NeuroMolecular Medicine

Copyright © 2002 Humana Press Inc.

All rights of any nature whatsoever reserved.
ISSN1535-1084/02/02:151–165/$20.00

Axonal Transport, Tau Protein, and Neurodegeneration

in Alzheimer’s Disease

Dick Terwel, Ilse Dewachter, and Fred Van Leuven*

Experimental Genetics Group, K.U.Leuven, B-3000 Leuven, Belgium
Received May 21, 2002; Accepted May 22, 2002

Abstract
The molecular causes and the genetic and environmental modifying factors of the sporadic
form of Alzheimer’s disease (AD) remain elusive. Extrapolating from the known mutations that
cause the rare familial forms and from the typical post-mortem pathological lesions in all AD
patients—e.g., amyloid plaques and neurofibrillary tangles (NFTs)—the evident molecular can-
didates are amyloid precursor protein (APP), presenilin, and tau protein. To include ApoE4 as
the only certain genetic modifier known leaves us to face the challenge of implementing these
very different molecules into an evident pathological partnership. In more than one respect,
the proposition of disturbed axonal transport appears attractive with more details becoming
available on APP processing and microtubular transport and also of the pathology in the model
systems—e.g., transgenic mice expressing APP or protein tau. Conversely, the resistance of APP-
transgenic mice with full-blown amyloid pathology to also develop tau-related neurofibrillar
pathology is a major challenge for this hypothesis. From the most relevant data discussed here,
we conclude that the postulate of disturbed axonal transport as the primary event in AD is dif-
ficult to defend. On the other hand, failing axonal transport appears to be of major importance
in the later stages in AD, by further compromising tau protein, APP metabolism, and synaptic
functioning. Protein tau may thus be the central “executer” in the chain of events leading from
amyloid neurotoxicity to tau hyperphosphorylation, microtubular destabilization, disturbed
axonal transport, and synaptic failure to neurodegeneration. In order to identify normal phys-
iological processes and novel pathological targets, definition is needed—in molecular detail—
of the complex mechanisms involved.

Index Entries: axonal transport; Alzheimer’s disease; transgenic mice; protein tau; GSK-3β;
microtubules; amyloid.

*Authors to whom all correspondence and reprint requests should be addressed. E-mail: fredvl@med.kuleuven.ac.be;
www.med.kuleuven.ac.be/legtegg/

NeuroMolecular Medicine 151 Volume 2, 2002

152
Introduction
The Problem of Plaques and Tangles
In patients suffering from Alzheimer’s disease
(AD), the brain is histopathologically characterized
by the presence of extracellular plaques made up of
β-amyloid peptides (Aβs) and intraneuronal fibril-
lary tangles made up of aggregated tau protein. The
direct or indirect mechanistical relationship between
tangles and plaques remains elusive.
In a small percentage of AD cases (less than 1%),
the disease is strictly and dominantly inherited
through mutations in the genes that code for amy-
loid precursor protein (APP) or for one of the pre-
senilins. The Aβs are liberated from APP by the
action of two distinct proteinases—e.g., the first
cleaving at the N-terminal side of the amyloid
sequence, known as the β-site of APP, giving rise to
the names β-secretase or BACE, and the second
named γ-secretase, an activity closely involving pre-
senilins (for reviews see Esler and Wolfe, 2001; Siso-
dia and St George-Hyslop, 2002). Whether presenilin
is molecularly identical to γ-secretase or an essen-
tial associated cofactor is a matter of debate. Nev-
ertheless, all known mutations in APP and in
presenilins that cause early-onset familial AD (FAD)
increase either the total levels of Aβ or the ratio of
Aβ42/Aβ40 in the brain, thereby tilting the balance
toward amyloid deposits.
No mutations that cause AD have been described
in the gene that codes for tau protein (chromosome
17). Conversely, mutations in the coding region of
tau protein or in intron 10 of the tau gene are associ-
ated with a group of diseases known as frontotem-
poral dementia (FTD) with parkinsonism linked to
chromosome 17 (FTDP-17), presenting with abnor-
mal tau aggregates only (reviewed by Delacourte and
Buée, 2000). Depending on the particular mutation
in the tau protein, the tau filaments formed in neu-
rons in the brains of these patients can be either
straight, twisted or paired helical filaments (PHFs).
However, the fact that mutant tau proteins form
aggregates and are the primary cause of these neu-
rodegenerative diseases has sparked renewed inter-
est into the role of tau protein in normal physiology
and in the pathology of AD.
Tau protein is a typical microtubule-associated
protein (MAP) and thus is directly implicated in
maintaining the integrity and stability of the micro-
tubules and involved in axonal transport. On the
NeuroMolecular Medicine
Terwel et al.
other hand, recent findings propose a direct role for
APP in axonal transport, as APP can link to kinesins
moving along the microtubules (Kamal et al., 2001).
This evidently brings many factors involved in AD
in close proximity to each other, and the findings
may alter our perspective of the role of axonal trans-
port in AD, which is the topic of this article.
Although numerous reviews have been dedicated
to the amyloid aspect of the disease, this article
focuses on the pathological role of tau and related
mechanisms, as studied in our transgenic mice. We
place these findings into the context of recent find-
ings that might link both sides of the medal, and
may even incorporate the mysterious role of ApoE4.
Tau Protein: Structure and Function
Tau protein is predominantly expressed in neu-
rons of both the central and peripheral nervous
system. It is encoded by a single gene with 16 exons
located on the long arm of chromosome 17 (for
review see Buée et al., 2000). By alternative splicing,
the primary transcript can produce up to six dif-
ferent mRNAspecies in the brain, coding for six tau-
protein isoforms ranging from 352 to 441 amino
acids. These six isoforms contain either three or four
imperfect C-terminal tandem repeats (TRs) and
either one or two shorter N-terminal domains (N).
The shortest tau-protein isoform is the only one pre-
sent in the fetal human brain, and all six isoforms
are present in the adult human brain.
An important function of tau protein involves the
stabilization and spacing of microtubules, proba-
bly by linking to a number of tubulin subunits,
thereby preventing or slowing microtubule depoly-
merization (Drubin and Kirschner, 1986). The three
or four C-terminal repeats that mediate binding of
tau protein to the microtubules (Lee et al., 1989) con-
tain highly conserved 18-amino-acid stretches sepa-
rated by less conserved 13- or 14-amino acid
inter-repeat domains (Goedert et al., 1989). The clus-
ter of three or four microtubule-binding domains are
N-terminally flanked by a proline-rich region and
C-terminally by the so-called 5th repeat (also named
R′). Although the three or four C-terminal repeats are
essential for effective binding of tau protein to the
microtubules, the affinity of binding is further
increased by their flanking regions. This model or
mechanism is known as the “jaws-model”: the cen-
tral repeats enable rapid binding to microtubules,
and the flanking regions are considered to be
Volume 2, 2002
Axonal Transport in Alzheimer’s Disease
“embracing” domains, augmenting the affinity (Man-
delkow et al., 1995). It is to be remembered that tau
protein with four C-terminal domains (tau-4R) binds
more tightly to microtubules than three-repeat tau
(tau-3R) (Goedert and Jakes, 1990; Scott et al., 1991).
The microtubule-binding properties of tau protein
are believed to be important for a number of
processes, e.g., the formation and maintenance of
axons and for fast axonal transport (FAT). Down-
regulation of expression of tau protein in primary
cerebellar neurons with anti-sense oligonucleotides
decreases their ability to generate axons (Caceras and
Kosik, 1990; Caceras et al., 1991). Similarly, tau anti-
sense oligonucleotides produces retraction of neu-
rites in PC12 cells treated with nerve growth factor
(Hanemaaijer and Ginzburg, 1991). These observa-
tions suggest that protein tau is involved in assem-
bling neuronal microtubules into the organized
arrays required for the elaboration and maintenance
of axons. Injection of tau protein in Sf9 insect cells
has induced axon-like processes containing polar-
ized bundles of microtubules as found in axons (Baas
et al., 1991; Biernat and Mandelkow, 1999). By con-
trast, protein tau-deficient mice display a very light
phenotype and only very modest alterations of their
microtubule organization in small-caliber axons. This
can be explained by the increased expression of
microtubule-associated protein (MAP) 1A, that
would compensate for the absent tau protein (Harada
et al., 1994).
One disruptive effect or toxic gain function of tau
protein in intracellular transport has become evident
from studies on stably transfected Chinese hamster
ovary (CHO) cells and differentiated neuroblastoma
(N2a) cells (Ebneth et al., 1998; Trinczek et al., 1999).
Special mention should be made of primary cultures
of dorsal root ganglion (DRG) cells from our trans-
genic mice that overexpress 4R human tau (htau-4R).
In these cells, the overexpression of tau interfered
with anterograde intracellular transport of cyto-
plasmic organelles and intermediate filaments medi-
ated by kinesin-like motor proteins along
microtubules (Nuydens et al., 2002). Transgenic mice
that overexpress tau and tau-kinases have been gen-
erated, and have revealed relevant aspects of
(dys)function of tau protein, as discussed further.
Tau Phosphorylation and Tangle
Formation in AD
Tau protein is a phosphoprotein: the longest brain
isoform has 79 putative serine and threonine phos-
NeuroMolecular Medicine
153
phorylation sites, and at least 30 of these were effec-
tively identified experimentally as being phospho-
rylated in one system or another (reviewed by Buée
et al. 2000). Many phosphorylation sites are serine
or threonine followed by proline, and most are
located outside the microtubule-binding repeat
region. Among the specific “proline-dependent”
kinases that can phosphorylate protein tau at dif-
ferent sites in vitro, special attention is dedicated to
glycogen-synthase kinase-3β (GSK-3β) (Michel et
al., 1998), mitogen-activated protein kinase (MAPK)
(Drewes et al., 1992), stress-activated protein kinases
(SAPKs) (Goedert et al., 1997) and cyclin-dependent
kinases (CDKs) including cdc2 and cdk5 (Baumann
et al., 1993; Patrick et al., 1999).
Other serine and threonine residues, not followed
by proline, are phosphorylated by other protein
kinases, including microtubule-affinity-regulating
kinase (MARK) (Drewes et al., 1993), calcium/
calmodulin kinase II (CAMKII), cAMP-dependent
kinase (PKA) (Johnson et al., 1992), and casein kinase
II (Greenwood et al., 1994). Tau protein is rapidly
dephosphorylated by endogenous phosphatases
such as protein phosphatases 1, 2A, and 2B (cal-
cineurin) that are all present in the brain, and effec-
tively dephosphorylate tau protein in vitro. Since
endogenous, active tau protein is constitutively
phosphorylated at many sites, and rather similar to
tau proteins isolated from PHF in the AD brain
(Matsuo et al., 1994), it is accepted that phospho-
rylation and dephosphorylation is the control mech-
anism that regulates binding and dissociation of tau
protein to and from microtubules.
Protein tau is “hyper”-phosphorylated in the
brains of patients suffering from AD or from other
neurodegenerative diseases characterized by neu-
ronal tau protein aggregates (Buée et al., 2000).
Apparently, this is why hyperphosphorylation of
protein tau has been considered important for its
disturbed function and for its self-aggregation,
although it is not clear whether the higher number
of phosphate groups or the phosphorylation of spe-
cific sites is more important. It is also not clear
whether the two phenomena—e.g., hyperphos-
phorylation and aggregation—are inherently and
functionally linked, or whether they are the result
of the same structural feature(s) of tau protein, as
disturbed by phosphorylation. However, it is clear
that aggregation per se will prevent effective bind-
ing to microtubules, and effectively lowers the con-
centration of free, active tau protein. As far as we
Volume 2, 2002
154
know, aggregation of tau protein is not known to
be a control mechanism in its own right.
Following the observation that increased phos-
phorylation of tau protein is an early change in those
brain areas where tangle pathology is manifested
in the AD brain (Braak et al., 1994; Delacourte et al.,
1999) and because hyperphosphorylated tau pro-
tein has a reduced affinity for microtubules (Iqbal,
1986; Yoshida and Ihara, 1993) a model was pro-
posed to link both. Essentially, increased phospho-
rylation of tau protein by different kinases is
believed to be the primary event(s) preventing
hyperphosphorylated tau to bind to the micro-
tubules, causing their destabilization and interfer-
ing with axonal transport. It has been proposed that
tau hyperphosphorylation results from oxidative
stress and covalent modification of tau by the lipid
peroxidation product 4-hydroxynonenal (Mattson
et al., 1997). The gradual increase of the cytoplas-
mic levels of hyperphosphorylated tau in the
affected neurons progressively constitutes a pool of
substrate from which the PHFs form and aggregate
into neurofibrillary tangles (NFTs).
Causes and Consequences
of Hyperphosphorylation of Tau Protein
This model does not offer an explanation for the
cause or for the identity of the increased kinase activ-
ity, and has also been challenged by the observa-
tion that phosphorylation of tau protein reduces its
aggregation in the presence of cofactors such as
heparin or other polyanions (Friedhoff et al., 2000).
However, in the absence of any cofactors, phos-
phorylation of tau protein was necessary and suf-
ficient for its self-assembly into paired helical or
straight filaments (Alonso et al., 2001). It is clear
that the process of tau self-aggregation is not well-
understood, not even in vitro with recombinant tau
protein. The excess phosphate groups may simply
neutralize charges, shift properties of tau protein in
solution from a rather basic to a more acid protein,
and thereby render it from a hydrophilic to a more
hydrophobic protein. Indeed, occupation of only
nine phosphorylation sites could be sufficient to
reduce or neutralize the basic character of tau pro-
tein (Ruben et al., 1997). Thus, the conglomeration
of tau protein and other proteins into the massive
NFTs, can be the result of subsequent modifications,
such as the transglutamination and glycosylation
(Norlund et al., 1999; Wang et al., 1996), which are
NeuroMolecular Medicine
Terwel et al.
processes that cannot and need not be discussed
here.
However, an even greater enigma remains—the
relation of tau pathology to processing of APP into
its metabolites, mainly the Aβs and C99 fragments.
Indeed, it is totally uncertain whether the increased
Aβ level, which is accepted by many as the primary
event in the pathology of AD, is in any way related
to increased phosphorylation of tau protein. The
most intricate aspects are not examined here because
of limited space. We must consider two essential
options: (i) a co-effect, or the same triggers that
derange APP processing also govern hyperphos-
phorylation of tau protein, or alternatively (ii) a
sequential process, whereby unknown causes lead
to mis-processing of APP, which in turn deranges
the mechanisms that control tau protein. Evidently,
this neither grants—nor denies—hyperphosphory-
lation of tau the role of executer!
The model of hyperphosphorylation and PHF
and NFT formation can only fit with deterioration
of axonal transport, if less active tau protein will
effectively be available for stabilization of the micro-
tubules. Evidently, this occurs only when tran-
scriptional and translational activity and thereby
its synthesis remains unaltered by the post-
translational modifications—e.g., hyperphospho-
rylation of tau. This is a matter of debate, as it has
not been unequivocally demonstrated in the brain
of AD patients. In addition, however, hyperphos-
phorylated tau can sequester normal tau protein
into filaments and thereby promote destabilization
and disassembly of microtubules, as shown at least
in vitro (Alonso et al., 1996). Finally, the tau fila-
ments and aggregates can form physical barriers or
otherwise interfere physically with anterograde and
retrograde axonal transport, and thereby pose sec-
ondary problems.
Transgenic Mice with Tau
and Tau-Kinases
Disruption of Axonal Transport
by Excess Tau
The remarkable effect of excess tau protein on
axonal transport has been originally exemplified in
vivo in our transgenic mice that overexpress the
longest four-repeat human tau protein isoform
(htau-2N, 4R) specifically in neurons (Spittaels et
al., 1999). These transgenic mice developed axonal
Volume 2, 2002
Axonal Transport in Alzheimer’s Disease 155

Fig. 1. Selected examples of axonal problems in the brain of tau-4R-transgenic mice. (A) Beginning dilatation
(arrow-heads) close to the cell body. (B) Overview of cortex showing numerous dilatations (arrow-heads). (C)
Dilatations of different size, revealed by silver staining. (D) Close-up view of a spheroid from panel B. (E) Ultra-
structural view of the contents of a spheroid: mitochondria (triangles), neurofilaments (small arrow-heads), micro-
tubule (small arrow), and unidentified vesicle (larger arrow). (A, B, D) the immunohistochemistry was performed
with antibody SMI32 directed against neurofilaments. Color image available for viewing at www.humanapress.com

degeneration in the brain and spinal cord that was
characterized by axonal dilations with accumula-
tion of neurofilaments (NFs), microtubules, mito-
chondria, and different types of vesicles (Fig. 1).
Subsequently, Wallerian degeneration and severe
muscle wasting and motoric problems demon-
strated the extensive neurodegeneration caused by
the overexpression of human tau protein in a gene-
dosage fashion.
This phenotype is explained by the fact that the
mouse thy-1 gene promoter did express tau protein
in all central neurons, including the motor neurons
in the spinal cord, which are known to be particu-
larly sensitive to stress. In this respect, these find-
ings are of considerable interest in the field of
amyotrophic lateral sclerosis (ALS). The dilated
axons or spheroids much resemble those observed
in ALS and ALS/Parkinson dementia complex
(Nakazato et al., 1984; Shankar et al., 1989). In ALS,
accumulation of NFs is a prominent feature
(Rouleau et al., 1996), and it has been demonstrated
that NFs contribute heavily to the axonopathy of
tau transgenic mice (Ishihara et al., 2001a). How-
ever, problems with NFs are observed in AD only
in later stages, and a direct role of NFs and axonal
dilations in the early ontogeny of AD seems less
likely.
Most recently, our tau transgenic mice have
proved to be relevant for prion-induced neurode-
generation (Künzi et al., 2002). Tau transgenic mice
infected with prions accumulated intraneuronal
deposits of hyperphosphorylated tau protein,
resembling Gerstmann-Sträussler-Scheinker syn-
drome. These findings identify tau pathology as a
possible end stretch of prion-induced neurodegen-
eration (Künzi et al., 2002)
For the sake of completeness, we also refer to tau-
3R transgenic mice that developed another type of
pathology in the hippocampus, e.g., straight fila-
ments formed in aged mice older than 18 mo (Ishi-
hara, 2001b), which was proposed to be relevant for
AD, given the age-dependence.

NeuroMolecular Medicine Volume 2, 2002

156
Specified mutations in the tau gene, including
those in intron 10, cause neurodegeneration and
have a remarkable molecular effect—only an
increase in the levels of the functionally normal
tau-4R protein isoform (Hutton et al., 1998; Spillan-
tini et al., 1998; Hong et al., 1998; Hasegawa et al.,
1998). Evidently, this argues for critical levels of
protein tau-4R in the pathology of FTD and by exten-
sion, in AD. As stated previously, we have unequiv-
ocally documented in the transgenic mice in vivo
that elevated concentrations of tau protein blocked
traffic of vesicles and organelles (Fig. 1) (Spittaels
et al., 1999; 2000).
The following proposition has been recently reit-
erated that axonal transport in AD could become
disrupted by increased neuronal concentrations of
tau protein, now based on evidence of a cellular par-
adigm, also demonstrating an effect on transport of
APP (Stamer et al., 2002). These results could help
to explain the increased Aβ production in the trans-
Golgi network. The amount of Aβ produced could
be altered by delayed axonal transport, as well as
the precise species of metabolites of APP produced—
e.g., Aβ40 or 42, monomeric Aβ, or Aβ-oligomers or
Aβ-derived diffusible ligands (ADDLs) (Lambert et
al., 1998; Walsh et al., 2000). Moreover, the C99 frag-
ments resulting from β-cleavage of APP and the
obligate precursors of the Aβs may be included
among the toxic metabolites of APP, as we have
recently demonstrated in vivo in aged mice that
were neuronally deficient in PS1, and thereby in
γ-secretase activity (Dewachter et al., 2002).
Unfortunately, evidence for increased intracel-
lular tau concentrations in the brain of AD patients
is to be taken with caution, since it represents data
on total tau concentration, including deposited or
“tangle” tau (Khatoon et al. 1992), while concen-
trations of soluble tau in the AD brain have been
found to be reduced (Ksiezak-Reding et al., 1988).
Rescue of Axonopathy in Tau Transgenic
Animals by Co-Expression of GSK-3β
Vesicle stalling would fit well with the more clas-
sical type of view on how hyperphosphorylation of
tau protein causes problems with axonal transport.
In addition to a role in destabilization of micro-
tubules (as discussed here), a role for GSK-3β has
been documented in regulating kinesin-based motil-
ity (Morfini et al., 2002). Evidence that GSK-3β phos-
phorylates kinesin light chain was presented
NeuroMolecular Medicine
Terwel et al.
together with an increased accessibility of the
J-domain motifs of the KLCs to HSC70, leading
to removal of kinesin from vesicles and their
rapid degradation. Therefore, increased GSK-3β
activity—and/or decreased de-phosphorylation—
can be anticipated to jeopardize FAT in two differ-
ent ways.
Direct evidence for a role of GSK-3β in tau
protein-related processes stems from our findings
with single- and double-transgenic mice that over-
express htau-4R and/or GSK-3β (Spittaels et al.,
1999, 2000). Neuronally overexpressed GSK-3β was
demonstrated to hyperphosphorylate tau protein
in vivo in the brain and spinal cord of double-
transgenic mice (Spittaels et al., 2000) thereby
reducing the amount of protein tau associated with
microtubules by 50% (Spittaels et al., 2000). More-
over, the unbound tau protein was hyperphospho-
rylated and especially at the AD-2 epitope, e.g., an
epitope shown to contain serine 396 and serine 404.
The region surrounding these serines in tau protein
is thus demonstrated to contain essential determi-
nants governing tau protein—microtubule interac-
tion in vivo.
Surprisingly, the axonopathy was rescued in the
tau x GSK-3β double-transgenic mice, together
with a near-total normalization of the functional
disabilities. The explanation we proposed was
that increased phosphorylation of tau protein by
the extra GSK-3β activity did effectively lower the
actual concentration of tau protein that was able
to bind to the microtubules, and thereby reduced
the interference with transport by the kinesin
motors (Fig. 2). More recently, we observed that
the overexpression of GSK-3β on itself caused a
light form of microcephaly in the GSK-3β-
transgenic mice. This appeared to be caused by a
decrease in the caliber of the neuronal processes
(Spittaels et al., 2002). Although much further work
is needed, these additional data demonstrate that
GSK-3β is intimately involved in the architecture
of axons and other neuronal processes, providing
indirect support for its role in tau-mediated con-
trol of axonal transport.
The theory that phosphorylation of tau protein
reduces binding to microtubules and restores axonal
transport was further supported by studies in DRG
cells cultured from the transgenic mice that over-
express GSK-3β and htau (Nuydens et al., 2002).
DRG cells from htau-transgenic mice showed
reduced sprouting capacity, while density and sta-
Volume 2, 2002
Axonal Transport in Alzheimer’s Disease 157

Fig. 2. Schematic representation of impaired axonal transport by excess tau protein, and its rescue by
co-expression of GSK-3β. In essence, excess tau overexpressed in tau-4R-transgenic mice (e.g., tau in middle panel)
binds to microtubules and prevents kinesin motors to move, as they normally do (e.g., wild-type in upper panel).
Co-expression of GSK-3β phosphorylates tau protein, and thereby prevents it from binding to the microtubules
restoring normal axonal transport (lower panel). The schemes are based on extensive experimental data (Spittaels
et al., 1999, 2000, 2002), as discussed in the text. Color image available for viewing at www.humanapress.com

bility of microtubules in the axonal processes were
increased. Video-enhanced contrast microscopy
demonstrated a dramatic inhibition of FAT. Coex-
pression of GSK-3β increased tau protein phos-
phorylation and reversed the effects on microtubule
stability and saltatory motion (Nuydens et al., 2002).
These studies demonstrate that GSK-3β is an
effective tau-kinase in vivo, and that this phospho-
rylation is directly involved in the regulation of
binding of tau protein to microtubules in vivo. Nev-
ertheless, despite the high levels of hyperphos-
phorylated tau protein in the brain of the tau x
GSK-3β double-transgenic mice, no signs of
PHF formation was evident in these studies, indi-
cating that other factors are involved (Spittaels et
al., 1999, 2000).
Pathology in Mice that Co-Express
Tau Protein and APP
If axonal transport is problematic in AD, then
alterations are also to be expected in APP transgenic
mouse models. We have analyzed this aspect in our
APP[V717I]-transgenic mice, which neuronally
overexpress the London mutant of APP and thereby
recapitulate—robustly and progressively—the
amyloid pathology that is very reminiscent of AD
patients, including vascular angiopathy (Moechars
et al., 1999; Van Dorpe et al., 2000). A role for an
axonal-transport deficit in the development of
pathology was not directly obvious, however
(Moechars et al., 1999; Van Dorpe et al., 2000;
Dewachter et al., 2000; and unpublished results).
Eventually, this could be explained by the lesser vul-
nerability of mouse neurons to Aβs (Geula et al.,
1998). Whether this is related to the fact that the
APP[V717I] mice fail to develop tangle pathology,
despite the presence of typical tau epitopes in
swollen neurites surrounding the plaques
(Moechars et al., 1999; Van Dorpe et al., 2000;
Dewachter et al., 2000), is a distinct possibility.
In all transgenic mice that we generated to over-
express APP, tau protein-4R, GSK-3β, ApoE4, and
others, the transgene was controlled by the mouse
thy-1 gene promoter, which allows high expression
in all central neurons, including those of the spinal
cord (Moechars et al., 1999; Spittaels et al., 1999,
2000; Tesseur et al., 2000a,b). Therefore, we antici-
pated that if mere expression of APP in the spinal
cord were to jeopardize axonal transport, that
would have been manifested by motoric problems,
as seen in the tau-4R- and ApoE4-transgenic mice
(Spittaels et al., 1999, 2000; Tesseur et al., 2000a,b).
In sharp contrast to results obtained with overex-
pression of APP in Drosophila larva (Gunawardena
and Goldstein, 2001) this theory was not shown to
be accurate.

NeuroMolecular Medicine Volume 2, 2002

158
To investigate the synergism of APP and tau in
vivo, we have generated APP[V717I] x tau-4R
double-transgenic mice. Surprisingly, the resulting
offspring suffered from a very high premature mor-
tality rate, with lifespans of between 6 and 20 wk,
preventing an extensive analysis of the phenotype
of adult, and especially aged mice (unpublished
results). In the absence of any major organic defects,
we have now confirmed that the APP x tau double-
transgenic mice appear to be extremely sensitive to
external environmental stress, not unlike some
patients. A possible related vulnerability to stress
was previously reported in a study of one line of
APP-mutant-transgenic mice (Pedersen et al., 1999).
Analysis of young double-transgenic mice did not
demonstrate a major synergistic effect of pathology,
which was also largely lacking in the parent strains
at young ages (less than 3 mo, unpublished results).
Others have demonstrated an interaction
between APP and tau protein in double-transgenic
mice, using a mutant form of tau protein, as opposed
to the wild-type human tau that we have used (Lewis
et al., 2001). These mice develop signs of additional
if not synergistic pathology, whereby NFTs
become apparent in regions of the brain that are not
affected in single-tau-mutant mice. The effect
appeared to be mediated through projection of
axonal extensions, which was interpreted to mean
that Aβ-mediated neurotoxicity by axonal projec-
tion, implicating axonal transport in increasing vul-
nerability of neurons to Aβ in a “feed-forward”
fashion. How this would result in increased tangle
formation is open for debate and more experimen-
tation. It is in agreement with the proposal that
plaque and tangle formation can begin in different
brain areas, and the presence of amyloid causes
spreading of tangle pathology along neuronal tracts
in a hierarchal fashion—or vice versa (Braak and
Braak, 1991; Delacourte et al., 1999).
A Role for APP in Axonal Transport—
Or for Axonal Transport
in APP Metabolism?
The biological function of APP remains an
enigma, since it was identified as the precursor of
the Aβs and already then shown to be a proteinase
inhibitor, at least in regard to the isoforms that con-
tain the Kunitz inhibitor domain. Without entering
NeuroMolecular Medicine
Terwel et al.
into the fine details, mice deficient in APP have not
helped very much to solve the mystery, probably
because of redundancy of the APP-like proteins.
However, it is believed that the secreted form of
APP plays important roles in neuronal plasticity
and survival (Mattson, 1997).
Most recently, this role became even more puz-
zling, since APP was implemented as a kinesin-1
receptor (Kamal et al., 2000; 2001). APP apparently
binds to the light chain of kinesin-1, which itself is
responsible for anterograde axonal transport and
consists of two light chains (KLC) associated with
two heavy chains (KIF5B). KIF5B contains ATP- as
well as microtubule-binding motifs that are essen-
tial for vesicle transport. KLC associates with KIF5B
and tethers membrane vesicles containing a subset
of proteins. Kinesin takes care of anterograde axonal
transport and dynein is responsible for retrograde
axonal transport, but the structure of dynein is not
examined here.
An axonal membrane compartment was isolated
from sciatic nerve and corpus callosum by gradient
centrifugation and binding to an antibody against
APP that contained BACE and presenelin-1 (PS1), as
well as the degradation products of APP, Aβ40, and
Aβ42 (Kamal et al., 2001). Isolated fragments of sci-
atic nerve, when incubated at 37°C, were shown to
generate Aβ40 and Aβ42 and increased soluble
kinesin-1 and particulate Aβ40 and Aβ42 were found
in sciatic nerve segments after ligation. The combined
data are taken as strong evidence that processing of
APP to Aβ can occur in axonal vesicles transported
by kinesin-1 (Kamal et al, 2001). This would consti-
tute the first link of axonal transport involving APP
and Aβ generation on the one hand and the micro-
tubular molecular motors on the other, indirectly
implicating tau protein.
In a completely different model, deletion of the
APP-like gene (Appl) or overexpression of human
APP or APPL in Drosophila caused axonal-
transport phenotypes similar to kinesin and dynein
mutants (Gunawardena and Goldstein, 2001). Dele-
tion of APPL resulted in vesicle or organelle stalling
in axons of Drosophila larva. Overexpression of
APP695 or APPL led to non-APP-vesicle stalling
and axonal accumulations. Deletion of the
C-terminus of APP695 or APPL, including the
kinesin-binding region, disrupted axonal transport
of APP695 and APPL.
Although these findings are most interesting and
above all intriguing, they do not yet offer a solution
Volume 2, 2002
Axonal Transport in Alzheimer’s Disease
to the impasse we entered: the physiological role
of mammalian APP. On the contrary. If APP medi-
ates its own transport and that of its complete
Aβ-metabolizing machinery—e.g., BACE and PS1—
in vesicles along the axons, the question is, for what
purpose? One possibility is that the destination of
the transported vesicles is determined by proteol-
ysis of APP, which only complicates matters further
because local signals along the axon will have to be
invoked.
Is an Axonal Transport Deficit
Inherently Involved in AD Pathology?
Axonal transport cannot be investigated directly
in humans, but neuropathological observations of
AD patients have been interpreted to suggest a prob-
lem. Electron microscopy showed a disturbed neu-
ronal cytoskeleton in AD brain (Perry et al., 1991;
and references therein), which evidently could be
consequent to the pathology, not its cause or medi-
ator. The very few studies that used biopsy mater-
ial provided shreds of evidence for a reduction in
the number of microtubules in affected neurons,
circumventing at least partially the postmortem
problem of dissolution of microtubules.
In the absence of direct evidence in patients, the
model systems have offered possibilities as to how
axonal transport might become impaired in AD: (i)
by increased APP production, (ii) by increased tau
protein production, (iii) by increased protein tau
phosphorylation, or (iv) by neuronal expression of
apolipoprotein (ApoE) (see Fig. 3). The first two pos-
sibilities have not been substantiated unequivocally
in the AD brain in later stages, let alone identified
as the primary event. Thus, it must be considered
that normal expression of tau and APP in combi-
nation with age-related changes is sufficient to cause
disturbances in axonal transport in AD, shifting the
problem again to be defined as “age-related
changes” in molecular terms!
Still, other arguments that have been put forth as
to why the transport function of APP could be related
to the initiation of AD pathology are: (i) apposition
of APP, BACE, and PS in the same vesicle, (ii) extra
APP gene copy in Down syndrome, (iii) dystrophic
neurites in Drosophila that overexpress APP (Gold-
stein et al., 2001). Although these arguments point
to and offer possibilities, none separately or in com-
bination are totally convincing. The extra gene copy
NeuroMolecular Medicine
159
and production of APP in Down’s syndrome is in
perfect agreement with the amyloid-β cascade
hypothesis, which demands merely increased Aβ
production but not necessarily disturbed axonal
transport. Other systems and arguments have already
been addressed. Direct interaction between APP and
tau in the same neuron does not appear to be a pre-
requisite for the combined lesions to occur. There-
fore, the synergism between tangle and amyloid
pathology may well be regarded as “distant,” as sup-
ported by some experimental evidence (Lewis et al.,
2001; Götz et al., 2001) as well as by neuropatholog-
ical observations in AD (Braak and Braak, 1994).
A Concluding, Not Conclusive
Scheme
A convincing scenario could consist of an initial
separate formation of Aβ peptides and hyper-
phosphorylation of tau, eventually resulting in
plaques and tangles respectively, but essentially
by a “spreading” phenomenon via axonotoxicity
of Aβ peptides in any form in which they can
exist, Aβ-(proto)fibrils, Aβ-oligomers, or ADDLs
(Lambert et al., 1998; Koo et al., 1999; Walsh et al.,
2000), synergizing both aspects of the pathology.
Inherently, the increased Aβ toxicity somehow leads
to increased tangle formation—e.g., by hyperphos-
phorylation of tau. This further disrupts axonal
transport and results in increased neuronal vul-
nerability, and evidently, this is a vicious circle (see
Figs. 3A, 4).
Some evidence for this scheme has been found
in hippocampal cell cultures derived from tau
protein-knockout and tau-transgenic mice, demon-
strating that tau was needed for Aβ toxicity in
hippocampal cultures (Rapoport et al., 2002). Phar-
macological inhibition of MAPK blocked neurite
retraction of neurons exposed to Aβ (Rapoport and
Ferreira, 2000), and brain slices from Fyn-knockout
mice were completely protected from ADDL toxic-
ity (Lambert et al., 1998)—an interesting link, since
MAPK is positioned downstream from tyrosine
kinase Fyn (Williamson et al., 2002). Another inter-
esting link is offered by Niemann-Pick disease, since
these patients have tangles that are indistinguish-
able from the NFTs in AD. The product of the
responsible Niemann-Pick gene (NPC) regulates
cholesterol homeostasis, and transgenic mice over-
expressing NPC1 show cognitive defects, motoric
Volume 2, 2002
160 Terwel et al.

Fig. 3. Schematic representation of axonal transport of vesicles, incorporating the major molecular players
involved in AD. (A) A vesicle containing APP, BACE, and PS1 is transported along the axonal microtubules, even-
tually delivering Aβ peptides and/or aggregates to the synapse (based on Kamal et al., 2001). The resulting neu-
rotoxic signals are mediated via different pathways into nearby and distant neurons to which they project, and
trigger hyperphosphorylation of tau by specified kinases (see text for details). The depletion of active tau protein
destabilizes the microtubular network and further disturbs axonal transport. (B) In normal conditions (upper)
diverse organelles (red circles) and APP-carrying vesicles (open circles) are axonally transported and arrive at the
synapse. Increasing the levels of tau protein inhibits kinesin motors to bind to the microtubules, and less vesicles
of all types arrive at the synapse: axonal transport is impaired (middle). When the levels of APP are increased, it
is anticipated (Kamal et al., 2001) that vesicles carrying APP would preferentially bind to kinesin motors, and
thereby less organelles and more vesicles carrying APP, with higher loads of Aβ, will arrive at the synapse (lower).
Color image available for viewing at www.humanapress.com

NeuroMolecular Medicine Volume 2, 2002

Axonal Transport in Alzheimer’s Disease 161

Fig. 4. Scheme relating unknown causes of sporadic AD to the pathological lesions and impaired axonal trans-
port. The unknown causes of sporadic AD, represented by X and Y, affect separately or combined the processing
of APP and the control of phosphorylation of tau protein, leading respectively to increased production of amyloid-
fragments of APP (e.g., C99, Aβ40, and Aβ42) and to hyperphosphorylation of tau protein (tau-P). These events
give rise in the first (or in last) instance to the well-known amyloid and tau-pathology of plaques and tangles that
are the postmortem hallmarks of all AD patients. The link of hyperphosphorylation of tau to microtubular desta-
bilization and impaired axonal transport can be taken established, as opposed to the inference of C99 and/or Aβ
peptides with tau-phosphorylation, or vice versa (indicated by the double-headed arrow with question mark). The
theory that impaired axonal transport negatively contributes to the pathology by augmenting the generation of Aβs
has not yet been proven, as indicated by the question mark. Other question marks refer to problems and questions
raised and expanded in the text. Color image available for viewing at www.humanapress.com

problems with hyperphosphorylation of tau, and
a 10-fold increase in MAPK activity (Sawamura et
al., 2001; Voikar et al., 2002). It appears unlikely
that the known cholesterol export problem in these
patients relates to the problem caused by the ApoE4
allele in AD patients. On the other hand, the direct
test of the hypothesis of neuronal expression of
ApoE4 in transgenic mice resulted in very similar
motoric problems as in tau-4R transgenic mice,
including the progressive accumulation of hyper-
phosphorylated tau with age (Tesseur et al.,
2000a,b). This unexpected finding must be further
analyzed to define the signaling pathways involved,
especially with regards to the ApoER2/LRP8
receptor as essential reelin receptor during
development.
Further experimental evidence is needed to
determine whether the signaling pathways and
kinases discussed here are indeed as prominently
involved as we have proposed. Nevertheless, a
major problem remains—not if, but how early
tangle formation involves GSK-3β. The mouse
models do not unequivocally answer the question
(Lucas et al., 2001; Spittaels et al., 2000). In the brain
of AD patients, active GSK-3β is associated with
the earliest signs of tangle formation (Pei et al.,
1999) and in a paradigm of neuronal apoptosis
in vivo—e.g., hypometabolism induced by a
glutamate antagonist, GSK-3β activation preceded
tau hyperphosphorylation (Elyaman et al., 2002).
Most recently, co-overexpression of tau and shaggy,
the GSK-3β homolog in Drosophila, resulted in
neurofibrillary pathology in transgenic flies
(Jackson et al., 2002). The circumstances that
increase or deregulate GSK-3β activity in AD are
unclear, but are likely related to cellular stress asso-
ciated mainly with aging as the major risk factor
for AD. This brings us back to defining the “aging”
problem in molecular terms.
In conclusion, although it is most attractive
to postulate that the pathology in AD is from the
onset accompanied by disturbed axonal transport,

NeuroMolecular Medicine Volume 2, 2002

162
its acceptance as a primary event appears dif-
ficult to defend. On the other hand, the secondary
effects on metabolism and synaptic function make
it a likely partner in the subsequent stages of the
pathology of AD and other neuro-degenerative dis-
orders. Whereas a stronger case can be made now
than before for linking Aβ toxicity with tau phos-
phorylation and axonal transport disturbances,
future research should be directed at understand-
ing the mechanisms involved, to eventually allow
some remediation.
Acknowledgments
This investigation was supported by Fonds voor
Wetenschappelijk Onderzoek-Vlaanderen (FWO-
Vlaanderen); the EEC-5th Framework Program;
the Rooms-Fund; the Janssen Research Foundation;
the KULeuven GOA-Research Fund and KULeu-
ven R&D; the Institute voor Aanmoediging van
Wetenschappelijk en Technisch Onderzoek (IWT).
Ilse Dewachter is a post-doctoral fellow of FWO-
Vlaanderen and Dick Terwel is a post-doctoral
fellow of IWT.
References
Alonso A. C., Grundke-Iqbal I., and Iqbal K. (1996)
Alzheimer ’s disease hyperphosphorylated tau
sequesters normal tau into tangles of filaments and
disassembles microtubules. Nat. Med. 2, 783–787.
Alonso A. C., Zaidi T., Novak M., Grundke-Iqbal I.,
and Iqbal K. (2001) Hyperphosphorylation induces
self-assembly of tau into tangles of paired helical
filaments/straight filaments. PNAS 98, 6923–6928.
Baas P. W., Pienkowski T. P., and Kosik K. S. (1991)
Processes induced by tau expression in Sf9 cells
have an axon-like microtubule organization. J. Cell
Biol. 115, 1333–1344.
Baumann K., Mandelkow E.-M., Biernat J., Piwnica-
Worms H., and Mandelkow E. (1993) Abnormal
Alzheimer-like phosphorylation of protein tau
by cyclin-dependent kinases cdk2 and cdk5. FEBS
Lett. 336, 417–424.
Biernat J. and Mandelkow E.-M. (1999) The develop-
ment of cell processes induced by protein tau
requires phosphorylation of serine 262 and 356
in the repeat domain and is inhibited by phospho-
rylation in the proline-rich domains. Mol. Biol. Cell
10, 727–740.
NeuroMolecular Medicine
Terwel et al.
Braak E. and Braak E. (1991) Neuropathological stage-
ing of Alzheimer-related changes. Acta Neuropathol.
82, 239–259.
Braak E., Braak H., and Mandelkow E.-M. (1994) A
sequence of cytoskeletal changes related to the for-
mation of neurofibrillary tangles and neuropil
threads. Acta Neuropathol. 87, 554–567.
Buée L., Bussière T., Buée-Scherrer V., Delacourte A.,
and Hof P. R. (2000) Tau protein isoforms, phos-
phorylation and role in neurogenerative disorders.
Brain Res. Rev. 33, 95–130.
Caceras A. and Kosik K. S. (1990) Inhibition of neurite
polarity by tau antisense oligonucleotides in pri-
mary cerebellar neurons. Nature 343, 461–463.
Caceras A., Potrebic S., and Kosik K. S. (1991) The effect
of tau antisense oligonucleotides on neurite for-
mation of cultured cerebellar macroneurons.
J. Neurosci. 11, 1515–1523.
Delacourte A. and Buée L. (2000) Tau pathology: a
marker of neurodegenerative disorders. Curr. Opin.
Neurol. 13, 371–376.
Delacourte A., David J. P., Sergeant N., et al. (1999) The
biochemical pathway of neurofibrillary degenera-
tion in aging and Alzheimer’s disease. Neurology
52, 1158–1165.
Dewachter I., Van Dorpe J., Smeijers L., Gilis M., Kuiperi
C., Laenen I., et al. (2000) Aging increased amyloid
peptide and caused amyloid plaques in brain of
old APP/V717I transgenic mice by a different
mechanism than mutant presenilin 1. J. Neurosci. 20,
6452–6458.
Dewachter I., Reverse D., Caluwaerts N., Ris L., Kuipéri
C., Van den Haute C., et al. (2002) Neuronal
deficiency of presenilin 1 inhibits amyloid plaque
formation and corrects hippocampal long-
term potentiation but not a cognitive defect of
amyloid precursor protein [V717I] transgenic mice.
J. Neurosci. 22, 3445–3453.
Drewes G., Lichtenberg-Kraag B., Doring F., Man-
delkow E.-M., Biernat J., Goris J., et al. (1992) Mito-
gen activated protein (MAP) kinase transforms
protein tau into an Alzheimer-like state. EMBO J.
11, 2131–2138.
Drewes G., Mandelkow E.-M., Baumann K., Goris J.,
Merlevede W., and Mandelkow E. (1993) Dephos-
phorylation of protein tau and Alzheimer
paired helical filaments by calcineurin and
phosphatase-2A. FEBS Lett. 336, 424–432.
Drubin D. G. and Kirschner M. W. (1986) Protein tau
function in living cells. J. Cell Biol. 103, 2739–2746.
Ebneth A., Godeman R., Stamer K., Illenberger S.,
Trinczek B., Mandelkow E.-M., et al. (1998)
Overexpression of protein tau inhibits kinesin-
Volume 2, 2002
Axonal Transport in Alzheimer’s Disease
dependent trafficking of vesicles, mitochondria, and
endoplasmatic reticulum: implications for
Alzheimer’s disease. J. Cell Biol. 143, 777–794.
Esler W. P. and Wolfe M. S. (2001) A portrait of
Alzheimer secretases—new features and familiar
faces. Science 293, 1449–1454.
Elyaman W., Terro F., Wong N. S., and Hugon J. (2002)
In vivo activation and nuclear translocation of
phosphorylated glycogen synthase kinase-3β in
neuronal apoptosis: links to tau phosphorylation
Eur. J. Neurosci. 15, 651–660.
Friedhoff P., von Bergen M., Mandelkow E.-M., and
Mandelkow E. (2000) Structure of protein tau and
assembly into paired helical filaments. BBA 1502,
122–132.
Geula C., Wu C. K., Saroff D., Lorenzo A., Yuan M.,
and Yankner B. A. (1998) Aging renders the brain
vulnerable to amyloid-β protein neurotoxicity. Nat.
Med. 4, 827–831.
Goedert M., Spillantini M. G., Potier M. C., Ulrich
J., and Crowther R. A. (1989) Cloning and sequenc-
ing of the cDNAencoding an isoform of microtubule-
associated protein tau containing four tandem
repeats: differential expression of protein tau
mRNAs in human brain. EMBO J. 8, 393–399.
Goedert M. and Jakes R. (1990) Expression of separate
isoforms of human tau protein: correlation with the
tau pattern in brain and effects on tubulin poly-
merisation. EMBO J. 9, 4225–4230.
Goedert M., Hasegawa M., Jakes R., Lawler S., Cuenda
A., and Cohen P. (1997) Phosphorylation of
microtubule-associated protein tau by stress-
activated protein kinases. FEBS Lett. 409, 57–62.
Goldstein L. S. B. (2001) Kinesin molecular motors:
transport pathways, receptors, and human disease.
PNAS 98, 6999–7003.
Götz J., Chen F., Van Dorpe J., and Nitsch R. M. (2001)
Formation of neurofibrillary tangles in P301L tau
transgenic mice induced by Aβ42 fibrils. Science 293,
1491–1495.
Greenwood J. A., Scott C. W., Spreen R. C., Caputo
C. B., and Johnson G. V. (1994) Casein kinase II pref-
erentially phosphorylates tau isoforms containing
an amino-terminal insert. Identification of threo-
nine 39 as the primary phosphate acceptor. J. Biol.
Chem. 269, 4373–4380.
Gunawardena S. and Goldstein L. S. B. (2001) Dis-
ruption of axonal transport and neuronal viability
by amyloid precursor protein mutations in
Drosophila. Neuron 32, 389–401.
Hanemaaijer R. and Ginzburg I. (1991) Involvement
of mature tau isoforms in the stabilization of neu-
rites in PC12 cells. J. Neurosci. 30, 163–171.
NeuroMolecular Medicine
163
Harada A., Oguchi K., Okabe S., Kuno J., Terada S.,
Oshima T., et al. (1994) Altered microtubule orga-
nization in small-calibre axons of mice lacking tau
protein. Nature 369, 488–491.
Hasegawa M., Smith M. J., and Goedert M. (1998) Tau
proteins with FTDP-17 mutations have reduced abil-
ity to promote microtubule assembly. FEBS Lett. 437,
207–210.
Hong M., Zhukareva V., Vogelsberg-Ragaglia V.,
Wszolek Z., Reed L., Miller B. I., et al. (1998)
Mutation-specific functional impairments in dis-
tinct tau isoforms of hereditary FTDP-17. Science
282, 1914–1917.
Hutton M., Lendon C. L., Rizzu P., Baker M., Froelich
S., Houlden H., et al. (1998) Association of
missense and 5′-splice-site mutations in tau with
the inherited dementia FTDP-17. Nature 393,
702–705.
Iqbal K., Grundke-Iqbal I., Zaidi T., Merz P. A., Wen
G. Y., Shaikh S. S., et al. (1986) Defective brain micro-
tubule assembly in Alzheimer’s disease. Lancet 2,
421–426.
Ishihara T., Higuchi M., Zhang B., Yoshiyama Y., Hong
M., Trojanowski J. Q., et al. (2001a) Attenuated neu-
rodegenerative disease phenotype in tau transgenic
mouse lacking neurofilaments. J. Neurosci. 21,
6026–6035.
Ishihara T., Zhang B., Higuchi M., Yoshiyama Y.,
Trojanowski J. Q., and Lee V. M.-Y. (2001b) Age-
dependent induction of congophilic neurofibrillary
tau inclusions in tau transgenic mice. Am. J. Pathol.
158, 555–562.
Jackson G. R., Wiedau-Pazos M., Sang T.-K.,
Wagle N., Brown C. A., Massachi S., et al. (2002)
Human wild-type tau interacts with wingless
pathway components and produces neuro-
fibrillary pathology in Drosophila. Neuron 34,
509–519.
Johnson G. V. (1992) Differential phosphorylation of
tau by cyclic AMP-dependent protein kinase and
Ca2+/calmodulin-dependent protein kinase II: meta-
bolic and functional consequences. J. Neurochem. 59,
2056–2062.
Kamal A., Stokin G. B., Yang Z., Xia C.-H., and Gold-
stein S. B. (2000) Axonal transport of amyloid pre-
cursor protein is mediated by direct binding to the
kinesin light chain subunit of kinesin-1. Neuron 28,
449–459.
Kamal A., Almenar-Queralt A., Leblanc J. F., Roberts
E. A., and Goldstein L. S. B. (2001) Kinesin-mediated
axonal transport of a membrane compartment con-
taining β-secretase and presenilin-1 requires APP.
Nature 414, 643–647.
Volume 2, 2002
164
Koo E. H., Lansbury P. T. Jr, and Kelly J. W. (1999) Amy-
loid diseases: abnormal protein aggregation in neu-
rodegeneration. PNAS 96, 9989–9990.
Khatoon S., Grundke-Iqbal I., and Iqbal K. (1992) Brain
levels of microtubule-associated protein tau are ele-
vated in Alzheimer’s disease: a radioimmunoslot-
blot assay for nanograms of the protein. J. Neurochem.
59, 750–753.
Ksiezak-Reding H., Binder L. I., and Yen S. H. (1988)
Immunochemical and biochemical characterization
of tau protein in normal and Alzheimer’s disease
brains with Alz 50 and Tau-1. J. Biol. Chem. 263,
7948–7953.
Künzi V., Glatzel M., Nakano M. Y., Greber U. F.,
Van Leuven F., and Aguzzi A. (2002) Unhampered
prion neuroinvasion despite impaired fast axonal
transport in transgenic mice overexpressing four-
repeat tau. J. Neurosci. 22, 7471–7477.
Lambert M. P., Barlow A. K., Chromy B. A., Edwards
C., Freed R., Liosatos M., et al. (1998) Diffusible,
nonfibrillar ligands derived from Aβ1-42 are potent
central nervous system neurotoxins. PNAS 95,
6448–6453.
Lee G., Neve R. L., and Kosik K. S. (1989) The micro-
tubule binding domain of tau protein. Neuron 2,
1615–1624.
Lewis J., Dickson D. W., Lin W.-L., Chisholm L., Corral
A., Jones G., Yen S.-H., et al. (2001) Enhanced
neurofibrillary degeneration in transgenic mice
expressing mutant tau and APP. Science 293,
1487–1491.
Lucas J. J., Hernandez F., Gomez-Ramos P., Moran
M. A., Hen R. and Avila J. (2001) Decreased nuclear
β-catenin, tau hyperphosphorylation and neu-
rodegeneration in GSK-3β conditional transgenic
mice. EMBO J. 20, 27–39.
Mandelkow E.-M., Biernat J., Drewes G., Gustke N.,
Trinczek B., and Mandelkow E. (1995) Tau domains,
phosphorylation, and interaction with micro-
tubules. Neurobiol. Aging 16, 355–363.
Matsuo E. S., Shin R.-W., Billingsley M. L., Van de Voorde
A., O´Connor M., Trojanowski J. Q., et al. (1994)
Biopsy-derived adult human tau is phosphorylated
at many of the same sites as Alzheimer’s disease
paired helical filament tau. Neuron 13, 989–1002.
Mattson M. P. (1997) Cellular actions of beta-amyloid
precursor protein and its soluble and fibrillogenic
derivatives. Physiol. Rev. 77, 1081–1132.
Mattson M. P., Fu W., Waeg G., and Uchida K. (1997)
4-Hydroxynonenal, a product of lipid peroxida-
tion, inhibits dephosphorylation of the micro-
tubule-associated protein tau. Neuroreport 8,
2275–2281.
NeuroMolecular Medicine
Terwel et al.
Michel G., Mercken M., Murayama M., Noguchi K.,
Ishiguro K., Imahori K., et al. (1998) Characteriza-
tion of tau phosphorylation in glycogen synthase
kinase-3β and cyclin dependent kinase-5 activator
(p23) transfected cells. BBA 1380, 177–182.
Moechars D., Dewachter I., Lorent K., Reverse D.,
Baekelandt V., Naidu A., et al. (1999) Early pheno-
typic changes in transgenic mice that overexpress
different mutants of amyloid precursor protein in
brain. J. Biol. Chem. 274, 6483–6492.
Morfini G., Szebenyi G., Elluru R., Ratner N., and Brady
S. T. (2002) Glycogen synthase kinase 3 phospho-
rylates kinesin light chains and negatively regulates
kinesin-based motility. EMBO J. 21, 281–293.
Nakazato Y., Sasaki A., Hirato J., and Ishida Y. (1984)
Immunohistochemical localization of neurofila-
ment protein in neuronal degenerations. Acta
Neuropathol. (Berl.) 64, 30–36.
Norlund M. A., Lee J. M., Zainelli G. M., and Muma
N. A. (1999) Elevated transglutaminase-induced
bonds in PHF tau in Alzheimer’s disease. Brain Res.
851, 154–163.
Nuydens R., Van den Kieboom G., Nolten C., Verhulst
C., Van Osta P., Spittaels K., et al. (2002) Coexpres-
sion of GSK-3β corrects phenotypic aberrations of
dorsal root ganglion cells, cultured from adult
transgenic mice overexpressing human protein tau.
Neurobiol. Dis. 9, 38–48.
Patrick G. N., Zukerberg L., Nikolic M., De La Monte
S., Dikkes P., and Tsai L. H. (1999) Conversion of
p35 to p25 deregulates cdk5 activity and promotes
neurodegeneration. Nature 402, 615–622.
Pedersen W. A., Culmsee C., Ziegler D., Herman J. P.,
and Mattson M. P. (1999) Aberrant stress response
associated with severe hypoglycemia in a trans-
genic mouse model of Alzheimer’s disease. J. Mol.
Neurosci. 13, 159–165.
Pei J. J., Braak E., Braak H., Grundke-Iqbal I., Iqbal K.,
Winblad B., et al. (1999) Distribution of active glyco-
gen synthase kinase-3β (GSK-3β) in brains staged
for Alzheimer disease neurofibrillary changes.
J. Neuropathol. Exp. Neurol. 58, 1010–1019.
Perry G., Kawai M., Tabaton M., Onorato M., Mulvihill
P., Richey P., et al. (1991) Neuropil threads
of Alzheimer’s disease show a marked alteration of
the normal cytoskeleton. J. Neurosci. 11, 1748–1755.
Rapoport M. and Ferreira A. (2000) PD98059 pre-
vents neurite degeneration induced by fibrillar
β-amyloid in mature hippocampal neurons.
J. Neurochem. 74, 125–133.
Rapoport M., Dawson H. N., Binder L. I., Vitek M. P.,
and Ferreira A. (2002) Tau is essential to β-amyloid-
induced neurotoxicity. PNAS 99, 6364–6369.
Volume 2, 2002
Axonal Transport in Alzheimer’s Disease
Rouleau G. A., Clarke A. W., Rooke K., Pramatarova
A., Krizus A., Suchowersky O., et al. (1996) SOD1
mutation is associated with accumulations of neu-
rofilaments in amyotrophic lateral sclerosis. Ann.
Neurol. 39, 128–131.
Ruben G. C., Ciardelli T. L., Grundke-Iqbal I., and Iqbal
K. (1997) Alzheimer disease hyperphosphorylated
tau aggregates hydrophobically. Synapse 27,
208–229.
Sawamura N., Gong J.-S., Garver W. S., Heidenreich
R. A., Ninomiya H., Ohno K., et al. (2001) Site-
specific phosphorylation of tau accompanied by
activation of mitogen-activated protein kinase
(MAPK) in brains of Niemann-Pick type C mice.
J. Biol. Chem. 276, 10,314–10,319.
Scott C. W., Blowers D. P., Barth P. T., Lo M. M., Salama
A. I., and Caputo C. B. (1991) Differences in the abil-
ities of human tau isoforms to promote microtubule
assembly. J. Neurosci. Res. 30, 154–162.
Shankar S. K., Yanagihara R, Garruto R. M., Grundke-
Iqbal I, Kosik K. S., and Gajdusek D. C. (1989)
Immunocytochemical characterization of neu-
rofibrillary tangles in amyotrophic lateral sclerosis
and parkinsonism-dementia of Guam. Ann. Neurol.
25, 146–151.
Sisodia S. S. and St. George-Hyslop P. H. (2002)
γ-secretase, Notch, Aβ and Alzheimer’s disease:
where do the presenilins fit in? Nature Rev. Neurosci.
3, 281–290.
Spillantini M. G., Murrell J. R., Goedert M., Farlow
M. R., Klug A., and Ghetti B. (1998) Mutation in the
tau gene in familial multiple system tauopathy with
presenile dementia. PNAS 95, 7737–7741.
Spittaels K., Van den Haute C., Van Dorpe J., Bruynseels
K., et al. (1999) Prominent axonopathy in the brain
and spinal cord of transgenic mice overexpressing
four-repeat human tau protein. Am. J. Pathol. 155,
2153–2165.
Spittaels K., Van den Haute C., Van Dorpe J., Geerts
H., Mercken M., Bruynseels K., et al. (2000) Glyco-
gen synthase kinase-3β phosphorylates tau and res-
cues axonopathy in the central nervous system of
human four-repeat tau transgenic mice. J. Biol. Chem.
275, 41,340–41,349.
Spittaels K., Van den Haute C., Van Dorpe J., Terwel
D., Vandezande K., Lasrado R., et al. (2002) Neona-
tal neuronal over-expression of glycogen synthase
kinase-3β reduces brain-size in transgenic mice. 113,
797–808.
NeuroMolecular Medicine
165
Stamer K., Vogel R., Thies E., Mandelkow E., and
Mandelkow E.-M. (2002) Tau blocks traffic of
organelles, neurofilaments, and APP vesicles in
neurons and enhances oxidative stress. J. Cell Biol.
156, 1051–1063.
Tesseur I., Van Dorpe J., Spittaels K., Van den Haute
C., Moechars D., and Van Leuven F. (2000) Expres-
sion of human apolipoprotein E4 in neurons causes
hyperphosphorylation of protein tau in the brains
of transgenic mice. Am. J. Pathol. 156, 951–964.
Tesseur I., Van Dorpe J., Bruynseels K., Bronfman, F.,
Sciot R., Van Lommel A., et al. (2000) Prominent
axonopathy and disruption of axonal transport in
transgenic mice expressing human apolipoprotein
E4 in neurons of brain and spinal cord. Am. J. Pathol.
157, 1495–1510.
Trinczek B., Ebneth A., Mandelkow E.-M., and Man-
delkow E. (1999) Tau regulates the attachment/
detachment but not the speed of motors in
microtubule-dependent transport of single
vesicles and organelles. J. Cell Sci. 112, 2355–2367.
Van Dorpe J., Smeijers L., Dewachter I., Nuyens D., Spit-
taels K., Van den Haute C., et al. (2002) Promiment
cerebral amyloid angiopathy in transgenic mice
overexpressing the London mutant of human APP
in neurons. Am. J. Pathol. 157, 1283–1298.
Voikar V, Rauvala H, and Ikonen E. (2002) Cognitive
deficit and development of motor impairment in a
mouse model of Niemann-Pick type C disease.
Behav. Brain Res. 132, 1–10.
Walsh D. M., Tseng B. P., Rydel R. E., Podlisney
M. B., and Selkoe D. J. (2000) The oligomerization
of amyloid β-protein begins intracellularly in cells
derived from human brain. Biochemistry 39,
10,831–10,839.
Wang J. Z., Grundke-Iqbal I., and Iqbal K. (1996) Gly-
cosylation of microtubule-associated protein tau:
an abnormal posttranslational modification in
Alzheimer’s disease. Nat. Med. 2, 871–875.
Williamson R., Scales T., Clark B. R., Gibb G., Reynolds
C. H., Kellie S., et al. (2002) Rapid tyrosine phos-
phorylation of neuronal proteins including tau and
focal adhesion kinase in response to amyloid-β pep-
tide exposure: involvement of src family protein
kinases. J. Neurosci. 22, 10–20.
Yoshida H. and Ihara Y. (1993) Tau in paired helical fil-
aments is functionally distinct from fetal tau: assem-
bly incompetence of paired helical filament tau.
J. Neurochem. 61, 1183–1186.
Volume 2, 2002