Advertisement

Protein Kinase CK2 Is an Inhibitor of the Neuronal Cdk5 Kinase*

  • Anthony C.B. Lim
    Affiliations
    Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609
    Search for articles by this author
  • Zhibo Hou
    Affiliations
    Department of Biochemistry, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China
    Search for articles by this author
  • Choon-Peng Goh
    Affiliations
    Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609
    Search for articles by this author
  • Robert Z. Qi
    Correspondence
    To whom correspondence should be addressed. Tel.: 852-2358-7273; Fax: 852-2358-1552;
    Affiliations
    Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609

    Department of Biochemistry, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China
    Search for articles by this author
  • Author Footnotes
    * This work was supported in parts by the Research Grants Council of Hong Kong (HKUST6128/04M), the Area of Excellence Scheme established under the University Grants Committee of Hong Kong (AoE/B-15/01), and the Agency for Science, Technology and Research of Singapore. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Open AccessPublished:September 01, 2004DOI:https://doi.org/10.1074/jbc.M404760200
      The complex of Cdk5 and its neuronal activator p35 is a proline-directed Ser/Thr kinase that plays an important role in various neuronal functions. Deregulation of the Cdk5 enzymatic activity was found to associate with a number of neurodegenerative diseases. To search for regulatory factors of Cdk5-p35 in the brain, we developed biochemical affinity isolation using a recombinant protein comprising the N-terminal 149 amino acids of p35. The catalytic α-subunit of protein kinase CK2 (formerly known as casein kinase 2) was identified by mass spectrometry from the isolation. The association of CK2 with p35 and Cdk5 was demonstrated, and the CK2-binding sites were delineated in p35. Furthermore, CK2 displayed strong inhibition toward the Cdk5 activation by p35. The Cdk5 inhibition is dissociated from the kinase function of CK2 because the kinase-dead mutant of CK2 displayed the similar Cdk5 inhibitory activity as the wild-type enzyme. Further characterization showed that CK2 blocks the complex formation of Cdk5 and p35. Together, these findings suggest that CK2 acts as an inhibitor of Cdk5 in the brain.
      Cdk5 was identified independently on the basis of its sequence similarity to the family of Cdks, its interaction with cyclin D, and its protein kinase activity toward a proline-directed Ser/Thr sequence (
      • Meyerson M.
      • Enders G.H.
      • Wu C.L.
      • Su L.K.
      • Gorka C.
      • Nelson C.
      • Harlow E.
      • Tsai L.H.
      ,
      • Lew J.
      • Beaudette K.
      • Litwin C.M.
      • Wang J.H.
      ,
      • Xiong Y.
      • Zhang H.
      • Beach D.
      ). Despite having 60% sequence identity with Cdk1 and Cdk2, a role for Cdk5 in cell cycle regulation has yet to be identified. Nevertheless, Cdk5 has been implicated in the regulation of neuronal differentiation, degeneration, and cytoskeletal dynamics (
      • Dhavan R.
      • Tsai L.H.
      ). Monomeric Cdk5 shows no enzymatic activity because its activation is dependent on its association with the regulatory subunit p35 or p39, neither of which is a cyclin protein (
      • Lew J.
      • Huang Q.Q.
      • Qi Z.
      • Winkfein R.J.
      • Aebersold R.
      • Hunt T.
      • Wang J.H.
      ,
      • Tang D.
      • Yeung J.
      • Lee K.Y.
      • Matsushita M.
      • Matsui H.
      • Tomizawa K.
      • Hatase O.
      • Wang J.H.
      ,
      • Tsai L.H.
      • Delalle I.
      • Caviness V.S.J.
      • Chae T.
      • Harlow E.
      ). Mammalian p35 and p39 are primarily expressed in post-mitotic neurons, thereby restricting Cdk5 activity predominantly to the nervous system (
      • Ko J.
      • Humbert S.
      • Bronson R.T.
      • Takahashi S.
      • Kulkarni A.B.
      • Li E.
      • Tsai L.H.
      ). Cdk5 and p35 are essential for proper brain development. Mice deficient of Cdk5 or p35 display cell-positioning defects in the cerebral cortex (
      • Chae T.
      • Kwon Y.T.
      • Bronson R.
      • Dikkes P.
      • Li E.
      • Tsai L.H.
      ,
      • Ohshima T.
      • Ward J.M.
      • Huh C.G.
      • Longenecker G.
      • Veeranna Pant H.C.
      • Brady R.O.
      • Martin L.J.
      • Kulkarni A.B.
      ). In the brain, Cdk5-p35 phosphorylates a number of cytoskeletal proteins that are thought to play important roles in the reassembly of cytoskeletal elements, thereby mediating neurite outgrowth and neuronal migration during the development (
      • Baumann K.
      • Mandelkow E.M.
      • Biernat J.
      • Piwnica-Worms H.
      • Mandelkow E.
      ,
      • Grant P.
      • Sharma P.
      • Pant H.C.
      ,
      • Ackerley S.
      • Thornhill P.
      • Grierson A.J.
      • Brownlees J.
      • Anderton B.H.
      • Leigh P.N.
      • Shaw C.E.
      • Miller C.C.
      ,
      • Rashid T.
      • Banerjee M.
      • Nikolic M.
      ,
      • Nikolic M.
      • Chou M.M.
      • Lu W.
      • Mayer B.J.
      • Tsai L.H.
      ,
      • Xie Z.
      • Sanada K.
      • Samuels B.A.
      • Shih H.
      • Tsai L.H.
      ). In addition, there are several lines of evidence implicating aberrant regulation of Cdk5 in neurodegeneration and cell death (
      • Lim A.C.
      • Qi R.Z.
      ). Indeed, p25, a truncated C-terminal fragment of p35, was found to accumulate in Alzheimer's disease brains; its associated-Cdk5 kinase activity was shown to lead to cytoskeletal disruption, morphological degeneration, and apoptosis (
      • Lee M.S.
      • Kwon Y.T.
      • Li M.
      • Peng J.
      • Friedlander R.M.
      • Tsai L.H.
      ,
      • Ahlijanian M.K.
      • Barrezueta N.X.
      • Williams R.D.
      • Jakowski A.
      • Kowsz K.P.
      • McCarthy S.
      • Coskran T.
      • Carlo A.
      • Seymour P.A.
      • Burkhardt J.E.
      • Nelson R.B.
      • McNeish J.D.
      ,
      • Noble W.
      • Olm V.
      • Takata K.
      • Casey E.
      • Mary O.
      • Meyerson J.
      • Gaynor K.
      • LaFrancois J.
      • Wang L.
      • Kondo T.
      • Davies P.
      • Burns M.
      • Veeranna Nixon R.
      • Dickson D.
      • Matsuoka Y.
      • Ahlijanian M.
      • Lau L.F.
      • Duff K.
      ,
      • Cruz J.C.
      • Tseng H.C.
      • Goldman J.A.
      • Shih H.
      • Tsai L.H.
      ).
      Cdk5 is involved in the regulation of many vital functions in the brain and, hence, its activity needs to be tightly controlled in the cells. Upon association with its activator p35 or p39, Cdk5 is regulated through various means, and many of its regulatory properties are distinct from those of the authentic Cdk-cyclin. First, p35 is an unstable protein with a half-life of 20–30 min during which it is multi-ubiquitylated and degrades through the ubiquitin-proteosome pathway (
      • Patrick G.N.
      • Zhou P.
      • Kwon Y.T.
      • Howley P.M.
      • Tsai L.H.
      ,
      • Saito T.
      • Onuki R.
      • Fujita Y.
      • Kusakawa G.
      • Ishiguro K.
      • Bibb J.A.
      • Kishimoto T.
      • Hisanaga S.
      ). Secondly, in contrast to Cdk1 and Cdk2, the Cdk5 phosphorylation at Tyr-15, which is catalyzed by c-Abl and Fyn, is stimulatory but not inhibitory to Cdk5 kinase activity (
      • Zukerberg L.R.
      • Patrick G.N.
      • Nikolic M.
      • Humbert S.
      • Wu C.L.
      • Lanier L.M.
      • Gertler F.B.
      • Vidal M.
      • Van Etten R.A.
      • Tsai L.H.
      ,
      • Sasaki Y.
      • Cheng C.
      • Uchida Y.
      • Nakajima O.
      • Ohshima T.
      • Yagi T.
      • Taniguchi M.
      • Nakayama T.
      • Kishida R.
      • Kudo Y.
      • Ohno S.
      • Nakamura F.
      • Goshima Y.
      ). Next, the KIP and INK families of Cdk inhibitors do not show any activity toward Cdk5 (
      • Lee M.H.
      • Nikolic M.
      • Baptista C.A.
      • Lai E.
      • Tsai L.H.
      • Massague J.
      ). Biochemical studies of Cdk5 in a bovine brain extract suggested the existence of inhibitory factor(s) together with Cdk5 and p35 in macromolecular protein complexes (
      • Lee K.Y.
      • Rosales J.L.
      • Tang D.
      • Wang J.H.
      ). More recently, protein C42, which was isolated in a yeast two-hybrid search, was reported as a down-regulator of Cdk5 activity (
      • Ching Y.P.
      • Pang A.S.
      • Lam W.H.
      • Qi R.Z.
      • Wang J.H.
      ).
      A growing body of evidence has implicated Cdk5-p35 as a multifunctional enzyme that exists in many protein complexes in the cells (
      • Lim A.C.
      • Qu D.
      • Qi R.Z.
      ). To learn more about the functional and regulatory properties of this kinase, a recombinant p35 fragment was utilized in the affinity isolation of p35-binding proteins from rat brain homogenates. Isolated proteins were identified by tandem mass spectrometry. Here, we present the catalytic α-subunit of protein kinase CK2 as one of the proteins isolated. CK2α is able to interact physically with p35 and Cdk5. Using deletion mutants, CK2 is shown to bind to two regions located separately at the N- and C-terminal halves of p35. Moreover, the binding of CK2 to p35 or Cdk5 blocks the interaction between p35 and Cdk5, thereby inhibiting the p35-associated activation of Cdk5.

      EXPERIMENTAL PROCEDURES

      Plasmid Constructs and Recombinant Proteins—Various fragments of p35, namely p10 (amino acids 1–98 of p35), p16 (amino acids 1–149 of p35), p35(53–149) (amino acids 53–149 of p35), p35(150–307) (amino acids 150–307 of p35), and p25 (amino acids 99–307 of p35), were engineered into pGEX vectors by polymerase chain reactions. CK2α and its kinase-dead mutant CK2αK68A were cloned into pET32a, and human CK2β was cloned into pQE30 (
      • Lim A.C.
      • Tiu S.Y.
      • Li Q.
      • Qi R.Z.
      ).
      Recombinant proteins tagged with GST
      The abbreviations used are: GST, glutathione S-transferase; MOPS, 4-morpholinepropanesulfonic acid; siRNA, small interfering RNA.
      or His6 were expressed in bacteria and purified as described in earlier reports (
      • Lim A.C.
      • Tiu S.Y.
      • Li Q.
      • Qi R.Z.
      ,
      • Qu D.
      • Li Q.
      • Lim H.Y.
      • Cheung N.S.
      • Li R.
      • Wang J.H.
      • Qi R.Z.
      ). Escherichia coli BL21(DE3) was used for protein expression from pGEX and pET32a constructs, and E. coli M15 was the expression host for the pQE30 construct. To purify GST proteins, bacteria were lysed by sonication in Buffer A (50 mm Tris-HCl, pH 7.4, 150 mm NaCl, 1% Triton X-100, 1 mm EDTA, 1 mm dithiothreitol, and the Roche Applied Science protease inhibitor mixture) and clarified by centrifugation. After protein binding to GSH-Sepharose (Amersham Biosciences), the beads were washed with Buffer A supplemented with 0.5 m NaCl. GST proteins were eluted by 5 mm GSH in Buffer B (25 mm Tris-HCl, pH 7.4, 100 mm NaCl, 1 mm EDTA and 1 mm dithiothreitol) and then dialyzed against Buffer B to remove GSH. Bacteria expressing His6-tagged proteins were lysed in Buffer C (50 mm sodium phosphate, pH 8.0, 150 mm NaCl, 20 mm imidazole, 10 mm β-mercaptoethanol, and the Roche Applied Science EDTA-free protease inhibitor mixture) supplemented with 1% Triton X-100. After protein binding to a nickel-nitrilotriacetic acid column (Qiagen), unbound proteins were washed off using Buffer C. His6-tagged proteins were eluted with 250 mm imidazole in Buffer C and then dialyzed in Buffer B.
      Isolation and Identification of p35-binding Proteins—The isolation of p35-binding proteins from rat brains was carried out essentially as described in an earlier report with modifications (
      • Qu D.
      • Li Q.
      • Lim H.Y.
      • Cheung N.S.
      • Li R.
      • Wang J.H.
      • Qi R.Z.
      ). Rat brains were homogenized in Buffer D (25 mm Tris-HCl, pH 7.5, 100 mm NaCl, 1 mm EDTA, 1 mm dithiothreitol, 0.1% Nonidet P-40, 1 mm phenylmethylsulfonyl fluoride, 0.1 mm benzamidine, 3 μg/ml leupeptin, and 3 μg/ml pepstatin). After centrifugation, the lysate was incubated at 4 °C for 3 h with GSH-Sepharose pre-coupled with 100 μg of GST-p16 or GST. The beads were washed extensively using Buffer D before the elution of associated proteins by Buffer D plus 1 m NaCl. Eluents were subjected to acetone precipitation and SDS-PAGE analysis. Protein bands that appeared in the GST-p16 pull-down but not in that of GST were excised for in-gel tryptic digestion (
      • Shevchenko A.
      • Wilm M.
      • Vorm O.
      • Mann M.
      ). Recovered peptides were analyzed on a quadrupole/time-of-flight hybrid mass spectrometer (QSTAR-Pulsar, Applied Biosystems/Sciex) equipped with a nanoelectrospray ion source. Protein identities were revealed by interrogating sequence databases with peptide sequence tags generated from tandem mass spectra (
      • Mann M.
      • Wilm M.
      ,
      • Wilm M.
      • Shevchenko A.
      • Houthaeve T.
      • Breit S.
      • Schweigerer L.
      • Fotsis T.
      • Mann M.
      ).
      Binding Assay—5 μg of GST-tagged proteins were incubated with 10 μg of CK2 proteins as indicated in 300 μl of Buffer D for 1 h at room temperature. The GST-tagged proteins were retrieved by further incubation with GSH-Sepharose for 1 h at 4 °C. The beads were washed four times with Buffer D. Bound proteins were released by boiling in the SDS-PAGE sample loading buffer and analyzed by immunoblotting. Antibodies recognizing p35 (C-19), Cdk5 (C-8), CK2α, and CK2β were purchased from Santa Cruz Biotechnology.
      Protein Size Exclusion Chromatography—Rat brains were homogenized in Buffer E (25 mm HEPES, pH 7.3, 150 mm NaCl, 1 mm EDTA, 1 mm dithiothreitol, 5 μg/ml leupeptin, 5 μg/ml aprotinin, and 5 μg/ml antipain). Crude homogenate was centrifuged at 100,000 × g for 1 h, and the supernatant was then collected carefully. Using the fast protein liquid chromatography system (Amersham Biosciences), the lysate containing 10-mg protein was applied onto a Superdex-200 column (HR16/50, Amersham Biosciences) pre-equilibrated in Buffer E. Proteins eluted by Buffer E were collected as 1-ml fractions. The whole procedure was carried out at 4 °C. Proteins in alternate fractions were concentrated by acetone precipitation and subjected to immunoblotting analysis using antibodies as indicated. The Superdex-200 column was calibrated under the same chromatography condition using a molecular mass calibration kit (Amersham Biosciences) containing blue dextran (2,000 kDa), thyroglobulin (670 kDa), ferritin (440 kDa), aldolase (158 kDa), and albumin (67 kDa).
      Cdk5 Kinase Assay—Cdk5 kinase assay was performed according to the method described previously with minor modifications (
      • Qi Z.
      • Huang Q.Q.
      • Lee K.Y.
      • Lew J.
      • Wang J.H.
      ). Briefly, Cdk5 activity was measured by phosphorylating 100 μm histone H1 peptide HS(9–18) (PKTPKKAKKL) in 30 mm MOPS, pH 7.4, 10 mm MgCl2, and 100 μm [γ-32P]ATP (∼400 dpm/pmol) at 30 °C for 10 min. Reactions were terminated with the addition of trichloroacetic acid to 15% in the final volume. The samples were then set on ice for 10 min before being clarified by centrifugation. Supernatants were spotted onto P81 paper squares (Whatman). After the washing of the paper squares, phosphate incorporated into the histone H1 peptide was measured in a scintillation counter.
      Cell Transfection, RNA Interference, and Immunoprecipitation— COS-7 and SH-SY5Y cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Transient transfection was carried out using Effectene (Qiagen) for 24 h. CK2α and CK2β were cloned into the bicistronic vector pBudCE4.1 (Invitrogen) for co-expression of the two proteins. The siRNA sequence used against human CK2α/α′ was described in an earlier report (
      • Lim A.C.
      • Tiu S.Y.
      • Li Q.
      • Qi R.Z.
      ). 20 μm CK2α/α′ siRNA or a scrambled siRNA sequence (Dharmacon) was applied into transfections using the TransIT-TKO reagent (Mirus). After transfection, the cells were further incubated for 48 h. For immunoprecipitation, rat brains or the transfected cells were lysed in Buffer F (25 mm Tris-HCl, pH 7.4, 100 mm NaCl, 1 mm EDTA, 0.5% Triton X-100, and the Roche Applied Science protease inhibitor mixture). After clarification, lysates were incubated overnight with antibodies as indicated at 4 °C before protein A beads (Amersham Biosciences) were added for a further incubation of 3 h at 4°C. The beads were retrieved and washed extensively with Buffer F. Proteins bound on the beads were resolved by SDS-PAGE and immunoblotted using antibodies as indicated. Antibodies against the V5 and c-Myc tags were from Invitrogen and Santa Cruz Biotechnology, respectively.

      RESULTS

      Identification of CK2α as a p35-binding Protein—To uncover regulators and effectors of the Cdk5-p35 kinase, we performed biochemical isolation of p35-interacting proteins from rat brains. Previous studies had implicated the importance of the N-terminal sequence of p35 in the formation of macromolecular complexes in the brain (
      • Lim A.C.
      • Qu D.
      • Qi R.Z.
      ). Therefore, GST-p16, which comprises the N-terminal region next to the Cdk5-binding domain of p35, was constructed to serve as the bait in the isolation. Proteins bound to p16 from the brain extract were released under the high salt condition (1 m NaCl) and subsequently resolved by SDS-PAGE. From several independent pull-downs, a protein band at ∼45 kDa was observed consistently from the GST-p16 pull-downs, but not in those of GST (Fig. 1A). The protein band was excised and subjected to in-gel digestion and mass spectrometric analysis. A search in sequence databases with tandem mass spectrometric data generated from several prominent peptide signals showed that the sequences of these peptides match those from the catalytic α-subunit of rat protein kinase CK2 (data not shown). To confirm this result, samples from a subsequent pull-down were used for immunoblotting analysis. As shown in Fig. 1B, the immunoblot clearly indicated that CK2α was specifically pulled down by GST-p16. Interestingly, a shorter N-terminal fragment of p35, namely GST-p10 (amino acids 1–98 of p35), failed to pull down CK2α from the brain extract (Fig. 1B), implying that the region spanning the residues 99–149 of p35 is necessary for the association with CK2α.
      Figure thumbnail gr1
      Fig. 1Isolation of CK2α as a p35-binding protein from rat brains. A, GST-p16 was used in the pull-downs from rat brain lysate. Proteins co-precipitated from the lysate through GST-p16 or GST were separated by SDS-PAGE. The protein band at ∼45 kDa in the GST-p16 lane was identified by mass spectrometry as rat CK2α. B, Pull-downs of the rat brain lysate using GST, GST-p10, and GST-p16 were analyzed by anti-CK2α immunoblotting.
      CK2α Exists in Complexes with p35 and Cdk5 in the Rat Brain—Rat brain extract was fractionated by size exclusion chromatography using a Superdex-200 column. Fractions were collected and immunoblotted to examine the elution patterns of the Cdk5, p35, and CK2 proteins. Fig. 2A shows the elution profiles derived. Almost all of p35 was found to be eluted in the macromolecular weight fractions, migrating nearly at the void volume of the column. Cdk5 was eluted across a fairly broad range of elution positions, which is characteristic of various Cdk5 complexes present in the brain extract. Substantial amounts of Cdk5 were found in the macromolecular weight fractions containing p35 as well as in the low molecular weight fractions corresponding to the molecular mass of ∼30 kDa, characteristic of its monomeric form. CK2α immunoreactivity was detected in the macromolecular weight fractions of p35 and Cdk5, with the peak staining closely aligned with that of p35 (Fig. 2A). Intriguingly, the macromolecular weight fractions containing Cdk5, p35, and CK2α were completely devoid of CK2β (Fig. 2A), suggesting that CK2α but not the holoenzyme exists with Cdk5 and p35 in large protein complexes in the brain. To further demonstrate the association of CK2α with Cdk5 and p35, co-immunoprecipitation was performed using the rat brain lysate. As shown in Fig. 2B, CK2α was specifically detected in the immunocomplexes resulting from anti-p35 and anti-Cdk5 immunoprecipitations. Indeed, both p35 and Cdk5 were also observed in the CK2α immunoprecipitates (Fig. 2C), providing evidence for the existence of CK2α in the complexes of p35 and Cdk5.
      Figure thumbnail gr2
      Fig. 2CK2α exists in complexes with Cdk5 and p35 in the rat brain. A, rat brain lysate containing 10 mg of protein was analyzed by gel filtration chromatography using a Superdex-200 column. Proteins in alternate fractions were precipitated by acetone and subsequently immunoblotted with antibodies recognizing Cdk5, p35, CK2α, and CK2β. B and C, co-immunoprecipitation of Cdk5 and p35 with CK2α. Rat brain lysate was subjected to immunoprecipitation (IP) with the p35- and Cdk5-specific antibodies (B) and the CK2α-specific antibody (C), followed by immunoblotting of the immunoprecipitates. Nonspecific rabbit IgG was used in the control immunoprecipitation at equal amounts of the specific antibodies.
      Direct Association of CK2 with p35 and Cdk5—We next examined the physical association of CK2 with p35 and Cdk5 using purified recombinant proteins. The CK2 holoenzyme was reconstituted from individual recombinant α- and β-subunits. GST-tagged p35 and Cdk5 were individually incubated with CK2α, CK2β, or the holoenzyme. The GST-tagged proteins were subsequently retrieved using GSH beads to analyze coprecipitated proteins by immunoblotting. CK2α and CK2β immunoblots clearly showed the physical association of p35 with CK2α and the holoenzyme but not with CK2β alone (Fig. 3A). In addition to p35, Cdk5 displayed direct binding to CK2α, although it failed to interact with the CK2 holoenzyme (Fig. 3A). Taken together, p35 is able to associate with the CK2 holoenzyme via its interaction with CK2α, whereas Cdk5 binds only to CK2α.
      Figure thumbnail gr3
      Fig. 3Direct binding of CK2 to p35 and Cdk5. A, CK2 interacts with p35 and Cdk5. 10 μg of GST-p35 or GST-Cdk5 were incubated individually with CK2α (0.5 μg), CK2β (0.5 μg), and the CK2 holoenzyme (1 μg). After retrieval of the GST proteins using GSH beads, bound proteins were immunoblotted using a mixture of the CK2α and CK2β antibodies. B, mapping CK2-binding sites in p35. 10 μg of GST-tagged p35 fragments or GST were incubated with CK2α (0.5 μg). GSH beads were then used to retrieve the GST proteins, and retrieved proteins were analyzed by anti-CK2α immunoblotting. p35 constructs used in the binding assay are illustrated below the immunoblot.
      Upon identifying CK2α as a binding partner of p35, we proceeded to delineate the CK2-binding region(s) in p35. Various fragments of p35 were constructed as shown in Fig. 3B and prepared as recombinant GST-tagged proteins. p16, but none of its further truncated constructs, interacted with CK2α in the binding assays, indicating that further truncations of p16 abolished its CK2-binding capability (Fig. 3B). Unexpectedly, p25 and a truncated C-terminal fragment of p25, namely p35(150–307), exhibited interaction with CK2α, although their bindings were relatively weak (Fig. 3B). In p35, the Cdk5-binding and -activating domain is located at the region spanning amino acids 150–292 (
      • Tang D.
      • Chun A.C.
      • Zhang M.
      • Wang J.H.
      ,
      • Poon R.Y.
      • Lew J.
      • Hunter T.
      ). Therefore, CK2 appeared to interact with two regions of p35 at different affinities; one of them overlaps the Cdk5-binding domain.
      CK2 Inhibits Cdk5 Activation—Cdk5 can be highly activated by reconstitution with p35 or its truncated fragment, p25 (
      • Qi Z.
      • Huang Q.Q.
      • Lee K.Y.
      • Lew J.
      • Wang J.H.
      ). To understand the functionality of the CK2 interaction with p35 and Cdk5, we carried out Cdk5 activation using recombinant proteins and examined potential CK2 effects on Cdk5 kinase activity. Kinase assays showed that the reconstitution of Cdk5 and p35 in the presence of CK2α or the CK2 holoenzyme resulted in inhibition of the Cdk5 activity in a CK2 dose-dependent manner (Fig. 4A). Compared with CK2α, the CK2 holoenzyme exhibited a slightly higher potency in the Cdk5 inhibition (Fig. 4A). Because p25 constitutes the second binding site for CK2, we went on to examine the effect of CK2 on the p25 activation of Cdk5. Similarly, CK2α and the CK2 holoenzyme strongly inhibited the Cdk5 activity when they were included in the reconstitution mixture of Cdk5 and p25 (Fig. 4B). Hence, both the p35 and the p25 activation of Cdk5 can be readily blocked by CK2α or the CK2 holoenzyme.
      Figure thumbnail gr4
      Fig. 4CK2 inhibits Cdk5 activation. A, CK2 inhibits the p35 activation of Cdk5. GST-Cdk5 (0.28 μm) was incubated with GST-p35 (0.27 μm) in the presence of an increasing amount of CK2α or the CK2 holoenzyme. Samples were incubated for 1 h at room temperature before being assayed for Cdk5 kinase activity. B, CK2 inhibits the p25 activation of Cdk5. GST-Cdk5 (0.28 μm) was incubated with GST-p25 (0.33 μm) and CK2α (3.79 μm) or the CK2 holoenzyme (1.19 μm) for 1 h at room temperature. The kinase activity of Cdk5 was then determined and expressed as percentages of that in the absence of the CK2 proteins. Data shown here are averages of three separate experiments.
      To further demonstrate Cdk5 inhibition by CK2, we introduced p35 and CK2 into cultured COS-7 cells by transient transfection. Cdk5 is a ubiquitously expressed protein. Western blots detected expressed p35 and CK2 proteins as well as endogenous Cdk5 in the cells (Fig. 5). p35 was immunoprecipitated to measure its associated Cdk5 activity. When p35 was transfected without CK2, a high kinase activity of Cdk5 was detected from the p35 immunoprecipitate (Fig. 5). In the double transfection of p35 and the construct expressing CK2α or the CK2 holoenzyme (CK2α/CK2β), the p35-associated Cdk5 activity was significantly reduced (Fig. 5). In agreement with the in vitro reconstitution assays, the expression of the CK2 holoenzyme yielded a slightly stronger Cdk5 inhibition compared with that of CK2α (Fig. 5).
      Figure thumbnail gr5
      Fig. 5Cdk5 is inhibited by CK2 in transfected COS-7 cells. COS-7 cells were double transfected with the p35 plasmid and the plasmid expressing CK2α or the holoenzyme CK2α/CK2β (CK2holo). The vector of CK2 was co-transfected with p35 in the control experiment. Expressed CK2α and CK2β proteins contain V5 and c-Myc tags, respectively. Protein expression was detected in the cell lysates by immunoblots against V5, c-Myc, p35, and Cdk5. Following anti-p35 immunoprecipitation, the activity of co-precipitated Cdk5 was determined and expressed as percentages of that in the control sample. The plot is the average of three individual experiments.
      To evaluate the physiological role of CK2 on Cdk5 activity, we knocked down CK2α/α′ in human neuroblastoma SH-SY5Y cells using a siRNA duplex. As shown by cell lysate immunoblots, the introduction of CK2α/α′-specific siRNA led to a significant decrease of the cellular CK2α protein level (Fig. 6A), whereas the protein contents of Cdk5 and p35 were not affected (data not shown). To assess the knockdown effect on Cdk5 activity, p35 was immunoprecipitated, and its associated Cdk5 activity was assayed. The introduction of CK2α/α′-specific siRNA yielded a notably higher Cdk5 activity as compared with that of the control sample (scrambled siRNA sequence) (Fig. 6B), implicating the Cdk5 inhibitory effect of CK2 in the cells.
      Figure thumbnail gr6
      Fig. 6CK2 inhibits Cdk5 activity in neuroblastoma cells. SH-SY5Y cells were introduced with the siRNA of human CK2α/α′ or a scrambled sequence. A, the knockdown of CK2α/α′ was monitored by anti-CK2α immunoblotting. B, following the immunoprecipitation of p35, the activity of co-precipitated Cdk5 was determined. Data shown in the graph are the averages of three individual experiments.
      Cdk5 Inhibition by CK2 Is Phosphorylation-independent— CK2 is a protein Ser/Thr kinase that phosphorylates a multitude of proteins (
      • Meggio F.
      • Pinna L.A.
      ). To investigate the role of CK2 kinase function in the Cdk5 inhibition, a kinase-dead mutant of CK2α, namely CK2αK68A, was employed. Analogous to wild-type CK2α, CK2αK68A strongly blocked the p35 activation of Cdk5 in the reconstitution assays (Fig. 7). The wild-type and the kinase-dead mutant of CK2α displayed similar Cdk5 inhibitory activity (Fig. 7). Additionally, Cdk5 inhibition was achieved using the kinase-dead holoenzyme of CK2, which was reconstituted from CK2αK68A and CK2β (Fig. 7). From these results, it can be concluded that the CK2 inhibition of Cdk5 is independent of the CK2 kinase activity.
      Figure thumbnail gr7
      Fig. 7CK2 kinase activity is not required for the Cdk5 inhibition. GST-Cdk5 (0.5 μg) was mixed with GST-p35 (0.5 μg) and one of the following proteins (5 μg): CK2α, CK2αK68A, the wild-type holoenzyme CK2α/CK2β, or the kinase-dead holoenzyme CK2αK68A/CK2β. Samples were incubated for 1 h at room temperature before the Cdk5 kinase assay. Data shown in the graph are representatives of three separate experiments.
      CK2 Blocks the Association of Cdk5 and p35—To understand how CK2 exerts its inhibition toward Cdk5, we utilized biochemical binding assays to examine if CK2 affects the interaction between Cdk5 and p35. When GST-Cdk5 was reconstituted with p35-His6 in the presence of varying amounts of CK2α, it was found that CK2α strongly reduced the binding of p35 to Cdk5 (Fig. 8A, top). The amount of p35 co-precipitated with GST-Cdk5 correlated inversely with the input of CK2α. Similarly, when GST-p35 was incubated with His6-Cdk5 and an excess of CK2α, anti-Cdk5 Western blots of retrieved GST-p35 showed the dramatically reduced co-precipitation of Cdk5 (Fig. 8A, bottom). These results indicated that the association of Cdk5 and p35 were blocked by CK2α. Furthermore, we went on to analyze whether the association of CK2α with Cdk5 and p35 can be competed off by excessive amounts of p35 and Cdk5, respectively. In a binding competition, Cdk5 was applied at varying amounts to pre-complexed CK2α and GST-p35. The subsequent pull-down of GST-p35 showed that CK2α was gradually displaced by Cdk5 from the p35 complex (Fig. 8B, top). Similarly, when various amounts of p35 were applied to the pre-formed complex of CK2α and GST-Cdk5, p35 could displace CK2α from the complex (Fig. 8B, bottom).
      Figure thumbnail gr8
      Fig. 8Cdk5, p35 and CK2α compete to form heterodimeric complexes. A, CK2α blocks the interaction between Cdk5 and p35. Top, GST-Cdk5 (1 μg) was incubated with p35-His6 (0.1 μg) in the presence of CK2α as indicated for 1 h at 4 °C. GST-Cdk5 was subsequently retrieved using GSH beads. Bound p35-His6 was detected by immunoblotting. Bottom, GST-p35 (1 μg) was incubated with His6-Cdk5 (1 μg) in the presence or absence of CK2α (5 μg) for 1 h at 4 °C. GST-p35 was then retrieved to detect bound Cdk5 by immunoblotting. B, CK2α can be displaced from its complexes with Cdk5 and p35. Top, GST-p35 (1 μg) was pre-incubated with CK2α (1 μg) for 1 h at 4 °C. An increasing amount of His6-Cdk5 was then added for a further incubation of 1 h. After GST-p35 was retrieved, bound CK2α was detected by immunoblotting. Bottom, GST-Cdk5 (1 μg) was pre-incubated with CK2α (1 μg) for 1 h at 4 °C. Various amounts of p35-His6 were then added as indicated, and the samples were further incubated for 1 h. GST-Cdk5 was retrieved to detect bound CK2α by immunoblotting.

      DISCUSSION

      Protein kinase CK2 is a Ser/Thr kinase highly conserved in eukaryotic cells. It is a ubiquitously expressed and pleiotropic enzyme that is involved in the control of various cellular processes such as cell cycle, apoptosis, transcriptional regulation, and signal transduction (
      • Meggio F.
      • Pinna L.A.
      ,
      • Allende J.E.
      • Allende C.C.
      ,
      • Pinna L.A.
      • Meggio F.
      ,
      • Litchfield D.W.
      ). Like Cdk5, CK2 is much more abundant in the brain than in any other tissue. In neural cells, there appears to be a myriad of CK2 substrates that have clear implications in neural development, neuritogenesis, synaptic transmission, and plasticity (
      • Blanquet P.R.
      ). The CK2 holoenzyme is generally composed of two catalytic α- or α′-subunits and two regulatory β-subunits to form a tetrameric structure through the dimerization of the two β subunits (
      • Niefind K.
      • Guerra B.
      • Ermakowa I.
      • Issinger O.G.
      ). The α- and α′-subunits are catalytically active, whereas the β-subunit is inactive but appears to modulate enzyme activity and substrate specificity by targeting the enzyme to its substrates (
      • Allende J.E.
      • Allende C.C.
      ,
      • Pinna L.A.
      • Meggio F.
      ). In addition, several lines of evidence indicated that the α- and β-subunits can exist individually in vivo to interact with other cellular proteins, implying that the CK2 subunits may have biological functions other than those assigned to the holoenzyme (
      • Litchfield D.W.
      ,
      • Chen M.
      • Li D.
      • Krebs E.G.
      • Cooper J.A.
      ,
      • Chen M.
      • Cooper J.A.
      ,
      • Heriche J.K.
      • Lebrin F.
      • Rabilloud T.
      • Leroy D.
      • Chambaz E.M.
      • Goldberg Y.
      ,
      • Guerra B.
      • Siemer S.
      • Boldyreff B.
      • Issinger O.G.
      ).
      Evidence presented here identifies the catalytic α-subunit of CK2 as an inhibitor of Cdk5. It was shown from the co-immunoprecipitation results that CK2α exists in complexes with p35 and Cdk5 in the brain. Direct interaction between CK2α and Cdk5 or p35 was demonstrated by biochemical binding assays using purified recombinant proteins. Interestingly, based on the protein co-elution profiles of the rat brain lysate, it is CK2α but not the CK2 holoenzyme that coexists with p35 and Cdk5 in the brain, even though the holoenzyme displayed p35 binding capability in vitro. Furthermore, interaction characterization indicated that CK2α, Cdk5, and p35 compete with one another in the formation of heterodimeric complexes. The binding of CK2 to p35 or Cdk5 blocks the association of Cdk5 with p35, thereby inhibiting the Cdk5 activation by p35. In agreement with the results from the in vitro inhibition assays, the coexpression of CK2 and p35 in COS-7 cells inhibited the Cdk5 activation. In addition, the knockdown of CK2α/α′ significantly enhanced the p35-associated Cdk5 activity in the neuroblastoma cells. Although the Cdk5-inhibitory mechanism of CK2 is distinctive to the KIP/CIP family of Cdk inhibitors, it is reminiscent of the Cdk4/6 inhibition by the INK4 family members, which interact with both Cdk4/6 and cyclin D to block the Cdk4/6 association with cyclin D (
      • Russo A.A.
      • Tong L.
      • Lee J.O.
      • Jeffrey P.D.
      • Pavletich N.P.
      ,
      • Guan K.L.
      • Jenkins C.W.
      • Li Y.
      • O'Keefe C.L.
      • Noh S.
      • Wu X.
      • Zariwala M.
      • Matera A.G.
      • Xiong Y.
      ,
      • Hirai H.
      • Roussel M.F.
      • Kato J.Y.
      • Ashmun R.A.
      • Sherr C.J.
      ).
      CK2 was found to bind p35 in two regions located separately at the N- and C-terminal halves. Given the facts that the C-terminal binding site of CK2 lies within the Cdk5-binding domain and that CK2 displayed inhibitory activity toward the p25-associated Cdk5 activation, it is tempting to conclude that the binding of CK2 to the Cdk5-binding domain blocks the association of p35 with Cdk5. Perhaps also, the binding of CK2 to the N-terminal half of p35 causes a p35 conformation change to facilitate blocking of the interaction between p35 and Cdk5. The binary complexes of p35-CK2α and Cdk5-CK2α can be dissociated by excessive amounts of Cdk5 and p35, respectively. Upon the dissociation of CK2 from p35 or Cdk5 and the complex formation of Cdk5 and p35, Cdk5 becomes enzymatically active and catalyzes the phosphorylation of p35, promoting p35 degradation via the proteasome-dependent pathway (
      • Patrick G.N.
      • Zhou P.
      • Kwon Y.T.
      • Howley P.M.
      • Tsai L.H.
      ). Conceivably, the CK2 association prevents unwanted Cdk5 activation and subsequent p35 turnover.
      CK2 is generally thought to execute its functions through phosphorylation of a myriad of proteins that includes Cdk1 (
      • Russo G.L.
      • Vandenberg M.T.
      • Yu I.J.
      • Bae Y.S.
      • Franza Jr., B.R.
      • Marshak D.R.
      ). We have examined the involvement of CK2 kinase activity in the inhibition of Cdk5. Our assays using the kinase-dead mutant of CK2 indicated that the Cdk5-inhibiting activity of CK2 is independent of its kinase activity. This was corroborated with the competitive protein binding results, which were achieved under non-phosphorylation conditions. It became clear that CK2 imparts the Cdk5 inhibition through its physical association of Cdk5 or p35 but not through any enzymatic action. The results presented here have revealed a new CK2 function that is dissociated from its intrinsic kinase property.
      In summary, this study has presented the identification and characterization of a novel inhibitor of Cdk5. To date, all identified functions of Cdk5 are inextricably linked to the phosphorylation of its substrates. Our results revealed a control mechanism of Cdk5 activation and, thereby, its actions in the brain. Moreover, the aberrant Cdk5 activation by p25, which is a proteolytic product of p35 under neurotoxic conditions, occurs in degenerating neurons, and the deregulated Cdk5 activity is detrimental to cell survival (
      • Nguyen M.D.
      • Lariviere R.C.
      • Julien J.P.
      ,
      • Patrick G.N.
      • Zukerberg L.
      • Nikolic M.
      • de La Monte S.
      • Dikkes P.
      • Tsai L.H.
      ). CK2 exhibited a strong inhibition of the Cdk5 activation by p25, suggesting that the Cdk5 inhibition by CK2 might be one of the deregulated mechanisms during degeneration. It has been proposed that CK2 is involved in the processes underlying cell survival with anti-apoptotic functions (
      • Litchfield D.W.
      ,
      • Pinna L.A.
      ,
      • Ahmed K.
      • Gerber D.A.
      • Cochet C.
      ). Conceivably, CK2 may act to protect the cells against neurotoxicity, at least through its inhibition of Cdk5. Thus, the function described here for CK2 also facilitates our understanding of how CK2 supports cell viability in the nervous system.

      Acknowledgments

      We thank Drs. Jerry H. Wang and Nancy Y. Ip for critical reading of the manuscript. We are also indebted to Drs. Walter Hunziker and Alice Tay for their support and Sock-Yeen Tiu for technical assistance.

      References

        • 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
        • Lew J.
        • Beaudette K.
        • Litwin C.M.
        • Wang J.H.
        J. Biol. Chem. 1992; 267: 13383-13390
        • Xiong Y.
        • Zhang H.
        • Beach D.
        Cell. 1992; 71: 505-514
        • Dhavan R.
        • Tsai L.H.
        Nat. Rev. Mol. Cell Biol. 2001; 2: 749-759
        • Lew J.
        • Huang Q.Q.
        • Qi Z.
        • Winkfein R.J.
        • Aebersold R.
        • Hunt T.
        • Wang J.H.
        Nature. 1994; 371: 423-426
        • Tang D.
        • Yeung J.
        • Lee K.Y.
        • Matsushita M.
        • Matsui H.
        • Tomizawa K.
        • Hatase O.
        • Wang J.H.
        J. Biol. Chem. 1995; 270: 26897-26903
        • Tsai L.H.
        • Delalle I.
        • Caviness V.S.J.
        • Chae T.
        • Harlow E.
        Nature. 1994; 371: 419-423
        • Ko J.
        • Humbert S.
        • Bronson R.T.
        • Takahashi S.
        • Kulkarni A.B.
        • Li E.
        • Tsai L.H.
        J. Neurosci. 2001; 21: 6758-6771
        • Chae T.
        • Kwon Y.T.
        • Bronson R.
        • Dikkes P.
        • Li E.
        • Tsai L.H.
        Neuron. 1997; 18: 29-42
        • Ohshima T.
        • Ward J.M.
        • Huh C.G.
        • Longenecker G.
        • Veeranna Pant H.C.
        • Brady R.O.
        • Martin L.J.
        • Kulkarni A.B.
        Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 11173-11178
        • Baumann K.
        • Mandelkow E.M.
        • Biernat J.
        • Piwnica-Worms H.
        • Mandelkow E.
        FEBS Lett. 1993; 336: 417-424
        • Grant P.
        • Sharma P.
        • Pant H.C.
        Eur. J. Biochem. 2001; 268: 1534-1546
        • Ackerley S.
        • Thornhill P.
        • Grierson A.J.
        • Brownlees J.
        • Anderton B.H.
        • Leigh P.N.
        • Shaw C.E.
        • Miller C.C.
        J. Cell Biol. 2003; 161: 489-495
        • Rashid T.
        • Banerjee M.
        • Nikolic M.
        J. Biol. Chem. 2001; 276: 49043-49052
        • Nikolic M.
        • Chou M.M.
        • Lu W.
        • Mayer B.J.
        • Tsai L.H.
        Nature. 1998; 395: 194-198
        • Xie Z.
        • Sanada K.
        • Samuels B.A.
        • Shih H.
        • Tsai L.H.
        Cell. 2003; 114: 469-482
        • Lim A.C.
        • Qi R.Z.
        J. Alzheimers Dis. 2003; 5: 329-335
        • Lee M.S.
        • Kwon Y.T.
        • Li M.
        • Peng J.
        • Friedlander R.M.
        • Tsai L.H.
        Nature. 2000; 405: 360-364
        • Ahlijanian M.K.
        • Barrezueta N.X.
        • Williams R.D.
        • Jakowski A.
        • Kowsz K.P.
        • McCarthy S.
        • Coskran T.
        • Carlo A.
        • Seymour P.A.
        • Burkhardt J.E.
        • Nelson R.B.
        • McNeish J.D.
        Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 2910-2915
        • Noble W.
        • Olm V.
        • Takata K.
        • Casey E.
        • Mary O.
        • Meyerson J.
        • Gaynor K.
        • LaFrancois J.
        • Wang L.
        • Kondo T.
        • Davies P.
        • Burns M.
        • Veeranna Nixon R.
        • Dickson D.
        • Matsuoka Y.
        • Ahlijanian M.
        • Lau L.F.
        • Duff K.
        Neuron. 2003; 38: 555-565
        • Cruz J.C.
        • Tseng H.C.
        • Goldman J.A.
        • Shih H.
        • Tsai L.H.
        Neuron. 2003; 40: 471-483
        • Patrick G.N.
        • Zhou P.
        • Kwon Y.T.
        • Howley P.M.
        • Tsai L.H.
        J. Biol. Chem. 1998; 273: 24057-24064
        • Saito T.
        • Onuki R.
        • Fujita Y.
        • Kusakawa G.
        • Ishiguro K.
        • Bibb J.A.
        • Kishimoto T.
        • Hisanaga S.
        J. Neurosci. 2003; 23: 1189-1197
        • Zukerberg L.R.
        • Patrick G.N.
        • Nikolic M.
        • Humbert S.
        • Wu C.L.
        • Lanier L.M.
        • Gertler F.B.
        • Vidal M.
        • Van Etten R.A.
        • Tsai L.H.
        Neuron. 2000; 26: 633-646
        • Sasaki Y.
        • Cheng C.
        • Uchida Y.
        • Nakajima O.
        • Ohshima T.
        • Yagi T.
        • Taniguchi M.
        • Nakayama T.
        • Kishida R.
        • Kudo Y.
        • Ohno S.
        • Nakamura F.
        • Goshima Y.
        Neuron. 2002; 35: 907-920
        • Lee M.H.
        • Nikolic M.
        • Baptista C.A.
        • Lai E.
        • Tsai L.H.
        • Massague J.
        Proc Natl. Acad. Sci. U. S. A. 1996; 93: 3259-3263
        • Lee K.Y.
        • Rosales J.L.
        • Tang D.
        • Wang J.H.
        J. Biol. Chem. 1996; 271: 1538-1543
        • Ching Y.P.
        • Pang A.S.
        • Lam W.H.
        • Qi R.Z.
        • Wang J.H.
        J. Biol. Chem. 2002; 277: 15237-15240
        • Lim A.C.
        • Qu D.
        • Qi R.Z.
        Neurosignals. 2003; 12: 230-238
        • Lim A.C.
        • Tiu S.Y.
        • Li Q.
        • Qi R.Z.
        J. Biol. Chem. 2004; 279: 4433-4439
        • Qu D.
        • Li Q.
        • Lim H.Y.
        • Cheung N.S.
        • Li R.
        • Wang J.H.
        • Qi R.Z.
        J. Biol. Chem. 2002; 277: 7324-7332
        • Shevchenko A.
        • Wilm M.
        • Vorm O.
        • Mann M.
        Anal. Chem. 1996; 68: 850-858
        • Mann M.
        • Wilm M.
        Anal. Chem. 1994; 66: 4390-4399
        • Wilm M.
        • Shevchenko A.
        • Houthaeve T.
        • Breit S.
        • Schweigerer L.
        • Fotsis T.
        • Mann M.
        Nature. 1996; 379: 466-469
        • Qi Z.
        • Huang Q.Q.
        • Lee K.Y.
        • Lew J.
        • Wang J.H.
        J. Biol. Chem. 1995; 270: 10847-10854
        • Tang D.
        • Chun A.C.
        • Zhang M.
        • Wang J.H.
        J. Biol. Chem. 1997; 272: 12318-12327
        • Poon R.Y.
        • Lew J.
        • Hunter T.
        J. Biol. Chem. 1997; 272: 5703-5708
        • Meggio F.
        • Pinna L.A.
        FASEB J. 2003; 17: 349-368
        • Allende J.E.
        • Allende C.C.
        FASEB J. 1995; 9: 313-323
        • Pinna L.A.
        • Meggio F.
        Prog. Cell Cycle Res. 1997; 3: 77-97
        • Litchfield D.W.
        Biochem. J. 2003; 369: 1-15
        • Blanquet P.R.
        Prog. Neurobiol. 2000; 60: 211-246
        • Niefind K.
        • Guerra B.
        • Ermakowa I.
        • Issinger O.G.
        EMBO J. 2001; 20: 5320-5331
        • Chen M.
        • Li D.
        • Krebs E.G.
        • Cooper J.A.
        Mol. Cell. Biol. 1997; 17: 1904-1912
        • Chen M.
        • Cooper J.A.
        Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 9136-9140
        • Heriche J.K.
        • Lebrin F.
        • Rabilloud T.
        • Leroy D.
        • Chambaz E.M.
        • Goldberg Y.
        Science. 1997; 276: 952-955
        • Guerra B.
        • Siemer S.
        • Boldyreff B.
        • Issinger O.G.
        FEBS Lett. 1999; 462: 353-357
        • Russo A.A.
        • Tong L.
        • Lee J.O.
        • Jeffrey P.D.
        • Pavletich N.P.
        Nature. 1998; 395: 237-243
        • Guan K.L.
        • Jenkins C.W.
        • Li Y.
        • O'Keefe C.L.
        • Noh S.
        • Wu X.
        • Zariwala M.
        • Matera A.G.
        • Xiong Y.
        Mol. Biol. Cell. 1996; 7: 57-70
        • Hirai H.
        • Roussel M.F.
        • Kato J.Y.
        • Ashmun R.A.
        • Sherr C.J.
        Mol. Cell. Biol. 1995; 15: 2672-2681
        • Russo G.L.
        • Vandenberg M.T.
        • Yu I.J.
        • Bae Y.S.
        • Franza Jr., B.R.
        • Marshak D.R.
        J. Biol. Chem. 1992; 267: 20317-20325
        • Nguyen M.D.
        • Lariviere R.C.
        • Julien J.P.
        Neuron. 2001; 30: 135-147
        • Patrick G.N.
        • Zukerberg L.
        • Nikolic M.
        • de La Monte S.
        • Dikkes P.
        • Tsai L.H.
        Nature. 1999; 402: 615-622
        • Pinna L.A.
        J. Cell Sci. 2002; 115: 3873-3878
        • Ahmed K.
        • Gerber D.A.
        • Cochet C.
        Trends Cell Biol. 2002; 12: 226-230