Advertisement

Reconstitution of Neuronal Cdc2-like Kinase from Bacteria-expressed Cdk5 and an Active Fragment of the Brain-specific Activator

KINASE ACTIVATION IN THE ABSENCE OF Cdk5 PHOSPHORYLATION *
  • Zhong Qi
    Affiliations
    Medical Research Council Group in Signal Transduction, Department of Medical Biochemistry, University of Calgary, Calgary, Alberta, T2N 4N1 Canada
    Search for articles by this author
  • Qi-Quan Huang
    Affiliations
    Medical Research Council Group in Signal Transduction, Department of Medical Biochemistry, University of Calgary, Calgary, Alberta, T2N 4N1 Canada
    Search for articles by this author
  • Ki-Young Lee
    Affiliations
    Medical Research Council Group in Signal Transduction, Department of Medical Biochemistry, University of Calgary, Calgary, Alberta, T2N 4N1 Canada
    Search for articles by this author
  • John Lew
    Footnotes
    Affiliations
    Medical Research Council Group in Signal Transduction, Department of Medical Biochemistry, University of Calgary, Calgary, Alberta, T2N 4N1 Canada
    Search for articles by this author
  • Jerry H. Wang
    Correspondence
    Alberta Heritage Foundation for Medical Research Scientist. To whom correspondence and reprint requests should be addressed: Dept. of Medical Biochemistry, Faculty of Medicine, University of Calgary, 3330 Hospital Dr., N.W., Calgary, Alberta T2N 4N1, Canada . Tel.: 403-220-3023; Fax: 403-283-4841
    Affiliations
    Medical Research Council Group in Signal Transduction, Department of Medical Biochemistry, University of Calgary, Calgary, Alberta, T2N 4N1 Canada
    Search for articles by this author
  • Author Footnotes
    * This work was supported by operating grants from the Medical Research Council of Canada, Alzheimer's Society of Canada, and the National Cancer Institute of Canada. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
    Present address: Dept. of Chemistry, University of California at San Diego, La Jolla, California 92093-0654.
Open AccessPublished:May 05, 1995DOI:https://doi.org/10.1074/jbc.270.18.10847
      Neuronal Cdc2-like kinase is a heterodimer of Cdk5 and a 25-kDa subunit which is derived from a brain-specific 35-kDa novel protein, p35 (Lew, J., Huang, Q.-Q., Qi, Z., Winkfein, R. J., Aebersold, R., Hunt, T., and Wang, J. H. (1994) Nature 371, 423-426). Three truncated forms of p35 including the one corresponding to the 25-kDa subunit of the kinase have been expressed in Escherichia coli and shown to activate a bacteria-expressed Cdk5 with equal efficacy. The shortest truncated form of p35, p21, spanning amino acid residues 88 to 291, has been used to reconstitute active Cdk5 kinase and to characterize the activation reaction. The purified kinase displays similar specific enzyme activity and similar phosphorylation site specificity as the neuronal Cdc2-like kinase purified from bovine brain. Bovine brain extract contains Cdk5 uncomplexed with p35 or p25 which has also been found to be activated by p21 or p25. The results substantiate the previous suggestion that p35 is a specific Cdk5 activator. Several observations suggest that, unlike other well characterized Cdc2-like kinases whose activities depend on the phosphorylation of the catalytic subunits at a specific site by a distinct kinase, the reconstituted Cdk5/p21 does not depend on the phosphorylation of Cdk5 for activity. The reconstitution of the highly active Cdk5 kinase was achieved without requiring any other kinase in the reconstitution reaction. The possibility of autophosphorylation of Cdk5 on the putative activation site has been ruled out as no phosphorylation occurred on Cdk5 during the enzyme reaction. The rate and extent of the kinase reconstitution were not significantly affected by Mg2+ATP.

      INTRODUCTION

      The regulation of cell division cycle in eukaryotes is governed by a universal mechanism in which a family of closely related kinases play rate-limiting functions at discrete phases of the cell cycle (for reviews, see Refs.
      • Murray A.W.
      • Kirschner M.W.
      ,
      • Hunt T.
      ,
      • Draetta G.
      ,
      • Nurse P.
      ,
      • Pines J.
      • Hunter T.
      ,
      • Maller J.L.
      ,
      • Norbury C.
      • Nurse P.
      ). In yeast cells, p34cdc2 associates with phase-specific cyclins, G1 or M phase cyclins, to regulate cell cycle progression at specific phases (
      • Nurse P.
      ,
      • Nash R.
      • Tokiwa G.
      • Anand S.
      • Erickson K.
      • Futcher A.B.
      ,
      • Hadwiger J.A.
      • Wittenberg C.
      • Richardson H.
      • DeBaros Lopes M.
      • Reed S.I.
      ,
      • Fantes P.
      ) . In animal cells, while the regulation of the entry and exit of mitotic phase by p34cdc2 is conserved, a distinct Cdc2 homolog, Cdk2 is primarily involved in the control of G1 phase (
      • Fang F.
      • Newport J.W.
      ,
      • Girard F.
      • Strausfeld U.
      • Fernandez A.
      • Lamb N.J.C.
      ,
      • Pagano M.
      • Pepperkok R.
      • Verde F.
      • Ansorge W.
      • Draetta G.
      ,
      • Rosenblatt J.
      • Gu Y.
      • Morgan D.O.
      ,
      • Koff A.
      • Giordano A.
      • Dipty D.
      • Yamashita K.
      • Harper J.W.
      • Elledge S.
      • Nishimoto T.
      • Morgan D.O.
      • Franza B.R.
      • Roberts J.M.
      ,
      • Tsai L.-H.
      • Lees E.
      • Harlow E.
      • Riabowol K.
      ) . In addition, recent studies have revealed the existence of extended families of both Cdc2 homologous proteins and cyclins (
      • Lee M.G.
      • Nurse P.
      ,
      • Hunter T.
      • Pines J.
      ,
      • Meyerson M.
      • Enders G.H.
      • Wu C.-L.
      • Su L.-K.
      • Gorka C.
      • Nelson C.
      • Harlow E.
      • Tsai L.-H.
      ,
      • Sherr C.J.
      ) . Several Cdc2 family members with molecular masses in a range of 33 kDa to 35 kDa have been suggested to depend on cyclin for activity and designated as cyclin-dependent kinases, Cdks; individual members of this group are distinguished by numbers, such as Cdk1 (Cdc2), Cdk2 etc. (
      • Soloman M.J.
      • Glotzer M.
      • Lee T.H.
      • Philippe M.
      • Kirschner M.W.
      ,
      • Xiong Y.
      • Zhang H.
      • Beach D.
      ,
      • Pines J.
      ,
      • Soloman M.J.
      ) . Members of the cyclin family, on the other hand, are categorized into subfamilies by letters as, cyclin A, cyclin B etc. A large number of active protein kinases may be derived from heterologous interaction between members of the two protein families. Cell cycle progression of animal cells appears to be coordinated by the concerted action of many of these kinases.
      In addition to depending on cyclin for activity, Cdc2 and Cdk2 are regulated by complex phosphorylation mechanisms involving a network of protein kinases and phosphatases (
      • Soloman M.J.
      • Glotzer M.
      • Lee T.H.
      • Philippe M.
      • Kirschner M.W.
      ,
      • Gould K.L.
      • Nurse P.
      ,
      • Krek W.
      • Nigg E.A.
      ,
      • Krek W.
      • Nigg E.A.
      ,
      • Ducommun B.
      • Brambilla P.
      • Felix M.A.
      • Franza B.J.
      • Karsenti E.
      • Draetta G.
      ,
      • Norbury C.
      • Blow J.
      • Nurse P.
      ). The phosphorylation of a specific threonine residue, Thr161/Thr160, by a distinct kinase, Cdk activating kinase, is essential for, or markedly enhances, the activity of the Cdc2/Cdk2 kinase (
      • Desai D.
      • Gu Y.
      • Morgan D.O.
      ,
      • Solomon M.J.
      • Lee T.
      • Kirschner M.W.
      ). Cdk activating kinase has recently been identified to contain a Cdc2 homologous protein, MO15, as catalytic subunit and a cyclin partner (
      • Fesquet D.
      • Labbe J.C.
      • Derancourt J.
      • Capony J.P.
      • Galas S.
      • Girard F.
      • Lorca T.
      • Shuttleworth J.
      • Doree M.
      • Cavadore J.C.
      ,
      • Poon R.Y.C.
      • Yamashita K.
      • Adamczewski J.P.
      • Hunt T.
      • Shuttleworth J.
      ,
      • Solomon M.J.
      • Harper J.W.
      • Shuttleworth J.
      ,
      • Fisher R.P.
      • Morgan D.O.
      ,
      • Makela T.P.
      • Tassan J.-P.
      • Nigg E.A.
      • Frutiger S.
      • Hughes G.J.
      • Weinberg R.A.
      ). On the other hand, phosphorylation of a threonine, Thr14 or a tyrosine, Tyr15, in the amino-terminal region results in inactivation of the Cdc2/Cdk2 kinase (
      • Gould K.L.
      • Nurse P.
      ,
      • Krek W.
      • Nigg E.A.
      ,
      • Norbury C.
      • Blow J.
      • Nurse P.
      ). A protein kinase catalyzing the inhibitory tyrosine phosphorylation has been identified and shown to be related to the yeast wee1 gene product (
      • Russell P.
      • Nurse P.
      ,
      • Featherstone C.
      • Russell P.
      ,
      • Parker L.L.
      • Atherton-Fessler S.
      • Lee M.S.
      • Ogg S.
      • Falk J.L.
      • Swenson K.I.
      • Piwnica-Worms H.
      ,
      • Parker L.L.
      • Piwnica-Worms H.
      ), whereas the enzyme responsible for the phosphorylation of Cdc2/Cdk2 kinase on Thr14 has not been identified. Since the regulatory phosphorylation sites are mostly conserved, or conservatively substituted among Cdc2 family members, it seems possible that some other Cdc2-like kinases are also regulated by similar phosphorylation mechanisms. For example, a recent study has shown that the activation of Cdk4 by cyclin D is also dependent on an activating kinase (
      • Kato J.-Y.
      • Matsuoka M.
      • Strom D.K.
      • Sherr C.J.
      ,
      • Matsuoka M.
      • Kato J.-Y.
      • Fisher R.P.
      • Morgan D.O.
      • Sherr C.J.
      ).
      Not all Cdc2-like kinases are cell cycle regulators. Many Cdc2-related proteins are present in tissues where proliferative activity is very low (
      • Meyerson M.
      • Enders G.H.
      • Wu C.-L.
      • Su L.-K.
      • Gorka C.
      • Nelson C.
      • Harlow E.
      • Tsai L.-H.
      ,
      • Draetta G.
      • Beach D.
      • Moran E.
      ,
      • Okuda T.
      • Cleveland J.L.
      • Downing J.R.
      ). Neurons of mammalian central nervous system contain a Cdc2-like kinase, called neuronal Cdc2-like kinase, which has been implicated in the regulation of neurocytoskeleton dynamics (
      • Hellmich M.R.
      • Pant H.C.
      • Wada E.
      • Battey J.F.
      ,
      • Shetty K.T.
      • Link W.T.
      • Pant H.C.
      ,
      • Paudel H.K.
      • Lew J.
      • Ali Z.
      • Wang J.H.
      ). Purified active neuronal Cdc2-like kinase is a heterodimer of Cdk5 and a 25-kDa regulatory protein (
      • Lew J.
      • Beaudette K.
      • Litwin C.M.E.
      • Wang J.H.
      ,
      • Lew J.
      • Winkfein R.J.
      • Paudel H.K.
      • Wang J.H.
      ,
      • Ishiguro K.
      • Kobayashi S.
      • Omori A.
      • Takamatsu M.
      • Yonekura S.
      • Anzai K.
      • Imahori K.
      • Uchida T.
      ). Molecular cloning studies have revealed that the 25-kDa protein is derived from a novel 35-kDa protein, p35 (
      • Lew J.
      • Huang Q.-Q.
      • Qi Z.
      • Winkfein R.J.
      • Aebersold R.
      • Hunt T.
      • Wang J.H.
      ). The mRNA transcript of p35 has been shown to be specifically expressed in the central nervous system. The primary structure of p35 shows only limited homology to members of the cyclin family; yet, the protein appears to perform a cyclin-like role in neuronal Cdc2-like kinase. A bacteria-expressed, truncated form of p35 has been found to activate a bacteria-expressed Cdk5.
      In the present study, we demonstrate the activation of Cdk5 by a number of bacterially expressed, truncated forms of p35, including one corresponding to the 25-kDa subunit of neuronal Cdc2-like kinase purified from bovine brain, and carry out detailed characterization of the kinase activation reaction. The study has revealed a major difference between the mechanism of Cdk5 activation and that of Cdc2 or Cdk2 activation. Although the amino acid sequence of Cdk5 contains a phosphorylatable residue, serine, at the position corresponding to the positive regulatory threonine of Cdc2 or Cdk2, the activation of Cdk5 does not appear to be dependent on the presence of a Cdk activating kinase. The active form of the kinase does not catalyze autophosphorylation of Cdk5, thus ruling out the possibility that the specific serine residue is phosphorylated by an autocatalytic reaction.

      MATERIALS AND METHODS

      Fusion Plasmid Construction and Escherichia coli Expression

      Four sets of polymerase chain reaction primers were designed according to the full sequence of neuronal regulatory subunit described previously (
      • Lew J.
      • Huang Q.-Q.
      • Qi Z.
      • Winkfein R.J.
      • Aebersold R.
      • Hunt T.
      • Wang J.H.
      ) for plasmid constructions to express the full-length or truncated forms of the neuronal regulatory subunit (Fig. 1). Primer 1, 5’-CCCGGGGATCCATGGGCACGGTGCTG-3’, and primer 3, 5’-CCCGGGAATTCTCACCGGTCCAGCCC-3’, are for the full-length protein of 307 amino acids, p35. Primer 2, 5’-CCGGGGATCCGCCCAGCCCCCGCCGG-3’, and primer 3 are for the 25-kDa protein (p25) found in the purified neuronal Cdc2-like kinase. Primer 4, 5’-CCCGGGGATCCAAGTCGCTGTCGTGC-3’, and primer 6, 5’-CCCGGGAATTCTAGTTCTTCAGGTCGGAGA-3’, are for the 23-kDa truncated form (p23) of p35, which encodes amino acids 88-291. Primer 5, 5’-CCCGGATCCTGTCCCCCTGCCAGC-3’, and primer 6 are for the 21-kDa fragment (p21) of p35, which encodes amino acids 108-291. These expressed fragments as well as the 35-kDa full-length sequence are schematically shown in Fig. 1. The polymerase chain reaction products were cloned in pGEX-2T vector at BamHI and EcoRI sites.
      Figure thumbnail gr1
      Figure 1:Schematic representation of p35 and the various truncated forms of p35.
      The recombinant pGEX plasmid constructs, encoding glutathione S-transferase (GST,
      The abbreviations used are: GST
      glutathione S-transferase
      DTT
      dithiothreitol
      MOPS
      4-morpholinepropanesulfonic acid
      PBS
      phosphate-buffered saline
      PAGE
      polyacrylamide gel electrophoresis
      FPLC
      fast protein liquid chromatography.
      ) 26 kDa) fusion proteins, GST-p21, GST-p23, GST-p25, GST-Cdk5, and GST-cyclin D1, were individually expressed in E. coli strain BL21 (DE3). The host cells were cultured in 1 liter of LB medium containing 100 μg/ml ampicillin to A600 of 0.5-0.8 at 37°C. The synthesis of GST recombinant protein was induced with 0.4 mM isopropyl-1-thio-β-D-galactopyranoside (Life Technologies, Inc.) for a subsequent incubation of 2-3 h. The cells were harvested by centrifugation at 2000 × g. The cell pellets were washed with PBS and pelleted again for subsequent purification of the recombinant proteins.

      Purification of the Recombinant Proteins

      Purification of the GST fusion proteins was carried out essentially as described by Smith and Johnson (
      • Smith D.B.
      • Johnson K.S.
      ). The cell pellets were suspended in 20 ml of 50 mM Tris-HCl (pH 7.4) buffer containing 2 mM EDTA, 1% Tween 20, 1 mM DTT, 0.25 mM phenylmethylsulfonyl fluoride, and 1 μg/ml each of leupeptin, aprotinin, and pepstatin. All procedures for the purification were carried out at 0-5°C. The cells were lysed using a French press at 1000 p.s.i. and then centrifuged at 18,000 × g for 15 min. The supernatant was applied onto a 2-ml column of glutathione-agarose (Sigma) which had been equilibrated with 20 mM Tris-HCl (pH 7.4) supplemented with 0.5 M NaCl, 0.5% Nonidet P-40, 1 mM EDTA and DTT, and the mixture of protease inhibitors described above. The column was sealed and allowed to sit on an end-over-end shaker for 1 h. After loading, the column was washed with 20 bed volumes of the same buffer, followed by 5 bed volumes of PBS supplemented with 1 mM DTT. The GST fusion protein was eluted by 5 mM reduced glutathione (Sigma) in 50 mM Tris-HCl, pH 8.0, and 1 mM DTT. To remove glutathione, the pooled samples were dialyzed against PBS containing 1 mM EDTA and DTT and then concentrated to a small volume by dialysis against Aquacide II (Calbiochem). His6-tagged cyclin A was expressed and purified as described by Poon et al. (
      • Poon R.Y.C.
      • Yamashita K.
      • Adamczewski J.P.
      • Hunt T.
      • Shuttleworth J.
      ) except that cells were lysed by passing through a French Press in the lysis buffer without lysozyme.
      For further purification of GST-p21 to remove most of the proteolyzed products, the concentrated GST-p21 was clarified by centrifugation and loaded on an FPLC Superose 12 gel filtration column (Prep grade, Pharmacia Biotech Inc.) equilibrated in PBS with 1 mM EDTA and 1 mM DTT and eluted at a flow rate of 0.5 ml/min. Fractions containing the 45-kDa polypeptide of GST-p21, as determined by SDS-PAGE, were pooled and then concentrated by dialysis against Aquacide II to a small volume.

      Reconstitution and Purification of Neuronal Cdc2-like Kinase

      2 mg of GST-Cdk5 affinity-purified by a glutathione-agarose column was mixed with 4 mg of GST-p21 purified from a glutathione-affinity column and gel filtration in a volume of 5 ml in PBS buffer containing 1 mM EDTA and 1 mM DTT. The mixture was allowed to sit at 30°C for 1 h for reconstitution and subsequently treated with human thrombin (Sigma) in a concentration of 10 units of thrombin per mg of recombinant protein at room temperature for 1 h to cut out the GST fusion part. The GST-free complex of Cdk5 and p21 was dialyzed against 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1 mM DTT, 0.25 mM phenylmethylsulfonyl fluoride, and 100 μg/ml benzamidine at 4°C for 2 h. The dialyzed sample was clarified by centrifugation and then loaded on an FPLC Mono S column (HR5/5, Pharmacia) equilibrated with the same buffer at a flow rate of 0.5 ml/min. The column was washed with 10 bed volumes of the same buffer, and the bound kinase activity was eluted by a linear gradient of NaCl (0-0.3 M/10 ml). Protein kinase activity was monitored by phosphorylation assay of the histone peptide.

      Isolation of the Native Cdk5 Monomer

      The bovine brain extract in Buffer A (25 mM Hepes (pH 7.2), 1 mM EDTA, 1 mM DTT, 0.6 mM phenylmethylsulfonyl fluoride, 0.3 mg/ml benzamidine, 1 μg/ml leupeptin, 2 μg/ml antipain) supplemented with 0.1 mg/ml soybean trypsin inhibitor was clarified by centrifugation at 100,000 × g. The supernatant containing 40 mg of protein was injected at the flow rate of 0.5 ml/min into an FPLC Mono S column (HR 5/5, Pharmacia) equilibrated with Buffer A. The column was washed with 20 ml of Buffer A and eluted with a salt gradient (0-0.5 M NaCl) in 30 ml of Buffer A. The fractions containing Cdk5 monomer were identified by immunoblots with the Cdk5 and p25 antibodies. The pooled sample was concentrated by Amicon Centricon-10 concentrator and then applied to an FPLC Superose 12 column (HR 10/30, Pharmacia) equilibrated in Buffer A supplemented with 0.15 M NaCl at 0.5 ml/min. This column was calibrated in the same condition using a Bio-Rad calibration kit (bovine thyroglobulin, 670 kDa; bovine γ-globulin, 158 kDa; chicken ovalbumin, 44 kDa; horse myoglobin, 17 kDa; vitamin B12, 1.35 kDa). Immunoblotting and reconstitution assay with GST-p21 or GST-p25 were performed with the column fractions.

      Protein Kinase Activity Measurement

      The protein kinase activity and substrate specificity were determined according to the procedure described previously (
      • Lew J.
      • Beaudette K.
      • Litwin C.M.E.
      • Wang J.H.
      ,
      • Beaudette K.N.
      • Lew J.
      • Wang J.H.
      ). The histone H1-derived peptide, HS (
      • Murray A.W.
      • Kirschner M.W.
      ,
      • Hunt T.
      ,
      • Draetta G.
      ,
      • Nurse P.
      ,
      • Pines J.
      • Hunter T.
      ,
      • Maller J.L.
      ,
      • Norbury C.
      • Nurse P.
      ,
      • Nash R.
      • Tokiwa G.
      • Anand S.
      • Erickson K.
      • Futcher A.B.
      ,
      • Hadwiger J.A.
      • Wittenberg C.
      • Richardson H.
      • DeBaros Lopes M.
      • Reed S.I.
      ,
      • Fantes P.
      ,
      • Fang F.
      • Newport J.W.
      ,
      • Girard F.
      • Strausfeld U.
      • Fernandez A.
      • Lamb N.J.C.
      ,
      • Pagano M.
      • Pepperkok R.
      • Verde F.
      • Ansorge W.
      • Draetta G.
      ,
      • Rosenblatt J.
      • Gu Y.
      • Morgan D.O.
      ,
      • Koff A.
      • Giordano A.
      • Dipty D.
      • Yamashita K.
      • Harper J.W.
      • Elledge S.
      • Nishimoto T.
      • Morgan D.O.
      • Franza B.R.
      • Roberts J.M.
      ,
      • Tsai L.-H.
      • Lees E.
      • Harlow E.
      • Riabowol K.
      ,
      • Lee M.G.
      • Nurse P.
      ,
      • Hunter T.
      • Pines J.
      ), KTPKKAKKPKTPKKAKKL, was used as substrate for the protein kinase activity assay (
      • Lew J.
      • Beaudette K.
      • Litwin C.M.E.
      • Wang J.H.
      ). This sequence contains a repeat of the proline-directed phosphorylation site. A set of peptide analogues, derived from the parent peptide HS (
      • Hadwiger J.A.
      • Wittenberg C.
      • Richardson H.
      • DeBaros Lopes M.
      • Reed S.I.
      ,
      • Fantes P.
      ,
      • Fang F.
      • Newport J.W.
      ,
      • Girard F.
      • Strausfeld U.
      • Fernandez A.
      • Lamb N.J.C.
      ,
      • Pagano M.
      • Pepperkok R.
      • Verde F.
      • Ansorge W.
      • Draetta G.
      ,
      • Rosenblatt J.
      • Gu Y.
      • Morgan D.O.
      ,
      • Koff A.
      • Giordano A.
      • Dipty D.
      • Yamashita K.
      • Harper J.W.
      • Elledge S.
      • Nishimoto T.
      • Morgan D.O.
      • Franza B.R.
      • Roberts J.M.
      ,
      • Tsai L.-H.
      • Lees E.
      • Harlow E.
      • Riabowol K.
      ,
      • Lee M.G.
      • Nurse P.
      ,
      • Hunter T.
      • Pines J.
      ), PKTPKKAKKL, was used for determination of the substrate specificity (
      • Beaudette K.N.
      • Lew J.
      • Wang J.H.
      ). The kinase reaction was performed at 30°C, and the reaction mixture contained 20 mM MOPS, pH 7.4, 5 mM MgCl2, 100 μM [γ-32P]ATP (~300 dpm/pmol), peptide substrate as indicated, and the protein kinase sample. When assay of purified kinase was performed, 0.5 mg/ml bovine serum albumin was included in the reaction mixture for enzyme stability. Phosphate incorporation into substrate peptide was quantitated by liquid scintillation using a scintillation counter (Beckman LKB 1215).

      Protein Concentration Determination

      Protein concentration was measured either by absorbance at A280 or by the method of Bradford (
      • Bradford M.M.
      ). The Bradford assay was performed in an alternate way. The samples were mixed with Bradford reagent in a microtiter plate and read at A570 using a Dynatec MR 600 microtiter plate reader. The amount of pure enzyme in the purified sample was estimated by scanning densitometry of Coomassie Blue-stained gels using a Pharmacia LKB 2202 Ultrascan laser densitometer. The stained protein bands were compared with carbonic anhydrase as the protein standard.

      SDS-PAGE and Immunoblot

      SDS-PAGE was performed by the method of Laemmli (
      • Laemmli U.K.
      ) in 10% vertical slab gels, and gels were stained with silver (
      • Wray W.
      • Boulikas T.
      • Wray V.
      • Hankcock R.
      ) or 0.2% Coomassie Brilliant Blue. For immunoblot, proteins were transferred to polyvinylidene difluoride membranes (Millipore) and immunostained with the indicated antibodies. Antibodies against Cdk5 and the p35 derivatives were rabbit polyclonal antibody preparations raised against the bacteria-expressed Cdk5 and 21-kDa fragment of p35, respectively.

      RESULTS

      Protein Kinase Activity of Cdk5 and p35-related Proteins

      The expression of GST-fused Cdk5 and p35, as well as p25, p23, and p21 truncated forms of p35 (Fig. 1) were attempted in E. coli strain BL21 (DE3). The bacteria lysates were centrifuged to separate into a soluble and a particulate fraction, and expression of the recombinant proteins was analyzed by SDS-PAGE and immunoblot. The expression of the full-length p35 fusion protein was not successful; essentially no fusion protein was detectable in the bacteria lysate (data not shown). Most of the expressed truncated forms of p35 were found in the insoluble fraction. In all cases, the soluble recombinant proteins appeared to have undergone considerable proteolytic degradation. Lowering the induction temperature from 37°C to room temperature or decreasing the protein induction time to 1 h did not significantly affect the extent of the protein degradation. Attempts to express the full-length p35 using a different vector, PQE30 (QIAGEN), was also unsuccessful.
      The expressed proteins were purified from the soluble fractions of bacteria lysates by glutathione affinity column chromatography. As shown previously (
      • Lew J.
      • Huang Q.-Q.
      • Qi Z.
      • Winkfein R.J.
      • Aebersold R.
      • Hunt T.
      • Wang J.H.
      ), the expressed Cdk5 fusion protein did not exhibit any kinase activity by itself, nor did any of the p35-related fusion proteins. Protein kinase activity, however, could be readily detected upon mixing Cdk5 with the fusion proteins of the truncated forms of p35, and the fusion proteins of p21, p23, and p25 showed similar activity in Cdk5 activation (Fig. 2 A). To achieve maximal activation, the mass ratio of activating proteins to Cdk5 of 8:1 was used in these experiments, and the protein mixtures were preincubated for 30 min before kinase assay. Dose-dependent activation of Cdk5 by p21 indicated that close to maximal activation could be approached at a mass ratio of Cdk5/p21 of 1:1.5 (Fig. 2 B). The optimal activation of Cdk5 by the truncated forms of p35 also requires preincubation of the proteins. For example, the time course of the kinase reaction of a mixture of Cdk5 and p21 displays low kinase activity and pronounced upward curvature indicative of a slow activation of the kinase during the reaction. If the two proteins have been incubated for 1 h before the kinase reaction is initiated, a linear initial time course with greatly increased kinase activity will result (Fig. 2 C). Since our attempts in bacteria expression of the full-length p35 had not been successful, a similar reconstitution experiment with the full-length protein was not possible. p25 corresponds to the 25-kDa subunit of the purified bovine brain neuronal Cdc2-like kinase which exhibits specific kinase activities similar to or higher than those of the fully activated Cdc2 and Cdk2 kinases. Thus, the results of Fig. 2 suggest that the three truncated forms of p35 contain the complete Cdk5-activating domain. The activation of Cdk5 by p35 derivatives appeared to be specific, the bacteria-expressed His6-tagged cyclin A and GST-cyclin D1 could not substitute for the p35 derivatives in Cdk5 activation under the same condition (Fig. 2 A).
      Figure thumbnail gr2
      Figure 2:Activation of bacteria-expressed Cdk5. A, 0.5 μg of GST-Cdk5 and 5 μg each of GST-p21, GST-p23, GST-p25, His6-cyclin A, and GST-cyclin D1 were preincubated alone or together at 30°C for 30 min in a volume of 15 μl of PBS buffer containing 1 mM EDTA and 1 mM DTT. 10× concentrated protein kinase assay mixture was added to each sample to make the ATP and substrate peptide HS(1-18) concentration 100 μM and the Mg2+ concentration 5 mM in the kinase reaction. B, dose-dependent activation of Cdk5 by p21. 2 μg of GST-Cdk5 was reconstituted with various amounts of GST-p21 at 30°C for 1 h, other conditions were the same as in A. The kinase reaction was performed at 30°C for 20 min as described under “Materials and Methods.” C, effect of preincubation on kinase activity. GST-Cdk5 (10 μg) and GST-p21 (20 μg) in PBS buffer containing 1 mM EDTA and 1 mM DTT were mixed in a final volume of 150 μl. Kinase reaction was initiated by the addition of [γ-32P]ATP, Mg2+, and peptide substrate immediately (○- - -○) or after incubation of the mixture at 30°C for 1 h (●–●). At the indicated time after initiation of the kinase reaction, aliquots (15 μl) were withdrawn for the analysis of phosphate incorporation.
      Bovine brain extract contains a high concentration of Cdk5 that does not complex with p25 or p35. The p25/p35-free Cdk5 in the 100,000 × g fraction can be separated from the complexed forms by FPLC on a Mono S column (see “Materials and Methods”). When the Mono S column-purified Cdk5 fraction was analyzed on an FPLC Superose 12 column, all Cdk5 appeared in a single peak (as revealed by Western immunoblot analysis) at a position indicative of Cdk5 monomer (Fig. 3 A). While the fractions containing Cdk5 displayed no histone H1 peptide kinase activity, high kinase activity could be demonstrated after incubation of the fractions with the bacterially expressed p21 for 40 min. The activation of brain Cdk5 shows characteristics similar to that of the recombinant Cdk5 expressed in bacteria. p21 and p25 have been found to activate the brain Cdk5 with essentially identical efficiency (Fig. 3 B).
      Figure thumbnail gr3
      Figure 3:Analysis of the p25/p35-free Cdk5 from bovine brain extract. A, Superose 12 gel filtration chromatography. The crude brain extract was applied on Mono S column and the fractions containing Cdk5 which was uncomplexed with p25 or p35 were pooled and concentrated. 0.5 ml of the sample was injected into the FPLC Superose 12 column (bed volume = 25 ml) equilibrated in Buffer A with 0.15 M NaCl. The column was precalibrated in the same condition as described under “Materials and Methods.”∆–∆, A280. Fractions were collected in 0.5 ml. The column fractions were incubated with (●–●) or without (○–○) GST-p21 at 30°C for 40 min before the histone peptide kinase activity was assayed. The inset is the immunoblots with antibodies against Cdk5 (top panel) and p25/p35 (bottom panel), respectively. B, activation of the monomeric Cdk5 isolated from brain extract by GST-p21 and GST-p25. Aliquots (5.4 μg) of the sample pooled from the Superose 12 collections containing the monomeric Cdk5, was incubated alone or reconstituted with 1 μg of GST-p21 or GST-p25 at 30°C for 40 min. Histone peptide kinase activity was assayed at 30°C for 20 min as described under “Materials and Methods.”
      Specific kinase activity of the close-to-homogeneous neuronal Cdc2-like kinase from bovine brain is about 4 μmol of PO4/min/mg of protein using histone H1 peptide HS (
      • Murray A.W.
      • Kirschner M.W.
      ,
      • Hunt T.
      ,
      • Draetta G.
      ,
      • Nurse P.
      ,
      • Pines J.
      • Hunter T.
      ,
      • Maller J.L.
      ,
      • Norbury C.
      • Nurse P.
      ,
      • Nash R.
      • Tokiwa G.
      • Anand S.
      • Erickson K.
      • Futcher A.B.
      ,
      • Hadwiger J.A.
      • Wittenberg C.
      • Richardson H.
      • DeBaros Lopes M.
      • Reed S.I.
      ,
      • Fantes P.
      ,
      • Fang F.
      • Newport J.W.
      ,
      • Girard F.
      • Strausfeld U.
      • Fernandez A.
      • Lamb N.J.C.
      ,
      • Pagano M.
      • Pepperkok R.
      • Verde F.
      • Ansorge W.
      • Draetta G.
      ,
      • Rosenblatt J.
      • Gu Y.
      • Morgan D.O.
      ,
      • Koff A.
      • Giordano A.
      • Dipty D.
      • Yamashita K.
      • Harper J.W.
      • Elledge S.
      • Nishimoto T.
      • Morgan D.O.
      • Franza B.R.
      • Roberts J.M.
      ,
      • Tsai L.-H.
      • Lees E.
      • Harlow E.
      • Riabowol K.
      ,
      • Lee M.G.
      • Nurse P.
      ,
      • Hunter T.
      • Pines J.
      ) as substrate (
      • Lew J.
      • Beaudette K.
      • Litwin C.M.E.
      • Wang J.H.
      ). This value is similar to the specific kinase activity of purified Cdk2/cyclin A protein kinase (
      • Connell-Crowley L.
      • Solomon M.J.
      • Wei N.
      • Harper J.W.
      ). The maximal kinase activity obtained by the mixture of recombinant Cdk5 and p21, however, was only several tens of nanomoles per min per mg of protein. A number of factors may contribute to the low specific activity displayed by the reconstituted enzyme. More than 50% of the expressed GST-Cdk5 and GST-p21 proteins purified by glutathione beads were degraded forms which, upon being separated from the intact protein by gel filtration, were found to have little p21-activated histone kinase activity and Cdk5-activating activity, respectively. Moreover, gel filtration analysis showed that only the trailing edge of the protein peak enriched with the intact Cdk5 or p21 fusion proteins displayed activities (data not shown), suggesting that most of the proteins in the affinity-purified GST-Cdk5 and GST-p21 were present in incorrectly folded states.

      Purification and Characterization of the Active Form of Reconstituted Cdk5/21-kDa Kinase

      Since the low specific activity of the reconstituted enzyme appeared to be due, at least in part, to the existence of a large amount of degraded and incorrectly folded recombinant proteins, attempts were made to purify the active form of the enzyme from the reconstitution reaction. It was found that a single step FPLC Mono S column chromatography of an optimally reconstituted and thrombin-treated sample could achieve a 65-fold purification of the active kinase (for detailed procedures, see “Materials and Methods”). Most of the protein in the reconstitution mixture did not bind to the Mono S column, whereas about 85% of the total kinase activity bound. The bound activity could be quantitatively eluted as a sharp peak, along with a small amount of protein from the column by a salt gradient (Fig. 4 A).
      Figure thumbnail gr4
      Figure 4:FPLC Mono S chromatography of the reconstituted kinase. The reconstitution and purification of the Cdk5-p21 complex were carried out as described in detail under “Materials and Methods.” After loading and column washing, the collection was started by fractions of 0.5 ml. A: ○- - -○, histone peptide phosphorylation activity; ●–●, A570. Protein was determined by the Bradford method performed in a microtiter plate. B, SDS-PAGE of the indicated fractions from A. os is the original sample loaded on the column. ft is the flow-through of the loaded sample, and w is the collected column wash after sample loading. The gel was stained with silver.
      Analysis by SDS-PAGE revealed that the 33-kDa and 21-kDa proteins were abundant in the histone kinase active fractions (Fig. 4 B), which were identified as Cdk5 and p21, respectively, by immunoblots (data not shown). Densitometric scanning of the Coomassie Blue-stained gels indicated that Cdk5 and p21 were about 50% of the total protein, and the molar ratio of the 21-kDa protein to the 33-kDa protein was about 1.3. Purification data for a typical preparation are summarized in Table I. Protein concentration of the activity peak fraction (Fraction 14) was determined by both scanning densitometry of the Coomassie Blue-stained gels and the Bradford assays performed in a microtiter plate. The purified, reconstituted Cdk5/p21 kinase had a specific kinase activity of 3.8 μmol of PO4/min/mg of protein, 65-fold higher than that of the crude reconstituted enzyme. Since about 50% of the proteins of the purified sample were represented by Cdk5 and p21, a homogeneous active enzyme is expected to have a specific kinase activity of about 7 μmol of PO4/min/mg of protein. Thus, the active form of the kinase reconstituted from bacteria-expressed Cdk5 and a truncated form of p35 has a specific activity as high as that of neuronal Cdc2-like kinase purified from bovine brain, about a few micromoles of PO4 transferred per min per mg of enzyme.
      Table IPurification of the reconstituted kinase by Mono S chromatography
      We have previously used a set of peptide analogues to elucidate the substrate structural determinants for bovine brain neuronal Cdc2-like kinase (
      • Beaudette K.N.
      • Lew J.
      • Wang J.H.
      ). These peptides were prepared by systematic substitution of a parent peptide derived from the Cdc2 kinase phosphorylation site of histone H1. Kinetic characterization of neuronal Cdc2-like kinase-catalyzed peptide phosphorylation reactions was carried out, and the results were used to assess the contribution of individual residues of the parent peptide to substrate activity. To further characterize the enzymological property of the reconstituted kinase, the phosphorylation reactions of these peptides by the reconstituted kinase were examined. The kinetic parameters were determined for these reactions and compared to those of the bovine brain enzyme-catalyzed reactions. Table II shows that the reconstituted kinase displayed very similar, if not identical, Kmvalues for these peptides as the enzyme purified from brain.
      Table IIKinetic parameters of NCLK purified from bovine brain and the reconstituted kinase
      The kinetic parameters were determined as detailed under “Materials and Methods” with the purified enzymes. Those parameters for the kinase purified from bovine brain were determined by Beaudette et al. (53).

      The Activation of Cdk5 by Reconstitution with p21 Is Independent of Cdk5 Phosphorylation

      The activation of Cdc2 by cyclin B shows an absolute dependence on the phosphorylation of Thr161 of Cdc2 by Cdk activating kinase (
      • Desai D.
      • Gu Y.
      • Morgan D.O.
      ,
      • Solomon M.J.
      • Lee T.
      • Kirschner M.W.
      ,
      • Connell-Crowley L.
      • Solomon M.J.
      • Wei N.
      • Harper J.W.
      ). Although Cdk2 exhibits some activity upon activation by cyclin A, the activity can be greatly increased by the Cdk activating kinase-catalyzed phosphorylation of the protein at Thr160(
      • Connell-Crowley L.
      • Solomon M.J.
      • Wei N.
      • Harper J.W.
      ). Sequence alignment of Cdk5 with Cdc2 and Cdk2 has shown that the residue corresponding to Thr161/Thr160 of Cdc2/Cdk2 is substituted in Cdk5 by a serine, Ser159. Since sequence surrounding Ser159 is highly conserved in Cdk5 in comparison to those of Cdc2 and Cdk2, it has been suggested that Cdk5 may also be regulated by Cdk activating kinase (
      • Lew J.
      • Winkfein R.J.
      • Paudel H.K.
      • Wang J.H.
      ). The isolation of the highly active, reconstituted Cdk5-p21 complex, however, appears to argue against such a suggestion.
      While the reconstitution of highly active kinase from bacteria-expressed protein components may rule out the involvement of an activating kinase in the enzyme reconstitution, the possibility that Ser159 phosphorylation is required in the enzyme activation cannot be ruled out completely. Thus, the enzyme may undergo autophosphorylation at this serine residue during reconstitution to result in increased kinase activity. Since the time dependence of Cdk5 activation by p21 can be readily monitored (see Fig. 2C), one approach to test the autoactivation mechanism is to examine the effect of Mg2+ATP in the preincubation on the kinase activation reaction. Fig. 5 A shows that the time course of the activation of recombinant Cdk5 by p21 was not significantly affected by Mg2+ATP. The activation of brain Cdk5 monomer by the recombinant p21 appears to be depressed somewhat by the addition of Mg2+ATP in the activation incubation (Fig. 5 B). The reason for this is not clear. In any case, the results of Fig. 5 argue against the notion that serine 159 of Cdk5 is phosphorylated by an autophosphorylation reaction during the reconstitution reaction.
      Figure thumbnail gr5
      Figure 5:Mg2+ATP effect on Cdk5/p21 reconstitution and kinase activation. A, time dependence of Cdk5/p21 kinase reconstitution. GST-Cdk5 (20 μg) and GST-p21 (40 μg) were mixed in a volume of 300 μl in PBS buffer containing 1 mM EDTA, 1 mM DTT, and 0.5 mg/ml bovine serum albumin. The preincubation was performed at 30°C with (●–●) or without (○–○) Mg2+ATP at a concentration of 100 μM. Aliquots (30 μl) were removed at indicated intervals for the measurement of protein kinase activity. Protein kinase reaction time was 5 min. B, dose-dependent activation of the isolated native Cdk5 by GST-p21. 7.5 μg of the Superose 12 column pooled sample containing monomeric Cdk5 from brain extract (see “Materials and Methods”) was reconstituted with various amounts of GST-p21 at 30°C for 1 h with (●–●) or without (○–○) 100 μM Mg2+ATP. There was GST protein instead of GST-p21 in the control tubes (∆–∆). Histone H1 peptide phosphorylation assays were performed at 30°C for 15 min as described under “Materials and Methods.”
      To probe further whether or not the reconstituted kinase could undergo autophosphorylation on Cdk5, a sample of the purified, reconstituted kinase was incubated with [γ-32P]ATP and Mg2+ at 30°C, aliquots of the sample were withdrawn at various intervals and analyzed by SDS-PAGE and autoradiography. Fig. 6 A shows that while p21 was rapidly phosphorylated, no radioactivity was seen to associate with Cdk5 even after prolonged incubation. Similar results were also obtained with a partially purified kinase reconstituted from GST-p21 and bovine brain Cdk5 (Fig. 6 B). The possibility that autophosphorylation on Cdk5 could occur during the kinase reaction has also been examined and ruled out. Fig. 6, A and B, shows that kinase samples taken from kinase reactions were autophosphorylated on the 21-kDa subunit but not on Cdk5. The ability of the reconstituted kinase to autocatalyze the phosphorylation of p21 is compatible with results of a previous study showing that the regulatory subunit of bovine brain neuronal Cdc2-like kinase is phosphorylated under autophosphorylation conditions (
      • Lew J.
      • Beaudette K.
      • Litwin C.M.E.
      • Wang J.H.
      ). The phosphorylation of p21 does not appear to affect the activity of the enzyme. As shown in Fig. 6 C, the phosphorylation of p21 in the kinase reaction was significantly suppressed in the presence of the substrate peptide. While the extent of p21 phosphorylation decreased with the increase in the peptide substrate concentration, there was no indication of substrate-dependent activation or inactivation for this enzyme.
      Figure thumbnail gr6
      Figure 6:Autophosphorylation of the reconstituted Cdk5/p21 kinase. Time course of the autophosphorylation reaction is shown in A and B. A, aliquots (1 μg) of the reconstituted enzyme purified by Mono S column from the bacterially expressed proteins were incubated at 30°C with 100 μM [γ-32P]ATP (~5000 dpm/pmol) in 20 mM MOPS (pH 7.4), 5 mM MgCl2 for different intervals as indicated. B, 0.4 mg of GST-p21 and 10 mg of the brain Cdk5 sample partially purified by Mono S and Superose 12 columns were mixed and incubated at 30°C for 1 h. The mixture was then mixed with the GSH-agarose beads at 4°C for 30 min. After the beads were washed intensely, Cdk5/GST-p21 was released by 5 mM glutathione in 50 mM Tris (pH 8.0) and 1 mM DTT. Aliquots containing 2.6 μg of the affinity-purified Cdk5/GST-p21 were incubated at 30°C with 20 μM [γ-32P]ATP (~5000 dpm/pmol) in 20 mM MOPS (pH 7.4) and 5 mM MgCl2 for different intervals as indicated. The immunoblots were done with the anti-Cdk5 (I) and anti-p21 (II) antibodies individually. C, substrate effect on autophosphorylation. Aliquots (0.4 μg) of the reconstituted enzyme which was purified with the bacterially expressed Cdk5 and p21 were incubated at 30°C for 5 min in the same condition as above with substrate peptide HS(1-18) of 0 μM, 50 μM, 100 μM, 500 μM, and 1000 μM. Reactions were stopped with SDS-PAGE sample buffer, boiled at 95°C for 5 min, and then analyzed by SDS-PAGE. Phosphorylated proteins were visualized by an autoradiogram which was carried out at −70°C for 3 h with intensifying screens.
      To further substantiate the suggestion that the activation of Cdk5 by p21 does not depend on the phosphorylation of Cdk5, the possibility, albeit remote, that the bacterially expressed Cdk5 used for reconstitution had already been phosphorylated was tested by examining the effect of protein phosphatase on the reconstituted kinase. The protein phosphatase 2A catalytic subunit, which had been shown to inactivate and dephosphorylate the isolated p34cdc2-cyclin complex previously (
      • Lee T.H.
      • Solomon M.J.
      • Mumby M.C.
      • Kirschner M.W.
      ) , was applied in such experiments. The bacterially expressed GST-Cdk5 and GST-p25 were mixed and reconstituted at 30°C for 1 h in the presence of 0.5 mg/ml bovine serum albumin. 20-μl aliquots containing 1 μg of GST-Cdk5 and 1.5 μg of GST-p21 were then treated with 20 ng of protein phosphatase 2A catalytic subunit at 30°C for 30 min. Okadaic acid (100 nM in final concentration) was then added to the mixture to stop the phosphatase activity before the kinase activity was measured. The kinase reaction was carried out for a short interval, 6 min, to minimize the possibility of autophosphorylation during the reaction. It was found that the kinase activity of the phosphatase-treated sample was identical with that of the control samples which had been treated in the same way except that okadaic acid was added before the addition of the phosphatase.

      DISCUSSION

      When neuronal Cdc2-like kinase was first purified and shown to be a heterodimer of a Cdc2-related 33-kDa and a 25-kDa subunit (
      • Lew J.
      • Beaudette K.
      • Litwin C.M.E.
      • Wang J.H.
      ,
      • Lew J.
      • Winkfein R.J.
      • Paudel H.K.
      • Wang J.H.
      ,
      • Ishiguro K.
      • Kobayashi S.
      • Omori A.
      • Takamatsu M.
      • Yonekura S.
      • Anzai K.
      • Imahori K.
      • Uchida T.
      ), it was speculated that the 25-kDa protein played the role of an essential activating subunit. This suggestion is supported by the recent observation that histone H1 kinase activity could be obtained from mixing a bacteria-expressed truncated form of the 25-kDa protein (p21) with a bacteria-expressed Cdk5 (
      • Lew J.
      • Huang Q.-Q.
      • Qi Z.
      • Winkfein R.J.
      • Aebersold R.
      • Hunt T.
      • Wang J.H.
      ). The reconstituted kinase, however, displayed very low specific kinase activity, about 2 orders lower than that of the neuronal Cdc2-like kinase purified from bovine brain. In this study, we have extended the previous observation and carried out detailed characterization of the kinase reconstitution reaction. The results have shown that the main reason for the lower-than-expected kinase activity may be attributed to the fact that most of the bacteria-expressed proteins are present in the degraded and/or incorrectly folded forms. While this finding is hardly surprising, we have succeeded, by removing the inactive protein derivatives, in the purification of a highly active reconstituted kinase which has similar specific activity, as well as similar phosphorylation site specificity as the enzyme purified from brain. This has provided a convenient source for neuronal Cdc2-like kinase. The amount of reconstituted kinase obtained from a typical preparation with 1 liter each of the bacteria culture of the protein components is equivalent to 2 preparations of the brain enzyme, each with 500 g of tissue. Purification of the enzyme from bovine brain routinely requires about 8 days to complete.
      Three truncated forms of p35, p21, p23, and p25, have been expressed as GST fusion proteins in E. coli and tested as Cdk5 activator. p25 has the amino acid sequence corresponding to the 25-kDa subunit of the purified bovine brain neuronal Cdc2-like kinase. The facts that all these p35 derivatives showed similar Cdk5 activation activity and that the purified neuronal Cdc2-like kinase containing the 25-kDa subunit is highly active suggest strongly that all the derivatives contain complete Cdk5-activating domain. Although the purified reconstituted Cdk5 kinase used for most of the detailed characterizations was the Cdk5-p21 heterodimer, other reconstituted kinases, Cdk5-p23 and Cdk5-p25, are expected to have similar general characteristics. Similarly, although the recombinant and bacterially expressed Cdk5 was used for the reconstitution of the highly active Cdk5 kinase, Cdk5 monomer isolated from bovine brain could also be activated by the truncated forms of p35.
      The best characterized members of the Cdc2-like kinase family, Cdc2/cyclin B and Cdk2/cyclin A have been shown by many investigators to depend on the phosphorylation of Thr161/Thr160 of Cdc2/Cdk2 by a distinct kinase, Cdk activating kinase, for full kinase activity (
      • Fesquet D.
      • Labbe J.C.
      • Derancourt J.
      • Capony J.P.
      • Galas S.
      • Girard F.
      • Lorca T.
      • Shuttleworth J.
      • Doree M.
      • Cavadore J.C.
      ,
      • Poon R.Y.C.
      • Yamashita K.
      • Adamczewski J.P.
      • Hunt T.
      • Shuttleworth J.
      ,
      • Solomon M.J.
      • Harper J.W.
      • Shuttleworth J.
      ,
      • Fisher R.P.
      • Morgan D.O.
      ,
      • Makela T.P.
      • Tassan J.-P.
      • Nigg E.A.
      • Frutiger S.
      • Hughes G.J.
      • Weinberg R.A.
      ,
      • Connell-Crowley L.
      • Solomon M.J.
      • Wei N.
      • Harper J.W.
      ). More recently, the kinase, Cdk4/cyclin D has also been documented to depend on an activating kinase for full activity (
      • Kato J.-Y.
      • Matsuoka M.
      • Strom D.K.
      • Sherr C.J.
      ,
      • Matsuoka M.
      • Kato J.-Y.
      • Fisher R.P.
      • Morgan D.O.
      • Sherr C.J.
      ). Although the amino acid residue in Cdk5 corresponding to Thr161/Thr160 of Cdc2/Cdk2, Ser159, is also a potential kinase phosphorylation residue, several observations of the present study suggest that the activity of neuronal Cdc2-like kinase does not depend on phosphorylation of Cdk5. A highly active kinase could be isolated from the reconstitution reaction of Cdk5 and p21 without requiring any additional kinase. The enzyme reconstitution and Cdk5 activation were not affected by Mg2+ATP. Moreover, no phosphorylation of Cdk5 could be detected when the highly active, reconstituted enzyme was incubated under phosphorylation conditions, ruling out the possibility that Cdk5 is phosphorylated at Ser159 by an autocatalytic reaction. Similar results were obtained with either the bacterially expressed recombinant Cdk5 or the partially purified brain monomeric Cdk5 indicating that the difference in the mechanism of Cdk5 and Cdc2/Cdk2 activation cannot be attributed to an artifact arising from the use of bacterially expressed Cdk5. Since we were unable to express the full-length p35 to test how it might activate Cdk5, the possibility that the activation of Cdk5 by the full-length protein is dependent on the phosphorylation of Cdk5 at Ser159 cannot be completely ruled out. In addition, our results do not exclude the possibility that the highly active, reconstituted neuronal Cdc2-like kinase can be further activated by a Cdk activating kinase-like kinase.
      The recently elucidated crystal structure of Cdk2 shows that Thr160 is in a large loop, T-loop, expanding residues 152-170 (
      • DeBondt H.L.
      • Rosenblatt J.
      • Jancarik J.
      • Jones H.D.
      • Morgan D.O.
      • Kim S.H.
      ). Comparing the structure of Cdk2, an inactive form of kinase, with the crystal structure of the catalytic unit of cAMP-dependent protein kinase, a fully active kinase has revealed that T-loop is a region of major conformational difference between the two structures. Like that in Cdc2 or Cdk2, this region of cAMP-dependent kinase contains a regulatory phosphorylation residue, Thr199; autophosphorylation of the enzyme at this site brings about ionic interactions between the phosphothreonine residue with a number of cationic residues: His87, Arg165, and Lys189 to stabilize the active conformation. It is suggested that cyclin activation of Cdk2 involves a displacement of the loop to result in a conformation similar to cAMP-dependent kinase. Phosphorylation of Thr160 may bring about ionic interactions to stabilize the active conformation. Two of the three cationic amino acids in cAMP-dependent kinase that undergo ionic interaction with P-Thr199 are conserved in Cdk2. The observation that Cdk5-21-kDa protein displays high kinase activity in the absence of phosphorylation of Cdk5 is somewhat unexpected. Amino acid sequence of Cdk5 in the T-loop region is highly homologous to those of Cdc2 and Cdk2, and the two arginines postulated to be involved in interacting with P-Thr160 of Cdk2 are also conserved in Cdk5.
      It is tempting to suggest that the unique, phosphorylation-independent kinase activation is due to unique structural properties of the regulatory component of the protein complex, as p35 is only remotely related to cyclins. This suggestion, unfortunately, is difficult to test, since Cdk5 has not been shown to be activated by any cyclins. The fact that p35 expression is restricted to neurons of central nervous systems, whereas Cdk5 is widely distributed among human tissues (
      • Lew J.
      • Huang Q.-Q.
      • Qi Z.
      • Winkfein R.J.
      • Aebersold R.
      • Hunt T.
      • Wang J.H.
      ), suggests that there might be other protein activators of Cdk5. It is significant to determine whether these protein activators activate Cdk5 in a phosphorylation-dependent or independent manner. Similarly, if further studies uncover other Cdc2-like kinases which are activated by p35, it will be important to determine whether those kinase activations are phosphorylation-dependent or independent. In any case, detailed structural information about the Cdk5-p21 protein complex may shed important new light on the molecular basis of the phosphorylation-dependent activation of Cdc2-like kinases. In addition, Cdk5-associated activity, measured as histone H1 kinase, has been found only in brain extract, thus raising the possibility that the kinase specificity of Cdk5 is dictated by its regulatory partner.

      Acknowledgments

      We thank Drs. Li-Huei Tsai and Ed Harlow (Massachusetts General Hospital) for providing the GST-Cdk5 plasmid, Dr. Tim Hunt (ICRF Clare Hall Laboratories, United Kingdom) for the GST-cyclin D1 and hexahistidine-tagged cyclin A plasmids, and Dr. Anna A. DePaoli-Roach (Indiana University School of Medicine) for phosphatase 2A.

      REFERENCES

        • Murray A.W.
        • Kirschner M.W.
        Science. 1989; 246: 614-621
        • Hunt T.
        Curr. Opin. Cell Biol. 1989; 1: 268-274
        • Draetta G.
        Trends Biochem. Sci. 1990; 15: 378-383
        • Nurse P.
        Nature. 1990; 344: 503-508
        • Pines J.
        • Hunter T.
        New Biol. 1990; 2: 389-401
        • Maller J.L.
        Curr. Opin. Cell Biol. 1991; 3: 269-275
        • Norbury C.
        • Nurse P.
        Annu. Rev. Biochem. 1992; 61: 441-470
        • Nash R.
        • Tokiwa G.
        • Anand S.
        • Erickson K.
        • Futcher A.B.
        EMBO J. 1988; 7: 4335-4346
        • Hadwiger J.A.
        • Wittenberg C.
        • Richardson H.
        • DeBaros Lopes M.
        • Reed S.I.
        Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 6255-6259
        • Fantes P.
        Curr. Opin. Cell Biol. 1989; 1: 250-255
        • Fang F.
        • Newport J.W.
        Cell. 1991; 66: 731-742
        • Girard F.
        • Strausfeld U.
        • Fernandez A.
        • Lamb N.J.C.
        Cell. 1991; 67: 1169-1179
        • Pagano M.
        • Pepperkok R.
        • Verde F.
        • Ansorge W.
        • Draetta G.
        EMBO J. 1992; 11: 961-971
        • Rosenblatt J.
        • Gu Y.
        • Morgan D.O.
        Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 2824-2828
        • Koff A.
        • Giordano A.
        • Dipty D.
        • Yamashita K.
        • Harper J.W.
        • Elledge S.
        • Nishimoto T.
        • Morgan D.O.
        • Franza B.R.
        • Roberts J.M.
        Science. 1992; 257: 1689-1694
        • Tsai L.-H.
        • Lees E.
        • Harlow E.
        • Riabowol K.
        Oncogene. 1993; 8: 1593-1602
        • Lee M.G.
        • Nurse P.
        Nature. 1987; 327: 31-35
        • Hunter T.
        • Pines J.
        Cell. 1991; 66: 1071-1074
        • Meyerson M.
        • Enders G.H.
        • Wu C.-L.
        • Su L.-K.
        • Gorka C.
        • Nelson C.
        • Harlow E.
        • Tsai L.-H.
        EMBO J. 1992; 11: 2909-2917
        • Sherr C.J.
        Cell. 1993; 73: 1059-1065
        • Soloman M.J.
        • Glotzer M.
        • Lee T.H.
        • Philippe M.
        • Kirschner M.W.
        Cell. 1990; 63: 1013-1024
        • Xiong Y.
        • Zhang H.
        • Beach D.
        Cell. 1993; 71: 505-514
        • Pines J.
        Trends Biochem. Sci. 1993; 18: 195-197
        • Soloman M.J.
        Curr. Opin. Cell Biol. 1993; 5: 180-186
        • Gould K.L.
        • Nurse P.
        Nature. 1989; 342: 39-45
        • Krek W.
        • Nigg E.A.
        EMBO J. 1991; 10: 305-316
        • Krek W.
        • Nigg E.A.
        EMBO J. 1991; 10: 3331-3341
        • Ducommun B.
        • Brambilla P.
        • Felix M.A.
        • Franza B.J.
        • Karsenti E.
        • Draetta G.
        EMBO J. 1991; 10: 3311-3319
        • Norbury C.
        • Blow J.
        • Nurse P.
        EMBO J. 1991; 10: 3321-3329
        • Desai D.
        • Gu Y.
        • Morgan D.O.
        Mol. Biol. Cell. 1992; 3: 571-582
        • Solomon M.J.
        • Lee T.
        • Kirschner M.W.
        Mol. Biol. Cell. 1992; 3: 13-27
        • Fesquet D.
        • Labbe J.C.
        • Derancourt J.
        • Capony J.P.
        • Galas S.
        • Girard F.
        • Lorca T.
        • Shuttleworth J.
        • Doree M.
        • Cavadore J.C.
        EMBO J. 1993; 12: 3111-3121
        • Poon R.Y.C.
        • Yamashita K.
        • Adamczewski J.P.
        • Hunt T.
        • Shuttleworth J.
        EMBO J. 1993; 12: 3123-3132
        • Solomon M.J.
        • Harper J.W.
        • Shuttleworth J.
        EMBO J. 1993; 12: 3133-3142
        • Fisher R.P.
        • Morgan D.O.
        Cell. 1994; 78: 713-724
        • Makela T.P.
        • Tassan J.-P.
        • Nigg E.A.
        • Frutiger S.
        • Hughes G.J.
        • Weinberg R.A.
        Nature. 1994; 371: 254-257
        • Russell P.
        • Nurse P.
        Cell. 1987; 49: 559-567
        • Featherstone C.
        • Russell P.
        Nature. 1991; 349: 808-811
        • Parker L.L.
        • Atherton-Fessler S.
        • Lee M.S.
        • Ogg S.
        • Falk J.L.
        • Swenson K.I.
        • Piwnica-Worms H.
        EMBO J. 1991; 10: 1255-1263
        • Parker L.L.
        • Piwnica-Worms H.
        Science. 1992; 257: 1955-1957
        • Kato J.-Y.
        • Matsuoka M.
        • Strom D.K.
        • Sherr C.J.
        Mol. Cell. Biol. 1994; 14: 2713-2721
        • Matsuoka M.
        • Kato J.-Y.
        • Fisher R.P.
        • Morgan D.O.
        • Sherr C.J.
        Mol. Cell. Biol. 1994; 14: 7265-7275
        • Draetta G.
        • Beach D.
        • Moran E.
        Oncogene. 1988; 2: 553-557
        • Okuda T.
        • Cleveland J.L.
        • Downing J.R.
        Oncogene. 1992; 7: 2249-2258
        • Hellmich M.R.
        • Pant H.C.
        • Wada E.
        • Battey J.F.
        Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 10867-10871
        • Shetty K.T.
        • Link W.T.
        • Pant H.C.
        Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 6844-6848
        • Paudel H.K.
        • Lew J.
        • Ali Z.
        • Wang J.H.
        J. Biol. Chem. 1993; 268: 23512-23518
        • Lew J.
        • Beaudette K.
        • Litwin C.M.E.
        • Wang J.H.
        J. Biol. Chem. 1992; 267: 13383-13390
        • Lew J.
        • Winkfein R.J.
        • Paudel H.K.
        • Wang J.H.
        J. Biol. Chem. 1992; 267: 25922-25926
        • Ishiguro K.
        • Kobayashi S.
        • Omori A.
        • Takamatsu M.
        • Yonekura S.
        • Anzai K.
        • Imahori K.
        • Uchida T.
        FEBS Lett. 1994; 342: 203-208
        • Lew J.
        • Huang Q.-Q.
        • Qi Z.
        • Winkfein R.J.
        • Aebersold R.
        • Hunt T.
        • Wang J.H.
        Nature. 1994; 371: 423-426
        • Smith D.B.
        • Johnson K.S.
        Gene (Amst.). 1988; 67: 31-40
        • Beaudette K.N.
        • Lew J.
        • Wang J.H.
        J. Biol. Chem. 1993; 268: 20825-20830
        • Bradford M.M.
        Anal. Biochem. 1976; 72: 248-254
        • Laemmli U.K.
        Nature. 1970; 227: 680-685
        • Wray W.
        • Boulikas T.
        • Wray V.
        • Hankcock R.
        Anal. Biochem. 1981; 118: 197-203
        • Connell-Crowley L.
        • Solomon M.J.
        • Wei N.
        • Harper J.W.
        Mol. Biol. Cell. 1993; 4: 79-92
        • DeBondt H.L.
        • Rosenblatt J.
        • Jancarik J.
        • Jones H.D.
        • Morgan D.O.
        • Kim S.H.
        Nature. 1993; 363: 595-602
        • Lee T.H.
        • Solomon M.J.
        • Mumby M.C.
        • Kirschner M.W.
        Cell. 1991; 64: 415-423