- Email: [email protected]
Characterization of tau phosphorylation in glycogen synthase kinase-3β and cyclin dependent kinase-5 activator (p23) transfected cells
Biochimica et Biophysica Acta 1380 Ž1998. 177–182
Characterization of tau phosphorylation in glycogen synthase kinase-3b and cyclin dependent kinase-5 activator žp23 / transfected cells Gilles Michel, Marc Mercken, Miyuki Murayama, Kaori Noguchi, Koichi Ishiguro, Kazutomo Imahori, Akihiko Takashima ) Mitsubishi Kasei Institute of Life Sciences, 11 Minamiooya, Machida-shi, Tokyo 194, Japan Received 30 September 1997; accepted 21 October 1997
Abstract One of the histopathological markers in Alzheimer’s disease is the accumulation of hyperphosphorylated tau in neurons called neurofibrillary tangles ŽNFT. composing paired helical filaments ŽPHF.. Combined tau protein kinase II ŽTPK II., which consists of CDK5 and its activator Žp23., and glycogen synthase kinase-3b ŽGSK-3b . phosphorylate tau to the PHF-form in vitro. To investigate tau phosphorylation by these kinases in intact cells, the phosphorylation sites were examined in detail using well-characterized phosphorylation-dependent anti-tau antibodies after overexpressing the kinases in COS-7 cells with a human tau isoform. The overexpression of tau in COS-7 cells showed extensive phosphorylation at Ser-202 and Ser-404. The p23 overexpression induced a mobility shift of tau, but most of the phosphorylation sites overlapped the endogenous phosphorylation sites. GSK-3b transfection showed the phosphorylation at Ser-199, Thr-231, Ser-396, and Ser-413. Triplicated transfection resulted in phosphorylation of tau at 8 observed sites ŽSer-199, Ser-202, Thr-205, Thr-231, Ser-235, Ser-396, Ser-404, and Ser-413.. q 1998 Elsevier Science B.V. Keywords: GSK-3b; p23; CDK5; Tau; Phosphorylation site; Phosphorylation-dependent anti-tau antibody
1. Introduction Alzheimer’s diseaseŽ AD. is histochemically characterized by the accumulation of amyloid b protein ŽAb . in senile plaques w1x and paired helical filamentsŽPHF. composed of hyperphosphorylated tau in neurofibrillary tangles ŽNFT. w1–5x. Tau is a group of microtubule-associated proteins expressed predominantly in axons w6,7x that promote microtubule assembly in vitro w8x and that stabilize microtubules in vivo )
Corresponding author. Present address: Lab. for Alzheimer’s Disease, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wakoshi, Saitama 350-01, Japan. Fax: q81-48-462-4796; E-mail: [email protected]
w9x. Hyperphosphorylation of tau abrogates its ability to form microtubules, leading to the instability of cytoskeleton w10–12x. Since this instability may cause the degeneration of neurons in AD brain, the mechanism of hyperphosphorylation of tau is important for understanding the pathogenesis of AD. In vitro studies have shown that tau can be a substrate for various protein kinasesw11–21x. We previously identified tau protein kinase I ŽTPK I. which is identical to glycogen synthase kinase-3b ŽGSK-3b . w17xand tau protein kinase II ŽTPK II., which consists of CDK5 and its activator Žp23. w22,23x. GSK-3b required prior phosphorylation by CDK5rp23 w24x for the complete phosphorylation of tau in vitro. Combined, these kinases hyperphosphorylated tau which carries the
0304-4165r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 3 0 4 - 4 1 6 5 Ž 9 7 . 0 0 1 3 9 - 6
178
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
epitope of PHF-tau. To examine tau phosphorylation in intact cells, we transfected these kinases into COS-7 cells, and the phosphorylation sites were examined in detail by using phosphorylation-dependent anti-tau antibodies.
2. Materials and methods 2.1. Antibodies Polyclonal antisera for each phosphopeptide were raised in rabbits by using corresponding synthetic phosphopeptides conjugated to keyhole limpet hemocyanin. In the previous study, we confirmed that produced antibodies reacted specifically for each corresponding phosphorylated site w25x. The following antibodies were used for analyzing the phosphorylation sites of tau: PS199 Ž1:200. directed against phosphorylated Ser-199 of tau Žin the numbering of htau 40 w26x., PS202 Ž 1:200. directed against phosphorylated Ser-202 of tau, PT205 Ž1:200. directed against phosphorylated Thr-205 of tau, PT231 Ž 1:200. directed against phosphorylated Thr-231 of tau, PS235 Ž1:200. directed against phosphorylated Ser-235 of tau, PS396 Ž1:200. directed against phosphorylated Ser-396 of tau, PS404 Ž1:200. directed against phosphorylated Ser-404 of tau, and PS413 Ž 1:30. directed against phosphorylated Ser-413 of tau. JM Ž1:5000. was used for detecting both phosphorylated and nonphosphorylated tau. Non-phosphorylated Ser-199 and Ser-202 of tau were recognized by monoclonal antibody Tau 1 Ž1:10000. Ž Boehringer.. The levels of tau phosphorylation kinases were analyzed by using polyclonal antibodies for p23 Ž1:400. w23x and GSK3b Ž1:400. w27x, and a monoclonal antibody against CDK-5 Ž1:1000. Ž Santa Cruz.. 2.2. Cell culture COS-7 cells were routinely cultured in DMEM containing 10% fetal bovine serum. For transient expression, 1 = 10 6 cells were plated on 75 cm2 flasks and grown overnight. After being washed with DMEM without serum, cells were incubated with DMEM containing 10% Nu serumŽ Gibco BRL., 0.5 mgrml DEAE-dextran, 5 mgrml of each plasmid DNA Žhtau40 w26x, human TPKI ŽL33801. and bovine
p23 ŽX78934.. for 4 h in a CO 2 incubator Ž378C, 5% CO 2 . . Then, they were washed by DMEM and cultured in routine medium for 72 h. 2.3. Plasmid construction The plasmid constructs were obtained after PCR amplification of corresponding full-length cDNA. Amplified PCR products for htau40 and human TPKI were further subcloned into the mammalian expression vector pCR3 using the Eucaryotic TA cloning Kit ŽInvitrogen.. The PCR product for the bovine p23 coding sequence was inserted at the XbaI site of the pEF-BOS vector. For TPKI, a Ser-9 to Ala mutation was introduced to assure a constitutively active expression of the kinase w28x. All of these plasmids contain a SV40 origin supporting their replication and short-term high level protein expression in COS-7 cells. 2.4. Immunoprecipitation of the reconstituted heterodimer (p23 r cdk-5) Cells were collected, and lysed in Tris-buffer ŽpH7.4. containing 1% Triton X100 and protease inhibitors Ž1 mM phenylmethylsulfonyl fluoride and leupeptin, pepstatin and aprotinin each at 1 mgrml.. From the cell lysate, 200 mg of protein was incubated with antibody for p23 and shaken gently for 8 h at 48C. Then, 5 ml of Protein-G–Sepharose 4 Fast Flow suspension ŽPharmacia. was added, and the mixture was incubated for 2 h at 48C on a rocking platform. The tertiary protein G–antigen–antibody complexes were then collected by centrifugation and washed five times with Tris-buffer containing 1% Triton X100. After addition of SDS-PAGE sample buffer and boiling for 3 min, the p23-associated CDK5 was analyzed by SDS polyacrylamide gel electrophoresis. For the detection of Cdk5, a mouse monoclonal antibody ŽSanta Cruz. was used at a dilution of 1r500. 2.5. Enzyme actiÕity After transfection, cells were collected and homogenized in 10 mM Tris, 150 mM NaCl, 2 mM EGTA, 2 mM DTT, protease inhibitors, and 0.1 mM sodium vanadate. The cell homogenates were centrifuged at
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
179
18000 = g for 30 min at 48C. Protein concentration in the supernatants was determined by the Bradford method, and 6 mg samples were used to determine each enzyme activity. TPK IrGSK-3b activities were assayed by immunoprecipitation assay w29x. CDK activities were assayed as described w15x using substrate peptide PKTPKKAKKL ŽHSŽ9-18.. w30x 2.6. Analysis of tau phosphorylation
Fig. 1. Analysis of expression of GSK-3b in transfected COS-7 cells. ŽA. GSK-3b expression in COS-7 cells. After transfection, each cell lysate Ž10 mg. was analyzed by western blotting using anti-GSK-3b antibody. Lane 1; mock transfected cells, Lane 2; GSK-3b transfected cells. ŽB. GSK-3b activity in COS-7 cells. The GSK-3b activity was expressed as the level of 32 P uptake into substrate Žmean " S.D. Ž ns 4.., ), P - 0.001 Žtwo tailed t-test..
Cells were harvested and lysed in RIPA buffer Ž50 mM Tris–Cl, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 5 mgrml pepstatin, 10 mgrml leupeptin, and 1 mgrml PMSF. 72 h after transfection. The lysates were centrifuged in a microcentrifuge at 15,000 rpm at 48C, and then the supernatant was collected. To analyze tau phosphorylation, tau is partially purified as a heat stable soluble fraction. Since the efficiency of transfection varies slightly from one experiment to another, the study of the phosphorylation of tau was carried out by normalizing the content of tau in the heat-stable fraction. The amount of tau in every sample was determined after
Fig. 2. Analysis of expression of GSK-3b and p23 in transfected COS-7 cells. The p23 expression ŽA. and CDK5 expression in the transfected COS-7 cells. Cell lysate Ž10 mg. of each transfected cells was analyzed by western blotting as described in the Section 2 using anti-p23 antibody ŽA. and anti-CDK5 antibody ŽB. Lane 1: control Lane 2: p23-transfected cell. Lane 3: GSK-3b transfected cells detection in control cell lysate. ŽC. CDK activity in p-23 transfected cells. The activity of CDK was expressed as the level of 32 P uptake into substrate Žmean " S.D. Ž n s 4.., ), P - 0.001 Žtwo tailed t-test.. ŽD. Immunoprecipitation of the reconstituted Žp23rCdk5. heterodimer. The immunocomplex of p23 and CDK5 was precipitated with anti-p23-protein-G–sepharose from each cell lysates Ž200 mg., and analyzed with anti-CDK5 antibody: Lane 1: mock transfected cell. Lane 2: p23 transfected cell.
180
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
Western blotting using the phosphorylation-independent antibody JM labeled by 125 I-protein A, and the quantitation of tau was performed with a laser image analyzer ŽBAS 2000, Fuji.. The equal amount of tau in each lane was then analyzed as to the phosphorylation level by using phosphorylation-dependent anti-tau antibodies. 2.7. EÕaluation of GSK-3b and p23-CDK5 actiÕator expression The total protein concentration of the cell extract was determined. Equivalent amounts were electrophoresed on SDS-PAGE gels and analyzed by Western blotting using anti-GSK-3b rabbit polyclonal antibody for GSK-3b and anti-p23 rabbit antibody for p23 as the primary antibody. HRP-conjugated goat anti-rabbit IgG ŽH q L. Ž 1:1000. or rabbit anti-mouse IgG Ž1:2000. was used as the secondary antibody, and the reacted bands were visualized with the ECL detection system ŽAmersham..
To elucidate whether these kinases can phosphorylate tau in vivo, we co-transfected them with tau. Only tau transfection in COS-7 cells showed two bands with different mobility ŽFig. 3.. In the case of co-transfection of tau and p23, tau showed the same two bands as tau transfection alone. The intensity of the upper band was stronger in p23-co-expressed COS-7 cells than in only tau-expressing cells. The tau with slower mobility seemed to be a phosphorylated species. Thus, tau was phosphorylated by endogenous kinases in COS-7 cells, and the p23 expression increased with the phosphorylation of tau. The GSK-3b expression induced 4 bands of tau with different mobilities ŽFig. 3. as in previous reports w28,31,32x. The transfection of 3 cDNA Žtau, p23, and GSK-3b . into COS-7 cells produced the same tau species as the result of tau plus GSK-3b transfection. The immunoreactivity for Tau-1 antibody was lost in COS-7 cells transfected with each kinase, supporting that each kinase could phosphorylate tau in intact cells. The degree of phosphorylation was different with the introduced kinases.
3. Results COS-7 cells extensively expressed p23 and GSK3b by transient transfection ŽFig. 1ŽA. and Fig. 2Ž A.. . COS-7 cells expressed endogenous GSK-3b, and transient transfection of GSK-3b resulted in an increased level of GSK-3b compared with the endogenous level ŽFig. 1ŽA.., leading to an approximately 4.5 fold activation of GSK-3b relative to non-transfected COS-7 cells ŽFig. 1ŽB... The p23 protein was not observed in non-transfected or GSK-3b transfected cells ŽFig. 2Ž A.. , lane 1 and 3.. After transient transfection of p23 in COS-7 cells, the expression of p23 was observed in cell lysate Ž Fig. 2Ž A. , lane2. . The level of CDK5, which is a catalytic subunit of p23, was not affected by p23 or GSK-3b overexpression ŽFig. 2ŽB... However, p23 expression increased CDK activity to a level more than 10 fold that of non-transfected cells Ž Fig. 2Ž C.. . The immunoprecipitation by anti-p23 appeared to show a direct association of p23 and CDK5 in p23 transfected cells Ž Fig. 2ŽD... An increase of CDK activity due to p23 overexpression and the forming of complex of p23 and CDK5. Thus, the expressed p23 associated with and activated CDK5 in COS-7 cells as indicated by an in vitro study w30x.
Fig. 3. Phosphorylation of tau in transfected COS-7 cells. Phosphorylation of tau was analyzed by two anti-tau antibodies. JM Župper panel. recognizes both phosphorylated and non-phosphorylated tau, and tau 1 Žlower panel. recognizes non-phosphorylated tau. Immunoreactive bands were visualized and quantified by laser image analysis ŽFuji BAS 2000.. Lane 1: recombinant tau, Lane 2: lysate of tau expression, Lane 3: lysate of tau and p23 expression, Lane 4: lysate of tau and GSK-3b expression, Lane 5: tau, p23, and GSK-3b expression.
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
To analyze the phosphorylation sites of tau protein in transfected cells, we produced phosphorylation-dependent antibodies, which specifically recognized each corresponding phosphorylation site Ž Fig. 4. . The endogenous kinases phosphorylated tau at Ser-202 and Ser-404 extensively, and Ser-205 of the tau protein was slightly phosphorylated. The p23 expression, increasing CDK activity, revealed tau phosphorylation at almost the same sites as the endogenous kinases, which showed extensive phosphorylation at Ser-202 and Ser-404, and a slight phosphorylation at Ser-205, Ser-231, and Ser-235. These findings sug-
181
gested that the complex of p23 and CDK5 phosphorylated tau in intact cells but that the phosphorylation sites were almost overlapped with those of endogenous kinases. In GSK-3b or combination GSK-3b plus p23-expressed cells, Ser-199, Thr-231, Ser-396, and Ser-413 were extensively phosphorylated. These sites were reported as GSK-3b phosphorylation sites in vitro and were only slightly or not at all phosphorylated by p23rCDK5 or endogenous kinases. The phosphorylation of Ser-202 and Ser-404 were relatively inhibited by the expression of GSK-3b. GSK-3b may inhibit the activity of some endogenous kinases. The combination of GSK-3b and p23 expression showed the phosphorylation of all 8 sites Ž Ser-199, Ser-202, Thr205, Thr-231, Ser-235, Ser-396, Ser-404, and Ser413.. Thus the expression of both kinases was required for phosphorylation of these sites.
4. Discussion
Fig. 4. Analysis of phosphorylation sites by using specific phosphorylation-dependent anti-tau antibodies in transfected cells. Antibodies used are indicated in the left panel. PS199 Ž1:2000., PS202 Ž1:200., PS235 Ž1:100., PS396 Ž1:200., PS404 Ž1:200., PS413 Ž1:30., PT205 Ž1:100., and PT231 Ž1:200. are rabbit polyclonal antibodies which recognize serine and threonine residues phosphorylated at the corresponding sites on the longest human tau isoform Žhtau40.. Lane 1: recombinant tau, Lane 2: lysate of tau expression, Lane 3: lysate of tau and p23 expression, Lane 4: lysate of tau and GSK-3b expression, Lane 5: tau, p23, and GSK-3b expression.
Recently, it was noted that 19 sites of PHF-tau are phosphorylated w21x. Our study showed that endogenous kinases in COS-7 cells could phosphorylate tau. Strong phosphorylation was shown at Ser-202, Thr205 and Ser-404 in tau transfected COS-7 cells. These sites are phosphorylated by CDK5rp23, cdc2 and MAP kinase in vitrow13,19x. CDK5rp23 were not present in COS-7 cells, but the other kinases work as endogenous kinases. These kinases may explain the phosphorylation of tau in transfected cells. The reported phosphorylation sites of CDK5rP23, which are Ser-202, Thr-205, Ser-235, Ser-404, overlap with those of cdc2 or MAP kinase. This might indicate that p23 overexpression induced the activation of CDK and that the increased activity induced tau phosphorylation at the same phosphorylation sites as endogenous kinases. GSK-3b overexpression appeared to increase the phosphorylation of Ser-199, Thr-231, Ser-396, and Ser-413, sites which were reported as the specific phosphorylation sites of GSK-3b in vitro w25x. Prior-phosphorylation by CDK5rp23 is required in vitro for tau phosphorylation by GSK-3b w24x. However, in intact cells, tau is already phosphorylated by endogenous kinases at the same sites as CDK5rp23, so that GSK-3b could
182
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
phosphorylate tau. GSK-3b recognizes phospho-Ser residues, and phosphorylates Ser or Thr residue around prior-phosphorylated Ser residues. This characteristic of GSK-3b suggests that GSK-3b expression may phosphorylate tau sites other than those reported here. GSK-3b was able to phosphorylate tau in specific sites. However, the involvement of other kinases in the formation of PHF-tau could not be ruled out, since GSK-3b could not phosphorylate Ser-202 and Ser-404 in transfected cells. The exact sites of phosphorylation by these kinases in intact cells remain to be analyzed.
w14x w15x w16x w17x
w18x
w19x w20x
References w1x D.J. Selkoe, Neuron 6 Ž1991. 487–498. w2x I. Grundke-Iqbal, K. Iqbal, M. Quinlan, Y.C. Tung, M.S. Zaidi, H.M. Wisniewski, J. Biol. Chem. 261 Ž1986. 6084– 6089. w3x I. Grundke-Iqbal, K. Iqbal, Y.C. Tung, M. Quinlan, H.M. Wisniewski, L.I. Binder, Proc. Natl. Acad. Sci. U.S.A. 83 Ž1986. 4913–4917. w4x Y. Ihara, N. Nukina, R. Miura, M. Ogawara, J. Biochem 99 Ž1986. 1807–1810. w5x V.M.-Y. Lee, B.J. Balin, L. Otvos, J.Q. Trojanowski, Science 251 Ž1991. 675–678. w6x L.I. Binder, A. Frankfurter, L.I. Rebhun, J. Cell Biol. 101 Ž1985. 1371–1378. w7x A. Watanabe, M. Hasegawa, M. Suzuki, K. Takio, M. Morishima-Kawashima, K. Titani, T. Arai, K.S. Kosik, Y. Ihara, J. Biol. Chem. 268 Ž1993. 25712–25717. w8x Y. Kanai, R. Takemura, T. Oshima, H. Mori, Y. Ihara, M. Yanagisawa, T. Masaki, N. Hirokawa, J. Cell Biol. 109 Ž1989. 1173–1184. w9x D.G. Drubin, M.W. Kirschner, J. Cell Biol. 103 Ž1986. 2739–2746. w10x H. Yoshida, Y. Ihara, J. Neurochem. 61 Ž1993. 1183–1186. w11x A.C. Alonso, T. Zaidi, I. Grundke-Iqbal, K. Iqbal, Proc. Natl. Acad. Sci. U.S.A. 91 Ž1994. 5562–5566. w12x J. Biernat, N. Gustke, G. Drewes, E.-M. Mandelkow, E. Mandelkow, Neuron 11 Ž1993. 153–163. w13x G. Drewes, B. Lichtenberg-Kraag, F. Doring, E.-M. Man-
w21x
w22x w23x w24x w25x w26x w27x
w28x w29x w30x w31x
w32x
delkow, J. Biernat, J. Goris, M. Doree, E. Mandelkow, EMBO J. 11 Ž1992. 2131–2138. I. Correas, J. Diaz-Nido, J. Avila, J. Biol. Chem. 267 Ž1992. 15721–15728. K. Ishiguro, A. Omori, K. Sato, K. Tomizawa, K. Imahori, T. Uchida, Neurosci. Lett. 128 Ž1991. 195–198. K. Ishiguro, A. Omori, M. Takamatsu, K. Sato, M. Arioka, T. Uchida, K. Imahori, Neurosci. Lett. 148 Ž1992. 202–206. K. Ishiguro, A. Shiratsuchi, S. Sato, A. Omori, M. Arioka, S. Kobayashi, T. Uchida, K. Imahori, FEBS Lett. 325 Ž1993. 167–172. C.W. Scott, R.C. Spreen, J.L. Herman, F.P. Chow, M.D. Davison, J. Young, C.B. Caputo, J. Biol. Chem. 268 Ž1993. 1166–1173. R. Vulliet, S.M. Halloran, R.K. Braun, A.J. Smith, G. Lee, J. Biol. Chem. 267 Ž1992. 22570–22574. S.-D. Yang, J.-S. Song, J.-S. Yu, S.-G. Shiah, J. Neurochem. 61 Ž1993. 1742–1747. M. Morishima-Kawashima, M. Hasegawa, K. Takio, M. Suzuki, H. Yoshida, K. Titani, Y. Ihara, J. Biol. Chem. 270 Ž1995. 823–829. L.H. Tsai, I. Dellale, V.R. Caviness Jr., T. Chae, E. Harlow, Nature 371 Ž1994. 419–423. T. Uchida, K. Ishiguro, J. Ohnuma, M. Takamatsu, S. Yonekura, K. Imahori, FEBS Lett. 355 Ž1994. 35–40. M. Arioka, M. Tsukamoto, K. Ishiguro, R. Kato, K. SAto, K. Imahori, T. Uchida, J. Neurochem. 60 Ž1993. 461–468. K. Ishiguro, K. Sato, M. Takamatsu, J. Park, T. Uchida, K. Imahori, Neurosci. Lett. 202 Ž1995. 81–84. M. Goedert, M.G. Spillantini, R. Jakes, D. Rutherford, R.A. Crowther, Neuron 3 Ž1989. 519–526. M. Takahashi, K. Tomizawa, R. Kato, K. Sato, T. Uchida, S.C. Fujita, K. Imahori, J. Neurochem. 64 Ž1994. 1759– 1768. B.R. Sperber, S. Leight, M. Goedert, V.M.-Y. Lee, Neurosci. Lett. 197 Ž1995. 149–153. J.V. Lint, R.L. Khandelwal, W. Merlevede, J.R. Vandenheede, Anal. Biochem. 208 Ž1993. 132–137. Z. Qi, Q.-Q. Huang, K.-Y. Lee, J. Lew, J.H. Wang, J. Biol. Chem. 270 Ž1995. 10847–10854. S. Lovestone, C.H. Reynolds, D. Latimer, D.R. Davis, B.H. Anderton, J.-M. Gallo, D. Hanger, S. Mulot, B. Marquarrdt, S. Stabel, J.R. Woodgett, C.C.J. Miller, Current Biol. 4 Ž1994. 1077–1086. L. Baum, R. Seger, J.R. Woodgett, S. Kawabata, K. Maruyama, M. Koyama, J. Silver, T. Saitoh, Mol. Brain Res. 34 Ž1995. 1–17.
Characterization of tau phosphorylation in glycogen synthase kinase-3b and cyclin dependent kinase-5 activator žp23 / transfected cells Gilles Michel, Marc Mercken, Miyuki Murayama, Kaori Noguchi, Koichi Ishiguro, Kazutomo Imahori, Akihiko Takashima ) Mitsubishi Kasei Institute of Life Sciences, 11 Minamiooya, Machida-shi, Tokyo 194, Japan Received 30 September 1997; accepted 21 October 1997
Abstract One of the histopathological markers in Alzheimer’s disease is the accumulation of hyperphosphorylated tau in neurons called neurofibrillary tangles ŽNFT. composing paired helical filaments ŽPHF.. Combined tau protein kinase II ŽTPK II., which consists of CDK5 and its activator Žp23., and glycogen synthase kinase-3b ŽGSK-3b . phosphorylate tau to the PHF-form in vitro. To investigate tau phosphorylation by these kinases in intact cells, the phosphorylation sites were examined in detail using well-characterized phosphorylation-dependent anti-tau antibodies after overexpressing the kinases in COS-7 cells with a human tau isoform. The overexpression of tau in COS-7 cells showed extensive phosphorylation at Ser-202 and Ser-404. The p23 overexpression induced a mobility shift of tau, but most of the phosphorylation sites overlapped the endogenous phosphorylation sites. GSK-3b transfection showed the phosphorylation at Ser-199, Thr-231, Ser-396, and Ser-413. Triplicated transfection resulted in phosphorylation of tau at 8 observed sites ŽSer-199, Ser-202, Thr-205, Thr-231, Ser-235, Ser-396, Ser-404, and Ser-413.. q 1998 Elsevier Science B.V. Keywords: GSK-3b; p23; CDK5; Tau; Phosphorylation site; Phosphorylation-dependent anti-tau antibody
1. Introduction Alzheimer’s diseaseŽ AD. is histochemically characterized by the accumulation of amyloid b protein ŽAb . in senile plaques w1x and paired helical filamentsŽPHF. composed of hyperphosphorylated tau in neurofibrillary tangles ŽNFT. w1–5x. Tau is a group of microtubule-associated proteins expressed predominantly in axons w6,7x that promote microtubule assembly in vitro w8x and that stabilize microtubules in vivo )
Corresponding author. Present address: Lab. for Alzheimer’s Disease, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wakoshi, Saitama 350-01, Japan. Fax: q81-48-462-4796; E-mail: [email protected]
w9x. Hyperphosphorylation of tau abrogates its ability to form microtubules, leading to the instability of cytoskeleton w10–12x. Since this instability may cause the degeneration of neurons in AD brain, the mechanism of hyperphosphorylation of tau is important for understanding the pathogenesis of AD. In vitro studies have shown that tau can be a substrate for various protein kinasesw11–21x. We previously identified tau protein kinase I ŽTPK I. which is identical to glycogen synthase kinase-3b ŽGSK-3b . w17xand tau protein kinase II ŽTPK II., which consists of CDK5 and its activator Žp23. w22,23x. GSK-3b required prior phosphorylation by CDK5rp23 w24x for the complete phosphorylation of tau in vitro. Combined, these kinases hyperphosphorylated tau which carries the
0304-4165r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 3 0 4 - 4 1 6 5 Ž 9 7 . 0 0 1 3 9 - 6
178
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
epitope of PHF-tau. To examine tau phosphorylation in intact cells, we transfected these kinases into COS-7 cells, and the phosphorylation sites were examined in detail by using phosphorylation-dependent anti-tau antibodies.
2. Materials and methods 2.1. Antibodies Polyclonal antisera for each phosphopeptide were raised in rabbits by using corresponding synthetic phosphopeptides conjugated to keyhole limpet hemocyanin. In the previous study, we confirmed that produced antibodies reacted specifically for each corresponding phosphorylated site w25x. The following antibodies were used for analyzing the phosphorylation sites of tau: PS199 Ž1:200. directed against phosphorylated Ser-199 of tau Žin the numbering of htau 40 w26x., PS202 Ž 1:200. directed against phosphorylated Ser-202 of tau, PT205 Ž1:200. directed against phosphorylated Thr-205 of tau, PT231 Ž 1:200. directed against phosphorylated Thr-231 of tau, PS235 Ž1:200. directed against phosphorylated Ser-235 of tau, PS396 Ž1:200. directed against phosphorylated Ser-396 of tau, PS404 Ž1:200. directed against phosphorylated Ser-404 of tau, and PS413 Ž 1:30. directed against phosphorylated Ser-413 of tau. JM Ž1:5000. was used for detecting both phosphorylated and nonphosphorylated tau. Non-phosphorylated Ser-199 and Ser-202 of tau were recognized by monoclonal antibody Tau 1 Ž1:10000. Ž Boehringer.. The levels of tau phosphorylation kinases were analyzed by using polyclonal antibodies for p23 Ž1:400. w23x and GSK3b Ž1:400. w27x, and a monoclonal antibody against CDK-5 Ž1:1000. Ž Santa Cruz.. 2.2. Cell culture COS-7 cells were routinely cultured in DMEM containing 10% fetal bovine serum. For transient expression, 1 = 10 6 cells were plated on 75 cm2 flasks and grown overnight. After being washed with DMEM without serum, cells were incubated with DMEM containing 10% Nu serumŽ Gibco BRL., 0.5 mgrml DEAE-dextran, 5 mgrml of each plasmid DNA Žhtau40 w26x, human TPKI ŽL33801. and bovine
p23 ŽX78934.. for 4 h in a CO 2 incubator Ž378C, 5% CO 2 . . Then, they were washed by DMEM and cultured in routine medium for 72 h. 2.3. Plasmid construction The plasmid constructs were obtained after PCR amplification of corresponding full-length cDNA. Amplified PCR products for htau40 and human TPKI were further subcloned into the mammalian expression vector pCR3 using the Eucaryotic TA cloning Kit ŽInvitrogen.. The PCR product for the bovine p23 coding sequence was inserted at the XbaI site of the pEF-BOS vector. For TPKI, a Ser-9 to Ala mutation was introduced to assure a constitutively active expression of the kinase w28x. All of these plasmids contain a SV40 origin supporting their replication and short-term high level protein expression in COS-7 cells. 2.4. Immunoprecipitation of the reconstituted heterodimer (p23 r cdk-5) Cells were collected, and lysed in Tris-buffer ŽpH7.4. containing 1% Triton X100 and protease inhibitors Ž1 mM phenylmethylsulfonyl fluoride and leupeptin, pepstatin and aprotinin each at 1 mgrml.. From the cell lysate, 200 mg of protein was incubated with antibody for p23 and shaken gently for 8 h at 48C. Then, 5 ml of Protein-G–Sepharose 4 Fast Flow suspension ŽPharmacia. was added, and the mixture was incubated for 2 h at 48C on a rocking platform. The tertiary protein G–antigen–antibody complexes were then collected by centrifugation and washed five times with Tris-buffer containing 1% Triton X100. After addition of SDS-PAGE sample buffer and boiling for 3 min, the p23-associated CDK5 was analyzed by SDS polyacrylamide gel electrophoresis. For the detection of Cdk5, a mouse monoclonal antibody ŽSanta Cruz. was used at a dilution of 1r500. 2.5. Enzyme actiÕity After transfection, cells were collected and homogenized in 10 mM Tris, 150 mM NaCl, 2 mM EGTA, 2 mM DTT, protease inhibitors, and 0.1 mM sodium vanadate. The cell homogenates were centrifuged at
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
179
18000 = g for 30 min at 48C. Protein concentration in the supernatants was determined by the Bradford method, and 6 mg samples were used to determine each enzyme activity. TPK IrGSK-3b activities were assayed by immunoprecipitation assay w29x. CDK activities were assayed as described w15x using substrate peptide PKTPKKAKKL ŽHSŽ9-18.. w30x 2.6. Analysis of tau phosphorylation
Fig. 1. Analysis of expression of GSK-3b in transfected COS-7 cells. ŽA. GSK-3b expression in COS-7 cells. After transfection, each cell lysate Ž10 mg. was analyzed by western blotting using anti-GSK-3b antibody. Lane 1; mock transfected cells, Lane 2; GSK-3b transfected cells. ŽB. GSK-3b activity in COS-7 cells. The GSK-3b activity was expressed as the level of 32 P uptake into substrate Žmean " S.D. Ž ns 4.., ), P - 0.001 Žtwo tailed t-test..
Cells were harvested and lysed in RIPA buffer Ž50 mM Tris–Cl, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 5 mgrml pepstatin, 10 mgrml leupeptin, and 1 mgrml PMSF. 72 h after transfection. The lysates were centrifuged in a microcentrifuge at 15,000 rpm at 48C, and then the supernatant was collected. To analyze tau phosphorylation, tau is partially purified as a heat stable soluble fraction. Since the efficiency of transfection varies slightly from one experiment to another, the study of the phosphorylation of tau was carried out by normalizing the content of tau in the heat-stable fraction. The amount of tau in every sample was determined after
Fig. 2. Analysis of expression of GSK-3b and p23 in transfected COS-7 cells. The p23 expression ŽA. and CDK5 expression in the transfected COS-7 cells. Cell lysate Ž10 mg. of each transfected cells was analyzed by western blotting as described in the Section 2 using anti-p23 antibody ŽA. and anti-CDK5 antibody ŽB. Lane 1: control Lane 2: p23-transfected cell. Lane 3: GSK-3b transfected cells detection in control cell lysate. ŽC. CDK activity in p-23 transfected cells. The activity of CDK was expressed as the level of 32 P uptake into substrate Žmean " S.D. Ž n s 4.., ), P - 0.001 Žtwo tailed t-test.. ŽD. Immunoprecipitation of the reconstituted Žp23rCdk5. heterodimer. The immunocomplex of p23 and CDK5 was precipitated with anti-p23-protein-G–sepharose from each cell lysates Ž200 mg., and analyzed with anti-CDK5 antibody: Lane 1: mock transfected cell. Lane 2: p23 transfected cell.
180
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
Western blotting using the phosphorylation-independent antibody JM labeled by 125 I-protein A, and the quantitation of tau was performed with a laser image analyzer ŽBAS 2000, Fuji.. The equal amount of tau in each lane was then analyzed as to the phosphorylation level by using phosphorylation-dependent anti-tau antibodies. 2.7. EÕaluation of GSK-3b and p23-CDK5 actiÕator expression The total protein concentration of the cell extract was determined. Equivalent amounts were electrophoresed on SDS-PAGE gels and analyzed by Western blotting using anti-GSK-3b rabbit polyclonal antibody for GSK-3b and anti-p23 rabbit antibody for p23 as the primary antibody. HRP-conjugated goat anti-rabbit IgG ŽH q L. Ž 1:1000. or rabbit anti-mouse IgG Ž1:2000. was used as the secondary antibody, and the reacted bands were visualized with the ECL detection system ŽAmersham..
To elucidate whether these kinases can phosphorylate tau in vivo, we co-transfected them with tau. Only tau transfection in COS-7 cells showed two bands with different mobility ŽFig. 3.. In the case of co-transfection of tau and p23, tau showed the same two bands as tau transfection alone. The intensity of the upper band was stronger in p23-co-expressed COS-7 cells than in only tau-expressing cells. The tau with slower mobility seemed to be a phosphorylated species. Thus, tau was phosphorylated by endogenous kinases in COS-7 cells, and the p23 expression increased with the phosphorylation of tau. The GSK-3b expression induced 4 bands of tau with different mobilities ŽFig. 3. as in previous reports w28,31,32x. The transfection of 3 cDNA Žtau, p23, and GSK-3b . into COS-7 cells produced the same tau species as the result of tau plus GSK-3b transfection. The immunoreactivity for Tau-1 antibody was lost in COS-7 cells transfected with each kinase, supporting that each kinase could phosphorylate tau in intact cells. The degree of phosphorylation was different with the introduced kinases.
3. Results COS-7 cells extensively expressed p23 and GSK3b by transient transfection ŽFig. 1ŽA. and Fig. 2Ž A.. . COS-7 cells expressed endogenous GSK-3b, and transient transfection of GSK-3b resulted in an increased level of GSK-3b compared with the endogenous level ŽFig. 1ŽA.., leading to an approximately 4.5 fold activation of GSK-3b relative to non-transfected COS-7 cells ŽFig. 1ŽB... The p23 protein was not observed in non-transfected or GSK-3b transfected cells ŽFig. 2Ž A.. , lane 1 and 3.. After transient transfection of p23 in COS-7 cells, the expression of p23 was observed in cell lysate Ž Fig. 2Ž A. , lane2. . The level of CDK5, which is a catalytic subunit of p23, was not affected by p23 or GSK-3b overexpression ŽFig. 2ŽB... However, p23 expression increased CDK activity to a level more than 10 fold that of non-transfected cells Ž Fig. 2Ž C.. . The immunoprecipitation by anti-p23 appeared to show a direct association of p23 and CDK5 in p23 transfected cells Ž Fig. 2ŽD... An increase of CDK activity due to p23 overexpression and the forming of complex of p23 and CDK5. Thus, the expressed p23 associated with and activated CDK5 in COS-7 cells as indicated by an in vitro study w30x.
Fig. 3. Phosphorylation of tau in transfected COS-7 cells. Phosphorylation of tau was analyzed by two anti-tau antibodies. JM Župper panel. recognizes both phosphorylated and non-phosphorylated tau, and tau 1 Žlower panel. recognizes non-phosphorylated tau. Immunoreactive bands were visualized and quantified by laser image analysis ŽFuji BAS 2000.. Lane 1: recombinant tau, Lane 2: lysate of tau expression, Lane 3: lysate of tau and p23 expression, Lane 4: lysate of tau and GSK-3b expression, Lane 5: tau, p23, and GSK-3b expression.
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
To analyze the phosphorylation sites of tau protein in transfected cells, we produced phosphorylation-dependent antibodies, which specifically recognized each corresponding phosphorylation site Ž Fig. 4. . The endogenous kinases phosphorylated tau at Ser-202 and Ser-404 extensively, and Ser-205 of the tau protein was slightly phosphorylated. The p23 expression, increasing CDK activity, revealed tau phosphorylation at almost the same sites as the endogenous kinases, which showed extensive phosphorylation at Ser-202 and Ser-404, and a slight phosphorylation at Ser-205, Ser-231, and Ser-235. These findings sug-
181
gested that the complex of p23 and CDK5 phosphorylated tau in intact cells but that the phosphorylation sites were almost overlapped with those of endogenous kinases. In GSK-3b or combination GSK-3b plus p23-expressed cells, Ser-199, Thr-231, Ser-396, and Ser-413 were extensively phosphorylated. These sites were reported as GSK-3b phosphorylation sites in vitro and were only slightly or not at all phosphorylated by p23rCDK5 or endogenous kinases. The phosphorylation of Ser-202 and Ser-404 were relatively inhibited by the expression of GSK-3b. GSK-3b may inhibit the activity of some endogenous kinases. The combination of GSK-3b and p23 expression showed the phosphorylation of all 8 sites Ž Ser-199, Ser-202, Thr205, Thr-231, Ser-235, Ser-396, Ser-404, and Ser413.. Thus the expression of both kinases was required for phosphorylation of these sites.
4. Discussion
Fig. 4. Analysis of phosphorylation sites by using specific phosphorylation-dependent anti-tau antibodies in transfected cells. Antibodies used are indicated in the left panel. PS199 Ž1:2000., PS202 Ž1:200., PS235 Ž1:100., PS396 Ž1:200., PS404 Ž1:200., PS413 Ž1:30., PT205 Ž1:100., and PT231 Ž1:200. are rabbit polyclonal antibodies which recognize serine and threonine residues phosphorylated at the corresponding sites on the longest human tau isoform Žhtau40.. Lane 1: recombinant tau, Lane 2: lysate of tau expression, Lane 3: lysate of tau and p23 expression, Lane 4: lysate of tau and GSK-3b expression, Lane 5: tau, p23, and GSK-3b expression.
Recently, it was noted that 19 sites of PHF-tau are phosphorylated w21x. Our study showed that endogenous kinases in COS-7 cells could phosphorylate tau. Strong phosphorylation was shown at Ser-202, Thr205 and Ser-404 in tau transfected COS-7 cells. These sites are phosphorylated by CDK5rp23, cdc2 and MAP kinase in vitrow13,19x. CDK5rp23 were not present in COS-7 cells, but the other kinases work as endogenous kinases. These kinases may explain the phosphorylation of tau in transfected cells. The reported phosphorylation sites of CDK5rP23, which are Ser-202, Thr-205, Ser-235, Ser-404, overlap with those of cdc2 or MAP kinase. This might indicate that p23 overexpression induced the activation of CDK and that the increased activity induced tau phosphorylation at the same phosphorylation sites as endogenous kinases. GSK-3b overexpression appeared to increase the phosphorylation of Ser-199, Thr-231, Ser-396, and Ser-413, sites which were reported as the specific phosphorylation sites of GSK-3b in vitro w25x. Prior-phosphorylation by CDK5rp23 is required in vitro for tau phosphorylation by GSK-3b w24x. However, in intact cells, tau is already phosphorylated by endogenous kinases at the same sites as CDK5rp23, so that GSK-3b could
182
G. Michel et al.r Biochimica et Biophysica Acta 1380 (1998) 177–182
phosphorylate tau. GSK-3b recognizes phospho-Ser residues, and phosphorylates Ser or Thr residue around prior-phosphorylated Ser residues. This characteristic of GSK-3b suggests that GSK-3b expression may phosphorylate tau sites other than those reported here. GSK-3b was able to phosphorylate tau in specific sites. However, the involvement of other kinases in the formation of PHF-tau could not be ruled out, since GSK-3b could not phosphorylate Ser-202 and Ser-404 in transfected cells. The exact sites of phosphorylation by these kinases in intact cells remain to be analyzed.
w14x w15x w16x w17x
w18x
w19x w20x
References w1x D.J. Selkoe, Neuron 6 Ž1991. 487–498. w2x I. Grundke-Iqbal, K. Iqbal, M. Quinlan, Y.C. Tung, M.S. Zaidi, H.M. Wisniewski, J. Biol. Chem. 261 Ž1986. 6084– 6089. w3x I. Grundke-Iqbal, K. Iqbal, Y.C. Tung, M. Quinlan, H.M. Wisniewski, L.I. Binder, Proc. Natl. Acad. Sci. U.S.A. 83 Ž1986. 4913–4917. w4x Y. Ihara, N. Nukina, R. Miura, M. Ogawara, J. Biochem 99 Ž1986. 1807–1810. w5x V.M.-Y. Lee, B.J. Balin, L. Otvos, J.Q. Trojanowski, Science 251 Ž1991. 675–678. w6x L.I. Binder, A. Frankfurter, L.I. Rebhun, J. Cell Biol. 101 Ž1985. 1371–1378. w7x A. Watanabe, M. Hasegawa, M. Suzuki, K. Takio, M. Morishima-Kawashima, K. Titani, T. Arai, K.S. Kosik, Y. Ihara, J. Biol. Chem. 268 Ž1993. 25712–25717. w8x Y. Kanai, R. Takemura, T. Oshima, H. Mori, Y. Ihara, M. Yanagisawa, T. Masaki, N. Hirokawa, J. Cell Biol. 109 Ž1989. 1173–1184. w9x D.G. Drubin, M.W. Kirschner, J. Cell Biol. 103 Ž1986. 2739–2746. w10x H. Yoshida, Y. Ihara, J. Neurochem. 61 Ž1993. 1183–1186. w11x A.C. Alonso, T. Zaidi, I. Grundke-Iqbal, K. Iqbal, Proc. Natl. Acad. Sci. U.S.A. 91 Ž1994. 5562–5566. w12x J. Biernat, N. Gustke, G. Drewes, E.-M. Mandelkow, E. Mandelkow, Neuron 11 Ž1993. 153–163. w13x G. Drewes, B. Lichtenberg-Kraag, F. Doring, E.-M. Man-
w21x
w22x w23x w24x w25x w26x w27x
w28x w29x w30x w31x
w32x
delkow, J. Biernat, J. Goris, M. Doree, E. Mandelkow, EMBO J. 11 Ž1992. 2131–2138. I. Correas, J. Diaz-Nido, J. Avila, J. Biol. Chem. 267 Ž1992. 15721–15728. K. Ishiguro, A. Omori, K. Sato, K. Tomizawa, K. Imahori, T. Uchida, Neurosci. Lett. 128 Ž1991. 195–198. K. Ishiguro, A. Omori, M. Takamatsu, K. Sato, M. Arioka, T. Uchida, K. Imahori, Neurosci. Lett. 148 Ž1992. 202–206. K. Ishiguro, A. Shiratsuchi, S. Sato, A. Omori, M. Arioka, S. Kobayashi, T. Uchida, K. Imahori, FEBS Lett. 325 Ž1993. 167–172. C.W. Scott, R.C. Spreen, J.L. Herman, F.P. Chow, M.D. Davison, J. Young, C.B. Caputo, J. Biol. Chem. 268 Ž1993. 1166–1173. R. Vulliet, S.M. Halloran, R.K. Braun, A.J. Smith, G. Lee, J. Biol. Chem. 267 Ž1992. 22570–22574. S.-D. Yang, J.-S. Song, J.-S. Yu, S.-G. Shiah, J. Neurochem. 61 Ž1993. 1742–1747. M. Morishima-Kawashima, M. Hasegawa, K. Takio, M. Suzuki, H. Yoshida, K. Titani, Y. Ihara, J. Biol. Chem. 270 Ž1995. 823–829. L.H. Tsai, I. Dellale, V.R. Caviness Jr., T. Chae, E. Harlow, Nature 371 Ž1994. 419–423. T. Uchida, K. Ishiguro, J. Ohnuma, M. Takamatsu, S. Yonekura, K. Imahori, FEBS Lett. 355 Ž1994. 35–40. M. Arioka, M. Tsukamoto, K. Ishiguro, R. Kato, K. SAto, K. Imahori, T. Uchida, J. Neurochem. 60 Ž1993. 461–468. K. Ishiguro, K. Sato, M. Takamatsu, J. Park, T. Uchida, K. Imahori, Neurosci. Lett. 202 Ž1995. 81–84. M. Goedert, M.G. Spillantini, R. Jakes, D. Rutherford, R.A. Crowther, Neuron 3 Ž1989. 519–526. M. Takahashi, K. Tomizawa, R. Kato, K. Sato, T. Uchida, S.C. Fujita, K. Imahori, J. Neurochem. 64 Ž1994. 1759– 1768. B.R. Sperber, S. Leight, M. Goedert, V.M.-Y. Lee, Neurosci. Lett. 197 Ž1995. 149–153. J.V. Lint, R.L. Khandelwal, W. Merlevede, J.R. Vandenheede, Anal. Biochem. 208 Ž1993. 132–137. Z. Qi, Q.-Q. Huang, K.-Y. Lee, J. Lew, J.H. Wang, J. Biol. Chem. 270 Ž1995. 10847–10854. S. Lovestone, C.H. Reynolds, D. Latimer, D.R. Davis, B.H. Anderton, J.-M. Gallo, D. Hanger, S. Mulot, B. Marquarrdt, S. Stabel, J.R. Woodgett, C.C.J. Miller, Current Biol. 4 Ž1994. 1077–1086. L. Baum, R. Seger, J.R. Woodgett, S. Kawabata, K. Maruyama, M. Koyama, J. Silver, T. Saitoh, Mol. Brain Res. 34 Ž1995. 1–17.