betaII protein kinase C is required for the G2/M phase transition of cell cycle.

Entry into mitosis requires the coordinated action of multiple mitotic protein kinases. In this report, we investigate the involvement of protein kinase C in the control of mitosis in human cells. Treatment of synchronized HL60 cells with the highly selective protein kinase C (PKC) inhibitor chelerythrine chloride leads to profound cell cycle arrest in G2 phase. The cellular effects of chelerythrine are not due to either direct or indirect inhibition of the known mitotic regulator p34(cdc2)/cyclin B kinase. Rather, several lines of evidence demonstrate that chelerythrine-mediated G2 phase arrest results from selective inhibition and degradation of betaII protein kinase C. First, chelerythrine causes dose-dependent inhibition of betaII PKC in vitro with an IC50 identical to that for G2 phase blockade in whole cells. Second, chelerythrine specifically inhibits betaII PKC-mediated lamin B phosphorylation and mitotic nuclear lamina disassembly. Third, chelerythrine leads to selective loss of betaII PKC during G2 phase in synchronized cells. Fourth, chelerythrine mediates activation-dependent degradation of PKC, indicating that betaII PKC is selectively activated during G2 phase of cell cycle. Taken together, these data demonstrate that betaII PKC activation at G2 phase is required for mitotic nuclear lamina disassembly and entry into mitosis and that betaII PKC-mediated phosphorylation of nuclear lamin B is important in these events.

Entry into mitosis requires the coordinated action of multiple mitotic protein kinases. In this report, we investigate the involvement of protein kinase C in the control of mitosis in human cells. Treatment of synchronized HL60 cells with the highly selective protein kinase C (PKC) inhibitor chelerythrine chloride leads to profound cell cycle arrest in G 2 phase. The cellular effects of chelerythrine are not due to either direct or indirect inhibition of the known mitotic regulator p34 cdc2 /cyclin B kinase. Rather, several lines of evidence demonstrate that chelerythrine-mediated G 2 phase arrest results from selective inhibition and degradation of ␤ II protein kinase C. First, chelerythrine causes dose-dependent inhibition of ␤ II PKC in vitro with an IC 50 identical to that for G 2 phase blockade in whole cells. Second, chelerythrine specifically inhibits ␤ II PKC-mediated lamin B phosphorylation and mitotic nuclear lamina disassembly. Third, chelerythrine leads to selective loss of ␤ II PKC during G 2 phase in synchronized cells. Fourth, chelerythrine mediates activation-dependent degradation of PKC, indicating that ␤ II PKC is selectively activated during G 2 phase of cell cycle. Taken together, these data demonstrate that ␤ II PKC activation at G 2 phase is required for mitotic nuclear lamina disassembly and entry into mitosis and that ␤ II PKC-mediated phosphorylation of nuclear lamin B is important in these events.
In higher eukaryotes, entry into mitosis is characterized by a dramatic structural reorganization of the cell. Mitotic events include chromosome condensation, nuclear lamina disassembly, and cytokinesis. Recent studies have demonstrated that multiple protein kinases play key regulatory roles in mitosis (1)(2)(3)(4). The p34 cdc2 /cyclin B kinase is required for the G 2 /M phase transition in both yeast and higher eukaryotes (5,6). Introduction of p34 cdc2 /cyclin B kinase into mammalian fibroblasts leads to mitotic chromosome condensation and cytoskeletal rearrangements, but not to mitotic nuclear lamina disassembly (7). Consistent with its critical role in mitotic events, p34 cdc2 /cyclin B kinase directly phosphorylates many mitotic phosphoproteins (3). However, the fact that p34 cdc2 /cyclin B kinase is not sufficient to mediate all aspects of mitosis demonstrates that other mitotic regulators are required (7).
Increases in cytosolic free calcium have been implicated in the regulation of mitosis in a number of cell systems (8 -12). Intracellular calcium release triggers nuclear envelope breakdown in sea urchin embryos, and calcium transients are required for nuclear envelope breakdown in mammalian fibroblasts (8 -12). Potential targets for the increase in intracellular calcium at the G 2 /M phase transition include the calcium/ calmodulin-dependent protein kinases and the calcium/phospholipid dependent protein kinase (PKC). 1 Accumulating evidence has implicated both of these protein kinases in the control of a number of mitotic events (reviewed in Ref. 4). PKC has been implicated in the regulation of mitosis in a number of systems (4,(13)(14)(15)(16)(17)(18)(19). Genetic evidence indicates that the yeast PKC homolog, PKC1, is an essential gene required for cell cycle (13). Deletion of PKC1 leads to recessive lethality and a mutant cell cycle phenotype characterized by a block in G 2 phase (13). PKC has also been implicated in cell cycle progression in mammalian cells (15,16,19). Staurosporine, a potent but relatively nonselective protein kinase C inhibitor, arrests cells at two cell cycle phases, G 1 and G 2 /M phase, depending on the concentration used (15). Several staurosporine analogues and other structurally distinct PKC inhibitors also lead to inhibition of cell cycle at G 2 /M phase (16,19). However, neither the individual PKC isotypes required during G 2 /M phase nor the mechanism by which inhibition of PKC leads to G 2 /M phase arrest have been elucidated. In the present study, we utilize the highly selective PKC inhibitor chelerythrine to demonstrate that PKC activity is required for entry of human cells into mitosis. Furthermore, we show that PKC activation is required for phosphorylation of key mitotic sites on nuclear lamin B previously implicated in mitotic nuclear lamina disassembly (17,18). Finally, we find that ␤ II PKC is a critical PKC isotype involved in the G 2 /M phase transition.

EXPERIMENTAL PROCEDURES
Cell Synchronization and Flow Cytometric Analysis-Human promyelocytic leukemia (HL60) cells were synchronized in G 1 /S phase with 2 g/ml aphidicolin (Sigma) at a cell density of 1 ϫ 10 6 /ml for 18 h as described previously (18). Cells were released from the G 1 /S blockade (time ϭ 0) by removal of aphidicolin and resuspended in Iscove's medium (Life Technologies, Inc.) containing 10% iron-supplemented calf serum (Hyclone). In some cases, nocodazole, chelerythrine, or staurosporine was added to the cells during mid to late S phase (5-7 h after release from aphidicolin) at the concentrations indicated in the figure legends. Cells were sampled at the indicated time points after release from aphidicolin, fixed in 90% methanol, stained with propidium iodide, and analyzed for cell cycle progression using flow cytometry as described previously (18).
Immunofluorescence Staining for Lamin B 1 and Detection of DNA-HL60 cells were fixed in 90% methanol and then incubated in phosphate-buffered saline containing 1% bovine serum albumin (PBS-BSA) at room temperature for one hour. Fixed cells were incubated for 2 h in a 1:100 dilution of a mouse monoclonal antibody to human lamin B 1 (Matritech) in PBS-BSA, washed three times with PBS-BSA, and incubated for 1 h with a 1:50 dilution of an fluorescein isothiocyanateconjugated goat anti-mouse antibody (Kirkegaard and Perry Laboratories) in PBS-BSA. The stained cells were washed three times in PBS-BSA, once with PBS, and then incubated with 0.5 g/ml of 4Ј,6diamidino-2-phenylindole (DAPI) in PBS for 5 min. Double stained cells were mounted onto microscope slides and observed using epifluorescence illumination with the appropriate excitation filters. Images were captured using a Photometrics CH250 CCD camera and aligned and colorized using Adobe Photoshop on a Power Macintosh.
Isolation of Lamin B 1 from Synchronized HL60 Cells-HL60 cells were synchronized by sequential treatment with aphidicolin and either chelerythrine or nocodazole as described above. For analysis of lamin B 1 solubility, chelerythrine-treated cells, nocodazole-treated cells or asynchronous cells were pelleted at 800 ϫ g for 10 min and washed twice with PBS. Washed cell pellets were lysed into buffer containing 2% Nonidet P-40 with gentle vortexing. Cell lysates were centrifuged at 12,000 ϫ g for 10 min yielding an Nonidet P-40-soluble supernatant and an Nonidet P-40-insoluble pellet. The supernatants were prepared for SDS-PAGE by addition of 3 ϫ SDS sample buffer followed by boiling for 5 min. Pellets were solubilized directly into SDS sample buffer, boiled for 5 min, and sonicated to reduce viscosity prior to SDS-PAGE in 8% acrylamide gels. Resolved proteins were transferred to nitrocellulose and subjected to immunoblot analysis for lamin B 1 using a monoclonal lamin B 1 antibody as described previously (18).
For radiolabeling studies, cells were synchronized with aphidicolin as described above and released into fresh medium. Seven hours after release from blockade, cells were washed three times with phosphatefree RPMI medium and resuspended in phosphate-free medium containing 100 Ci/ml carrier-free [ 32 P]orthophosphate (Amersham Corp.) and either 20 M chelerythrine or 40 ng/ml nocodazole. At 15 h after aphidicolin blockade, cells were pelleted and washed three times with cold PBS. Nuclear envelopes and lamin B 1 were isolated from chelerythrine-treated cells as described previously (18). Mitotic lamin B was isolated from nocodazole treated cells by extraction into lysis buffer containing 2% Nonidet P-40 as described above. Nonidet P-40 supernantants and isolated nuclear envelopes were resolved by SDS-PAGE and lamin B 1 identified by autoradiography and immunoblot analysis using a monoclonal lamin B 1 antibody as described above. Bands corresponding to lamin B 1 were excised and subjected to comparative phosphopeptide mapping as described previously (20).
Affinity Purification of p34 cdc2 /Cyclin B Kinase-HL60 cells were synchronized by sequential treatment with aphidicolin and either nocodazole or chelerythrine as described above or were obtained directly from asynchronous cultures (interphase control). Cells (5 ϫ 10 7 ) were lysed at a concentration of 2 ϫ 10 7 cells/ml in bead buffer (50 mM Tris-HCl, pH 7.4, 250 mM NaCl, 15 mM EGTA, 1 mM DTT, 5 mM MgCl 2 , 0.1% Nonidet P-40, 20 mM NaF, 50 mM ␤-glycerophosphate, 1 mM phenylmethylsulfonyl fluoride, and 20 g/ml leupeptin), and insoluble material was removed from the lysate by centrifugation at 14,000 ϫ g for 5 min. 50 l of p13 suc1 -agarose beads (Oncogene Science, Inc., 5 mg of p13 suc1 /ml of agarose) equilibrated in bead buffer were added to the cell lysates and incubated with gentle rocking for 1 h at 4°C. Beads were pelleted by centrifugation, washed three times in bead buffer, and stored in 50% glycerol for histone kinase assays.
In Vitro Kinase Assays-Recombinant human ␣, ␤ II , and PKC were obtained using the baculovirus insect cell expression system as described previously (21). Human p34 cdc2 /cyclin B kinase was obtained from HL60 cells by either p13 suc1 affinity as described above or by immunoprecipitation with an antipeptide antibody to p34 cdc2 (Life Technologies, Inc.) or human cyclin B (amino acids 257-267). Kinase reactions were performed in standard assay buffers for each kinase as described previously (18,21) using 10 g of histone H1 (Boehringer Mannheim) as the substrate for 15 min at 37°C in a total reaction volume of 40 l. In some reactions, the cyclin-dependent kinase inhibitor olomoucine was included at a concentration of 70 M. Protein kinase activity was quantitated by phosphorimaging analysis (Molecular Dynamics) and cyclin-dependent kinase activity was calculated as total histone kinase activity minus kinase activity in the presence of olomoucine. Olomoucine typically inhibited 80 -95% of the total histone kinase activity.
Immunoblot Analysis of Protein Kinase C Isotype Degradation-HL60 cells were synchronized with aphidicolin and treated with either chelerythrine or diluent as described above. Total cell lysates were prepared by lysis into 0.5 ml of boiling Laemmli sample buffer followed by sonication to reduce viscosity. For analysis of protein kinase C isotype expression and degradation, 5 ϫ 10 5 cell equivalents were resolved by 8% SDS-PAGE, transferred to nitrocellulose (Schleicher and Schuell), and blocked for 1 h in PBS, 0.05% Tween 20 (w/v), 5% non-fat dry milk (PBS/Tween/milk). Nitrocellulose sheets were incubated with PKC isotype-specific antibodies directed against ␣, ␤ II , or PKC as described previously (18,22). Antigen-antibody complexes were detected using horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence (ECL, Amersham) as described previously (18,22). Protein levels were quantitated by densitometry (Molecular Dynamics).
The effect of PKC activation on chelerythrine-induced degradation was assessed by treating asynchronous cells with either 20 M chelerythrine, 20 M dioctanoylglycerol (diC 8 ), or the combination of both compounds for up to 9 h. At the indicated times, total cell lysates were subjected to immunoblot analysis for ␣, ␤ II , and PKC as described above.

Chelerythrine Induces Dose-dependent G 2 /M Phase Arrest in Synchronized Human
Cells-Previous studies demonstrated that staurosporine, a potent protein kinase inhibitor, induces cell cycle arrest in both G 1 and G 2 /M phase (15), suggesting a role for protein kinase C in cell cycle progression. However, staurosporine is a relatively nonselective protein kinase inhibitor, making it difficult to attribute its cellular effects to inhibition of PKC rather than any of a number of other cellular kinases. Of particular concern, staurosporine has also been shown to be an effective inhibitor of p34 cdc2 /cyclin B kinase, a kinase whose activity is required for entry into mitosis (15). Therefore, we chose to investigate the effects of the highly selective PKC inhibitor, chelerythrine chloride, on the cell cycle of human leukemia (HL60) cells. Chelerythrine exhibits little inhibitory effect on cyclic AMP-dependent protein kinase, Ca 2ϩ /calmodulin-dependent kinase or tyrosine kinases (23) making it useful for these studies.
In order to assess the effects of chelerythrine on cell cycle progression, HL60 cells were synchronized in G 1 /S phase with aphidicolin and allowed to progress through cell cycle. These cells progress synchronously through cell cycle, returning to G 1 phase by 12 h after removal of aphidicolin (Fig. 1A). In contrast, cells released into medium containing chelerythrine progress to G 2 /M phase, but fail to return to G 1 phase. A similar cell cycle arrest in the G 2 /M phase was seen in cells treated with either staurosporine, a nonselective protein kinase inhibitor that is known to arrest cells in G 2 /M phase (15), or with nocodazole, a mitotic spindle poison that arrests cells in midmetaphase (Fig. 1A). As has been reported with staurosporine (15), chelerythrine leads to a secondary cell cycle arrest in G 1 phase, as indicated by the accumulation of a smaller G 1 phase population in cells treated with these compounds. Little or no cell cycle progression was observed at 15 and 24 h, demonstrating that these compounds induce cell cycle blockade at G 2 /M phase rather than slowed cell cycle progression (data not shown). The cell cycle arrest induced by chelerythrine is dosedependent with an apparent IC 50 of about 16 M (Fig. 1B).
Chelerythrine Blocks Cells in G 2 Phase Prior to Mitosis-Flow cytometric analysis of DNA content revealed that both chelerythrine and nocodazole lead to significant arrest in G 2 /M phase (69 and 67% G 2 /M phase cells, respectively). However, flow cytometry does not discriminate between cells in G 2 phase and those in mitosis. Therefore, we determined the mitotic index of cells arrested with either chelerythrine or nocodazole. In chelerythrine-treated cells, the mitotic index was 4%, while in nocodazole-treated cells, the mitotic index was 65%. Comparison of the mitotic indices and flow cytometric data revealed that whereas nocodazole arrested cells in mitosis (67% G 2 /M phase ϭ 65% mitotic index ϩ 2% G 2 phase cells), chelerythrine led to arrest in G 2 phase (69% G 2 /M phase ϭ 65% G 2 phase ϩ 4% mitotic index).
A number of morphological hallmarks have been identified for the G 2 /M phase transition. These include the condensation of the cellular chromatin into highly condensed, paired chromosomes, the loss of heterochromatin and nucleolar structures characteristic of the interphase nucleus, and the solubilization of the polymeric nuclear lamina. To further characterize the cell cycle status of chelerythrine-and nocodazole-treated cells, we assessed cells treated with these agents for characteristics associated with the G 2 /M phase transition. Chromatin morphology was determined in chelerythrine-and nocodazoletreated cells by fluorescence staining with the DNA-specific dye DAPI (Fig. 2). Staining of nocodazole-treated cells with DAPI reveals the presence of highly condensed metaphase chromosomes ( Fig. 2A). In contrast, DNA staining of chelerythrinetreated cells demonstrates that the chromatin is largely decondensed with prominent nucleolar structures typical of interphase (Fig. 2C). Chelerythrine-treated Cells Contain an Intact Nuclear Lamina-We next assessed the structural organization of the nuclear lamina of chelerythrine-and nocodazole-treated cells. For this purpose, cells were subjected to indirect immunofluorescence analysis using a monoclonal antibody specific for human lamin B 1 . Immunofluorescence staining of nocodazole-treated cells reveals that lamin B 1 is distributed in a vesicular pattern throughout the cytosol with little or no staining associated with the condensed chromosomal mass (Fig. 2B). This distribution pattern indicates that the nuclear lamina, which in these cells consists of a single lamin species, lamin B 1 (27), has undergone mitotic disassembly and that solubilized lamin B 1 is associated with cytoplasmic vesicles derived from the nuclear membrane. In contrast, immunofluorescence staining of chelerythrinetreated cells reveals a nuclear rim staining pattern characteristic of the interphase nuclear lamina (Fig. 2D). The spatial relationship between the nuclear lamina and the interphase chromatin in chelerythrine-treated cells is revealed when both DNA and lamin B 1 staining are visualized together (Fig. 2E).
These results indicate that the nuclear lamina of chelerythrinetreated cells is intact and forms a continuous ring surrounding the nuclear contents. In contrast, the nuclear lamina in nocodazole-treated cells has undergone mitotic disassembly as evidenced by the vesicular, cytosolic distribution of lamin B 1 (Fig. 2B).
In interphase cells, the nuclear lamina is highly polymerized, and the nuclear lamins are insoluble in buffers containing non-ionic detergents (24). During mitosis, however, the nuclear lamina disassembles, and the resultant lamin dimers are soluble in non-ionic detergents (24). Lamin B 1 solubility was therefore used to determine the polymeric state of the nuclear lamina in chelerythrine-and nocodazole-treated cells (Fig. 3). Treated cells were fractionated into Nonidet P-40 soluble and insoluble fractions, and the presence of lamin B 1 in these fractions was assessed by immunoblot analysis using a monoclonal lamin B 1 antibody. As expected, lamin B 1 is recovered in the Nonidet P-40 insoluble (nuclear) pellet from unsynchronized cells (Fig. 3A), reflecting the fact that the vast majority of these cells are in interphase. After nocodazole treatment, a significant proportion of the lamin B is recovered in the Nonidet P-40-soluble supernatant with the remainder found in the insoluble pellet (Fig. 3B). In contrast, essentially all the lamin B from chelerythrine-treated cells is recovered in the Nonidet P-40-insoluble pellet (Fig. 3C). Densitometric analysis of the immunoblots reveals that about 45% of the total lamin B in nocodazole-treated cells is Nonidet P-40-soluble with the remainder recovered in the insoluble pellet, whereas Ͻ5% of the total lamin B from chelerythrine-treated cells is Nonidet P-40soluble. These data are in good agreement with the observed mitotic indices of nocodazole-and chelerythrine-treated cells. Therefore, by both morphologic and biochemical criteria, chelerythrine-treated cells contain a structurally intact, highly polymerized nuclear lamina which has not undergone mitotic nuclear lamina disassembly. In contrast, nocodazole-treated cells exhibit metaphase morphology characterized by a depolymerized, vesicularized nuclear lamina that is soluble in Nonidet P-40, indicative of mitotic nuclear lamina disassembly. Therefore, in the presence of chelerythrine, cells complete S phase and arrest in G 2 phase prior to mitosis.
Chelerythrine Directly Inhibits PKC but Not p34 cdc2 /Cyclin B Kinase in Vitro-Given the G 2 phase cell cycle blockade induced by chelerythrine, we wished to assess the mechanism by which it elicits this effect. Chelerythrine has been reported to be a potent and highly selective inhibitor of protein kinase C (23). To confirm the specificity of this compound for inhibition of protein kinase C, we first assessed the effect of chelerythrine on purified recombinant ␣, ␤ II , and PKC (the three PKC isotypes expressed in HL60 cells; Ref. 21), and on cyclin-dependent kinases affinity-purified from mitotic cells with p13 agarose beads in vitro (Fig. 4A). Chelerythrine leads to dosedependent inhibition of ␣, ␤ II , and PKC activity. Of the three PKC isotypes, ␤ II PKC is most sensitive to chelerythrine inhibition (IC 50 ϳ15 M), followed by ␣ (IC 50 ϳ25 M) and PKC (IC 50 Ͼ40 M). These IC 50 values, particularly those for ␣ and ␤ II PKC, are in close agreement with the IC 50 for G 2 phase blockade in whole cells (ϳ16 M; Fig. 1B). In contrast, chelerythrine had little or no inhibitory effect on affinity-purified mitotic cyclin dependent kinase activity even at concentrations of 40 M. The predominant cyclin-dependent kinase activity from mitotic cells is p34 cdc2 /cyclin B (1). These results indicate that chelerythrine does not directly inhibit p34 cdc2 /cyclin B kinase activity at concentrations that cause profound G 2 phase arrest in whole cells. Similar results were obtained with p34 cdc2 /cyclin B kinase immunopurified using antibody to the carboxyl terminus of p34 cdc2 (data not shown). We conclude that the cell cycle effects of chelerythrine are not mediated through direct inhibition of p34 cdc2 /cyclin B kinase, but rather correlate directly with the inhibitory activity of chelerythrine against conventional PKC isotypes.
p34 cdc2 /Cyclin B Kinase Is Fully Active in Chelerythrinetreated Cells-During the cell cycle, a number of kinases and phosphatases coordinate to activate p34 cdc2 /cyclin B kinase through a series of phosphorylation/desphosphorylation events (1). Inhibition of any one of these enzymes could potentially result in indirect inhibition (or lack of activation) of mitotic p34 cdc2 /cyclin B kinase. In order to determine whether chelerythrine prevents normal mitotic activation of p34 cdc2 /cyclin B kinase in whole cells, we compared the activities of p34 cdc2 / cyclin B kinase from nocodazole-treated (M phase), cheleryth-rine-treated (G 2 phase), and asynchronous (interphase) cells (Fig. 4B). p34 cdc2 /cyclin B kinase isolated from both nocodazoleand chelerythrine-treated cells exhibits similar high cyclin-dependent histone kinase activity. In contrast, interphase cells exhibit low cyclin-dependent histone kinase activity, consistent with the low levels of p34 cdc2 /cyclin B kinase activity present during interphase (1). Identical results are obtained when p34 cdc2 /cyclin B kinase is isolated by p13 suc1 affinity or by immunoprecipitation using p34 cdc2 or cyclin B-specific antibodies (Fig. 4B). The p34 cdc2 /cyclin B kinase activity associated with chelerythrine-treated cells is slightly lower than that of mitotic cells, perhaps because chelerythrine-treated cells have not yet entered mitosis when p34 cdc2 /cyclin B kinase activity is at its peak. These data demonstrate that the effects of chelerythrine do not appear to be due to either direct or indirect inhibition of p34 cdc2 /cyclin B kinase.
Chelerythrine Inhibits Mitotic Phosphorylation of Lamin B 1 at Ser 405 , a ␤ II PKC-mediated Phosphorylation Site-Our recent studies using human erythroleukemia (K562) cells demonstrated that ␤ II PKC is a lamin kinase that directly phosphorylates lamin B 1 at a prominent mitotic phosphorylation site, Ser 405 (18). Furthermore, we demonstrated that ␤ II PKC is selectively activated at the nucleus during G 2 phase of cell cycle where it phosphorylates Ser 405 on lamin B 1 (18). Therefore, we assessed the phosphorylation status of lamin B 1 in cheleryth- FIG. 2. Chelerythrine leads to cell cycle blockade in G 2 phase. HL60 cells were synchronized with aphidicolin and treated with either nocodazole (A and B) or chelerythrine (C-E) as described in the legend to Fig. 1. 15 h after release from aphidicolin, cells were fixed in methanol and stained with 0.5 g/ml DAPI to visualize DNA morphology and with a monoclonal antibody to human lamin B and fluorescein isothiocyanate-labeled secondary antibody to visualize nuclear lamin B as described under "Experimental Procedures." Cells treated with nocodazole (A and B) exhibit a morphology typical of mid-metaphase arrest. The cellular chromatin is highly condensed in the form of metaphase chromosomal masses (A). Lamin B is distributed throughout the cytoplasm in a vesicular pattern indicating that mitotic nuclear lamina disassembly has occurred (B). Chelerythrine-treated cells (C-E) exhibit a typical interphase morphology. The chromatin is decondensed and nuclei contain recognizable nucleolar structures (C). Lamin B localizes in a distinct nuclear rim staining pattern at the nuclear periphery indicative of an intact, polymeric nuclear envelope (D). Double staining reveals the spatial relationship of the chromatin and nuclear lamina staining in chelerythrine-treated cells (E). rine-treated (G 2 phase) and nocodazole-treated (mitotic) HL60 cells (Fig. 5).
Lamin B 1 from mitotic cells is highly phosphorylated on three tryptic phosphopeptides (Fig. 5A, labeled 1-3). This pattern is identical to that observed in mitotic lamin B 1 from K562 cells and corresponds to phosphorylation within the carboxylterminal domain of lamin B 1 (17,18). Lamin B 1 from chelerythrine-treated cells is highly phosphorylated on one peptide, phosphopeptide 1, whereas phosphorylation on phosphopeptides 2 and 3 is drastically reduced (Fig. 5B). Recombinant ␤ II PKC phosphorylates lamin B 1 on two phosphopeptides (Fig.  5C), corresponding to phosphorylation of the consensus PKC phosphorylation sites Ser 395 (peptide 4) and Ser 405 (peptide 2) (17,18). The identity of phosphopeptide 2 was confirmed by mixing lamin B 1 phosphopeptides from mitotic cells with those from either chelerythrine-treated cells (Fig. 5D) or lamin B 1 phosphorylated by recombinant ␤ II PKC in vitro (Fig. 5E). Inhibition of Ser 405 phosphorylation (peptide 2) is indicative of direct inhibition of ␤ II PKC-mediated phosphorylation of lamin B 1 in the presence of chelerythrine. Inhibition is selective, since phosphorylation of phosphopeptide 1, which is not attributable to PKC, is unaffected by chelerythrine treatment. Phosphorylation of peptide 3 is also inhibited in the presence of chelerythrine. However, phosphorylation of this peptide cannot be attributed to the direct action of ␤ II PKC, since it is not phosphorylated by recombinant ␤ II PKC in vitro ( Fig. 5C; Refs. 17 and 18). It is possible that chelerythrine directly inhibits the kinase responsible for phosphorylation of peptide 3. Alternatively, peptide 3 may be phosphorylated by a kinase whose activity is modulated by or is dependent upon ␤ II PKC activation in vivo.
Chelerythrine Results in Selective Degradation of the ␤ II PKC Isotype at G 2 Phase-Attempts at rescuing chelerythrinetreated cells by extensive washing and/or treatment with phorbol myristate acetate or bryostatin, both potent activators of PKC, failed to stimulate entry into mitosis. The irreversibility of the chelerythrine-induced G 2 phase blockade was not due to cytotoxicity since chelerythrine-treated cells maintain high viability (as measured by trypan blue exclusion) in the G 2 -arrested state for at least 48 h after treatment. Our inability to reverse the effects of chelerythrine led us to examine the levels of PKC isotypes in chelerythrine-treated cells (Fig. 6).
Total cell lysates from synchronized and chelerythrinetreated cells were obtained during the time period when synchronized cells traverse mitosis (11-15 h after release from aphidicolin; see Fig. 1). Total cell extracts were subjected to immunoblot analysis using a panel of isotype-specific antibodies against ␣, ␤ II , and PKC. Immunoblot analysis reveals that ␤ II PKC is selectively degraded in chelerythrine-treated cells, while ␣ and PKC levels remain unchanged (Fig. 6A). The degradation of ␤ II PKC is not a consequence of the cell synchronization procedure or the transit of cells through mitosis, since Purified ␣, ␤ II , and PKC expressed in the baculovirus system and human p34 cdc2 /cyclin B kinase isolated from mitotic HL60 cells by p13 suc1 affinity were assayed for histone kinase activity as described under "Experimental Procedures." Chelerythrine was included in the assays at the indicated concentrations. Results are plotted as percent histone kinase activity in the absence of chelerythrine. Data represent the mean of triplicate experiments Ϯ S.D. Some error bars are not visible, since they are smaller than the data point symbols. B, p34 cdc2 / cyclin B kinase from chelerythrine-treated cells is highly active. p34 cdc2 / cyclin B kinase was affinity purified from asynchronous interphase (I), chelerythrine-blocked (G 2 ), and nocodazole-blocked (M) HL60 cells by either p13 suc1 -agarose affinity (p13 suc1 ) or immunoprecipitation with antibody to p34 cdc2 kinase (␣ cdc2) or cyclin B (␣ cyclin B) and assayed for histone kinase activity as described under "Experimental Procedures." Specific, olomoucine-inhibitable cyclin-dependent histone kinase activity is plotted as a percentage of the activity present in mitotic cells.
it is not seen in synchronized cells allowed to complete mitosis in the absence of chelerythrine (Fig. 6A). In order to determine the temporal relationship between chelerythrine-induced degradation of ␤ II PKC and progression through mitosis, ␤ II PKC levels in synchronized, chelerythrine-treated cells were assessed along with cell cycle progression through G 2 /M phase in parallel cultures of synchronized control cells (Fig. 6B). As can be seen, chelerythrine-induced degradation coincides with cell cycle progression through G 2 /M phase. Degradation is not dependent upon the length of exposure to chelerythrine, since it is observed specifically at G 2 /M phase whether chelerythrine is added from 1 to 5 h before mitosis, yet it is not observed when chelerythrine is added after mitosis has occurred (data not shown). These results indicate that ␤ II PKC becomes highly sensitive to chelerythrine-mediated degradation specifically during G 2 phase prior to mitosis, a finding that is interesting in light of our recent demonstration that ␤ II PKC is selectively translocated and activated at the nucleus during G 2 phase (18). The selective degradation of ␤ II PKC may account for the irreversibility of the chelerythrine-induced G 2 blockade.
Chelerythrine Induces Activation-dependent Degradation of PKC Isotypes in Vivo- Fig. 6 demonstrates that chelerythrine induces rapid and selective degradation of ␤ II PKC in G 2 phase. Our previous studies showed that ␤ II PKC, but not ␣ or PKC, is selectively activated at the nucleus during G 2 phase (18). These results suggested that chelerythrine may induce activation-dependent degradation of PKC. In order to test this hypothesis, we assessed the effect of chelerythrine on the stability of the ␣, ␤ II , and PKC isotypes in unsynchronized cells in the presence and absence of PKC activators (Fig. 7). Incubation of unsynchronized cells with chelerythrine leads to little or no change in stability of the ␣ and PKC isotypes and to a slow loss of ␤ II PKC over a 9-h period (Fig. 7). Likewise, activation of PKC by addition of diC 8 has little effect on the stability of ␣, ␤ II , and PKC. However, when cells are exposed to both chelerythrine and diC 8 , rapid degradative loss of the ␣ and ␤ II PKC isotypes is observed (Fig. 7). Interestingly, PKC is not degraded under these conditions, consistent with the observation that PKC does not bind, and is not activated by, conventional PKC activators (26). Similar results were obtained using phorbol myristate acetate and bryostatin to activate ␣ and ␤ II PKC, indicating that this effect is not specific for diC 8 -mediated activation (data not shown).
The rates of degradation of ␣ and ␤ II PKC in the presence of chelerythrine and PKC activators are similar. However, some degradation of ␤ II PKC is observed even in the absence of exogenous activator. It is not apparent whether this reflects a

FIG. 5. Chelerythrine inhibits mitotic phosphorylation of lamin B.
HL60 cells were synchronized with aphidicolin, radiolabeled in the presence of [ 32 P]orthophosphoric acid and treated with either nocodazole or chelerythrine as described under "Experimental Procedures." At 15 h after release from aphidicolin blockade, cells were lysed, and lamin B was isolated and subjected to tryptic phosphopeptide analysis as described under "Experimental Procedures." Alternatively, lamin B was phosphorylated by purified ␤ II PKC in vitro. A, phosphopeptide map of mitotic lamin B from nocodazoletreated cells. B, phosphopeptide map of lamin B from G 2 phase cells blocked with chelerythrine. C, phosphopeptide map of lamin B phosphorylated by ␤ II PKC in vitro. D, mixture of A and B. E, mixture of A and C.
FIG. 6. Chelerythrine induces selective degradation of ␤ II PKC during G 2 phase in synchronized cells. A, HL60 cells were synchronized and treated with either chelerythrine (ϩ) or diluent (Ϫ) as described above. Cells were harvested at the indicated times and total cell lysates were subjected to immunoblot analysis for ␣, ␤ II , and PKC as described under "Experimental Procedures." B, temporal relationship between ␤ II PKC degradation and progression through G 2 /M phase. HL60 cells were synchronized with aphidicolin and treated with either chelerythrine or diluent as described above. Cellular ␤ II PKC levels were determined in chelerythrine-treated cells by densitometry of immunoblots at the indicated times. Cell cycle progression through the G 2 /M phase was assessed by flow cytometry of parallel cultures of synchronized cells in the absence of chelerythrine. Results are plotted as percent of control ␤ II PKC levels and percent cells in G 2 /M phase versus time after release from aphidicolin. difference in the activation status of the ␣ and ␤ II PKC isotypes in unsynchronized cells or an intrinsic difference in the susceptibility of these PKC isotypes to chelerythrine-induced degradation. The degradation of ␣ and ␤ II PKC differs in another respect. Immunoreactive ␤ II PKC disappears completely, whereas ␣ PKC undergoes "laddering" as the full-length protein is degraded into lower molecular weight fragments prior to complete loss of immunoreactive ␣ PKC. Fig. 7 demonstrates that chelerythrine is capable of activationdependent degradation of both ␣ and ␤ II PKC. However, in synchronized cells a selective loss of the ␤ II PKC isotype is observed during G 2 phase. These results are consistent with our previous finding that ␤ II PKC, but not ␣ PKC, is selectively activated during G 2 phase in synchronized cells (18). DISCUSSION The protein kinase C family has been implicated in the regulation of differentiation and proliferation in human leukemia cells (20,22,(27)(28)(29)(30). We previously demonstrated that the ␣ and PKC isotypes are involved in cellular differentiation, whereas the ␤ II PKC isotype is required for proliferation (22). ␤ II PKC levels correlate directly with proliferative capacity, and inhibition of ␤ II PKC expression using antisense oligonucleotides leads to profound inhibition of cellular proliferation (22). The current study provides a plausible mechanism by which ␤ II PKC participates in the proliferative pathway.
Earlier fractionation studies demonstrated that ␤ II PKC is selectively translocated and activated at the nucleus of human leukemia cells in response to proliferative stimuli (20,30). At the nucleus, ␤ II PKC directly phosphorylates a major nuclear envelope component, lamin B 1 (17,18,20,21,(27)(28)(29)(30). Phosphorylation site mapping identified the ␤ II PKC phosphorylation sites on lamin B as Ser 395 and Ser 405 (17,18). Furthermore, phosphorylation of these two sites by purified ␤ II PKC leads to nuclear lamina disassembly in vitro (17). Finally, combined cellular fractionation and phosphorylation studies demonstrated that ␤ II PKC is selectively translocated and activated at the nucleus of synchronized cells at the time of the G 2 /M phase transition and that the ␤ II PKC phosphorylation site on lamin B, Ser 405 , becomes prominently phosphorylated during the G 2 /M phase transition (18). We have not detected phosphorylation of Ser 395 in synchronized cells during any phase of cell cycle, suggesting that this site may not be of physiologic relevance. 2 Our previous data suggested a role for ␤ II PKC activation and lamin B phosphorylation in nuclear events during G 2 /M phase of cell cycle. In the present report, we provide compelling evidence that ␤ II PKC activation is a requisite event in the G 2 /M phase transition.
Treatment of synchronized cell populations with the highly selective PKC inhibitor chelerythrine leads to profound cell cycle arrest in G 2 phase. This blockade is characterized by the presence of a fully replicated genome, decondensed nuclear chromatin morphology including dense nucleolar structures, and the presence of an intact, highly polymerized nuclear lamina. Inhibition studies demonstrate that chelerythrine does not inhibit the prominent mitotic protein kinase, p34 cdc2 /cyclin B, either in vitro or in whole cells. Phosphopeptide mapping demonstrates that chelerythrine inhibits ␤ II PKC-mediated phosphorylation of lamin B 1 at Ser 405 , a site previously implicated in mitotic nuclear lamina disassembly (17,18). In addition, a second, unidentified mitotic phosphorylation site on lamin B 1 was also inhibited by chelerythrine treatment. Phosphorylation of this site does not appear to be directly mediated by either ␤ II PKC or p34 cdc2 /cyclin B kinase, since neither kinase phosphorylates this site in vitro (18). It is possible that chelerythrine directly inhibits the kinase that phosphorylates this site or that phosphorylation at this site requires ␤ II PKC activity. Current studies are ongoing to identify this site, the kinase responsible for its phosphorylation, and the effect of ␤ II PKC activation on phosphorylation.
The cell cycle effects of chelerythrine are not readily reversible. However, this effect does not appear to be due to cytotoxicity, but rather to the selective degradation of ␤ II PKC at G 2 phase in chelerythrine-treated cells. Chelerythrine-induced degradation of PKC is not peculiar to ␤ II PKC, since chelerythrine leads to degradation of both ␣ and ␤ II PKC in unsynchronized cells in the presence of PKC activators. Several observations suggest that chelerythrine-induced PKC degradation is activation-dependent. First, both ␣ and ␤ II PKC are rapidly degraded when HL60 cells are treated with both chelerythrine and a PKC activator. Second, PKC is not degraded in these cells in the presence of chelerythrine and either diC 8 , phorbol myristate acetate, or bryostatin. It is well established that PKC, unlike ␣ and ␤ II PKC, is not activated by these conventional PKC activators (31). The lack of degradation of PKC is not due to an intrinsic insensitivity to chelerythrine inhibition, since chelerythrine inhibits PKC activity in vitro. Third, ␤ II PKC is selectively degraded during G 2 phase in synchronized cells treated with chelerythrine, consistent with our previous observation that ␤ II PKC, but not ␣ or PKC, is selectively activated during this phase of cell cycle (18).
Our data are interesting in light of the proposed mechanism of action of chelerythrine (23). Kinetic studies have determined that chelerythrine inhibition is substrate-competitive and that chelerythrine binding likely involves the active site of PKC (23). According to the prevailing model of PKC regulation, inactive, cytosolic PKC assumes a conformation in which an amino-terminal pseudosubstrate domain occupies the active site of the enzyme, making it inaccessible to substrates (32). Binding of co-factors and activators to the regulatory domain leads to a conformational change that displaces the pseudosubstrate from the active site, thereby allowing substrates to bind and leading to enzyme activation (33). Since chelerythrine binding to PKC involves the active site and is competitive with 2 L. J. Thompson and A. P. Fields, unpublished results. substrate, PKC activation may lead to enhanced binding of chelerythrine.
It is well established that the active conformation of PKC is highly susceptible to proteolytic degradation, which appears to be the molecular basis underlying phorbol ester-induced PKC degradation and down-regulation (33,34). Phorbol esters activate PKC by binding to diacylglycerol binding sites within the regulatory domain. The high affinity of phorbol esters for PKC causes the enzyme to remain in its active conformation for extended periods of time, leading to chronic activation of the enzyme and proteolytic degradation (33,34). Our data suggest that chelerythrine may act in a manner similar to the phorbol esters. In the case of chelerythrine, binding of endogenous activators induces the active conformation. Release of the pseudosubstrate domain from the active site allows subsequent binding of chelerythrine at or near the active site. High affinity binding of chelerythrine may hold PKC in its active conformation, leading to increased susceptibility to proteolytic degradation. Further studies will be required to determine whether this proposed mechanism is responsible for the activation-dependent degradation of PKC by chelerythrine.
The present study provides compelling evidence that ␤ II PKC activation is a requisite step for entry into mitosis. ␤ II PKC activation appears to be necessary, either directly or indirectly, for a number of mitotic events including chromosome condensation and mitotic nuclear lamina disassembly. Our results indicate that ␤ II PKC is a cell cycle regulator distinct from p34 cdc2 /cyclin B kinase. Many cell cycle regulatory proteins, including p34 cdc2 /cyclin B kinase and PKC, are highly conserved evolutionarily. The yeast PKC homolog, PKC1, is required for the G 2 /M phase transition in yeast (13). Ablation of PKC1 leads to recessive lethality and arrest of cells in G 2 phase prior to mitosis (13). Functionally, PKC1 appears to regulate aspects of the osmotic stability of the yeast cell wall necessary for completion of the cell division cycle (35) and is a mitotic checkpoint independent from the yeast homolog of p34 cdc2 / cyclin B kinase. Our data are consistent with those obtained in yeast, since the activity of p34 cdc2 /cyclin B kinase is unaffected by chelerythrine either in vitro or in whole cells, suggesting that ␤ II PKC may lie in a distinct pathway from that of p34 cdc2 / cyclin B kinase. Furthermore, our data indicate that mitotic activation of p34 cdc2 /cyclin B kinase is not sufficient for entry into mitosis, consistent with the data of others (7).
Given the critical role that PKC plays in the regulation of mitosis in species as diverse as yeast and man, it is tempting to speculate that the pathways in which PKC participates are likewise conserved. PKC1 has been shown to regulate the activity of a number of downstream protein kinases including the yeast homologs of the mammalian dual specificity kinase, mitogen-activated protein kinase kinase, and the serine/threonine kinase, mitogen-activated protein kinase (36,37). Current studies are under way to determine whether these kinases lie downstream of ␤ II PKC in mammalian cells during the G 2 /M phase transition.
The present data, coupled with our previous studies, indicate that lamin B 1 is a relevant and perhaps critical target for direct, ␤ II PKC-mediated phosphorylation during the G 2 /M phase transition. Phosphorylation of lamin B 1 by purified ␤ II PKC leads to nuclear lamina disassembly in vitro and the ␤ II PKC phosphorylation site on lamin B 1 , Ser 405 , is prominently phosphorylated in mitotic cells (17,18), suggesting that phosphorylation of Ser 405 plays a key role in mitotic nuclear lamina disassembly in vivo. Indeed, chelerythrine inhibits phosphorylation of Ser 405 (and a second unidentified site) on lamin B 1 and inhibits mitotic nuclear lamina disassembly. These data suggest that one mechanism by which chelerythrine blocks entry into mitosis is by preventing phosphorylation of a site or sites on lamin B 1 required for mitotic nuclear lamina disassembly. Clearly, chelerythrine inhibits other aspects of mitosis including chromosome condensation and mitotic spindle formation. Therefore, ␤ II PKC represents an important G 2 /M phase cell cycle regulator that is required for multiple aspects of mitosis.
Although the mechanism by which ␤ II PKC is translocated and activated at the nucleus during G 2 phase is not known, it has recently been shown that the nucleus has a distinct phosphoinositide cycle that is responsive to both mitogenic stimuli and intrinsic cell cycle regulation (38 -41). In regenerating rat hepatocytes, nuclear phosphoinositide turnover increases during G 2 phase, leading to an increase in nuclear diacylglycerol levels and the translocation of PKC to the nucleus (41). In addition, a nuclear membrane lipid factor, termed NMAF, has been identified from human leukemia cells that selectively activates ␤ II PKC at the nuclear membrane (21). Therefore, it is likely that nuclear phospholipid metabolism can generate signals for the nuclear activation of ␤ II PKC during G 2 phase of cell cycle. In this regard, it has recently been reported that in Saccharomyces cerevisiae the yeast homolog of phosphatidylinositol 4-kinase, STT4, lies upstream of PKC1 (42). The existence of cell cycle-regulated pathways involving PKC in both yeast and mammals suggests that PKC is a phylogenetically conserved regulator of cell cycle events, particularly those associated with the G 2 /M phase transition.