Seminars in Cell & Developmental Biology 15 (2004) 4549

Tau protein and neurodegeneration

Michel Goedert
Medical Research Council Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH, UK

Abstract
Tau protein is the major component of the intracellular filamentous deposits that define a number of neurodegenerative diseases. They
include the largely sporadic Alzheimers disease, progressive supranuclear palsy (PSP), corticobasal degeneration (CBD), Picks disease
(PiD), argyrophilic grain disease, as well as the inherited frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17).
The identification of mutations in Tau as the cause of FTDP-17 established that dysfunction or misregulation of tau protein is sufficient to
cause neurodegeneration and dementia. At an experimental level, the new understanding is leading to the development of good transgenic
animal models of the tauopathies.
2004 Elsevier Ltd. All rights reserved.
Keywords: Alternative mRNA splicing; Amyloid; Frontotemporal dementia; Microtubule assembly; Neurodegeneration; Tau protein

1. Introduction
Recent progress in our understanding of some of the most
common neurodegenerative diseases was made possible by
the coming together of two independent lines of research.
First, the biochemical study of the neuropathological lesions
that define these diseases led to the identification of their
molecular components. Second, the study of familial forms
of disease led to the identification of gene defects that cause
the inherited variants of the different diseases. Remarkably,
in most cases, the defective genes were found to encode or
increase the expression of the main components of the neuropathological lesions. It has therefore been established that
the basis of the familial forms of these diseases is a toxic
property conferred by mutations in the proteins that make
up the filamentous lesions. A corollary of this insight is that
a similar toxic property might also underlie the sporadic
disease forms.
Alzheimers disease (AD) is the most common neurodegenerative disease. Neuropathologically, it is defined by the
presence of abundant extracellular neuritic plaques made
of the -amyloid peptide and intraneuronal neurofibrillary
lesions made of the microtubule-associated protein tau [1].
Similar tau lesions, in the absence of extracellular deposits,
are also the defining characteristic of a number of other
neurodegenerative diseases, the best known of which are
progressive supranuclear palsy (PSP), corticobasal degen-

Tel.:

+44-1223-402036; fax: +44-1223-402197.

E-mail address: mg@mrc-lmb.cam.ac.uk (M. Goedert).

1084-9521/$ see front matter 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.semcdb.2003.12.015

eration (CBD) and Picks disease (PiD) [2]. Until recently,

there was no genetic evidence implicating tau protein in the
neurodegenerative process. This changed with the discovery
of tau gene mutations in a familial form of frontotemporal dementia and parkinsonism [35]. Here I review the
evidence implicating tau protein in this and other neurodegenerative diseases.

2. Tau isoforms in human brain and their

interactions with microtubules
Tau is a microtubule-binding protein that is believed to
be important for the assembly and stabilization of microtubules. In nerve cells, tau is normally found in axons, but
in the tauopathies it is redistributed to the cell body and
dendrites. In normal adult human brain, there are six isoforms of tau, produced from a single gene by alternative
mRNA splicing [6]. They differ from one another by the
presence or absence of a 29- or 58-amino acid insert in the
amino-terminal half of the protein and by the inclusion, or
not, of a 31-amino acid repeat, encoded by exon 10 of Tau,
in the carboxy-terminal half of the protein. The exclusion
of exon 10 leads to the production of three isoforms, each
containing three repeats, and its inclusion leads to a further
three isoforms, each containing four repeats. The repeats
constitute the microtubule-binding region of tau protein. In
normal adult human cerebral cortex, there are similar levels
of three-repeat and four-repeat isoforms. In developing human brain, only the shortest isoform (three repeats and no
amino-terminal inserts) is expressed.

46

M. Goedert / Seminars in Cell & Developmental Biology 15 (2004) 4549

The tau molecule can be subdivided into an amino-terminal

domain that projects from the microtubule surface and
the carboxy-terminal microtubule-binding domain. Recent
structural work using gold-labelled tau co-assembled with
microtubules has begun to shed light on the way that
tau protein interacts with microtubules. When bound to
taxol-stabilized microtubules, tau was found to localize
along the outer ridges of the microtubule protofilaments [7].
A different conclusion was reached when tau was bound
to microtubules in the absence of taxol. The tau repeats
were now found to bind to the inner surface of the microtubule, close to the taxol-binding site on -tubulin [8].
Taxol binds to a site on -tubulin, where -tubulin has a
conserved extra loop of eight amino acids. Interestingly,
this extended loop has significant sequence homology with
the tau repeats. Since tubulin cannot have evolved to bind
taxol, this work may answer the question of what type of
natural substrate binds to this pocket in -tubulin. In this
model, part of the proline-rich region of tau must provide
the link between the amino-terminal projection domain on
the outside of the microtubule and the repeat motifs on
the inside surface. It could thread through one of the holes
between protofilaments.

their tau isoform compositions vary between diseases [2],

it is the repeat region of tau that forms the core of the filament, with the amino- and carboxy-terminal regions forming
a fuzzy coat around the filament [9].
The discovery that incubation of bacterially expressed
human tau with sulphated glycosaminoglycans leads to
bulk assembly of tau filaments [10,11] made it possible
to obtain structural information [12]. Filaments assembled
from either three- or four-repeat tau showed cross- structure by selected area electron diffraction, X-ray diffraction
from macroscopic fibres and Fourier transform infrared
spectroscopy. This work was extended to PHFs and SFs extracted from diseased human brain. There had been controversy in the literature with regard to the internal molecular
fine structure of these filaments. The difficulty had been to
produce from human brain pure preparations of filaments
for analysis. This problem was circumvented by using selected area diffraction from small groups of filaments of
defined morphology. Using this approach, PHFs and SFs
had a clear cross- structure, which is the defining feature
of amyloid fibres. They share this structure with the extracellular deposits present in the systemic and organ-specific
amyloid diseases. It is therefore appropriate to consider the
tauopathies a form of brain amyloidosis.

3. Tau filaments as nerve cell amyloid

4. Mutations causing tauopathy
In human diseases with tau pathology, the normally soluble tau protein is present in an abnormal filamentous form
(Fig. 1). It is also hyperphosphorylated. In AD, tau filaments
consist principally of paired helical filaments (PHFs), with
straight filaments (SFs) being a minority species. Electron
micrographs of negatively stained isolated filaments show
images in which the width of the filament varies between
about 8 and 20 nm with a spacing between cross-overs of
about 80 nm [1]. Although the filament morphologies and

Fig. 1. Electron micrograph of a negatively stained preparation of tau

filaments of the Alzheimer-type. Examples of the characteristic PHF (P)
and SF (S) are indicated. Scale bar, 100 nm.

Frontotemporal dementias occur as familial forms and,

more commonly, as sporadic diseases. They are characterized by a remarkably circumscribed atrophy of the frontal
and temporal lobes of the cerebral cortex, often with
additional, subcortical changes. In 1994, an autosomaldominantly inherited form of frontotemporal dementia with
parkinsonism was linked to chromosome 17q21.2 [13].
Subsequently, other familial forms of frontotemporal dementia were found to be linked to this region, resulting in
the denomination frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17) for this class of
disease. All cases of FTDP-17 have so far shown a filamentous pathology made of hyperphosphorylated tau protein. In
June 1998, the first mutations in Tau in FTDP-17 patients
were reported [35]. Currently, 31 different mutations have
been described in over 80 families with FTDP-17.
Tau mutations are either missense, deletion or silent mutations in the coding region, or intronic mutations located
close to the splice-donor site of the intron following the alternatively spliced exon 10 (Fig. 2). Functionally, they fall
into two largely non-overlapping categoriesthose whose
primary effect is at the protein level and those that influence the alternative splicing of tau pre-mRNA. Most missense mutations reduce the ability of tau protein to interact
with microtubules, as reflected by a reduction in the ability
of mutant tau to promote microtubule assembly [14,15]. A
number of mutations in Tau may cause FTDP-17, at least in
part, by promoting the aggregation of tau protein [16,17].

M. Goedert / Seminars in Cell & Developmental Biology 15 (2004) 4549

47

Fig. 2. Mutations in the tau gene in frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17). (a) Schematic diagram of the six
tau isoforms (352441 amino acids) that are expressed in adult human brain, with mutations in the coding region indicated using the numbering of the
441 amino acid isoform. Nineteen missense mutations, two deletion mutations and three silent mutations are shown. The six tau isoforms are produced
by alternative mRNA splicing from a single gene. They differ by the presence or absence of three inserts, shown in red (encoded by exon 2), green
(encoded by exon 3) and yellow (encoded by exon 10), respectively. (b) Stem-loop structure in the pre-mRNA at the boundary between exon 10 and
the intron that follows it. Nine mutations are shown, two of which (S305N and S305S) are located in exon 10. They destabilize the stem-loop structure,
resulting in increased inclusion of exon 10 and the relative overproduction of four-repeat tau. Exon sequences are boxed and shown in capital, with
intron sequences being shown in lower-case letters. The two-dimensional representation of the tau exon 10 splicing regulatory element RNA is based on
the three-dimensional structure determined in [34].

Most of these mutations lead to the formation of nerve cell

inclusions that consist of filaments made of all six tau isoforms. Mutations P301L and P301S in exon 10 are exceptions, since they lead to the formation of neuronal and glial
inclusions that are made of tau isoforms with four-repeats.
The intronic mutations and most coding region mutations
in exon 10 increase the splicing of exon 10, thus changing the
ratio between three- and four-repeat tau isoforms, resulting
in the overproduction of soluble four-repeat tau [4,5]. This
leads to the formation of filaments made of four-repeat tau
in both nerve cells and glial cells. Approximately half of the
known Tau mutations have their primary effect at the RNA
level. They affect splicing enhancer or silencer sequences
in exon 10 or destabilise a stem-loop structure located at
the boundary between exon 10 and the intron that follows it
(Fig. 2b). Thus, to a significant degree, FTDP-17 is a disease
of the alternative mRNA splicing of exon 10 of the tau gene.
It follows that a correct ratio of three-repeat to four-repeat
tau isoforms is essential for preventing neurodegeneration
and dementia in mid-life.

5. Relevance for the sporadic tauopathies

The study of FTDP-17 has established that dysfunction
or misregulation of tau protein can cause neurodegeneration

and dementia. It follows that tau protein is most probably

also of central importance in the pathogenesis of diseases,
such as AD, PSP, CBD, and PiD. This is further underlined
by the fact that the aforementioned diseases are partially or
completely phenocopied by cases of FTDP-17 [2].
Eight missense mutations in Tau have been shown to give
rise to a clinical and neuropathological phenotype that is
closely related to PiD. The finding that overproduction of
four-repeat tau causes disease and leads to the assembly of
tau isoforms with only four repeats in nerve cells and glial
cells may shed light on the pathogenesis of PSP and CBD.
Both diseases are characterized by a neuronal and glial tau
pathology, with the filaments comprising only four-repeat
tau [2]. An association between progressive supranuclear
palsy and a dinucleotide repeat polymorphism in the intron
between exons 9 and 10 of Tau has been described [18].
The alleles at this locus carry 1115 repeats. The A0 allele,
with 11 repeats, has a frequency of over 90% in patients
with PSP and about 70% in controls. Subsequently, two
common tau haplotypes that differ at the nucleotide level,
but not at the level of the protein coding sequence, have
been reported [19]. Homozygosity of the more common allele H1 predisposes to PSP and CBD, but not to AD or PiD
[1922]. This work strongly suggests that dysfunction of
tau protein is of central importance in PSP and CBD, as it is
in PiD.

48

M. Goedert / Seminars in Cell & Developmental Biology 15 (2004) 4549

6. Transgenic animal models of the tauopathies

Animal models of the tauopathies are essential for elucidating the mechanisms by which dysfunction of tau
protein leads to neurodegeneration. In addition, they may
prove useful for the development and testing of novel
therapies.
The discovery of mutations in Tau in FTDP-17 is leading
to the production of transgenic mouse lines that express
mutant human tau protein in nerve cells and glial cells.
Abundant tau filaments were described in lines expressing
four-repeat tau with either the P301L or the P301S mutation
[23,24]. Filamentous tau protein was hyperphosphorylated
in a similar way to the human diseases and hyperphosphorylation at most sites appeared to precede the assembly of tau
into filaments. However, it remains to be seen whether hyperphosphorylation of tau is necessary for filament assembly
in brain and spinal cord. Widely differing conclusions have
been drawn from in vitro studies [10,25,26]. In a recent
study in transgenic mice expressing human P301L tau, an
increase in the phosphorylation of soluble tau resulted in increased filament formation, suggesting that phosphorylation
of tau can drive filament assembly in the brain [27]. In the
transgenic mice, filamentous tau deposits contained mostly
the mutant human tau protein, in line with in vitro findings showing that the P301L and P301S mutations strongly
enhance the heparin-induced assembly of tau protein into
filaments. In these lines, non-apoptotic nerve cell loss was
detected in the spinal cord, with signs of a neurogenic muscle atrophy. The mice suffered from a severe paraparesis.
In a mouse line expressing human P301L tau in oligodendrocytes, co-expression of mutant human -synuclein
resulted in the appearance of thioflavin S-positive staining
that was not observed with the single transgenic lines [28].
Moreover, in vitro experiments showed that -synuclein can
induce the formation of tau filaments, giving a possible explanation for the co-occurrence of tau and -synuclein inclusions in some neurodegenerative diseases.
In human P301L tau mouse lines, co-expression of mutant
human amyloid precursor protein or the intracerebral injection of -amyloid fibrils was reported to increase the number
of tangle-bearing nerve cells [29,30]. It thus appears that extracellular -amyloid deposits can exacerbate the intraneuronal pathology caused by the expression of mutant human
tau protein. In contrast to these findings, -amyloid deposits
failed to induce the formation of tau filaments in mice expressing wild-type human tau protein. Based on these findings, it would appear that -amyloid can promote, but not
induce, tau filament formation.
The existing transgenic mouse models of tauopathies
indicate a connection between the development of tau filaments and nerve cell degeneration. This contrasts with
Caenorhabditis elegans and Drosophila melanogaster,
where overexpression of mutant human tau resulted in
nerve cell degeneration, in the apparent absence of tau fil-

aments [3133]. It suggests that conformationally altered,

non-filamentous human tau protein can be neurotoxic, at
least in an invertebrate context.
The transgenic mouse lines described so far exhibit the
essential features of a human tauopathy, including the formation of abundant filaments made of hyperphosphorylated
tau protein and neurodegeneration. They will be invaluable
for a better understanding of the molecular mechanisms by
which the dysfunction of tau protein causes the death of
nerve cells and this may in turn lead to novel therapeutic
strategies. In particular, the administration of specific protein
kinase inhibitors to these mice ought to establish whether
hyperphosphorylation of tau is either necessary or sufficient
for filament formation or neurodegeneration.

Acknowledgements
I thank Dr R.A. Crowther for Figure 1.

References
[1] St George-Hyslop PH, Farrer LA, Goedert M. Alzheimer disease
and the frontotemporal dementias: diseases with cerebral deposition
of fibrillar proteins. In: Scriver CR, Beaudet AL, Sly WS, Valle D,
editors. The metabolic & molecular bases of inherited disease. New
York: McGraw-Hill; 2001. p. 587599.
[2] Lee VMY, Goedert M, Trojanowski JQ. Neurodegenerative
tauopathies. Annu Rev Neurosci 2001;24:112159.
[3] Poorkaj P, Bird TD, Wijsman E, Nemens E, Garruto RM, Anderson
L, et al. Tau is a candidate gene for chromosome 17 frontotemporal
dementia. Ann Neurol 1998;43:81525.
[4] Hutton M, Lendon CL, Rizzu P, Baker M, Froelich S, Houlden H, et
al. Association of missense and 5 -splice-site mutations in tau with
the inherited dementia FTDP-17. Nature 1998;393:7026.
[5] Spillantini MG, Murrell JR, Goedert M, Farlow MR, Klug A, Ghetti
B. Mutation in the tau gene in familial multiple system tauopathy
with presenile dementia. Proc Natl Acad Sci USA 1998;5:773741.
[6] Goedert M, Spillantini MG, Jakes R, Rutherford D, Crowther RA.
Multiple isoforms of human microtubule-associated protein tau: sequences and localization in neurofibrillary tangles in Alzheimers
disease. Neuron 1989;3:51926.
[7] Al-Bassam J, Ozer RS, Safer D, Halpain S, Milligan R. MAP2
and tau bind longitudinally along the outer ridges of microtubule
protofilaments. J Cell Biol 2002;157:118796.
[8] Kar S, Fan J, Smith MJ, Goedert M, Amos LA. Repeat motifs of tau
bind to the insides of microtubules in the absence of taxol. EMBO
J 2003;22:707.
[9] Wischik CM, Novak M, Thogersen HC, Edwards PC, Runswick MJ,
Jakes R, et al. Isolation of a fragment of tau derived from the core
of the paired helical filament of Alzheimer disease. Proc Natl Acad
Sci USA 1988;85:450610.
[10] Goedert M, Jakes R, Spillantini MG, Hasegawa M, Smith MJ,
Crowther RA. Assembly of microtubule-associated protein tau into
Alzheimer-like filaments induced by sulphated glycosaminoglycans.
Nature 1996;383:5503.
[11] Perez M, Valpuesta JM, Medina M, Montejo de Garcini E, Avila J.
Polymerization of tau into filaments in the presence of heparin: the
minimal sequence required for tautau interactions. J Neurochem
1996;67:118390.

M. Goedert / Seminars in Cell & Developmental Biology 15 (2004) 4549

[12] Berriman J, Serpell LC, Oberg KA, Fink AL, Goedert M, Crowther
RA. Tau filaments from human brain and from in vitro assembly
of recombinant protein show cross- structure. Proc Natl Acad Sci
USA 2003;100:90348.
[13] Wilhelmsen KC, Lynch T, Pavlou E, Higgins M, Nygaard TG. Localization of disinhibition-dementia-Parkinsonism-amyotrophy complex
to 17q21-22. Am J Hum Genet 1994;55:115965.
[14] Hasegawa M, Smith MJ, Goedert M. Tau proteins with FTDP-17
mutations have a reduced ability to promote microtubule assembly.
FEBS Lett 1998;437:20710.
[15] Hong M, Zhukareva V, Vogelsberg-Ragaglia V, Wszolek Z, Reed L,
Miller BI, et al. Mutation-specific functional impairments in distinct
tau isoforms of hereditary FTDP-17. Science 1998;282:19147.
[16] Nacharaju P, Lewis J, Easson C, Yen S, Hackett J, Hutton M, et
al. Accelerated filament formation from tau protein with specific
FTDP-17 mutations. FEBS Lett 1999;447:1959.
[17] Goedert M, Jakes R, Crowther RA. Effects of frontotemporal dementia FTDP-17 mutations on heparin-induced assembly of tau filaments. FEBS Lett 1999;450:30611.
[18] Conrad C, Andreadis A, Trojanowski JQ, Dickson DW, Kang D, Chen
X, et al. Genetic evidence for the involvement of tau in progressive
supranuclear palsy. Ann Neurol 1997;41:27781.
[19] Baker M, Litvan I, Houlden H, Adamson J, Dickson D, Perez-Tur
J, et al. Association of an extended haplotype in the tau gene with
progressive supranuclear palsy. Hum Mol Genet 1999;8:7115.
[20] Di Maria E, Tabaton M, Vigo T, Abbruzzese G, Bellone E, Donati C,
et al. Corticobasal degeneration shares a common genetic background
with progressive supranuclear palsy. Ann Neurol 2000;47:3747.
[21] Russ C, Powell JF, Zhao J, Baker M, Hutton M, Crawford F, et al. The
microtubule-associated protein tau gene and Alzheimers diseasean
association study and meta-analysis. Neurosci Lett 2001;314:926.
[22] Russ C, Lovestone S, Baker M, Pickering-Brown SM, Andersen PM, Furlong R, et al. The extended haplotype of the
microtubule-associated protein tau gene is not associated with Picks
disease. Neurosci Lett 2001;99:1568.
[23] Lewis J, McGowan E, Rockwood J, Melrose H, Nacharaju P, Van
Slegtenhorst M, et al. Neurofibrillary tangles, amyotrophy and progressive motor disturbance in mice expressing mutant (P301L) tau
protein. Nat Genet 2000;25:4025.

49

[24] Allen B, Ingram E, Takao M, Smith MJ, Jakes R, Virdee K, et

al. Abundant tau filaments and nonapoptotic neurodegeneration in
transgenic mice expressing human P301S tau protein. J Neurosci
2002;22:934051.
[25] Schneider A, Biernat J, von Bergen M, Mandelkow E, Mandelkow
EM. Phosphorylation that detaches tau protein from microtubules
(Ser262, Ser214) also protects it against aggregation into Alzheimer
paired helical filaments. Biochemistry 1999;38:354958.
[26] Alonso AC, Zaidi T, Novak M, Grundke-Iqbal I, Iqbal K. Hyperphosphorylation induces self-assembly of tau into tangles of paired
helical/straight filaments. Proc Natl Acad Sci USA 2001;98:69238.
[27] Noble W, Olm V, Takata K, Casey EOM, Meyerson J, Gaynor K,
et al. Cdk5 is a key factor in tau aggregation and tangle formation
in vivo. Neuron 2003;38:55565.
[28] Giasson BI, Forman MS, Higuchi M, Golbe LI, Graves CL,
Kotzbauer PT, et al. Initiation and synergistic fibrillization of tau
and alpha-synuclein. Science 2003;300:63640.
[29] Lewis J, Dickson DW, Lin WL, Chisholm L, Corral A, Jones G, et al.
Enhanced neurofibrillary degeneration in transgenic mice expressing
mutant tau and APP. Science 2001;293:148791.
[30] Gtz J, Chen F, van Dorpe J, Nitsch RM. Formation of neurofibrillary tangles in P301L tau transgenic mice induced by A42 fibrils.
Science 2001;293:14915.
[31] Wittmann CW, Wszolek MF, Shulman JM, Salvaterra PM, Lewis J,
Hutton M, et al. Tauopathy in Drosophila: neurodegeneration without
neurofibrillary tangles. Science 2001;293:7114.
[32] Jackson GR, Wiedau-Pazos M, Sang TK, Wagle S, Brown CA, Massachi S, et al. Human wild-type tau interacts with wingless pathway
components and produces neurofibrillary pathology in Drosophila.
Neuron 2002;34:50919.
[33] Kraemer BC, Zhang B, Leverenz JB, Thomas JH, Trojanowski JQ,
Schellenberg GD. Neurodegeneration and defective neurotransmission in a Caenorhabditis elegans model of tauopathy. Proc Natl Acad
Sci USA 2003;100:99805.
[34] Varani L, Hasegawa M, Spillantini MG, Smith MJ, Murrell JR,
Ghetti B, et al. Structure of tau exon 10 splicing regulatory element
RNA and destabilization by mutations of frontotemporal dementia
and parkinsonism linked to chromosome 17. Proc Natl Acad Sci
USA 1999;96:822934.