- Email: [email protected]
cyclin B through phosphorylation and activation of the cdc25C phosphatase
Cellular Signalling 12 (2000) 405–411 http://www.elsevier.com/locate/cellsig
The human polo-like kinase, PLK, regulates cdc2/cyclin B through phosphorylation and activation of the cdc25C phosphatase Amy K. Roshak*, Elizabeth A. Capper✩, Christina Imburgia, James Fornwald#, Gilbert Scott#, Lisa A. Marshall The Department of Immunology and the #Department of Biotechnology and Genetics, SmithKline Beecham Pharmaceuticals, King of Prussia, PA 19406, USA Received 8 January 2000; accepted 21 March 2000
Abstract Entry into mitosis by mammalian cells is triggered by the activation of the cdc2/cyclin B holoenzyme. This is accomplished by the specific dephosphorylation of key residues by the cdc25C phosphatase. The polo-like kinases are a family of serine/ threonine kinases which are also implicated in the control of mitotic events, but their exact regulatory mechanism is not known. Recently, a Xenopus homologue, PLX1, was reported to phosphorylate and activate cdc25, leading to activation of cdc2/cyclin B. Jurkat T leukemia cells were chemically arrested and used to verify that PLK protein expression and its phosphorylation state is regulated with respect to cell cycle phase (i.e., protein is undetectable at G1/S, accumulates at S phase and is modified at G2/M). Herein, we show for the first time that endogenous human PLK protein immunoprecipitated from the G2/M-arrested Jurkat cells directly phosphorylates human cdc25C. In addition, we demonstrate that recombinant human (rh)PLK also phosphorylates rhcdc25C in a time- and concentration-dependent manner. Phosphorylation of endogenous cdc25C and recombinant cdc25C by PLK resulted in the activation of the phosphatase as assessed by dephosphorylation of cdc2/cyclin B. These data are the first to demonstrate that human PLK is capable of phosphorylating and positively regulating human cdc25C activity, allowing cdc25C to dephosphorylate inactive cdc2/cyclin B. As this event is required for cell cycle progression, we define at least one key regulatory mode of action for human PLK in the initiation of mitosis. 2000 Published by Elsevier Science Inc. All rights reserved. Keywords: Cyclin-dependent kinase; Phosphatase; Proliferation; Mitosis; Cell cycle
1. Introduction Control of the eukaryotic cell cycle is mediated through a series of highly coordinated biochemical processes, which include a number of complex phosphorylation/dephosphorylation cascades. In particular, cell entry into mitosis is controlled by the M phase promoting factor (MPF); a complex of the cyclin dependent kinase, cdc2; and its regulatory subunit, cyclin B. It is well established in both higher and lower eukaryotes that the activity of the complex is tightly regulated by reversible phosphorylation events [1]. Phosphorylation of threonine 14 and tyrosine 15 on cdc2 by the myt1 and wee-1 kinases renders cdc2/cyclin B inactive at the G2/M border [2]. The dual specificity phosphatase cdc25C is responsible for dephosphorylation of both residues, resulting in acti✩ Amy K. Roshak and Elizabeth A. Capper contributed equally to the manuscript. * Corresponding author. Tel.: 610-270-4969; fax: 610-270-5381. E-mail address: [email protected] (A.K. Roshak)
vation of the cdc2 kinase and cell cycle progression into mitosis [3]. Consistent with the tight regulation of this process, cdc25C activity is also controlled by phosphorylation [4,5]. Cdc2/cyclin B has been reported to phosphorylate cdc25C in vitro, suggesting the existence of an autoamplification loop [4,6,7]. However, phosphorylation and activation of cdc25C occurs in the absence of cdc2 kinase activity, indicating that another kinase is involved in the initial activation of cdc25C [8]. The kinase responsible for this activation is not known. The polo kinases are a family of distinct serine/threonine protein kinases which also play an important role in cell cycle control. The activity of these kinases has been shown to be required for regulating the mechanics of mitosis in multiple species. For example, Drosophila polo mutations are embryonic lethal due to aberrant mitotic phenotypes, while disruption of yeast cdc5 activity leads to mitotic arrest [9,10,11]. Recently, Kumagai and Dunphy demonstrated in Xenopus oocyte extracts that the Xenopus homologue PLX1 regulates the cdc25
0898-6568/00/$ – see front matter 2000 Published by Elsevier Science Inc. All rights reserved. PII: S0898-6568(00)00080-2
406
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
phosphatase activity directly by phosphorylating its amino terminus [12]. This corresponded with dephosphorylation and activation of the cdc2/cyclin B complex. These data would suggest a direct role for Xenopus PLX1 in the regulation of MPF and mitotic progression in the oocyte. A human polo kinase homologue, PLK, exists; its mRNA expression correlates with cellular mitotic index [13]. Immunolocalization studies show PLK associated with several components of the mitotic spindle apparatus, and antibody microinjection has confirmed its importance in mitotic events and cytokinesis [14,15,16]. In cell cycle phase arrested human cells, PLK protein is low to undetectable at G1/S, accumulates during the S phase, is modified by phosphorylation during G2/M, and is rapidly degraded after mitosis [14]. In addition, activation of PLK is reported to occur in a time frame close to the onset of cdc2/cyclin B activation [14]. While studies by several investigators have clearly demonstrated the critical role of human PLK in the successful completion of chromosomal segregation, the potential involvement of human PLK in the initiation of the G2/M transition is not known. Although the recent findings in Xenopus support PLX1 regulation of G2/M transition in embryonic extracts, controversy exists as to whether the same regulatory mechanisms exist in human somatic cells [12]. With the goal of elucidating the exact mechanism for human PLK regulation of mitotic progression, we have conducted studies utilizing recombinant, as well as human cell-derived, PLK and cdc25C. Herein, we show for the first time that human cdc25C is a target of human PLK phosphorylation. Further, we demonstrate that this interaction results in phosphatase activation and subsequent dephosphorylation of human cdc2/cyclin B, which is known to result in mitotic progression.
2. Materials and methods 2.1. Jurkat T cell culture, chemical cell cycle synchronization and flow cytometry analysis Jurkat T cell lymphoma was acquired from American Type Culture Collection (Rockville, MD). The cells were maintained at a density of 1 to 5 ⫻ 105 cells/mL and grown as suspension cultures in RPMI 1640 with 2 mM L-glutamine, 1.5 g/L NaHCO3, and 10% FBS. For drug-induced cell synchronization, Jurkat cells at 3 ⫻ 105 cells/mL were treated with 3 mM hydroxyurea for 16 h (G1/S arrest) and 10 M nocodazole for 12 h (G2/ M arrest). The cells were also treated with 2 mM thymidine for 17 h, released for 6 h in culture media (no thymidine), then re-treated with 2 mM thymidine for 15 h (broad S-phase arrest). Following the treatments, the cells were centrifuged, the media were removed, and the cells were resuspended to 2 ⫻ 107 in 1% paraformaldehyde in PBS. They were placed at 4⬚C for 15
min, then resuspended in 1 mL of a propidium iodide solution containing 250 g/mL propidium iodide and 250 g/mL RNAse A. The stained cells were vortexed, then incubated at room temperature for 30 min in darkness. The fluorescent intensities were measured using a Becton-Dickinson FACs scanner. Forward scatter versus side light scatter and fluorescent area versus width were used to gate for intact, single cells to evaluate for DNA content. Identification of haploid/diploid cell populations were based on 20,000 gated cell events. 2.2. Immunoprecipitation of human PLK and cdc25C from synchronized Jurkat cells Protein G agarose beads (Gibco-BRL, Gaithersburg, MD) were washed in immunoprecipitation buffer containing 0.1% Triton X-100, 5 mM NaF, 5 mM EDTA, and 5 mM EGTA in TBS (pH 7.4) with protease inhibitors (1 mM PMSF, 200 M leupeptin, 20 g/mL aprotinin and 20 g/mL soybean trypsin inhibitor). Beads were linked with rabbit anti-human PLK antibody (Zymed Laboratories Inc., San Francisco, CA) or rabbit anti-human cdc25C antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Antibodies were linked to beads by incubation with 20 mM dimethylpimelimadate (DMP) for 30 min at room temperature. Reactions were stopped by washing in 0.2 M ethanolamine (pH 9.0). Unbound antibodies were removed by washing with 100 mM glycine (pH 3.0). Linked beads were stored at 4⬚C in PBS with 0.1% merthiolate until used. Antibody complexed beads were equilibrated in homogenization buffer at a volume of 5 times the bead volume. Complexed beads were placed in cytosol preparations (30– 100 g, prepared as previously described [17] of the cell cycle-arrested Jurkat cells and incubated at 4⬚C for 5 h with continual end-over-end rocking. Suspensions were cleared by centrifugation at 5000 ⫻ g for 2 min and the immunodepleted cytosols were removed. Beads were washed 3 times in immunoprecipitation buffer, as above. For evaluation of co-immunoprecipitating proteins, the beads were pelleted and complexes were solubilized by the addition of 2X Laemmeli sample buffer. For in vitro kinase assays, beads were washed by a 10-min incubation in homogenization buffer at 4⬚C, pelleted as above, resuspended to a 50% slurry in assay buffer, and stored on ice until assays were run. A volume of final bead preparation equal to that used in the assays was taken and analyzed by Western blot (as described below) to verify the effectiveness of the immunoprecipitation. 2.3. Immunoblot analysis Jurkat cytosols (50 g protein), immunoprecipitates or dephosphorylation assays were analysed by SDSPAGE (10% tris-glycine gels; Biorad, Hercules, CA). Proteins were transferred to nitrocellulose paper, incubated with either rabbit anti-human PLK antibody (1:1000; Zymed Laboratories Inc.), rabbit anti-human cdc25C antibody (1:500), mouse anti-human cdc2 anti-
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
body (1:1000) or mouse anti-cyclinB antibody (1:1000) (Santa Cruz Biotechnology, Inc.) and then incubated with either rabbit anti-mouse or donkey anti-rabbit IgG conjugated to horseradish peroxidase (1:3000; Amersham, Arlington Heights, IL). Detection of immunoreactive bands was carried out using the ECL Western blotting system (Amersham). 2.4. Expression and purification of human cell cycle proteins 2.4.1. Human gst-PLK A gst-PLK expression vector was constructed which included the glutathione-S-transferase gene fused to the amino terminus of the full-length human PLK gene via a linker containing a thrombin cleavage site. The construct was cloned into the baculovirus expression vector, pFASTBAC, which was used to generate viral stock for subsequent infection. Spodoptera frugiperda cells (Sf9) were infected with virus-expressing gst-PLK and grown for three days prior to harvesting. Cell pellets were lysed in 20 mM Tris-HCl (pH 8.0), 137 mM NaCl, 2 mM EDTA, 10% glycerol, 1% Triton X-100, 5 mM DTT, 2 g/mL E-64, 1 mM AEBSF, and 1 g/mL pepstatin A (lysis buffer) by an Avestin cell disrupter. Cell lysates were clarified by centrifugation at 28,000 ⫻ g at 4⬚C for 1 h. The supernatant was mixed with 10 mL of glutathione Sepharose 4B (Pharmacia Biotech, Sweden) and gently shaken at 4⬚C for 2 h. The resin was packed onto a Pharmacia XK26/50 column and unbound material was washed out with 20 mM Tris-HCl (pH 7.0), 100 mM NaCl, 12 mM MgCl2 and 1 mM DTT (wash buffer). Proteins bound to the resin were then eluted with 50 mM Tris-HCl (pH 8.0), 10 mM glutathione, and 0.05% Brij-35 (elution buffer). Fractions containing proteins were pooled and concentrated using an Amicon ultrafiltration system with YM30 membrane. Based on SDSPAGE analysis and Bio-Rad protein assay, the PLK kinase was purified to ⵑ80% purity, and the identity of the purified protein was confirmed by N-terminal sequence analysis. Remaining impurities consisted predominantly of free gst. 2.4.2. Human gst-cdc25C The gst-cdc25C E. coli expression construct was the generous gift of Dr. Laurent Meijer (Roscoff University, Roscoff, France). Briefly, the construct contained the glutathione-S-transferase gene fused to the amino terminus of full-length cdc25C by a linker containing a thrombin cleavage site. This was subcloned into the baculovirus expression vector, pFASTBAC, which was used to generate viral stock for subsequent infection as described above. The gst-cdc25C fusion protein was purified by the same procedures used for gst-PLK purification. SDS-PAGE analysis indicated that the band corresponding to the predicted molecular weight of fulllength gst-cdc25C was ⵑ30% of the mixture. The identity of this protein was confirmed by N-terminal amino
407
acid sequence analysis. The remaining contaminants were primarily composed of free gst and cdc25C degradation products. A ⵑ55kDa protein present in the preparation was confirmed by sequencing as a C-terminal truncated form of gst-cdc25C. For some studies the gst tag was removed from rhcdc25C by thrombin cleavage using the Novagen Thrombin Cleavage Kit as per the manufacturer’s instructions. 2.5. PLK kinase assay Unless otherwise indicated, each kinase assay contained either gst-rhcdc25C (1 g), gst cleaved rhcdc25C (1 g), or cdc25C immunoprecipitated from 50 g of G1/S arrested cell cytosol as substrate, to which 1-2 Ci/ assay of [32P]-␥ATP, 10 M cold ATP and kinase reaction buffer containing 20 mM HEPES (pH 7.4), 50 mM KCL, 10 mM MgCl2, 1 mM EGTA, 0.5 mM DTT were added. The reaction was initiated by addition of the indicated amount of either rhPLK or immunoprecipitated PLK from nocodazole-treated, G2/M synchronized Jurkat cells and was allowed to proceed for the indicated time. Kinase activity was terminated by the addition of EDTA to 25 mM. SDS-PAGE sample buffer was added, reactions were boiled for 5 min and samples were electrophoresed through 10% tris-glycine gels. Gels were exposed to phosphorscreens and images obtained by phosphorimager (Storm 860, Molecular Dynamics, Sunnyvale, CA). 2.6. Cdc2/cyclin B dephosphorylation assays A catalytically inactive, doubly phosphorylated recombinant human (rh)cdc2/cyclin B complex was prepared by treatment with rh-myt1 and rh-wee1 kinases. Briefly, 8 ng/l of rhcdc2/cyclin B was incubated with 3 ng/l rh-myt1 kinase and 8 ng/l rh-wee1 kinase with 1 mM ATP for 1 h at 37⬚C in buffer containing 50 mM HEPES, 10 mM MgCl2 and 1 mM DTT. For assays 80 ng total cdc2/cyclin B complex in reaction mix was incubated with or without rhPLK (2–4 g), rhcdc25C (4 g) or cdc25C immunoprecipitated from 50 g cytosol of hydroxyurea-treated Jurkat T cells at 37⬚C in PLK kinase assay buffer with 10 M ATP as described above. Reactions were stopped by the addition of EDTA to 25 mM. SDS-Sample buffer was added, and samples were boiled and subjected to SDS-PAGE and Western blotting as described above. All experiments were conducted 2 times and figures show results of one representative study. 3. Results and discussion 3.1. Recombinant human (rh) cdc25C is a substrate for rhPLK Recombinant human cdc25C and PLK gst-fusion proteins (ⵑ87 kDa and 97 kDa, respectively) were generated and production of active PLK kinase was con-
408
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
Fig. 1. Recombinant human (rh)cdc25C is a substrate for rhPLK. rhPLK (0.5 g /assay; 25 nM) was incubated alone or in the presence of increasing concentrations of rhcdc25C (0.1–3.0 g/reaction) for 60 min at 37⬚C. Reactions were terminated by the addition of 25 mM EDTA, resolved by SDS-PAGE, and substrate phosphorylation was visualized by phosphorimager.
firmed by evaluating phosphorylation of ␣-casei, a ubiquitous substrate, as has been previously reported (data not shown) [15]. Significant phosphorylation of gst alone was not observed (data not shown). RhPLK1 was then incubated with rhcdc25C as described under Materials and Methods. Fig. 1 shows that addition of increasing amounts of rhcdc25C (0.1–3.0 g/reaction) to the kinase reaction resulted in a marked concentration-dependent increase in rhcdc25C phosphorylation. In addition to phosphorylation of the full-length gst-cdc25C fusion protein, rhPLK also actively phosphorylated the ⵑ55 kDa C- terminal truncated cdc25C protein. Incubation of rhPLK1 (0.5 g/reaction; 25nM) alone appears to result in some autophosphorylation which is consistent with the findings of other investigators [15]. Fig. 2 shows that rhPLK phosphorylation of rhcdc25C increases over time and can be inhibited by the addition of the nonhydrolyzable ATP analogue, ATP␥S. Together, these data demonstrate that rhcdc25C is a substrate for rhPLK and provide the first evidence that human PLK may participate in the regulation of cdc25C phosphatase activity. 3.2. Evaluation of PLK and cdc25C protein expression in cell cycle phase arrested Jurkat T cells Human Jurkat T leukemia cells were chemically synchronized at either G1/S, S phase, or G2/M as described in Materials and Methods and cell cycle phase was con-
firmed by flow cytometry (Fig. 3a). Each population was then fractionated into a cytosol and 100,000 ⫻ g particulate fraction and PLK and cdc25C levels were evaluated in the cytosol by Western analysis. Fig. 3c shows that, as has been previously described, cdc25C protein expression does not change over the course of the cell cycle [5]. Instead, the phosphatase is modified by multiple phosphorylation events, which is evident by its reduced mobility (Fig. 3c) [5]. Consistent with reports in other human cell types [14,16], PLK protein is low to undetectable at G1/S (Fig. 3b). Protein levels begin to accumulate at S phase and are prominent at G2/M. The mobility of PLK in G2/M arrested fractions is retarded compared to the other cell cycle fractions, indicating protein modification. Human PLK is thought to be modified and activated by phosphorylation by an asyet unidentified kinase [14]. This is consistent with the higher molecular weight band observed in the G2/M arrested fraction and its enhanced activity in cells at the G2/M transition [14,15]. This fraction is, therefore, a good source for activated native human PLK. Human PLK and cdc25C were immunoprecipitated from the cytosol (100 g) of nocodazole treated, G2/M arrested Jurkat cells. Anti-PLK antibody was able to efficiently precipitate PLK (ⵑ67 kDa) from these cells, but no cdc25C was detectable in PLK immune complexes by Western analysis (Fig. 4a). Similarly, cdc25C, but not PLK, was found in the complexes immunoprecipitated with anti-cdc25C linked beads (Fig. 4b). No cdc2 or cyclin B was detectable via Western blot in either the PLK or cdc25C immune complexes, yet they were detectable in an amount of total cytosolic extract equal to that which was immunodepleted (Fig. 4c and d). These data indicate that while these proteins may participate in a regulatory amplification loop, they do not appear to exist in a tightly associated multimeric complex. 3.3. Native human PLK phosphorylates cdc25C Native PLK protein immunoprecipitated from 50 g of G2/M arrested Jurkat cytosol was evaluated for its ability to phosphorylate rhcdc25C. In contrast to the recombinant kinase, incubation of immunoprecipitated
Fig. 2. rhPLK phosphorylation of rhcdc25C increases with time. rhPLK (0.5 g/assay; 25 nM) was incubated with rhcdc25C (1 g/assay) for the indicated times at 37⬚C in the presence or absence of ATP␥S (1mM). Reactions were terminated by the addition of 25 mM EDTA, resolved by SDS-PAGE and substrate phosphorylation was visualized by phosphorimager.
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
409
Fig. 3. Evaluation of human PLK and cdc25C protein levels in cell cycle phase synchronized Jurkat T cells. (a) Asynchronous, G1/S, S or G2/M phase synchronized Jurkat cells were prepared as described under Materials and Methods. Cells were fixed and stained with propidium iodide; cell cycle phase was evaluated by flow cytometry. (b) shows Western immunoblot analysis of PLK protein levels in either asynchronous or synchronized Jurkat cell cytosols (50 g/Lane). (c) shows cdc25C protein levels in the same cell cycle synchronized Jurkat cell cytosols (50 g/ Lane).
Fig. 4. Immunoprecipitation of PLK and cdc25C from cell cycle synchronized Jurkats. Anti-human PLK or anti-human cdc25C linked protein G agarose beads were used to immunoprecipitate PLK or cdc25C from 100 g G2/M synchronized Jurkat cell cytosols as described in Materials and Methods. A control precipitation with no primary antibody was also conducted. Precipitated complexes were resolved by SDS-PAGE, and PLK (a), cdc25C (b), cdc2 (c), or cyclin B (d) protein levels were monitored by western immunoblot analysis.
PLK alone in the kinase reaction did not result in detectable autophosphorylation (Fig. 5), suggesting that the kinase in this G2/M cell preparation is fully phosphorylated. This finding is consistent with the Western blot analysis of the nocodazole treated (Fig. 3b) fraction that indicates PLK is phosphorylated and thus activated at the G2/M border. A mock immunoprecipitation of the same amount of G2/M arrested cytosol using rabbit IgG control antibodies was also performed. This reaction was used as a control to evaluate non-specific phosphorylation events. Fig. 5 shows that addition of the mock immunoprecipitate to gst-cleaved-rhcdc25C in a kinase reaction did not result in phosphorylation of the phosphatase. In contrast, incubation of immunoprecipitated PLK with gst-rhcdc25C (Fig. 5, 1 g/reaction) resulted in a marked phosphorylation of the cdc25C substate. As observed with the recombinant PLK, native PLK efficiently phosphorylated both the 87 kDa gst-fusion protein and the 55 kDa C-terminal truncated cdc25C protein present as a contaminant in the gst-rhcdc25C preparation. When native PLK was incubated with the gst-cleaved cdc25 the appropriately sized, full-length cdc25C (57 kDa) was phosphorylated (Fig. 5). These data confirm and strengthen our findings using recombinant PLK and are the first data utilizing human-cell derived PLK to demonstrate that cdc25C is a biologi-
410
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
Fig. 5. Native human PLK phosphorylates rhcdc25C. Anti-human PLK antibody or non-specific rabbit IgG antibody-linked protein G agarose beads were incubated with G2/M arrested Jurkat cytosols (50 g/reaction). PLK immunoprecipitates were incubated alone or in the presence of gst-rhcdc25C (1 g/reaction) or gst-cleaved rhcdc25C (1 g/reaction) for 60 min at 37⬚C in kinase reaction buffer. IgG control immunoprecipitations were incubated with gst-cleaved rhcdc25C (1 g/reaction) for 60 min at 37⬚C in kinase reaction buffer. Reactions were terminated by the addition of 25 mM EDTA, resolved by SDS-PAGE and substrate phosphorylation visualized by phosphorimager.
cally relevant substrate of human PLK. Moreover, these data strongly suggest that the human polo-like kinase may participate in cdc2/cyclin B regulation through phosphorylation and subsequent activation of cdc25C. 3.4. PLK phosphorylation of human cdc25C is required for activation of the phosphatase To assess the potential biological relevance of PLK phosphorylation of cdc25C, studies were conducted to evaluate the ability of PLK-treated and -untreated cdc25C to dephosphorylate its substrate, cdc2/cyclin B. Inactive threonine 14- and tyrosine 15-phosphorylated cdc2/cyclin B (designated cdc2-PP, Fig. 6a and b) was prepared by incubation with both the myt1 and wee1 kinases as described in Materials and Methods. The doubly phosphorylated cdc2/cyclin B complex was incubated (30 min at 37⬚C) alone or together with either
rhPLK, rhcdc25C or both proteins together. Activation of cdc25C was assessed by dephosphorylation of cdc2, which was monitored by a change in electrophoretic mobility. Fig. 6a shows that the rhPLK alone had no effect on the phosphorylation state of the cdc2 molecule. Incubation with the rhcdc25C alone resulted in dephosphorylation (60% as assessed by densitometry) of the cdc2 protein, yielding both the single and doubly dephosphorylated forms (designated cdc2-P and cdc2 respectively; Fig. 6a and b). This result is not surprising, as the rhcdc25C phosphatase undergoes some activation by the insect host cells during expression prior to purification. Treatment of rhcdc25C with PLK resulted in a substantial potentiation of the phosphatase activity of cdc25C, as evidenced by significantly more cdc2 protein in the dephosphorylated state (80% as assessed by densitometry). This activation was found to be time depen-
Fig. 6. PLK phosphorylation of rhcdc25C or native cdc25C results in phosphatase activation. (a) Catalytically inactive, doubly phosphorylated cdc2/cyclin B complexes (80 ng/reaction) were incubated for 30 min at 37⬚C either alone, with 4 g of rhPLK alone, 4 g rhcdc25C alone, or the two enzymes together. (b) Inactive cdc25C was immunoprecipitated from G1/S arrested Jurkat cytosols (50 g/ reaction) and incubated for 2 h at 37⬚C with cdc2/cyclin B complexes either alone, with 2 g rhPLK, or with rhPLK in the presence of 5 mM NaF. All reactions were terminated by the addition of 25 mM EDTA and resolved by SDS-PAGE. The phosphorylation state of cdc2 was assessed by Western immunoblot analysis.
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
dent with maximal activity observed between 30 and 60 min (data not shown). To assess whether phosphorylation by PLK is required for activation of the endogenous, inactive phosphatase, cdc25C was immunoprecipitated from the cytosol of G1/S arrested Jurkats (50 g) and incubated with the fully phosphorylated inactive cdc2/cyclin B complex with or without rhPLK (30 min at 37⬚C). Fig. 6b clearly demonstrates that incubation of the inactive cdc2/cyclin B complex with G1/S-derived cdc25C alone had no effect on the phosphorylation state of cdc2. This was expected, as the cdc25C phosphatase is inactive at this phase of the cell cycle. However, addition of rhPLK to the reaction induced a marked dephosphorylation of the cdc2 protein. The dephosphorylation could be completely abrogated by the addition of the phosphatase inhibitor, NaF. These findings demonstrate that native human cdc25C is directly activated by PLK phosphorylation, which leads to the dephosphorylation of cdc2/cyclin B, an absolute requirement for G2/M transition. In summary, we have shown for the first time that human cdc25C is a substrate for the human polo-like kinase, PLK. Moreover, this interaction was shown to be biologically relevant as it results in the activation of the cdc25C phosphatase and subsequent dephosphorylation of cdc2/cyclin B. These observations suggest that human PLK is instrumental not only in the orchestration of chromosomal segregation but also the initiation of the G2/M transition in somatic cells. Cdc25C catalyzed dephosphorylation of the threonine 14 and tyrosine 15 residues on cdc2 is known to trigger activation of the MPF (cdc2/cyclin B) and initiate a cell’s entry into mitosis. While cdc2/cyclin B is also capable of phosphorylating and activating cdc25C, phosphatase activation occurs in the absence of cdc2, indicating the existence of an additional kinase in cdc25C regulation [8]. It has been suggested that a “trigger kinase” is responsible for the initial activation of cdc25C and the MPF autoamplification loop. Taken together, the data presented herein
411
strongly support a hypothesis for human PLK as the key activator of cdc25C phosphatase, an obligatory step in the initiation of G2/M transition and progression into mitosis. Acknowledgments The authors wish to thank Shannon Reed for technical assistance, Dr. Jeffrey R Jackson and Maureen Ho for providing inactive cdc2/cyclin B complex and Drs. Ruth Mayer and Patty Podolin for critical review. References [1] Piwnica-Worms H. J Lab Clin Med 1996;128:350–4. [2] Mueller PR, Coleman TR, Kumigai A, Dunphy WG. Science 1995;270:86–90. [3] Kumagai A, Dunphy WG. Cell 1991;64:903–14. [4] Hoffmann I, Clarke PR, Marcote MJ, Karsenti E, Draetta G. EMBO 1993;12:53–63. [5] Gabrielli BG, Clark J, McCormack AK, Ellem, KAO. J Bio Chem 1997;272:28607–14. [6] Izumi T, Maller JL. Mol Bio Cell 1993;4:1337–50. [7] Stausfeld U, Fernandez A, Capony J-P, Girard F, Lautredou N, Derancourt J, Labbe J-C, Lamb, NJC. J Bio Chem 1994;269: 5989–6000. [8] Izumi T, Maller JL. Mol Bio Cell 1995;6:215–26. [9] Sunkel CE, Glover DM. J Cell Sci 1988;89:25–38. [10] Kitada K, Johnson A, Johnston L, Sugino A. Mol Cell Bio 1993; 13;4445–57. [11] Llamazares S, Moreira A, Tavares A, Girdham C, Spruce BA, Gonzalez C, Karess RE, Glover DM, Sunkel CE. Genes and Dev 1991;5:2153–65. [12] Kumagai A, Dunphy WG. Science 1996;273:1377–1380. [13] Holtrich U, Wolf G, Brauninger A, Karn T, Bohme B, Rubsamen-Waigmann H, Strebhardt K. Proc Natl Acad Sci USA 1994;91:1736–40. [14] Hamanaka R, Smith MR, O’Connor PM, Maloid S, Mihalic K, Spivak JL, Longo DL, Ferris DK. J Bio Chem 1995;270:21086–91. [15] Golsteyn RM, Mundt KE, Fry AM, Nigg EA. J Cell Bio 1995;129:1617–28. [16] Lee KS, Yuan YLO, Kuriyama R, Erikson RL. Mol Cell Bio 1995;15:7143–51. [17] Roshak AK, Jackson JR, McGough K, Chabot-Fletcher M, Mochan E, Marshall LA. J Bio Chem 1996;271:31496–501.
The human polo-like kinase, PLK, regulates cdc2/cyclin B through phosphorylation and activation of the cdc25C phosphatase Amy K. Roshak*, Elizabeth A. Capper✩, Christina Imburgia, James Fornwald#, Gilbert Scott#, Lisa A. Marshall The Department of Immunology and the #Department of Biotechnology and Genetics, SmithKline Beecham Pharmaceuticals, King of Prussia, PA 19406, USA Received 8 January 2000; accepted 21 March 2000
Abstract Entry into mitosis by mammalian cells is triggered by the activation of the cdc2/cyclin B holoenzyme. This is accomplished by the specific dephosphorylation of key residues by the cdc25C phosphatase. The polo-like kinases are a family of serine/ threonine kinases which are also implicated in the control of mitotic events, but their exact regulatory mechanism is not known. Recently, a Xenopus homologue, PLX1, was reported to phosphorylate and activate cdc25, leading to activation of cdc2/cyclin B. Jurkat T leukemia cells were chemically arrested and used to verify that PLK protein expression and its phosphorylation state is regulated with respect to cell cycle phase (i.e., protein is undetectable at G1/S, accumulates at S phase and is modified at G2/M). Herein, we show for the first time that endogenous human PLK protein immunoprecipitated from the G2/M-arrested Jurkat cells directly phosphorylates human cdc25C. In addition, we demonstrate that recombinant human (rh)PLK also phosphorylates rhcdc25C in a time- and concentration-dependent manner. Phosphorylation of endogenous cdc25C and recombinant cdc25C by PLK resulted in the activation of the phosphatase as assessed by dephosphorylation of cdc2/cyclin B. These data are the first to demonstrate that human PLK is capable of phosphorylating and positively regulating human cdc25C activity, allowing cdc25C to dephosphorylate inactive cdc2/cyclin B. As this event is required for cell cycle progression, we define at least one key regulatory mode of action for human PLK in the initiation of mitosis. 2000 Published by Elsevier Science Inc. All rights reserved. Keywords: Cyclin-dependent kinase; Phosphatase; Proliferation; Mitosis; Cell cycle
1. Introduction Control of the eukaryotic cell cycle is mediated through a series of highly coordinated biochemical processes, which include a number of complex phosphorylation/dephosphorylation cascades. In particular, cell entry into mitosis is controlled by the M phase promoting factor (MPF); a complex of the cyclin dependent kinase, cdc2; and its regulatory subunit, cyclin B. It is well established in both higher and lower eukaryotes that the activity of the complex is tightly regulated by reversible phosphorylation events [1]. Phosphorylation of threonine 14 and tyrosine 15 on cdc2 by the myt1 and wee-1 kinases renders cdc2/cyclin B inactive at the G2/M border [2]. The dual specificity phosphatase cdc25C is responsible for dephosphorylation of both residues, resulting in acti✩ Amy K. Roshak and Elizabeth A. Capper contributed equally to the manuscript. * Corresponding author. Tel.: 610-270-4969; fax: 610-270-5381. E-mail address: [email protected] (A.K. Roshak)
vation of the cdc2 kinase and cell cycle progression into mitosis [3]. Consistent with the tight regulation of this process, cdc25C activity is also controlled by phosphorylation [4,5]. Cdc2/cyclin B has been reported to phosphorylate cdc25C in vitro, suggesting the existence of an autoamplification loop [4,6,7]. However, phosphorylation and activation of cdc25C occurs in the absence of cdc2 kinase activity, indicating that another kinase is involved in the initial activation of cdc25C [8]. The kinase responsible for this activation is not known. The polo kinases are a family of distinct serine/threonine protein kinases which also play an important role in cell cycle control. The activity of these kinases has been shown to be required for regulating the mechanics of mitosis in multiple species. For example, Drosophila polo mutations are embryonic lethal due to aberrant mitotic phenotypes, while disruption of yeast cdc5 activity leads to mitotic arrest [9,10,11]. Recently, Kumagai and Dunphy demonstrated in Xenopus oocyte extracts that the Xenopus homologue PLX1 regulates the cdc25
0898-6568/00/$ – see front matter 2000 Published by Elsevier Science Inc. All rights reserved. PII: S0898-6568(00)00080-2
406
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
phosphatase activity directly by phosphorylating its amino terminus [12]. This corresponded with dephosphorylation and activation of the cdc2/cyclin B complex. These data would suggest a direct role for Xenopus PLX1 in the regulation of MPF and mitotic progression in the oocyte. A human polo kinase homologue, PLK, exists; its mRNA expression correlates with cellular mitotic index [13]. Immunolocalization studies show PLK associated with several components of the mitotic spindle apparatus, and antibody microinjection has confirmed its importance in mitotic events and cytokinesis [14,15,16]. In cell cycle phase arrested human cells, PLK protein is low to undetectable at G1/S, accumulates during the S phase, is modified by phosphorylation during G2/M, and is rapidly degraded after mitosis [14]. In addition, activation of PLK is reported to occur in a time frame close to the onset of cdc2/cyclin B activation [14]. While studies by several investigators have clearly demonstrated the critical role of human PLK in the successful completion of chromosomal segregation, the potential involvement of human PLK in the initiation of the G2/M transition is not known. Although the recent findings in Xenopus support PLX1 regulation of G2/M transition in embryonic extracts, controversy exists as to whether the same regulatory mechanisms exist in human somatic cells [12]. With the goal of elucidating the exact mechanism for human PLK regulation of mitotic progression, we have conducted studies utilizing recombinant, as well as human cell-derived, PLK and cdc25C. Herein, we show for the first time that human cdc25C is a target of human PLK phosphorylation. Further, we demonstrate that this interaction results in phosphatase activation and subsequent dephosphorylation of human cdc2/cyclin B, which is known to result in mitotic progression.
2. Materials and methods 2.1. Jurkat T cell culture, chemical cell cycle synchronization and flow cytometry analysis Jurkat T cell lymphoma was acquired from American Type Culture Collection (Rockville, MD). The cells were maintained at a density of 1 to 5 ⫻ 105 cells/mL and grown as suspension cultures in RPMI 1640 with 2 mM L-glutamine, 1.5 g/L NaHCO3, and 10% FBS. For drug-induced cell synchronization, Jurkat cells at 3 ⫻ 105 cells/mL were treated with 3 mM hydroxyurea for 16 h (G1/S arrest) and 10 M nocodazole for 12 h (G2/ M arrest). The cells were also treated with 2 mM thymidine for 17 h, released for 6 h in culture media (no thymidine), then re-treated with 2 mM thymidine for 15 h (broad S-phase arrest). Following the treatments, the cells were centrifuged, the media were removed, and the cells were resuspended to 2 ⫻ 107 in 1% paraformaldehyde in PBS. They were placed at 4⬚C for 15
min, then resuspended in 1 mL of a propidium iodide solution containing 250 g/mL propidium iodide and 250 g/mL RNAse A. The stained cells were vortexed, then incubated at room temperature for 30 min in darkness. The fluorescent intensities were measured using a Becton-Dickinson FACs scanner. Forward scatter versus side light scatter and fluorescent area versus width were used to gate for intact, single cells to evaluate for DNA content. Identification of haploid/diploid cell populations were based on 20,000 gated cell events. 2.2. Immunoprecipitation of human PLK and cdc25C from synchronized Jurkat cells Protein G agarose beads (Gibco-BRL, Gaithersburg, MD) were washed in immunoprecipitation buffer containing 0.1% Triton X-100, 5 mM NaF, 5 mM EDTA, and 5 mM EGTA in TBS (pH 7.4) with protease inhibitors (1 mM PMSF, 200 M leupeptin, 20 g/mL aprotinin and 20 g/mL soybean trypsin inhibitor). Beads were linked with rabbit anti-human PLK antibody (Zymed Laboratories Inc., San Francisco, CA) or rabbit anti-human cdc25C antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Antibodies were linked to beads by incubation with 20 mM dimethylpimelimadate (DMP) for 30 min at room temperature. Reactions were stopped by washing in 0.2 M ethanolamine (pH 9.0). Unbound antibodies were removed by washing with 100 mM glycine (pH 3.0). Linked beads were stored at 4⬚C in PBS with 0.1% merthiolate until used. Antibody complexed beads were equilibrated in homogenization buffer at a volume of 5 times the bead volume. Complexed beads were placed in cytosol preparations (30– 100 g, prepared as previously described [17] of the cell cycle-arrested Jurkat cells and incubated at 4⬚C for 5 h with continual end-over-end rocking. Suspensions were cleared by centrifugation at 5000 ⫻ g for 2 min and the immunodepleted cytosols were removed. Beads were washed 3 times in immunoprecipitation buffer, as above. For evaluation of co-immunoprecipitating proteins, the beads were pelleted and complexes were solubilized by the addition of 2X Laemmeli sample buffer. For in vitro kinase assays, beads were washed by a 10-min incubation in homogenization buffer at 4⬚C, pelleted as above, resuspended to a 50% slurry in assay buffer, and stored on ice until assays were run. A volume of final bead preparation equal to that used in the assays was taken and analyzed by Western blot (as described below) to verify the effectiveness of the immunoprecipitation. 2.3. Immunoblot analysis Jurkat cytosols (50 g protein), immunoprecipitates or dephosphorylation assays were analysed by SDSPAGE (10% tris-glycine gels; Biorad, Hercules, CA). Proteins were transferred to nitrocellulose paper, incubated with either rabbit anti-human PLK antibody (1:1000; Zymed Laboratories Inc.), rabbit anti-human cdc25C antibody (1:500), mouse anti-human cdc2 anti-
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
body (1:1000) or mouse anti-cyclinB antibody (1:1000) (Santa Cruz Biotechnology, Inc.) and then incubated with either rabbit anti-mouse or donkey anti-rabbit IgG conjugated to horseradish peroxidase (1:3000; Amersham, Arlington Heights, IL). Detection of immunoreactive bands was carried out using the ECL Western blotting system (Amersham). 2.4. Expression and purification of human cell cycle proteins 2.4.1. Human gst-PLK A gst-PLK expression vector was constructed which included the glutathione-S-transferase gene fused to the amino terminus of the full-length human PLK gene via a linker containing a thrombin cleavage site. The construct was cloned into the baculovirus expression vector, pFASTBAC, which was used to generate viral stock for subsequent infection. Spodoptera frugiperda cells (Sf9) were infected with virus-expressing gst-PLK and grown for three days prior to harvesting. Cell pellets were lysed in 20 mM Tris-HCl (pH 8.0), 137 mM NaCl, 2 mM EDTA, 10% glycerol, 1% Triton X-100, 5 mM DTT, 2 g/mL E-64, 1 mM AEBSF, and 1 g/mL pepstatin A (lysis buffer) by an Avestin cell disrupter. Cell lysates were clarified by centrifugation at 28,000 ⫻ g at 4⬚C for 1 h. The supernatant was mixed with 10 mL of glutathione Sepharose 4B (Pharmacia Biotech, Sweden) and gently shaken at 4⬚C for 2 h. The resin was packed onto a Pharmacia XK26/50 column and unbound material was washed out with 20 mM Tris-HCl (pH 7.0), 100 mM NaCl, 12 mM MgCl2 and 1 mM DTT (wash buffer). Proteins bound to the resin were then eluted with 50 mM Tris-HCl (pH 8.0), 10 mM glutathione, and 0.05% Brij-35 (elution buffer). Fractions containing proteins were pooled and concentrated using an Amicon ultrafiltration system with YM30 membrane. Based on SDSPAGE analysis and Bio-Rad protein assay, the PLK kinase was purified to ⵑ80% purity, and the identity of the purified protein was confirmed by N-terminal sequence analysis. Remaining impurities consisted predominantly of free gst. 2.4.2. Human gst-cdc25C The gst-cdc25C E. coli expression construct was the generous gift of Dr. Laurent Meijer (Roscoff University, Roscoff, France). Briefly, the construct contained the glutathione-S-transferase gene fused to the amino terminus of full-length cdc25C by a linker containing a thrombin cleavage site. This was subcloned into the baculovirus expression vector, pFASTBAC, which was used to generate viral stock for subsequent infection as described above. The gst-cdc25C fusion protein was purified by the same procedures used for gst-PLK purification. SDS-PAGE analysis indicated that the band corresponding to the predicted molecular weight of fulllength gst-cdc25C was ⵑ30% of the mixture. The identity of this protein was confirmed by N-terminal amino
407
acid sequence analysis. The remaining contaminants were primarily composed of free gst and cdc25C degradation products. A ⵑ55kDa protein present in the preparation was confirmed by sequencing as a C-terminal truncated form of gst-cdc25C. For some studies the gst tag was removed from rhcdc25C by thrombin cleavage using the Novagen Thrombin Cleavage Kit as per the manufacturer’s instructions. 2.5. PLK kinase assay Unless otherwise indicated, each kinase assay contained either gst-rhcdc25C (1 g), gst cleaved rhcdc25C (1 g), or cdc25C immunoprecipitated from 50 g of G1/S arrested cell cytosol as substrate, to which 1-2 Ci/ assay of [32P]-␥ATP, 10 M cold ATP and kinase reaction buffer containing 20 mM HEPES (pH 7.4), 50 mM KCL, 10 mM MgCl2, 1 mM EGTA, 0.5 mM DTT were added. The reaction was initiated by addition of the indicated amount of either rhPLK or immunoprecipitated PLK from nocodazole-treated, G2/M synchronized Jurkat cells and was allowed to proceed for the indicated time. Kinase activity was terminated by the addition of EDTA to 25 mM. SDS-PAGE sample buffer was added, reactions were boiled for 5 min and samples were electrophoresed through 10% tris-glycine gels. Gels were exposed to phosphorscreens and images obtained by phosphorimager (Storm 860, Molecular Dynamics, Sunnyvale, CA). 2.6. Cdc2/cyclin B dephosphorylation assays A catalytically inactive, doubly phosphorylated recombinant human (rh)cdc2/cyclin B complex was prepared by treatment with rh-myt1 and rh-wee1 kinases. Briefly, 8 ng/l of rhcdc2/cyclin B was incubated with 3 ng/l rh-myt1 kinase and 8 ng/l rh-wee1 kinase with 1 mM ATP for 1 h at 37⬚C in buffer containing 50 mM HEPES, 10 mM MgCl2 and 1 mM DTT. For assays 80 ng total cdc2/cyclin B complex in reaction mix was incubated with or without rhPLK (2–4 g), rhcdc25C (4 g) or cdc25C immunoprecipitated from 50 g cytosol of hydroxyurea-treated Jurkat T cells at 37⬚C in PLK kinase assay buffer with 10 M ATP as described above. Reactions were stopped by the addition of EDTA to 25 mM. SDS-Sample buffer was added, and samples were boiled and subjected to SDS-PAGE and Western blotting as described above. All experiments were conducted 2 times and figures show results of one representative study. 3. Results and discussion 3.1. Recombinant human (rh) cdc25C is a substrate for rhPLK Recombinant human cdc25C and PLK gst-fusion proteins (ⵑ87 kDa and 97 kDa, respectively) were generated and production of active PLK kinase was con-
408
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
Fig. 1. Recombinant human (rh)cdc25C is a substrate for rhPLK. rhPLK (0.5 g /assay; 25 nM) was incubated alone or in the presence of increasing concentrations of rhcdc25C (0.1–3.0 g/reaction) for 60 min at 37⬚C. Reactions were terminated by the addition of 25 mM EDTA, resolved by SDS-PAGE, and substrate phosphorylation was visualized by phosphorimager.
firmed by evaluating phosphorylation of ␣-casei, a ubiquitous substrate, as has been previously reported (data not shown) [15]. Significant phosphorylation of gst alone was not observed (data not shown). RhPLK1 was then incubated with rhcdc25C as described under Materials and Methods. Fig. 1 shows that addition of increasing amounts of rhcdc25C (0.1–3.0 g/reaction) to the kinase reaction resulted in a marked concentration-dependent increase in rhcdc25C phosphorylation. In addition to phosphorylation of the full-length gst-cdc25C fusion protein, rhPLK also actively phosphorylated the ⵑ55 kDa C- terminal truncated cdc25C protein. Incubation of rhPLK1 (0.5 g/reaction; 25nM) alone appears to result in some autophosphorylation which is consistent with the findings of other investigators [15]. Fig. 2 shows that rhPLK phosphorylation of rhcdc25C increases over time and can be inhibited by the addition of the nonhydrolyzable ATP analogue, ATP␥S. Together, these data demonstrate that rhcdc25C is a substrate for rhPLK and provide the first evidence that human PLK may participate in the regulation of cdc25C phosphatase activity. 3.2. Evaluation of PLK and cdc25C protein expression in cell cycle phase arrested Jurkat T cells Human Jurkat T leukemia cells were chemically synchronized at either G1/S, S phase, or G2/M as described in Materials and Methods and cell cycle phase was con-
firmed by flow cytometry (Fig. 3a). Each population was then fractionated into a cytosol and 100,000 ⫻ g particulate fraction and PLK and cdc25C levels were evaluated in the cytosol by Western analysis. Fig. 3c shows that, as has been previously described, cdc25C protein expression does not change over the course of the cell cycle [5]. Instead, the phosphatase is modified by multiple phosphorylation events, which is evident by its reduced mobility (Fig. 3c) [5]. Consistent with reports in other human cell types [14,16], PLK protein is low to undetectable at G1/S (Fig. 3b). Protein levels begin to accumulate at S phase and are prominent at G2/M. The mobility of PLK in G2/M arrested fractions is retarded compared to the other cell cycle fractions, indicating protein modification. Human PLK is thought to be modified and activated by phosphorylation by an asyet unidentified kinase [14]. This is consistent with the higher molecular weight band observed in the G2/M arrested fraction and its enhanced activity in cells at the G2/M transition [14,15]. This fraction is, therefore, a good source for activated native human PLK. Human PLK and cdc25C were immunoprecipitated from the cytosol (100 g) of nocodazole treated, G2/M arrested Jurkat cells. Anti-PLK antibody was able to efficiently precipitate PLK (ⵑ67 kDa) from these cells, but no cdc25C was detectable in PLK immune complexes by Western analysis (Fig. 4a). Similarly, cdc25C, but not PLK, was found in the complexes immunoprecipitated with anti-cdc25C linked beads (Fig. 4b). No cdc2 or cyclin B was detectable via Western blot in either the PLK or cdc25C immune complexes, yet they were detectable in an amount of total cytosolic extract equal to that which was immunodepleted (Fig. 4c and d). These data indicate that while these proteins may participate in a regulatory amplification loop, they do not appear to exist in a tightly associated multimeric complex. 3.3. Native human PLK phosphorylates cdc25C Native PLK protein immunoprecipitated from 50 g of G2/M arrested Jurkat cytosol was evaluated for its ability to phosphorylate rhcdc25C. In contrast to the recombinant kinase, incubation of immunoprecipitated
Fig. 2. rhPLK phosphorylation of rhcdc25C increases with time. rhPLK (0.5 g/assay; 25 nM) was incubated with rhcdc25C (1 g/assay) for the indicated times at 37⬚C in the presence or absence of ATP␥S (1mM). Reactions were terminated by the addition of 25 mM EDTA, resolved by SDS-PAGE and substrate phosphorylation was visualized by phosphorimager.
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
409
Fig. 3. Evaluation of human PLK and cdc25C protein levels in cell cycle phase synchronized Jurkat T cells. (a) Asynchronous, G1/S, S or G2/M phase synchronized Jurkat cells were prepared as described under Materials and Methods. Cells were fixed and stained with propidium iodide; cell cycle phase was evaluated by flow cytometry. (b) shows Western immunoblot analysis of PLK protein levels in either asynchronous or synchronized Jurkat cell cytosols (50 g/Lane). (c) shows cdc25C protein levels in the same cell cycle synchronized Jurkat cell cytosols (50 g/ Lane).
Fig. 4. Immunoprecipitation of PLK and cdc25C from cell cycle synchronized Jurkats. Anti-human PLK or anti-human cdc25C linked protein G agarose beads were used to immunoprecipitate PLK or cdc25C from 100 g G2/M synchronized Jurkat cell cytosols as described in Materials and Methods. A control precipitation with no primary antibody was also conducted. Precipitated complexes were resolved by SDS-PAGE, and PLK (a), cdc25C (b), cdc2 (c), or cyclin B (d) protein levels were monitored by western immunoblot analysis.
PLK alone in the kinase reaction did not result in detectable autophosphorylation (Fig. 5), suggesting that the kinase in this G2/M cell preparation is fully phosphorylated. This finding is consistent with the Western blot analysis of the nocodazole treated (Fig. 3b) fraction that indicates PLK is phosphorylated and thus activated at the G2/M border. A mock immunoprecipitation of the same amount of G2/M arrested cytosol using rabbit IgG control antibodies was also performed. This reaction was used as a control to evaluate non-specific phosphorylation events. Fig. 5 shows that addition of the mock immunoprecipitate to gst-cleaved-rhcdc25C in a kinase reaction did not result in phosphorylation of the phosphatase. In contrast, incubation of immunoprecipitated PLK with gst-rhcdc25C (Fig. 5, 1 g/reaction) resulted in a marked phosphorylation of the cdc25C substate. As observed with the recombinant PLK, native PLK efficiently phosphorylated both the 87 kDa gst-fusion protein and the 55 kDa C-terminal truncated cdc25C protein present as a contaminant in the gst-rhcdc25C preparation. When native PLK was incubated with the gst-cleaved cdc25 the appropriately sized, full-length cdc25C (57 kDa) was phosphorylated (Fig. 5). These data confirm and strengthen our findings using recombinant PLK and are the first data utilizing human-cell derived PLK to demonstrate that cdc25C is a biologi-
410
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
Fig. 5. Native human PLK phosphorylates rhcdc25C. Anti-human PLK antibody or non-specific rabbit IgG antibody-linked protein G agarose beads were incubated with G2/M arrested Jurkat cytosols (50 g/reaction). PLK immunoprecipitates were incubated alone or in the presence of gst-rhcdc25C (1 g/reaction) or gst-cleaved rhcdc25C (1 g/reaction) for 60 min at 37⬚C in kinase reaction buffer. IgG control immunoprecipitations were incubated with gst-cleaved rhcdc25C (1 g/reaction) for 60 min at 37⬚C in kinase reaction buffer. Reactions were terminated by the addition of 25 mM EDTA, resolved by SDS-PAGE and substrate phosphorylation visualized by phosphorimager.
cally relevant substrate of human PLK. Moreover, these data strongly suggest that the human polo-like kinase may participate in cdc2/cyclin B regulation through phosphorylation and subsequent activation of cdc25C. 3.4. PLK phosphorylation of human cdc25C is required for activation of the phosphatase To assess the potential biological relevance of PLK phosphorylation of cdc25C, studies were conducted to evaluate the ability of PLK-treated and -untreated cdc25C to dephosphorylate its substrate, cdc2/cyclin B. Inactive threonine 14- and tyrosine 15-phosphorylated cdc2/cyclin B (designated cdc2-PP, Fig. 6a and b) was prepared by incubation with both the myt1 and wee1 kinases as described in Materials and Methods. The doubly phosphorylated cdc2/cyclin B complex was incubated (30 min at 37⬚C) alone or together with either
rhPLK, rhcdc25C or both proteins together. Activation of cdc25C was assessed by dephosphorylation of cdc2, which was monitored by a change in electrophoretic mobility. Fig. 6a shows that the rhPLK alone had no effect on the phosphorylation state of the cdc2 molecule. Incubation with the rhcdc25C alone resulted in dephosphorylation (60% as assessed by densitometry) of the cdc2 protein, yielding both the single and doubly dephosphorylated forms (designated cdc2-P and cdc2 respectively; Fig. 6a and b). This result is not surprising, as the rhcdc25C phosphatase undergoes some activation by the insect host cells during expression prior to purification. Treatment of rhcdc25C with PLK resulted in a substantial potentiation of the phosphatase activity of cdc25C, as evidenced by significantly more cdc2 protein in the dephosphorylated state (80% as assessed by densitometry). This activation was found to be time depen-
Fig. 6. PLK phosphorylation of rhcdc25C or native cdc25C results in phosphatase activation. (a) Catalytically inactive, doubly phosphorylated cdc2/cyclin B complexes (80 ng/reaction) were incubated for 30 min at 37⬚C either alone, with 4 g of rhPLK alone, 4 g rhcdc25C alone, or the two enzymes together. (b) Inactive cdc25C was immunoprecipitated from G1/S arrested Jurkat cytosols (50 g/ reaction) and incubated for 2 h at 37⬚C with cdc2/cyclin B complexes either alone, with 2 g rhPLK, or with rhPLK in the presence of 5 mM NaF. All reactions were terminated by the addition of 25 mM EDTA and resolved by SDS-PAGE. The phosphorylation state of cdc2 was assessed by Western immunoblot analysis.
A.K. Roshak et al. / Cellular Signalling 12 (2000) 405–411
dent with maximal activity observed between 30 and 60 min (data not shown). To assess whether phosphorylation by PLK is required for activation of the endogenous, inactive phosphatase, cdc25C was immunoprecipitated from the cytosol of G1/S arrested Jurkats (50 g) and incubated with the fully phosphorylated inactive cdc2/cyclin B complex with or without rhPLK (30 min at 37⬚C). Fig. 6b clearly demonstrates that incubation of the inactive cdc2/cyclin B complex with G1/S-derived cdc25C alone had no effect on the phosphorylation state of cdc2. This was expected, as the cdc25C phosphatase is inactive at this phase of the cell cycle. However, addition of rhPLK to the reaction induced a marked dephosphorylation of the cdc2 protein. The dephosphorylation could be completely abrogated by the addition of the phosphatase inhibitor, NaF. These findings demonstrate that native human cdc25C is directly activated by PLK phosphorylation, which leads to the dephosphorylation of cdc2/cyclin B, an absolute requirement for G2/M transition. In summary, we have shown for the first time that human cdc25C is a substrate for the human polo-like kinase, PLK. Moreover, this interaction was shown to be biologically relevant as it results in the activation of the cdc25C phosphatase and subsequent dephosphorylation of cdc2/cyclin B. These observations suggest that human PLK is instrumental not only in the orchestration of chromosomal segregation but also the initiation of the G2/M transition in somatic cells. Cdc25C catalyzed dephosphorylation of the threonine 14 and tyrosine 15 residues on cdc2 is known to trigger activation of the MPF (cdc2/cyclin B) and initiate a cell’s entry into mitosis. While cdc2/cyclin B is also capable of phosphorylating and activating cdc25C, phosphatase activation occurs in the absence of cdc2, indicating the existence of an additional kinase in cdc25C regulation [8]. It has been suggested that a “trigger kinase” is responsible for the initial activation of cdc25C and the MPF autoamplification loop. Taken together, the data presented herein
411
strongly support a hypothesis for human PLK as the key activator of cdc25C phosphatase, an obligatory step in the initiation of G2/M transition and progression into mitosis. Acknowledgments The authors wish to thank Shannon Reed for technical assistance, Dr. Jeffrey R Jackson and Maureen Ho for providing inactive cdc2/cyclin B complex and Drs. Ruth Mayer and Patty Podolin for critical review. References [1] Piwnica-Worms H. J Lab Clin Med 1996;128:350–4. [2] Mueller PR, Coleman TR, Kumigai A, Dunphy WG. Science 1995;270:86–90. [3] Kumagai A, Dunphy WG. Cell 1991;64:903–14. [4] Hoffmann I, Clarke PR, Marcote MJ, Karsenti E, Draetta G. EMBO 1993;12:53–63. [5] Gabrielli BG, Clark J, McCormack AK, Ellem, KAO. J Bio Chem 1997;272:28607–14. [6] Izumi T, Maller JL. Mol Bio Cell 1993;4:1337–50. [7] Stausfeld U, Fernandez A, Capony J-P, Girard F, Lautredou N, Derancourt J, Labbe J-C, Lamb, NJC. J Bio Chem 1994;269: 5989–6000. [8] Izumi T, Maller JL. Mol Bio Cell 1995;6:215–26. [9] Sunkel CE, Glover DM. J Cell Sci 1988;89:25–38. [10] Kitada K, Johnson A, Johnston L, Sugino A. Mol Cell Bio 1993; 13;4445–57. [11] Llamazares S, Moreira A, Tavares A, Girdham C, Spruce BA, Gonzalez C, Karess RE, Glover DM, Sunkel CE. Genes and Dev 1991;5:2153–65. [12] Kumagai A, Dunphy WG. Science 1996;273:1377–1380. [13] Holtrich U, Wolf G, Brauninger A, Karn T, Bohme B, Rubsamen-Waigmann H, Strebhardt K. Proc Natl Acad Sci USA 1994;91:1736–40. [14] Hamanaka R, Smith MR, O’Connor PM, Maloid S, Mihalic K, Spivak JL, Longo DL, Ferris DK. J Bio Chem 1995;270:21086–91. [15] Golsteyn RM, Mundt KE, Fry AM, Nigg EA. J Cell Bio 1995;129:1617–28. [16] Lee KS, Yuan YLO, Kuriyama R, Erikson RL. Mol Cell Bio 1995;15:7143–51. [17] Roshak AK, Jackson JR, McGough K, Chabot-Fletcher M, Mochan E, Marshall LA. J Bio Chem 1996;271:31496–501.