J. Biochem. 121, 179-188 (1997)

Physiology and Pathology of Tau Protein Kinases in Relation to

Alzheimer's Disease

Kazutomo Imahori1 and Tsuneko Uchida

Mitsubishi Kasei Institute of Life Sciences, 11 Minamiooya, Machida, Tokyo 194

Received for publication, November 27, 1996

Alzheimer's disease (AD) is characterized by neuronal cell death and two kinds of deposits,

neurofibrillary tangles (NFT) and senile plaques. The main component of NFT is paired

helical filaments (PHF), which mainly consist of hyperphosphorylated tau protein. Tau

protein kinases I and II were found as candidate enzymes responsible for hyperphos

phorylation of tau to induce the formation of PHF. Since prior phosphorylation of tau by

TPKII strongly enhanced the action of TPKI, it was thought that TPKII was involved in the

formation of PHF-tau in concert with TPKI. After cloning, TPKI was found to be identical

with glycogen synthase kinase 3ƒÀ (GSK3A), while TPKII consists of a novel 23kDa protein

activator and a catalytic subunit that is identical with cyclin-dependent kinase 5 (CDK5).

The phosphorylation sites on tau by TPKI and TPKII could account for the most, but not all,

of the major phosphorylation sites of fetal tau and PHF-tau. An antibody for a site

specifically phosphorylated by TPKI (Ser413) could identify all three neurofibrillary

lesions in the AD brain, and double staining for either TPKI or TPKII and NFT in the brain

of Down's syndrome patients clearly demonstrated that TPKI and TPKII are both associat

ed with NFT in vivo, suggesting that the level of TPKI or TPKII is elevated in AD brain by

some mechanism. On the other hand, the levels of both TPKs change developmentally, being

high in the neonatal period when the phosphorylation of fetal tau proceeds actively,

suggesting that the TPKI/TPKII cooperative system has an important physiological role in

the formation of neural networks. In AD brain, aberrant accumulation of amyloid-ƒÀ

protein (A,6) occurs ahead of the accumulation of PHF in NFT. When a primary culture of

embryonic rat hippocampus was treated with 20ƒÊM AƒÀ, induction of TPKI, extensive

phosphorylation of tau and then programmed cell death were observed, indicating that

TPKI induced by AƒÀ phosphorylates tau, followed by disruption of axonal transportation

and finally cell death. By using a yeast two hybrid system, TPKI was found to interact with

pyruvate dehydrogenase (PDH), which is a key enzyme in the glycolytic pathway. PDH was

phosphorylated in vitro by TPKI to reduce the activity converting pyruvate into acetyl

CoA, which is required for acetylcholine synthesis. In a primary culture of rat hippocampal

cells treated with AƒÀ, PDH was inactivated in inverse relation to the activation of TPKI,

resulting in accumulation of pyruvate or lactate, energy failure induced by the disturbance

of glucose metabolism, and a shortage of acetylcholine owing to deficiency of acetyl-CoA, all

of which are characteristic of AD brain. In cholinergic neurons such as those of the septum,

non-aggregated A,8, specifically AƒÀ (1-42), not AƒÀ (1-40), caused a shortage of acetylcholine

by activation of TPKI and inactivation of PDH without cell death.

Key words: Alzheimer's disease, amyloid-ƒÀ protein, phosphorylation of tau, pyruvate

dehydrogenase, tau protein kinases I and II.

Alzheimer's disease (AD) is a heterogeneous group of dementias that share common clinical symptoms, involving progressive cognitive impairments. This form of senile

dementia is characterized by two kinds of pathological

deposits in specific areas of the brain, called senile plaques

and neurofibrillary tangles (NFTs), and finally by severe

atrophy of the brain resulting from the extensive loss of

neuronal cells. The relationships among these three charac

teristics have not yet been established, and further patho

logical and biochemical evidence, is required to clarify the

etiology of Alzheimer's disease.

Senile plaques contain extracellular deposits of ƒÀ-amy

loid protein (ƒ¿ƒÀ) associated with degenerating nerve

processes known as dystrophic neurites. Initial deposits are

non-fibrillar (a diffuse plaque), but are progressively trans

formed into fibrils, giving rise to the characteristic amyloid

1 To whom correspondence should be addressed. Tel: +81-427.24

6222, Fax: +81-427-29-1252

Abbreviations: Af, amyloid-ƒÀ protein or J4-amyloid protein; Ach,

acetyl choline; AD, Alzheimer's disease; CDK5, cyclin dependent

kinase 5; DS, Down's syndrome; FITC, fluorescein isothiocyanate;

GSK3ƒ¿(ƒÀ), glycogen synthase kinase 3ƒ¿(ƒÀ); MAPK, mitogen

activated protein kinase; MARK, microtubule affinity regulating

kinase; NFT, neurofibrillary tangle; NT, neuropil thread; NP,

neuritic plaque; PDH, pyruvate dehydrogenase; PHF, paired helical

filament; TPK, tau protein kinase.

Vol. 121, No. 2, 1997 179

180 K. Imahori and T. Uchida

plaques. Neurofibrillary tangles form inside neuronal cells that die during the course of the disease and consist

primarily of abnormal paired helical filaments (PHFs). However, large numbers of senile plaques are found in some cognitively normal individuals, suggesting that not only their presence, but also the coexistence of NFTs is required for dementia. On the other hand, large numbers of NFTs in cerebral cortex and hippocampus closely correlate with the degree of dementia in Alzheimer's disease (1). Thus, the accumulation of neurofibrillary lesions may represent a final pathway that leads to neuronal cell death and neurodegeneration. Therefore, our attention was first directed to the formation and accumulation of PHF.

PHF consists of hyperphosphorylated tau (PHF-tau) (2, 3), and contains a small amount of ubiquitin (4). Tau is one of the microtubule-associated proteins and is specifically localized in the neuron. In the normal neuron, tau has the physiological role of promoting the assembly and stability of axonal microtubules, which are involved in intraneuronal transport (5, 6). During the course of Alzheimer's disease, tau becomes hyperphosphorylated at multiple sites and integrated into PHFs, losing its physiological functions. No protein kinase phosphorylating tau to give PHF-like state was known at that time. Such an enzyme, if found, could serve as a probe for investigating the mechanism of the formation and accumulation of PHF, as wwell as the changes of cellular metabolism involved in the degenerative process leading to neuronal death. Therefore, our initial goal was identification and characterization of the candidate kinase.

Identification of a candidate kinase for hyperphos

phorylation of tau leading to a PIIF-like stateThe characteristics of PHF-tau described in early reports

were an increase of apparent molecular weight (Mr) on SDS-gel owing to hyperphosphorylation (3), and the acquisition of immunoreactivity with anti-PHF polyclonal antibody (2) or the loss of immunoreactivity with anti-non

phosphorylated tau monoclonal antibody (tau-1), in addition to high resistance to proteolysis (7). Thus, three criteria were set up for identifying a candidate kinase as mentioned above: i.e., it must be capable of (1) phosphorylating tau, (2) inducing the mobility shift of tau on SDSpolyacrylamide gel electrophoresis, and (3) forming a PHF epitope on the phosphorylated tau. These criteria proved to be reasonable, because PHF-like structure was later shown to be constructed from highly phosphorylated whole tau

protein (A68) alone (8). Protein kinase activity meeting all three criteria was found in the microtubule proteins fraction of rat or bovine brain extracts and partially purified (9). This kinase is a serine/threonine kinase, which phosphorylates tau to form a PHF epitope, easily binds to tau, and does not depend on second messengers, such as cyclic AMP, cyclic GMP, Ca2+, and calmodulin. The phosphorylation catalyzed by this kinase was stimulated by tubulin under conditions of microtubule formation. This novel enzyme, different from known protein kinases, was designated as tau protein kinase [EC 2.7.1.135] by the Nomenclatture Committee of the International Union of Biochemistry (10). It is abbreviated as TPK.

On further purification, tau protein kinase was separated into two kinds of activities with different modes of phosphorylation of tau. These were purified to homogeneity and designated as TPKI and TPKII, respectively (11).

Relation between TPKI and TPKII

The two enzymes differed in the mode of phosphorylation

of tau. TPKI can phosphorylate native tau isolated from

normal brain, which is already phosphorylated to some

extent, but not completely dephosphorylated tau. In con

trast with TPKI, TPKII can phosphorylate both. TPKI can

induce an apparent molecular weight shift of tau and form

a PHF epitope by phosphorylating native tau, but TPKII

can not do so. Apparently, TPKI converts native tau to

PHF-tau, whereas TPKII seems to have no relation to PHF

formation (11). When dephosphorylated tau is initially

phosphorylated by TPKII, however, TPKI can then phos

phorylate it as well as native tau to generate a PHF epitope,

indicating that prior phosphorylation of tau by TPKII

enhances phosphorylation by TPKI (12). TPKII may

regulate the phosphorylation state of tau not only in normal

brain, but also in Alzheimer's disease brain, in concert with

TPKI.

Chemical nature of TPKI and TPKII

TPKI consists of a single polypeptide of 45kDa. Se

quence analyses of its cDNA proved that TPKI is identical

with "glycogen synthase kinase 3,8 (GSK3ƒÀ)" (13), one of

two isoforms of GSK3 that are encoded by different genes

(14). Mammalian GSK3ƒÀ is homologous to the product of

the Drosophila gene termed shaggy (sgg) or zeste-white 3

(zw-3), and is reported to rescue the sgg/zw-3 mutant

phenotype connected with wingless (wg) signal transduc

tion (15) and with Notch signalling (16). Furthermore,

GSK3 was recently shown to be required for ventral

differentiation (17). Therefore, mammalian GSK3 is also

considered to be involved in embryogenesis and neurogene

sis.

TPKII is a complex composed of two subunits, a catalytic

subunit of 30kDa and a regulatory subunit of 23kDa. Since

TPKII extensively phosphorylated histone H1 and MAP2

as well as tau, this enzyme was expected to be similar to the

cdc2 enzyme. The catalytic subunit of TPKII was proved by

sequence analysis of the cDNA to be identical with a "cdc2 -related kinase , PSSALRE/Cdk5" (18), which was

first found as one of 7 novel cdc2-related kinases cloned

from a human cell line (19). TPKII is an active complex of

CDK5 with a 23kDa protein; CDK5 monomer alone is

inactive (20). Therefore, the 23kDa protein (p23), not

cyclin, is regarded as a CDK5 activator in neuronal cells.

Cloning and sequencing of the 23kDa subunit revealed it to

be a C-terminal region of a novel protein with a molecular

weight of 34kDa (20). The sequence of the 34kDa protein,

p34, contains that of p23 and an additional 98 amino acids

at the N-terminus of p23, suggesting that p34 is a precursor

of the regulatory subunit, p23 (21). Certainly, no active

complex of CDK5 with p34 was detected in bovine brain

extract. However, it might be argued that p34 itself is a

CDK5 activator and p23 is the proteolytic derivative,

because recombinant p34 can be reconstituted with bacter

ially synthesized CDK5 to result in high kinase activity

(22). In the purification process of TPKII, enzymatic

activity is often localized in the high-molecular region.

These results suggest that TPKII may form a larger active

complex in vivo.

Our results on p23 are in good agreement with those on

the regulatory subunit, p25 of neuronal cdc2-like kinase

(nclk) (23), formerly called bovine proline-directed kinase

J. Biochem.

Pathology of Tau Protein Kinases in Alzheimer's Disease 181

(BPDK), which was purified from bovine brain extract as a

neurofilament kinase. These enzymes, TPK II and nclk, are

considered to be identical.

Phosphorylation of tau with TPKI and TPKII

The phosphorylation sites in tau protein by TPKI and

TPKII were first examined chemically. Protease digests of

tau maximally phosphorylated by the kinase in the pres

ence of [ƒÁ-32P]ATP, were separated by HPLC using a C4

column. In the case of TPKI, tau which had first been fully

phosphorylated with TPKII and cold ATP was used. The

sequences of phosphopeptides were determined using an

improved protocol in a gas-phase sequencer (24), and the

phosphorylation sites of tau by TPKII were identified as

Ser202, Thr205, Ser235, and Ser404 (25), while those by

TPKI were Ser199, Thr231, Ser396, and Ser413 (number

ing of amino acids is based on the longest human tau) (26)

(Table I and Fig. 1). We omitted ambiguous sites which

exhibited very low levels of phosphorylation or which were

localized in the extreme C-terminal region of long peptide

fragments.

Phosphorylation sites by TPKII have a common sequence

of SerPro or ThrPro, indicating that TPKII is a proline

directed kinase, though not all Ser/ThrPro sequences in tau

protein, were phosphorylated. Although TPKI/GSK3ƒÀ is

also regarded as a member of the class of proline-directed

kinases, one of the 4 TPKI sites, Ser413, is non-proline

directed site.

As a next step, we aimed to prepare antibodies specific

for these phosphorylation sites of TPKI and TPKII by using

chemically synthesized phosphopeptides as antigens, and to

confirm that TPKI and TPKII act at the phosphorylation

sites determined chemically (12, 27). Antigen dodecapep

tides containing phosphoserine or phosphothreonine in the

center were synthesized by the use of a phenyl group (28)

or a cyclohexyl group (in the case of aromatic or sulfur

containing amino acids) (29) as a protecting group for the

phosphate group. In addition, antibodies raised against

peptides phosphorylated at Ser262 or Ser422 were prepar

ed in the method described above. Each phosphopeptide

was named based on the species and the number of the

phosphorylated amino acid, e.g., PS199 or PT205. The

obtained antibodies against the phosphorylated 4 TPKI

sites, 4 TPKII sites, Ser262 and Ser422 were specific for

the corresponding phosphopeptide, and did not cross-react

with other phosphopeptides. Human recombinant tau

(htau40) phosphorylated by TPKII, reacted with all anti

bodies specific for the 4 TPKII sites, anti-PS202, anti-

TABLE I. Phosphorylation sites of tau in vivo (A) and those generated by various kinases reported so far as candidates to act in PHF formation (B).

aCited from f. 33. bCited from Ref. 34. cCited from Ref. 38. dSer205 was reported as a PHF-site by others (35, 36).

Fig. 1. Phosphorylation sites on tau by TPKI and II. The primary structure of the longest human tau is presented here. •«:

Phosphorylation sites by TPKI and TPKII are indicated by 1 and 2, respectively. •ª: Phosphorylation sites in PHF-tau (33). Rn: 31 or 32 amino

acid repeat region. •¡: 18 amino acid microtubule -binding element.

Vol. 121, No. 2, 1997

182 K. Imahori and T. Uchida

PT205, anti-PS235, and anti-PS404, confirming the results

of chemical analysis. TPKI in the presence of TPKII

incorporated phosphate into tau at 4 TPKI sites, Ser199,

Thr231, Ser396, and Ser413, inducing the largest shift-up

of apparent molecular weight of tau. Unexpectedly, TPKI

slightly phosphorylated tau at Ser262, but not at Ser422.

Accordingly, immunoblots of recombinant tau phosphor

ylated by TPKI and/or TPKII with these antibodies specific

for phosphorylation sites, clearly confirmed the 4 TPKI and

4 TPKII sites identified from the chemical analysis, as well

as the enhancement of TPKI activity after prior phosphor

ylation of tau with TPKII, that is, the TPKII/TPKI co

operative phosphorylation system (27). This immunoblot

method is advantageous to detect low levels of phosphor

ylation and changes of phosphorylation state in vitro and in

vivo (24).

These 8 TPKs sites are localized in the amino and

carboxyl-terminal flanking regions of the microtubule

binding domain of tau, as shown in Fig. 1. The microtubule

binding domain consists of the tandem 31 or 32 amino acid

repeat region, which includes 4 repeats in adult tau and 3

repeats in fetal tau (30). The sequences flanking the repeat

were indicated by binding studies to enhance the affinity of

this domain for microtubules, though they do not bind

directly with microtubules (31). In particular, the proline

rich region upstream of the repeats, which contains a

cluster of Ser/ThrPro phosphorylation sites, appears to be

essential for tubulin binding in vivo (32). Thus, phosphor

ylation by TPKI and TPKKKII markedly reduces the tubulin

assembly-promoting activity (26). Through its moduration

of the extent of phosphorylation of tau, TPKII appears to be

involved in the regulation of microtubule dynamics in

normal brain. In addition, TPKII sites, particularly in the

proline-rich region, are phosphorylated to a two to three

fold higher level at postnatal week 1 as compared with the

level in adult rat, suggesting that TPKII is developmentally

regulated in rat brain (12).

Recently, phosphorylation sites in PHF-tau from the

brain of Alzheimer's disease patients, in fetal tau from

neonatal rat brain (33) and in adult tau from adult rat brain

(34) have been determined by identifying phosphopeptides

by means of ion spray mass spectrometry followed by

sequencing of the ethanethiol-modified peptides, as shown

in Table I and Fig. 1. The degree of phosphorylation

increases in the order of normal adult tau <fetal tau

<PHF-tau. Nineteen sites identified in PHF-tau are all

localized in the N. and C-terminal regions of the micro

tubule binding domain except for Ser262. The extent of

phosphorylation at each site varies differently. Major

phosphorylation sites shown by double circles in Table I are

10 sites of the 19 sites. All the TPKI and TPKII sites except

Ser205 are included in these major sites. Ser205 was

identified as a PHF site by other workers (35, 36), in

accordance with our result. Three non-proline-directed

sites, Ser198, Ser400, and Ser409 could be phosphorylated

by TPKI/GSK3ƒÀ if Ser202, Ser404, and Ser413 were first

phosphorylated, as in fetal tau. Ser262 is considered to be

very important for microtubule-binding because it is locat

ed in the first repeat, having the highest affinity for

microtubules compared with the other three repeats. The

possibility that TPKI/GSK3ƒÀ is responsible for phosphor

ylation of tau at Ser262 in vivo still remains, because

TPKI/GSK3ƒÀ certainly phosphorylates a trace amount of

tau at Ser262 in vitro and this would be enhanced somewhat

in the presence of TPKII/(CDK5+p23) (27), and further,

the activity of GSK3ƒ¿, another isoform of GSK3ƒÀ, to

phosphorylate tau at Ser262, Ser324, and Ser356 is strong

ly stimulated in the presence of heparin (37). However, it

seems most probable at present that p110/MARK is the

kinase that phosphorylates Ser262 (38), because PKA,

another candidate, phosphorylates tau so weakly. Thus,

most of the phosphorylation sites in fetal tau and in

PHF-tau are modified by TPKI/TPKII, but one of the

major sites, Ser422, in PHF-tau is not notably phosphor

ylated by TPKI alone or by the TPKI/TPKII cooperative

system (27). Thus, it is reasonable to consider that other

kinases are also involved in the hyperphosphorylation of

tau to induce PHF-tau. Considering that all tau molecules in

PHF are not homogeneously hyperphosphorylated and that

the non-proline-directed sites that distinguish PHF-tau

from fetal tau are mostly minor sites, it is questionable

whether just phosphorylation at a crucial site or several

sites is essential for PHF formation. Nevertheless some

hyperphosphorylation, particularly in the two flanking

regions of the microtubule-binding domain, seems to be

essential at least for dissociation from the microtubules, in

addition to the initiation of self-assembly of PHF-tau.

Physiological roles of TPKI and TPKII

No information was available on the function of TPKI/

GSK3ƒÀ and TPKII/(CDK5 + p23) in normal brain when we

started our studies. We first demonstrated that TPKII sites

in tau are maximally phosphorylated at the first week

postnatally in rat brain (12). Phosphorylation sites in fetal

tau were mostly modulated by the TPKI/TPKII coopera

tive system as described above. We then examined de

velopmental change of both kinases in rat brain.

Northern blot analysis showed that CDK5 was expressed

predominantly in brain, though it is also ubiquitous in other

tissues, while p34, the precursor of the regulatory subunit

p23, is expressed specifically in brain (21). Other groups

also showed, using in situ hybridization of embryonic

mouse tissues, that p35 expression was restricted to the

central nervous system (CNS) (22), but CDK5 was ex

pressed in both the CNS and the peripheral nervous system

(39), indicating that the localizations of the two subunits

are slightly different.

Expression of p34 mRNA rapidly increased during

gestation day 15 to 18, reached the highest level just at

birth, remained abundant for 2 weeks in neonatal brain,

and then decreased gradually to about 10% of the maximum

in aged brain. Expression of CDK5 mRNA also changed

developmentally in a manner similar to p34 mRNA expres

sion, except that it remained at about 50% of the maximal

level in aged brain. In contrast, TPKII activity develop

mentally changed just in parallel with expression of p34

mRNA, but not of CDK5 mRNA , suggesting that it is controlled by p34 expression. Immunohistochemical

studies and in situ hybridization showed that both CDK5

and p34 are detectable from gestational day 12, when

proliferation of preneuronal cells comes to an end, and

CDK5 is expressed only in postmitotic neurons, but not in

glial cells or mitotically active cells (22). Therefore,

TPKII/(CDK5+p23) is involved in neurogenesis, but not

in cell division, particularly in sprouting for neural network

formation, which is a common feature of fetal brain (40, 41)

J. Biochem.

Pathology of Tau Protein Kinases in Alzheimer's Disease 183

and Alzheimer's disease brain (42). Recently, the subcell

ular localization of CDK5 and p34 proteins was examined

by indirect immunocytochemistry and confocal micros

copy, using cultured primary neurons derived from em

bryonal day 17-18 rat cortices (43). It was found that both

CDK5 and p34 are present in the growth cones of develop

ing neurons. Furthermore, expression of dominant-nega

tive mutants of CDK5 inhibited neurite outgrowth, which

was rescued by co-expression of the wild-type CDK5 and

p34. These observations suggest that TPKII/(CDK5 + p34)

plays a critical role in neurite outgrowth during neuronal

differentiation.

Developmental change of TPKI in rat brain was ex

amined by immunoblot analysis using monoclonal antibody

(T1.7) raised against TPKI (44) and polyclonal antibody

(C-1) raised against a C-terminal peptide of TPKI/GSK3ƒÀ

(45), which specifically recognized GSK3ƒÀ, but not

GSK3ƒ¿. The level of TPKI changed developmentally in a

similar way to that of TPKII, that is, it increased from

gestational day 12 (E12) to birth, was maintained at a high

level for 2 weeks after birth, and decreased in aged brain

(45), though TPKI reached the maximum level several

days later than TPKII. Immunohistochemistry with rabbit

antibody C-1 revealed that staining intensity in embryos

from E9 to E10 increased in both presumptive neural and

non-neural tissues. A further increase in staining, however,

was asssciatedd only with differentiating central and periph

eral neurons (44). In hippocampus and neocortex of 3 to

11-day-old rats, TPKI immunoreactivity was observed in

young axonal tracts and growing neuropil, where phos

phorylated Ser396 was also detected, while it declined to a

low level in a sections of adult brain, where phosphorylated

Ser-immunoreactivity was undetectable (46). These find

ings indicate that TPKI is also involved in the processes

leading to differentiation of neuronal cells and in the

extension of neuronal processes. Therefore, the TPKI/

TPKII cooperative system may play an immportant physi

ological role in the formation of neural networks.

Involvement of TPKs in PHF formation

In order to confirm that both TPKs are involved in PHF

formation, we carried out the following experiments. As

described above, about 19 sites are phosphorylatable in

PHF-tau (33) and 10 of them are major sites. Among these

sites, Ser413 is unique because it could be phosphorylated

only by TPKI, but not by any other kinase tested so far. In

other words, phosphorylation of Ser413 is the fingerprint of

TPKI. Probably Thr205 is the fingerprint of cyclin-depen

dent kinase. Both fingerprints are found in PHF. As the

next step, in order to show that TPKs are operating in PHF

formation, we examined the staining of AD brain using

anti-PS413 and anti-PT205, to detect the fingerprints of

TPKI and TPKII, respectively. Twelve AD brains and nine

control brains with no evidence of Alzheimer-like changes

were examined by staining for PS413 and PT205 epitopes

in the CAI subfield of the hippocampus, where neurode

generative change in AD brain is most prominent. In serial

vibrotome sections of formalin-fixed hippocampus, anti

PS413 and anti-PT205 identified three major lesions

associated with neurofibrillary degeneration (i. e., NFT,

NP, and NT) involving pyramidal neurons of the CAI

region in all twelve AD cases (47). In control experiments,

preabsorption of antisera with PS413 or PT205 abolished

immunostaining. On the other hand, the brains of non-de

mented individuals were faintly stained and the number of

stained neurons was small, being in agreement with the

previous report (48). These results suggested strongly that

both TPKs are operating in AD brain to generate their

specific epitopes.

Up-regulation of TPKs in AD brain

The preferential appearance of epitopes of TPKI and

TPKII in AD brain suggested up-regulation of these kinases

in AD brain. In order to test this, we tried to raise antibody

against TPKI/GSK3ƒÀ. We obtained antibodies against

synthetic peptides representing carboxyl-terminal frag

ments of rat TPKI/GSK3ƒÀ and its isoform GSK3ƒ¿, where

the difference in homology between these isoforms is most

marked (45). After brief microwave heating, intense

staining of pyramidal neurons of focal groups in AD brain

and their surrounding neuropil made a vivid contrast with

the weaker, diffuse reaction of the neighboring cells (Fig.

2A). The same region of the age-matched control brains was

also stained, but only faintly, and the number of stained

cells was small (Fig. 2B). Similar strong staining was

observed in the prefrontal cortex, entorhinal cortex, and

amygdala, but not in the cerebellum, which lacks AD

pathology. Preabsorption of the anti-serum with TPKI

peptide completely inhibited the immunostaining, but

preabsorption with GSK3ƒ¿-peptide had no effect (47).

Similar experiments were carried out on the brain of a

Down's syndrome (DS) patient (49). In this experiment,

colocalization of TPKI/GSK3ƒÀ was clearly revealed by

double immunostaining for TPKI/GSK3ƒÀ (labeled with

FITC) and tau-1 (labeled with Texas red). The size and

shape of the areas of strong TPKI labeling were quite

similar to those of tau-1-positive NFT. Neuropil threads

and degenerating neurites of senile plaque were also TPKI

positive in DS brain (49) (Fig. 2C).

As mentioned earlier, TPKII consists of two subunits and

the larger one is identical with CDK5. We therefore stained

AD brain by using commercially available anti-CDK5

antibody. The results obtained were very similar to those in

the case of TPKI. Double staining clearly indicated that

tau-l-positive NFT were also positive for CDK5. Neuropil

threads were negative for CDK5 in contrast with the case of

TPKI (49) (Fig. 2D). However our results clearly indicate

that the level of TPKI or TPKII is elevated in AD brain.

Induction of TPKI and neuronal cell death

The above results suggested to us that some factor(s)

causing induction of TPKs in neuronal cells may appear in

AD brain. There is a long-standing hypothesis that aberrant

accumulation of AƒÀ occurs in AD brain and is associated

with NFT formation and neuronal death (50, 51). In order

to test whether AƒÀ might be the putative inducer, we

adopted an in vitro approach. Primary cultures of embry

onic rat hippocampus were used for this purpose. This

tissue was treated with or without 20ƒÊM AƒÀ. At this

concentration, AƒÀ molecules aggregated and the solution

was turbid. After 3h, the tissue was stained by anti-TPKI

antibody and also by anti-PS396 antibody. Both antibodies

stained AƒÀ-treated cells strongly, whereas non-treated

cells were scarcely stained (27, 52). This suggests that in 3

h, AƒÀ induced TPKI, which in turn phosphorylated tau

proteins. The activity of TPKI was assayed quantitatively

Vol. 121, No. 2, 1997

184 K . Imahori and T. Uchida

Fig. 2. Immunocytochemistry of anti-TPKs in the human hippocampal CAI subfield. (A) AD brain stained with anti-TPKI

antibody. (B) Control brain with slight AD-like changes stained with anti-TPKI antibody. In pyramidal cells, the cell body, apical dendrite , basal neurite and proximal axon are stained. (C) Staining of Down's

syndrome brain with anti-TPKI antibody (top) , tau-1 (middle), and double exposure (bottom) . (D) Staining of Down's syndrome brain

with anti CDK5 antibody (top) , tau-1 and double exposure (bottom). Tau-I stained dephosphorylated NFT . Overlapping of blue and red

gave yellow.

J. Biochem.

Pathology of Tau Protein Kinases in Alzheimer's Disease 185

after immunoprecipitation with anti-TPKI . The specific activity of TPKI was elevated to more than two times the con

trol by the treatment with AƒÀ. As shown in Table I there are

several candidate kinases which might be involved in PHF

formation through phosphorylation of tau . We examined whether some of them, such as MAPK and GSK3a

, might also be induced by AƒÀ treatment. However neither of them

showed any increase in activity. Thus , TPKI appears to be

primarily involved in AƒÀ-induced phosphorylation of tau.

In parallel with the enzyme assay, we examined the

survival of the cells. After incubation for 24 h , the survival of the cells were reduced to less than 10% of that in the

control culture (culture without AƒÀ), suggesting neurotox

icity of AƒÀ, as reported by several authors (53, 54). In

order to clarify the nature of TPKI activation, we synthe

sized both sense and antisense oligonucleotides of the TPKI

gene based on the cDNA sequence (52). Sense or antisense

oligonucleotide was added to an incubation mixture con

taining A6. After incubation for 3 h with antisense oligonu

cleotide, the activity of TPKI was suppressed to the same

level as that of the control. On the other hand, sense

oligonucleotide had no ability to suppress TPKI activity

(52). Antisense oligonucleotide was also effective to rescue

the cells from death. After 24 h, survival of the cells

incubated with both AƒÀ, and antisense oligonucleotide was

about 80% of the control while sense oligonucleotide had no

effect on cell survival (52). These results indicate that (1)

AƒÀ induces net synthesis of TPKI and (2) TPKI thus

expressed acts to kill the cell. When actinomycin D or

cycloheximide was added to the medium containing A13 to

inhibit protein biosynthesis, cell death was prevented.

Thus, it was suggested that rat hippocampal neurons

succumb to programmed cell death triggered by AƒÀ and

that TPKI participates in this process in some way.

It is clear that induction of TPKI phosphorylates tau on

the one hand and kills the cell on the other. However, an

anti-phosphopeptide antibody such as anti-PS396 stained

the cell body uniformly, not just special structures such as

NFT. Nevertheless, the cells died. Thus, PHF formation is

not requisite for cell death, but phosphorylation of tau

protein itself results in cell death. Tau protein is a kind of

microtubule-associated protein, forming cross bridges

between microtubule molecules and stabilizing them. Thus,

phosphorylation near or at the three or four repeats

sequence inhibits tau functions by disfavoring binding of

tau with tubulin, which results in disruption of axonal

transportation and finally cell death.

Inactivation of pyruvate dehydrogenase by TPKI/

GSK3ƒÀ

The phosphorylation of tau protein may lead to cell death

via some indirect way such as destabilization of micro

tubules. On the other hand, since TPKI is an isoform of

GSK, it is conceivable that TPKI might interact with some

key enzyme of the glycolytic pathway. Thus, we searched

for other substrates of TPKI using a yeast two-hybrid

system. We constructed the fusion protein plasmid be

tween TPKI and GAL4 DNA-binding domain. This hybrid

protein was employed to screen a human brain cDNA

library tagged by GAL4 transcriptional activation domain.

By screening 4 million transformants, we obtained clones

encoding pyruvate dehydrogenase (PDH). Then we ex

amined whether TPKI could phosphorylate PDH (55). As

expected, incubation of porcine heart PDH complex with

purified TPKI resulted in ATP-dependent phosphorylation

of PDH. It is quite striking that, in parallel with this

phosphorylation, the activity of PDH decreased. Other

kinases, including TPKII, failed to inactivate PDH.

Since these results were obtained in an in vitro system we

examined the situation in an in vivo system using primary

cultures of rat hippocampal neurons, as used in the above

experiments. Since PDH exists inside mitochondria we first

confirmed that TPKI is present in mitochondria, both

biochemically and morphologically (56). On treatment with

AƒÀ, TPKI activity in mitochondria increased rapidly,

reached the maximum at 6 h after AƒÀ exposure and then

leveled off. PDH was inactivated in inverse proportion to

the AƒÀ-induced TPKI activity (Fig. 3A). After 12h, the

remaining activity of PDH was only 20% of the control.

When radioactive orthophosphate was present in the incu

bation mixture, a 6h AƒÀ-treatment resulted a 2-3-fold

increase of radioactive phosphate in PDH. Moreover, when

TPKI synthesis was blocked by the antisense oligonucleo

tide as described above, PDH activity showed a 2-3-fold

increase, corresponding to the reduction in the amount of

TPKI by the antisense oligonucleotide (55). Inactivation of

PDH should result a reduction of in the level of acetyl-CoA,

which is a key substance for ATP synthesis through the

TCA cycle. The level of ATP in AƒÀ-treated cells was indeed

remarkably lowered.

Fig. 3. Activation of TPKI/GSK3ƒÀ, inactiva

tion of PDH, and inhibition of Ach synthesis in

septum cells cultured with amyloid 8. (A)

Dependency of TPKI activation and PDH inactiva

tion on the concentration of AƒÀ (1-42). The

activity was measured after 12h incubation with

various concentrations of AƒÀ (1-42) as shown on

the ordinate. (B) Effects of AƒÀ (1-42) and AƒÀ (1

40) on the suppression of Ach synthesis. Intracel

lular level of Ach was determined by HPLC with

an electrochemical detector system after 12h

culture with AƒÀ.

Vol. 121, No. 2, 1997

186 K. Imahori and T. Uchida

The above results are extremely significant since they

can explain many phenomena observed in AD patients.

1) Since PDH complex is essential to energy formation,

especially in brain, where energy is supplied by glu

cose, abnormal activation of TPKI will severely reduce

brain energy, as reported for AD brain.

2) In AD brain, a 44% reduction in cerebral metabolic rate

of glucose and a fourfold increase of lactate production

were observed, though the cerebral blood flow and

cerebral metabolic rate of oxygen were found to be

unaltered (57, 58). Those results can be well under

stood if PDH activity is blocked.

3) In postmortem studies, PDH activity was found to be

greatly decreased in AD brain (59-61).

These considerations lead to the following hypothesis for

AD pathogenesis in terms of the pathological function of

TPKI. AƒÀ interacts with neurons and activates TPKI. The

activated TPKI causes extensive phosphorylation of tau

and destabilizes microtubules, resulting in impaired axonal

transport and neuronal death. On the other hand, the

activated TPKI in mitochondria will phosphorylate and

inactivate PDH, resulting in dysfunction of glucose metabo

lism, which also contributes to neuronal death through

energy failure.

Inhibition of acetylcholine synthesis by TPKI

Acetyl-CoA is also important for acetylcholine (Ach)

production, and in cholinergic neurons, the inactivation of

PDH will cause a reduction of Ach synthesis, which will

impair release of Ach and synapse-forming ability. Thus,

we examined the relation between PDH inactivation and

Ach synthesis, using primary cultures of cholinergic neu

rons such as rat septum. As expected, Ach level in the

cultured cells declined to 14% relative to the control after

AƒÀ treatment for 24 h. Under this condition AƒÀ did not

affect choline acetyltransferase or choline esterase activity

in cultured rat septum, indicating that the decreased level

of acetyl-CoA consequent upon inactivation of PDH caused

the reduction of Ach synthesis.

In this system, aggregation of AƒÀ to a fibrillary state was

not necessary and a concentration of 10 nM was enough to

cause a significant increase in TPKI activity and a signifi

cant reduction in the intracellular Ach level, when AƒÀ (1

42) was used. On the other hand, AƒÀ (1-40) was completely

ineffective at this concentration (62) (Fig. 3B). AƒÀ (1-42)

was not toxic for the cells under this condition. Soluble AƒÀ

appears specifically to activate mitochondrial TPKI and to

inactivate PDH, without accelerating tau phosphorylation.

It is reported that, in AD brain, there is a severe selective

loss of cholinergic neurons, mostly due to the decline of Ach

level (59, 63). Considering that a specific increase in

soluble AƒÀ (1-42) occurs in AD brain before amyloid

plaque formation, soluble AƒÀ (1-42) may act as an acti

vator of TPKI in the pathogenesis of AD.

Conclusion

We set out to identify the enzyme responsible for the

hyperphosphorylation of tau protein, which is believed to

trigger the formation of PHF structure. We purified two

candidate enzymes, tau protein kinases I and II, which could

convert tau protein into a PHF-like state. The sites phos

phorylated by these kinases were confirmed to be among

the phosphorylated sites of PHF. Although TPKI and II

have different substrate specificity, they operate in a

concerted manner. Unless stripped tau protein has been

first phosphorylated by TPKII, TPKI cannot react with it.

This suggested that TPKII behaves as regulator of TPKI. In

addition TPKII itself is a kind of regulatory enzyme. TPKII

consists of two subunits and the catalytic subunit (30kDa)

is inactive in the absence of the other subunit (23kDa).

TPKI and II cannot account for all the phosphorylated

sites of PHF. Nevertheless these kinases seem to play a

crucial role in PHF formation. First, although about 19 sites

are phosphorylated in PHF, they are not equally phos

phorylated. Among them, about 10 sites are regarded as

major phosphorylation sites. TPKI and II can act at most of

them. In addition, one of the major sites, Ser413, could be

phosphorylated only by TPKI, so far as tested.

Involvement of TPKs in NFT or PHF formation was

confirmed by immunohistochemistry. Antibody raised

against an antigen-peptide containing the fingerprint amino

acid in the center stained NFT very strongly. In addition,

antibody raised against TPKI stained pyramidal cells in the

CA1 subfield of AD brain, but scarcely stained those of the

normal control. These facts indicate that TPKI is up

regulated in AD brain. Since similar results were obtained

for CDK5, it is conceivable that TPKII is also up-regulated

in AD brain. If TPKs were activated, there should be some

factor in AD brain which induces or activates TPKs. We

consider that AƒÀ is a likely candidate. Indeed, when

hippocampal cells were cultured with AƒÀ, the activity of

TPKI increased to more than two times the control, which

in turn resulted in phosphorylation of tau proteins.

Since tau protein binds to and stabilizes microtubules its

moderate phosphorylation might be necessary to make

microtubules flexible, as required for plasticity. This may

explain the physiological function of TPKs, but hyperphos

phorylation of tau would result in neuronal pathogenesis by

converting tau protein into PHF-like structure.

TPKI phosphorylates not only tau proteins, but also

pyruvate dehydrogenase, which results in its inactivation.

Since PDH is a key enzyme in the glycolytic pathway its

inactivation will have many serious consequences. It will

result in failure of the TCA cycle and of ATP production,

which, of course, is very harmful to the cells. Accumulation

of lactate will occur, resulting in acidosis, as is actually seen

in AD brain. However the most serious effect will be the

inhibition of acetylcholine synthesis. This would affect

cholinergic neurons, and might easily lead to loss of

memory and cognitive ability, which is characteristic of

AD.

In summary, TPKs, which were isolated and character.

ized as the mediators of the hyperphosphorylation of tau

protein accounted for not only PHF formation, at least in

part, but also cell death, which characterizes AD brain. In

addition, our finding that TPKI can inactivate PDH by

phosphorylation could also explain the poor energy metabo

lism, accumulation of lactate and shortage of acetylcholine,

which are well known symptoms of AD. Thus, TPKs may be

the key to the pathogenesis of AD from the biochemical

point of view.

REFERENCES

1. Arriagada, P.A. , Growdon, J.H., Hedley-White, E.T., and Hyman, B.T. (1992) Neurofibrillary tangles but not senile

J. Biochem.

Pathology of Tau Protein Kinases in Alzheimer's Disease 187

plaques parallel duration and severity of Alzheimer's disease.Neurology 42, 631-639

2. Ihara, Y., Nukina, N., Miura, R., and Ogawara, M. (1986) Phosphorylated tau protein is integrated into paired helical filaments in Alzheimer's disease. J. Biochem. 99, 1807-1810

3. Grundke-Igbal, I., Iqbal, K., Tung, Y.-C., Quinlan, M., Wisniewski, H.M., and Binder, L.I. (1986) Abnormal phosphorylation of

the microtubule -associated protein r(tau) in Alzheimer cyto skeletal pathology. Proc. Natl. Acad. Sci. USA 83, e4913-4917

4. Mori, H., Kondo, J., and Ihara, Y. (1987) Ubiquitin is a component of paired helical filament in Alzheimer's disease. Science

235,1641-16445. Weingarten, M.D., Lockwood, A.H., Hwo, S.Y., and Kirschner,

M.W. (1975) A protein factor essential for microtubule assembly. Proc. Natl. Acad. Sci. USA 72, 1858-1862

6. Cleveland, D.W., Hwo, S.Y., and Kirschner, M.W. (1977) Purification of tau, a microtubule -associated protein that induces assembly of microtubules from purified tubulin. J. Mol. Biol. 116,207-225

7. Kondo, J., Honda, Y., Mori, H., Hamada, Y., Miura, R., Ogawara, R., and Ihara, Y. (1988) The carboxyl third of tau is tightly bound to paired helical filaments. Neuron 1, 827-834

8. Lee, V.M.-Y., Balin, B.J., Otvos, L., Jr., and Trojanowski, J.O. (1991) A68: a major subunit of paired helical filaments and derivatized forms of normal tau. Science 251, 675-678

9. Ishiguro, K., Ihara, Y., Uchida, T., and Imahori, K. (1988) A novel tubulin -dependent protein kinase forming a paired helical filament epitope on tau. J. Biochem. 104, 319-321

10. Webb, E.C. (1990) Enzyme nomenclature. Recommendations 1984. Supplement 3: corrections and additions. Eur. J. Biochem. 187,263-281

11. Ishiguro, K., Takamatsu, M., Tomizawa, K., Omori, A., Takahashi, M., Arioka, M., Uchida, T., and Imahori, K. (1992) Tau protein kinase I converts normal tau protein into A68-like component of paired helical filaments. J. Biol. Chem. 267, 10897-10901

12. Arioka, M., Tsukamoto, M., Ishiguro, K., Kato, R., Sato, K., Imahori, K., and Uchida, T. (1993) r protein kinase 11 is involved in the regulation of the normal phosphorylation state of r protein. J. Neurochem. 60, 461-468

13. Ishiguro, K., Shiratsuchi, A., Sato, S., Omori, A., Arioka, M., Kobayashi, S., Uchida, T., and Imahori, K. (1993) Glycogen synthase kinase 3(3 is identical to tau protein kinase I generating several epitopes of paired helical filaments. FEES Lett. 325, 167-172

14. Woodgett, J.R. (1990) Molecular cloning and expression of glycogen synthase kinase-3/Factor A. EMBO J. 9, 2431-2438

15. Siegfried, E., Chou, T.-B., and Perrimon, N. (1992) wingless Signaling acts through zeste-white 3, the Drosophila homolog of glycogen synthase kinase-3, to regulate engrailed and establish cell fate. Cell 71, 1167-1179

16. Ruel, L., Bourouis, M., Heitzler, P., Pantesco, V., and Simpson, P. (1993) Drosophila shaggy kinase and rat glycogen synthase kinase-3 have conserved activities and act downstream of Notch.

Nature 362, 557-56017. He, X., Saint-Jeannet, J.-P., Woodget, J.R., Varmus, H.E., and

Dawid, I.B. (1995) Glycogen synthase kinase-3 and dorsoventral patterning in Xenopus embryos. Nature 374, 617-62218. Kobayashi, S., Ishiguro, K., Omori, A., Takamatsu, M., Arioka,

M., Imahori, K., and Uchida, T. (1993) A cdc2-related kinase PSSALRE/cdk5 is homologous with the 30 kDa subunit of tau

protein kinase II, a proline directed protein kinase. associated with microtubule. FEBS Lett. 335, 171-175

19. Meyerson, M., Enders, G.H., Wu, C.-L., Su, L.-K., Gorka, C., Nelson, C., Harlow, E., and Tsai, L.-H. (1992) A family of human cdc2-related protein kinases. EMBO J. 11, 2909-2917

20. Ishiguro, K., Kobayashi, S., Omori, A., Takamatsu, M., Yonekura, S., Anzai, K., Imahori, K., and Uchida, T. (1994) Identification of the 23kDa subunit of tau protein kinase II as a putative

activator of cdk5 in bovine brain. FEBS Lett. 342, 203-20821. Uchida, T., Ishiguro, K., Ohnuma, J., Takamatsu, M., Yonekura,

S., and Imahori, K. (1994) Precursor of cdk5 activator, the 23

kDa subunit of tau protein kinase II: its sequence and develop

mental change in brain. FEBS Lett. 355, 35-40

22. Tsai, L.-H., Delalle, I., Caviness, V.S., Jr., Chae, T., and Harlow,

E. (1994) p35 is a neural-specific regulatory subunit of cyclin

dependent kinase 5. Nature 371, 419-423

23. Lew, J., Huang, Q.-Q., Qi, Z., Winkfein, R.J., Aebersold, R.,

Hunt, T., and Wang, J.H. (1994) A brain-specific activator of

cyclin -dependent kinase 5. Nature 371, 423-426

24. Uchida, T., Omori, A., Ishiguro, K., and Sato, K. (1993)

Sequence analysis of phosphopeptides and its application for the

determination of phosphorylated sites of proteins in Methods in

Protein Sequence Analysis (Imahori, K. and Sakiyama, F., eds.)

pp. 199-206, Plenum Press, New York and London

25. Ishiguro, K., Omori, A., Sato, K., Tomizawa, K., Imahori, K.,

and Uchida, T. (1991) A serine/threonine proline kinase activity

is included in the tau protein kinase fraction forming a paired

helical filament epitope. Neurosci. Lett. 128, 195-198

26. Ishiguro, K., Omori, A., Takamatsu, M., Sato, K., Arioka, M.,

Uchida, T., and Imahori, K. (1992) Phosphorylation sites on tau

by tau protein kinase I, a bovine derived kinase generating an

epitope of paired helical filaments. Neurosci. Lett. 148, 202-206

27. Ishiguro, K., Sato, K., Takamatsu, M., Park, J., Uchida, T., and

Imahori, K. (1995) Analysis of phosphorylation of tau with

antibodies specific for phosphorylation sites. Neurosci. Lett. 202,

81-84

28. Tsukamoto, M., Kato, R., Ishiguro, K., Uchida, T., and Sato, K.

(1991) Improved protective groups for phosphate of o-phospho

serine useful for the solid-phase peptide synthesis. Tetrahedron

Lett. 32, 7083-7086

29. Sato, K., Ishiguro, K., Omori, A., Kato, R., Kim, J.I., Uchida, T.,

Saruta, K., and Wakamiya, T. (1994) Solid phase synthesis of

phosphopeptides related to tau protein kinase I in Peptide

Chemistry 1993 (Okada, Y,, ed.) pp. 109-112, Protein Research

Foundation, Osaka

30. Goedert, M. (1993) Tau protein and the neurofibrillary pathology

of Alzheimer's disease. TINS 16, 460-465

31. Butner, K.A. and Kirschner, M.W. (1991) Tau protein binds to

microtubules through a flexible array of distributed weak sites. J.

Cell Biol. 115, 717-730

32. Kanai, Y., Chen, J., and Hirokawa, N. (1992) Microtuhule

bundling by tau protein in vivo: Analysis of functional domains.

EMBO J. 11, 3953-3961

33. Morishima-Kawashima, M., Hasegawa, M., Takio, K., Suzuki,

M., Yoshida, H., Titani, K., and Ihara, Y. (1995) Proline-directed

and non-proline-directed phosphorylation of PHF-tau. J. Biol.

Chem. 270, 823-829

34. Watanabe, A., Hasegawa, M., Suzuki, M., Takio, K., Morishima

Kawashima, M., Titani, K., Arai, T., Kosik, K.S., and Ihara, Y.

(1993) In vivo phosphorylation sites in fetal and adult rat tau. J.

Biol. Chem. 268, 25712-25717

35. Paudel, H.K., Lew, J., Ali, Z., and Wang, J,H. (1993) Brain

praline-directed protein kinase phosphorylates tau on sites that

are abnormally phosphorylated in tau associated with Alzhei

mer's paired helical filaments. J. Biol. Chem. 268, 23512-23518

36. Vulliet, R., Haloran, S.M., Braun, R.K., Smith, A.J., and Lee, G.

(1992) Proline-directed phosphorylation of human tau protein. J.

Biol. Chem. 267, 22570-22574

37. Yang, S.-D., Yu, J.-S., Shiah, S.-G., and Huang, J.-J. (1994)

Protein kinase FA/glycogen synthase kinase-3ƒ¿ after heparin

potentiation phosphorylates r on sites abnormally phosphorylat

ed in Alzheimer's disease brain. J. Neurochem. 63, 1416-1425

38. Drewes, G., Trinczk, B., Illenberger, S., Biernat, J., Schmitt

Ulmms, G., Meyer, H.E., Mandelkow, E.-M., and Mandelkow, E.

(1995) Microtubule -associated protein/microtubule affinity

regulating kinase (p110mark). J. Biol. Chem. 270, 7679-7688

39. Ino, H., Ishizuka, T., Chiba, T., and Tatibana, M. (1994)

Expression of CDK5 (PSSALRE kinase), a neural cdc2-related

protein kinase, in the mature and developing mouse central and

peripheral nervous system. Brain Res. 661, 196-206

40. Altman, J. (1972) Postnatal development of the cerebellar cortex

in the maturation of Purkinje cells and of the

molecular layer. J. Comp. Neurol. 145, 399-464

Vol. 121, No. 2, 1997

188 K . Imahori and T. Uchida

41. Bayer, S.A. (1980) Development of the hippocampal region in

the rat. I. Neurogenesis examined with 3H-thymidine autoradi

ography. II. Morphogenesis during embryonic and early postnatal

life. J. Comp. Neurol. 190, 87-114, 115-134

42. Ihara, Y. (1988) Massive somatodendritic sprouting of cortical

neurons in Alzheimer's disease. Brain Res . 459, 138-144

43. Nikolic, M., Dudek, H., Kwon, Y.T., Ramos, Y.F.M., and Tsai,

L.-H. (1996) The cdk5/p35 kinase is essential for neurite

outgrowth during neuronal differentiation . Genes Dev. 10, 816

- 825

44. Fujita, S.C., Takahashi, M., Hasegawa, J., and Imahori, K. (

1995) Immunohitochemistry of tau protein kinase I in the fetal

rat nervous system. Neurochem. Res. 20, 327

45. Takahashi, M., Tomizawa, K., Kato, R., Sato, K., Uchida, T.,

Fujita, S.C., and Imahori, K. (1994) Localization and develop

mental changes of r protein kinase I/Glycogen synthase kinase.

3,6 in rat brain. J. Neurochem. 63, 245-255

46. Takahashi, M., Tomizawa, K., Ishiguro, K ., Takamatsu, M.,

Fujita, S.C., and Imahori, K. (1995) Involvement of r protein

kinase I in paired helical filament-like phosphorylation of the

juvenile z in rat brain. J. Neurochem. 64, 1759-1768

47. Shiurba, R.A., Ishiguro, K., Takahashi, M., Sato, K., Spooner,

E.T., Mercken, M., Yoshida, R., Wheelock, T.R., Yanagawa, H., I

mahori, K., and Nixon, R.A. (1996) Immunocytochemistry of

tau phosphoserine 413 and tau protein kinase I in Alzheimer

pathology. Brain Res. 737, 119-132

48. Plyte, S.E., Hughes, K., Nikolakaki, E., Pulverer, B.J., and

Woodgett, J.R. (1992) Glycogen synthase kinase-3: functions in

oncogenesis and development. Biochim. Biophys . Acta 1114,

147-162

49. Yamaguchi, H., Ishiguro, K., Uchida, T., Takashima, A., Le

mere, C.A., and Imahori, K. (1996) Preferential labeling of

Alzheimer neurofibrillary tangles with antisera for protein kinase

(TPK)I/GSK-3ƒÀ and cdk5, a component of TPKII. Acta Neuro

pathol. 92, 232-241

50. Hardy, J.A. and Higgins, G.A. (1992) Alzheimer's disease: The

amyloid cascade hypothesis. Science 256, 184-185

51. Selkoe, D.J. (1989) Biochemistry of altered brain proteins in

Alzheimer's disease. Annu. Rev. Neurosci. 12, 463-490

52. Takashima, A., Noguchi, K., Sato, K., Hoshino, T., and Imahori,

K. (1993) Tau protein kinase I is essential for amyloid ƒÀ-protein

induced neurotoxicity. Proc. Natl. Acad. Sci. USA 90, 7789

- 7793

53. Yankner, B.A., Caceres, A., and Duffy, L.K. (1990) Nerve

growth factor potentiates the neurotoxicity of ƒÀ-amyloid. Proc.

Natl. Acad. Sci. USA 87, 9020-9023

54. Cotman, C.W., Pike, C.J., and Copani, A. (1992) ƒÀ-Amyloid

neurotoxicity-a discussion of in vitro findings. Neurobiol. Aging

13,587-590

55. Hoshi, M., Takashima, A., Noguchi, K., Murayama, M., Sato,

M., Kondo, S., Saitoh, Y., Ishiguro, K., Hoshino, T., andlmahori,

K. (1996) Regulation of mitochondrial pyruvate dehydrogenase

activity by tau protein kinase I/glycogen synthase kinase 3ƒÀ in

brain. Proc. Natl. Acad. Sci. USA 93, 2719-2723

56. Hoshi, M., Sato, M., Kondo, S., Takashima, A., Noguchi, K.,

Takahashi, M., Ishiguro, K., and Imahori, K. (1995) Different

localization of tau protein kinase I/glycogen synthase kinase-3ƒÀ

from glycogen synthase-3ƒ¿ in cerebrum mitochondria. J. Bio

chem. 118, 683-685

57. Hoyer, S., Oesterreich, K., and Wagner, O. (1988) Glucose

metabolism as the site of the primary abnormality in early-onset

dementia of Alzheimer type? J. Neurol. 235, 143-148

58. Parnetti, L., Gaiti, A., Brunetti, M., Avellini, L., Polidori, C.,

Cecchetti, R., Palumbo, B., and Senin,U. (1995) Increased CSF

pyruvate levels as a marker of impaired energy metabolism in

Alzheimer's disease. J. Am. Geriatr. Soc. 43, 316-318

59. Perry, E.K., Gibson, P.H., Blessed, G., Perry, R.H., and Tomlin

son, B.E. (1977) Neurotransmitter enzyme abnormalities in

senile dementia: Choline acetyltransferase and glutamic-acid

decarboxylase activities in necropsy brain tissue. J. Neurol. Sci.

34,247-266

60. Sorbi, S., Bird, E.D., and Blass, J.P. (1982) Decreased pyruvate

dehydrogenase complex activity in Huntington and Alzheimer

brain. Ann. Neurol. 13, 72-78

61. Yates, C.M., Butterworth, J., Tennant, M.C., and Gordon, A.

(1990) Enzyme activities in relation to pH and lactate in post

mortem brain in Alzheimer-type and other dementias. J.

Neurochem. 55, 1624-1630

62. Hoshi, M., Murayama, M., Yasutake, K., Hoshino, T., Ishiguro,

K., Takashima, A., and Imahori, K. (1997) Effect of amyloid,8

peptidei.42 on cholinergic neurons: Suppression of acetylcholine

synthesis precedes neurodegeneration. J. Biol. Chem . 272, 2038

- 2041

63. Whitehouse, P.J., Price, D.L., Struble, R.G., Clark, A.W., Coyle, J

.T., and DeLong, M.R. (1982) Alzheimer's disease and senile

dementia: loss of neurons in the basal forebrain. Science 215, 1237

-1239

J. Biochem.