or other
physiological manipulations, etc.) and then retested with either additional hidden
platform tests or probe trials. Thus, the effects of the experimental manipulations
on all processes used to solve the task, with the exception of learning and memory
formation, may be studied. Namely, processes such as memory retrieval and spatial
bias, as well as motoric, sensory, and motivational effects of the manipulations may
be delineated.
13.3.2 PLATFORM DISCRIMINATION PROCEDURES

These methods require rats to discriminate between two visible platforms in order
to successfully escape the water (Figure 13.4, left). One of the platforms is rigid and
Water Maze Variations
Day 1 Day 3
Day 4 Day 2
Discrimination Working Memory

Tasks Tasks
FIGURE 13.4 Illustration of platform discrimination and working memory procedures in

the Morris water maze.

able to sustain the weight of the rat, while the other platform is floating (often made
of styrofoam) and not able to sustain the rat’s weight. Both spatial and nonspatial
versions of this task have been used. In the spatial version of the task, the platforms
appear identical (visually) and rats are required to discern the viable platform by
learning its location relative to distal visual cues in the room. In the nonspatial ver-
sion of the task, the rats learn to visually discriminate between two platforms of
different appearance. For example, discrimination between platforms may be engen-
dered via a difference in shape, brightness, or painted pattern. Curtains are drawn to
exclude the influence of extramaze cues.
13.3.3 WORKING MEMORY PROCEDURES

Working memory procedures in the MWM (sometimes referred to as spatial “match-
ing to sample” procedures) generally involve a two-trials-per-day paradigm in which
a hidden platform is located in one of four quadrants and randomly relocated on each
of several subsequent days of testing (Figure 13.4, right). The assumption drawn is
that each rat will obtain information regarding the location of the platform on the
first trial, which will be of benefit for discerning its position on trial two. The ITI can
be manipulated in order to alter the difficulty of the task.
13.3.4 EXTINCTION
While not commonly used for this purpose, the behavioral process known as extinc-
tion can also be assessed in the MWM. Extinction occurs when the relations among
stimuli recognized during acquisition are no longer valid and the previously estab-
lished responses are suppressed. Accordingly, preferences for a spatial location
decrease in the water maze as the animal learns that the cues no longer predict
the location of the hidden platform.30 Extinction in this context is considered a type
of cognitive flexibility, a form of fluid intelligence that encompasses the ability to
inhibit strong response preferences in order to explore alternative solution paths.31
In contrast to the more common MWM studies where acquisition (hidden platform
testing) and retention (probe trials) are the focus, in extinction experiments subjects
are first trained in the hidden platform test to an asymptotic level of performance
(defined as a latency to find the hidden platform of less than 10–15 sec for four con-
secutive trials). We have found in unpublished studies that 10–12 days (two trials per
day) is more than sufficient to reach this asymptotic level in young vehicle-control
subjects. Subsequently (i.e., on the following day after the last hidden platform test),
four or more consecutive probe trials are conducted to assess the subject’s ability to
decrease (i.e., extinguish) its spatial bias for the previous platform location.
13.4 SUMMARY AND CONCLUSIONS

The MWM equipped with a video tracking system has become a commonly used
and well-accepted behavioral task for rodents. It is quite easy to set up in a relatively
small laboratory, is comparatively less expensive to operate than many types of
behavioral tasks, and is easy to master by research and technical personnel. It uses a
number of mnemonic processes in rats that are relevant to the study of human learn-

ing and memory and disorders thereof. In addition, it is a very versatile paradigm,
which can be used to study both spatial and nonspatial (discriminative) learning,
as well as working memory processes and extinction, and offers several means of
delineating and dissociating confounding non-mnemonic processes.
REFERENCES
1. Glaser, OC. 1910. The formation of habits at high speed. J. Comp. Neurol., 20,
165–184.
2. Wever, EG. 1932. Water temperature as an incentive to swimming activity in the rat. J.
Comp Psychol, 14, 219–224.
3. Waller, MB, Waller, PF, and Brewster, LA. 1960. A water maze for use in studies of
drive and learning. Psychol Rep, 7, 99–102.
4. Woods, PJ, Davidson, EH and Peters, R.J., 1964. Instrumental escape conditioning in
a water tank: effects of variation in drive stimulus intensity and reinforcement magni-
tude. J. Comp. Psychol, 57, 466–470.
5. Morris, RGM. 1984. Development of a water-maze procedure for studying spatial
learning in the rat. J Neurosci Meth 11, 47–60.
6. McNamara RK, and Skelton RW, 1993. The neuropharmacological and neurochemical
basis of place learning in the Morris water maze. Brain Res Rev. 18, 33–49.
7. Perry E, Walker M, Grace J, Perry R. 1999. Acetylcholine in mind: a neurotransmitter
correlate of consciousness? Trends Neurosci, 22, 273–280.
8. Francis PT, Palmer AM, Snape M, Wilcock GK. 1999. The cholinergic hypothesis of
Alzheimer’s disease: a review of progress. J Neurol Neurosurg Psychiatry 66,137–147.
9. Whitehouse PJ, Hedreen JC, White CL 3d, Price DL. 1983. Basal forebrain neurons in
the dementia of Parkinson disease. Ann Neurol 13, 243–248.
10. Perry EK, Curtis M, Dick DJ, Candy JM, Atack JR, Bloxham CA, Blessed G, Fairbairn
A, Tomlinson BE, Perry RH. 1985. Cholinergic correlates of cognitive impairment in
Parkinson’s disease: comparisons with Alzheimer’s disease. J Neurol Neurosurg Psy-
chiatry 48, 413–421.
11. Sunderland T, Tariot PN, Newhouse PA. 1988. Differential responsivity of mood,
behavior, and cognition to cholinergic agents in elderly neuropsychiatric populations.
Brain Res 472, 371–389.
12. Ebert U, Kirch W. 1998. Scopolamine model of dementia: electroencephalogram find-
ings and cognitive performance. Eur J Clin Invest 28, 944–949.
13. McDonald RJ, and White NM. 1995. Hippocampal and nonhippocampal contributions
to place learning in rats. Behavi Neurosci; 109, 579–593.
14. Terry RD, Katzman R. Senile dementia of the Alzheimer type. Ann Neurol 14, 497-506,
1983.
15. Mann DM. 1991. The topographic distribution of brain atrophy in Alzheimer’s disease.
Acta Neuropathol (Berl) 83, 81–86.
16. Scheibel AB. 1979. The hippocampus: organizational patterns in health and senes-
cence. Mech Ageing Dev. 9, 89-102.
17. Eslinger PJ, Benton AL. 1983. Visuoperceptual performances in aging and dementia:
clinical and theoretical implications. J Clin Neuropsychol 5, 213–220.
18. Huber SJ, Shuttleworth EC, Freidenberg DL. 1989. Neuropsychological differences
between the dementias of Alzheimer’s and Parkinson’s diseases. Arch Neurol 46,
1287–1291.
19. Morris JC, McKeel DW Jr, Storandt M, Rubin EH, Price JL, Grant EA, Ball MJ, Berg
L. 1991. Very mild Alzheimer’s disease: informant-based clinical, psychometric, and
pathologic distinction from normal aging. Neurology 41, 469–478.

20. Henderson VW, Mack W, Williams BW. 1989. Spatial disorientation in Alzheimer’s
disease. Arch Neurol 46, 391–394.
21. Mendez MF, Tomsak RL, Remler B. 1990. Disorders of the visual system in Alzheim-
er’s disease. J Clin Neuroophthalmol 10, 62–69.
22. Terry, A.V., Jr., M. Gattu, M., Buccafusco, J.J., J.W. Sowell, J.W., and Kosh, J.W. 1999.
Ranitidine Analog, JWS-USC-75IX, Enhances Memory-Related Task Performance in
Rats. Drug Dev Res, 47, 97–106.
23. Prendergast, M.A., Terry, A.V., Jr., and Buccafusco, J.J. 1997. Chronic, low-level expo-
sure to diisopropylfluorophosphate causes protracted impairment of spatial navigation
learning. Psychopharmacol, 129, 183–191.
24. Brandeis R, Brandys Y, Yehuda S. L. 1989. The use of the Morris Water Maze in the
study of memory and learning. Int J Neurosci 48, 29–69.
25. Astur RS, Taylor LB, Mamelak AN, Philpott L, Sutherland RJ. 2002. Humans with
hippocampus damage display severe spatial memory impairments in a virtual Morris
water task. Behav Brain Res 132:77-84.
26. Vorhees CV, Williams MT. 2006. Morris water maze: procedures for assessing spatial
and related forms of learning and memory. Nat Protoc 1: 848–58.
27. Lattal KM, Abel T. 2001. Different requirements for protein synthesis in acqui-
sition and extinction of spatial preferences and context-evoked fear. J Neurosci
21(15):5773–5780.
28. Beversdorf DQ, Hughes JD, Steinberg BA, Lewis LD, Heilman KM. 1999. Nor-
adrenergic modulation of cognitive flexibility in problem solving. Neuroreport
10(13):2763–2767.
29. Hernandez, C.M, and Terry, A.V., Jr. 2005. Repeated Nicotine Exposure in Rats: Effects
on Memory Function, Cholinergic Markers and Nerve Growth Factor. Neuroscience
130:997–1012.
30. Terry, A.V., Jr. 2001. “Spatial Navigation (Water Maze) Tasks” in J. J. Buccafusco
(Ed.) Behavioral Methods in Neuroscience. CRC Press: Boca Raton, Chapter 10, pages
153–166.
31. Terry, AV., Jr., Parikh V, Gearhart DA, Pillai, Hohnadel EJ, Warner, S, Nasrallah, HA,
and Mahadik SP. 2006. A Time Dependent Effects of Haloperidol and Ziprasidone on
Nerve Growth Factor, Cholinergic Neurons, and Spatial Learning in Rats. Journal of
Pharmacology and Experimental Therapeutics 318:709–724.

14 Water Maze
Tasks in Mice
Special Reference to
Alzheimer’s Transgenic Mice
Dave Morgan
CONTENTS
14.1 Introduction................................................................................................. 281
14.2 Methods....................................................................................................... 283
14.2.1 Animal Subjects............................................................................... 283
14.2.2 Equipment ........................................................................................284
14.2.3 Working Memory Procedure ...........................................................284
14.2.4 Reference Memory Procedure ......................................................... 285
14.2.5 Visible Platform in an Open Pool .................................................... 286
Acknowledgments.................................................................................................. 289
References.............................................................................................................. 289
14.1 INTRODUCTION
Water maze tasks have been used for over a quarter century in testing rodent spatial
navigation memory.1 Although initially developed for rats, they have also been useful in
evaluating memory in mice, often using scaled down pool sizes. The major advantage
of the water maze tasks over dry mazes is increased motivation to escape, and hence
more rapid performance within the maze. Typical trials in water mazes are limited to
60 sec, while dry maze trials often last much longer. This permits higher throughput
and increased efficiency when large numbers of animals require evaluation.
The original Morris maze used an open pool with a hidden platform just below
the water level midway between the pool wall and the center of the pool. The rodent
is placed in the pool, typically facing the wall, in one of four arbitrarily defined
quadrants, and permitted to explore the pool. Extramaze cues surround the pool
to orient the rodent as it navigates within the pool. Initially animals stumble upon
the platform, climb onto it and are forced to remain for a short period before being
removed to consolidate the experience. They are then removed from the platform and
placed into a different quadrant from the first trial and again given the opportunity
281

to explore the pool. Once again the platform is encountered and the animal escapes
the pool by climbing onto the platform. Often a third and in some versions up to
six trials are performed. The time spent prior to finding the platform is recorded as
“latency to escape” and is averaged for each day’s performance. The procedure is
repeated over 2–14 days with the platform remaining in the same location each day,
making this a reference memory task.
After 3–10 days of training, the rodents are administered a “probe” trial (actually
an extinction trial) in which the platform is removed and memory for platform loca-
tion assessed. A number of measurements have been proposed to infer the strength
of the “memory” of the mouse for the platform location. The simplest is time spent
in the “target” quadrant (i.e., the one that previously contained the platform). More
elaborate measurements include the average swim distance from the previous plat-
form location or the number of crossings over the exact location of the platform.
Many of these measurements involve videotaping of the rodent’s performance and
application of computerized software to analyze the performance. However, given
that the mice must still be shuttled manually into the pool and off the platform (and
rescued if drowning), there is no meaningful personnel efficiency achieved by the use
of computerized analysis (the behaviorist must remain at the pool for each rodent).
There are many variations on these procedures. Some have used intermediate
probe trials after days 3, 6, and 9, for example, and used comparisons of these per-
formances as an index of rate of learning.2 Others have converted the normal refer-
ence memory version of the water maze to a working memory model by measuring
the number of trials to reach a latency criterion at one location and then measuring
trials to criterion at a new platform location.3 In general, this open pool Morris water
maze approach is useful in discriminating memory dysfunction in amyloid precur-
sor protein (APP) transgenic mice.4–8 In our own work with the water maze task,
we found in some cases that mice were impaired when measuring latency, but not
on the probe trial.9 In the Barnes maze, these mice showed no significant deficits.
However, when the same mice were tested on the radial arm water maze, the APP
transgenic mice were significantly impaired. Although we originally included both
the open pool Morris water maze and the radial arm water maze (and Barnes maze)
as components of a 6-wk behavioral test battery,10 we have now abbreviated this to
a 2-wk battery in which a shortened version of the radial arm water maze is the pri-
mary cognitive task.11
The radial arm water maze involves the imposition of a radial arm maze onto a
pool. This approach was first developed for rats.12,13 This is implemented by inser-
tion of triangular wedges into the pool that reach above the surface of the water,
forming swim alleys surrounding a central open region (Figure 14.1). The platform
is placed within one of the alleys (goal arm) and the mouse is started in one of the
other swim arms. Although our initial work focused on a working memory version
of this task,14,15 we have found a reference memory variant of the radial arm water
maze that can consistently reveal deficits in transgenic mouse memory performance
with as few as 2 days of training, increasing the flexibility of scheduling behavioral
testing. This latter adaptation, developed in consultation with David Diamond, takes
advantage of optimal spacing of trials to minimize the time needed for acquisition
of the task. Moreover, the measurement of errors does not require use of video cam-

Water Maze Tasks in Mice 283
FIGURE 14.1 The radial arm water maze. Shown are the pool used for behavioral testing
with metal inserts/dividers in place forming swim alleys and a central swim area. The plat-
form shown is the visible platform that protrudes just slightly above the water and is striped for
salience. Also note the visual cues outside the pool consisting of two shower curtains (visible),
a plain wall (to the right outside of frame), and the remainder of the room (behind camera).
eras or computers to obtain reliable data regarding performance. Depending on the

pool size, up to 16 arms can been used and the contingencies organized to separately
detect reference versus working memory errors.16 The design flexibility of the dry
radial arm maze can be combined with the motivational advantages of the water
maze format.
14.2 METHODS
Our work has focused exclusively on transgenic mouse models of amyloid deposition
or tau pathology. Our APP mice have been derived from the Tg2576 line4 bred with
a mutant PS1 line 5.117 as described by Holcomb et al.18 As a result these mice are of
a mixed genetic background. It should be noted that many inbred mouse lines carry
a retinal degeneration gene mutation19,20 that does not impair performance on many
murine behavioral tests (including the visual cliff 21), but does cause severe deficits in
spatial navigation tasks.22 The JAX laboratories Web site indicates whether a given
strain is known to possess one of these rd mutations and provides primer pairs to
detect the most prevalent rd1 mutation in individual mice when genotyping. It is
essential that mouse lines either be investigated for presence of a background strain
carrying an rd mutation, or that individual mice be tested for this mutation. Our
Tg2576 line carried the rd mutation via inclusion of the SJL strain. However, geno-
typing and selective breeding have eliminated this mutation from the background of
our APP animals.
Mice should be generally healthy and free of open wounds that might become
infected by exposure to water. Thus, we do not test mice within 7 days of a surgical
procedure. For the APP-only mice, we can detect behavioral deficits in modest-sized

cohorts (6–8) around 12–15 mo of age. For many studies we prefer older mice (20–
24 mo) as this more closely resembles conditions of aged Alzheimer’s disease (AD)
patients. We always include a cohort of untreated nontransgenic littermate mice in
any drug/therapy trials to act as a positive control (if the nontransgenic mice fail to
learn, there is a problem with the behavioral testing procedure).
14.2.2 EQUIPMENT
Pool size is not an essential variable, but should be constrained for practical reasons
(e.g., ability of the experimenter to reach all parts of the pool). For mice we have
used a 1 m pool that is 30 cm deep. We constructed pie-shaped wedges out of a sheet
of stainless steel (plastic or sheet aluminum may also be suitable) that was 24 cm
wide and 60 cm long. These were then bent at the center of the long axis to form a
60° angle. They were placed into the pool to form a “V” 24 cm high with a vertex 30
cm from the edge of the pool. By equally spacing six of these inserts into the pool,
we form six swim alleys of 30 cm in length with a 40-cm wide central region (Fig-
ure 14.1). The pool is then filled with water to a depth of 14 cm (10 cm from the top
of the inserts) at a temperature of 20.5°C. A platform should be placed in one swim
alley just below the water line. We use inverted terra cotta pots (10 cm diameter) that
are painted the same color as the inside of the pool and positioned 1 cm beneath the
surface of the water. A pool liner may be used instead of paint to achieve a uniform
color that can be replaced instead of cleaned and repainted (black works very nicely).
We do not find it necessary to add paint or milk to the water to increase opacity.
14.2.3 WORKING MEMORY PROCEDURE

Prior to cognitive testing mice are administered a small neurological test battery
consisting of wire hang and balance beam (day 1), Y-maze testing (day 2), and 3 days
of accelerating rotorod testing. Mice failing to perform adequately in these tasks are
removed from the cognitive testing group. All of these tasks are administered by
the same individual who will perform the cognitive testing. Mice appear somewhat
more sensitive to changes in experimenter than rats are, suggesting this induces a
stress response that interferes with normal cognitive abilities. We have found that
simply transporting mice up and down an elevator impairs performance on these
cognitive function tasks.
The general radial arm water maze procedure involves placing the mouse into
one arm of the maze other than the goal arm, and releasing the mouse to begin
swimming. Most mice swim readily and explore the maze. When a mouse enters a
swim arm other than the goal arm, the mouse is charged with one error (a mouse is
considered to enter the arm when all four limbs move into the swim alley). Occasion-
ally, mice stop swimming and float, or they swim in the central regions without mak-
ing an arm entry. For each 15-sec period a mouse fails to enter an arm for whatever
reason it is charged one error. In this manner, a mouse failing to swim accumulates
four errors, which is a score typical of mice that have not learned the platform loca-
tion. Mice that consistently fail to swim are removed from the study. The trial con-
tinues for 60 sec or until the mouse ascends the platform. If a mouse does not locate
the platform within 60 sec, it is guided to the platform. The mouse is removed after

15 sec on the platform and either started in another trial, or dried with a towel and
placed in its home cage (with a heat lamp available in one corner). Both error number
and latency to find the platform are recorded.
In order to test working memory, the goal arm location within the maze was
changed each day. Within each day, a mouse was given four consecutive 60-sec
acquisition trials, followed 30 min later by a fifth (retention) trial. The next day, the
platform was moved to a new location, and the mouse had to learn the new platform
location. The rationale for the 30-min delay on the retention trial was that short-term
forgetting is common in AD patients. It was hoped that the mice would learn the new
platform location during the first four trials when the inter-trial interval was 15 sec
(registration of the material to be learned), and then demonstrate poor performance
at the 30 min time point (the recall point in testing for memory deficits clinically).
Thus far we have not found mice that learned location by trial 4 but failed to remem-
ber on trial 5, as we had hoped might occur. Instead we find that APP transgenic
mice fail to improve over the acquisition trials, and, as expected, perform poorly on
the retention trial as well.
One of the limitations to this procedure was that mice were slow to acquire
the procedural aspects of the testing (understanding there was a platform and that
the platform moved each day). This may be a result of the ethologically unlikely
possibility that escape location in a natural environment would change daily. As a
general criterion, we felt that when the mice as a group reached a criterion of one or
fewer errors on trials 4 and 5, they had learned the task. On some occasions, non-
transgenic cohorts would reach this criterion within 10 days of continuous testing.
Other cohorts could require 15 days to reach this learning criterion. Thus, in order
to maintain consistent performance, the same investigator must be available for a
period of up to 2 wk for 3–4 hr at the same time of day. Although it is conceivable
that training could be suspended for the weekend, we never fully investigated this
variable. Instead, to simplify the testing procedure we opted to examine the refer-
ence memory version of the maze described below.
14.2.4 REFERENCE MEMORY PROCEDURE

For reference memory testing we began by running mice for 15 trials on each of 2
days. The goal arm was constant for these 2 days, and the mouse waas placed pseu-
dorandomly (no repeats) in a different start arm for each trial. Moreover, the trials on
day 1 alternated between using a visible platform above the water and a hidden plat-
form in the same arm. On day 2, all trials used the hidden platform. Moreover, the
goal arm for each mouse was different (to minimize possible effects of odor trails).
This was predetermined on a score sheet that the experimenter used to determine the
start arm and goal arm for each trial for each mouse (examples of these score sheets
are available from the author).
One problem with 15 trials per day is that older mice can get fatigued from
swimming for this long a period without rest. Thus we designed a testing schedule
whereby the mice had a rest period after each trial. Mice were assigned to groups of
four (treatment conditions should be equally distributed in these testing groups). Typ-
ically two groups of four mice were tested in parallel. First, mouse 1 of group 1 was

administered trial 1, then mouse 2 of group 1, mouse 3, and mouse 4. After mouse 4
of group 1 was tested on trial 1, mouse 1 of group 1 was administered trial 2. Mice
2–4 of that group then followed. After each trial, mice were returned to their home
cages with a heat lamp available in the corner (one lamp served all four cages).
After all mice in group 1 received six trials, a second group of four mice were
administered their first six trials in the same fashion as group 1. First, mouse 1, trial
1, then mouse 2, trial 1, etc. After all four mice were administered six trials, the first
group of four mice was administered trials 7–12. Then group 2 was administered
trials 7–12. Finally, group 1 was administered trials 13–15, and then group 2 was
administered trials 13–15. The entire process can be accomplished in 3–4 hr. This
permits a second series of eight mice to be tested on the same day.
On day 2 the entire process was repeated, except the platform was hidden for
all trials. We have never fully investigated whether the alternation of visible and
hidden platforms on day 1 is essential for good learning to occur, thus this may be
considered optional.
For most cohorts of mice, this resulted in average scores of one error or less for
the “positive” control groups (usually nontransgenic mice). On some occasions the
control mice may not have reached this criterion (for example old mice or occasion-
ally some inbred lines). In these circumstances we ran a reversal trial on day 3 (a
new goal arm location for each mouse not adjacent to the initial goal arm; all trials
used the hidden platform). This also sometimes revealed deficits in performing the
reversal task in treatment groups that were not easily distinguished in the first 2 days
of testing. If performance is still poor after the first day of reversal testing, a second
day of reversal testing can be performed.
The 2-day reference memory version of the radial arm water maze is the most
efficient method we have found for testing in this procedure. Most cohorts of control
mice learn the task within 30 trials. For the working memory version, 50–75 tri-
als are necessary for the mice to demonstrate solid learning of platform location.
Similar numbers apply to the Morris open pool version of the water maze. We feel
that this procedure optimally spaces trials so that mice have some immediate recol-
lection of the events and a rest period so that fatigue is not a factor, and that longer
rest periods during testing permit some consolidation to occur within the day, rather
than between days.
14.2.5 VISIBLE PLATFORM IN AN OPEN POOL

Irrespective of whether reversal training is performed, the last day of testing used a
visible platform in an open pool (inserts removed). The visible platform has ensigns
located above the water while the platform remains slightly below the water. The
mouse was started in the same location for each trial, and the location of the visible
platform was moved for each mouse. Latency to reach the platform was recorded.
Mice were shuttled just as on day 1 of radial arm maze testing (mouse 1, trial 1; mouse
2, trial 1; etc.). Fifteen trials were performed. The purpose of the visible platform task
is to assess if mice have the performance skills necessary for the water maze tasks.
Mice failing to reach a criterion of 20 sec latency for the last three trials of the visible
platform task were considered to be impaired. These mice may have been blind or


FIGURE 14.2 Typical data for the working memory version of the radial arm water maze.
Fifteen-mo-old nontransgenic or APP+PS1 transgenic mice were given five trials daily as
described in the text. Four trials were continuous (solid line) and the fifth trial was adminis-
tered after a 30-min delay (dashed line). Data are presented as the number of errors for each
trial averaged over 3-day blocks. The learning criterion of one error is shown as the dashed
horizontal line across the graph. *P < 0.01.
impaired motorically and thus not capable of being evaluated for cognitive function.
Very few mice completing the testing protocol fail to meet this criterion.

Figure 14.2 shows results from the working memory version of the radial arm water
maze. Shown are three blocks of results summed over 3 days each. The first block
represents data from days 1–3, the second block from days 4–6, and the third block
from days 7–9. On the first trial of each block, mice are performing at chance levels
as the platform is in a new location. In all cases, the transgenic mice show only mod-
est (and not significant) improvement in their performance over the five trials (the
last trial being a retention trial). However, by the last block of trials (days 7–9) we
found that the nontransgenic mice reached the criterion of less than one error (on the
retention trial 5). At this point there was also a significant difference between the
transgenic and nontransgenic mice.
Figure 14.3 shows the results from the 2-day reference memory version of the
water maze. Note that on day 1 both groups improve slightly in their performance.
However, by day 2 the nontransgenic mice are able to find the platform with few
errors. The performance of the nontransgenic mice is significantly better than the
transgenic mice on blocks 6–10 of the radial arm water maze task.

One of the issues regarding the radial arm maze results is that individual mouse data
is noisy. Some mice simply by fortune find the platform on their first arm entry of the

App 25 mo Non Tg 25 mo
8
7
6
5
Errors
4
3
2
1
0
1 2 3 4 5 6 7 8 9 10
Day 1 Day 2
Blcok Number
FIGURE 14.3 Typical data for the 2-day reference version of the radial arm water maze.
Data shown are for 25-mo-old nontransgenic (NonTg) and untreated APP Tg 2576 derived
mice (APP). Data presented are averages of three trial blocks. The horizontal dashed line rep-
resents the one error per trial learning criterion. On day 2, nontransgenic mice perform sig-
nificantly better than transgenic mice for all trial blocks (P < 0.01). Source: Data are redrawn
from Wilcock, D. M., Rojiani, A., Rosenthal, A., et al. 2004. Passive immunotherapy against
Abeta in aged APP-transgenic mice reverses cognitive deficits and depletes parenchymal
amyloid deposits in spite of increased vascular amyloid and microhemorrhage. J. Neuroin-
flammation 1:24, with permission.
first day. As a result, the data are portrayed most favorably with some degree of sum-
mation over trials and/or days. If large sample sizes are used (> 25 mice), it would
seem plausible to include data for every trial when presenting the data. However, in
most studies using transgenic or aged mice, numbers are a limiting resource. While
we aim for a sample size of 10 in most experimental designs, we sometimes are
limited to final sample sizes as low as five after attrition because of death, or culling
because of motoric deficiencies (rare) or skin lesions (more common in older mice).
There are many ways of accomplishing this averaging for presentation purposes. In
Figure 14.2, we collapse over 3-day blocks in the working memory version of the
maze, as each trial reflects a different stage of learning. In most of our published
studies with this method,9,10,14,15,21,23,24 we averaged the final 2–3 days of testing after
the nontransgenic mice had reached the learning criterion. We typically discarded
the first 5–10 days of training from data analysis as this was the time when the
mice were still acquiring procedural aspects of the task. However, this is not the
only method to demonstrate clear differences in transgenic mice. The group led by
Ottavio Arancio averages all days for each of the five trials, and still observes clear
differences caused by the presence of amyloid.25,26 Thus, there can be several alter-
native means of presenting these results. Although we collect the latency data, we
feel latencies are affected by variables other than mnemonic functions (swim speed)
and reporting significant effects in latency when there were no differences in errors
seems deceptive to us.

For the 2-day water maze, we average over blocks of three trials for each mouse.
These three trial blocks become 10 data points used for statistical analysis. For all
studies (working and reference), we first perform a repeated measures analysis of
variance (ANOVA) to seek a main effect of genotype and trials. We then perform
post-hoc means comparisons using Fischer’s LSD test with the statistical program
Statview (SAS) to identify group differences on specific trials or blocks.
A final comment on statistical analysis regards experiments testing for drug or
other therapeutic effects in mouse models of neurodegeneration. We emphasize in
these studies that the inclusion of the positive control group (in our case, nontrans-
genic mice) is simply to validate the success of the behavioral testing process. How-
ever, we do not include these mice in the statistical analysis to determine the effect
of the therapeutic modality. These analyses should directly compare the treated and
untreated disease model mice, without reference to the nontransgenic data. Only if
the treated and untreated transgenic groups differ is there truly an effect of the treat-
ment. We have witnessed and reviewed manuscripts that show a significant difference
between untreated disease model mice (transgenic) and positive control (nontrans-
genic) mice, but fail to reach significance in comparing treated transgenic mice and
nontransgenic mice. The authors sometimes attempt to conclude (erroneously) that
there was then an effect of their treatment. This violates a cardinal rule of statistics
that failure to detect a difference does not mean there is no difference, only that
the study was unable to detect it. Statistically significant differences in performance
between treated and untreated disease model groups are essential to argue for benefits
of the therapy.
ACKNOWLEDGMENTS
We thank David Diamond and Gary Arendash for years of assistance in develop-
ing these methods and collecting data relevant to this technique. We thank Jennifer
Alamed, our laboratory behaviorist, for collecting data and aiding in the figures for
the manuscript. DGM is supported by the following awards from the National Insti-
tutes of Health: AG04418, AG15490, AG18478, AG 25509, AG25711, and NS48355,
and is a supervisor for AG031291.
REFERENCES
1. Morris, R. G., Garrud, P., Rawlins, J. N., and O’Keefe, J. 1982. Place navigation
impaired in rats with hippocampal lesions. Nature 297:681–83.
2. Westerman, M. A., Cooper-Blacketer, D., Mariash, et al. 2002. The relationship between
Abeta and memory in the Tg2576 mouse model of Alzheimer’s disease. J. Neurosci.
22:1858–67.
3. Chen, G., Chen, K. S., Knox, J., et al. 2000. A learning deficit related to age and beta-
amyloid plaques in a mouse model of Alzheimer’s disease. Nature 408:975–79.
4. Hsiao, K., Chapman, P., Nilsen, S., et al. 1996. Correlative memory deficits, Abeta
elevation, and amyloid plaques in transgenic mice. Science 274:99–102.
5. Puolivali, J., Wang, J., Heikkinen, T., et al. 2002. Hippocampal A beta 42 levels cor-
relate with spatial memory deficit in APP and PS1 double transgenic mice. Neurobiol.
Dis. 9:339–47.

6. Palop, J. J., Jones, B., Kekonius, L., et al. 2003. Neuronal depletion of calcium-depen-
dent proteins in the dentate gyrus is tightly linked to Alzheimer’s disease-related cog-
nitive deficits. Proc. Nat. Acad. Sci. USA 100:9572–77.
7. Kelly, P. H., Bondolfi, L., Hunziker, D., et al. 2003. Progressive age-related impairment
of cognitive behavior in APP23 transgenic mice. Neurobiol. Aging 24:365–78.
8. Jankowsky, J. L., Melnikova, T., Fadale, D. J., et al. 2005. Environmental enrichment
mitigates cognitive deficits in a mouse model of Alzheimer’s disease. J. Neurosci.
25:5217–24.
9. Arendash, G. W., King, D. L., Gordon, M. N., et al. 2001. Progressive behavioral
impariments in transgenic mice carrying both mutant APP and PS1 transgenes. Brain
Res. 891:45–53.
10. Austin, L., Arendash, G. W., Gordon, M. N., et al. 2003. Short-term beta-amyloid vac-
cinations do not improve cognitive performance in cognitively impaired APP + PS1
mice. Behav. Neurosci. 117:478–84.
11. Wilcock, D. M., Rojiani, A., Rosenthal, A., et al. 2004. Passive immunotherapy against
Abeta in aged APP-transgenic mice reverses cognitive deficits and depletes parenchy-
mal amyloid deposits in spite of increased vascular amyloid and microhemorrhage. J.
Neuroinflammation 1:24.
12. Diamond, D. M., Park, C. R., Heman, K. L., and Rose, G. M. 1999. Exposing rats to a
predator impairs spatial working memory in the radial arm water maze. Hippocampus
9:542–51.
13. Bimonte, H. A., and Denenberg, V. H. 1999. Estradiol facilitates performance as work-
ing memory load increases. Psychoneuroendocrinology 24:161–73.
408:982–85.
15. Gordon, M. N., King, D. L., Diamond, D. M., et al. 2001. Correlation between cog-
nitive deficits and Aß deposits in transgenic APP+PS1 mice. Neurobiology of Aging
22:377–85.
16. Bimonte, H. A., Hyde, L. A., Hoplight, B. J., and Denenberg, V. H. 2000. In two spe-
cies, females exhibit superior working memory and inferior reference memory on the
water radial-arm maze. Physiol. Behav. 70:311–17.
17. Duff, K., Eckman, C., Zehr, C., et al. 1996. Increased amyloid-beta42(43) in brains of
mice expressing mutant presenilin 1. Nature 383:710–13.
18. Holcomb, L., Gordon, M. N., McGowan, E., et al. 1998. Accelerated Alzheimer-type
phenotype in transgenic mice carrying both mutant amyloid precursor protein and pre-
senilin 1 transgenes. Nat. Med. 4:97–100.
19. Chang, B., Hawes, N. L., Hurd, R. E., Davisson, M. T., Nusinowitz, S., and Heckenliv-
ely, J. R. 2002. Retinal degeneration mutants in the mouse. Vision Res. 42:517–25.
20. Clapcote, S. J., Lazar, N. L., Bechard, A. R., Wood, G. A., and Roder, J. C. 2005. NIH
Swiss and Black Swiss mice have retinal degeneration and performance deficits in cog-
nitive tests. Comp. Med. 55:310–16.
21. Garcia, M. F., Gordon, M. N., Hutton, M., et al. 2004. The retinal degeneration (rd)
gene seriously impairs spatial cognitive performance in normal and Alzheimer’s trans-
genic mice. NeuroReport 15:73–77.
22. Alamed, J., Wilcock, D. M., Diamond, D. M., Gordon, M. N., and Morgan, D. 2006.
Two-day radial-arm water maze learning and memory task: Robust resolution of amy-
loid-related memory deficits in transgenic mice. Nat. Protoc. 1:1671–79.
23. Morgan, D. 2003. Learning and memory deficits in APP transgenic mouse models of
amyloid deposition. Neurochem. Res. 28:1029–34.

24. Joseph, J. A., Denisova, N. A., Arendash, G., et al. 2003. Blueberry supplementation
enhances signaling and prevents behavioral deficits in an Alzheimer disease model.
Nutr. Neurosci. 6:153–62.
25. Gong, B., Vitolo, O. V., Trinchese, F., Liu, S., Shelanski, M., and Arancio, O. 2004.
Persistent improvement in synaptic and cognitive functions in an Alzheimer mouse
model after rolipram treatment. J. Clin. Invest. 114:1624–34.
26. Trinchese, F., Liu, S., Battaglia, F., Walter, S., Mathews, P. M., and Arancio, O. 2004.
Progressive age-related development of Alzheimer-like pathology in APP/PS1 mice.
Ann. Neurol. 55:801–14.

15 Behavioral
of Zebrafish
Neuroscience
Edward D. Levin and Daniel T. Cerutti
CONTENTS
15.1 Introduction................................................................................................. 293
15.2 Use of Fish Models in Behavioral Neuroscience ........................................ 294
15.3 Procedures and Processes ........................................................................... 295
15.3.1 Assessment of Swimming Activity in Newly Hatched Zebrafish ... 297
15.3.2 Reflexes and Habituation ................................................................. 297
15.3.3 Pavlovian Conditioning....................................................................300
15.3.4 Operant Conditioning and Mazes....................................................300
15.3.5 Testing Anxiety and Stress Response.............................................. 303
15.4 Conclusions .................................................................................................306
Acknowledgement..................................................................................................306
References..............................................................................................................306
15.1 INTRODUCTION
Models are used to represent complex problems in simplified forms—physics, chem-
istry, and biology all make good use of models. The most familiar are the mathemati-
cal sorts that form the basis of natural science theory. In the life sciences, the concept
of modeling can extend further to include experimental procedures and nonhuman
subjects. For example, a neuroscientist might employ a rat running in a radial-arm
maze to study working memory processes, or a mouse in an open-field test to study
anxiety. The value of a model is primarily a function of its fidelity: in the case of a
theoretical model, fidelity is measured in terms of predicted findings; in the case of
biological models, the issue is couched in terms of validity. It is this second kind of
model that concerns us in this chapter on neuroscience methods, where the challenge
of model species is particularly acute because behavioral and brain processes are
both extraordinarily complex, and the problem is to find species that display both
interesting behavior and easily accessible neural processes.
Rats and mice, unquestionably the most successful models in neuroscience,
have been extremely effective in helping determine which mammalian brain regions
and neurotransmitter systems involved in cognition, learning, and other varieties of
behavioral function. But the invertebrate Aplysia, a marine mollusk, has also served
as a molecular model of memory processes.1 Such seemingly unrelated model spe-
cies are useful to the extent that they balance external validity, simplicity, and cost.
Most recently these considerations have led researchers in behavioral neuroscience
293

to use fish, a sort of middle ground between rodents and mollusks. In this chapter we
review progress in the behavioral neuroscience of the diminutive zebrafish (Danio
rerio), a species that has already firmly established itself as a model of vertebrate
development, and now opens new doors for the investigation of brain mechanisms.
Zebrafish are sometimes identified as an alternative model (relative to classic
rodent models), but the term complementary model might be more appropriate since
it addresses the use of fish in addition to classic mammalian models. Some ques-
tions, such as about the role of frontal cortical and hippocampal structures in learn-
ing and memory, cannot be studied with fish since these are not evident (but see2).
But other attributes of fish make them valuable models in behavioral neuroscience
research. Developmental processes can be continuously visualized in species that
have a clear chorion (egg sack). Reporter systems can highlight specific neural sys-
tems so that their proliferation, differentiation, migration, and projections can be
easily discerned. Reversible genetic suppression through the morpholino technique
can determine the importance of specific molecular mechanisms for neurodevelop-
ment. Numerous mutants available also help with the evaluation of molecular mech-
anisms throughout life. Finally, fish are easily bred in great numbers and develop
rapidly, reducing the cost of experimentation and significantly increasing research
throughput—potentially, more experiments can be run in less time to answer any
number of questions.
The merit of fish models is now a matter of record. Zebrafish, in particular, have
been well used in genetics, neuroscience, pharmacology, and toxicology (e.g., see3–7).
The next and ongoing step is to extend the zebrafish model to pursue questions of
behavioral neuroscience, an undertaking that requires valid, reliable, and efficient
methods of behavioral assessment.
15.2 USE OF FISH MODELS IN BEHAVIORAL NEUROSCIENCE

Fish are the obvious ancestral form of existing tetrapods, so it is not terribly surpris-
ing to find that they show most of the behavior seen in terrestrial species in some
form or other. In social behavior alone, there are species known to show monoga-
mous mating for life (e.g., angelfish8), individual recognition of conspecifics by sight
or odor,9 socially mediated learning,10 intricate mate-selection strategies,11 ritualized
displays of aggression,12 and communication of danger.13,14 With respect to cogni-
tion and adaptive behavior, fish show highly developed spatial navigation abilities,15
nonassociative learning such as habituation,16,17 precise timing abilities,18–20 Pavlov-
ian conditioning (e.g., see21), operant behavior motivated by aversive stimuli such as
shuttle box behavior,22 negatively reinforced avoidance,23 and food-reinforced lever
pressing positively reinforced responding.24 In terms of sensory processes, fish have
excellent color vision;25 some species generate and detect weak electrical currents,
a sense that they use to detect predators and prey;26 and have lateral-line organs that
allow them to resolve the location, size, and features of distal objects by sensing their
pressure shadows.
Behavioral research with fish began with ethologists and comparative psycholo-
gists asking questions about the evolution of learning, cognition,27–29 and brain func-
tion.30–33 As in other species, the understanding of the teleost brain has been driven

Behavioral Neuroscience of Zebrafish 295
in large part by the development of appropriate behavioral assays (e.g., see31). The
extent to which basic behavioral and brain processes in mammals and fish are analo-
gous remains an open question—there are clear similarities and differences—and, as
with all animal models, the validity of a fish model hinges on the particular question
being asked. Many species of fish have been used in models of cognitive impairment,
for example, the Japanese medaka (Oryzias latipes) is being used in toxicological
studies on effects of the insecticide diazinon (e.g., see34), and walleye (Stizostedion
vitreum) have been used to demonstrate the adverse impacts of insecticides on cho-
linergic systems.35 Goldfish (Carassius auratus) have historically been used to study
learning and memory processes.26,36,37 Fish have not been widely used in pharmacol-
ogy but there is no reason to believe that they would not be suitable (e.g., see3).
Zebrafish have rapidly become a prominent model for studying the molecular
basis of vertebrate neurodevelopment.4,38,39 The scientific potential of the zebrafish
was discovered by George Streisinger.40 The clear chorion of the zebrafish allows
continuous visualization of neuroanatomy; their rapid development and accessibil-
ity to genetic analysis make the zebrafish an excellent model system for molecular
and mechanistic studies of neurodevelopment. Since its introduction, many genetic
mutants have become available, including varieties that can help determine the
molecular mechanisms of neurobehavioral function. More recently, the availability
of morpholino techniques, whereby specific parts of the genome can be reversibly
suppressed during early development, provides a unique way to explore the molecular
biology of development. Zebrafish have been critical in the identification of a vari-
ety of genes affecting various aspects of neural development and function (e.g., see
partial list in41). As a result, the genetics and physiology of learning and memory are
now being more widely studied in zebrafish (e.g., see42). Many tasks are now able to
tap behavioral processes previously only studied with rodents and goldfish.3,25,43–48
15.3 PROCEDURES AND PROCESSES

It might seem a simple matter to develop a valid battery of behavioral tests to study
learning and cognition in fish, but it has not been so. One problem is translating
between terrestrial and aquatic habitats; another problem is finding reliable and
valid dependent measures; nor are theorists always in agreement about how to clas-
sify behavioral processes.49 A somewhat simpler question involves distinguishing
between procedures: those that involve stimulus presentations (e.g., to study reflexes
and fixed-action patterns), and those that involve arranging consequences for behav-
ior (e.g., to study instrumental or operant behavior). All other behavioral preparations
derive from these—Pavlovian conditioning involves signaling a stimulus presenta-
tion (e.g., a tone that signals a shock), and operant discrimination involves signaling
a consequence (e.g., a color that signals which arm of a maze contains food).
Further subtle variants of these basic procedures can answer any number of
questions about processes;50 for example, a rat’s visual contrast sensitivity can be
tested with great accuracy by arranging an operant discrimination between sin-wave
gratings and gray patches.51 Four cases of apparatuses adapted to study behavioral
processes in zebrafish are illustrated in Figure 15.1. At the top left (1) is an aquatic
version of the T-maze, an operant task used to study problems in discrimination

1. T-Maze 2. Escape a.
b.
Start
Favorable
Habitat Rotation
4. Place Preference
3. Bite Test d.
c.
g.
e.
f.
FIGURE 15.1 Four apparatuses that have been used to study learning and memory in the
zebrafish. (1) T-maze. The T-maze can be used to study a variety of questions in learning and
cognition including discrimination,25 and spatial and nonspatial navigation (e.g., with gold-
fish52). The version shown here was employed by Darland and Dowling3 with zebrafish in an
experiment in which the primary datum was latency to reach the favorable habitat. (2) Visual
escape. Li and Dowling3,53 used elicited escape from a moving stimulus (a) to study visual
function in zebrafish. In this apparatus, rotation of the dark band (a) surrounding the swim
area elicits defensive hiding behavior behind a central pole (b). (3) Exploratory biting. In this
procedure, a zebrafish is trained to enter the raised platform through the door (c) to explore a
small, submerged stimulus (d) such as a colored bead. Miklosi and Andrew17 employed this
apparatus to study lateralization in the zebrafish and found habituation of biting and explor-
atory behavior elicited by a bead. (4) Place preference. The place-preference procedure is
used to assess affinity for conditioned stimuli. In a typical procedure, a test space is divided
into two distinctive halves (e and f) with a partition between them (g); the subject is exposed
to an unconditioned stimulus in one half, and then later with the partition opened, is tested for
side preference. For example, Darland and Dowling3 found that zebrafish show a preference
for a conditioned stimulus previously paired with cocaine.
including attention, memory, and reinforcement. In the case of fish, the T-maze has
been used to study color discrimination in zebrafish,25 problems in navigation (e.g.,
in goldfish52), and effects of genetic manipulations on habitat selection as in the
maze shown here in which the dependent measure was latency to reach the favor-
able habitat.3 The top-right illustration (2) depicts a rotating drum apparatus that has
been used to study reflexive escape (a variant of the opto-kinetic reflex test). In this
test, a typical fish will flee the rotating band (a) by hiding behind the central pole
(b), a visually guided escape taxis.3,53 The lower-left illustration (3) depicts a setup
used to study novelty-elicited exploratory behavior. A fish placed in the tank will
visit the raised platform through the door (c) to explore a small, submerged stimulus
(d) such as a colored bead;17 exploration shows habituation to familiar stimuli and
dishabituation with the introduction of new stimuli. The lower-right illustration (4)

depicts an aquatic version of the place-preference procedure used to measure condi-

tioned appetitive stimuli. In this test, a subject is first exposed to an unconditioned
stimulus, e.g., cocaine,3 in one of two distinctive halves (e and f) of the tank, and then
later with the partition opened, it is given a preference test. In the sections that follow
we present these tests and others in greater detail, emphasizing behavioral processes
as much as procedures.
15.3.1 ASSESSMENT OF SWIMMING ACTIVITY IN NEWLY HATCHED ZEBRAFISH

It is important to determine the motor behavior function in young zebrafish for stud-
ies of development as well as for higher throughput tests of the adverse effects of
early toxicant exposure. Figure 15.2 (top inset) shows how a dissecting microscope
can be used to image the movement of newly hatched zebrafish. Either through
manual scoring of videotapes using a grid system or a computerized digital video
tracking system, the swimming activity of newly hatched zebrafish can be reliably
indexed. The lower graphs in Figure 15.2 shows the significant reduction in swim-
ming activity with 100 ng/mL of chlorpyrifos from fertilization through hatching on
day 5 when the behavioral test was conducted on day 6 or day 9.54
15.3.2 REFLEXES AND HABITUATION

Much research on the physiological mechanisms of zebrafish behavior has focused
on sensory-motor development (e.g., vision, swimming, and touch-elicited reflexes)
in larvae or young fish (review in55,56). The simplest investigations are those of the
tap-elicited startle reflex, the so called “C-start” response, which has been found
to show an increased latency with early alcohol exposure.57,58 The development of
touch-elicited escape behavior has been detailed by Granato et al.55: “Although the
embryo is resting most of the time, touching the tail tip induces a fast and straight
movement away from the stimulus source. In contrast, mechanical stimuli near the
head of the embryo induce a fast escape response, where the embryo turns 180°
along its horizontal body axis. At 96 hours the larva is freely swimming, changes
swimming directions spontaneously, and is able to direct its swimming towards tar-
gets” (p. 399).
We have recently examined habituation of tap-elicited swimming in unrestrained
zebrafish to evaluate the effects of toxicants and drugs on a nonassociative learning
process. Figure 15.3 (upper inset) illustrates a fully automated procedure using com-
mercially available video tracking software (Ethovision, Noldus, Inc., Wageningen,
The Netherlands). Fish were studied individually in a test battery consisting of eight
50-mm diameter tanks. The test arranged a “step up” transition in stimulus rates,
with 20 taps presented with an inter-stimulus interval of 10 sec, followed imme-
diately by 20 additional taps with an inter-stimulus interval of 20 sec. The graph
depicts the effect of scopolamine (fish were immersed for 5 min in 200 mg/L of
scopolamine prior to testing) on swim distance in the 5 sec after each tap (previously
unpublished data). Both control and scopolamine-treated fish showed habituation of
tap-elicited swimming, but only the control fish showed a reliable recovery in the
taps immediately following the “step up” transition, a finding in the scopolamine

%!")% +), )*""-.* +"(&),".*
,)'"-+% %"0
&))+
#+%!
)*%"0
"+4,$'+2)(% $&)+*2+%*$),1*),.+"
3" -,)(- $&%(#0%''%(# -%/%-2
2 2
*
*
"#'"(-+),,%(#,

(#' (#'
FIGURE 15.2 Swim test to evaluate motor function in newly hatched zebrafish. The left
side of the upper inset (A) shows a video microscope setup with five arenas; the right side (B)
shows a close-up of the cylindrical arena and grid pattern used to measure distance traveled
(segment crossings). The graphs show results of a study on chlorpyrifos effects on swimming
on newly hatched fish, control versus 100 ng/mL on day 6 (p < 0.01) and on day 9 (p < 0.05)
after fertilization.54
treated fish that is consistent with a selective disruption in short-term memory (for a
theoretical treatment of short- and long-term memory in habituation, see59).
Zebrafish show a highly developed visually guided escape reflex, which may be
related to the opto-kinetic response in other species,53 escaping a stimulus behind
a place of concealment. This concealment reflex may be analogous to the targeted

%!( $ &#(!

+"'( "&
/ $&-$#(!
+" &#
&&,
)'$!#$
%'$!#$' %#'"
$%$!"#0('$#$&(&""$&,

..' .. '
*&+"'(# '

$#(&$!..'
"$% ..'
$#(&$!.. '
"$% .. '

&!' %!$ '
FIGURE 15.3 Tap-elicited swim test used to study habituation in the zebrafish. The left
side of the upper inset (A) shows a horizontal array of eight arenas below a digital camera;
the right side shows the push-solenoid used to deliver sharp taps under the cylindrical arenas.
The dependent measure is swim distance in the 5 sec following taps, as measured by a video
tracking system. The graph compares habituation in 16 control fish and 16 fish following a 10-
min immersion in 200 mg/L of scopolamine on habituation. In this procedure, the fish were
exposed to 20 taps with an inter-stimulus interval (ISI) of 10 sec, immediately followed by 20
more taps with an ISI of 20 sec. The control fish, but not the scopolamine fish, show a brief
recovery at the transition (t-test, one tail, p < .025), suggesting that scopolamine selectively
disrupted short-term memory.
response concealment behavior described in mice by Blanchard,60 where if mice are

familiarized with a container containing a place of concealment, they flee directly
to that place when threatened. Figure 15.1(2) shows an apparatus developed to study
this visually guided escape reflex in zebrafish.3,53 Fish are tested by rotating the outer
cylinder of the apparatus, which contains a vertical black band (a), and observing the
subject’s orientation with respect to the band and a central cylinder (b) behind which

it can hide. The test can be adapted to test visual function61,62 and has been used to
measure visual contrast sensitivity of zebrafish.63
Exploratory behavior in novel environments has been used to assay anxiety in
rodent models (e.g., see64–66) and analogous procedures have entered the zebrafish
literature. Several experiments show that the fish first explores a stimulus predomi-
nantly using the right eye and subsequently approaches the stimulus favoring the
left eye.67 Figure 15.1(3) shows an apparatus employed by Miklosi and Andrew17 to
study lateralization of visual exploration in the zebrafish. Subjects are first trained
to visit a box suspended in their home tanks by entering a door (c) to eat. After they
reliably enter the box, a colored bead (d) is lowered into the water with the behavior
of the fish recorded on video. Right eye use and biting were highly probable the first
time a stimulus was presented and declined in probability in two subsequent trials,
demonstrating habituation (see also16,68–70).
15.3.3 PAVLOVIAN CONDITIONING

Zebrafish have shown Pavlovian learning in several experiments. Figure 15.1(4)
illustrates a “place-preference” task used by Darland and Dowling3 to screen zebraf-
ish for cocaine sensitivity (see also71). The apparatus consists of a tank divided into
two distinctive chambers by a screen. During training, the screen is sealed and a
zebrafish is exposed to cocaine in one of the chambers. In subsequent preference
tests, the fish showed an appetitive conditioning effect by approaching and staying in
the chamber in which they had previously received cocaine.
Several studies with zebrafish have used shuttle-box procedures in which they
learn to avoid an aversive conditioned stimulus by swimming to alternate sides of an
elongated tank.44,45,72 Among the earliest demonstrations of associative learning in
zebrafish used a shock deletion procedure to reinforce swimming away from a shock
signal.73 More recently, Pradel et al.74 used a shuttle box and shock avoidance to study
the role of cell adhesion molecules in memory consolidation. Technically speaking,
swimming in these procedures resembles an operant response—the Pavlovian pro-
cess involves the pairing of the conditioned stimulus with shock. However, without
additional investigation it is difficult to say whether the “avoidance” swimming is
elicited or operant in nature. This ambiguity also appears in appetitive procedures
such as in the food-reinforced T-maze discrimination tasks (e.g., see25).
Suboski and colleagues13,14 demonstrated Pavlovian conditioning of fear by pair-
ing morpholine and alarm substance (a chemical secreted by frightened or injured
fish) and subsequently showing conditioned fear to morpholine alone.14 The Pavlov-
ian nature of their learning was later confirmed by showing that the conditioned
alarm response could also be transferred between stimuli by sensory- and second-
order conditioning. The last finding in particular highlights the subtlety of learning
possible in this unassuming, diminutive fish.
15.3.4 OPERANT CONDITIONING AND MAZES

Although it might be assumed that this predominance of aversive procedures is
because aversive procedures are more rapid than appetitive procedures, there are
exceptions such as that shown by Williams et al.46 who trained fish to alternate

between two feeding sites in an average of 14 trials. The task is essentially an appe-
titive version of the shuttle box discussed above, except that trials are initiated by
the experimenter tapping on the center of the tank and 5 sec later dropping a small
amount of food in one end of the tank (the location of food is alternated between tri-
als); the dependent measure is the position of the fish immediately before the deliv-
ery of food. Carvan et al.57 used this task to show dose-dependent detrimental effects
of ethanol on learning and memory in zebrafish.
Perhaps the earliest example of behavioral research with zebrafish is a maze
learning study in which negotiation of a left-right-left-right maze and approach to
black or white stimuli was trained by eliciting an anode galvanotaxic reflex that elic-
ited approach to the target stimulus, a procedure that is somewhat difficult to imple-
ment.75 Colwill et al.25 recently trained color discrimination in zebrafish by placing
different colors at the end of each arm of a T-maze (green vs. purple and red vs.
blue) and feeding the fish only at one arm. These researchers unambiguously dem-
onstrated discrimination of color by arranging discrimination reversals (i.e., a cross-
over design) and experimenter-blind testing. In a similar T-maze apparatus as shown
in Figure 15.1(1), Darland and Dowling3 reinforced choice of one arm by providing
it with a goal box containing deep water, artificial grass, and marbles (however, the
dependent measure was the reduction in latency to reach the enriched arm, a result
that could be caused by habituation of fear in the novel maze apparatus).
The three-chamber maze shown in Figure 15.4 was developed by Arthur and
Levin43 to assess learning and memory in the zebrafish. The three-chamber maze
can be thought of as a simplification of the T-maze, but one in which aversive con-
sequences follow errors. The start area is the middle “start chamber,” and there are
vertically sliding doors on either side of this central start area leading to left and
right choice areas. At the outset of a trial the fish is placed in the start chamber and
allowed to move about for a brief period. In the choice phase, the vertical sliding
doors to the left- and right-choice chambers are opened and the fish is allowed time
to swim to one or the other; if it persists in the start chamber, a fish net is waved
in the chamber (a “threatening stimulus”) until it makes a choice. After making a
choice, both vertical sliding doors are closed. If the choice is correct (i.e., to the goal
side) the fish is permitted to swim for a short period of time; if the choice is incorrect
the sliding partition is moved to the “restricting position” for a short period of time.
This procedure is repeated for a fixed number of trials. Dependent measures in the
three-chamber shuttle maze include latency to escape the start chamber and correct
choices.7,43 Initial tests with the maze43 showed that zebrafish could be trained to turn
in a particular direction (spatial learning) or to approach a particular color regardless
of location (nonspatial learning).
We have used the three-chamber maze to show that the delayed spatial alter-
nation behavior is a sensitive index of the persisting cognitive impairment caused
by developmental exposure to the organophosphate pesticide chlorpyrifos (Fig-
ure 15.5).7 A parallel line of investigation (Figure 15.6)76 found that acute nicotine
administration causes a significant improvement in delayed spatial alternation at low
doses but impairs performance at high doses. The biphasic effect of nicotine improv-
ing memory function at low doses and having less improvement at higher doses is a
common finding across a wide variety of species including rats, mice, monkeys, and

A. Choice Phase B. Correct Choice C. Incorrect Choice
Vertical
sliding Vertical
Sliding
door sliding Sliding
partition
(open) door partition
and track
(closed) in restriction
Left-Choice Start Right-Choice
Chamber Chamber position
Chamber
1. Spatial Discrimination 2. Red-Blue Color Discrimination

Learning Learning Reversal
4 6
Linear Trend ***p<0.0005 vs. Ses 1–2
**p<0.005 vs. Ses 13–14
Escape/Avoid Latency (secs)
p<0.05 4
Escape/Avoid Latency (secs)
3 *p<0.01 vs. Rev 1–2

**
2
2
0
1
–2 *
0
–4 ***
–1 –6
–2 –8
1–2 3–4 5–6 7–8 9–10 1–2 3–4 5–6 7–8 9–10 11– 13– Rev Rev
12 14 1–2 3–4
Session Block Session Block
FIGURE 15.4 Learning and memory assessment in zebrafish with the three-chamber shut-
tle maze.7,43,76,82 As shown in the top inset (A), trials begin with the fish in the “start chamber”;
during a choice phase, both vertical sliding doors are opened. After a choice the sliding
doors are closed; if the fish chooses the correct chamber (as in B) it is allowed to swim freely
for a short time, but incorrect choices are punished by sliding the partition to restrict the
swimming of the fish (as in C). Graphs present data from studies with zebrafish in the three-
chamber maze. The lower-left graph presents findings from a study on spatial discrimination
learning,43 and the lower-right graph presents findings from a study on color discrimination
and reversal learning.43 Data points are average performances and standard errors.
humans.77–79 The fact that the same effect was seen in zebrafish points to similari-
ties of nicotinic effects on memory with mammalian species. This similarity can
be advantageous because molecular studies of neural function can be more easily
studied in zebrafish than mammals. The inexpensive zebrafish and a rapid, short, six-
trial spatial discrimination test was useful in determining the time-effect function
for nicotine-induced accuracy improvements.

#!% $ "

&! #"%

!!

# !

%
FIGURE 15.5 Early exposure to chlorpyrifos produces a persisting effect on delayed spa-
tial alternation in zebrafish tested in the three-chamber shuttle maze.7 Both 10 ng/mL (p <
0.05) and 100 ng/mL (p < 0.01) of chlorpyrifos during the first 5 days of development cause
reduced spatial alternation accuracy in adult zebrafish. Bars show average performances and
standard errors.
15.3.5 TESTING ANXIETY AND STRESS RESPONSE

Anxiety and stress response can be tested in zebrafish by indexing their diving in
a novel tank—many fish, including the zebrafish, show a defensive diving response
in response to threat. When zebrafish are placed in a novel tank they tend to dive to
the bottom of the tank, dwelling there and gradually rising to the upper levels over
a period of minutes. This is similar to thigmotaxis or time spent near the walls of
an open field in rodents. We have developed an automated assessment of the diving
response using digital video tracking, illustrated in Figure 15.7 (upper inset). The
graph in Figure 15.7 shows results of a recent study examining the acute nicotine
effect on the stress-diving response.80 Control zebrafish stayed in the bottom third
of the tank for nearly all of the first minute in a novel tank, and then over the rest
of the 5-min session spent progressively less time in the bottom. Acute nicotine
treatment significantly reduced this effect with 50 mg/mL significantly reducing the
initial diving and 100 mg/mL eliminating preference for the bottom. Interestingly,
this effect of nicotine on the diving response did not seem to be caused by drug-
induced confusion since the same dose significantly improved accuracy in the maze
test. In addition, nicotine effects on swimming activity were also congruent with the
changes in bottom dwelling.

& *.&)"2" .-*)"'1"!+.&'

'.",).&*)&)",4-%", ").*,," .

&)","2" .
*0",!*-"
&2",") "#,*(*).,*'
+
", ").*,," .

3
3

& *.&)"($'&.", (&)
& *.&)"2" .-*)*-&.&*)

&- ,&(&).&*)- +"*-&)$.*"-.&)$
)0",0'2" .#*,($& *.&)"

).",0'2" .-

", ").*,," .
&)",+
/!,.& +

& *.&)"0-*).,*'
+

+

).",0'(&)
FIGURE 15.6 Effects of acute nicotine on zebrafish performance in the three-chamber

maze. The top graph presents findings showing that nicotine produces a significant (p <
0.005) linear dose-effect function on spatial alternation: low doses facilitate performance
and high doses impair performance.76 The bottom graph shows the nicotine time-effect func-
tion on spatial discrimination in which significant improvement in spatial alternation is seen
20–40 min after dosing at 100 mg/L for 3 min.82 Data points are average performances, with
bars showing standard error of the mean.

7 "-/& (
2&)/*' &!"+
--4 )"-
+),0/"-&5"!
&!"+- '&*$
4./")
& +/&*"6" /.+*+//+)2"((&*$

&*"-8.%+1"-&)"

& +/&*"
)$#+- )&*
" +*!./+//+),"-&*0/"

& 3&*

*/"- /&+*
,

&),("&*6" /.
& +/&*"1.+*/-+(
, ,

,

&*0/"
FIGURE 15.7 A zebrafish diving response test used to measure fear—stress response—to
a novel environment. The top inset shows an array of tanks used to study the novelty-induced
diving response in the zebrafish. In the study, fish are individually introduced into a tank and
a digital tracking system is used to measure the duration of bottom dwelling. The graph shows
results of a study examining the acute nicotine effect on the diving response.80 Immersion in
50 mg/L of nicotine for 3 min significantly (p < .05) reduced bottom-dwelling time in the first
minute, and 100 mg/L significantly reduced bottom-dwelling throughout the test (p-values
ranging from 0.05 to 0.0001).

15.4 CONCLUSIONS
Despite the small size of the zebrafish, a number of promising behavioral assays have
appeared in the literature—it is now clear that the zebrafish model of development
can be used in studies of learning, memory, and cognition. There are both appetitive
and aversive techniques, and they test a range of behavior, from simple reflexes81 and
fear conditioning,72 to visual discrimination25 and spatial orientation.43
In the development and use of animal models of behavioral dysfunction, it is
important to develop complementary models to take advantage of the unique advan-
tages of the different species. Non-mammalian vertebrates such as zebrafish provide
the opportunity to directly observe neurodevelopmental processes and determine
the impact of developmental permutations on learning and memory. Zebrafish are
particularly valuable because of the availability of morpholino techniques to tran-
siently suppress specific parts of genomic expression. The development of new meth-
ods for high-throughput tests of cognitive function for fish can provide means for
rapid screening of potential toxic agents as well as promising therapeutic agents. It is
equally important to develop specific tests of various aspects of cognitive function,
including habituation, associative learning, memory, and attention, as well as to be
able to differentiate changes in sensorimotor function from cognition. Key in the use
of zebrafish models is the determination of which mechanisms of cognitive function
are similar to mammals and which are different. Non-mammalian models can be
used in concert with classic mammalian models to determine the neural bases of
cognitive function and discovery of toxicants and potential therapeutic agents.
ACKNOWLEDGEMENT
Research was supported by NIH ES10356.
REFERENCES
1. Kandel, E. R. 2004. The molecular biology of memory storage: A dialog between genes
and synapses, Bioscience Report 24:475–522.
2. Rodríguez, F., López, J. C., Vargas, J. P., Broglio, C., Gómez, Y., and Salas, C. 2002.
Spatial memory and hippocampal pallium through vertebrate evolution: Insights from
reptiles and teleost fish. Brain Research Bulletin 57:499–503.
3. Darland, T., and Dowling, J. E. 2001. Behavioral screening for cocaine sensitivity in
mutagenized zebrafish. Proceedings of the National Academy of Sciences of the United
States of America 98(20):11,691–96.
4. Guo, S. 2001. Linking genes to brain, behavior and neurological disease: What can we
learn from zebrafish. Genes, Brain and Behavior 3:63–74.
5. Gerlai, R., Lahav, M., Guo, S., and Rosenthal, A., 2000. Drinks like a fish: Zebra fish
(Danio rerio) as a behavior genetic model to study alcohol effects. Pharmacology,
Biochemistry & Behavior 67(4):773–82.
6. Linney, E., Upchurch, L., and Donerly, S., 2004. Zebrafish as a neurotoxicological
model. Neurotoxicology and Teratology 27(1):709–718.
7. Levin, E. D., Crysthansis, E., Yacisin, K., and Linney, E. 2003. Chlorpyrifos exposure
of developing zebrafish: Effects on survival and long-term effects on response latency
and spatial discrimination. Neurotoxicology and Teratology 25:51–57.

8. Barlow, G. W. 2002. The cichlid fishes: Nature’s grand experiment in evolution. New
York: DaCapo Press.
9. Tebbich, S., Bshary, R., and Grutter, A. S. 2002. Cleaner fish Labroides dimidiatus
recognise familiar clients. Animal Cognition 5(3):139–45.
10. Reader, S. M., Kendal, J. R., and Laland, K. N., 2003. Social learning of foraging sites
and escape routes in wild Trinidadian guppies. Animal Behaviour 66(4):729–39.
11. Santangelo, N., and Itzkowitz, M. 2004. Sex differences in the mate selection process of
the monogamous, biparental convict cichlid, Archocentrus nigrofasciatum. Behaviour
141(8):1041–59.
12. Payne, R. J. H. 1998. Gradually escalating fights and displays: The cumulative assess-
ment model. Animal Behaviour 56(3):651–62.
13. Suboski, M. D. 1988. Acquisition and social communication of stimulus recognition by
fish. Behavioural Processes 16(3):213–44.
14. Suboski, M. D., Bain, S., Carty, A. E., and McQuoid, L. M. 1990.Alarm reaction in
acquisition and social transmission of simulated predator recognition by zebra danio
fish. Journal of Comparative Psychology 104:101–12.
15. Braithwaite, V. A., Armstrong, J. D., McAdam, H. M., and Huntingford, F. A., 1996.
Can juvenile Atlantic salmon use multiple cue systems in spatial learning? Animal
Behaviour 51(6):1409–15.
16. Miklosi, A., Andrew, R. J., and Savage, H. 1997. Behavioural lateralisation of the tetra-
pod type in the zebrafish (Brachydanio rerio). Physiology & Behavior 63(1):127–35.
17. Miklosi, A., and Andrew, R. J. 1999. Right eye use associated with decision to bite in
zebrafish. Behavioral Brain Research 105:199–205.
18. Drew, M. R., Zupan, B., Cooke, A., Couvillon, P. A., and Balsam, P. D. 2005. Temporal
control of conditioned responding in goldfish. Journal of Experimental Psychology:
Animal Behavior Processes 31(1):31–9.
19. Reebs, S. G. 1996. Time-place learning in golden shiners (Pisces: Cyprinidae). Behav-
ioural Processes 36(3):253–62.
20. Higa, J. J., and Simm, L. A. 2004. Interval timing in Siamese fighting fish (Betta splen-
dens). Behavioural Processes 67:501–9.
21. Hollis, K. L. 1999. The role of learning in the aggressive and reproductive behav-
ior of blue gouramis, Trichogaster trichopterus. Environmental Biology of Fishes
54(4):355–69.
22. Behrend, E. R., and Bitterman, M. E. 1964. Avoidance-conditioning in the fish: Further
studies of the CS-US interval. American Journal of Psychology 77(1):15–28.
23. Behrend, E. R., and Bitterman, M. E. 1963. Sidman avoidance in the fish. Journal of the
Experimental Analysis of Behavior 6(1):47–52.
24. Talton, L. E., Higa, J. J., and Staddon, J. E. R. 1999. Interval schedule performance in
the goldfish Carassius auratus. Behavioural Processes 45(1–3):193–206.
25. Colwill, R. M., Raymond, M. P., Ferreira, L., and Escudero, H. 2005. Visual discrimi-
nation learning in zebrafish (Danio rerio). Behavioural Processes 70(1):19–31.
26. Reebs, S. 2001. Fish behavior in the aquarium and in the wild. Ithaca, NY: Comstock
Pub. Assoc.
27. Bitterman, M. E. 1965. Phyletic differences in learning. American Psychologist
20:396–410.
28. Tinbergen, N. 1951. The study of instinct. Oxford, UK: Oxford University Press.
29. Hollis, K. L., and Overmier, J. B. 1978. The function of the telost telencephalon in
behavior: A reinforcer mediator. In ed. D I. Mostofsky, The behavior of fish and other
aquatic animals, 1371–95. New York: Academic Press.
30. Wullimann, M. F., and Mueller, T. 2004. Teleostean and mammalian forebrains
contrasted: Evidence from genes to behavior. Journal of Comparative Neurology
475(2):143–62.

31. Overmier, J. B., and Papini, M. R. 1986. Factors modulating the effects of telost tel-
encephalon ablation on retention, relearning, and extinction of avoidance behavior.
Behavioral Neuroscience 100:190–99.
32. Salas, C., Broglio, C., Rodriguez, F., Lopez, J. C., Portavella, M., and Torres, B., 1996.
Telencephalic ablation in goldfish impairs performance in a ‘spatial constancy’ prob-
lem but not in a cued one. Behavioral Brain Research 79(1–2):193–200.
33. Broglio, C., Rodriguez, F., and Salas, C. 2003. Spatial cognition and its neural basis in
teleost fishes. Fish and Fisheries 4:247–55.
34. Shin, S. W., Chung, N. I., Kim, J. S., et al. 2001. Effect of diazinon on behavior of
Japanese medaka (Oryzias latipes) and gene expression of tyrosine hydroxylase as a
biomarker. Journal of Environmental Science & Health—Part B: Pesticides, Food
Contaminants, & Agricultural Wastes 36(6):783–95.
35. Phillips, T. A., Summerfelt, R. C., and Atchison, G. J. 2002. Environmental, biological,
and methodological factors affecting cholinesterase activity in walleye (Stizostedion
vitreum). Archives of Environmental Contamination & Toxicology 43(1):75–80.
36. Brookshire, K. H., and Hognander, O. C. 1968. Conditioned fear in the fish. Psycho-
logical Reports 22:75–81.
37. Pinckney, G. A. 1967. Avoidance learning in fish as a function of prior fear condition-
ing. Psychological Reports 20:71–4.
38. Fetcho, J. R., and Liu, K. S. 1998. Zebrafish as a model system for studying neuronal
circuits and behavior. Annals of the New York Academy of Sciences 860:333–45.
39. Penberthy, W. T., Shafizadeh, E., and Lin, S. 2002. The zebrafish as a model for human
disease. Frontiers in Bioscience 7:D1439–53.
40. Streisinger, G., Walker, C., Dower, N., Knauber, D., and Singer, F. 1981. Production of
clones of homozygous diploid zebra fish (Brachydanio rerio). Nature 291:293–96.
41. Schier, A. F. 1997. Genetics of neural development in zebrafish. Current Opinion in
Neurobiology 7(1):119–26.
42. Anichtchik, O. V., Kaslin, J., Peitsaro, N., Scheinin, M., and Panula, P. 2004. Neuro-
chemical and behavioural changes in zebrafish Danio rerio after systemic administra-
tion of 6-hydroxydopamine and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. Journal
of Neurochemistry 88(2):443–53.
43. Arthur, D., and Levin, E. D. 2001. Spatial and non-spatial discrimination learning in
zebrafish. Animal Cognition 4:125–31.
44. Hall, D., and Suboski, M. D. 1995. Visual and olfactory stimuli in learned release of
alarm reactions by zebra danio fish (Brachydanio rerio). Neurobiology of Learning &
Memory 63(3):229–40.
45. Hall, D., and Suboski, M. D. 1995. Sensory preconditioning and second-order condi-
tioning of alarm reactions in zebra danio fish (Brachydanio rerio). Journal of Com-
parative Psychology 109(1):76–84.
46. Williams, F. E., White, D., and Messer, W. S. 2002. A simple spatial alternation task for
assessing memory function in zebrafish. Behavioural Processes 58(3):125–32.
47. Williams, F. E., and Messer, W. S. Jr. 1998. Memory function and muscarinic receptors
in zebrafish. Society for Neuroscience Abstracts 24(46):182.
48. Mueller, T., Vernier, P., and Wullimann, M. F. 2004. The adult central nervous choliner-
gic system of a neurogenetic model animal, the zebrafish Danio rerio. Brain Research
1011(2):156–69.
49. Donahoe, J. W., and Palmer, D. C. 1993. Learning and complex behavior. Needham
Heights, MA: Allyn & Bacon.
50. Shettleworth, S. J. 1998. Cognition, evolution, and behavior. New York: Oxford Uni-
versity Press.

51. Keller, J., Strasburger, H., Cerutti, D. T., and Sabel, B. A. 2000. Assessing spatial
vision—automated measurement of the contrast-sensitivity function in the hooded rat.
Journal of Neuroscience Methods 97:103–10.
52. Rodriguez, F., Duran, E., Vargas, J. P., Torres, B., and Salas, C. 1994. Performance of
goldfish trained in allocentric and egocentric maze procedures suggests the presence of
a cognitive mapping system in fishes. Animal Learning & Behavior 22(4):409–20.
53. Li, L., and Dowling, J. E. 1997. A dominant form of inherited retinal degeneration
caused by a non-photoreceptor cell-specific mutation. Proceedings of the National
Academy of Sciences of the United States of America 94(21):11645–50.
54. Levin, E. D., Swain, H. A., Donerly, S., and Linney, E., 2004. Developmental chlorpy-
rifos effects on hatchling zebrafish swimming behavior. Neurotoxicology & Teratology
26(6):719–23.
55. Granato, M., van Eeden, F. J., Schach, U., et al. 1996. Genes controlling and mediating
locomotion behavior of the zebrafish embryo and larva. Development 123:399–413.
56. Brustein, E., Saint-Amant, L., Buss, R. R., Chong, M., McDearmid, J. R., and Drapeau,
P. 2003. Steps during the development of the zebrafish locomotor network. Journal of
Physiology—Paris 97(1):77–86.
57. CarvanI, M. J., Loucks, E., Weber, D. N., and Williams, F. E. 2004. Ethanol effects on
the developing zebrafish: Neurobehavior and skeletal morphogenesis. Neurotoxicology
and Teratology 26(6):757–68.
58. Carvan, M. J. 2003. Mechanistic observations on the effects of neurotoxicants on
zebrafish behavior. Neurotoxicology and Teratology 25(3):383.
59. Staddon, J. E. R., and Higa, J. J. 1996. Multiple time scales in simple habituation. Psy-
chological Review 103:720–33.
60. Blanchard, R. J., Hebert, M. A., Ferrari, P., et al. 1998. Defensive behaviors in wild
and laboratory (Swiss) mice: The Mouse Defense Test Battery. Physiology & Behavior
65(2):201–9.
61. Neuhauss, S. C. F. 2003. Behavioral genetic approaches to visual system development
and function in zebrafish. Journal of Neurobiology 54(1):148–60.
62. Li, L. 2001. Zebrafish mutants: Behavioral genetic studies of visual system defects.
Developmental Dynamics 221(4):365–72.
63. Rinner, O., Rick, J. M., and Neuhauss, S. C. F. 2005. Contrast sensitivity, spatial and
temporal tuning of the larval zebrafish optokinetic response. Investigative Ophtalmol-
ogy & Visual Science 137–42.
64. Sagvolden, T., Russell, V. A., Aase, H., Johansen, E. B., and Farshbaf, M. 2005.
Rodent models of attention-deficit/hyperactivity disorder. Biological Psychiatry
57(11):1239–47.
65. Trinh, J. V., Nehrenberg, D. L., Jacobsen, J. P. R., Caron, M. G., and Wetsel, W. C.
2003. Differential psychostimulant-induced activation of neural circuits in dopamine
transporter knockout and wild type mice. Neuroscience 118(2):297–310.
66. Jacobsen, J. P. R., Rodriguiz, R. M., Mork, A., and Wetsel, W. C. 2005. Monoaminergic
dysregulation in glutathione-deficient mice: Possible relevance to schizophrenia? Neu-
roscience 132(4):1055–72.
67. Miklosi, A., Andrew, R. J., and Savage, H. 1998. Behavioral lateralization of the tetra-
pod type in the zebrafish (Brachiodanio rerio). Physiology & Behavior 63:127–35.
68. Barth, K. A., Miklosi, A., Watkins, J., Bianco, I. H., Wilson, S. W., and Andrew, R. J.
2005. fsi zebrafish show concordant reversal of laterality of viscera, neuroanatomy, and
a subset of behavioral responses. Current Biology 15(9):844–50.
69. Bisazza, A., Rogers, L. J., and Vallortigara, G. 1998.The origins of cerebral asymme-
try: A review of evidence of behavioural and brain lateralization in fishes, reptiles and
amphibians. Neuroscience and Biobehavioral Reviews 22(3):411–26.

70. Peitsaro, N., Kaslin, J., Anichtchik, O. V., and Panula, P. 2003. Modulation of the his-
taminergic system and behaviour by alpha-fluoromethylhistidine in zebrafish. Journal
of Neurochemistry 86(2):432–41.
71. Swain, H. A., Sigstad, C., and Scalzo, F. M. 2004. Effects of dizocilpine (MK-801) on
circling behavior, swimming activity, and place preference in zebrafish (Danio rerio).
Neurotoxicology and Teratology 26(6):725–29.
72. Pradel, G., Schmidt, R., and Schachner, M., 2000. Involvement of L1.1 in memory con-
solidation after active avoidance conditioning in zebrafish. Journal of Neurobiology
43:389–403.
73. Gleason, P. E., Weber, P. G., and Weber, S. P. 1977. Effect of group size on avoidance
learning in zebra fish: Brachydanio rerio (Pisces: Cyprinidae). Animal Learning &
Behavior 5(2):213–16.
74. Pradel, G., Schachner, M., and Schmidt, R. 1999. Inhibition of memory consolidation
by antibodies against cell adhesion molecules after active avoidance conditioning in
zebrafish. Journal of Neurobiology 39(2):197–206.
75. Flanigan, W. F., and Caldwell, W. E., 1971. Galvanotaxic behavior and reinforcement of
fish Brachydanio rerio. Genetic Psychology Monographs 84(1):35–71.
76. Levin, E. D., and Chen, E. 2004. Nicotinic involvement in memory function in zebraf-
ish. Neurotoxicology and Teratology 26:731–35.
77. Levin, E. D., Decker, M. W., and Butcher, L. L. 1992. Neurotransmitter interactions
and cognitive function. Boston: Berkhäuser.
78. Levin, E. D., and Simon, B. B. 1998. Nicotinic acetylcholine involvement in cognitive
function in animals. Psychopharmacology 138:217–30.
79. Levin, E. D., and Slotkin, T. A. 1998. Developmental neurotoxicity of nicotine. In
Handbook of developmental neurotoxicology, eds. W. Slikker Jr. and L. W. Chang,
587–615. San Diego: Academic Press.
80. Levin, E. D., Bencan, Z., and Cerutti, D. T. 2007. Anxiolytic effects of nicotine in
zebrafish. Physiology and Behavior 90:54–58.
81. Carvan, M. J. III, Loucks, E., Weber, D. N., and Williams, F. E. 2004. Ethanol effects
on the developing zebrafish: Neurobehavior and skeletal morphogenesis. Neurotoxicol-
ogy and Teratology 26(6):757–68.
82. Levin, E. D., Limpuangthip, J., Rachakonda, T., and Peterson, M. 2006. Timing of
nicotine effects on learning in zebrafish. Psychopharmacology 184:547–52.

16 Caenorhabditis
elegans Model for
Initial Screening and
Mechanistic Evaluation
of Potential New
Drugs for Aging and
Alzheimer’s Disease
Yuan Luo, Yanjue Wu, Marishka
Brown, and Christopher D. Link
CONTENTS
16.1 Introduction................................................................................................. 312
16.2 Methods....................................................................................................... 314
16.2.1 Animal Subjects............................................................................... 314
16.2.2 Equipment ........................................................................................ 314
16.2.3 Procedure 1: AG-Induced Paralysis Behavior Assay ....................... 315
16.2.4 Procedure 2: Chemotaxis Behavior Assay ...................................... 316
16.2.5 Procedure 3: Serotonin Sensitivity Assay and Egg-Laying
Assay................................................................................................ 317
16.2.6 Procedure 4: Pharyngeal Pumping Assay and Life Span Assay ..... 318
16.3.1 Target Identification and Validation ................................................ 319
16.3.2 Mechanism of Action Using Mutant Worms and RNAi Feeding ... 319
16.3.3 Rapid Toxicity Assessment for Pharmaceutical Compounds .......... 321
16.5.1 Screening Compounds that Affect AG-Induced Pathological
Behaviors ......................................................................................... 323
16.5.2 Dose-Response of Epigallocatechin Gallate on Pharyngeal
Pumping in C. elegans..................................................................... 323
16.5.3 Life Span Extension by EGb 761 and Other Compounds................ 323
311

16.5.4 Testing Compounds that Delay Age-Associated Muscle

Degeneration .................................................................................... 323
Acknowledgments.................................................................................................. 325
References.............................................................................................................. 325
16.1 INTRODUCTION
Alzheimer’s disease (AD) is one of the most devastating CNS disorders of old age,
which has led to a serious public health problem. Currently about five million Amer-
icans are affected at a proximate cost of up to $100 billion per year. It has been esti-
mated that by the year 2050, the number of AD patients will be over 14 million if no
new treatments are developed.1 The existing two classes of Food and Drug Adminis-
tration (FDA)-approved drugs for treatment of AD—acetylcholineesterase (AChE)
inhibitors and N-methyl-D-aspartate (NMDA) receptor antagonists—only provide
symptomatic benefit for some mild to moderate AD patients. The means to prevent
or reduce the rate of this disorder is a high priority for medical research. Develop-
ment of new drugs with disease-modifying or preventive properties for this devastat-
ing disease continues to be a challenging job. The apparent difficulty is because of
the lack of validated therapeutic targets and adequate animal models.2
The transgenic mouse model of AD has been useful for understanding the
mechanisms by which specific mutations might lead to an AD-related behavior phe-
notype and for testing possible therapeutics.3–6 Despite the availability of a vari-
ety of transgenic mouse models of AD,7,8 drug evaluations with these animals are
expensive and time consuming. It is recognized that assays based on inexpensive
in vivo models amenable to high-throughput screening (HTS), such as the round
worm Caenorhabditis elegans9 and the fruit fly Drosophila melanogaster10 pro-
vide attractive platforms for streamlining efficient drug discovery and drug target
identification.11
Modeling AD in a simple microscopic organism such as C. elegans is a rela-
tively new approach.12 Worms genetically engineered to express the human AG1–42
peptide in muscle cells accumulate both immunoreactive deposits of AG1–42 and
insoluble G-amyloid, as is observed in senile plaques in AD brains. (This muscle
cell expression model may more directly model inclusion body myositis, a severe
human myopathy associated with intramuscular accumulation of AG.) Accumula-
tion of AG1–42 in this model is associated with a progressive paralysis, providing
a simple biological readout of AG toxicity.9 The relatively short life span of the
worms (they only live about 20 days) allows us to evaluate the time sequence
of events in these animals during their entire lives. Thus, transgenic C. elegans
expressing AG42 have been used extensively for the mechanistic study of AG-42tox-
icity (1) because of its ability to express muscle-specific human AG peptide, which
forms intracellular AG deposits;13 and (2) its up-regulated stress response genes,14
which are known to be elevated in the AD brain.15,16 Most importantly, the trans-
genic strain develops a concomitant progressive paralysis phenotype (CL4176),17
which can be easily scored and quantified.
To understand the early events in the progression of AD for development of ther-
apeutic strategies, it is fundamentally important to determine the temporal sequence

Caenorhabditis elegans Model for Initial Screening 313
of events leading to neurodegeneration. C. elegans is a suitable tool for mechanistic

examination of the transgene products, as well as for pharmacological analysis of
time course and the kinetics of drug effects.18 For example, a relationship between
AG amino-acid sequence and amyloid aggregation was established using this model.
Yatin et al.19 showed that methionine (Met35) is critical for free radical production
by AG1–42, and it is also critical for G-sheet formation in the transgenic C. elegans
lines.20
Although the worms may not be universally accepted as having a direct relevance
for AD pathology, they are well suited for validation of target AG toxicity in vivo.12,21
The absence of endogenous AG production in the worms offers an opportunity to
find a direct role of the AG involvement in pathological behaviors.22 In addition, pre-
dominantly intracellular expression of AG provides another tool to address specific
roles of intracellular AG in relation to its toxicity. Substantial evidence implicates
intracellular AG oligomers in early events related to AD.16 Intracellular AG has also
been observed in human brain neurons23 and in triple transgenic AD mouse models,
where its accumulation preceded neurofibrillary tangle formation.24 This evidence
supports the notion that AG toxicity assayed in the worm model reflects AG toxic-
ity in mammalian neurons. A recent study indicates that the transgenic C. elegans
model may be generally relevant to the proteotoxicity underlying neurodegenerative
diseases.25 Additionally, the strain has been used to investigate the role of insulin-
like signaling and heat-shock factor in AG proteotoxicity,26,27 providing excellent
examples for the relevance of the C. elegans model to AD.
There are several advantages of C. elegans over the mouse model for initial drug
screening and target characterization. First, there are highly conserved biochemi-
cal pathways between worms and humans. Second, established transgenic mutant
linking of human AG expression with pathological behavioral phenotypes are easy
to score. The worms have a relatively low cost of cultivation because of their small
size, rapid life cycle, and short life span,28 which allow screening of thousands of
animals over multiple generations on microtiter plates. The simple structure of its
nervous system, consisting of only 302 neurons in an adult nematode, makes it valu-
able for screening drugs against age-associated neurodegeneration and the ease of
genetic manipulations, which is evident in the availability of mutants and application
of RNA interference (RNAi) knockdown. Several examples illustrate the power of
C. elegans in screening for new drugs,29 including many known human drugs.30,31
Some lead molecules originating from worm-based screening assays are in advanced
stages of drug discovery.11
Using the C. elegans model in the past years, we have uncovered effects of natu-
ral compounds on extension of the worms’ life span;32 on a stress response protein,
the small heat-shock protein hsp-16.2;33 on age-related behavioral declines;34 on mus-
cle degeneration;35 and on AG-expression-induced pathological behaviors.22 Most of
those experiments would be difficult and might be impossible to perform in mice. In
this chapter, we describe methods we have employed for compound screening and
pharmacological evaluations of potential AD drugs using the C. elegans model.

16.2 METHODS
Since Sydney Brenner introduced the soil nematode C. elegans into laboratory prac-
tices in the 1960s, it has continuously proven its merit as a model organism in scien-
tific research. C. elegans’ diet consists primarily of bacteria (Escherichia coli), and
under optimal laboratory conditions the nematode has a 3-day growth-reproduction
cycle. C. elegans has two sexes: hermaphrodite and male. An adult hermaphrodite
grows to ~1 mm in length and 80 μm in diameter and produces close to 300 progeny
by self-fertilization. Besides the ability to grow a homogenous population, the worm
is small enough that it can be handled in large numbers and it is easily cultivated in
the laboratory. It is also amendable to genetic manipulations, which is evidenced by
the large mutant library. The fully sequenced worm genome revealed 60%–80% of
the genes shared with humans (available at the Wormbase: http://www.wormbase.
org). There are many biochemical pathways that have been evolutionarily conserved
between the nematode and humans that make it an ideal model for discovering novel
drug targets.
16.2.2 EQUIPMENT
1. Temperature-controlled incubator (Sheldon Manufacturing, Cornelius,
Oregon, USA)
2. Automated timer
3. Platinum wire loop or inoculating loop
4. NGM-Nematode Growth Media (for 500 mL)
a. Before autoclaving, add:
i. 9.5 g agar
ii. 1.25 g peptone
iii. 1.5 g NaCl
iv. 487.5 mL distilled H2O
v. 0.5 mL cholesterol (5 mg/mL in ethanol)
vi. Autoclave mixture for 30 min
b. After autoclaving, add:
i. 0.5 mL CaCl2 (1 M)
ii. 0.5 mL MgSO4 (1 M)
iii. 12.5 mL KH2PO4 (1 M) (pH 6.0)
5. Microscope: Nikon SMZ1000 stereoscopic zoom microscope
6. Petri dishes (60 × 15 mm)
7. OP50 food source (E. coli) (to make 500 mL)
a. 5.0 g tryptone
b. 2.5 g NaCl
c. 500 mL distilled H2O
Note: Add tryptone and NaCl to a 500 mL glass bottle. Pour in 500 mL of distilled
H2O and add a stirring rod. Stir on hot plate without heat until the tryptone is dis-
solved. Place the culture media and two empty covered 500 mL Erlenmeyer flasks in

the autoclave for 30 min. Leave for several hours to cool at room temperature. After
the culture media has reached room temperature, place 5 mL into two 15 mL prest-
erilized centrifuge tubes. Using aseptic technique, sterilize an inoculating loop and
carefully place colonies of E. coli into each tube. Let incubate at 37°C for 18–24 hr.
Place 250 mL of the culture media into each sterilized flask. Add the 5 mL of liquid
OP50 to each flask. Incubate at 37°C for 18–24 hr. Pour OP50 into 50 mL presteril-
ized tubes and keep at 4°C until needed.
8. M9 buffer (to make 1 L)
a. 3 g KH2PO4
b. 6 g Na2HPO4
c. 5 g NaCl
d. 1 mL MgSO4 (1 M)
e. 1000 mL distilled H2O
f. Autoclave mixture
16.2.3 PROCEDURE 1: AG-INDUCED PARALYSIS BEHAVIOR ASSAY

Unlike human amyloid precursor protein (APP) gene, C. elegans homologue gene
apl-1 cannot produce neurotoxic peptide AG1–42.36 However, AG toxicity can be
investigated by engineering C. elegans to express human AG1–42 in either neurons
[12] or body wall muscle.9 In developing an inducible transgenic C. elegans AD
model, the transgene plasmid was introduced into C. elegans by gonad microinjec-
tion with visible marker gene rol-6. Abnormally long 3' untranslated region (UTR)
in the transgene plasmid makes transgene expression under the control of messenger
RNA (mRNA) surveillance system. At 16°C, the mRNA surveillance system in the
transgenic strain monitors the expression of abnormally long 3' UTR sequence in the
AG plasmid and degrades these transcripts. After temperature upshifts to 23°C, the
mRNA surveillance system in C. elegans becomes inactive, which allows the stabi-
lized expression of the AG plasmid.14 After AG expression in the muscle cells for 24
hr, transgenic worms gradually become paralyzed9 (Figure 16.1A).
Transgenic C. elegans CL4176 (smg-1ts [myo-3/AG1–42 long 3'-UTR]) and control
strain CL1175 (smg-1ts) are propagated at 16°C on solid nematode growth medium
(NGM) plates seeded with 100 μL OP50. To prepare age-synchronized animals,
transgenic C. elegans are transferred to fresh NGM plates upon reaching reproduc-
tive maturity at L4 stage and allowed to lay eggs for 4–6 hr on a food lawn contain-
ing compounds or vehicle. Young larvae from the synchronized eggs are cultured at
16°C in a temperature-controlled incubator (Sheldon Manufacturing, Model 2005,
Cornelius, Oregon, USA). AG transgene expression in muscle cells is induced by
upshifting the temperature from 16°C to 23°C at the 36th hour after egg laying and
lasts until the end of the paralysis assay. Usually after temperature upshifts for 24 hr,
the number of paralyzed worms is scored under the microscope (Nikon SMZ1000,
Japan) at 1-hr intervals until the last worm becomes paralyzed. To identify the paral-
ysis, each worm will be gently touched with a platinum loop (VWR, Bridgeport,
New Jersey, USA). The worm is considered paralyzed if it moves its head only or
does not move after touching. Paralysis time course is plotted thereafter. To quantify


FIGURE 16.1 (a) Images of AG-induced paralysis in the transgenic CL4176 strain without
transgene expression (no AG), untreated with EGb 761 (Ctrl, left panel), and in the transgenic
CL4176 strain (muscle AG strain) fed with (EGb, right) or without EGb 761 (Ctrl, middle) at
36 hr after temperature up-shift. (b) Schematic diagram of the chemotaxis assay. Worms are
transferred to the center of the plate. After attractant and control odorant mixed with sodium
azide are dropped on the plate, C.elegans start to move toward odorant source until they are
paralyzed by sodium azide.
the paralysis, PT50 is used to measure the rate of the paralysis. It is defined as a mean
time duration at which 50% worms are paralyzed.37
Note: It is important to keep the upshift temperature consistent (23°C). Even a dif-
ference by 1°C would significantly affect the onset and the duration of paralysis.
16.2.4 PROCEDURE 2: CHEMOTAXIS BEHAVIOR ASSAY

C. elegans can detect attractants or repellents with chemosensory neurons. The che-
motaxis response is mediated by activation of several sensory neurons and interneu-
rons to stimulate the motor neurons.38 The chemotaxis assay was used to identify a
neuronal behavioral phenotype in the transgenic C. elegans strain CL2355 in which
AG is expressed in neurons.21 Before chemotaxis assay, synchronized transgenic
C.elegans CL2355 (smg-1ts [snb-1/AG1–42 long 3'-UTR]) and its control strain CL2122
(mtl-2/green fluorescent protein) are cultured in 16°C for 36 hr and then in 23°C
for another 36 hr. The worms are then collected and washed with M9 buffer three
times and transferred to 100 × 15 mm plates (VWR, Bridgeport, New Jersey, USA)
containing 1.9% agar, 1 mM CaCL2, 1 mM MgSO4, and 25 mM phosphate buffer,
pH 6.0. Twenty worms are placed to the center of the plate. After all animals are
transferred to the plate, 1 μL of odorant (0.1% benzaldehyde in 100% ethanol) along
with 1 μL of 1 M sodium azide is added to the original position. On the opposite

side of the attractant, 1 μL drop of sodium azide and 1 μL of control odorant (100%
ethanol) are added. Assay plates are incubated at 23°C for 1 hr (Figure 16.1B). The
chemotaxis behavior is scored and expressed as chemotaxis index (CI). CI is defined
as (number of worms at the attractant location − number of worms at the control
location)/total number of worms on the plate.
Note: The chemotaxis behavior is affected by age. Therefore, it is critical to com-
pare CI in a given age, e.g., at L4 stage.
16.2.5 PROCEDURE 3: SEROTONIN SENSITIVITY ASSAY AND EGG-LAYING ASSAY

C. elegans has the ability to take up exogenous serotonin (5-HT). 5-HT sensitivity
assay is used to determine whether 5-HT mediated neurotransmission is affected
by AG depositions in the neurons. 5-HT is a key neurotransmitter that modulates
several behaviors of C. elegans, including egg laying, locomotion, and olfactory
learning.39 This 5-HT sensitivity assay has previously been used to identify 5-HT
hypersensitive mutants, which revealed the relationship of the genes involved in 5-
HT signaling.40,41
Synchronized transgenic worms (CL2355) and control strain (CL2122) are col-
lected at 36 hr after temperature upshift. 5-HT (serotonin, creatinine sulfate salt;
Sigma) is dissolved in M9 buffer to 1 mM. Twenty worms in each group are washed
with M9 three times and transferred into 200 μL serotonin solution in a 96-well assay
plate (VWR, Bridgeport, New Jersey, USA). The number of worms is scored after 5
min as active or paralyzed (immobile for 5 sec) (Figure 16.2). Data are expressed as
percent not paralyzed.
Egg laying, a 5-HT controlled behavior in C. elegans, has been used as a simple
genetic system to identify and characterize the action of drugs on neurotransmitter
pathways that modulate a particular behavior.42 Egg laying requires the harmoni-
ous orchestration of several different neurons, neurotransmitters, and muscles. Neu-
rotransmitters activate the vulval muscles and the fertilized eggs are ejected out of
#

&
Vehicle or compounds in food "#)$$$$)
#
&#''%(%)$$)
) #! * Vehicle or compounds
#' #
) !#
!!!%'$$)
Vehicle or compounds
FIGURE 16.2 Summary of experiments for paralysis, survival, and biochemical assays in
C. elegans

the uterus by contractions of these muscles. The rate of egg laying is controlled by
the availability of food. If the animals are starved, then there is a cessation of egg
laying and development continues to occur inside the adult worm.
To perform the egg-laying assay, age-synchronized, well-fed L4 young larvae are
transferred to fresh plates seeded with OP50 and allowed to develop ~20 hr at 20°C.
The resultant young adults will be used in the assay. To test the response to drugs,
individual young adults (that are fed our drug) will be transferred to a 96-well plate
containing 100 μL of the following drug concentrations: 5 mg/mL solution of 5-HT,
serotonin creatinine sulfate complex (Sigma); 0.75 mg/mL imipramine; 0.5 mg/mL
fluoxetine. All drugs are dissolved in M9 buffer. The number of eggs released at
room temperature will be scored after 60 min. The egg-laying assay protocol and
concentrations are well established.40
16.2.6 PROCEDURE 4: PHARYNGEAL PUMPING ASSAY AND LIFE SPAN ASSAY

The neuromuscular pharynx allows the worm to ingest the bacteria. Rhythmic con-
traction and relaxation of this organ transports food from the mouth to the intes-
tines. This rate of pharyngeal pumping varies between individual worms and the
environment that the worms are in. If there is an abundant food source, the rate of
pharyngeal pumping can exceed 250 pumps/min-1. However, when there is a pau-
city of resources, i.e., food source, then the rate of pharyngeal pumping declines.
The pumping rates also progressively decline as the animals age. Recent research
has demonstrated that pharyngeal pumping rates can be pharmacologically modu-
lated.43 Administration of Epigallocatechin gallate (EGCG), a component of green
tea, delayed the age-related functional decline of pharyngeal pumping rates34 (Fig-
ure 16.2). Pharyngeal pumping spans have been positively correlated with the life
span of the animal.
To measure pharyngeal pumping, individual worms are placed on NGM agar
plates at room temperature containing a lawn of OP50. Assessment of pump-
ing behavior was done by observing the number of times the terminal bulb of the
pharynx contracted over a 1-min interval. Worms that display no pharyngeal con-
tractions in the intervening time period are classified as not pumping. Animals that
display 1–24 contractions are classified as slow pumping, and 25 or more pharyngeal
contractions are classified as fast pumping. Pharyngeal pumping is recorded every
day until the death of the worms. Adult day 1 animals are divided into a control
group and a treated group. Each NGM plate contains one worm and there are 12 total
plates for each group. The number of pharyngeal pumps is counted for each group
and an average is obtained.
For life span assay, adult hermaphrodites are allowed to lay eggs for 2–4 hr on a
lawn of OP50. The age-synchronized young larvae (L1) are transferred to plates that
contain either the vehicle and/or other compounds and are maintained at 20°C. The
worms are then transferred to fresh plates on the fourth day after hatching. Once the
worms reach adulthood, they should be transferred daily for six consecutive days
until the cessation of egg laying to avoid overlapping generations. After these six
consecutive days, the worms can then be transferred every other day. Worms are
counted around the same time every day and scored as dead if they do not respond

to a touch stimulus by a platinum wire. This is repeated until all animals have died
(Figure 16.3A).

16.3.1 TARGET IDENTIFICATION AND VALIDATION
An example of using the worms for target identification was described by Ashrafi et
al. using a genome-wide RNAi library to screen for genes that regulate fat storage in
order to identify potential targets for the treatment of obesity.44,45 This approach has
recently been reviewed in depth.18
The AG-expressing worm bearing pathological behavior9 serves as an excellent
example of target validation for AG, which has become well accepted as a disease-
modifying target for AD. The amyloid cascade hypothesis predicts that increased
production, aggregation, and accumulation of AG lead to senile plaques, neuro-
toxicity, and the clinical symptoms of AD. Accordingly, many attempts have been
made to target AG aggregates as a disease-modifying therapeutic strategy in AD.6,46
Increasing experimental evidence suggests that specific oligomeric forms of AG con-
stitute the toxic species.47,48 Thus, specific inhibition of the toxic species of AG is of
importance for therapeutic development of new drugs for AD,43 which was validated
in the transgenic C.elegans.37
16.3.2 MECHANISM OF ACTION USING MUTANT WORMS AND RNAI FEEDING

Using mutant worms can help us determine the mechanism of action of a particular
drug or compound. The insulin/IGF-1 (insulin growth factor-1) pathway has been
well studied in the nematode. Mutations in this pathway have been shown to prolong
the life span of the worms or accelerate aging. Amazingly enough, certain natural
compounds have also been shown to have life-span-extending properties that are
related to caloric restriction, the process of reducing food intake, which has been
shown to increase the life spans of several species. One study demonstrated that
resveratrol, a component found in the skin of grapes, was able to extend the life
span of the worms by activating SIR2-like proteins (sirtuins), a family of NAD+
dependent deacetylases that are evolutionarily conserved from bacteria to humans.49
Mutant worms that lacked this particular gene, SIR2.1, failed to have their life spans
extended in response to resveratrol. In fact, the adult life span of the sir 2.1–null
mutants was not significantly different from that of the wild-type worms. Another
study demonstrated the effect of blueberry polyphenols on C. elegans life span and
used mutants to show that this life span extension required the CaMKII pathway.50
Treatment of several mutants with blueberry phenol, including sir 2.1, the daf-16
mutant, which promotes an increased life span and stress resistance, and skn-1,
which promotes resistance to oxidative stress, all had an increase in longevity. From
these results it was determined that these gene products were not required for the
beneficial effects of the blueberry polyphenols.
Using mutant animals is also a very effective way to determine the site of drug
action of compoundsDWSUHRUSRVWV\QDSWLFFRPSRQHQWV)RUH[DPSOHWKHtph-1
PXWDQWLVGH¿FLHQWLQSUHV\QDSWLFVHURWRQLQELRV\QWKHVLV7KHmod-5 gene encodes


+)$ 45

')%*&'+)$./
7 2

',)* +)%()+,)(*"# ++' 1

"%'+-#*&-$

+)$ 45 +)$ 45 5 53 5 56 33 9 :
'0+)#& ,)'&$0+)#&

'&+)'$ 455 ;8

455 ;8 455 ;8

,%(#&!+%#&

,$+!.*

FIGURE 16.3

WKHRQO\NQRZQSUHV\QDSWLF+7WUDQVSRUWHULQWKHZRUP 6(57 7KHmod-1ser-

1DQGser-4PXWDQWVDOOHQFRGHSRVWV\QDSWLF+7UHFHSWRUVLQC.elegans. If a drug
DFWVSUHV\QDSWLFDOO\WKHQWKHRXWFRPHRIWKH+7FRQWUROOHGHJJOD\LQJEHKDYLRU
would be unchanged in the mod-1 ser-1DQGser-4 mutants fed with that drug. If the
GUXJDFWVSRVWV\QDSWLFDOO\WKHQWKHHJJOD\LQJUHVXOWVZRXOGGLIIHULQWKHVHZRUPV
compared to the wild-type worms.
Addressing whether a particular receptor is necessary for AG-induced patho-
logical behavior can be done in either of two ways: (1) down-regulate the receptor
expression using RNAi followed by testing pathological behavior; or, (2) develop a
new double transgenic strain by crossing a transgenic strain with a receptor mutant
strain, to be used in a defined behavior assay.7KH51$LWHFKQLTXHFDQEHHPSOR\HG
WR NQRFN GRZQ FHUWDLQ JHQHV LQ D 7J C. elegans strain. This newly constructed
strain will allow an examination of the relationship between the receptors and AG
expression by observing their pathological behavioral phenotype.27 RNAi can be
performed in C. elegans by feeding the worms with double-stranded RNA (dsRNA)-
containing bacteria. Knock down rate of the target gene will be confirmed by reverse
transcriptase polymerase chain reaction (RT-PCR). ,I51$LFDQQRWEHFRQ¿UPHG
E\3&5DJUHHQÀXRUHVFHQWSURWHLQ *)3 UHSRUWHUZLOOEHXVHGWRFRQ¿UPUHGXFHG
JHQHH[SUHVVLRQ%HFDXVHRIWKHLQKHUHQWOLPLWDWLRQVRI51$L ZHDNHIIHFWVIRUQHX-
URQDOJHQHVSRVVLELOLW\RIRIIWDUJHWHIIHFWV DOWHUQDWLYH³ORVVRIIXQFWLRQ´H[SHUL-
ments can be validated with genetic mutants by using standard genetic crosses.
16.3.3 RAPID TOXICITY ASSESSMENT FOR PHARMACEUTICAL COMPOUNDS

When interpreting the pharmacological experiments with C. elegans, it is impor-
tant to ensure that the compounds are not toxic to the organism. Lethality can be
determined as a simple signal for toxicity after treatment with specific compounds
FIGURE 16.3 (a) Time course of paralysis assays in CL4176 fed with different drugs. Syn-
chronized eggs of CL4176 C. elegans were maintained at 16°C, on the 35 × 10 mm culture
plates (~ 35 eggs/plate) containing vehicle (Ctrl), EGb 761 (100 μg/mL), or Congo red (CR,
139 μg/mL). The hatched worms were grown for 38 hr at 16°C followed by upshifting the
temperature to 23°C to induce transgene expression. The paralysis was scored at 60-min
intervals. Data are expressed as percentage of nonparalyzed worms from at least three inde-
pendent assays of 100 worms in each experiment. Paralysis in the transgenic strain CL4176 is
caused by AG expression as compared with the control strain CL1175, which does not express
AG transgene (filled triangles). (b) Assay for chemotaxis behavior in neuronal AG-expressing
strain CL2355. Chemotaxis behavior in neuronal strain CL2355 (neuronal AG strain) was
reduced compared with the transgenic control strain CL2122 (no AG strain). Feeding with
100 μg/mL EGb 761 for 4 days alleviated this descent in the transgenic strain, but not on the
control strains (n = 4, *P < 0.05). (c) Pharyngeal pumping assay in C. elegans. N2 worms were
treated with 10–100 μM Epigallocatechin gallate (EGCG) starting from day 1 young larvae
(L1). Pharyngeal pumps were scored every day from adult day 1 until the end of their life span
to assay functional decline of the pharynx muscle (n = 12 animals for each group). The rates
of decline of the muscle contractions in the animals fed with higher concentrations (50 μM
and 100 μM) of EGCG were significantly attenuated.

or their combinations. Toxicity assessment of pharmaceutical compounds using C.

elegans is valid for predicting mammalian toxicity.51 To determine lethality (LD50)
of compounds to C. elegans, survival assays will be conducted31 in a multi-well
microplate in solutions containing concentrations that are 10- to 20-fold of the effec-
tive concentration (1 μM). If no lethality is observed, egg-laying behavior assays will
be performed, since egg laying was much more (30–50 times) sensitive than lethal-
ity in the characterization of toxicity in C. elegans.51,52 In addition, stress response
using GFP-hsp1632,33 can also be used as a more subtle indicator of toxicity. This
approach was successfully employed for fast ranking toxicity of receptor tyrosine
kinase inhibitors.53

Although there are many potential advantages to using C. elegans for drug screen-
ing (particularly for high-throughput studies), there are also scientific and technical
limitations inherent in using C. elegans for this purpose. In the case of the AD
model worms described here, it is reasonable to question whether AG toxicity in
worms is directly (or even indirectly) relevant to AG toxicity in people. Although
the best-supported mechanisms of AG toxicity (e.g., membrane disruption, oxidative
damage, etc.) would be expected to be operational in invertebrates, it is probably
true that the actual relevance of this model will only be known retrospectively, when
the mechanisms of AD pathology are fully understood. Given this scientific limita-
tion, it is clear that drug screening in this model is best employed either for in vivo
characterization of modes of action of compounds already suggested to have activi-
ties in mammalian models, or for broad initial screening of compounds destined to
be further validated in mammalian models. A practical disadvantage of this latter
approach, at least from the perspective of pharmaceutical development, is that the
multiple levels of validation required for compounds found to be active in a worm
model may involve time and expense investments that outweigh the advantages of
this model system.
The technical limitations of using C. elegans as a target organism for drug
development stem primarily from the potential difficulty of getting compounds into
worms and/or knowing what tissue concentration results from a given compound
dose. C. elegans is surrounded by a fairly impermeant cuticle, which is composed
primarily of collagen fibers surrounded by a lipid-rich epicuticle. Although there has
been no comprehensive study of which compounds can or cannot diffuse across the
cuticle, anecdotal evidence suggests that many compounds that readily enter tissue
culture cells do not similarly diffuse into C. elegans. It is also true that compounds
that do not directly diffuse into worms can enter the animals by oral uptake during
feeding. While this may be an effective way for drugs to get into worms, it precludes
establishing any general relationships between compound exposure and resultant
tissue concentrations.
Pharyngeal pumping in C. elegans requires the presence of particulate food,
which is typically E. coli. Levels of worm food can therefore become a variable
affecting compound uptake. In addition, the use of live E. coli as a food source (the
typical approach for C. elegans cultivation) can obviously influence drug exposure if

the bacteria degrade, modifies, or concentrates the added drug. It is therefore impor-
tant that compound exposure tests employ standardized amounts of inviable (but not
lysed) bacteria, and that C. elegans cultures are maintained monoaxenically.

16.5.1 SCREENING COMPOUNDS THAT AFFECT A'-
INDUCED PATHOLOGICAL BEHAVIORS
Using the model organism C.elegans and EGb 761 as a toll, we were able to associate
AG species with AG-induced pathological behaviors. We reported that EGb 761 and
one of its components, ginkgolide A, alleviates AG-induced pathological behaviors,
including paralysis and chemotaxis behavior in a transgenic C. elegans model (Fig-
ure 16.3A and B). Interestingly, feeding the worms with antioxidants did not delay
paralysis, suggesting that reducing oxidative stress is not the mechanism by which
EGb 761 suppresses AG-induced toxicity.37
16.5.2 DOSE-RESPONSE OF EPIGALLOCATECHIN GALLATE

ON PHARYNGEAL PUMPING IN C. ELEGANS
In a series of experiments, EGCG and F-lipoic acid were administered to C. elegans,

and the ability of these antioxidants to modulate several characteristic C. elegans
behaviors was examined, including pharyngeal pumping, chemotaxis behavior, life
span, and amyloid G–associated pathological behavior. We demonstrate that both
antioxidants attenuate the levels of hydrogen peroxide in C. elegans, but their effects
on the behaviors are different. EGCG, but not F-lipoic acid, enhances pharyngeal
pumping behavior in C. elegans in a biphasic dose-dependent manner (Figure 16.3C).
In contrast, F-lipoic acid, but not EGCG, facilitates the chemotaxis behavior in C.
elegans.34
16.5.3 LIFE SPAN EXTENSION BY EGB 761 AND OTHER COMPOUNDS

Given the extremely long life span (4000 years) of the Ginkgo biloba tree, the effect
of EGb 761 and its constituents on life span of C. elegans was examined. In three
independent experiments, with the total number of worms equal to 100, a signifi-
cant difference (P = 0.033) between the survival curves in the worms fed with and
without EGb 761 were observed (Figure 16.4A). EGb 761 moderately increased the
median and maximum life span by 1 day each (less than 10%).32
16.5.4 TESTING COMPOUNDS THAT DELAY AGE-

ASSOCIATED MUSCLE DEGENERATION
As a follow up study of life span extension by EGb 761, pharmacological modulation
of age-dependent muscle degeneration, or sacropenia, was determined. Transgenic
C. elegans strain (PD4251) expressing GFP-MYO-3 protein, localized in body wall
muscles and vulval muscle nuclei, were fed with EGb 761 or Wisconsin Ginseng, and
muscle integrity was analyzed by quantification of GFP fluorescence. Both EGb 761

Life Span Assay Survival Assay

1.0 1.0
No EGb Wild type
Fraction Surviving
EGb No EGb
Fraction Surviving
EGb
0.5 0.5
0.0 0.0
0 1 2 3 4
5 10 15 20
Age (days) Exposure to Juglone (h)
(a) (b)
Age-Dependent Muscle Cell Decline (GFP-Reporter)

Age 4d 7d 10d 15d
Control
70 EGb
Number of Nuclear GFP/Worm
WG
60
*
50 *
40
Day 4 Day 7 Day 10 Day 15

Age (days)
(c)
FIGURE 16.4 (A) Survival curves of the wild-type adult worms, which were fed with E.
coli (OP50) supplemented with either 100 μg/mL EGb 761 (filled circles), or with the vehicles
(ethanol and Tween-80) at appropriate concentrations. Standard errors for each data point
were calculated using the GraphPad Prism software. (B) Age-associated deterioration of C.
elegans body wall muscle shown as green fluorescent protein (GFP) fluorescence declines.
(C) Inset—representative images of the anterior body of the transgenic C. elegans (PD4251)
expressing GFP in the nuclei of body wall muscles on ages indicated. Fluorescent images
were captured in live worms under a 4X objective of the microscope. Graph—quantification
of the number of GFP labeled nuclei in the transgenic worms treated with either vehicle (open
bars), or EGb 761 (filled bars), or WG (dashed bars). The decline of GFP signals in control
worms over time was faster than that of drug-treated animals. Error bars indicate the standard
errors. EGb-treated day 10 animals retain significantly more GFP labeled nuclei of muscle
cells than other animals (*P < 0.05). The delay of the muscle cell deterioration in worms fed
with WG is significantly different from the control animals at day 15 (*P < 0.05).

and Wisconsin Ginseng significantly delayed sarcopenia (Figure 16.4C). Ginseng

was more effective in worms of advanced age. Furthermore, both agents ameliorated
age-associated decline of locomotive behaviors, including locomotion, body bend,
and pharyngeal pumping, suggesting that pharmacological extension of life span is
a consequence of maintaining functional capacity of the tissue, and that C. elegans
is a valid model system for testing therapeutic intervention for delaying the progress
of sarcopenia.35
ACKNOWLEDGMENTS
We would like to thank National Institutes of Health grants for supporting the C.
elegans work (AG012423 to Link lab, and AT001928 to Luo lab). We also thank Dr.
Yves Christen (IPSEN, France) for inspiring the initial studies in the nematode.
REFERENCES
1. Brookmeyer, R., Gray, S., and Kawas, C. 1998. Projections of Alzheimer’s disease in
the United States and the public health impact of delaying disease onset. Am. J. Public
Health 88(9):1337–42.
2. Fillit, H. M., O’Connell, A. W., Brown, W. M., et al. 2002. Barriers to drug discovery
and development for Alzheimer disease. Alzheimer Dis. Assoc. Disord. 16(suppl 1):
S1–8.
3. Hsiao, K. 1998. Transgenic mice expressing Alzheimer amyloid precursor proteins.
Ex.p Gerontol. 33(7–8):883–89.
4. Duff, K., Eckman, C., Zehr, C., et al. 1996. Increased amyloid-beta42(43) in brains of
mice expressing mutant presenilin 1. Natur, 383(6602):710–13.
5. Borchelt, D. R., Ratovitski, T., van Lare, J., et al. 1997. Accelerated amyloid deposition
in the brains of transgenic mice coexpressing mutant presenilin 1 and amyloid precur-
sor proteins. Neuron 19(4):939–45.
408(6815):982–85.
7. Price, D. L., Sisodia, S. S., and Borchelt, D. R. 1998. Alzheimer disease—when and
why? Nat. Genet. 19(4):314–16.
8. Higgins, G. A., and Jacobsen, H. 2003. Transgenic mouse models of Alzheimer’s dis-
ease: Phenotype and application. Behav. Pharmacol. 14(5–6):419–38.
9. Link, C. D. 1995. Expression of human beta-amyloid peptide in transgenic Caenorhab-
ditis elegans. Proc. Natl. Acad. Sci. USA 92(20):9368–72.
10. Iijima, K., Liu, H. P., Chiang, A. S., Hearn, S. A., Konsolaki, M., and Zhong, Y. 2004.
Dissecting the pathological effects of human Abeta40 and Abeta42 in Drosophila: A
potential model for Alzheimer’s disease. Proc. Natl. Acad. Sci. USA 101(17):6623–28.
11. Artal-Sanz, M., de Jong, L., and Tavernarakis, N. 2006. Caenorhabditis elegans: A
versatile platform for drug discovery. Biotechno.l ., (12):1405–18.
12. Link, C. D. 2006. C. elegans models of age-associated neurodegenerative diseases:
Lessons from transgenic worm models of Alzheimer’s disease. Exp. Gerontol.
41(10):1007–13.
13. Link, C. D., Johnson, C. J., Fonte, V., et al. 2001. Visualization of fibrillar amyloid
deposits in living, transgenic Caenorhabditis elegans animals using the sensitive amy-
loid dye, X-34. Neurobiol. Aging 22(2):217–26.

14. Link, C. D., Taft, A., Kapulkin, V., et al. 2003. Gene expression analysis in a transgenic
Caenorhabditis elegans Alzheimer’s disease model. Neurobiol. Aging 24(3):397–413.
15. Hensley, K., Hall, N., Subramaniam, R., et al. 1995. Brain regional correspondence
between Alzheimer’s disease histopathology and biomarkers of protein oxidation. J.
Neurochem. 65(5):2146–56.
16. Kienlen-Campard, P., Miolet, S., Tasiaux, B., and Octave, J. N. 2002. Intracellular amy-
loid-beta 1-42, but not extracellular soluble amyloid-beta peptides, induces neuronal
apoptosis. J. Biol. Chem. 277(18):15,666–70.
17. Drake, J., Link, C. D., and Butterfield, D. A. 2003. Oxidative stress precedes fibrillar
deposition of Alzheimer’s disease amyloid beta-peptide (1-42) in a transgenic Cae-
norhabditis elegans model. Neurobiol. Aging 24(3):415–20.
18. Kaletta, T., and Hengartner, M. O. 2006. Finding function in novel targets: C. elegans
as a model organism. Nat. Rev. Drug Discov. 5(5):387–98.
19. Yatin, S. M., Varadarajan, S., Link, C. D., and Butterfield, D. A. 1999. In vitro and in
vivo oxidative stress associated with Alzheimer’s amyloid beta-peptide (1-42). Neuro-
biol. Aging 20(3):325–30; discussion 339–42.
20. Fay, D. S., Fluet, A., Johnson, C. J., and Link, C. D. 1998. In vivo aggregation of beta-
amyloid peptide variants. J. Neurochem. 71(4):1616–25.
21. Link, C. D. 2005. Invertebrate models of Alzheimer’s disease. Genes Brain Behav.
4(3):147–56.
22. Wu, Y., and Luo, Y. 2005. Transgenic C. elegans as a model in Alzheimer’s research.
Curr. Alzheimer Res. 2(1):37–45.
23. Gouras, G. K., Tsai, J., Naslund, J., et al. 2000. Intraneuronal Abeta42 accumulation in
human brain. Am. J. Pathol. 156(1):15–20.
24. Oddo, S., Caccamo, A., Smith, I. F., Green, K. N., and LaFerla, F. M. 2006. A dynamic
relationship between intracellular and extracellular pools of Abeta. Am. J. Pathol.
168(1):184–94.
25. Link, C. D., Fonte, V., Hiester, B., et al. 2006. Conversion of green fluorescent protein
into a toxic, aggregation-prone protein by C-terminal addition of a short peptide. J.
Biol. Chem. 281(3):1808–16.
26. Hsu, A. L., Murphy, C. T., and Kenyon, C. 2003. Regulation of aging and age-related
disease by DAF-16 and heat-shock factor. Science 300(5622):1142–45.
27. Cohen, E., Bieschke, J., Perciavalle, R. M., Kelly, J. W., and Dillin, A. 2006. Opposing
activities protect against age-onset proteotoxicity. Science 313(5793):1604–10.
28. Luo, Y. 2004. Long-lived worms and aging. Redox Rep. 9(2):65–69.
29. Moy, T. I., Ball, A. R., Anklesaria, Z., Casadei, G., Lewis, K., and Ausubel, F. M. 2006.
Identification of novel antimicrobials using a live-animal infection model. Proc. Natl.
Acad. Sci. USA 103(27):10,414–19.
30. Gottschalk, A., Almedom, R. B., Schedletzky, T., Anderson, S. D., Yates, J. R. III, and
Schafer, W. R. 2005. Identification and characterization of novel nicotinic receptor-
associated proteins in Caenorhabditis elegans. Embo. J. 24(14):2566–78.
31. Crowder, C. M. 2004. Ethanol targets: A BK channel cocktail in C. elegans. Trends
Neurosci. 27(10):579–82.
32. Wu, Z., Smith, J. V., Paramasivam, V., et al. 2002. Ginkgo biloba extract EGb 761
increases stress resistance and extends life span of caenoraibditis elegans. Cell & Mol.
Biol. 48(6):725–31.
33. Strayer, A., Wu, Z., Christen, Y., Link, C. D., and Luo, Y. 2003. Expression of the small
heat-shock protein Hsp16-2 in Caenorhabditis elegans is suppressed by Ginkgo biloba
extract EGb 761. Faseb. J. 17(15):2305–7.
34. Brown, M. K., Evans, J. L., and Luo, Y. 2006. Beneficial effects of natural antioxidants
EGCG and alpha-lipoic acid on life span and age-dependent behavioral declines in
Caenorhabditis elegans. Pharmacol. Biochem. Behav. 85(3):620–28.

35. Cao, Z., Wu, Y., Curry, K.,Wu, Z., Christen, Y., and Luo, Y. 2007. Ginkgo biloba extract
EGb 761 and American ginseng delay sarcopenia in Caenorhabditis elegans. Journal
of Gerotology Biological Sciences 62(12):1337–45.
36. Daigle, I., and Li, C. 1993. apl-1, a Caenorhabditis elegans gene encoding a protein
related to the human beta-amyloid protein precursor. Proc. Natl. Acad. Sci. USA
90(24):12,045–49.
37. Wu, Y., Wu, Z. X., Christen,Y., et al. 2006. Amyloid beta-induced pathological behav-
iors are suppressed by Ginkgo biloba extract Egb 761 and ginkgolides in transgenic
Caenorhabditis elegans. J. Neurosci. 26:13,102–13.
38. Hobert, O. 2003. Behavioral plasticity in C. elegans: Paradigms, circuits, genes. J.
Neurobiol. 54(1):203–23.
39. Sawin, E. R., Ranganathan, R., and Horvitz, H. R. 2000. C. elegans locomotory rate
is modulated by the environment through a dopaminergic pathway and by experience
through a serotonergic pathway. Neuron 26(3):619–31.
40. Schafer, W. R., Sanchez, B. M., and Kenyon, C. J, 1996. Genes affecting sensitivity to
serotonin in Caenorhabditis elegans. Genetics 143(3):1219–30.
41. Ranganathan, R., Sawin, E. R., Trent, C., and Horvitz, H. R. 2001. Mutations in the
Caenorhabditis elegans serotonin reuptake transporter MOD-5 reveal serotonin-
dependent and -independent activities of fluoxetine. J. Neurosci. 21(16):5871–84.
42. Dempsey, C. M., Mackenzie, S. M., Gargus, A., Blanco, G., and Sze, J. Y. 2005. Serotonin
(5HT), fluoxetine, imipramine and dopamine target distinct 5HT receptor signaling to
modulate Caenorhabditis elegans egg-laying behavior. Genetics 169(3):1425–36.
43. Ashrafi, K., Chang, F. Y., Watts, J. L., et al. 2003. Genome-wide RNAi analysis of
Caenorhabditis elegans fat regulatory genes. Nature 421(6920):268–72.
44. Kamath, R. S., Fraser, A. G., Dong, Y., et al. 2003. Systematic functional analysis of the
Caenorhabditis elegans genome using RNAi. Nature 421(6920):231–37.
45. Schenk, D., Barbour, R., Dunn, W., et al. 1999. Immunization with amyloid-beta attenuates
Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400(6740):173–77.
46. Lambert, M. P., Barlow, A. K., Chromy, B. A., et al. 1998. Diffusible, nonfibrillar
ligands derived from Abeta1-42 are potent central nervous system neurotoxins. Proc.
Natl. Acad. Sci. USA 95(11):6448–53.
47. Walsh, D. M., Klyubin, I., Fadeeva, J. V., et al. 2002. Naturally secreted oligomers
of amyloid beta protein potently inhibit hippocampal long-term potentiation in vivo.
Nature 416(6880):535–39.
48. Luo, Y. B. P. 2005. Amylod-G peptide, a therapeutic target for Alzheimer’s disease? In ed.
L. Packer. Oxidative stress and age-related neurodegeneration, chap. 23. Boca Raton,
FL: CRC press.
49. Wood, J. G., Rogina, B., Lavu, S., et al. 2004. Sirtuin activators mimic caloric restric-
tion and delay ageing in metazoans. Nature 430(7000):686–89.
50. Wilson, M. A., Shukitt-Hale, B., Kalt, W., Ingram, D. K., Joseph, J. A., and Wolkow, C.
A. 2006. Blueberry polyphenols increase lifespan and thermotolerance in Caenorhab-
ditis elegans. Aging Cell 5(1):59–68.
51. Dhawan, R., Dusenbery, D. B., and Williams, P. L. 1999. Comparison of lethality,
reproduction, and behavior as toxicological endpoints in the nematode Caenorhabditis
elegans. J. Toxicol. Environ. Health A 58(7):451–62.
52. Anderson, G. L., Boyd, W. A., and Williams, P. L. 2001. Assessment of sublethal end-
points for toxicity testing with the nematode Caenorhabditis elegans. Environ. Toxi-
col. Chem. 20(4):833–38.
53. Dengg, M., and van Meel, J. C. 2004. Caenorhabditis elegans as model system for
rapid toxicity assessment of pharmaceutical compounds. J. Pharmacol. Toxicol. Meth-
ods 50(3):209–14.

17 The Revival of
Scopolamine Reversal
for the Assessment
of Cognition-
Enhancing Drugs
Jerry J. Buccafusco
CONTENTS
17.1 Introduction................................................................................................. 329
17.2 Methods....................................................................................................... 331
17.2.1 Nonhuman Primate Testing ............................................................. 332
17.2.2 Rat Testing ....................................................................................... 333
17.2.3 Statistical Analysis........................................................................... 336
17.3 Scopolamine Reversal................................................................................. 336
17.3.1 In Nonhuman Primates.................................................................... 336
17.3.2 In Rats .............................................................................................. 338
17.4 Discussion ...................................................................................................340
17.4.1 Effects of Scopolamine....................................................................340
17.4.2 Donepezil Reversal of Scopolamine-Induced Decreases in
Task Accuracies ...............................................................................340
References.............................................................................................................. 341
17.1 INTRODUCTION
Atropine and scopolamine are muscarinic receptor antagonists with amnestic prop-
erties that have been used for decades in experimental animals to induce impairment
in their performance of a variety of tasks requiring intact working and reference
memory.1–3 As long as 30 years ago, scopolamine had been used in clinical research
studies (e.g., see4). Scopolamine has also been used clinically (though less frequently
than in past years) as an adjunct to surgical or obstetric procedures to induce seda-
tion and post-procedural amnesia. Since the first reports of a central cholinergic
deficit associated with Alzheimer’s disease, the connection had been made between
the cognitive and memory deficits associated with this disease and the reversible
amnestic effects induced by centrally acting muscarinic cholinergic antagonists.
329

Indeed, blockade of central muscarinic receptors could induce a pattern of cogni-

tive impairment even in young subjects reminiscent of that observed in the aged,
or in individuals with Alzheimer’s disease. For many years, the amnestic action
produced in animals by the administration of centrally acting muscarinic cholin-
ergic antagonists, particularly scopolamine, has been a widely used model for the
characterization of potential cognition-enhancing drugs.5 For animal studies, how-
ever, pharmacological models of induced memory impairment have been suggested
to be of limited value because they fail to replicate the pathological aspects and the
progressive degenerative nature of Alzheimer’s disease.6 Despite this limitation, sco-
polamine-induced memory impairment, particularly when coupled with a version of
the inhibitory avoidance7 task provides a relatively rapid phenotypic screening tool
for drug discovery in the field of cognition enhancement.
A criticism often leveled at what might be termed the scopolamine reversal test
pertains to the lack of versatility in the model, since scopolamine’s actions are limited
to the blockade of brain function mediated via cholinergic (muscarinic) receptors. In
Alzheimer’s disease basal forebrain cholinergic neurons appear to be targeted pri-
marily in early stages of the disease. However, other neurotransmitter systems can be
affected.8,9 Alzheimer’s disease attacks cholinergic neurons in specific brain regions
where the disease process leads to the decreased expression of specific subtypes of
muscarinic cholinergic receptors. Presynaptic M2 receptors (located on cholinergic
basal forebrain projection neurons) are depleted more severely than postsynaptic
M1 receptors.10 Scopolamine, however, is relatively nonselective pharmacologically
with respect to receptor subtypes, and the drug does not discriminate very much
with respect to brain region. Scopolamine certainly would have little direct effect on
non-cholinergic neuronal pathways, although cholinergic neurons have functional
interactions with a wide variety of neurotransmitter systems that could be affected
indirectly by the drug. Despite these limitations, the amnestic responses elicited by
scopolamine in humans appear to mimic very closely the cognitive deficits asso-
ciated with Alzheimer’s disease.5 Another advantage of scopolamine reversal as a
model system for Alzheimer’s disease is that the induced memory impairment is
reversible, which lends itself well to screening methods in drug discovery. The util-
ity of the scopolamine model has been partially validated by the ability of centrally
acting cholinomimetic drugs to reverse the effects of scopolamine on memory task
performance. Studies with the centrally acting cholinesterase inhibitor physostig-
mine demonstrated consistent but partial reversal of scopolamine-induced memory
deficits.11,12 More recent studies with the currently prescribed Alzheimer’s drugs riv-
astigmine, donepezil, and galantamine have reported similar scopolamine reversal
properties in the rat.13–16 However, cholinomimetic drugs are not the only pharmaco-
logical class to reverse scopolamine-induced memory deficits,17–19 attesting perhaps
to the greater than expected versatility of the model.
Although the scopolamine reversal model is in wide use in preclinical stages of
drug development, some examples of the model have been applied to clinical studies
in young or healthy aged subjects. Indeed, a diverse set of compounds with unre-
lated pharmacological properties have shown efficacy in reversing the amnestic and
sedating actions of scopolamine in clinical trials (Table 17.1). In preclinical studies,
perhaps the two most widely used rodent tasks for studying or screening cognition-

The Revival of Scopolamine Reversal for the Assessment 331
TABLE 17.1
Reversal of Scopolamine-Induced Cognitive Impairment by Different
Classes of Pharmacological Agents in Human Trials
Reversing Agent Pharmacological Class Reference
Physostigmine Cholinesterase inhibitor 20
Donepezil Cholinesterase inhibitor 21
Eptastigmine Cholinesterase inhibitor 22
Arecoline Muscarinic receptor agonist 23
Nicotine Nicotinic receptor agonist 24
Choline Nicotinic receptor agonist 23
Caffeine Xanthine 24
Methamphetamine Dopaminergic agonist 12
Estrogen Steroid sex hormone 25
Thyrotropin-releasing hormone (TRH) Neuro(tri)peptide 26
enhancing drugs are the inhibitory avoidance and the water maze tasks. However,
clinical versions of these tasks are not well established. Computer-presented oper-
ant tasks designed for assessing the cognitive deficits associated with Alzheimer’s
disease are available such as the CogState™ product27,28 and the CANTAB™ prod-
uct.29,30 Computer presented cognitive test procedures are becoming more prevalent
in the clinical cognitive testing domain. Therefore, there would be some advantage
to having available preclinical models that are more relevant than those often used
for preclinical drug screening.
17.2 METHODS
In order to achieve the degree of relevance alluded to above, the preclinical approaches
should include a relevant animal species and a relevant behavioral task measuring
higher-level cognitive function, including memory. Translation of clinical efficacy
and drug toxicity from preclinical rodent models to humans has not always resulted
in a reliable degree of predictability for clinical outcome. Considering the complex-
ity of the human brain, and the high genetic homology between nonhuman primates
and human beings, our lab has elected to use macaque monkeys in their performance
of a computer-assisted delayed matching-to-sample (DMTS) task for evaluating new
cognition-enhancing compounds. Operant testing equipment designed for nonhu-
man primates is commercially available (e.g., Monkey CANTAB, http://www.camp-
deninstruments.com/animal_monkeycantab_main.html). We use our own in-house
equipment created from off-the-shelf computer supplies and controlled by propri-
etary software compiled in Microsoft Visual C++ for Windows. The only manufac-
tured component is the aluminum chassis that holds the touch-sensitive screen—a
15-in AccuTouch LCD Panelmount TouchMonitor (Elo Touch Systems, division of
Tyco Electronics, Menlo Park, California, USA)—and a food pellet dispenser unit

FIGURE 17.1 Computer-based nonhuman primate behavior testing equipment. Animal

subjects perform test sessions within their home cage by use of a removable touch-sensitive
screen and pellet dispenser housed within an aluminum chassis.
(ENV-203-300IR, Med Associates, St. Albans, Vermont, USA) with pellet recep-
tacle (Figure 17.1). One computer workstation controls three independent test panels.
We currently run six workstations, allowing the in-cage testing of up to 50 subjects
each weekday.
17.2.1 NONHUMAN PRIMATE TESTING

For the studies described below, male and female rhesus monkeys (Macaca mulatta)
were individually housed at the Animal Behavior Center of the Medical College of
Georgia in stainless steel cages composed of multiple 127 × 71 × 66 cm units. All
procedures employed during this study were reviewed and approved by the Medical
College of Georgia Institutional Animal Care and Use Committee and are consis-
tent with Association for Assessment and Accreditation of Laboratory Animal Care
guidelines. Measures were taken to minimize pain or discomfort in accordance with
the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory
Animals (NIH Publications No. 80-23), revised in 1996. Significant efforts were also
made to minimize the total number of animals used while maintaining statistically
valid group numbers.
Standard Methodology: To promote psychological well-being, toys and foraging
tubes are routinely provided and monkeys are allowed to observe television pro-
grams each afternoon after testing and on weekends. The subjects are well trained
(> 100 individual sessions) in the DMTS task. The animals are maintained on tap
water (unlimited) and standard laboratory monkey chow (Harlan Teklad Laboratory
monkey diet, Madison, Wisconsin, USA) supplemented with fruits and vegetables.
The animals are maintained on a feeding schedule in which all food is removed from

cages at about 0630 hr, and replaced after completion of testing of all subjects for the
day (at about 1630 hr). During testing additional food intake is derived from 300-mg
reinforcement food pellets (commercial composition of standard monkey chow and
banana flakes, Noyes Precision food pellets, P.J. Noyes Co., Lancaster, New Hamp-
shire, USA) obtained during experimental sessions. On weekends animals are fed
without time restrictions. The monkeys are maintained on a 12 hr light-dark cycle
and are tested each weekday between 0900 and 1400 hr. Room temperature and
humidity are maintained at 22oC ± 0.6oC and 52% ± 2%, respectively.
Test Panels Stimuli: Red, blue, or yellow 6 × 10 cm rectangles are presented
on a black background. A trial is initiated by presentation of a sample square com-
posed of one of the three colors. The sample rectangle (located above and centered
between the two choice rectangles) remains in view until the monkey touches the
screen within the borders of the sample rectangle to initiate a preprogrammed delay
(retention) interval. Touching a square gives the illusion that the figure was actually
depressed. Following the delay interval, the two choice rectangles are presented.
One of the two choice colors is presented so that the color of one rectangle matches
the color of the sample stimulus, whereas the color of the other rectangle (incorrect)
is presented as one of the two remaining colors. A correct (matching) choice is rein-
forced. Nonmatching choices are neither reinforced nor punished. The inter-trial
interval is 5 sec and each session consists of 96 trials. The presentation of stimulus
color, choice colors, and choice position are fully counterbalanced so as to relegate
nonmatching strategies to chance levels of accuracy. Three to five different presenta-
tion sequences are rotated through each daily session to prevent the subjects from
memorizing the first several trials. Delay intervals are established during numerous
nondrug or vehicle sessions prior to initiating the study. The duration for each delay
interval is adjusted for each subject until three levels of group performance accuracy
are approximated: zero delay (85%–100% of trials answered correctly); short delay
interval (75%–84% correct); medium delay interval (65%–74% correct); and long
delay interval (55%–64% correct). The assignment of retention intervals based upon
an individual’s baseline task accuracy is necessary to avoid ceiling effects in the
most proficient animals during drug studies, while also serving to ensure that each
animal begins testing at relatively the same level of task difficulty.31 In addition to
session accuracy, two response latencies are also measured: the “sample latency,”
which is the time between presentation of the sample color and the animal touching
the sample rectangle; and the “choice latency,” which is the time between presenta-
tion of the choice rectangles and the animal touching one of the choice rectangles.
Latencies are recorded as the nearest 100th of a sec. For further details see.32,33 A
3-min interval is allowed for the animal to respond after a sample or choice presenta-
tion. Failure to respond initiates the next trial in the sequence. Each trial that does
not receive a response is deemed not applicable and the percent of trials correct is
determined only from the total number of trials actually completed.
17.2.2 RAT TESTING

Male Long-Evans rats, 3 mo of age, are obtained from Harlan Sprague-Dawley, Inc.,
and are initially housed in pairs in polycarbonate cages with Bed-O-Cob bedding

in a temperature-controlled room (25ºC) with a 12-hr light/dark cycle. Upon arrival,

animals are provided with water and food (Teklad rodent feed® or Purina Rat Chow®)
ad libitum. All rats are handled daily for a minimum of 1 wk prior to behavioral test-
ing. One week prior to testing in the radial arm task (RAM), and throughout testing
in the RAM and the delayed stimulus discrimination task (DSDT), the animals are
housed individually and then food restricted to a daily feeding of approximately
80% of their ad libitum consumption. Additional food is provided on weekends and
holidays to maintain the weight of each rat at approximately its freely fed weight. As
with the primate studies, all procedures are reviewed and approved by the Medical
College of Georgia Committee on Animal Use for Research. The rats are trained 5
days per wk, Monday through Friday. Beginning 2 wk after arrival, they are trained
to lever press (without a presented stimulus) on a 30-trial continuous reinforcement
schedule (45 mg pellets, Bioserv) in which the doors are always open and either
lever press is rewarded. Once a particular rat is routinely successful at obtaining 30
pellets, it is graduated to a program in which 32 each of the light and tone configura-
tions are randomly distributed throughout 64 trials. The presentation of stimuli is
fully counterbalanced so as to relegate nonmatching strategies to chance levels of
accuracy. The doors close immediately before the stimulus presentation and remain
closed throughout a 1 sec delay, and only the correct lever selection is rewarded. The
delays are increased to 3 sec and then to 5 sec when accuracy approaches 90%. Once
the animals learn to discriminate correctly during trials involving 5-sec delays (cri-
terion = > 70% accuracy), two additional delays are employed and randomly distrib-
uted with the 5-sec delays throughout the session, for a total of 64 trials. The three
delays are designated as short, medium, and long, and are adjusted empirically (with
the exception of short delays, which are always 5 sec) and vary according to the abil-
ity of each individual rat to perform the task. The general criteria for length of delay
includes an accuracy of approximately 75%–85% correct at short delays, 65%–75%
correct at medium delays, and 55%–65% at long delays. When the rats are able to
maintain DSDT accuracy above chance at long delay intervals of at least 20 sec, with
stable baselines (defined as accuracy within the range defined above for each delay,
for a period of at least 10 consecutive sessions), saline injections are initiated. Daily
saline administration continues until baseline performance is unaffected by injec-
tion (typically 3–5 sessions). The rats are then considered ready for drug studies. In
the groups of rats trained thus far, a 6-mo period of training was necessary prior to
initiation of drug studies.
Rats are tested as described previously34,35 in operant chambers (27.9 × 29.2 × 30.5
cm) that are enclosed in ventilated, sound and light attenuated cubicles (Figure 17.2).
Dim chamber illumination is provided by a small, 3-watt house light located on the
cubicle ceiling. Each chamber is fully computer automated with levers centered on
each side of a feed box and separated 140 mm (center to center) apart. Both the levers
and the feed box are located 35 mm above the 1.6 cm mesh floor. Each lever, one
on the right and one on the left, can be depressed to earn food pellet rewards. Less
than 5 g of force is required to activate each lever. The designation of which lever is
associated with a food reinforcer is determined by the presentation of either a light
or a tone. The animals are trained to discriminate between the light and the tone,
i.e., the reward is provided only on the right side following a trial, which begins with

Delayed Stimulus Discrimination (DSDT)

Task Sequence
2 Doors Closed 2 Doors Closed
Response
Lever
Reward
Closed Tray
Door
Retention
Stimulus
Interval
(Tone or Light)
3 Doors Open 4 Doors Close After 5 sec
Open
Door
Response/Reward Intertrial
Interval (10 sec)
FIGURE 17.2 Schematic representation of the computer-based delayed stimulus discrimi-

nation task for rats. Pneumatically activated doors close during the retention (delay) interval
to prevent the animal from positioning in front of the correct response lever.
the presentation of a light, whereas for following trials beginning with the presenta-
tion of a tone, only a response to the left side is rewarded. The auditory stimulus is
produced by a 1900 Hz, 60–75 dB Sonalert7 device centered 10 mm from the top of
the panel. The visual stimulus is produced by a 6.3 V, 0.25 amp, 11.34 lumen light
located at the opposite end of the cage and centered between and near the top of two
retractable doors. The doors (a novel modification of this type of apparatus) are used
to diminish the ability of animals to position themselves on a particular side near one
of the levers after the stimulus. This arrangement is used to curtail the use of medi-
ating (spatial) strategies other than memory to solve the problem. The duration of
each stimulus is 3 sec. Immediately following the stimulus, variable delay intervals
(one of three), each associated an equal number of times with the light and the tone,
are presented repetitively to comprise a daily test session of 64 trials. During delay
intervals, the retractable doors remain closed, thus preventing access to the levers. At
the end of the delay the doors quickly open, allowing access for lever selection by the
rat. The doors remain open for 5 sec to allow time for the rat to choose a lever and, if
a correct choice is made, consume its reward. Finally the doors are gently closed for
a total inter-trial interval of 10 sec. If an incorrect choice is made, no reward is given
and the next trial is initiated.

If an occasion arises in which DSDT accuracy falls considerably outside the

performance criteria described above on a saline day, drugs will not be administered
on the following day. In addition, a minimum of 2 wk will be allowed as a washout
period between dose-effect studies and repeats of the optimum dose.
17.2.3 STATISTICAL ANALYSIS

DMTS and DSDT data for percent correct are subdivided according to delay inter-
val for each delay component of the session. All statistical analyses are performed
on raw data (percent trials correct). Data are analyzed by use of a multi-factorial
analysis of variance (ANOVA) with repeated measures (SAS, JMP statistical soft-
ware package). An orthogonal multi-comparison t-test is used to compare individual
means. Differences between means from experimental groups are considered sig-
nificant at the P < 0.05 level (2-sided test). Trends toward significance are considered
at the P < 0.10 but > 0.05.
17.3 SCOPOLAMINE REVERSAL

17.3.1 IN NONHUMAN PRIMATES
A series of preliminary experiments were performed to determine the most rele-
vant dose of scopolamine hydrobromide (a tertiary amine derivative of the drug)
to be used in the reversal studies. The drug produced a dose- and delay-dependent
decrease in task accuracies associated primarily with short and medium delay trials.
The dose producing the most reproducible decrement in task accuracy was deter-
mined to be 20 μg/kg. Note that when using scopolamine in any regimen to study its
actions on the central nervous system, particularly on memory, pretesting intervals
of 30–45 min should be used. Shorter pretest intervals can result in dominant periph-
eral anti-muscarinic side effects that can interfere with the animal’s ability to per-
form the task. Many operant tasks are food motivated (as are the tasks described in
this chapter). Scopolamine can decrease salivation, resulting in a loss of palatability
of the food reward. Other anti-muscarinic effects that can potentially interfere with
testing include blurry vision, sedation, increased locomotor activity, and increased
heart rate (also see36). In fact, part of the early controversy in using scopolamine
as a means of inducing memory impairment in animals was based on the apparent
“amnestic” actions of scopolamine methybromide—a quaternary amine analog of
the drug that does not readily enter the central nervous system.37,38
Each subject (N = 9) was tested during several sessions (replicates) of the stan-
dard DMTS task (no drug or vehicle pretreatment); after 20 μg/kg dose of scopol-
amine (vehicle pretreatment); and after various doses of donepezil (5–125 μg/kg)
were administered before 20 μg/kg of scopolamine. The dose-response data shown
in Figure 17.3A are plotted as task accuracy vs. dose of donepezil. Each of the four
delay intervals (zero, short, medium, and long) is shown separately. Horizontal lines
in each graph indicate task accuracies by the same animals with no drug pretreat-
ment. Overall there was a dose- and delay-dependent effect of drug treatment (F21,719
= 3.05, P < 0.0001). The ability of donepezil to partly reverse the decrements in task
accuracy produced by scopolamine was relegated to short delay trials. The effect

100
Accury (% trials correct)
95
90
85 Vehicle
20 μg/kg Scop
80 Best dose + Scop
75
70 100
65 Zero Delay
60
0 20 40 60 80 100 120 95
85
80 90
*
75 *
70
65 85
DMTS Accuracy (% trials correct)

60
55
80
50 Short Delay *
45
0 20 40 60 80 100 120 75
80
75
70 *
70
65
60 65
55
50
45 Medium Delay 60
40
0 20 40 60 80 100 120
55
80
75
70 50
65
60
55 45
Zero Short Medium Long
50
45 Long Delay Delay Interval
40 (b)
0 20 40 60 80 100 120
Dose Donepezil (μg/kg)
(a)
FIGURE 17.3 The ability of donepezil to reverse the decrements in delayed matching-to-
sample (DMTS) task accuracy by scopolamine in rhesus monkeys. Donepezil treatment pre-
ceded scopolamine by 10 min, and testing was initiated 30 min after scopolamine treatment.
(a) In each graph, the horizontal dashed line refers to control sessions in which saline (vehi-
cle) was administered 10 min before saline (vehicle). DMTS testing was always initiated 30
min after scopolamine injection. In each panel the 0 mg/kg dose indicates sessions in which
saline was administered 10 min before 20 μg/kg scopolamine. The remaining sessions were
performed by administering the indicated dose of donepezil 10 min before 20 μg/kg scopol-
amine. Each value represents the mean ± SEM. *Significantly different from respective mean
values in the vehicle (0 μg/kg) group, P < 0.05. (b) The ability of the individual best dose
(most effective in reversing the scopolamine-induced decrement in average task accuracy)
of donepezil to reverse the decrements in DMTS task accuracy by scopolamine in rhesus
monkeys. Donepezil treatment preceded scopolamine by 10 min, and testing was initiated 30
min after scopolamine treatment. Vehicle = control (vehicle–vehicle) sessions. 20 μg/kg Scop
= control scopolamine treatment group with saline pretreatment. Best Dose + Scop = the best
dose of donepezil + 20 μg/kg scopolamine. Each value represents the mean ± SEM. *Signifi-
cantly different from respective mean values in the 20 μg/kg Scop group, P < 0.05.

was statistically significant for the 10 μg/kg (t = 2.4, P = 0.019) and the 50 μg/kg (t =
2.5, P = 0.013) doses of donepezil.
Because of the complexity of the dose-response relationship, an individual best
dose (most effective in reversing the scopolamine decrement in average task accu-
racy) was selected for each subject. The distribution of best doses of donepezil was
as follows: 5 μg/kg—one subject; 10 μg/k—two subjects; 25 μg/kg—three subjects;
50 μg/kg—three subjects; providing an average best dose of 27.8 μg/kg. The accu-
racy data are presented in Figure 17.3B. In this instance, accuracy data are plotted as
a function of delay interval. There was a statistically significant effect of drug treat-
ment (F2,88 = 25.4, P < 0.0001). The decreases in task accuracy produced by scopol-
amine (t = 3.4, P = 0.001) were significantly reversed by the best dose of donepezil
during short (t = 3.2, P = 0.0019) and medium (t = 2.3, P = 0.024) delay trials. The
effect at zero delay trials was nearly significant (t = 1.7, P = 0.095). Even considering
the best dose of donepezil, the drug did not completely reverse the accuracy decre-
ments produced by scopolamine as there was still a significant difference between
the standard DMTS response and the donepezil best dose + scopolamine response
(t = 3.7, P = 0.0004).
17.3.2 IN RATS
For this study 12 rats were trained to criteria in the DSDT. Each rat received every
regimen in the protocol over the course of the study. For each regimen the two injec-
tions were spaced 15 min apart with saline or donepezil always following the sco-
polamine administration. Each regimen was replicated during sessions in which the
set of short delay trials preceded long delay trials, and during sessions in which the
set of long delay trials preceded the short delay trials. Thus each animal received
every drug regimen in the study four times, except for the saline-saline condition,
which was replicated five times over the course of the study to ensure a return to
baseline conditions after each regimen. The sequence of six regimens was random-
ized throughout the study. Long delay intervals for the subjects ranged from 20–60
sec, averaging 34.2 ± 5.2 sec.
The data for the effect of the drug regimens on accuracy in the DSDT are pre-
sented in Figure 17.4. Under control conditions (horizontal dotted lines) and for each
regimen in the study, there was a significant delay-dependent decline in task accuracy
from short (80% trials correct) to long (57% trials correct) delay trials (F1,937 = 245.0,
P < 0.0001). There also was a significant effect of drug treatment as an independent
factor (F1,937 = 59.3, P < 0.0001) and as an interaction with delay interval (F1,937 =
10.9, P < 0.0001). Scopolamine (0.05 mg/kg) produced a dramatic and significant
decrease in both short delay trial accuracy (t = 13.4, P < 0.0001) and long delay trial
accuracy (t = 4.9, P < 0.0001) relative to the saline-saline group. The decrement in
short delay task accuracy produced by scopolamine was significantly inhibited when
each of the four doses of donepezil was added to the regimen (t = 2.2, 5.8, 5.1, and
2.9, respectively, for the 0.05, 1, 2, and 3 mg/kg doses; P < 0.03). These data are plot-
ted as a function of dose of donepezil in Figure 17.4. Though donepezil significantly
reversed the amnestic action of scopolamine, the effect was not complete, amounting

85
DSDT Accuracy (% trials correct)

80
75
*
*
70
*
65
60
55 Short Delay
50
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
60
DSDT Accuracy (% trials correct)
58
56
54
52
50
48
46 Long Delay
44
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Dose Donepezil (mg/kg)
FIGURE 17.4 The effect of donepezil in scopolamine-pretreated rats on delayed stimu-

lus discrimination task (DSDT) accuracy plotted as a function of donepezil dose. In each
graph, the horizontal dashed line refers to control sessions in which saline (vehicle) was
administered 10 min before saline (vehicle). Delayed matching-to-sample (DMTS) testing
was always initiated 30 min after scopolamine injection. In each panel the 0 mg/kg dose
indicates sessions in which saline was administered 10 min before 0.05 mg/kg scopolamine.
The remaining sessions were performed by administering the indicated dose of donepezil 10
min before 0.05 mg/kg scopolamine. Short and Long refer to retention (delay) intervals. Each
value represents the mean ± SEM. *Significantly different from respective mean values in the
0.05 mg/kg Scop group, P < 0.05.
to only about 50% of task accuracy in the saline-saline group. The decrement in long
delay task accuracy produced by scopolamine was not significantly inhibited when
donepezil was added to the regimen, though there was a nearly significant effect in
the group that received the 2 mg/kg dose of donepezil (t = 1.7, P = 0.085).

17.4 DISCUSSION
17.4.1 EFFECTS OF SCOPOLAMINE
In both the monkey DMTS task and the rat DSDT task scopolamine produced pro-
found impairments of task accuracies, particularly during short delay trials. The
test dose used to produce these effects in the rat (50 μg/kg) was only slightly greater
than the dose of scopolamine used in the monkeys (20 μg/kg). The latter dose is
in keeping with the 2–10 μg/kg dose range used in healthy human volunteers.20–
22,26,39 Scopolamine-induced impairment in task acquisition in the proximate stages
of memory formation, i.e., attention and vigilance, is also in keeping with similar
effects reported in human subjects.40–44 In monkeys the memory retention curve
(Figure 17.1B) was shifted to the left, suggesting impaired attention, information
processing, and recall;31 the latter two memory impairments often being character-
istic of Alzheimer’s disease.5,45
17.4.2 DONEPEZIL REVERSAL OF SCOPOLAMINE-

INDUCED DECREASES IN TASK ACCURACIES
The dose of donepezil used in monkeys to reverse the effects of scopolamine was
quite similar (about 70 μg/kg) to that used in an acute dose scopolamine reversal
study in humans.21 We have used a computer-assisted version of the DMTS task for
over 15 yr to study the cognition-enhancing effects of drugs from several pharma-
cological classes in young and aged animals. The paradigm has proven successful
in predicting the utility of novel compounds in terms of clinical applications.46,47
The inclusion of scopolamine in the model provides an added degree of clinical rel-
evance, and the ability to reverse the effects of the muscarinic receptor antagonist is
not limited to cholinergic agonists (Table 17.1). In both the DSDT and DMTS tasks,
donepezil appeared to exert more selective improvement on short delay trials, with
little or no effect on long delay trials. The short delay profile suggests an ability of
the drug to affect the more proximate components of memory. These results suggest
that drugs like donepezil might have different mnemonic actions in normal subjects
or age-impaired subjects as opposed to the drug’s actions in demented subjects.
In none of the experiments did donepezil fully reverse the amnestic action of
scopolamine. Even in the rat studies, where the maximally effective dose of done-
pezil was about 20-fold that of scopolamine, only partial reversal of scopolamine’s
actions was noted. Thus the model confirms the clinical experience that cholines-
terase inhibitor therapy exerts widely variable responses in patient populations, and
that improvement in cognitive status in Alzheimer’s disease is limited. From the
data derived from these experiments it seems reasonable to conclude that scopol-
amine reversal has value as a relevant preclinical model for estimating the thera-
peutic potential of novel cognition-enhancing drugs. For translational effectiveness
the monkey DMTS task could provide a valuable step in determining lead com-
pound status prior to engaging in clinical evaluations, which incur the most risk from
both human safety and financial standpoints. If primate resources are not available,
operant tasks in rats such as the DSDT constitute perhaps the most desirable rodent
models. Useful features of the DSDT task include the ability to induce a mnemonic

demand (increasing delay intervals) and the ability to use the subjects repeatedly
(after washout) and as their own controls. Application of the scopolamine reversal
to both monkey and rodent tasks adds a relevant dimension to already very useful
models.
REFERENCES
1. Heise, G. A. 1984. Behavioral methods for measuring effects of drugs on learning and
memory in animals. Med. Res. Rev. 4:535–38.
2. Bartus, R. T. 1978. Evidence for direct cholinergic involvement in the scopolamine-
induced amnesia in monkeys: Effects of concurrent administration of physostigmine
and methylphenidate with scopolamine. Pharmacol. Biochem. Behav. 9:833–36.
3. Taffe, M. A., Weed, M. R., and Gold, L. H. 1999. Scopolamine alters rhesus mon-
key performance on a novel neuropsychological test battery. Cognitive Brain Res.
8:203–12.
4. Drachman, D. A., and Leavitt, J. 1974. Human memory and the cholinergic system.
Arch. Neurol. 30:113–21.
5. Ebert, U., and Kirch, W. 1998. Scopolamine models of dementia: Electroencephalo-
gram findings and cognitive performance. Eur. J. Clin. Invest. 28:944–49.
6. Van Dam, D., and De Deyn, P. P. 2006. Drug discovery in dementia: The role of rodent
models. Nature Rev. Drug Discovery 5:956–70.
7. Graham, J. H. and Buccafusco, J. J. 2001. Inhibitory avoidance behavior and memory
assessment. In Methods of behavior analysis in neuroscience, ed. Jerry J. Buccafusco.
Boca Raton, FL: CRC Press.
8. Rossor, M., and Iversen, L. L. 1986. Non-cholinergic neurotransmitter abnormalities in
Alzheimer’s disease. Brit. Med. Bull. 42:70–74.
9. Terry, A. V. Jr., Buccafusco, J. J., and Prendergast, M. A. 2003. The cholinergic hypoth-
esis of Alzheimer’s disease: Recent challenges and their implications for novel drug
development. J. Pharmacol. Exp. Ther. 306:821–27.
10. Aubert, I., Araujo, D. M., Cecyre, D., Robitaille, Y., Gauthier, S. and Quirion, R. 1992.
Comparative alterations of nicotinic and muscarinic binding sites in Alzheimer’s and
Parkinson’s diseases. J. Neurochem. 58:529–41.
11. Bartus, R. T. 1978. Evidence for direct cholinergic involvement in the scopolamine-
induced amnesia in monkeys: Effects of concurrent administration of physostigmine
and methylphenidate with scopolamine. Pharmacol. Biochem. Behav. 9:833–36.
12. Mewaldt, S. P., and Ghoneim, M. W. 1978. The effects and interactions of scopolamine,
physostigmine and methamphetamine on human memory. Pharmacol. Biochem.
Behav. 10:205–10.
13. Bejar, C., Wang, R. H., and Weinstock, M. 1999. Effect of rivastigmine on scopol-
amine-induced memory impairment in rats. Eur. J. Pharmacol. 383:231–40.
14. Takahata, K., Minami, A., Kusumoto, H., Shimazu, S., and Yoneda, F. 2005. Effects of
selegiline alone or with donepezil on memory impairment in rats. Eur. J. Pharmacol-
ogy 518:140–44.
15. van der Staay, F. J., and Bouger, P. C. 2005. Effects of the cholinesterase inhibitors
donepezil and metrifonate on scopolamine-induced impairments in the spatial cone
field orientation task in rats. Behav. Brain Res. 156:1–10.
16. de Bruin, N., and Pouzet, B. 2006. Beneficial effects of galantamine on performance
in the object recognition task in Swiss mice: Deficits induced by scopolamine and by
prolonging the retention interval. Pharmacol. Biochem. Behav. 85:253–60.

17. Misane, I., and Ogren, S. O. 2003. Selective 5-HT1A antagonists WAY 100635 and
NAD-299 attenuate the impairment of passive avoidance caused by scopolamine in the
rat. Neuropsychopharmacology 28:253–64.
18. Van Kampen, M., Selbach, K., Schneider. R., Schiegel, E., Boess, F., and Schreiber, R.
2004. AR-R 17779 improves social recognition in rats by activation of nicotinic alpha7
receptors. Psychopharmacology 172:375–83.
19. Lieben, C. K., Blokland, A., Sik, A., Sung, E., van Nieuwenhuizen, P., and Schreiber,
R. 2005. The selective 5-HT6 receptor antagonist Ro4368554 restores memory perfor-
mance in cholinergic and serotonergic models of memory deficiency in the rat. Neuro-
psychopharmacology 30:2169–79.
20. Preston, G. C., Brazell, C., Ward, C., Broks, P., Traub, M., and Stahl, S. M. 1988. The
scopolamine model of dementia: Determination of central cholinomimetic effects of
physostigmine on cognition and biochemical markers in man. J. Psychopharmacol.
2:67–79.
21. Snyder, P. J., Bednar, M. M., Cromer, J. R., and Maruff, P. 2005. Reversal of scopol-
amine-induced deficits with a single dose of donepezil, an acetylcholinesterase inhibi-
tor. J. Alzheimer’s Dis. Dementia 1:126–35.
22. Lines, C. R., Ambrose, J. H., Heald, A., and Traub, M. 1993. A double-blind, placebo-
controlled study of the effects of eptastigmine on scopolamine-induced cognitive defi-
cits in healthy male subjects. Human Psychopharmacol.: Clin. Exp. 8:271–78.
23. Sitaram, N., Weingartner, H., and Gillin, J. C. 1978. Human serial learning: Enhance-
ment with arecholine and choline impairment with scopolamine. Science 201:274–76.
24. Reidel, W., Hogervorst, E., Leboux, R., Verhey, F., van Praag, H., and Jolles, J. 1995.
Caffeine attenuates scopolamine-induced memory impairment in humans. Psycho-
25. Dumas, J., Hancur-Bucci, C., Naylor, M., Sites, C., and Newhouse, P. 2006. Estrogen
treatment effects on anticholinergic-induced cognitive dysfunction in normal post-
menopausal women. Neuropsychopharmacology 31:2065–78.
26. Molchan, S. F., Mellow, A. M., Lawlor, B. A., et al. 1990. TRH attenuates scopolamine-
induced memory impairment in humans. Psychopharmacology 101:84–89.
27. Collie, A., and Maruff, P. 2000. The neuropsychology of preclinical Alzheimer’s dis-
ease and mild cognitive impairment. Neurosci. Biobehav. Rev. 24:365–74.
28. Collie, A., Maruff, P., and Currie, J. 2002. Behavioral characterization of mild cogni-
tive impairment. J. Clin. Exp. Neuropsychol. 24:720–33.
29. Blackwell, A. D., Sahakian, B. J., Vesey, R., Semple, J., Robbins, T. W., and Hodges,
J. R. 2004. Detecting dementia: Novel neuropsychological markers of preclinical
Alzheimer’s disease. Dementia Geriatric Cognitive Disorders 17:42–48.
30. Beglinger, L. J., Tangphao-Daniels, O., Karaken, D., Zhang, L., Mohs, R., and Siemers,
E. R. 2005. Neuropsychological test performance in healthy elderly volunteers before
and after donepezil administration: A randomized controlled study. J. Clin. Psycho-
pharmacol. 25:159–65.
31. Paule, M. G., Bushnell, P. J., Maurissen, J. P.J., et al. 1998. Symposium overview: The
use of delayed matching-to-sample procedures in studies of short-term memory in ani-
mals and humans. Neurotoxicol. Teratol. 20:493–502.
32. Buccafusco, J. J., Terry, A. V. Jr., and Murdoch, P. B. 2002. A computer assisted cogni-
tive test battery for aged monkeys. J. Mol. Neurosci. 19:179–85.
33. Buccafusco, J. J., Jackson, W. J., Terry, A. V. Jr., Marsh, K. C., Decker, M. W., and
Arneric, S. P. 1995. Improvement in performance of a delayed matching-to-sample task
by monkeys following ABT-418: A novel cholinergic channel activator for memory
enhancement. Psychopharmacology 120:256–66.

34. Terry, A. V. Jr., Buccafusco, J. J., Jackson, W. J., Zagrodnik, S., Evans-Martin, F. F., and
Decker, M. W. 1996. Effects of stimulation or blockade of central nicotinic-cholinergic
receptors on performance of a novel version of the rat stimulus discrimination task.
35. Terry, A. V. Jr., Buccafusco, J. J., and Decker, M. W. 1997. Cholinergic channel activator,
ABT-418, enhances delayed-response accuracy in rats. Drug Devel. Res. 40:304–12.
36. Blokland, A. 1995. Acetylcholine: A neurotransmitter for learning and memory? Brain
Res. Brain Res. Rev. 21:285–300.
37. Andrews, J. S., Grutzner, M., and Stephens, D. N. 1992. Effects of cholinergic and
non-cholinergic drugs on visual discrimination and delayed visual discrimination per-
formance in rats. Psychopharmacology 106:523–30.
38. Buxton, A., Callan, O. A., Blatt, E. J., Wong, E. H., and Fontana, D. J. 1994. Choliner-
gic agents and delay-dependent performance in the rat. Pharmacol. Biochem. Behav.
49:1067–73.
39. Ray, P. G., Meador, K. J., Loring, D. W., Zamrini, E. W., Yang, X.-H., and Buccafusco,
J. J. 1992. Central anticholinergic hypersensitivity in aging. J. Geriatric Psychiatry
Neurol. 5:72–77.
40. Wesnes, K., and Warburton, D. M. 1984. Effects of scopolamine and nicotine on human
rapid information processing performance. Psychopharmacology 82:147–50.
41. Dunne, M. P., and Hartley, L. R. 1985. The effects of scopolamine upon verbal mem-
ory: Evidence for an attentional hypothesis. Acta. Psychol. 58:205–17.
42. Preston, G. C., Ward, C., Lines, C. R., Poppleton, P., Haigh, J. R., and Traub, M. 1989.
Scopolamine and benzodiazepine models of dementia: Cross-reversals by Ro 15-1788
and physostigmine. Psychopharmacology 98:487–94.
43. Sunderland, T., Weingartner, H., Cohen, R. M., et al. 1989. Low-dose oral lorazepam
administration in Alzheimer subjects and age-matched controls. Psychopharmacology
99:129–33.
44. Duka, T., Ott, H., Rohloff. A, and Voet, B. 1996. The effects of a benzodiazepine
receptor antagonist beta-carboline ZK-93426 on scopolamine-induced impairment on
attention, memory and psychomotor skills. Psychopharmacology 123:361–73.
45. Jones, R. W., Wesnes, K. A., and Kirby, J. 1991. Effects of NMDA modulation in sco-
polamine dementia. Annals NY Acad. Sci. 640:241–44.
46. Buccafusco, J. J., and Terry, A. V. Jr. 2000. Multiple CNS targets for eliciting beneficial
effects on memory and cognition. J. Pharmacol. Exp. Ther. 295:438–46.
47. Youdim, M. B. H., and Buccafusco, J. J. 2005. CNS targets for multi-functional drugs
in the treatment of neurodegenerative diseases. J. Neural. Transm. 112:519–37.

神经行为学 AVIOR ANALYSIS in NEUROSCIENCE, Second Edition

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

神经行为学 AVIOR ANALYSIS in NEUROSCIENCE, Second Edition

Uploaded by

Copyright:

Available Formats

Methods of

© 2009 by Taylor & Francis Group, LLC

Boca Raton London New York

CRC Press is an imprint of the

© 2009 by Taylor & Francis Group, LLC

© 2009 by Taylor & Francis Group, LLC

No claim to original U.S. Government works

International Standard Book Number-13: 978-1-4200-5234-3 (Hardcover)

Library of Congress Cataloging-in-Publication Data

Methods of behavior analysis in neuroscience / editor, Jerry J. Buccafusco. -- 2nd ed.

Visit the Taylor & Francis Web site at

© 2009 by Taylor & Francis Group, LLC

Acknowledgment and appreciation also goes to my office manager

Finally, I would like to express my appreciation to our excellent animal

© 2009 by Taylor & Francis Group, LLC

Chapter 1 Transgenic Mouse Models of Alzheimer’s Disease: Behavioral

Chapter 2 Cued and Contextual Fear Conditioning for Rodents ........................ 19

Chapter 3 Drug Discrimination .......................................................................... 39

Chapter 4 Conditioned Place Preference ............................................................ 59

Chapter 5 Anxiety-Related Behaviors in Mice................................................... 77

Chapter 6 Behavioral Assessment of Antidepressant Activity in Rodents....... 103

Chapter 7 Assessing Attention in Rodents........................................................ 119

Chapter 8 The Behavioral Assessment of Sensorimotor Processes in the

© 2009 by Taylor & Francis Group, LLC

Chapter 9 Intravenous Drug Self-Administration in Nonhuman Primates ...... 179

Chapter 10 Contextually Induced Drug Seeking During Protracted

Chapter 11 Operant Analysis of Fronto-striatal Function in Rodents................ 215

Chapter 12 Working Memory: Delayed Response Tasks in Monkeys ............... 247

Chapter 13 Spatial Navigation (Water Maze) Tasks ........................................... 267

Chapter 14 Water Maze Tasks in Mice: Special Reference to Alzheimer’s

Chapter 15 Behavioral Neuroscience of Zebrafish ............................................. 293

Chapter 16 Caenorhabditis elegans Model for Initial Screening and

Chapter 17 The Revival of Scopolamine Reversal for the Assessment of

© 2009 by Taylor & Francis Group, LLC

© 2009 by Taylor & Francis Group, LLC

© 2009 by Taylor & Francis Group, LLC

© 2009 by Taylor & Francis Group, LLC

© 2009 by Taylor & Francis Group, LLC

© 2009 by Taylor & Francis Group, LLC

© 2009 by Taylor & Francis Group, LLC

Peter J. Brasted, Ph.D.

Simon Brooks, Ph.D.

Kaitlin E. Browman, Ph.D.

Marishka Brown, B.S.

Kathryn J. Bryan, Ph.D.

Jerry J. Buccafusco, Ph.D.

© 2009 by Taylor & Francis Group, LLC

Philip J. Bushnell, Ph.D.

Gemma Casadesus, Ph.D.

Vincent Castagné, Ph.D.

Daniel T. Cerutti, Ph.D.

Jacqueline N. Crawley, Ph.D.

Máté D. Döbrössy, Ph.D.

Stephen B. Dunnett, Ph.D.

© 2009 by Taylor & Francis Group, LLC

William E. Fantegrossi, Ph.D.

Gerard B. Fox, Ph.D.

Leonard L. Howell, Ph.D.

John R. James, Ph.D.

Hyoung-gon Lee, Ph.D.

Edward D. Levin, Ph.D.

Christopher D. Link, Ph.D.

Yuan Luo, Ph.D.

© 2009 by Taylor & Francis Group, LLC