Accepted Manuscript

Differential use of danger and safety signals in an animal model

of anxiety vulnerability: The behavioral economics of avoidance

Kevin M. Spiegler, Ashley M. Fortress, Kevin C.H. Pang

PII: S0278-5846(17)30611-5
DOI: doi:10.1016/j.pnpbp.2017.11.015
Reference: PNP 9281
To appear in: Progress in Neuropsychopharmacology & Biological Psychiatry
Received date: 26 July 2017
Revised date: 8 November 2017
Accepted date: 18 November 2017

Please cite this article as: Kevin M. Spiegler, Ashley M. Fortress, Kevin C.H. Pang ,
Differential use of danger and safety signals in an animal model of anxiety vulnerability:
The behavioral economics of avoidance. The address for the corresponding author was
captured as affiliation for all authors. Please check if appropriate. Pnp(2017), doi:10.1016/
j.pnpbp.2017.11.015

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT

Differential use of danger and safety signals in an animal model of anxiety vulnerability: The Behavioral
Economics of Avoidance

Kevin M. Spiegler, BA*; Ashley M. Fortress, PhD*; Kevin C.H. Pang, PhD

* These authors contributed equally.

Affiliations: NeuroBehavioral Research Laboratory, Department of Veterans Affairs, New Jersey

T
Health Care System, East Orange, NJ (Fortress, Pang); Graduate School of Biomedical Sciences,

IP
Rutgers Biomedical and Health Sciences, Newark, NJ (Spiegler, Pang); Department of

CR
Pharmacology, Physiology and Neuroscience, New Jersey Medical School, Rutgers Biomedical

US
and Health Science (Pang) AN
Correspondence to: Kevin Pang, VAMC, New Jersey Health Care System, 385 Tremont Avenue,
M

Mailstop 15, East Orange, NJ 07018; kevin.pang@va.gov

ED
PT

Keywords: Negative Reinforcement; Anxiety; Avoidance; Behavioral economics;

motivation
CE
AC
 ACCEPTED MANUSCRIPT

Abstract

Differential processing of danger and safety signals may underlie symptoms of anxiety

disorders and posttraumatic stress disorder. One symptom common to these disorders is

pathological avoidance. The present study examined whether danger and safety signals influence

avoidance differently in anxiety-vulnerable Wistar-Kyoto (WKY) rats and Sprague Dawley (SD)

T
IP
rats. SD and WKY rats were tested in a novel progressive ratio avoidance task with and without

danger or safety signals. Two components of reinforcement, hedonic value and motivation, were

CR
determined by fitting an exponentiated demand equation to the data. Hedonic value of avoidance

US
did not differ between SD and WKY rats, but WKY rats had greater motivation to avoid than SD

rats. Removal of the safety signal reduced motivation to avoid in SD, but not WKY, rats.
AN
Removal of the danger signal did not alter avoidance in either strain. When danger and safety
M

signals were presented simultaneously, WKY rats responded to the danger signals, whereas SD

rats responded to the safety signal. The results provide evidence that 1) safety signals enhance
ED

motivation to avoid in SD rats, 2) both danger and safety signals influence motivation in WKY
PT

rats, and 3) danger signals take precedence over safety signals when presented simultaneously in

WKY rats. Thus, anxiety vulnerability is associated with preferential use of danger signals to
CE

motivate avoidance. The differential use of danger and safety signals has important implications
AC

for the etiology and treatment of pathological avoidance in anxiety disorders and posttraumatic

stress disorder.
 ACCEPTED MANUSCRIPT

1. Introduction

Anxiety disorders are a leading cause of disability with approximately 10% of the global

population affected by anxiety disorders at any given time (Baxter et al., 2014; Kessler et al.,

2009). All of the anxiety disorders, along with anxiety-related disorders such as posttraumatic

stress disorder (PTSD), share a common feature of pathological avoidance behavior (American

T
Psychiatric Association, 2013). While avoidance is typically adaptive by protecting an individual

IP
from potential danger, maladaptive avoidance occurs when it is too prevalent and interferes with

CR
normal functioning. Pathological avoidance predicts the severity of anxiety and treatment

resistance (Foa et al., 2006; Karamustafalioglu et al., 2006), and prevents an individual from

US
experiencing events that may lower anxiety, thereby contributing to the persistence of the
AN
disorder. Because of its role in prolonging anxiety, avoidance behavior is a target for clinical

treatment strategies such as Cognitive Behavioral Therapy (Hofmann & Smits, 2008; Olatunji et
M

al., 2010).
ED

Stimuli reliably occurring with aversive events can be used as predictive cues for danger

or safety (Zvolensky et al., 2000). Neutral stimuli that are associated with the onset of an
PT

aversive event become danger signals, and acquire the negative properties of the aversive events
CE

(Mowrer, 1951; Mowrer, 1960). Stimuli that are associated with the termination of an aversive

event or absence of danger can signal safety (Mowrer, 1951; Mowrer, 1960; Gray, 1987). Thus,
AC

danger signals enhance defensive behaviors, whereas the safety signals reduce these behaviors,

even in the presence of danger signals (summation test for conditioned inhibitors) (Christianson

et al, 2012).

Individuals with anxiety utilize danger and safety signals differently than those without

anxiety disorders. Individuals with anxiety attend more to danger and less to safety signals in
 ACCEPTED MANUSCRIPT

comparison to individuals without anxiety (Taylor & Rachman, 1994). The ineffectiveness of

safety signals in PTSD is demonstrated by the suppression of fear potentiated startle by safety

signals in healthy controls and the lack of suppression in those with PTSD (Jovanovic et al.,

2009; Jovanovic et al., 2010). Currently, the differential use of danger and safety signals in

anxious individuals has been demonstrated using fear conditioned stimuli. While fear is an

T
important component of some symptoms of PTSD, avoidance symptoms can be distinguished

IP
from fear symptoms (i.e. re-occurrence and hypervigilance; APA, 2013). Moreover, the

CR
neurobiology of fear conditioned response (freezing) and avoidance are different (Poremba &

Gabriel, 1999; Choi, et al., 2010). Therefore, the present study investigated whether a danger

US
signal bias was observed in avoidance behavior, a key symptom of anxiety disorders and PTSD.
AN
Danger and safety signals can influence avoidance behavior by acquiring reinforcement
M

properties. The termination of a danger signal can act as conditioned negative reinforcement,

whereas the onset of a safety signal can act as positive reinforcement (Gray, 1987). The literature
ED

supports this idea that both danger and safety signals are independently reinforcing (Taylor &
PT

Rachman, 1994; Dinsmoor, 2001), and can enhance operant avoidance (Fernando et al., 2014;

Christianson et al., 2012). However, it is still unclear whether a danger signal bias in anxiety-
CE

prone individuals translates to differential reinforcement by danger and safety signals, leading to
AC

pathological avoidance.

Positive reinforcement can be separated into hedonic (i.e. liking) and motivational (i.e.,

wanting) components (Berridge & Robinson, 1998; Salamone & Correa, 2012), and we have

recently shown that negative reinforcement is similar (Fragale et al., 2016). The idea of two

components of reinforcement is most attributed to the Incentive Salience Theory (Berridge &

Robsinson, 1998), although others have proposed similar concepts (Salamone et al., 2009). The
 ACCEPTED MANUSCRIPT

hedonic component refers to a physiological response to experiencing the reinforcer (requiring

minimal effort), whereas the motivation component refers to active behavior performed to

acquire the reinforcement (effortful). Because alterations in reinforcement are key features of a

number of psychopathological disorders, characterization of multiple components of

reinforcement may provide unique insights into similarities or differences in reinforcement-

T
related symptoms and their underlying neurobiology between clinical populations (Zald &

IP
Treadway, 2017).

CR
In the present study, a behavioral economic analysis was utilized to differentiate hedonic

and motivational components of reinforcement associated with danger and safety signals in an

US
animal model of anxiety vulnerability. The basic premise of behavioral economics is that
AN
consumption decreases as price increases (Sameulson, 2009), and data analysis allows for the

extraction of two components: Q0 and  (Hursh & Silberberg, 2008). Q0 is the consumption (or
M

demand) at a theoretical price of 0 and is related to the hedonic value of the reinforcer. Demand
ED

elasticity denoted by α relates consumption to price, and is inversely related to motivation.

PT

Behavioral economic analysis has been used to interpret sources of positive reinforcement

(Hursh & Silberberg, 2008; Hursh, 1984; Hursh et al., 1988; Bentzley et al., 2014; Porter-
CE

Stransky, 2015). Recently, we applied behavioral economic analysis to negative reinforcement

where animals could escape shock (Fragale et al., 2016). The hedonic value of shock termination
AC

(Q0 , “relief”) increased as the intensity of the shock increased. Moreover, motivation to escape

shock increased as shock intensity increased (α decreased). The results demonstrate that

negative reinforcement can be separated into hedonic value and motivation using behavioral

economics similar to positive reinforcement.

 ACCEPTED MANUSCRIPT

The Wistar-Kyoto Rat (WKY) is a unique model for anxiety vulnerability. They are

stress sensitive (Redei et al., 1994; De La Garza & Mahoney, 2004) and have a behaviorally

inhibited temperament (Pardon et al., 2002; Pare, 1989; Pare, 1994; Pare & Redei, 1993), which

are risk factors for anxiety disorders (Biederman et al, 1995; Biederman et al., 2001; Hirshfield

et al., 1992). WKY rats demonstrate pathological avoidance in that they acquire lever-press

T
avoidance more rapidly and to a heightened degree compared to a commonly used, outbred

IP
control strain Sprague-Dawley (SD) (Servatius et al., 2008; Beck et al., 2010; Beck et al. 2011;

CR
Jiao et al., 2011). Importantly, WKY rats fail to extinguish avoidance to the same degree as SD

rats when the footshock is removed (Servatius et al., 2008; Beck et al., 2011; Jiao et al., 2011),

demonstrating pathologically persistent avoidance.

US
AN
The present study investigated whether danger and safety signals were differentially

reinforcing in WKY and SD rats. Previous work demonstrated enhanced utilization of danger
M

signals and reduced effectiveness of safety signals in individuals with anxiety disorders, as
ED

assessed by classical conditioning (Jovanovic et al., 2009; Jovanovic et al., 2010). The results of

the present study extend these findings to demonstrate that danger and safety signals motivate
PT

avoidance differently in SD and WKY rats.

CE

2. Methods
AC

2.1 Subjects

Male Sprague Dawley (n = 24) and WKY (n = 24) were obtained from Envigo (Indianapolis,

IN) at 2.5-3 months of age. Three separate groups of rats were used for these studies. The first

group (n=8 per strain) of rats was tested for pain sensitivity, a second group (n=8 per strain) was

used in avoidance procedures to determine the use of danger and safety signals after acquisition
 ACCEPTED MANUSCRIPT

of avoidance, and a third group (n=8 per strain) was used to track signal bias during acquisition.

All rats were individually housed and provided with standard rodent chow and water ad libitum

in a colony room with a 12:12 light/dark cycle (lights on at 7:00 a.m.). All procedures followed

the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were

approved by the Institutional Animal Care and Use Committee at the VA New Jersey Health

T
Care System.

IP
CR
2.2 Behavioral Testing

US
2.2.1 Pain Sensitivity Test

Pain threshold testing occurred in an operant box fitted with grid floors to deliver
AN
scrambled footshock (Coulbourn Instruments, Whitehall, PA). Two thresholds were determined
M

by scoring vocalizations and flinching to incrementing footshock intensities (0.1 mA steps),

starting at 0.1 mA and ending at 1mA (10 s between shocks, 0.5 s shock duration, 10 steps total).
ED

Each shock was repeated twice. Vocalizations and body flinches were scored by two observers
PT

blind to the conditions. A flinch was recorded when the animal visibly tensed its limb muscles

immediately upon receiving the shock. A vocalization was recorded when the animal made an
CE

audible vocalization upon shock delivery. Threshold was determined to be the lowest shock
AC

intensity at which both observers independently agreed a vocalization or flinch was occurred.

2.2.2 Avoidance Acquisition (Standard FR1 procedure)

Rats were trained on a lever press escape-avoidance procedure modified from a

previously described protocol (Servatius et al., 2008; Beck et al., 2010; Beck et al., 2011; Jiao et
 ACCEPTED MANUSCRIPT

al., 2011). Training occurred in operant chambers with grid floors for delivering scrambled

footshock (Coulbourn Instruments). On one chamber wall was a response lever (10.5cm above

the floor), and on the opposite wall was a house light and speaker (26 cm above the floor). Each

operant box was housed in a sound-attenuating chamber and all chambers where controlled by

Graphic State Notation software (Version 2, Coulbourn Instruments).

T
Each trial started with the presentation of a danger signal (75dB tone). Footshocks (1.0

IP
mA, 0.5 s duration, 1 shock/3.5 s) commenced 72 s after the onset of the danger signal. The

CR
danger signal continued into the shock period. A lever press after the start of a trial and prior to

the start of the footshock was considered an avoidance response; this response terminated the

US
danger signal and initiated a 180 s intertrial interval (ITI). The ITI was signaled by a flashing
AN
light (5 Hz), which served as a safety signal. A single lever press after the start of footshock

terminated shock and danger signal and initiated the ITI (escape response). If the lever was not
M

pressed after 20 footshocks (72 s), the danger signal and shocks were terminated and the ITI
ED

commenced. This lack of response was considered a failure. The safety signal was present

during the ITI regardless of whether the animal avoided, escaped or failed the trial. Each
PT

training session consisted of 25 trials. In previous studies, rats were trained for 12 sessions prior
CE

to extinction (Cominski et al., 2014; Jiao et al., 2011). Because we were interested in whether

danger and safety signals contributed to perseverative avoidance during extinction, rats in the
AC

present study were trained for 12 sessions prior to testing in the progressive ratio procedure.

Sessions occurred 3 times per week with a minimum of 48 hours between sessions.

2.2.3 Modified Progressive Ratio Procedure (Standard PR procedure)

 ACCEPTED MANUSCRIPT

The procedure to assess hedonic value and motivation to avoid footshock was a modified

progressive ratio schedule, similar to that described for escape from footshock (Fragale et al.,

2017). Rats were trained in the same operant boxes as in the standard FR1 procedure. The

procedure was identical to that described for the standard FR1 procedure, except the number of

lever presses (price) to avoid footshock was increased in the following manner: 1, 2, 3, 5, 10, 18

T
and 32. Each price had 5 trials in each session, except for the initial price of 1 that consisted of

IP
10 trials. Because SD rats generally show suboptimal avoidance for the first 2-3 trials of a

CR
session (“warm-up”) (Servatius et al., 2008; Beck et al., 2010), the number of trials at the initial

price in the PR procedure was increased to reduce the effect of warm-up in SD rats. The price of

US
an avoidance response was increased regardless of performance. Throughout the session, a
AN
single lever press was necessary to terminate footshock during the shock period (escape

response).
M
ED

2.2.4 Signal Manipulation

Danger and safety signals were manipulated to determine whether SD and WKY rats
PT

utilize these signals differently. The signal manipulations were as follows: 1) PR procedure with
CE

danger, safety or both signals removed to determine the hedonic value and motivation of the

signals in avoidance, 2) FR1 procedure where the safety signal was terminated after the first
AC

minute of the ITI to determine the importance of the safety signal in motivating avoidance

responding, and 3) FR1 procedure where the duration of the danger signal was lengthened to

extend into the first minute of the ITI to assess the relative importance of co-occurring danger

and safety signals in avoidance responding. Because we were interested in how rats were using

the signals under standard FR1 and PR procedures, single probe sessions where the signals were
 ACCEPTED MANUSCRIPT

altered were intermingled with sessions of standard procedures. All comparisons of performance

after signal manipulation were made with respect to performance on the immediately preceding

session with the standard procedure. The order of behavioral testing is shown in Table 1.

2.2.5 Acquisition of Signal Bias

T
To investigate the development of signal biases, a third experiment was conducted. Naïve

IP
SD and WKY rats (n = 8/ strain) were trained in the standard FR1 Procedure for 12 sessions.

CR
Signal bias was evaluated by presenting both danger signal and safety signals together for the

first minute of the ITI (SS + DS trials). These trials represented one third of the trials in session

US
4, 8 and 12 (representing early, middle and late phases of acquisition, respectively). On the other
AN
two thirds of the trials, the standard FR1 procedure was performed (standard trials). The total

number of trials in these three sessions was increased from 25 to 30, with each session consisting
M

of 20 standard trials and 10 SS + DS trials presented in a pseudo-random order. The total number
ED

of lever presses during the first, second, and third minute of the ITI was compared between

standard and SS + DS trials, although the first minute of the ITI was the main comparison. In
PT

order to compare SS + DS trials to standard trials, the number of lever presses occurring on
CE

standard trials were divided by 2 because there were twice as many standard trials as SS + DS

trials.
AC

2.3 Data Analyses

The pain sensitivity test was analyzed using Mann-Whitney U Test. Avoidance

acquisition (FR1 procedure) was assessed in a 2 (strain) x 12 (session) mixed factor ANOVA
 ACCEPTED MANUSCRIPT

(SPSS for Windows, v. 22). FR1 manipulations were analyzed using paired-sample T-tests

during each ITI for each strain (SPSS).

An exponentiated demand equation (Equation 1) was fit to the data from the PR

procedure to determine measures of elasticity α and hedonic value Q0 (Koffarnus et al., 2015).

According to behavioral economics analysis, Q 0 represents hedonic value of shock avoidance

T
and α represents elasticity (inversely related to motivation).

IP
-aQ0C
-1)
Q = Q0 ´10k(e (1)

CR
The best fit of the exponentiated demand equation was performed using a GraphPad Prism (v.

US
6.0h) template modified from Institutes for Behavioral Resources, Inc.

(http://ibrinc.org/software/) (Fragale et al., 2016; Koffarnus et al., 2015). All fits of the equation
AN
to the actual data were excellent (r2 ≥ 0.95). Statistical interpretations of α and Q 0 were assessed

for significance using an Extra Sum of Squares F-test (GraphPad Prism v. 6.0h). Statistical
M

significance was reached when p < 0.05.

ED
PT

3. Results
CE

3.1 Pain Sensitivity

AC

SD and WKY rats were assessed for pain sensitivity to determine whether perception of

pain differed. The lowest shock intensity to elicit a flinch (U = 23.5, p = 0.34); Figure 1A) or

vocalization (U = 20, p = 0.25; Figure 1B) was not significantly different between strains,

suggesting no differences in pain sensitivity between the SD and WKY rats.

3.2 Avoidance Acquisition

 ACCEPTED MANUSCRIPT

WKY rats learned to avoid more rapidly than SD rats on an FR1 lever press avoidance

procedure as evidenced by a main effect of strain (F(1, 168) = 33.19; p <.001) and a strain x session

interaction (F(11, 168) = 1.914, p < .04; Figure 2A). Importantly, a main effect of session revealed

that all rats learned to avoid over the 12 acquisition sessions (F (11, 168) = 17.14, p < .001).

Furthermore, both strains reached the same asymptotic performance, as revealed by an analysis

T
of sessions A10 – A12 (main effect of strain: F(1,14) = .115, p = .74; strain X session interaction:

IP
F(1,28)=.673, p = .52).

CR
US
3.3 Behavioral Economics of Avoidance
AN
3.3.1 Standard PR procedure
M

Values of α for each of the three PR sessions varied by less than 6%; therefore, data from
ED

all 3 sessions were combined. In WKY rats, α values were lower compared to SD rats,

demonstrating that WKY rats were more motivated than SD rats to avoid shock (F(1,10) = 38, p <
PT

.001, Figures 2B, 2C). In contrast, Q 0 values were not different between strains, suggesting that
CE

the hedonic value of avoiding footshock was similar between SD and WKY rats (F (1,10) = 1.9, p =

.198, Figure 2D).

AC

3.3.2 Effect of removing danger or safety signals

Removal of the safety signal reduced motivation to avoid in SD but not WKY rats.

Overall analysis demonstrated significant differences in α (F (3,20) = 33, p < .001). Pairwise

comparisons revealed that removal of the safety signal significantly increased α (decreased
 ACCEPTED MANUSCRIPT

motivation to avoid) in SD (F(1,10) = 92, p < .001), but not WKY rats (F(1,10) = 0.57, p = .47)

(Figure 3A,B). Furthermore, removal of the safety signal did not alter hedonic value (Figure 3A).

The mean Q 0 for SD – Standard, SD – No safety signal, WKY – Standard and WKY – No safety

signal was 1.7 ± .14, 2.3 ± .85, 1.4 ± .05 and 1.3 ± .16, respectively. Q 0 was not different

between strains or conditions (F(3,20) = 1.8, p = .179). These results demonstrate that safety

T
signals contribute in motivating SD rats to avoid footshock, but this influence is not observed in

IP
WKY rats.

CR
Removal of the danger signal in the PR procedure did not alter  or Q 0 for either SD or

US
WKY rats (Figure 4A). Overall analysis demonstrated significant differences in α (F (3,20) = 11, p

< .001). However, the differences were due to strain differences and not due to removal of
AN
danger signals. Comparisons between the standard procedure and the procedure without the
M

danger signal were not different in SD (F (1,10) = 3.2, p = .10) or WKY rats (F(1,10) = 0.81, p = .388)

(Figure 4A,B). However, α was significantly smaller for WKY rats than for SD rats, suggesting
ED

increased motivation to avoid in WKY rats compared to SD rats during both the standard
PT

procedure (F(1,10) = 18, p = .002) and the manipulation without danger signal (F(1,10) = 13, p =

.005). Thus, removal of danger signal did not modify motivation to avoid footshock in either
CE

strain. Additionally, removal of danger signal did not alter hedonic value. Mean Q 0 for SD –
AC

Standard, SD – No danger signal, WKY – Standard and WKY – No danger signal were 1.1 ±

.08, 1.5 ± .27, 1.5 ± .15 and 1.3 ± .1, respectively. Differences in Q 0 were not statistically

different between strains or conditions (F (3,20) = 1.3, p = .291).

3.4 Influence of Danger and Safety Signals in an Avoidance FR1 procedure .

 ACCEPTED MANUSCRIPT

Results from the PR procedures suggest that SD and WKY rats are motivated differently

by danger and safety signals. To confirm these results, danger and safety signals were lengthened

and shortened, respectively, in the FR1 procedure. Terminating the safety signal after the first

minute of ITI did not alter ITI responding during the first minute (ITI-1) in either SD (t(7) = -1.82,

p = .111) or WKY rats (t(7) = 1.01, p = .344). However, the elimination of the safety signal

T
during the latter two minutes of the ITI increased ITI responding in SD rats, ITI-2 (t(7)=-3.93,

IP
p=.006) and ITI-3 (t(7)=-2.494, p=.041) as compared to the standard FR1 procedure (Figure 5B

CR
and Supplemental Figure S4A). In contrast to SD rats, ITI responding in WKY rats were not

affected by absence of the safety signal during ITI-2 (t(7) = .857, p = .42) or ITI-3 (t(7) = -1.246, p

US
= .253) (Figure 5C and Supplemental Figure S4B). These findings support the idea that the
AN
safety signal motivates SD rats but not WKY rats to lever press.
M

The next signal manipulation extended the danger signal into the first minute of the ITI

so that both danger and safety signals were presented during ITI-1. Lever presses during ITI-1
ED

were not altered in SD rats by the co-occurrence of danger and safety signals (t(7) = -0.799, p =
PT

.450; Figure 6B and Supplemental Figure S5A). In contrast, the co-occurrence of danger and

safety signals in ITI-1 caused WKY rats to increase their lever presses, as compared having only
CE

the safety signal present (standard FR1 procedure) (t(7) = -6.561, p < 0.001). During ITI-2, SD
AC

rats responded less in the extended danger signal procedure than in the standard procedure (t(7) =

3.53, p = .01), but this difference was not observed in ITI-3 (t(7) = .38, p = .712). The lengthened

danger signal did not alter responding of WKY rats in ITI-2 (t(7) = .454, p = .663) or ITI-3 (t(7) =

.364, p = .727) (Figure 6C and Supplemental Figure S5B). Together, the results of signal

manipulations in FR1 demonstrate that motivation to avoid in SD rats is enhanced by the

termination of the safety signal, consistent with our PR results. The PR results revealed that
 ACCEPTED MANUSCRIPT

WKY rats are motivated by both danger and safety signals; however, each signal was presented

alone in those sessions. The results from the FR1 procedure suggest that in a session containing

both danger and safety signals, WKY rats are more motivated by the danger signal than the

safety signal when making an avoidance response.

T
IP
3.5 Acquisition of signal bias in FR1 procedure

CR
The next experiment investigated the development of signal bias throughout avoidance

acquisition using the lengthened danger signal manipulation described above (Figure 6A).

US
Although the primary focus is on ITI-1 because this interval is when both safety and danger
AN
signals are presented in SS + DS trials, ITI-2 and ITI-3 are presented here completeness but not

shown in the figures. On session A4, neither SD nor WKY rats responded differently between
M

standard trials (safety signal alone) and SS + DS trials (safety and danger signal) during ITI-1
ED

(SD: t(7) = 1.539. p = .168; WKY: t(7) = 1.170, p = .28), ITI-2, (SD: t(7) = 1.085, p = .314; WKY:

t(7) = .123) or ITI-3 (SD: t(7) = .172, p =867; WKY: t(7) = 1.536, p =.169) (Figure 7). On session
PT

A8, responding during ITI-1 was higher during SS + DS trials than during standard trials in SD
CE

rats (t(7) = 4.793, p = .002), but this different was not observed for ITI-2 (t(7) = 1.000, p = .351) or

ITI-3 (t(7) = 1.657, p = .142). Similar to SD rats, WKY rats responded more during SS + DS
AC

trials compared to standard trials for ITI-1 (t(7) = 3.577, p = .009). Lever pressing of WKY rats

was higher during standard trials compared to SS + DS trials during ITI-2 (t(7) = 2.479, p = .042),

but no differences were seen during ITI-3 (t(7) = 1.700, p = .133). These results suggest both

strains had distinguished danger from safety signals, but were still responding to the danger

signal during SS + DS trials in session A8. On session A12, SD rats no longer showed increased

responding during SS + DS trials in ITI-1 (t(7) = .296, p = .776), demonstrating that safety signals
 ACCEPTED MANUSCRIPT

were now the predominant force in guiding responses when both safety and danger signals were

present. No differences were observed during ITI-2 (t(7) = 1.193, p = .272) and ITI-3 (t(7) =

1.861, p = .105) in SD rats. In contrast to SD rats, WKY rats continued to respond more during

SS + DS trials than standard trials during ITI-1 of session A12 (t(7) = 4.736, p = .002),

demonstrating that for WKY rats the danger signal was the main cue in guiding responses when

T
both signals were present. Responding during SS + DS and standard trials were not different for

IP
ITI-2 (t(7) = 1.492, p = .179) or ITI-3 (t(7) = 1.560, p =.163) in WKY rats. These data indicate that

CR
the transition from a danger to safety signal bias seen in SD rats is acquired relatively late in

training and may be impaired in WKY rats.

US
AN
4. Discussion

Avoidance behavior is a hallmark of anxiety disorders, as it heavily contributes to the

M

pathology and chronicity of these disorders. Anxiety disorders may develop from the over-
ED

prediction of danger and the underutilization of safety signals (Taylor & Rachman, 1994;

Jovanovic et al., 2009; Jovanovic et al., 2010). Differential use of danger and safety signals may
PT

result in pathological avoidance, increasing risk for developing an anxiety disorder. One way
CE

that these signals could influence avoidance behavior is through acquisition of secondary

reinforcement properties (Woody & Rachman, 1994). Therefore, the goals of the present study
AC

were two-fold. The first goal was to determine whether avoidance responding was more

reinforcing to an animal model of anxiety vulnerability, the WKY rat, as compared to an outbred

SD rat. The second goal was to determine the contribution of danger and safety signals to

avoidance behavior in WKY rats as compared to SD rats.

 ACCEPTED MANUSCRIPT

The current study employed a novel progressive ratio paradigm with behavioral

economics analysis to separate hedonic value and motivational aspects of avoidance of footshock

(Hursh & Silberberg, 2008; Koffarnus et al., 2015). Hedonic value of avoidance was not

different between SD and WKY rats in any of our experiments. However, WKY rats were more

motivated to avoid footshock than outbred SD rats. Previously, we found WKY rats were more

T
motivated to escape footshock than SD rats (Fragale et al., 2016). The enhanced motivation in

IP
WKY rats was not due to differences in pain sensitivity, as threshold for flinching and vocalizing

CR
to footshock were similar between strains. This pattern of results is similar to anxiety disorders

in which individuals are more preoccupied with threat than individuals without anxiety (Bar-

Haim et al., 2007).

US
AN
WKY have a behaviorally inhibited temperament, as displayed through reduced mobility
M

in the Forced Swim Task, Elevated Plus Maze, and Open Field (Pardon et al., 2002; Pare, 1989;

Pare, 1994; Pare & Redei, 1993). However, we have consistently found increased activity in
ED

WKY rats in our avoidance tasks compared to SD rats (Servatius et al., 2008; Beck et al., 2010;
PT

Beck et al. 2011; Jiao et al., 2011). In support, WKY rats in the present study were more

motivated to lever press to avoid footshock. This inconsistency may be explained by the ability
CE

to actively escape the stressor in avoidance. Further, motivation to lever press in WKY rats
AC

depends on the valence of the reinforcement. They display a negative bias, as they are more

motivated than SD rats to escape an aversive stimulus, yet are less motivated than SD rats to

acquire an appetitive stimulus (Fragale et al., 2016). A negative bias is similar expressed in

those with anxiety disorders (Eysenk et al., 2007).

In SD rats, the safety signal was more influential to avoidant behavior than the danger

signal. Removal of the safety signal reduced motivation to avoid in SD rats, providing evidence
 ACCEPTED MANUSCRIPT

that the safety signal had acquired reinforcing value and motivated SD rats to make an avoidance

response. By implication, this result also suggests that the danger signal acquired little to no

reinforcement properties as it was insufficient to maintain avoidance responding without the

safety signal. This conclusion is supported by the finding that removal of the danger signal did

not alter motivation in SD rats. Results from signal manipulation in the FR1 procedure also

T
support this conclusion. SD rats increased lever presses when the safety signal terminated

IP
prematurely during the ITI, suggesting that the safety signal denotes a safe period to SD rats, a

CR
time where lever pressing is not necessary. Further evidence that the safety signal represents a

time of safety is the finding that extending the danger signal into the first minute of the ITI did

US
not result in enhanced responding, presumably due to the simultaneous presentation of the safety
AN
signal. Interestingly, this use of the safety signal developed relatively late in the avoidance

acquisition process, somewhere between session 8 and 12 when avoidance acquisition was
M

asymptotic.
ED

On the surface, it may appear that the reduced responding after removal of the safety
PT

signal in our modified PR procedure is at odds with the enhanced responding when the safety

signal is shortened in the avoidance FR1 procedure. Our interpretation is that the reduced
CE

responding in a session without safety signal is due to decreased reinforcement, similar to an

AC

extinction procedure. Conversely, the enhanced responding when the safety signal is shortened

in the FR1 procedure is due to the frustration effect caused by a reduction of reinforcement due

to premature termination of the reinforcer (i.e. safety signal) (Amsel & Roussel, 1952; Bower,

1962). Together, our results point to an enhanced influence of the safety signal over that of the

danger signal in avoidance behavior of the SD rat.

 ACCEPTED MANUSCRIPT

In the anxiety vulnerable WKY rat, the influence of the danger signal predominated.

Unlike SD rats, presentation of the danger signal after removal of the safety signal was

sufficiently motivating to maintain avoidance responding in WKY rats. Similar to SD rats, the

safety signal was also motivating. Although removal of either safety or danger signals did not

alter avoidance responding, removal of both danger and safety signals reduced motivation for

T
WKY, as well as SD, rats, demonstrating that WKY rats could use either signal for optimal

IP
avoidance (Supplemental Figure S2). Evidence for preferential influence of the danger signal

CR
compared to the safety signal was obtained by manipulating the signals in the FR1 procedure.

WKY rats were unaffected by the shortening of the safety signal during the ITI. Furthermore,

US
extending the danger signal into the first minute of the ITI increased lever pressing in WKY rats
AN
despite the presence of the safety signal. In contrast to the SD rats, the bias for the danger signal

remained through 12 sessions of FR1 acquisition. These results provide evidence that the
M

influence of the danger signal predominates over the safety signal in WKY rats. Thus, in contrast
ED

with SD rats, WKY rats preferentially use danger signals in avoidance.

PT

A bias for danger signals in WKY rats is consistent with clinical populations. Individuals

with anxiety disorders express hypervigilance and exaggerated attention toward threat (Mogg &
CE

Bradley, 1998). It has been postulated that anxiety patients may underutilize safety signals in the
AC

presence of danger due to information processing bias, where danger signals take precedence

over safety signals (Telch et al., 1994). Processing bias toward threat may be a core component

across anxiety disorders (Bar-Haim et al., 2007) and may in fact cause anxiety- vulnerability

(Mathews & MacLeod, 2002). Moreover, low perceived safety is a risk factor for anxiety

disorders (Grieger et al., 2003; Grieger et al., 2004).

 ACCEPTED MANUSCRIPT

Safety signals reduce fear potentiated startle in healthy individuals, but the effect of

safety signal is impaired in PTSD patients (Jovanovic et al., 2009; Jovanovic et al., 2010). In

another study, high trait anxiety was associated with impaired ability of safety signals to reduce

shock expectancy, even though safety signals suppressed fear potentiated startle (Kindt & Soeter,

2014). Thus, the appreciation of safety signals may be impaired in high trait anxiety, even though

T
the influence of safety signals on reflex responses are normal. The present study extends these

IP
findings on reflex responses to demonstrate an anxiety-related bias toward danger signals in

CR
operant avoidance behavior. Given these results, anxiety disorders and PTSD may be associated

with an impaired ability to acquire conditioned inhibition. WKY rats continued lever pressing in

US
response to the danger signal despite the simultaneous presence of the safety signal, whereas in
AN
SD rats the presence of the safety signal to negated the influence of the danger signal. The

inability of the anxiety vulnerable to use the safety signal may contribute to the apparent bias
M

toward danger signals, and may worsen anxiety pathology.

ED

An alternative explanation of our results is that the differential use of danger and safety
PT

signals in the two strains is due to overtraining in WKY rats. The WKY rats learned the

avoidance response more rapidly than SD rats, although SD rats eventually reached the same
CE

asymptotic performance. Safety signals are theoretically learned after danger signals, as an
AC

individual must learn that something is dangerous before learning that a signal represents safety

from the danger stimulus (Lohr, 2007), and support for this idea is presented in our last

experiment (Figure 7). It would therefore be expected that the facilitated acquisition WKY rats

would enhance their use of safety signals compared to SD rats. Our findings do not support this

expectation. To further address this possibility, the SD rats were divided into the four best and

four worst avoidance learners. The best SD learners acquired the avoidance response similar to
 ACCEPTED MANUSCRIPT

WKY rats (Supplemental Figure S1). The best and worst SD learners did not differ in

motivation to avoid after removal of safety or danger signals, and both groups of SD rats were

different than WKY rats. Thus, the differential acquisition of avoidance is unlikely to account

for the strain differences in safety and danger signal use.

The present research is limited in that only male rats were investigated. Anxiety

T
IP
disorders are diagnosed more in females than in males (Baxter et al., 2014; Kessler et al., 2009;

CR
Tolin & Foa, 2006). Additionally, female rats acquire avoidance responses more readily than

male rats (Beck et al., 2010; Avcu et al., 2014) and a computational model suggest females

US
utilize safety signals differently than male rats (Radell et al., 2015). Given the results of the

present study, the role of danger and safety signals in avoidance in females should be
AN
investigated in future studies.
M

5. Conclusion
ED

WKY rats displayed increased motivation to avoid footshock as compared to SD rats

PT

despite valuing the hedonic aspect of avoidance similarly. Danger and safety signals contributed

to avoidance differently in SD and WKY rats. SD rats were most influenced by safety signals,
CE

whereas WKY rats were most influenced by danger signals. These results are the first
AC

demonstration of an anxiety-related danger signal bias in avoidance behaviors and suggest a

danger signal bias may motivate perseverative avoidance and be a vulnerability factor for the

development of anxiety disorders. The present study lays the groundwork to evaluate the role of

danger and safety signals in negative reinforcement and operant avoidance that can be pursued in

other animal models of anxiety and in clinical populations with anxiety disorders.
 ACCEPTED MANUSCRIPT

Contributors: K. Spiegler, A. Fortress and K. Pang designed the study. K. Spiegler acquired the

data, which K. Spiegler, A. Fortress and K. Pang analyzed. All authors wrote, reviewed and

approved the article for publication.

T
IP
Acknowledgements: This work was supported by the Biomedical Laboratory Research &

CR
Development and Department of Veterans Affairs Office of Research & Development [grants

I01BX000132 to KP and 5IK2BX003196 to AF]. Ian Smith and Jacquelyn Tomaio are

US
acknowledged for their assistance in data acquisition. The views expressed in this article are
AN
those of the authors and do not necessarily reflect the position or policy of the Department of

Veterans Affairs or the United States government.

M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

References

American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders
(DSM-5®). American Psychiatric Pub.

Amsel, A., Roussel, J., 1952. Motivational properties of frustration: I. Effect on a running
response of the addition of frustration to the motivational complex. J. Exp. Psychol. 43, 363.

Avcu, P., Jiao, X., Myers, C.E., Beck, K.D., Pang, K.C., Servatius, R.J., 2014. Avoidance as

T
expectancy in rats: sex and strain differences in acquisition. Frontiers in behavioral neuroscience

IP
8, 334.

CR
Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M.J., Van Ijzendoorn, M.H.,
2007. Threat-related attentional bias in anxious and nonanxious individuals: a meta-analytic
study. Psychol. Bull. 133, 1.

US
Baxter, A., Vos, T., Scott, K., Ferrari, A., Whiteford, H., 2014. The global burden of anxiety
disorders in 2010. Psychol. Med. 44, 2363-2374.
AN
Beck, K.D., Jiao, X., Pang, K.C., Servatius, R.J., 2010. Vulnerability factors in anxiety
determined through differences in active-avoidance behavior. Prog. Neuro-Psychopharmacol.
Biol. Psychiatry 34, 852-860.
M

Beck, K.D., Jiao, X., Ricart, T.M., Myers, C.E., Minor, T.R., Pang, K.C., Servatius, R.J., 2011.
ED

Vulnerability factors in anxiety: strain and sex differences in the use of signals associated with
non-threat during the acquisition and extinction of active-avoidance behavior. Prog. Neuro-
Psychopharmacol. Biol. Psychiatry 35, 1659-1670.
PT

Bentzley, B.S., Jhou, T.C., Aston-Jones, G., 2014. Economic demand predicts addiction- like
behavior and therapeutic efficacy of oxytocin in the rat. Proc. Natl. Acad. Sci. U. S. A. 111,
CE

11822-11827.

Berridge, K.C., Robinson, T.E., 1998. What is the role of dopamine in reward: hedonic impact,
reward learning, or incentive salience? Brain Res. Rev. 28, 309-369.
AC

Biederman, J., Hirshfeld-Becker, D.R., Rosenbaum, J.F., Hérot, C., Friedman, D., Snidman, N.,
Kagan, J., Faraone, S.V., 2001. Further evidence of association between behavioral inhibition
and social anxiety in children. Am. J. Psychiatry 158, 1673-1679.

Biederman, J., Rosenbaum, J.F., Chaloff, J., Kagan, J., 1995. Behavioral inhibition as a risk
factor for anxiety disorders. .

Bower, G.H., 1962. The influence of graded reductions in reward and prior frustrating events
upon the magnitude of the frustration effect. J. Comp. Physiol. Psychol. 55, 582.
 ACCEPTED MANUSCRIPT

Choi, J.S., Cain, C.K., LeDoux, J.E., 2010. The role of amygdala nuclei in the expression of
auditory signaled two-way active avoidance in rats. Learn. Mem. 17, 139-147.

Christianson, J.P., Fernando, A.B., Kazama, A.M., Jovanovic, T., Ostroff, L.E., Sangha, S.,
2012. Inhibition of fear by learned safety signals: a mini-symposium review. J. Neurosci. 32,
14118-14124.

Cominski, T.P., Jiao, X., Catuzzi, J.E., Stewart, A.L., Pang, K.C., 2014. The role of the
hippocampus in avoidance learning and anxiety vulnerability. Frontiers in behavioral
neuroscience 8, 273.

T
IP
De La Garza, R., Mahoney, J.J., 2004. A distinct neurochemical profile in WKY rats at baseline
and in response to acute stress: implications for animal models of anxiety and depression. Brain

CR
Res. 1021, 209-218.

Dinsmoor, J.A., 2001. Stimuli inevitably generated by behavior that avoids electric shock are

US
inherently reinforcing. J. Exp. Anal. Behav. 75, 311-333.

Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G., 2007. Anxiety and cognitive
performance: attentional control theory. Emotion 7, 336-353.
AN
Fernando, A.B., Urcelay, G.P., Mar, A.C., Dickinson, A., Robbins, T.W., 2014. Safety signals as
instrumental reinforcers during free-operant avoidance. Learn. Mem. 21, 488-497.
M

Foa, E.B., Stein, D.J., McFarlane, A.C., 2006. Symptomatology and psychopathology of mental
ED

health problems after disaster. J. Clin. Psychiatry 67, 15-25.

Fragale, J.E., Beck, K.D., Pang, K.C.H., 2017. Use of the exponential and exponentiated demand
equations to assess the behavioral economics of negative reinforcement. Front. Neurosci. 11, 77.
PT

Gray, J.A., 1987. The Psychology of Fear and Stress. CUP Archive.
CE

Grieger, T.A., Fullerton, C.S., Ursano, R.J., 2004. Posttraumatic stress disorder, depression, and
perceived safety 13 months after September 11. Psychiatric Services 55, 1061-1063.
AC

Grieger, T.A., Fullerton, C.S., Ursano, R.J., Reeves, J.J., 2003. Acute stress disorder, alcohol
use, and perception of safety among hospital staff after the sniper attacks. Psychiatric Services
54, 1383-1387.

Hirshfeld, D.R., Rosenbaum, J.F., Biederman, J., Bolduc, E.A., Faraone, S.V., Snidman, N.,
Reznick, J.S., Kagan, J., 1992. Stable behavioral inhibition and its association with anxiety
disorder. Journal of the American Academy of Child & Adolescent Psychiatry 31, 103-111.

Hofmann, S.G., Smits, J.A., 2008. Cognitive-behavioral therapy for adult anxiety disorders: a
meta-analysis of randomized placebo-controlled trials. J. Clin. Psychiatry 69, 621-632.
 ACCEPTED MANUSCRIPT

Hursh, S.R., 1984. Behavioral economics. J. Exp. Anal. Behav. 42, 435-452.

Hursh, S.R., Raslear, T.G., Shurtleff, D., Bauman, R., Simmons, L., 1988. A cost‐benefit
analysis of demand for food. J. Exp. Anal. Behav. 50, 419-440.

Hursh, S.R., Silberberg, A., 2008. Economic demand and essential value. Psychol. Rev. 115,
186.

Jiao, X., Paré, W.P., Tejani-Butt, S., 2003. Strain differences in the distribution of dopamine
transporter sites in rat brain. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 27, 913-919.

T
IP
Jovanovic, T., Keyes, M., Fiallos, A., Myers, K.M., Davis, M., Duncan, E.J., 2005. Fear
potentiation and fear inhibition in a human fear-potentiated startle paradigm. Biol. Psychiatry 57,

CR
1559-1564.

Jovanovic, T., Norrholm, S.D., Blanding, N.Q., Davis, M., Duncan, E., Bradley, B., Ressler,

US
K.J., 2010. Impaired fear inhibition is a biomarker of PTSD but not depression. Depress. Anxiety
27, 244-251.

Jovanovic, T., Norrholm, S.D., Fennell, J.E., Keyes, M., Fiallos, A.M., Myers, K.M., Davis, M.,
AN
Duncan, E.J., 2009. Posttraumatic stress disorder may be associated with impaired fear
inhibition: relation to symptom severity. Psychiatry Res. 167, 151-160.
M

Karamustafalioglu, O.K., Zohar, J., Güveli, M., Gal, G., Bakim, B., Fostick, L.,
Karamustafalioglu, N., Sasson, Y., 2006. Natural course of posttraumatic stress disorder: A 20-
ED

month prospective study of Turkish earthquake survivors. J. Clin. Psychiatry .

Kessler, R.C., Aguilar-Gaxiola, S., Alonso, J., Chatterji, S., Lee, S., Ormel, J., Üstün, T.B.,
PT

Wang, P.S., 2009. The global burden of mental disorders: an update from the WHO World
Mental Health (WMH) surveys. Epidemiol. Psichiatr. Soc. 18, 23-33.
CE

Kindt, M., Soeter, M., 2014. Fear inhibition in high trait anxiety. PloS one 9, e86462.

Koffarnus, M.N., Franck, C.T., Stein, J.S., Bickel, W.K., 2015. A modified exponential
behavioral economic demand model to better describe consumption data. Exp. Clin.
AC

Psychopharmacol. 23, 504.

Lohr, J.M., Olatunji, B.O., Sawchuk, C.N., 2007. A functional analysis of danger and safety
signals in anxiety disorders. Clin. Psychol. Rev. 27, 114-126.

Mathews, A., MacLeod, C., 2002. Induced processing biases have causal effects on anxiety.
Cognition & Emotion 16, 331-354.

Mogg, K., Bradley, B.P., 1998. A cognitive- motivational analysis of anxiety. Behav. Res. Ther.
36, 809-848.
 ACCEPTED MANUSCRIPT

Mowrer, O., 1960. Learning theory and behavior. .

Mowrer, O.H., 1951. Two-factor learning theory: summary and comment. Psychol. Rev. 58, 350.

Olatunji, B.O., Cisler, J.M., Deacon, B.J., 2010. Efficacy of cognitive behavioral therapy for
anxiety disorders: a review of meta-analytic findings. Psychiatr. Clin. North Am. 33, 557-577.

Pardon, M., Gould, G., Garcia, A., Phillips, L., Cook, M., Miller, S., Mason, P., Morilak, D.,
2002. Stress reactivity of the brain noradrenergic system in three rat strains differing in their
neuroendocrine and behavioral responses to stress: implications for susceptibility to stress-

T
related neuropsychiatric disorders. Neuroscience 115, 229-242.

IP
Paré, W.P., 1994. Open field, learned helplessness, conditioned defensive burying, and forced-

CR
swim tests in WKY rats. Physiol. Behav. 55, 433-439.

Paré, W.P., 1989. Stress ulcer susceptibility and depression in Wistar Kyoto (WKY) rats.

US
Physiol. Behav. 46, 993-998.

Pare, W., Redei, E., 1993. Depressive behavior and stress ulcer in Wistar Kyoto rats. Journal of
Physiology-Paris 87, 229-238.
AN
Poremba, A., Gabriel, M., 1999. Amygdala neurons mediate acquisition but not maintenance of
instrumental avoidance behavior in rabbits. J. Neurosci. 19, 9635-9641.
M

Porter‐Stransky, K.A., Bentzley, B.S., Aston‐Jones, G., 2015. Individual differences in orexin‐I
ED

receptor modulation of motivation for the opioid remifentanil. Addict. Biol. .

Radell, M.L., Beck, K.D., Pang, K.C., Myers, C.E., 2015. Using signals associated with safety in
PT

avoidance learning: computational model of sex differences. PeerJ 3, e1081.

Redei, E., Pare, W.P., Aird, F., Kluczynski, J., 1994. Strain differences in hypothalamic-
CE

pituitary-adrenal activity and stress ulcer. Am. J. Physiol. 266, R353-60.

Salamone, J.D., Correa, M., 2012. The mysterious motivational functions of mesolimbic
dopamine. Neuron 76, 470-485.
AC

Salamone, J.D., Correa, M., Farrar, A.M., Nunes, E.J., Pardo, M., 2009. Dopamine, behavioral
economics, and effort. Frontiers in behavioral neuroscience .

Samuelson, P., 2009. Economics: An Introductory Analysis, with Nordhaus WD (since 1985),
1948.

Servatius, R., Jiao, X., Beck, K., Pang, K., Minor, T., 2008. Rapid avoidance acquisition in
Wistar–Kyoto rats. Behav. Brain Res. 192, 191-197.
 ACCEPTED MANUSCRIPT

Taylor, S., Rachman, S., 1994. Stimulus estimation and the overprediction of fear. British
Journal of Clinical Psychology 33, 173-181.

Telch, M.J., Valentiner, D., Bolte, M., 1994. Proximity to safety and its effects on fear prediction
bias. Behav. Res. Ther. 32, 747-751.

Tolin, D.F., Foa, E.B., 2006. Sex differences in trauma and posttraumatic stress disorder: a
quantitative review of 25 years of research. Psychol. Bull. 132, 959.

Woody, S., Rachman, S., 1994. Generalized anxiety disorder (GAD) as an unsuccessful search

T
for safety. Clin. Psychol. Rev. 14, 743-753.

IP
Zald, D.H., Treadway, M.T., 2017. Reward Processing, Neuroeconomics, and Psychopathology.

CR
Annual Review of Clinical Psychology 13.

Zvolensky, M.J., Lejuez, C., Eifert, G.H., 2000. Prediction and control: Operational definitions

US
for the experimental analysis of anxiety. Behav. Res. Ther. 38, 653-663.
AN
The authors declare that the research was conducted without any commercial or financial
M

relationships that could be construed as a potential conflict of interest. All authors made
ED

significant contributions to the study and manuscript. All procedures followed the National

Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the
PT

Institutional Animal Care and Use Committee at the VA New Jersey Health Care System.
CE
AC
 ACCEPTED MANUSCRIPT

Table 1. Behavioral Testing

Standard FR1 Procedure 12 sessions

Standard PR Procedure 3 sessions

PR Procedure without Safety Signal 1 session

PR Procedure without Danger Signal 1 session

T
PR Procedure without Safety and Danger Signals 1 session

IP
FR1 Procedure with Short Safety Signal 1 session

CR
FR1 Procedure with Long Danger Signal 1 session

US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

Figure Captions

Figure 1. Sensitivity to footshock was similar in SD (n = 8) and WKY rats (n = 8). SD and WKY

rats were exposed to increasing intensities of footshock and monitored for flinching and

vocalization. SD and WKY rats did not differ in sensitivity to footshock as measured by the

minimum shock intensity (threshold) that elicits a flinch (A) or vocalization (B).

T
IP
Figure 2. WKY rats (n = 8) acquired avoidance more rapidly and were more motivated to avoid

CR
compared to SD rats (n = 8). (A) Rats were trained to avoid footshock using an FR1 schedule.

WKY rats acquired the avoidance response significantly faster compared to SD rats, but both SD

US
and WKY rats learned avoidance to the same level by the end of 12 sessions. (B) Motivation to
AN
avoid was significantly greater in WKY rats as demonstrated by the more inelastic demand curve

(shifted to the right) compared to SD rats. An exponentiated demand equation was fit to the data
M

and α, a measure of elasticity, and Q 0 , a measure of hedonic value, were obtained from the
ED

equation. (C) WKY rats had a significantly lower α than SD rats, signifying that WKY rats were

more motivated to avoid shock compared to SD rats. α is inversely related to motivation. (D)
PT

The increased motivation to avoid in WKY rats was not associated with differences in the
CE

hedonic value of avoiding shock, Q 0 . The same rats are shown in Figure 2 - 6. ***p < 0.001
AC

Figure 3. Removal of the safety signal reduced motivation to avoid footshock in SD but not

WKY rats. (A) Demand curves demonstrated significantly more inelasticity in WKY rats at

baseline and after removal of safety signal compared to SD rats. Further, elasticity was

significantly increased (leftward shift of the demand equation) following the removal of the

safety signal in SD but not WKY rats. (B) α was significantly less in WKY rats compared to SD

rats during the standard procedure, corresponding to the greater motivation to avoid footshock.
 ACCEPTED MANUSCRIPT

Removal of the safety signal significantly increased α in SD rats compared to the standard

procedure, demonstrating reduced motivation to avoid after removal of safety signal. Removal of

the safety signal had no effect on WKY rats. ***p < 0.001.

Figure 4. Motivation to avoid footshock was not affected by removal of the danger signal in SD

or WKY rats. (A) Whereas the demand curve was more inelastic for WKY rats compared to SD

T
IP
rats, removal of the danger signal had no effect on the demand curve in either strain. (B) α was

CR
significantly less in WKY rats compared to SD rats during the standard procedure and after

removal of danger signal, but removal of the danger signal had no effect on α in either strain. **p

US
< 0.01. AN
Figure 5. Shortening of the safety signal increased lever presses in SD but not WKY rats. (A)

Schematic illustrating the experimental manipulation: In the standard condition, the safety signal
M

is presented for the entire 180 s of the intertrial interval (ITI) (top). In the shortened safety signal
ED

condition, the safety signal was present only during the initial minute (ITI-1), leaving the last

two minutes of the ITI unsignaled (ITI-2, ITI-3) (bottom). (B) The absence of a safety signal
PT

during ITI-2 and ITI-3 significantly increased the number of lever presses in SD rats. (C) In
CE

contrast to SD rats, lever presses in WKY rats were not altered by the absence of a safety signal

during ITI-2 and ITI-3. *p < 0.05.

AC

Figure 6. Lengthening the danger signal increased lever presses in WKY but not SD rats. (A)

Schematic illustrating the experimental manipulation: In the standard condition, the 72 s danger

signal preceded the 180s safety signal (top). In the extended danger signal condition, the danger

signal was lengthened 60 s into the first minute of the ITI, resulting in both danger and safety

signals being present during ITI-1(bottom). (B) Co-presentation of the danger and safety signal
 ACCEPTED MANUSCRIPT

during ITI-1 did not alter lever pressing in the SD rats compared to the standard condition (only

safety signal presented during ITI-1). (C) Co-presentation of danger and safety signals in ITI-1

significantly increased lever press responses during ITI-1 in the WKY rats compared to the

standard condition. * p < 0.05, ** p < 0.01.

Figure 7. Acquisition of conditioned inhibitory properties by safety signals. (A) WKY rats

T
IP
acquired avoidance responding faster and to a greater degree than SD rats. Conditioned

CR
inhibition of safety signals was tested on one-third of the trials during A4, A8 and A12

(arrows). On these trials, danger and safety signals were presented during the first minute of ITI

US
(ITI-1, SS + DS) and compared to Standard trials where only safety signal was

presented. Responding on Standard and SS + DS trials were not different on session A4 for both
AN
SD (B) and WKY (C) rats. On session A8, responding on SS + DS trials was greater than on
M

Standard trials for both strains (B and C). On session A12, responding on SS + DS trials were

similar to Standard trials for SD rats, demonstrating that safety signals acquired conditioned
ED

inhibitory properties and effectively negated the presence of the danger signal in this strain
PT

(B). In contrast, WKY rats responded on SS + DS trials to a greater level than Standard trials on

A12, similar to A8 (C). These results demonstrate that the safety signal in WKY rats had not
CE

acquired conditioned inhibitory properties even through A12. These results demonstrate that the
AC

use of safety signals to suppress responding occurs relatively late in avoidance acquisition in SD

rats, and does not occur in WKY rats. **p<0.01 versus strain-matched group.
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

Highlights

 Behavioral economics of avoidance behavior was assessed in anxiety vulnerable Wistar-

Kyoto (WKY) rats and outbred Sprague-Dawley (SD) rats using a novel progressive ratio
task
 WKY rats demonstrate increased motivation to avoid
 Anxiety vulnerability was associated with danger signal bias in avoidance behavior

T
IP
CR
US
AN
M
ED
PT
CE
AC