Inhibition of Cell Growth by Conditional Expression of kpm, a Human Homologue of Drosophila warts/lats Tumor Suppressor*

kpm is a human serine/threonine kinase that is homologous to Drosophila tumor suppressor warts/lats and its mammalian homologue LATS1. In order to define the biological function of kpm, we generated stable transfectants of wild-type kpm (kpm-wt), a kinase-dead mutant ofkpm (kpm-kd), and luciferase in HeLa Tet-Off cells under the tetracycline-responsive promoter. Western blot analysis showed that high levels of expression of kpm-wt as well as kpm-kd with an apparent mass of 150 kDa were induced after the removal of doxycycline. Induction of kpm-wt expression resulted in a marked decline in viable cell number measured by both trypan blue dye exclusion and MTT assay, whereas that of kpm-kd or luciferase had no effect. We then analyzed the cell cycle progression and apoptosis upon induction of kpm expression. 2–3 days after removal of doxycycline, cells underwent G2/M arrest, demonstrated by flow cytometric analysis of propidium iodide incorporation and MPM-2 reactivity. In vitro kinase assay showed that induction of kpm-wt led to down-regulation of kinase activity of the Cdc2-cyclin B complex, which was accompanied by an increase in the hyperphosphorylated form of Cdc2 and a change of phosphorylation status of Cdc25C. Furthermore, both DAPI staining and TUNEL assay showed that the proportion of apoptotic cells increased as kpm expression was induced. Taken together, these results indicate that kpm negatively regulates cell growth by inducing G2/M arrest and apoptotic cell death through its kinase activity.

kpm (1) is a novel protein kinase that was molecularly cloned from a human myeloid precursor cell line KG-1a (2) by degenerate PCR targeted for the conserved serine/threonine kinase domain. The sequencing and homology search revealed that kpm is homologous to a Drosophila melangaster gene, warts, or alternatively lats (3,4). warts/lats was originally identified by mitotic recombination of somatic cells and screening for homozygous mutants with overproliferation phenotype. Somatic cells mutated for this gene undergo extensive proliferation and form large tumors with abnormal morphologies, suggesting that warts/lats functions as a tumor suppressor. It is now recognized that warts/lats belongs to a subfamily of protein kinases with a common characteristic structure in the kinase domain consisting of dbf2 of budding yeast (5), orb6 of fission yeast (6), cot-1 of Neurospora crassa (7), ndr of various species (8), myotonic dys-trophy protein kinase (DMPK) of human (9), Rho-associated kinases of mammalian (10), and some other related kinases. It is noted that all the members of this family have been shown to be involved in the regulation of cell cycle progression.
A mammalian homologue of warts/lats, named LATS1 (11,12), has been isolated and extensively studied. Introduction of human LATS1 into Drosophila warts/lats mutants could prevent tumor formation and support normal development, demonstrating that the functions of these genes are conserved from flies to humans. Moreover, mice deficient for mouse LATS1 developed soft tissue sarcomas and ovarian tumor (13), indicating that this gene functions as a tumor suppressor also in mammals. In terms of cell cycle regulation, LATS1 protein has been shown to be phosphorylated and bind to Cdc2 at early mitosis (11). These data suggest that LATS1 is an authentic homologue of Drosophila warts/lats and has crucial functions in regulation of cell growth.
kpm is distinct from LATS1 but these two genes are highly homologous to each other especially at the kinase domain. A cDNA identical to kpm was isolated by others and named as LATS2 (14), and we agree that it is likely that kpm is another mammalian homologue of Drosophila warts/lats. In the previous article, we showed that the kpm protein has autophosphorylation activity in vitro and undergoes M-phase-specific phosphorylation in vivo (1) as has been reported with LATS1. It is unclear which kinase is involved in the phosphorylation of kpm although the Cdc2-cyclin B complex has been reported to phosphorylate LATS1 (15). Overexpression of kpm resulted in an increase in cell population at the S/G 2 /M phase. However, in contrast to LATS1, our knowledge of kpm is still limited and its physiological function remains largely unclear. Thus, it is to be determined whether kpm plays a crucial role in regulation of cell growth as has been demonstrated with LATS1. In the present study, we generated HeLa-derived stable transfectants of kpm based on the tetracycline-responsive expression system and analyzed the function of kpm in terms of cell cycle progression and cell viability. Here we present data indicating that kpm negatively regulates cell growth by inducing G 2 /M arrest and apoptotic cell death through its kinase activity.

EXPERIMENTAL PROCEDURES
Cell Lines and Cell Culture Conditions-HeLa Tet-Off cells (16) were purchased from Clontech (Palo Alto, CA) and maintained in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal calf serum (FCS) (Invitrogen) and 100 g/ml G418 (Sigma Chemical Co.). For selection of double transfectants and induction of gene expression, tetracycline-free FCS (Clontech) was used.
Generation of Tetracycline-responsive Gene-inducible Cell Lines-The hemagglutinin-A (HA) 1 -tagged wild-type kpm cDNA and its ki-nase-dead mutant described previously (1) were recloned into pTRE vector (Clontech) downstream of the tetracycline-responsive promoter to make pTRE-kpm-wt and pTRE-kpm-kd, respectively. pTRE-Luc control response plasmid (16) was obtained from Clontech. HeLa Tet-Off cells were transfected with pTRE-kpm-wt, pTRE-kpm-kd, or pTRE-Luc together with pTK-Hyg (Clontech) by electroporation using a Gene Pulser (Bio-Rad Laboratories, Hercules, CA). After 2 days, transfected cells were subjected to selection with 200 g/ml hygromycin B (Invitrogen) in the presence of 10 ng/ml doxycycline (Clontech). Hygromycinresistant cell lines were screened for induction of kpm-wt or kpm-kd expression upon removal of doxycycline by Western blotting. Induction of luciferase activity in pTRE-Luc-transfectants was confirmed with the luciferase assay kit (Promega, Madison, WI) and luminometry (Berthold Australia Pty Ltd., Bundoora, Australia).
Cell Viability Analysis by MTT Assay and Trypan Blue Dye Exclusion-Both adherent and non-adherent cells were harvested by trypsinization, and the viable cell number as well as the cell viability was measured by microscopic examination with trypan blue dye exclusion. Cellular proliferation was measured by reduction of MTT, which corresponds to living cell number and metabolic activity (17). Cells were thoroughly washed, plated at 5ϫ10 4 cells/well in 24-well plates and incubated with or without 10 ng/ml doxycycline for various periods of time (for 1-5 days). 50 l of 1 mg/ml MTT solution (WST-8, Nacalai Tesque, Kyoto, Japan) was added to each well. After 1 h of incubation, the absorbance of each well was measured at 492 and 630 nm using a microplate reader Benchmark (Bio-Rad Laboratories) according to the manufacturer's protocol.
Western Blot Analysis-Cells were harvested and lysed in TG-VO4 solution (18) (1% Triton X-100, 10% glycerol, 0.198 trypsin inhibitor units (TIU) of aprotinin per milliliter of Dulbecco's phosphate-buffered saline lacking divalent cations with fresh 100 mM Na 3 VO 4 ) containing 0.1% phenylmethylsulfonyl fluoride, 1ϫ Complete protease inhibitors (Roche Applied Science). After centrifugation, the supernatants were collected, and the protein concentration of each cell lysate was measured. Adjusted amounts of cell lysates were separated on 7.5 or 12.5% SDS-polyacrylamide gels and transferred onto Immobilon-P polyvinylidene difluoride membranes (Millipore, Bedford, MA). After blocking, the membranes were incubated for 1 h with the first antibodies followed by incubation with peroxidase-conjugated second antibody for 1 h. The protein bands were detected using the ECL detection system (Amersham Biosciences) according to the manufacturer's instruction. The antibodies used for Western blotting were mouse anti-HA monoclonal antibody (12CA5) (Roche Applied Science), rabbit anti-Cdc2 polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA), rabbit anticyclin B polyclonal antibody (Santa Cruz Biotechnology), a specific anti-phospho-Cdc2-Y15 rabbit polyclonal antibody (Cell Signaling, Beverly, MA), rabbit anti-Cdc25C polyclonal antibody (Cell Signaling, Beverly, MA), and goat anti-actin polyclonal antibody (Santa Cruz Biotechnology).
Cell Cycle Analysis-kpm-wt-, kpm-kd-, and Luc-inducible HeLa Tet-Off cell lines were cultured in medium without doxycycline for the indicated periods of time, harvested and washed twice with ice-cold phosphate-buffered saline (PBS) containing 0.1% glucose. Cells were then fixed with 70% ethanol for 1 h and incubated in 1 ml of PBS containing 50 l/ml of propidium iodide (Sigma), and 66 units/ml RNase (Invitrogen) on ice for 30 min. DNA content analysis was performed by a FACScan with CellQuest software (BD Biosciences).
Cell populations at G 2 and M phases were distinguished by the reactivity with mitotic protein monoclonal 2 (MPM-2 mouse monoclonal antibody) (Upstate, Waltham, MA) as described (19 -21). In brief, cells were fixed in 70% methanol and stained with MPM-2 antibody followed by incubation with (FabЈ) 2 fraction of FITC-conjugated goat anti-mouse IgG (BIOSOURCE, Camarillo, CA). After washing, cells were incubated with propidium iodide for DNA staining and then analyzed by two-color flow cytometry using the FACScan (BD Biosciences). MPM-2 reactive cells were considered to be at the mitotic phase, and the percentage of this population represented the mitotic index.
In Vitro Kinase Assay of the Cdc2-Cyclin B Complex-Kpm-wt-, kpmkd-, and Luc-inducible HeLa Tet-Off cell lines were cultured for 48 h with or without doxycycline, and then treated with 0.5 g/ml nocodazole (Sigma) for the last 24 h (11,22,23). Both adherent and non-adherent cells were harvested by trypsinization, washed three times with PBS (Ϫ) for thorough removal of nocodazole, and cultured without nocodazole for 50 min. Fluorescent microscopy of DAPI-stained cells indicated that most cells were in prometaphase and metaphase at this time point. Cells were harvested by trypsinization, washed three times with PBS (Ϫ), and lysed with TG-VO 4 solution containing 0.1% phenylmethylsulfonyl fluoride, 1ϫ Complete protease inhibitors. The protein concentra-tions of the cell lysates were measured and adjusted equally. A part of each cell lysate was used for the study of whole cell expression of Cdc2, cyclin B, Cdc25C, and ␤-actin by Western blotting. The rest of the lysate was subjected to immunoprecipitation by anti-cyclin B monoclonal antibody (Santa Cruz Biotechnology) and protein G-Sepharose (Amersham Biosciences). Half of the immunoprecipitate was used to monitor Cdc2 co-immunoprecipitated with cyclin B by Western blotting. The Cdc2 fraction phosphorylated at Tyr-15 was detected by a specific anti-phospho-Cdc2-Y15 antibody. The remaining immunoprecipitate bound to beads was used for in vitro immune complex kinase assay as follows (18). The precipitated beads were washed five times with TG-VO 4 solution and once with kinase buffer (20 mM Tris-HCl, pH 7.5, 10 mM MgCl 2, 1 mM EGTA) and incubated in 20 l of reaction buffer (20 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , 1 mM EGTA, 2 mM dithiothreitol, 1 mM Na 3 VO 4 , 0.1 mM ATP, 10 Ci of [␣-32 P]ATP, 5 g of histone H1) at room temperature for 20 min. 20 l of 2ϫ Laemmli sample buffer was added and boiled for 5 min to stop the reaction. The phosphorylation of histone H1 (Sigma) was examined by separating the samples on a 15% SDS-PAGE gel and autoradiography.
DAPI Staining-To confirm the findings of TUNEL assay, apoptotic cells were also detected by DAPI staining (25). Cells were washed with PBS (Ϫ), transferred to 1.5-ml microtubes, and fixed with 1% glutaraldehyde at room temperature for 30 min. After washing with PBS (Ϫ), cells were resuspended in 20 l of PBS (Ϫ) and mixed with 5 l of 10 l/ml DAPI (Sigma). Cell suspensions were mounted on slide glasses and subjected to fluorescence microscopic examination.

Establishment of Stable Transfectants of kpm-wt, kpm-kd, and Luciferase under the Control of the Tetracycline-responsive
Promoter-HeLa Tet-Off cells were transfected with either pTRE-kpm-wt, pTRE-kpm-kd, or pTRE-Luc together with pTK-Hyg and subjected to selection with hygromycin B. Several stable transfectant lines of the three genes were expanded and screened for efficient gene induction by the removal of doxycycline. Representative transfectants of each of the three genes were compared with parental cells for the gene expression in the absence or presence of doxycycline. Western blot analysis showed that the representative lines of HeLa Tet-Offkpm-wt and -kpm-kd were induced to express a large amount of kpm protein when cultured without doxycycline (Fig. 1A). Expression of kpm was detected after 12 h and reached maximal levels after 48 h, dependent on the concentrations of doxycycline (Fig. 1B). Based on scanning densitometry, removal of doxycycline resulted in more than 100-fold induction of kpm-wt or kpm-kd by probing with anti-kpm polyclonal antibody, which recognize both endogenous and exogenous kpm (data not shown). Likewise, high levels of luciferase activity were in- Overexpression of kpm Inhibits Cell Growth-In order to explore the biological function of kpm, we first examined the effects of overexpression of kpm on cell viability and proliferation. HeLa Tet-Off-kpm-wt, -kpm-kd, and -Luc cells were switched into the culture without doxycycline, and the viable cell number was counted daily by trypan blue dye exclusion. As shown in Fig. 2A, induction of kpm-wt expression resulted in a decline in viable cell number after 2 days compared with noninduced culture. In contrast, cell growth of HeLa Tet-Offkpm-kd as well as -Luc was not affected by expression of these genes. In accordance with this, the MTT assay also showed that overexpression of kpm-wt suppressed cell proliferation while that of kpm-kd or luciferase had no effect (Fig. 2B). These results indicate that kpm is involved in either cell cycle progression or cell viability, and negatively regulates cell growth. Since overexpression of kpm-kd had no effect on cell viability or proliferation as that of luciferase, it is suggested that antiproliferative effect of kpm is dependent on its kinase activity and kd mutant does not function as a dominant negative form at least in this particular assay system.
Overexpression of kpm Induces G 2 /M Arrest-Since overexpression of kpm-wt resulted in inhibition of cell proliferation and a decline in viable cell number, a cell cycle arrest was suspected to have occurred. To determine at which stage of the cell cycle cells were arrested, we performed the cell cycle analysis in the three transfectant lines upon removal of doxycycline. As shown in Fig. 3A, overexpression of kpm-wt induced an increase in cell population in G 2 /M phase and a decrease in cell population in G 1 phase compared with non-induced culture. In contrast, there was no difference in the profile of cell cycle progression between non-induced and induced overexpression of kpm-kd or that of luciferase. To further analyze the kpm-induced cell cycle arrest and determine whether it was a or without doxycycline (right) for 3 days and harvested for DNA content analysis. Cells were stained with PI and subjected to flow cytometric analysis using a FACScan. Three independent experiments gave similar results and a histogram of a representative experiment is shown. B, to distinguish whether kpm-wt blocks G 2 /M transition or delays progress of mitosis, mitotic index was measured in asynchronized kpm-wtinduced and non-induced HeLa Tet-Off cells. Flow cytometric analysis clearly differentiates G 1 phase cells with low PI : DNA content ϭ 2 N and low MPM-2 expression, S phase cells with PI : 2 N Ͻ DNA content Ͻ 4 N and low MPM-2 expression, G 2 phase cells with high PI : DNA content ϭ 4 N and low MPM-2 expression, and mitotic phase cells with high PI : DNA content ϭ 4 N and high MPM-2 expression. This assay was repeated three times and gave similar results. Apoptotic cells and fragmented cell debris that appeared in the sub-G 1 region were not included in these assays. G 2 /M transition arrest or a mitotic arrest, we performed scoring of mitotic index by MPM-2 assay that had long been used to identify mitotic cells. Overexpression of kpm-wt increased the cell proportion in G 2 /M phase as has been shown but with no significant changes in cell proportion in mitotic phase expressing MPM-2 antigen (Fig. 3B), indicating that overexpression of kpm-wt resulted in a G 2 /M transition arrest.
Kpm Negatively Regulates the Kinase Activity of the Cdc2-Cyclin B Complex-Since the overexpression of kpm induced cell cycle arrest at the G 2 /M boundary, we next examined whether overexpression of kpm negatively regulated the kinase activity of the Cdc2-cyclin B complex. In parallel with the gene induction, cells were synchronized in prometaphase and metaphase by the nocodazole method. Western blotting with whole cell lysates showed that induction of kpm-wt, kpm-kd, or luciferase did not affect the total amounts of Cdc2 or cyclin B (Fig.  4A). Likewise, there was no particular difference in the amount of Cdc2 co-immunoprecipitated with cyclin B between noninduced and induced cells. However, the Cdc2-cyclin B complex of kpm-wt-induced cells showed much lower histone H1 phosphorylation acitivity than that of non-induced cells, while induction of kpm-kd or luciferase had no effect on the kinase activity (Fig. 4B). We repeated these experiments three times and obtained similar results. Synchronization by the double thymidine block did not work well with HeLa Tet-Off cells and gave only incomplete results (data not shown), which were nevertheless consistent with what we observed by the nocodazole method. As has been described elsewhere (26,27), Cdc2 co-immunoprecipitated with cyclin B consisted of a doublet of bands of which the upper one represented the hyperphosphorylated inactive form, and the lower one represented the dephosphorylated active form. As shown in Fig. 4B, the proportion of the hyperphosphorylated inactive form of Cdc2 was increased in kpm-wt-induced cells, which was also demonstrated by Western blotting using a specific anti-phospho-Cdc2-Y15 antibody. These data suggest that Cdc2 bound to cyclin B remained or was rendered inactive by the phosphorylation at Thr-14 and Tyr-15, which seems to be the major mechanism of the G 2 /M arrest.
Overexpression of kpm Affects the Phosphorylation Status of Cdc25C-We next investigated the possible involvement of Cdc25C in the down-regulation of the Cdc2-cyclin B kinase. It is known that in SDS-PAGE Cdc25C consists of an 85-kDa band of the hyperphosphorylated form with the highest phosphatase activity and a doublet of 60-and 57-kDa bands, which represent the Ser-216-phosphorylated inactive form and the dephosphorylated form with weak phosphatase activity, respectively (28,29). In parallel with gene induction, cells were synchronized in prometaphase and metaphase by the nocodazole method. Western blotting with whole cell lysates showed that the 85-kDa hyperphosphorylated active form as well as the doublet of Cdc25C were present in luciferase-or kpm-kdinduced cells as described above. In contrast, it was noted that the mitotic hyperphosphorylated form was almost undetectable and conversely the Ser-216-phosphorylated inactive form (the upper band of the doublet) was increased in kpm-wt-induced cells (Fig. 4C). These data suggest that Cdc25C remains or is rendered inactive by overexpression of kpm-wt resulting in the decrease in the activity of dephosphorylating Cdc2, which seems to be one of mechanisms of the inactivation of the Cdc2cyclin B kinase.
Overexpression of kpm Induces Apoptosis after G 2 /M Phase Arrest-Because many tumor suppressor genes are known to inhibit cell growth by inducing apoptosis as well as cell cycle arrest, we examined whether this was also the case with kpm. In fact, induction of apoptosis by kpm was already suggested by cell cycle analysis in which an increase in sub-G 1 phase cells with a DNA content less than 2 N was observed after an elongated induction of kpm-wt for more than 4 days (Table I).
To demonstrate that this population was generated as a result of apoptosis, cells after kpm induction were first subjected to TUNEL assay. As shown in Fig. 5A, an increase in TUNELpositive cells was clearly detectable in kpm-wt-induced but not non-induced cells. Induction of kpm-kd or luciferase elicited no change in TUNEL-positive cells. Presence of apoptotic cells was FIG. 4. Effects of overexpression of kpm on the Cdc2-cyclin B complex and its kinase activity. A, Western blot analysis of Cdc2 and cyclin B in whole cell lysates of non-induced or induced HeLa Tet-Off-Luc, -kpm-wt, and -kpm-kd 50 min after removal of nocodazole. Cdc2 consists of a doublet of bands of which the upper band presumably represents the phosphorylated form, and the lower band represents the dephosphorylated form. B, immunoprecipitation of the Cdc2-cyclin B complex and histone H1 kinase assay. The immunoprecipitates by anti-cyclin B were blotted by anti-cyclin B, anti-Cdc2, or a specific anti-phospho-Cdc2-Tyr15 antibody. The Cdc2-cyclin B complexes were assayed for histone H1 kinase activities. The phosphorylation of histone H1 was visualized by separating the samples on a 15% SDS-PAGE gel and autoradiography. These experiments were repeated three times and gave similar results. C, Western blot analysis of Cdc25C in whole cell lysates of non-induced or induced HeLa Tet-Off-Luc, -kpm-wt, and -kpm-kd cells at 50 min after removal of nocodazole. Cdc25C consists of 85-kDa hyperphosphorylated band and a doublet of 60-and 57-kDa bands, which represent Ser-216-phosphorylated inactive form and the dephosphorylated with weak phosphatase activity, respectively. These experiments were repeated three times and gave similar results. confirmed by DAPI staining and immunofluorescence microscopy. Induction of kpm-wt resulted in chromatin condensation and segregation characteristics of apoptotic cells whereas that of kpm-kd or luciferase did not (Fig. 5B). DISCUSSION We previously reported the molecular cloning of kpm, which encodes a putative human serine/threonine kinase homologous to warts/lats, a Drosophila tumor suppressor (1). Prior to our article, Tao et al. (11) described a human as well as mouse homologue of warts/lats named LATS1 that could functionally compensate the defect of warts/lats in Drosophila. In contrast to LATS1 that has been extensively studied, the function of kpm remains largely unknown. It is to be determined whether kpm has similar or unique function compared with LATS1. In the present study, we established HeLa-derived stable transfectants of wild-type kpm (kpm-wt), a kinase-dead mutant of kpm (kpm-kd), and luciferase under the control of tetracyclineresponsive promoter in order to define the biological function of kpm. Using this system (16,30), we demonstrated that overexpression of kpm-wt resulted in suppression of cell proliferation due to cell cycle arrest in G 2 /M phase and subsequent apoptotic cell death.
Cell cycle analysis combined with MPM-2 assay clearly showed that overexpression of kpm-wt induced a cell cycle arrest by blockade of G 2 /M transition rather than delaying progress of mitosis. Consistent with this, we showed that the histone H1 kinase activity of the Cdc2-cyclin B complex was markedly diminished in kpm-wt-induced cells. Furthermore, Cdc2 bound to cyclin B remained or was rendered phosphorylated at Tyr-15 in kpm-induced cells. It is well known that the transition between the G 2 phase and mitosis is regulated through inhibitory phosphorylation of the Cdc2 kinase (31,32). Since overexpression of kpm-wt did not change the protein levels of Cdc2 and cyclin B in the whole cell lysates as well as in the immunoprecipitates by anti-cyclin B, it seems likely that overexpression of kpm led to a cell cycle arrest at G 2 /M by increasing the ratio of the hyperphosphorylated inactive form of Cdc2.
We examined whether Cdc25C was involved in the inactivation of Cdc2 because it is established that phosphorylated Cdc2 is dephosphorylated by this dual-specific phosphatase. Western blot analysis showed that overexpression of kpm-wt resulted in a marked decrease in the hyperphosphorylated active form of Cdc25C and an increase in the Ser-216-phosphorylated inactive form. Considering that 14-3-3 proteins bind to phosphoserine 216 of Cdc25C and translocate it from the nucleus to the cytoplasm, the overall phosphatase activity of Cdc25C should be strongly down-regulated in kpm-wt-induced cells, which seems to be one of the mechanisms of the increase in phosphorylated inactive form of Cdc2. Although we do not exclude other possible mechanisms for the kpm-induced cell cycle arrest in G 2 /M phase, it is certain that the kinase activity of kpm itself plays the central role in such a putative phosphorylation-dephosphorylation cascade, because kpm-kd had no effect.
LATS1 has also been reported to inhibit cell growth and induce cell cycle arrest in G 2 /M (33,34). However, the mechanism of the cell cycle arrest in LATS1 overexpression is different from that of kpm described here. According to Tao et al. (11) LATS1 could associate with Cdc2 and competitively inhibit the binding of cyclin B to Cdc2, which resulted in a decrease in kinase activity of the Cdc2-cyclin B complex. In addition, ectopic expression of LATS1 in MCF-7 cells has been reported to induce specific down-regulation of protein levels of cyclin A and cyclin B, while no effect was found on cyclin E, Cdc2, CDK2, p27 KIP , and p21 CIP levels (34). The discrepancy between kpm TABLE I Sub-G 1 populations in Luc-, kpm-wt-, or kpm-kd-induced cells HeLa Tet-Off-Luc, -kpm-wt, or -kpm-kd cells were cultured for 5 days without doxycycline and then subjected to cell staining by propidium iodide. Sub-G 1 phase cells with a DNA content less than 2 N were counted by flow cytometry, which seemed to represent apoptotic cells and fragmented cell debris. FIG. 5. Apoptotic cell death induced by overexpression of kpm. A, HeLa Tet-Off-Luc (lower left), -kpm-wt (lower middle), and -kpm-kd (lower right) cells were cultured with or without doxycyline for 5 days and then subjected to TUNEL assay. TUNEL ϩ cells were quantified by flow cytometric analysis. The upper row indicates positive controls of parental HeLa Tet-Off cells treated with 3 M staurosporine (upper left) or DNA nuclease (upper right). Dotted lines indicate the histograms of non-treated or non-induced cells. ⌬MFI indicates the increase in mean fluorescence intensity. The difference between kpm-wt non-induced cells and kpm-wt-induced cells was significant using a Student's t test (p Ͻ 0.0005). B, to confirm the findings of TUNEL assay, cells were also analyzed by DAPI staining and fluorescence microscopy. Parental HeLa Tet-Off cells treated with staurosporine were included as a positive control. Condensed and segmented nuclei were detected only in staurosporine-treated cells and kpm-wt-induced cells. These experiments were repeated three times and gave similar results. and LATS1 may be simply because these two molecules are distinct from each other. The experimental systems were also different, in which we used a tetracycline-responsive gene expression system in HeLa-derived cells (30,35) whereas LATS1 overexpression was induced by transduction of fibroblasts and other cancer cells using adenovirus vectors (33,34). We do not exclude the possibility that kpm has the capacity to associate with Cdc2 although we have not been able to demonstrate the association of these molecules in vivo. However, the data presented here clearly indicated that, at least in this system and in HeLa cells, the cell cycle arrest at G 2 /M transition is not mediated by the competitive inhibition of binding between Cdc2 and cyclin B but rather by the increase in the ratio of phosphorylated inactive form of Cdc2 bound to cyclin B. Thus, the present study has revealed the presence of a novel pathway of G 2 /M regulation through kpm and the Cdc2-cyclin B complex.
It is likely that the function of kpm is not restricted to the regulation of the Cdc2-cyclin B kinase. In fact, we showed that overexpression of kpm induced apoptotic cell death after cell cycle arrest. With regard to this, our preliminary experiments have suggested that expression of Bcl-2 protein is specifically downregulated after 3 days of kpm-wt induction (data not shown) although the signaling cascade leading to apoptosis from kpm overexpression remains unclear. In particular, the relationship between the cell cycle arrest and the induction of apoptosis needs to be investigated. On the other hand, a new cytoplasmic protein named salvador (hWW45, a WW domain containing-gene) has recently been described that interacts with LATS1 to cause both cell cycle arrest in G 1 /S as well as G 2 /M and apoptotic cell death (36,37). kpm also has a PPXY motif (1,36,37) that is predicted to interact with WW domain, and actually we have recently found that salvador could be coimmunoprecipitated with kpm. 2 Further studies are required to molecularly define the function of kpm in terms of cell cycle regulation as well as induction of apoptosis.