If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
To whom correspondence should be addressed: Dept. of Biochemistry, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong . Tel.: 852-1358-8701; Fax: 852-2358-1552
Department of Biochemistry, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, theMedical Research Council Group in Signal Transduction, Department of Medical Biochemistry, University of Calgary, Calgary, Alberta T2N 4N1, Canada, the
* This work was supported in part by operating grants from the Medical Research Council 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.The nucleotide sequence(s) reported in this paper has been submitted to the GenBank™/EMBL Data Bank with accession number(s) U34051.
Neuronal Cdc2-like kinase is a heterodimer of Cdk5 and a 25-kDa subunit that is derived from a 35-kDa brain- and neuron-specific protein called the neuronal Cdk5 activator (p35/p25nck5a) (Lew, J., Huang, Q.-Q., Qi, Z., Winkfein, R. J., Aebersold, R., Hunt, T., and Wang, J. H.(1994) Nature 371, 423-426; Tsai, L. H., Delalle, I., Caviness, V. S., Jr., Chae, T., and Harlow, E.(1994) Nature 371, 419-423). Upon screening of a human hippocampus library with a bovine Nck5a cDNA, we uncovered a distinct clone encoding a 39-kDa isoform of Nck5a. The isoform, designated the neuronal Cdk5 activator isoform (p39nck5ai), showed a high degree of sequence similarity to p35nck5a with 57% amino acid identity. Northern blot analysis detected its mRNA transcript in bovine and rat cerebrum and cerebellum, but not in any other rat tissues examined. In situ hybridization showed that Nck5ai was enriched in CA1 to CA3 of the hippocampus, but absent in the fimbria of hippocampal formation. Among seven cell lines in proliferating cultures, only PC12 and N2A, two cell lines capable of differentiating into neuron-like cells, were found to contain Nck5ai mRNA. A 30-kDa truncated form of Nck5ai expressed as a glutathione S-transferase fusion protein in Escherichia coli] was found to associate with Cdk5 to form an active Cdk5 kinase. Thus, the isoform shares many common characteristics with p35nck5a, including Ckd5 activating activity and brain- and neuron-specific expression. Both proteins show limited sequence homology to cyclins, suggesting that they define a new family of cyclin-dependent kinase-activating proteins.
The cell division cycle gene, cdc2, in fission yeast performs rate-limiting functions in both G1/S and G2/M transitions. The protein product of the gene, p34cdc2, is a protein kinase catalytic subunit that associates with specific cyclins to form functional protein kinases, which are activated at discrete phases of the cell cycle(
). In mammalian cells, seven Cdc2 homologous proteins have been discovered; most have been shown to depend on cyclin for activity. These Cdc2 homologous proteins are called cyclin-dependent kinases (Cdks)
). In addition to Cdks, an extended family of cyclins has been demonstrated to exist in mammalian cells. Combinatorial association between members of these two protein groups gives rise to a large number of heterodimeric protein kinases that are called Cdc2-like kinases(
). Since this brain Cdc2-like kinase is considered neuron-specific on account of its regulatory subunit, it has been designated neuronal Cdc2-like kinase, and the regulatory subunits, neuronal Cdk5 activator (Nck5a or p25/35nck5a).
In addition to its tissue specificity, p25/35nck5a possesses certain unique properties that distinguish it from members of the cyclin family. The amino acid sequence of the protein displays minimal homology to cyclins(
). In this study, we document the discovery, cloning, and preliminary characterization of an isoform of Nck5a. The isoform, a 39-kDa protein, shows 57% amino acid sequence identity to Nck5a. The protein, tentatively designated the neuronal Cdk5 activator isoform (Nck5ai or p39nck5ai), is similar to Nck5a in displaying Cdk5 activation activity and brain- and neuron-specific expression, but distinct from Nck5a in cell culture distribution.
MATERIALS AND METHODS
cDNA Library Screen
A 2-year-old female human hippocampus cDNA library (Stratagene) was screened with a randomly labeled probe of bovine brain Nck5a according to the protocols of the manufacturer. 106 plaques were transferred to Hybond™ nylon membranes (Amersham Corp.); prehybridized at 42°C with prehybridization buffer containing 20 mM Pipes (Sigma), 0.8 mM NaCl, 50% formamide (Life Technologies, Inc.), 0.5% SDS, and 100 μg/ml denatured fragmented salmon sperm DNA for at least 2 h; and hybridized by addition of the probe, at a specific radioactivity of 106 cpm/ml, to prehybridization buffer at 42°C for 16 h. The membranes were washed twice in 2 × SSC (0.3 M NaCl, 0.03 M trisodium citrate), 0.1% SDS for 10 min each at room temperature and once in 0.2 × SSC, 0.1% SDS for 10 min at 42°C. Autoradiography was carried out at -70°C with a pair of amplifiers for 12 h.
Mouse fibroblast L929 cells were maintained in Joklik's modified minimum essential medium containing 5% fetal bovine serum. The other cell lines, A431 (a human epidermoid carcinoma cell line), HeLa, Jurkat, C6 (a glioma cell line), PC12, and Neuro-2A, were maintained in medium and serum as recommended(
). 5 μg of mRNA was separated in a 1.5% formaldehyde-agarose gel and transferred to a Hybond™ nylon membrane, which was then hybridized with buffer containing 6 × SSPE (1 M NaCl, 60 mM NaH2PO4, 6 mM EDTA), 50% formamide, 0.5% SDS, and 100 μg/ml boiled fragmented salmon sperm DNA for 16 h at 42°C. The membrane was washed twice in 2 × SSC, 0.5% SDS for 10 min each at room temperature and once in 0.2 × SSC, 0.5% SDS at 42°C for 15 min. Autoradiography was carried out at -70°C with amplifier screens for 20 h. The membrane was rehybridized with a new probe after soaking it twice in boiled 0.5% SDS. The random probes were labeled to 3 × 109 cpm/μg DNA(
). The p39nck5ai cRNA probes were prepared from rat cDNA clones. Specificity of these probes was verified by Northern blotting. Female Sprague-Dawley 3-month-old rats (Shizuoka Laboratory Center, Shizuoka, Japan) were deeply anesthetized with diethyl ether. Surgically obtained tissues were fixed in RNase-free 4% paraformaldehyde solution overnight; sequentially dehydrated with 70, 80, 90, and 100% ethanol; embedded in paraffin; and cut in sections of 4-μm thickness. Deparaffined tissue specimens were incubated for 15 min in Tris-HCl, pH 7.4, containing 0.01% proteinase K (Sigma) and acetylated for 15 min with 0.1 mol/liter triethanolamine containing 0.2% acetic acid. After sequential dehydration in alcohol solutions, each specimen was hybridized with a digoxigenin-labeled probe. After hybridization, the slides were incubated with RNase A (10 μg/ml) and washed with 2 × SSC and 0.2 × SSC. After thoroughly washing, the specimens were incubated in alkaline phosphatase-conjugated anti-digoxigenin antibody (Boehringer Mannheim GmbH) for 30 min. They were washed several times and developed by adding the substrate for alkaline phosphatase. As a control, the sense RNA probe was synthesized and used for hybridization at the same concentration and in the same way as described above.
The nucleotide sequence of nck5ai was determined by an automated DNA sequencer (Applied Biosystems, Inc.) using the DyeDeoxy™ terminator kit (Applied Biosystems, Inc.) and by the chain termination method using Sequenase Version 2.0 (U. S. Biochemical Corp.). For non-GC-rich regions, only one DNA strand was sequenced twice. For the 5′-GC-rich region, both DNA strands were sequenced several times. nck5ai was cloned in pBluescript KS+ and sequenced using both T3 (5′-ATTAACCCTCACTAAAG-3′) and T7 (5′-TAATACGACTCACTATAGGG-3′) primers.
Plasmid DNA used for nucleotide sequencing was purified by a modified mini-alkaline lysis/polyethylene glycol precipitation procedure as described by Applied Biosystems, Inc. (part 90149, rev. E). Other plasmid purifications were performed using the QIAprep plasmid purification kit (QIAGEN Inc.) according to the procedure recommended by the supplier.
The expression plasmids of GST-Cdk5 and GST-p21nck5a described previously (
) were used for protein expression. Escherichia coli strain BL21(DE3) was freshly transformed, and cells were cultured to A600 = 1.0 and then stimulated with isopropyl-β-D-thiogalactopyranoside at 25°C overnight. Cells were then lysed with a French press (1000 p.s.i.) in MTPBS (150 mM NaCl, 16 mM Na2HPO4, 4 mM NaH2PO4) containing 2 mM DTT, 2 μg/ml antipain, 2 μg/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride. After centrifugation at 10,000 × g for 5 min at 4°C, GST fusion protein was purified on glutathione-agarose (Sigma) as described previously(
). GST-Cdk5 was incubated with either GST-p21nck5a or GST-p30nck5ai at room temperature for 2 h in phosphate-buffered saline containing 1 mM EDTA and 1 mM DTT. The histone kinase activity of the reconstituted Cdk5 complex was measured in 30 mM Mops, pH 7.4, 10 mM MgCl2, 100 μM histone H1 peptide, and 100 μM [g-32P]ATP (1000 cpm/pmol) at 30°C for 30 min. The reaction was stopped by addition of acetic acid. Phosphate incorporation into the histone peptide was measured by a scintillation counter.
Proteins were separated by SDS-polyacrylamide gel electrophoresis (
) and transferred to a polyvinylidene difluoride membrane (Millipore Corp.). The membrane was blocked by 5% skim milk in 50 mM Tris, pH 7.5, 150 mM NaCl, and 0.1% Tween 20 and probed with a specific antibody.
Coprecipitation of Brain Cdk5 with GST Fusion Protein
Fresh bovine brain (1 kg) was homogenized in buffer containing 25 mM Hepes, pH 7.2, 1 mM EDTA, 1 mM DTT, 0.6 mM phenylmethylsulfonyl fluoride, 1 μg/ml leupeptin, 1 μg/ml antipain, 1 μg/ml aprotinin, 0.3 mg/ml benzamidine, and 0.1 mg/ml soybean trypsin inhibitor. The crude homogenate was centrifuged at 100,000 × g for 20 min, and the supernatant was incubated with GST or GST fusion proteins in homogenization buffer supplemented with 150 mM NaCl, 3 mM DTT at 4°C for 2 h. GST and GST fusion proteins were then precipitated with glutathione-agarose. The GST and GST fusion protein precipitates were either analyzed by Western blotting or assayed for histone kinase activity.
During the screening of a human hippocampus library with a randomly labeled DNA probe from a bovine Nck5a cDNA encoding the active domain of the protein(
), several clones (9I4, 3I4, 10I4, 15I23, and 7II2) were obtained. Restriction map analysis of the clones suggested that these clones represented two different mRNA transcripts. Partial nucleotide sequences and the deduced amino acid sequences of clones 15I23, 3I4, and 10I4 indicated that these clones encoded a protein that was identical to human Nck5a(
). On the basis of their nucleotide sequences and deduced amino acid sequences, clones 9I4 and 7II2 were found to encode a distinct Nck5a homologous protein. Clone 7II2 appeared to contain the complete protein reading frame. Fig. 1 shows the nucleotide sequence of clone 7II2 and the deduced amino acid sequence of the protein. The protein contains 367 amino acid residues with a calculated Mr of ~39,000. The assignment of the initiation codon is based on the observation that the sequence flanking the ATG codon contains an A at position -3 and a G at position +4, characteristic of a functional initiation codon(
). A GeneBank™ search with the nucleotide sequence and the protein sequence has shown that this is a novel protein.
The sequence alignment of p35nck5a and the isoform, p39nck5ai, is presented in Fig. 2A. The two proteins show 57% amino acid identity in the overall sequence, with a higher level of homology (65% amino acid identity) in the region of the protein corresponding to the 25-kDa subunit of the purified neuronal Cdc2-like kinase. This region of Nck5a contains the Cdk5-activating domain(
). The amino-terminal sequence of 9 amino acid residues and the carboxyl-terminal sequence of 3 residues of the two proteins are identical. The difference in molecular weights of the two proteins can be attributed mainly to the presence of a number of gaps in the Nck5a sequence when the proteins are aligned for optimal homology. Most of these gaps are located close to the protein termini. Like p35nck5a, p39nck5ai shows no overall amino acid sequence homology to cyclins. A small region of 17 residues in Nck5a and Nck5ai sequence is the only region with detectable sequence similarity to the equivalent region of the cyclin box consensus sequence (Fig. 2B). Amino acid identity of Nck5a and Nck5ai in this region is 88.2%.
In a previous study, Northern blot analysis of bovine brain mRNA using a Nck5a cDNA probe detected a 4-kb mRNA band, whereas human brain was found to contain two populations of mRNA of 4 and 2.4 kb(
). In the present study, mRNAs from bovine cerebrum and cerebellum and rat cerebrum and cerebellum were probed by the randomly labeled Nck5a cDNA. Fig. 3 shows that the two populations of mRNA could be detected in both bovine and rat cerebrum, with the Northern blot intensity of the 2.4-kb transcript much weaker than that of the 4-kb transcript. Bovine and rat cerebellum also contained the transcripts, but at much lower levels, and only the 4-kb transcript was clearly seen on the Northern blot. When the blots of rat and bovine brain mRNAs were probed by a randomly labeled Nck5ai cDNA probe, the same two populations of mRNA of 4.0 and 2.4 kb were detected in bovine and rat cerebrum. However, contrary to what was revealed by the Nck5a cDNA probe, the smaller transcript displayed far higher intensity than the larger transcript on the Northern blot probed by the Nck5ai cDNA (Fig. 3). Weak Northern blot signals of the 2.4-kb transcript could also be detected in bovine and rat cerebellum. whereas no signal was detected at the 4.0-kb mRNA positions (Fig. 3). These observations indicate that the 4.0- and 2.4-kb mRNAs represent Nck5a and Nck5ai transcripts, respectively.
Tissue distribution study in humans has revealed that the mRNA transcript of Nck5a is expressed specifically in the brain(
). Fig. 3 shows that Nck5a mRNA displays a similar brain-specific expression in rats. Essentially identical tissue expression patterns were observed for the 4- and 2.4-kb transcripts. Both transcripts were found to be expressed exclusively in brains, with levels in the cerebrum markedly higher than those in the cerebellum. No expression of either of the two transcripts was detected in non-neuronal tissues even upon increasing the exposure time by 10 times that used for the experiment of Fig. 3. The tissue distributions of Nck5a and Nck5ai mRNAs are markedly different from that of Cdk5. Previous studies showed that Cdk5 was widely distributed in humans as all human tissues examined were found to contain Cdk5 mRNA. Northern blot analysis of Cdk5 tissue distribution in rats showed that Cdk5 mRNA expression could also be readily detected in all the rat tissues examined (data not shown).
A number of proliferating cell lines were examined previously by Northern blot analysis and Western immunoblot analysis for Cdk5 distribution; all were found to contain Cdk5 mRNA and Cdk5 protein (
). To determine whether the cultured cells contained any of the neuronal Cdk5 activators, several cell lines were examined for the expression of p35nck5a and p39nck5ai mRNAs by Northern blot analysis. Fig. 4A shows that none of the cell lines expressed p35nck5amRNA, with the exception of Jurkat cells, a human T cell line, which showed a very low, nonetheless detectable Northern blot signal. On the other hand, N2A cells, a neuroblastoma cell line, and PC12 cells, a pheochromocytoma cell line that can be differentiated by nerve growth factor treatment into sympathetic neuron-like cells, expressed significant levels of p39nck5aimRNA (Fig. 4B). The level of the mRNA transcript was especially high in PC12 cells. None of the other cell lines, including a glioma cell line (C6), contained detectable p39nck5aitranscript. The observation that only cell lines of neural origin or with neuron characteristics express p39nck5aisuggests that the protein is neuron-specific.
In situ hybridization procedures were used to examine the regional distribution of p39nck5ai mRNA in the brain of a 3-month-old rat. The result indicated that the messenger transcript was highly expressed in CA1 to CA3 of hippocampal formation, an area enriched in neurons (Fig. 5). The presence of the Nck5ai mRNA in neurons can be clearly seen with a higher magnification of the CA3 region (Fig. 5C). Some expression of the p39nck5ai mRNA could also be seen in the dentate gyrus of the hippocampus. In contrast, no expression of the mRNA was detected in the fimbria hippocampi, an area containing axons of neurons and glial cells. These results on the regional distribution of p39nck5ai mRNA are compatible with the suggestion that p39nck5ai is a neuron-specific protein.
The active neuronal Cdc2-like kinase purified from bovine brain contains a 25-kDa subunit (p25nck5a)(
). Bacterially expressed p25nck5a or a further truncated form of 21 kDa (p21nck5a) can activate bacterially expressed monomeric Cdk5 to an activity similar to that of the purified neuronal Cdc2-like kinase(
). This observation suggests that p21nck5a contains the active domain essential for Cdk5 activation. A cDNA clone encoding a 30-kDa truncated form of p39nck5ai containing the region corresponding to p25nck5a was expressed as a GST fusion protein in E. coli. The expressed fusion protein, GST-p30nck5ai, was purified by glutathione affinity chromatography and tested for Cdk5 activating activity. Fig. 6 shows that the protein could activate bacterially expressed Cdk5 to a maximal activity about the same as that achieved by a GST-p21nck5a fusion protein. Three different preparations of GST-p30nck5ai were tested for the ability to activate Cdk5; essentially identical levels of maximal Cdk5 activation were obtained. However, the dose dependence of the Cdk5 activation by the bacterially expressed fusion protein varied from preparation to preparation, probably because different preparations contained different proportions of the active protein.
Bovine brain extract contains high levels of monomeric Cdk5. Monomeric brain Cdk5 has been partially purified and shown to associate readily with p25nck5a or p21nck5a to form active Cdk5 kinases. When the GST-p30nck5ai fusion protein was incubated with an aliquot of bovine brain extract and then affinity-precipitated by glutathione beads, Cdk5 in the brain extract was found to coprecipitate with the fusion protein as revealed by a specific Cdk5 antibody probe (Fig. 7A). The precipitate displayed good histone H1 peptide kinase activity, suggesting that brain Cdk5 could be activated by the fusion protein (Fig. 7B). In a separate experiment, a monomeric Cdk5 preparation partially purified from bovine brain was incubated with the GST-p30nck5ai fusion protein for 30 min, and the kinase activity of the sample was then determined. While neither monomeric Cdk5 nor the bacterially expressed fusion protein alone had any kinase activity, incubation of the two proteins together resulted in high histone H1 kinase activity (data not shown). This observation provides direct proof that Nck5ai is capable of activating the native form of Cdk5.
Until now, Nck5a has been the only protein known to be capable of reconstituting highly active Cdk5 kinases. In this study, we demonstrate the existence of an isoform of Nck5a, a 39-kDa protein that displays a high degree of amino acid sequence homology to p35nck5a with 57% amino acid identity. Like p35nck5a, this protein, called p39nck5ai, is capable of associating with Cdk5 to form functional kinases of high activity and shows brain-specific mRNA expression as revealed by the Northern blot analysis of its tissue distribution in both bovine and rat tissues. On the basis of detailed immunohistochemical studies and in situ hybridization analysis, p35nck5a has been suggested to be expressed specifically in neurons.
K. Tomizawa, H. Matsui, M. Matsushita, J. Lew, M. Tokuda, T. Itano, R. Konishi, J. H. Wang, and O. Hatase, submitted for publication.
At least two lines of evidence support the suggestion that p39nck5ai is also present specifically in animal brain neurons. Among a number of cell lines examined in culture, only N2A cells (a cell line of neural origin) and PC12 cells (a pheochromocytoma cell line capable of differentiating with nerve growth factor into neuron-like cells) were found to contain p39nck5ai mRNA. Significantly, a cell line of glial origin, the C6 glioma cell line, did not show detectable p39nck5ai mRNA signal on its Northern blot. The in situ hybridization patterns of Nck5ai indicate that p39nck5ai is greatly enriched in the CA1 to CA3 zone of hippocampal formation of adult rat brain (a region rich in neurons) and absent in the fimbria hippocampi, where glial cells predominate.
The existence of two brain- and neuron-specific Cdk5 activators has raised the question as to the physiological significance of the isoforms. As the brain is a highly complex organ containing many different cell types, one possible reason for the existence of isoforms of neuronal Cdk5 activators may be that the isoforms have distinct cell-type or brain regional distributions. Alternatively, the isoforms may coexist in the same neurons, but have differential subcellular localizations. We have recently observed that Cdk5 and p35nck5a are localized in different compartments in neurons of adult rat brains. While Cdk5 is enriched in the axons of central nervous neurons, p35nck5a has been seen mostly in the cell body.2 It is tempting to speculate that p39nck5ai may have an axonal localization so that the cell body and axonal function of Cdk5 in neurons are regulated by Nck5a and Nck5ai, respectively. The situation is somewhat similar to the regulation of Cdc2 by phase-specific cyclins in yeast during cell division cycle progression(
), except that the phase-specific cyclins are temporal specific regulators of Cdc2, whereas Nck5a and Nck5ai are spatial specific regulators of Cdk5. Work is under way to determine whether p39nck5ai indeed has an axonal localization.
The suggestion that p35nck5a and p39nck5ai have different functional roles in brain neurons appears to be supported by the observation that the two proteins are differentially expressed in cultured proliferating cells. Of all the cells examined, only the two cell lines with neuronal characteristics and capable of induced differentiation into neuron-like cells were found to contain p39nck5ai. While it is tempting to suggest that the protein contributes to the neuronal differentiation of the cells, supporting evidence for such a suggestion has not been forthcoming. In contrast to Nck5ai mRNA, Nck5a mRNA was not detected in any of the cell lines examined, except in the human T cell line (Jurkat cells), albeit at a very low level. The significance of the presence of p35nck5a in a human T cell line is not clear. It should be noted that neither p35nck5a nor p39nck5ai mRNA has been detected in bovine or rat thymus, an organ with concentrated T cells. The question as to why p35nck5a mRNA is absent in proliferating N2A and PC12 cells is not addressed in this study. The possibility that the protein and its mRNA are expressed in differentiated N2A and PC12 cells will be of considerable significance and interest.
Although p35nck5a and p39nck5ai are activators of a cyclin-dependent kinase, they display only marginal sequence homology to members of the cyclin family. The two proteins also appear to distinguish themselves from members of the cyclin family in the mechanism of kinase activation. It has been well documented that the activation of Cdk2 by cyclin A, Cdc2 by cyclin B, and Cdk4 by cyclin D is dependent on the phosphorylation of the Cdks on a threonine residue (Thr-160 of Cdc2 or its equivalent in the other Cdks) by a specific Cdk-activating kinase for full kinase activity(
). The unique amino acid sequences and distinctive activation characteristics together suggest that the two proteins may define a new family of kinase activator proteins. As none of the cyclins and no other proteins have been shown to activate Cdk5, this new family of kinase activators may be specific and unique for Cdk5 or Cdk5-like kinases. While Cdk5 has a wide tissue and cell distribution, only brain- and neuron-specific Cdk5 activators have been identified to date. It seems reasonable to suggest that there are additional members of this protein family that function as Cdk5 activators in non-neuronal cells. If so, degenerated polymerase chain reaction primers derived from some of the most conserved regions of these two neuronal Cdk5 activators may be useful in the search for additional Cdk5 kinase activators.