Tau Protein Kinase I Converts Normal Tau Protein into A68-like ...

THE JOURNAL OF BIOLOOICAL CHEMISTRY Q 1992 by The American Society for Biochemistry and Molecular Biology, h e .

Vol. 267, No. Issue of May 26, pp. 10897-10901,1992 Printed in U. S. A.

Tau Protein Kinase I Converts Normal Tau Protein into A68-like Component of Paired Helical Filaments*

(Received for publication, October 22, 1991)

Koichi IshiguroS, Masako Takamatsu, Kayoko Tomizawa, Akira Omori, Miho Takahashi, Manabu Arioka, Tsuneko Uchida, and Kazutomo Imahori From the Mitsubishi Kosei Institute of Life Sciences, 11 Minamiooya, Machida-shi, Tokyo 194, Japan

From bovine brain microtubules we purified tau pro- tein kinase I (TPKI, M, 46,000 on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- PAGE)) and tau protein kinase I1 (TPKII) whose activ- ity was attributed to a 30-kDa protein on SDS-PAGE by affinity-labeling using an ATP analog. Both kinases were activated by tubulin. TPKII, but not TPKI, phos- phorylated tau fragment peptides previously used for detection of a Ser/ThrPro kinase activity. Therefore, TPKII was considered to be the SerflhrPro kinase. TPKI was more effective than TPKII for producing the decrease of tau-1 immunoreactivity and mobility shift of tau on SDS-PAGE. Moreover, TPKI, but not TPKII nor other well-known protein kinases, gener- ated an epitope present on paired helical filaments. Them findings suggested that tau phosphorylated by TPKI resembled A-88, a component of paired helical filaments.

Paired helical filaments (PHF)’ are abnormal filaments, accumulated in Alzheimer’s disease (AD) brain. One compo- nent of PHF is tau protein, one of the neuron specific micro- tubule-associated proteins (MAPS) (1, 2). Using antibodies, several groups found that the tau in PHF was phosphorylated in a manner not recognized in a normal adult brain. Grundke- Iqbal et al. (3) reported that a monoclonal antibody to tau, tau-1, reacted with PHF, and more strongly reacted with dephosphorylated PHF. The tau-1 site was identified to be in the central part of tau (4). Ihara et d. ( 5 ) showed that anti- PHF polyclonal antibodies contained an antibody to react with ptau, but not with dephosphorylated tau. The antibody recognized the C-terminal part of tau, an extraordinarily insoluble part in the PHF core (2). In both cases the ptau had lower electrophoretic mobility on SDS-PAGE than non-phos- phorylated tau.

Several groups reported that calmodulin-dependent protein ~

* This work was supported by Grant-in-Aid for Scientific Research on Priority Areas 02240105 from the Ministry of Education, Science and Culture, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “oduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed Mitsubishi Kasei Institute of Life Sciences, 11 Minamiooya, Machida-shi, Tokyo 194, Japan.

The abbreviations used are: PHF, paired helical filaments; AD, Alzheimer’s disease; ptau, phosphorylated tau; PAGE, polyacrylamide gel electrophoresis; CMPK, Ca’+/calmcdulin-dependent protein kinase; FSBA, 5‘-p-fluorosulfonylbenzoyl adenosine; MES, 2-(N- morpho1ino)ethanesulfonic acid; MAP, microtubule-associated protein; SDS, sodium dodecyl sulfate; EGTA, [ethylene- bis(oxyethylenenitrilo)]tetraacetic acid.

kinase (CMPK) phosphorylated tau to induce the mobility shift and discussed the correlation between the kinase and the ptau in PHF (6,7). However, they did not prove that tau phosphorylated by CMPK had a phosphorylated PHF epi- tope. On the other hand, we reported a kinase activity phos- phorylating tau, forming a PHF epitope, and inducing a large mobility shift (8). Our kinase did not depend on the second messengers. The phosphorylation was stimulated by tubulin under conditions of microtubule formation. The kinase bound to tau. Considering these properties closely linked to tau, we designated this protein kinase as tau protein kinase (TPK) (EC 2.7.1.135).

We also found that the kinase fraction carried a protein kinase activity recognizing a serinelthreonine proIine ae- quence in chemically synthetic peptides (9). The Ser/ThrPro kinase phosphorylated tau at both the tau-1 site and the extraordinarily insoluble part in the C-terminal region. Other groups reported that phosphorylation sites in PHF had SerPro sequences (10-13). It was possible that tau phos- phorylated by the kinase was incorporated into PHF. But it remained unclear whether or not TPK showed the Ser/ ThrPro kinase activity because the purification was unsuc- cessful.

We later improved the purification procedures, succeeded in the purification of TPK, and consequently found that the TPK was separated from the Ser/ThrPro kinase activity. Now we propose to designate TPK as TPKI and the Ser/ ThrPro kinase as TPKII. In this report, we present the purification and characterization of TPKI and TPKII and also provide evidence that tau phosphorylated by TPKI re- sembles A68, a component of PHF (13).

EXPERIMENTAL PROCEDURES

Materials-Adult bovine brains were obtained from the Nakashi- betsu preparation center, Mitsubishi Kasei Co. Tau and tubulin were prepared from the bovine brain extracts (8). CMPKII and CAMP- dependent protein kinase were kind gifts from Dr. E. Miyamoto at the Kumamoto University. Protein kinase C was a kind gift from Dr. K. Chida at the University of Tokyo. Anti-ptau antibody was a kind gift from Dr. Y. Ihara at the University of Tokyo. Monoclonal antibody tau-1 was purchased from Boehringer Mannheim Yaman- ouchi. Polyclonal anti-tau antibody developed in rabbit was purchased from Sigma. Histones and caseins were also purchased from Sigma. Tau peptides K1, K2, and K3, were synthesized chemically as de- scribed in the previous paper (9). In accordance with the amino acid numbering of human tau (14), the sequences of K1, K2, and K3 were found in the regions of 168-182, 133-166, and 307-349, respectively.

Kinase Detection Assays-Kinase activity was measured by detec- tion of radioactive phosphate incorporation into tau from [Y-~~P]ATP (Amersham) as described previously (8). Kinase activity staining on SDS-PAGE was done by the method of Kameshita and Fujisawa (15) using tau at a protein concentration of 0.1 mg/ml as substrate. The kinase was also detected by affinity labeling with the ATP analog, [8-“C]FSBA (Du Pont-New England Nuclear) (16), separated by SDS-PAGE, and detected by autoradiography.

10897

10898 Tau Protein Kinase I Converts Tau into a PHF Component Purification of TPKZ and TPKZZ-Through detecting the phos-

phorylation activity causing the mobility shift of tau and forming a PHF epitope on tau, we purified TPKI from bovine brain microtubule proteins. TPKII was purified by detecting an activity phosphorylating the chemically synthetic peptides K1, K2, and K3. During the puri- fication after step 1, all buffer solutions contained 0.02% Tween 20, 10% glycerol, 5 mM 2-mercaptoethanol, 0.1 mM phenylmethylsulfonyl fluoride, 1 pg/ml leupeptin, 1 pg/ml pepstatin A, and 1 pg/ml antipain. When a detergent such as Tween 20 was omitted during the purifi- cation, the yield of kinases was very low.

Step 1: preparation of microtubule proteins. Brain extract was prepared from 15 bovine brains (about 6 kg), and microtubule proteins were prepared from the extract by temperature-dependent polymeri- zation and depolymerization of microtubule by a widely used proce- dure as modified by Ihara et al. (17). In this method, the polymerized microtubules after the second cycle were precipitated under a 25% glycerol cushion in order to separate them from soluble proteins overlaying the cushion. Protein yield was 3000 mg.

Step 2: ammonium sulfate fractionation. The microtubule proteins were diluted with buffer A (100 mM MES-NaOH, pH 6.5, 0.5 mM magnesium acetate, and 1 mM EGTA) to a protein concentration of 10 mg/ml, followed by ammonium sulfate fractionation (33-50%).

Step 3: phosphocellulose column chromatography. The resulting precipitate was dialyzed against PC buffer (20 mM MES-NaOH, pH 6.8,0.5 mM magnesium acetate, and 1 mM EGTA) and mixed with a P11 (Whatman) gel slurry previously equilibrated with PC buffer. Afterward, the gel was washed with PC buffer and 0.1 M NaCl until the ODm of the effluent became less than 0.01, and the gel was packed into a column (26 mm inner diameter X 180 mm). The adsorbed protein was eluted with a 800-ml linear gradient of 0.1-0.8 M NaCl in PC buffer. The kinases TPKI and TPKII were eluted a t 0.2-0.4 M NaCl together with tau.

Step 4 gel filtration. The active fractions (170 ml) were collected, concentrated by ultrafiltration with a PM-10 membrane (Amicon) to total volume of less than 10 ml, and applied to a G3000SW gel filtration column (21 mm inner diameter X 600 mm, Tosoh) previ- ously equilibrated with buffer A and 0.3 M NaCI. The kinases were eluted at M, 50,000 together with a main contaminating protein, p25, and separated from tau.

Step 5: hydroxyapatite column chromarography. The kinase frac- tions were applied to a hydroxyapatite column (7.5 mm inner diameter X 100 mm, Tonen) previously equilibrated with 12 mM sodium phosphate buffer, pH 6.8. The adsorbed proteins were eluted with a 10-ml linear gradient of 12-400 mM phosphate buffer. TPKII and TPKI activities were eluted at 0.13 and 0.15 M phosphate buffer, respectively.

Step 6 S-Sepharose chromatography. The TPKI fraction was dialyzed against PC buffer and applied to S-Sepharose column (5 mm inner diameter X 20 mm, Pharmacia LKB Biotechnologies Inc.) previously equilibrated with the PC buffer. The column was developed with a 7.5-ml pH gradient of 20 mM MES-NaOH, pH 6.8, to 20 mM HEPES-NaOH, pH 8.2, containing 0.5 mM magnesium acetate and 1 mM EGTA, and then with a 5-ml linear salt gradient of 0-200 mM NaCl in the buffer containing 20 mM HEPES-NaOH, pH 8.2,0.5 mM magnesium acetate, and 1 mM EGTA. TPKI was eluted at 150 mM NaCI, while the remaining TPKII was removed at 50 mM. When the elution was done without the pH gradient, the two kinases were not well separated.

Step 7: heparin column chromatography. The TPKI fraction was dialyzed against 20 mM sodium phosphate buffer, pH 7.6, and 1 mM EGTA, applied to TSKgel AF-heparin Toyopearl 650 column (4 mm inner diameter X 8 mm, Tosoh) and eluted with a 4-ml linear salt gradient from 0 to 200 mM NaCI. TPKI was eluted a t 80 mM NaC1.

Preparation of Substrates for TPKZ and TPKIZ-Acid-treated brain extract was prepared by the following: brain extract a t step 1 of the purification was precipitated in 5% trichloroacetic acid by centrifugation at 10,000 X g for 5 min, dissolved in buffer A, and dialyzed against the same buffer. This treatment inactivated endog- enous protein kinases and phosphatases.

MAP2 was prepared by the following: microtubule proteins were heated in 0.8 M NaCl and 0.5 M 2-mercaptoethanol a t 95 “C for 10 min, then the supernatant obtained by centrifugation at 10,000 X g for 10 min was applied to a gel filtration column (G3000SW) equili- brated with buffer A. Purified MAP2 was eluted at the void volume.

Estimation of M, of Proteins-Unless otherwise mentioned, pre- stained protein molecular weight standard (Bethesda Research Lab- oratories) was used for estimation of M, on SDS-PAGE. The apparent M, values of the prestained markers were determined by comparison

to molecular weight markers Daiichi I11 (Daiichi Pure Chemicals). The M, values used in this paper were as follows: lysozyme, 15,000; @-lactoglobulin, 19,000; carbonic anhydrase, 29,000; ovalbumin, 50,000; bovine serum albumin, 70,000; phosphorylase b, 110,000; myosin (H-chain), 200,000.

Others-SDS-PAGE was done by Laemmli’s method (18). Immu- noblotting was done according to the standard method of Vectastain ABC system (Vector Laboratories) using nitrocellulose membrane and 4-chloro-1-naphthol as the chromogenic substrate. Silver staining of protein on SDS-polyacrylamide gel was done with 2D-Silver Stain. I1 (Daiichi Pure Chemicals). For autoradiography, the gels were dried in vacuum, exposed to an Imaging Plate, and analyzed with a Bio- Image Analyzer BAS200 (Fuji Photo Film). Isoelectric points of protein were determined by isoelectric focusing column chromatog- raphy at pH 3.5-10.0 with Ampholine (LKB). Protein concentrations were determined with a Bio-Rad protein assay kit (Bio-Rad) (19).

RESULTS

Purification of TPKI and TPKII-Details of the purifica- tion were described under “Experimental Procedures.’’ Using the kinase fraction at step 4, we reported previously a protein kinase activity forming a PHF epitope (8) and a Ser/ThrPro kinase activity (9). This fraction contained primarily a 25- kDa protein (p25). Further purification revealed that the two kinase activities and p25 were separable.

At steps 5 and 6, the two kinase activities were separated. One kinase activity could generate a PHF epitope on tau but not phosphorylate the synthetic tau fragments K1, K2, and K3, which were used for detection of a Ser/ThrPro kinase activity (9). The other activity could phosphorylate these tau fragments but not form a PHF epitope on tau (Fig. 1). The former contained our desired TPK, which we call TPKI. The latter contained the Ser/ThrPro kinase, which was named

A fr . no 29 30 31 32 33 34 35 36 37 38

200K- llOK- 70K-

50K-

29K-

19K-

Ironl-

B pmol/mln 3 ’llm VI

c E O

C Ir . no 29 30 31 32 33 34 35 36 37 38

””*-.? -. 70K-

8

50K.

TPKII TPKI

FIG. 1. Separation of TPKI from TPKII by S-Sepharose column chromatography. The hydroxyapatite fraction containing TPKI activity was chromatographed on a S-Sepharose column as described under “Experimental Procedures.” Each 0.5 ml of eluate was collected. NaCl concentrations at fractions 32 and 36 were 50 and 150 mM, respectively. A, 5-pl aliquots were subjected to 13% SDS-PAGE and the gel was stained with silver staining. The M, of prestained markers are indicated at the .!eft. TPKI (45 kDa), TPKII (30 kDa), and p25 are indicated by the closed arrowhead, the open arrowhead, and the arrow at the right of the figure, respectively. B, the kinase activities were assayed with tau (200 pg/ml) (closed circles) and a synthetic peptide K3 (40 pg/ml) (open circles). C, purified tau protein was phosphorylated by the kinase activities in each fraction for 1 h and subjected to 9% SDS-PAGE. The separated proteins were electrotransferred onto nitrocellulose membrane and immunostained with anti-ptau antibody.

Tau Protein Kinase I Converts Tau into a PHF Component 10899

TPKII. At step 7 TPKI was separated from the contaminating p25 and other proteins. p25 is a novel protein which is phosphorylated by TPKII and specifically located in brain. Details about this protein were described in another report of ours (20).

This purification procedure produced about 10 pg of TPKI and 30 pg of TPKII from 15 bovine brains (about 6 kg) (Table I).

TPKI-The purified TPKI contained mostly a 45-kD.3 protein on SDS-PAGE (Fig. 2A) that proved to be a protein kinase by activity staining gel electrophoresis (15) using tau as a substrate (Fig. 2B). The specific activity of the kinase was 1000 nmol Pi/min/mg under optimum conditions (0 mM NaC1, 0.9 mM magnesium acetate, 0.4 mM ATP, 100 mM MES-NaOH, pH 6.4). This value was comparable to that of calmodulin-dependent protein kinase I1 (CMPKII) (21). These data suggested that the purification of TPKI was essentially complete and that the 45-kDa protein found on SDS-PAGE was the desired TPKI. Its PI was 9.4.

Phosphorylation of tau by TPKI caused a larger decrease in electrophoretic mobility and generated a stronger ptau epitope in PHF than any other protein kinases, such as TPKII, CPMKII, cyclic AMP-dependent protein kinase (Fig. 3), or protein kinase C (data not shown). The M, of the largest tau phosphorylated by TPKI was 68,000 on SDS-PAGE. Fig. 4 shows that phosphorylation of tau by either TPKI or TPKII caused a decrease in tau-1 immunoreactivity. Moreover, the largest tau phosphorylated with TPKI had no tau-1 epitope,

TABLE I Purification of TPKI and TPKII

The activities were measured by phosphorylation of tau under the standard condition reported previously (8), not the optimum condi- tion. Yields were estimated assuming that the activities in the frac- tions from gel filtration represent the total activities of TPKI and TPKII because the fractions prior to gel filtration still included many kinases capable of phosphorylating tau, in addition to TPKI and TPKII.

Step Fraction

1 Microtubule 2 (NHI)PSOI 3 Phosphocellulose 4 Gel filtration 5 Hydroxyapatite 6 S-Sepharose

TPKI TPKII

7 Heparin TPKI

Protein Activity Specific Yield

mg pmollmin pmolfminfmg %

activity

3,000 48,000 16 2,500 45,000 18

30 7,300 240 1 2,800 2,800 100 0.2 1,700 8,500 60

0.05 840 17,000 30 0.03 670 22.000 24

0.01 560 56,000 20

A B C m l 2

top- " - . - top- 20011- 11Ok-

70X-

top- 2OOk- 1101;-

70k-

97U- -

e e k - t - 4 % 42k-

3 0 , - i - - -3OU

SOL- -4%

29k- 2 9 ~ - I. -3OU -23);

2Ok-

14k-

1,011

IOU-

1Sk- 19k-

tront- Irm1-

FIG. 2. SDS-PAGE of TPKI and TPKII. A , TPKI (50 ng, lane 1 ) and TPKII (50 ng, lane 2 ) were separated by 13% SDS-PAGE and stained by silver staining. Lune m indicated M, markers (Daiichi 111). B, tau phosphorylating activity in TPKI fraction (50 ng) was detected by kinase activity staining. C, TPKII (50 ng) was affinity-labeled with [S-"C]FSBA. The autoradiogram is shown.

A m 1 2 3 4 5 6

B 1 2 3 4 5 6

70k-

50k-

FIG. 3. Tau phosphorylated by various protein kinases. Tau was phosphorylated by the TPK fraction purified up to step 4 (lane 2 ) , TPKII (lane 3 ) , TPKI (lane 4 ) , CMPKII (lane 5), and CAMP- dependent protein kinase (lane 6). One pg of non-phosphorylated tau (lane 1 ) and phosphorylated tau proteins were electrophoresed on a 9% SDS-PAGE. A, proteins were detected by silver staining. B, proteins were immunostained with a polyclonal anti-ptau antibody (x100 dilution). The M, markers are indicated in lane m.

A 1 2 3 4 5 -70k

-50k

1 2 3 4 5

FIG. 4. Disappearance of tau- 1 epitope accompanying phos- phorylation of tau by TPKI. Non-phosphorylated tau (lane I), tau phosphorylated by TPKI (lane 2) , tau phosphorylated by TPKII (lane 3 ) , TPKI alone (lane 4 ) , and TPKII alone (lane 5) were electrophoresed on a 9% SDS-PAGE, transferred to nitrocellulose membrane, and immunoblotted with polyclonal antibody anti-tau (A, x1000 dilution) or monoclonal antibody tau-1 (B , X2000 dilution). One pg of tau and phosphorylated tau were used on each immunoblot.

indicating that TPKI was more effective for the disappearance of the tau-1 epitope than TPKII. The disappearance of the tau-1 epitope, together with the large mobility shift and ptau epitope formation by TPKI, suggested that the tau phos- phorylated by TPKI was indeed identical to the one incor- porated into PHF.

This kinase was independent of well-known second messen- gers (data not shown). The phosphorylation of tau by TPKI was stimulated by the addition of tubulin (Fig. 5A) as we reported previously (8). As shown in lane 3 of Fig. 6A and C, the tubulin was almost pure. The stimulation by tubulin was also observed, when we used tubulin dialyzed against buffer A, to remove any low molecular weight substance(s). Tubulin without polymerizing activity did not stimulate the phos- phorylation. The critical concentration of tubulin for micro- tubule formation was 0.1 mg/ml.

The kinase did not phosphorylate histones and @-casein (Table 11). Among the many proteins in acid-treated brain extract, the kinase phosphorylated specifically MAP2 and two proteins (Mr 68,000 and 64,000) thought to be tau species (Fig. 6). These tau-like proteins had ptau epitopes (data not shown). It is not yet known why all tau species were not clearly phosphorylated in the brain homogenate.

Although TPKI could not phosphorylate K1, K2, and K3 peptides, the kinase phosphorylated the K1, K2, and K3 regions in tau (data not shown). The K2 region overlapped with the tau-1 site, and the K3 region contained the ptau

10900 Tau Protein Kinase I Converts Tau into a PHF Component

A

*u 0 0 0.5 1.0 1.5 2.0

Tubulin Concentratlon (rng I rnl)

B prnol I rnin I rnl

120

- 2 100 - 5

2

= 20

- 80

0 60 v)

40

0 0.5 1.0 1.5 2.0 (mg I rnl)

Tubulin Concentration FIG. 5. Stimulation of phosphorylation of tau by TPKI ( A )

and TPKII ( B ) in the presence of tubulin. Tau (0.2 mg/ml) was phosphorylated for 1 h in the presence (closed circles) or absence (open circles) of the kinases (1 pg/ml).

A B 1 2 3 45

i8k- 8 i8k- Iront- iront-

C D 1 2 3 4 5 1 2 3 4 5

200k- Z O O k ” 7 7 MOL- .. 70k-

50k-

29k-

k- t-

I9k-8 9 i s iront- Iron

FIG. 6. Substrate specificity of TPKI and TPKII. Phos- phorylation was carried out in the presence of no substrate (lane I), acid-treated brain extract (100 pg, lane 2), tubulin (3 pg, lane 3), MAP2 (1 pg, lane 4 ) and tau (2 pg, lane 5). These substrates were phosphorylated with [y3’P]ATP by TPKI ( A and B ) and TPKII (C and D), and electrophoresed on a 13% SDS-PAGE. The gels were stained with Coomassie Brilliant Blue R-250 (A and C) and autora- diographed ( B and D).

epitope in PHF (2). Work is in progress to determine the sites phosphorylated by TPKI.

TPKII-TPKII also phosphorylated tau to induce a mobil- ity shift but not to form the ptau epitope (Fig. 3). The phosphorylation of tau by the kinase reduced tk immuno- reactivity of the tau-1 epitope (Fig. 4), which was reasonable because TPKII phosphorylated the tau-1 site as mentioned

TABLE I1 Substrate specificity toward histones and caseins

0.1 mg/ml of substrates were used. Kinase activities are presented compared with that of tau as standard.

TPKI TPKII

Histone H1 0 734 Histone H2a 0 48 Histone H2b 0 34 Histone H3 0 9 a-Casein 6 0 j3-Casein 0 3 MAP2 42 105 Tau 100 100

before. The enzyme fraction contained two proteins, whose M, values were 30,000 and 23,000 on SDS-PAGE (Fig. 2A). The 30-kDa protein bound to an ATP analog (FSBA) (Fig. 2C), indicating that the 30-kDa protein contained the cata- lytic site of TPKII. It remains unknown whether the 30-kDa protein is a catalytic subunit and whether the 23-kDa protein is another subunit of TPKII. The PI of the 30-kDa protein was 8.5. The kinase was also activated by tubulin (Fig. 5B). The kinase phosphorylated remarkably histone H1, MAP2, and other histones as well as tau, but not caseins (Fig. 6, Table 11). These substrates were known to contain several Ser/ThrPro sequences (22,23).

DISCUSSION

We have purified a protein kinase which phosphorylates tau, makes the tau-1 epitope disappear, and forms the epitope recognized by an anti-ptau polyclonal antibody against PHF. We have named this protein kinase TPKI.

Other groups reported the possibility that tau in PHF was phosphorylated by casein kinase I1 (24) or CMPK (6, 7). TPKI is not thought to be one of these kinases. Tau phos- phorylated by TPKI produced a larger mobility shift and stronger immunoreactivity with anti-ptau than that which was phosphorylated by CMPKII (Fig. 3). TPKI was inde- pendent of calcium and calmodulin. Its partial sequence ob- tained so far also supports the conclusion that TPKI is different from casein kinase I1 and CMPKII?

Steiner et al. (7) reported that phosphorylation by CMPK in the C-terminal tail of tau induced the mobility shift, which we also observed using CMPKII as shown in Fig. 3. TPKI also phosphorylated the C-terminal region of whole molecule of tau, which is just corresponding to synthetic peptide K3. Phosphorylation of this region may be responsible for the shift. Determination of the phosphorylation site is now under investigation. In our preliminary study, the site was different from that reported by them and from the consensus site recognized by CMPKII.

Grundke-Iqbal et al. (3) reported that tau-1 site was phos- phorylated in PHF because they detected a decrease in im- munoreactivity with the monoclonal antibody, tau-1. The tau- 1 site was located at the middle of tau (4). Lee et al. (13) showed that this immunoreactivity disappeared completely in phosphorylated tau (A68), which is a component of PHF. Tau phosphorylated by TPKI resembles A68 in its electrophoretic mobility (Fig. 3) and because of the disappearance of immu- noreactivity with tau-1 (Fig. 4). Moreover, the ptau reacted with anti-ptau antibody prepared from anti-PHF antisera (5), which recognized the C-terminal region of tau (2). These data indicate that tau in PHF is phosphorylated by TPKI.

Flament et al. (25) showed that there were two abnormally phosphorylated tau proteins (Mr 69,000 and 64,000) in AD

A. Omori and K. Ishiguro, unpublished observation.

Tau Protein Kinase I Converts Tau into a PHF Component 10901

brain. TPKI phosphorylated two proteins (68- and 64-kDa protein) in brain homogenate (Fig. 6). It is quite possible that these two proteins are tau isoforms, identical to those reported by them. However, four phosphorylated bands were identified when purified tau was phosphorylated. This discrepancy be- tween crude and purified fractions may be explained by sug- gesting that contaminating proteins in the crude fraction disturbed the electrophoretic mobility of ptau or that contam- inating proteins prevent the phosphorylation of a species of tau. It is not clear which explanation is more appropriate.

The phosphorylation of tau might indicate a physiological change that can be correlated with AD and may reveal the etiology of AD. However, the biological significance of the phosphorylation still remains unclear. Based on the fact that this phosphorylation was stimulated by tubulin, we discussed previously (8) that the phosphorylation might be related with abnormal somatodendritic sprouting, which occurs in AD (26). Whether this hypothesis is viable or not will soon be clear as the kinase is now identified.

Previously, we reported that a Ser/ThrPro kinase activity was present in the tau protein kinase fraction at purification step 4 (9). This kinase, like TPKI, phosphorylated the tau-1 site and the C-terminal region on tau. After further purifica- tion, we found that the latter property was attributed not to TPKI but to TPKII. Tau phosphorylated by TPKII neither reacted with anti-ptau antibody nor induced an extreme mo- bility shift (Fig. 3). These do not support the suggestion that this kinase is involved in phosphorylation found in AD brain. However, phosphorylation sites are identified in the vicinity of those found in AD. As shown in Fig. 4, the immunoreactiv- ity to tau-1 decreased after phosphorylation with TPKII as observed in the phosphorylation by TPKI. The phosphoryla- tion by TPKII may be related to multiple phosphorylations at these sites.

The purified TPKII fraction contained a 30- and a 23-kDa protein. The former appears to be a kinase because it is an ATP-binding protein. A partial amino acid sequence of the 30-kDa protein resembles the consensus sequence of domain I of serine/threonine protein kinase (27) but was different from the sequence of the tentative domain I of TPKI? There- fore, TPKII is a different protein from TPKI, although both kinases have several similar properties. It is not yet known whether the 23-kDa protein is a subunit of the kinase or not. Should it be another subunit, it would make sense that TPKII eluted at the point consistent with a M, of 50,000 in gel filtration (step 4).

Acknowledgment-We thank Dr. Y. Ihara (University of Tokyo) for helpful discussions.

REFERENCES 1. Wischik, C. M., Novak, M., Thogersen, H. C., Edwards, P. C.,

Runswick, M. J., Jakes, R., Walker, J. E., Milstein, C., Roth, M., and Klug, A. (1988) Proc. Natl. Acad. Sci. U. S. A. 8 6 ,

2. Kondo, J., Honda, T., Mori, H., Hamada, Y., Miura, R., Ogawara, M., and Ihara, Y. (1988) Neuron 1,827-834

3. Grundke-Iqbal, I., Iqbal, K., Tung, Y.-C., Quinlan, M., Wis- niewski, H. M., and Binder, L. I. (1986) Proc. Natl. Aead. Sei.

4. Kosik, K. S., Orecchio, L. D., Binder, L., Trojanowski, J. Q., Lee,

5. Ihara. Y.. Nukina. N.. Miura. R.. and Ogawara. M. 11986) J.

4506-4510

U. S. A. 83,4913-4917

V. M.-Y., and Lee, G. (1988) Neuron 1,817-823

~ & h e A . (TO&) s6,1807-i81O I . .

6. Baudier. J.. and Cole. R. D. 11987) J. Biol. Chem. 262. 17577- 17583' '

7. Steiner, B., Mandelkow, E."., Biernat, J., Gustke, N., Meyer, H. E., Schmidt, B., Mieskes, G., Soling, H. D., Drechsel, D., Kirschner, M. W., Goedert, M., and Mandelkow, E. (1990)

8. Ishiguro, K., Ihara, Y., Uchida, T., and Imahori, K. (1988) J.

9. Ishiguro, K., Omori, A., Sato, K., Tomizawa, K., Imahori, K., and Uchida, T. (1991) Neurosci. Lett. 128,195-198

10. Coleman, M. P., and Anderton, B. H. (1990) J. Neurochem. 6 4 , 1548-1555

11. Iqbal, K., Grundke-Iqbal, I., Smith, A. J., George, L., Tung, Y.-

5650 C., and Zaidi, T. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,5646-

12. Lee, V. M.-Y., Otvos, L., Jr., Carden, M. J., Hollosi, M., Dietzschold, B., and Lazzarini, R. A. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,1998-2002

13. Lee, V. M.-Y., Balin, B. J., Otvos, L., Jr., and Trojanowski, Q. (1991) Science 2 5 1 , 675-678

14. Goedert, M., Wischik, C. M., Crowther, R. A., Walker, J. E., and Klug, A. (1988) Proc. Natl. Acad. Sci. U. S. A. 86,4051-4055

15. Kameshita, I., and Fujisawa, H. (1989) Anal. Biochem. 183,139- 143

16. Woodford, T. A., and Pardee, A. B. (1986) J. Biol. Chem. 2 6 1 ,

17. Ihara, Y., Fujii, T., Arai, T., Tanaka, R., and Kajiro, Y. (1979) J.

18. Laemmli, U. K. (1970) Nature 227,680-685 19. Bradford, M. M. (1976) Anal. Biochem. 7 2 , 248-254 20. Takahashi, M., Tomizawa, K., Ishipuro, K., Sato, K., Omori, A.,

. ,

EMBO J. 9,3539-3544

Biochem. (Tokyo) 104,319-321

4669-4676

Bwchem. (Tokyo) 86,587-590

Sato, S., Shiratsuchi, A., Uchida, T., and Imahori, K. (1991) FEBS Lett. 289.37-43

21. Yamamoto, H., Fukunaga, K., Tanaka, E., and Miyamoto, E.

22. Lewis, S. A., Wang, D., and Cowan, N. J. (1988) Science 242 ,

23. Coles, L. S., Robins, A. J., Madley, L. K., and Wells, J. R. E.

24. Iimoto, D. S., Masliah, E. M., DeTeresa, R., Terry, R. D., and

25. Flament, S., Delacourte, A., Hemon, B., and Defossez, A. (1989)

26. Ihara, Y. (1988) Brain Res. 4 6 9 , 138-144 27. Hanks, S. K., Quinn, A. M., and Hunter, T. (1988) Science 2 4 1 ,

(1983) J. Neurochem. 41, 1119-1125

936-939

(1987) J. Biol. Chem. 262,9656-9663

Saitoh, T. (1989) Brain Res. 507,273-280

J. Neurol. Sci. 9 2 , 133-141

42-51