Human Prk Is a Conserved Protein Serine/Threonine Kinase Involved in Regulating M Phase Functions*

Human prk encodes a novel protein serine/threonine kinase capable of strongly phosphorylating casein but not histone H1 in vitro. prk expression is tightly regulated at various levels during different stages of the cell cycle in lung fibroblasts. The Prk kinase activity is relatively low during mitosis, G1, and G1/S, and peaks during late S and G2 stages of the cell cycle. Recombinant human Prk expressed through the baculoviral vector system is capable of phosphorylating Cdc25C, a positive regulator for the G2/M transition. Human prk shares significant sequence homology with Saccharomyces cerevisiae CDC5 and Drosophila melanogaster polo, both of which are essential for mitosis and meiosis. Full-length prk transcripts greatly potentiate progesterone-induced meiotic maturation of Xenopus laevisoocytes. On the other hand, antisense prk transcripts significantly delay and reduce the rate of oocyte maturation. When expressed in a CDC5 mutant strain of S. cerevisiae, human Prk, but not a deletional mutant protein, fully rescues the temperature-sensitive phenotype of the budding yeast. Taken together,prk may represent a new protein kinase, playing an important role in regulating the onset and/or progression of mitosis in mammalian cells.

Cyclin-dependent kinases (CDKs) 1 control many phosphorylation events during the cell cycle, and are indispensable for eukaryotic cell division (1,2). Because of their importance in the cell cycle, CDKs are structurally and functionally conserved across a wide spectrum of evolution. For example, human p34 cdc2 is capable of functional complementation of temperature sensitive (ts) Cdc2 mutant strains of the fission yeast, Schizosaccharomyces pombe (3). During the past decade, significant progress has been made in structural and functional characterization of individual components such as cyclins and CDKs regulating the transit of cells through various checkpoints of the animal cell cycle (1)(2)(3)(4)(5). For example, p34 cdc2 plays a rate-limiting role in the transition from G 2 into M (6), and Cdc25C gene product, a dual specific protein phosphatase, dephosphorylates p34 cdc2 on threonine 14 and tyrosine 15 residues and thereby activates p34 cdc2 kinase activity (7). The Cdc25C phosphatase activity is also under regulation by reversible phosphorylation. It has been demonstrated that extensive phosphorylation of Cdc25C amino-terminal domain occurs at the onset of mitosis by a novel enzyme (8), and this phosphorylation strongly activates Cdc25C's phosphatase activity toward p34 cdc2 (8,9).
Recently, an emerging family of protein kinases (termed the polo kinase family) has been described in yeast, Drosophila, and higher animals. In Drosophila, the polo gene encodes a protein serine/threonine (Ser/Thr) kinase that is required for M phase functions (10,11). Mutants of polo induce hyper-condensed chromosomes, abnormal spindle formation, and polyploidy (10,11). A polo homolog encoded by CDC5 in Saccharomyces cerevisiae is required for nuclear division in the late mitotic stage (12). Deletion of this gene is lethal, resulting in a dumbbell-shaped terminal morphology with incomplete nuclear division (12). We have recently cloned a cDNA encoding a putative protein Ser/Thr kinase (13) named prk (proliferation-related kinase) which has a predicted molecular mass of 67.8 kDa. Prk shares an extensive homology with the polo family kinases, and the homology is not confined solely to the kinase domain (13). Expression of prk mRNA is inducible by growth factors or cytokines, and appears to be down-regulated in many lung carcinomas (13).

EXPERIMENTAL PROCEDURES
Cell Culture and Synchronization-Human A549 fibroblast cells originally derived from a lung carcinoma were cultured in Dulbecco's modified Eagle's medium (DMEM, Life Technologies, Inc.) supplemented with 10% fetal bovine serum (FBS, Hyclone, Logan UT) and antibiotics (100 units/ml penicillin and 50 g/ml streptomycin sulfate, Life Technologies, Inc.). G 1 cells were obtained by culturing in methionine-free DMEM containing 10% FBS for 48 h. Cells arrested at G 1 /S (some refer this stage as early S or S) (14,15) were achieved by the treatment with aphidicolin (5 g/ml) and hydroxyurea (0.5 mM) for 48 h. Later S (S/G 2 ) cells were obtained through washing the G 1 /S-arrested cells as above with phosphate-buffered saline and reculturing them in DMEM with 10% FBS for 6 h. To obtain mitotic metaphase cells, A549 cells were treated with nocodazole (0.4 g/ml) for 16 h. Aphidicolin, hydroxyurea, and nocodazole were purchased from Sigma.
RNA Isolation and Northern Blotting-Total RNA was isolated from A549 cells or Xenopus oocytes using the guanidine isothiocyanate method (16). Equal amounts of total RNA (15 g) were fractionated on 1% formaldehyde agarose gels. The fractionated RNA was transferred to Nytran Plus membranes. The RNA blots were baked for 2 h, prehybridized for 2 h, and hybridized with a 32 P-labeled prk or ␤-actin cDNA fragment overnight. After hybridization, the blots were washed and autoradiographed. High stringency hybridization and washing conditions were as described previously (17).
Antibody Production and Western Blotting-Both the N-terminal half (Prk-N, amino acids 20 -341) (13) and the C-terminal half (Prk-C, amino acids 334 -607) of Prk were expressed as a glutathione S-trans-ferase (GST) fusion protein using the plasmid pGEX-2T (Pharmacia Biotech Inc.). Fusion proteins were induced by isopropyl-␤-D-thiogalactopyranoside and purified through affinity column chromatography according to the protocol provided by the supplier. Polyclonal anti-Prk antisera were produced in rabbits using the purified Prk-C through a commercial source (Research Genetics, Inc., Atlanta, GA). For Western blotting, cells with various treatments were collected and lysed in a lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 1 mM EDTA, Na 3 VO 4 , 1 mM phenylmethylsulfonyl fluoride, 10 g/ml approtinin, 10 g/ml leupeptin). Proteins from cell lysates, representing equal numbers of cells, were analyzed through SDS-polyacrylamide gel electrophoresis (PAGE) followed by Western blotting. The protein blots were first probed with an anti-Prk or anti-Plk (Zymed Laboratories Inc., South San Francisco, CA) antibody raised in a rabbit and then with goat-anti-rabbit antibodies conjugated with horseradish peroxidase. The signals were detected by enhanced chemiluminescence (Amersham Corp.).
Immunoprecipitation and Immunocomplex Kinase Assays-A549 cells were collected and lysed in the lysis buffer as above. Protein lysates from about 3 ϫ 10 6 cells were preclarified by the addition of a normal rabbit serum (1:50 dilution) and 15 l of protein A-agarose bead slurry (Santa Cruz Inc., Santa Cruz, CA), and incubated at room temperature for 30 min. After centrifugation (12,500 ϫ g for 5 min), the supernatants were supplemented with an anti-Prk antiserum (1:100 dilution) or the normal rabbit serum as a control, and incubated at 4°C for 1 h. Protein A-agarose beads (15 l) were then added to each immunoprecipitation mixture, and the incubation was continued at 4°C for at least 8 h. Immunoprecipitates were collected, washed four times with the lysis buffer and once with a kinase buffer (buffer for casein: 10 mM HEPES, pH 7.4, 10 M MnCl 2 , 5 mM MgCl 2 ; buffer for histone H1: 25 mM HEPES, pH 7.4, 25 mM ␤-glycerophosphate, 10 mM MgCl 2 , 5 mM EGTA, 10 mM NaF, 1 mM dithiothreitol), and resuspended in 30 l of the respective kinase buffer. The kinase activity of immunoprecipitated Prk was assayed by the addition of a substrate (casein (20 g/reaction) or histone H1 (20 g/reaction), Upstate Biotechnology, Inc., Lake Place, NY)) and [␥-32 P]ATP (2 Ci) to each kinase mixture, and the reaction lasted for 30 min at 37°C. The kinase reaction mixtures were analyzed by SDS-PAGE followed by autoradiography. Purified recombinant His 6 -Prk expressed through the baculoviral expression system was also assayed for its in vitro kinase activity using substrates such as casein (15 g/reaction) and recombinant GST-Cdc25C (2.5 g/reaction). GST-Cdc25C was kindly provided by K. Galaktinov. The kinase reaction conditions were the same as above. The kinase reactions were terminated by the addition of 2 ϫ SDS-polyacrylamide gel sample buffer and analyzed on a 10% SDS-polyacrylamide gel followed by autoradiography. For phosphoamino acid analysis, the phosphorylated GST-Cdc25C was excised and eluted from the gel. The eluted GST-Cdc25C was then hydrolyzed in 6 N HCl for 60 min at 110°C. The hydrolyzed amino acids were analyzed by two-dimensional thin layer chromatography and autoradiography.
Expression and Purification of Recombinant Human Prk from Baculovirus-Full-length recombinant Prk was expressed using the baculoviral expression system (PharMingen, San Diego, CA) following the manufacturer's protocol. Briefly, a cDNA fragment containing the entire open reading frame of human prk was cloned into PVL-1393 transfer vector. To facilitate purification of recombinant Prk protein, a short nucleotide sequence coding for six histidine residues was inserted in-frame immediately after the ATG codon of prk cDNA. Baculoviral expression vector BaculoGold TM DNA and the transfer plasmid PVL-1393-Prk were then co-transfected into insect sf9 cells. Insect sf9 cells expressing recombinant Prk were harvested and lysed for purification of recombinant Prk using Ni-nitrilotriacetic acid resin (Qiagen, Chatworth, CA).
Oocyte Maturation Studies-The prk cDNA insert was cloned into pcDNA3 expression vector (Invitrogen, Inc., San Diego, CA). In vitro transcription was performed using T7 DNA-dependent RNA polymerase in the presence of a Cap analog (m7G(5Ј)ppp(5Ј)) according to the protocol provided by the supplier (Promega, Madison, WI). Antisense transcripts were synthesized with Sp6 DNA-dependent RNA polymerase in the absence of the Cap analog. In vitro transcribed RNAs were finally dissolved in diethylpyrocarboanate-treated water at a concentration of 1 ng/nl. Stage 5-6 oocytes from female Xenopus frogs (Xenopus I, Lansing, MI) were obtained and stored at 18°C in a modified Barth's solution (MBS) as described elsewhere (18). Microinjection was performed under an anatomic microscope and manipulated by a micromanipulator. Each oocyte was injected with 50 ng of transcripts by positive displacement using a 10-l micropipette. Oocytes injected with sense transcripts were incubated at 18°C for 60 h before induction experiments. Maturation was induced by a 30 min pulse exposure to progesterone (3 M for sense experiments or 30 M for antisense experiments) in MBS. After induction, oocytes were transferred to fresh MBS and incubated at room temperature for various lengths of time (1,2,3,4,5,6,7,8,9, and 10 h) unless otherwise specified. Meiotic maturation of oocytes was scored by observing the frequency of germinal vesicle breakdown (GVBD), characterized by the appearance of a white spot on the animal pole of the oocyte. Each experiment was repeated at least four times, and statistical confidence determined by the Student's t test.
PCR Amplification-Reverse transcriptase-mediated PCR was performed on Xenopus oocyte or A549 total RNA using a SuperScript TM preamplification system (Life Technologies, Inc.). A pair of oligonucleotide primers corresponding to two stretches of human prk cDNA sequences was synthesized by Oligo Etc. (Wilsonville, OR). The pair of prk primers has the following sequences: forward primer, 5Ј-AGCGGCCT-CATGCGCACA-3Ј, and reverse primer, 5Ј-AATCCATCTCCACTGCT-TGC-3Ј. Thirty cycles of PCR were performed with denaturing at 94°C for 1 min, annealing at 48°C for 1 min, and extension at 72°C for 1 min. As an amplification control, a pair of human ␤-actin primers were also used to amplify Xenopus and A549 cell cDNA. PCR products were analyzed on agarose gels and the fractionated PCR products were transferred to nitrocellulose membranes. Southern blots were probed with a 32 P-labeled human prk cDNA fragment. High stringency hybridization and washing conditions were as described previously (17).
Yeast Complementation Studies-Human prk cDNA was cloned into a yeast expression vector pADNS as described elsewhere (19). A short version of prk with a 23-amino acid deletion at the N terminus (Prk-st) was also cloned into the same vector. As a positive control, the wild-type CDC5 gene of S. cerevisiae was obtained via polymerase chain reaction and cloned into a pADNS expression vector. Original S. cerevisiae CDC5 mutants were as described previously (20), and a CDC5 mutant strain (2751-4-3a) was kindly provided by Dr. L. H. Hartwell at the University of Washington (Seattle, WA). The mutant strain had the following genotype: Mat␣, cdc5-1, leu2-3, 112 his7, and was maintained on YPD (yeast extract, peptone, dextrose media) plates. Various expression constructs as well as the parental plasmid were transformed into CDC5 mutants, and transformants were cultured at 22°C on drop-out base plates supplemented with a complete supplement mixture but without leucine (BIO 101, Inc., Vista, CA). Individual colonies from each transformation were then spread onto plates and cultured at 22°C for 3 days. Duplicate plates were cultured at 33°C for the same length of time.

RESULTS
Since prk mRNA expression is rapidly activated by serum or selected cytokines (13), we first examined the steady state level of prk mRNA during various stages of the cell cycle. Human A549 cells (21) were arrested at different stages of the cell cycle as described under "Experimental Procedures." Fig. 1A shows that a low level of prk transcript was present in G 1 cells (lane 3). The prk transcript level was significantly elevated at the G 1 /S junction (lane 4), peaking at late S and G 2 stages of the cell cycle (lane 5). Interestingly, prk mRNA level was markedly reduced during M phase (lane 6), indicating that prk mRNA expression was tightly regulated during the cell cycle. The cell cycle status in all cells was monitored by flow cytometric analysis of propidium iodide-stained cells (data not shown).
The Prk protein has two apparent domains with the approximate N-terminal half bearing the typical protein Ser/Thr kinase structure (13). We generated recombinant GST fusion proteins containing either the N-terminal half (Prk-N) or the C-terminal half (Prk-C) of Prk (data not shown). Since the C-terminal half of Prk shares no structural homology to any other known proteins except for polo family kinases, purified Prk-C was used to raise polyclonal anti-human Prk antibodies in rabbits. An anti-Prk antisera was first characterized for its specificity. Western blot analyses showed that two bands (with the molecular mass of about 68 and 64 kDa, respectively) were immunoreactive to an anti-Prk antiserum (Fig. 1B, lanes 1 and  2) but not to the preimmune serum (lanes 3 and 4). The 68-kDa band, but not the 64-kDa band, was competed by an excess amount of purified Prk-C protein (Fig. 1B, lanes 5 and 6), indicating that the band of 68 kDa represents Prk protein. Plk is another known member of the polo family kinases in the human, and it shares about 50% amino acid sequence identity with human Prk. Western blot analysis showed that Plk is also expressed in A549 cells, and it has a faster mobility of about 66 kDa than the 68-kDa Prk on denaturing gels (Fig. 1B, lanes 7  and 8). To determine Prk antigen levels, unsynchronized A549 cells as well as the cells synchronized at various stages of the cell cycle were analyzed for Prk expression by Western blotting using the anti-Prk anti-serum. Fig. 1C shows that Prk antigen levels were low in G 1 , but remained relatively constant throughout the rest of the cell cycle. The same blot was stripped and probed with an anti-Plk antibody. Fig. 1D shows that little or no Plk antigen was present in G 1 , G 1 /S, or S/G 2 (Fig. 1D,  lanes 2-4), but very high levels of the antigen were accumulated in M (lane 5). Both anti-Prk and anti-Plk antibodies recognized the murine counterparts (Figs. 1, C and D, lane 6).
To determine whether prk encodes an active kinase, Prk protein was immunoprecipitated from A549 cells and analyzed for its protein kinase activity in the presence of [␥-32 P]ATP and an in vitro substrate such as casein or histone H1. As a positive control for histone H1 kinase activity, p34 cdc2 kinase was also immunoprecipitated from A549 cell lysates. Fig. 2A shows that Prk immunoprecipitates catalyzed the incorporation of 32 P into casein (lane 2, arrowhead). No nonspecific incorporation of radioactivity into proteins of substrate size was observed with the preimmune serum control (lane 1), indicating that Prk is an active protein kinase. When histone H1 was used as a substrate, a very low Prk kinase activity was detected (lane 4). On the other hand, p34 cdc2 immunoprecipitates strongly catalyzed the incorporation of radioactivity into histone H1 (lane 5). To determine whether the Prk kinase activity is regulated during the cell cycle, cell lysates collected from defined stages of the cell cycle were immunoprecipitated with the anti-Prk antiserum. Immunocomplex kinase assays showed that Prk kinase activity was low during G 1 (Fig. 2B, lane 3) or G 1 /S (lane 4). The kinase activity dramatically increased during late S/G 2 (lane 5), and remained at a moderate level at metaphase (lane 6).
CDK's kinase activity absolutely depends on the association with a specific cyclin. To determine whether Prk's kinase activity requires the association with other protein(s), we expressed full-length Prk using the baculoviral expression system. Western blot analyses revealed that His 6 -Prk was strongly expressed (Fig. 3A, lanes 2 and 3) in sf9 cells infected with recombinant baculovirus. The purified His 6 -Prk (Fig. 3A,  lane 4) was analyzed for its kinase activity using casein as a substrate. Fig. 3B shows that His 6 -Prk strongly phosphorylates casein (lane 2, arrowhead casein), and the addition of staurosporine, a protein serine/threonine kinase inhibitor, drastically inhibited phosphorylation of casein by Prk (lane 6). Prk is capable of autophosphorylation (Fig. 3B, arrowhead Prk). The small arrowhead (Fig. 3B) denotes a degradation product of Prk, since it is immunoreactive to anti-Prk antibody (arrowhead in Fig. 3A). Since Cdc25C is extensively phosphorylated by a kinase other than Cdc2 and Cdk2 before the onset of mitosis (8), we tested whether Prk is capable of phosphorylating Cdc25C. Fig. 3B shows that His 6 -Prk (lane 3), but not its deletion mutant (lane 5), phosphorylates GST-Cdc25C. No phosphorylation was detected when GST alone was used as a substrate (data not shown), indicating the phosphorylation oc-  1 and 3), the anti-Prk antiserum (lanes 2 and 4), or an anti-p34 cdc2 antibody (lane 5). The immunoprecipitates were analyzed for in vitro protein kinase activity in the presence of [␥-32 P]]ATP and a substrate (casein or histone H1). Kinase reaction mixtures were then fractionated by SDS-PAGE followed by autoradiography. B, equal amounts of protein lysates from cells arrested at various stages of the cell cycle were immunoprecipitated with the anti-Prk antiserum (lanes 3-6). As a control, the protein lysates from unsynchronized cells (unsyn) were immunoprecipitated with the preimmune serum (unsyn/pre, lane 1) or the anti-Prk antiserum (unsyn, lane 2). The immunoprecipitates were analyzed for in vitro kinase activity using casein as a substrate. curs on Cdc25C moiety. We quantitated substrate phosphorylation by titrating the amount of Cdc25C used for the kinase reaction. His 6 -Prk phosphorylates Cdc25C to a stoichiometry of 4 mol of phosphate per mol of Cdc25C. To confirm the predicted Ser/Thr phosphorylation by Prk, GST-Cdc25C protein eluted from the gel was hydrolyzed as described under "Experimental Procedures" and analyzed for its phosphoamino acid content. As shown in Fig. 3C, both phosphoserine and phosphothreonine are present with a ratio of about 3 to 1.
The X. laevis oocyte is a useful bioassay system for assessing complex nuclear events (22)(23)(24). Immature oocytes are arrested at the G 2 /M border of meiosis I. When such oocytes are stimulated with progesterone, they undergo meiotic maturation which is visibly monitored by GVBD. Microinjection of Cdc2 activators such as cyclin B and Cdc25 also induce a rapid GVBD (23,24). Since prk homologs in yeast and Drosophila are involved in regulating mitotic and meiotic progression (10 -12), we asked whether prk is required for Xenopus oocyte maturation. Fig. 4A shows that injection of neither vehicle nor prk transcripts alone significantly stimulated the maturation rate. Progesterone treatment alone induced GVBD of about 40% oocytes. However, when treated with progesterone, all oocytes injected with the full-length, but not the short form, prk transcripts reached the GVBD stage (Fig. 4A). The short form refers to Prk with an in-frame deletion of amino acids 2-24 in the kinase domain as described elsewhere (13). Thus, injection of full-length prk transcripts greatly potentiated the progester-one-mediated oocyte maturation. To determine whether or not loss of function could affect induced GVBD, we tested the effect of antisense prk transcripts on the maturation of frog oocytes. Fig. 4B shows that injection of the vehicle or antisense prk transcripts alone had no effect on the maturation rate. However, injection of antisense prk transcripts significantly blocked the progesterone-induced GVBD (Fig. 4B). Thus, these oocytes experiments suggest that a progesterone-induced factor(s) is required for the activation of Prk or that activation of Prk downstream component(s) requires both Prk and a progesterone-induced factor. To gain insight into the specificity of antisense human prk transcripts, Xenopus oocyte RNA was analyzed for prk-specific transcripts using reverse transcriptase-PCR. Fig. 4C shows that human prk primers were capable of amplifying a Xenopus DNA product (lane 2) that has the same mobility as human prk fragment on an agarose gel (lane 3). High stringency Southern blotting analyses revealed that the amplified Xenopus cDNA fragment, which is outside the kinase domain, hybridized strongly with a human prk cDNA probe (Fig. 4D, lanes 2). Subsequent DNA sequencing analysis showed that Xenopus did contain prk gene and that there existed a very high homology (97%) between the amplified human and Xenopus prk cDNA fragments (data not shown). The slower mobility band in lane 3 (Fig. 4C) was most likely derived from the unspliced prk transcript. To further determine whether suppression of Prk expression would also delay their maturation, Xenopus oocytes injected with antisense prk transcripts or the vehicle were treated with progesterone and examined for GVBD at various time post treatment. It has been observed (Fig. 5) (CNTL). B, oocytes injected with antisense prk transcripts (Antisense) with or without progesterone (Prog.) treatment. Results from four experiments are shown with standard error bars. Star (*) denotes that the change in the maturation rate is statistically significant by the Student's t test analysis. C, agarose gel analysis of reverse transcriptase-PCR products amplified from Xenopus frog (lanes 2 and 4) or human (lanes 3 and 5) RNA using primer pairs corresponding to human prk (lanes 2 and 3) or human ␤-actin (lanes 4 and 5). Lane 1 (Mk) represents a 100-base pair molecular standard. D, high stringency Southern blot analysis of the fractionated PCR products as shown in C, using a human prk cDNA fragment as a probe.
hand, a small number of oocytes injected with antisense prk transcripts did not show any signs of maturation until 9 h post-progesterone treatment (Fig. 5, filled circles), suggesting that a loss Prk function also significantly delayed meiotic maturation of oocytes induced by progesterone.
Complementation studies with yeast conditional mutants have led to the cloning of many human genes conserved in cell cycle control and elucidation of their respective functions (3,25). The polo family kinases share structural homology, and the biological functions of the budding yeast CDC5 and Drosophila polo appear to be similar; ablation of either gene function arrests the cell at the G 2 /M phase (10 -12). The human Prk and yeast CDC5 share 38% amino acid residue identities which extend beyond the kinase domain (Fig. 6A). This structural conservation led us to ask whether human prk can complement a yeast CDC5 ts mutant strain. The wild-type CDC5 gene, used as a positive control, was cloned from the wild-type yeast genome via PCR. Fig. 6B shows that CDC5 ts mutant cells transformed with the vector alone remained temperature-sensitive. In contrast, the full-length human prk gene restored the capacity of mutants to grow at 33°C, and the growth rate was indistinguishable from CDC5 cells transformed with the wildtype CDC5 gene. In addition, a prk expression construct containing a short in-frame deletion in the conserved kinase domain (Prk-st) was unable to grow at 33°C (Fig. 6B), indicating that a functional prk is important in rescuing the mutant phenotype.

DISCUSSION
All proliferating cells undergo an orderly cyclic process that entails genome duplication and mitosis. Major checkpoints in the animal cell cycle control entry into DNA replication and completion of DNA replication before mitosis. Central to this regulation are families of Ser/Thr kinases (1)(2)(3)(4)(5)(6). CDKs were the first family of well documented protein kinases shown to play a major role in the regulation of G 1 , G 1 /S, and G 2 /M phase checkpoints (1)(2)(3)(4)(5)(6). Recently, the polo family of protein kinases has been implicated in the regulation of cell cycle progression at various stages (10,13,(25)(26)(27)(28)(29). There are at least three genes (fnk/prk, snk, and plk) in the mouse that show significant structural homology to Drosophila polo and S. cerevisiae CDC5 (25)(26)(27)(28)(29). In human, plk and prk represent the two known polo family kinases (13,26,28). Structurally, polo family kinases differ significantly from CDKs in that the former appears to  6. Rescuing the temperature-sensitive phenotype of a S. cerevisiae CDC5 mutant by human prk. A, computer-assisted amino acid sequence alignment of human Prk and the budding yeast CDC5. Darkened areas show the residue identity and shaded areas represent conservative substitutions. Criteria for conservative amino acid changes were as described previously (31). Two pairs of arrows indicate the kinase and putative regulatory domains, respectively. B, CDC5 mutants transformed with the vector, Prk, Prk-st, or the wild-type CDC5 gene were grown at 22 or 33°C, respectively. contain both kinase and regulatory domains (13), and their in vitro kinase activities do not require the association with other protein components (Fig. 3). Human Plk protein exhibits about 50% amino acid identity with human Prk (13). Plk has been shown to undergo dramatic redistribution as cells progress from metaphase to anaphase (29). Recently, Kumagai and Dunphy (30) have shown that Plx, a Xenopus counterpart of mammalian Plk, associates with, phosphorylates, and activates the XCdc25 (Xenopus Cdc25C) gene product. The activated XCdc25 in turn dephosphorylates and activates p34 cdc2 . We have demonstrated that human prk is capable of rescuing a yeast CDC5 mitotic mutant, indicating that human prk has a conserved function in regulating mitotic/meiotic progression and is a functional homolog of the yeast CDC5 gene.
Western blot analyses reveal that there are two major proteins immunoreactive with the anti-Prk antiserum (Fig. 1B). The band with a slower mobility on the denaturing gel appears to be Prk, since it has a molecular mass of about 68 kDa, which is consistent with what is predicted from its deduced amino acid sequence (13). In addition, the 68-kDa band, but not the 64-kDa band, is competed out with excess purified recombinant Prk-C (Fig. 1B). The 64-kDa band is not Plk since it is smaller than Plk and since it is not immunoreactive to the anti-Plk antibody (Fig. 2). Human prk encodes an active protein kinase capable of phosphorylating casein and Cdc25C in vitro (Figs. 2 and 3), and it phosphorylates histone H1 weakly (Fig. 2). The latter observation is consistent with early reports that histone H1 is a poor substrate for polo family kinases (29). We have shown that Prk kinase activity peaks at the late S and G 2 stages of the cell cycle (Fig. 2B). Considering that Prk antigen levels showed no significant increase in S/G 2 compared with that in G 1 /S and M (Fig. 1C), Prk kinase activity must be regulated by a post-translational mechanism.
In this report we also demonstrate that antisense human prk transcripts significantly block as well as delay the maturation of Xenopus oocytes induced by progesterone. These results suggest that prk is required for the oocyte meiotic maturation and that frog prk has functional as well as sequence homology to human prk. We have observed that there is about 40% overall homology between human Prk and Drosophila polo at the nucleotide level, and that the level of homology increases to 65% when only the kinase domains are aligned. Thus, it is reasonable to predict that human Prk should show stronger overall homology with the frog counterpart than with polo. Our PCR and Southern blotting analyses (Fig. 4, C and D) suggest that Xenopus has at least one prk gene. Our subsequent DNA sequencing analysis has confirmed the identity of the PCR product as a Xenopus prk fragment. Recently, it has been shown that Plx shares 80% amino acid sequence identity with human Plk (30), also indicating structural conservation of polo family kinases across species. We noted that the short form of human prk transcripts (Prk-st) stimulated to some degree GVBD in the presence of progesterone (Fig. 4A), whereas the same prk structure failed to rescue yeast CDC5 ts mutants (Fig. 6B). One interpretation is that Prk-st may possess a low level of biological activity that is capable of weakly regulating cellular targets in Xenopus. On the other hand, this level of activity is insufficient for complementing the deficiency in yeast because of the evolutionary distance.
p34 cdc2 kinase, a component of mitosis-promoting factor (6), plays an important role in the transition from G 2 into M. However, it remains unclear what molecular component(s) senses the completion of genome replication and initiates mitosis. Although Plx1 was recently shown to interact with and phosphorylate XCdc25 (30), it remains to be determined whether Plx1 kinase activity in vivo correlates with the activation of XCdc25 and therefore p34 cdc2 kinase, since early studies show that Plk kinase activity peaks during mitosis but not at the onset of mitosis and that Plk associates with mitotic spindle (29). In fact, we have also observed that the Plk antigen level is rather low during late S and G 2 stages of the cell cycle, and it reaches the highest level at the mitotic metaphase (Fig.  1D). On the other hand, we have demonstrated that prk mRNA expression and its kinase activity reach the peak level during late S and S/G 2 stages (Figs. 1 and 2), correlating with the completion of DNA synthesis and the activation of cyclin-dependent kinase p34 cdc2 . In addition, we have shown that Prk is capable of phosphorylating Cdc25C in vitro, suggesting that Prk may truly represent an important protein Ser/Thr kinase regulating the onset of mitosis/meiosis in animal cells.