Phosphorylation of Bcl-2 Protein by CDC2 Kinase during G 2 /M Phases and Its Role in Cell Cycle Regulation*

Although it has been reported that Bcl-2 phosphorylation is associated with certain types of apoptosis, there is much controversy over the functional significance of and the kinases responsible for the phosphorylation. In this study, we examined whether Bcl-2 is phosphorylated by CDC2 kinase, a master regulator of G 2 /M tran- sition in the eukaryotic cell cycle. When CDC2 was activated by okadaic acid in HL-60 cells, Bcl-2 phosphorylation was readily induced. The phosphorylation was correlated with the accumulation of cells in G 2 /M phases, but was not proportional to the level of apoptosis. Furthermore, we found that Bcl-2 was phosphorylated during G 2 /M phases of normal cell cycle. The ability of CDC2 to phosphorylate Bcl-2 was confirmed by in vitro kinase assay with a highly purified CDC2-cyclin B complex. Using synthetic peptides and mutant cell lines, we identified threonine 56, one of two consensus sites for CDC2 within the Bcl-2 sequence, as a residue phosphorylated by CDC2. Mutation at threonine 56 abrogated the cell cycle inhibitory effect of Bcl-2 without affecting anti-apoptotic function. These results suggest that two distinct functions of Bcl-2 (anti-apoptosis and cell cycle inhibition) are differentially regulated by post-translational mechanisms such as phosphorylation. CDC2-mediated phosphorylation of Bcl-2 may play some physiological roles in the negative regulatory events during mitosis. kinases including CDC2, was Nocodazole chem-icals Cell Culture and Separation— Human promyelocytic leukemia cell line HL-60 was maintained in RPMI 1640 medium supplemented with 10% fetal calf serum. Separation of cells enriched for each phase of the cell cycle was conducted with counterflow centrifugal elutriation using the SRR6Y system (Hitachi Koki Co., Tokyo, Japan) (21). Exponentially

Bcl-2 is a 26-kDa integral membrane protein, which is located on the cytoplasmic aspect of the mitochondrial outer membrane, endoplasmic reticulum, and nuclear envelope (1). It represents a family of proteins regulating apoptosis in either a positive or negative manner. The Bcl-2 family is divided into two subgroups: one with an anti-apoptotic property (Bcl-2, Bcl-xL Bcl-w) and the other with apoptosis inducing ability (Bax, Bak, Bok, Bad, Bik, and Bid) (2). It has been demonstrated that Bcl-2 can inhibit apoptosis in a variety of cell types (3,4). In addition to the anti-apoptotic effect, another important function of Bcl-2 was recently unveiled; it restrains cell cycle entry of quiescent cells (5)(6)(7)(8) or hastens withdrawal of proliferating cells from the cell cycle (9). Intriguingly, the cell cycle effect of Bcl-2 is shown to be genetically separable from its pro-survival function (10). Despite these investigations, however, the molecular basis of the cell cycle inhibitory effect of Bcl-2 is still unclear.
Although recent studies have established that Bcl-2 function is primarily modulated by heterodimerization with pro-apoptotic members of the Bcl-2 family (2), there may be other regulatory mechanisms such as phosphorylation. Phosphorylation of Bcl-2 was first demonstrated by Alnemri et al. (11) in Sf9 cells which ectopically express Bcl-2 with a baculovirus expression system. Since then, there are several reports dealing with Bcl-2 phosphorylation, but its functional significance remains controversial (12)(13)(14)(15)(16). Some researchers reported that anticancer drug-induced apoptosis was accompanied by Bcl-2 phosphorylation, suggesting that phosphorylation inactivates Bcl-2 function (12,13). A similar phenomenon was observed in Ras-induced apoptosis of Jurkat cells (14). By contrast, other groups demonstrated that Bcl-2 phosphorylation is essential for anti-apoptotic activity of bryostatin-1, an activator of protein kinase C, or cytokines such as interleukin-3 and erythropoietin (15,16). In any case, Bcl-2 phosphorylation could serve as a pivotal mechanism for integrating extracellular signals and apoptotic pathways. Therefore, establishing the molecular identity of protein kinases executing Bcl-2 phosphorylation is of paramount importance to our understanding of apoptosis and other cellular events regulated by Bcl-2.
In the present study, with this background in mind, we examined the possibility of Bcl-2 phosphorylation by CDC2 kinase, a master regulator of G 2 /M transition in eukaryotic cells (17,18) and a possible byplayer for apoptosis (19,20). We found that CDC2 phosphorylates Bcl-2 protein at threonine 56 during G 2 /M phases of the cell cycle, and modulates the cell cycle-regulatory function of Bcl-2 without affecting the antiapoptotic property.

EXPERIMENTAL PROCEDURES
Reagents-Okadaic acid (provided by Dr. Hirota Fujiki, National Cancer Center, Tokyo, Japan) was dissolved in dimethyl sulfoxide at a concentration of 100 g/ml and stored at Ϫ20°C until use. Butyrolactone I, a specific inhibitor of cyclin-dependent kinases including CDC2, was prepared and used as described previously (20). Nocodazole was obtained from Janssen Biotech N. V. (Olen, Belgium). All other chemicals were purchased from Sigma. Peptides were ordered from Asahi Emers Co., Ltd. (Tokyo, Japan).
Cell Culture and Separation-Human promyelocytic leukemia cell line HL-60 was maintained in RPMI 1640 medium supplemented with 10% fetal calf serum. Separation of cells enriched for each phase of the cell cycle was conducted with counterflow centrifugal elutriation using the SRR6Y system (Hitachi Koki Co., Tokyo, Japan) (21). Exponentially growing HL-60 cells were separated into 6 fractions, followed by flow cytometric analysis to monitor the status of separation. The cells were synchronized in M phase by nocodazole as described previously (22).
Flow Cytometry-The cell cycle profile was analyzed by staining intracellular DNA with propidium iodide in preparation for a flow cytometry with the FACScan/CellFIT system (Becton-Dickinson, San Jose, CA). DNA histogram was obtained by analyzing 25,000 cells with the ModFIT program (Becton-Dickinson). The proportion of apoptotic cells was simultaneously calculated by this method (20).
In Vitro Phosphorylation of Synthetic Peptides-High performance liquid chromatography-purified synthetic peptides (2 mM) were incubated at 30°C in 30 l of a reaction mixture containing 50 mM Tris-HCl (pH 7.5), 1 mM EGTA, 1 mM dithiothreitol, 100 M sodium orthovanadate, 30 mM MgCl 2 , and 0.1 mM [␥-32 P]ATP (200 Ci/ml) for 30 min. Then, each sample was spotted onto a phosphocellulose paper disc, washed twice with 1% acetic acid, and washed three times in distilled water. The radioactivity on each disc was determined by scintillation counting.
Metabolic Labeling and Immunoprecipitation of Bcl-2 Protein-Cells were washed three times in phosphate-free RPMI 1640 medium (Sigma), adjusted to a concentration of 2.5 ϫ 10 7 cells/ml, and incubated at 37°C for 16 h after adding [ 32 P]orthophosphoric acid at 1 mCi/ml. The cells in G 0 /G 1 and G 2 /M phases were separated by centrifugal elutriation and lysed in EBC buffer. Bcl-2 protein was immunoprecipitated as described above, and its phosphorylation state was analyzed by autoradiography.

Bcl-2 Protein Was Phosphorylated in HL-60 Cells Treated with Okadaic Acid Concomitantly with the Activation of CDC2
Kinase-Okadaic acid, a potent inhibitor of phosphatase 2A, is known to activate CDC2 kinase by inducing phosphorylation of CDC25 (24,25). To test the possibility that CDC2 is involved in phosphorylation of Bcl-2, we cultured HL-60 cells with 500 nM okadaic acid, a dose sufficient for CDC2 activation, and examined the expression and phosphorylation of Bcl-2 by immunoblotting. First, we confirmed that CDC2 was really activated by okadaic acid in our system. In proliferating HL-60 cells, CDC2 protein was discernible as 3 bands of different mobilities on SDS-polyacrylamide gels: the fastest migrating band corresponds to dephosphorylated, active CDC2; and the other two slower migrating bands represent inactive CDC2 in which tyrosine 14 and threonine 15 are phosphorylated (26) (Fig. 1, second row). After 4 h of treatment with okadaic acid, the phosphorylated CDC2 species completely disappeared and only the active form remained. This change was accompanied by an increase in histone H1 kinase activity (data not shown). Coincident with the activation of CDC2, a slowly migrating form of Bcl-2 appeared (Fig. 1, uppermost panel). This form was previously shown to be the phosphorylated Bcl-2 species by 32 P labeling experiments and phosphatase treatment (11,12). Neither CDC2 activation nor mobility shift of Bcl-2 was observed in HL-60 cells cultured with dimethyl sulfoxide (carrier) alone. These results suggest that CDC2 is involved in phosphorylation of Bcl-2 protein.
CDC2 Was Able to Phosphorylate Bcl-2 Protein in Vitro and in Vivo-To convincingly demonstrate that Bcl-2 is phosphorylated by CDC2, we carried out in vitro kinase assays using a highly purified CDC2-cyclin B complex. This complex efficiently phosphorylated histone H1, a specific substrate of CDC2 (27), in a dose-dependent manner, confirming that its activity is maintained in vitro ( Fig. 2A). Then, we examined whether this complex could phosphorylate Bcl-2 protein immunoprecipitated from HL-60 cells. As shown in Fig. 2B, CDC2cyclin B complex induced phosphorylation of Bcl-2 with similar kinetics to that of histone H1, indicating that CDC2 kinase can directly phosphorylate Bcl-2. To verify the relevance of CDC2 in phosphorylating Bcl-2 under in vivo conditions, we examined the effects of butyrolactone I, a cell permeable inhibitor of CDC2 kinase (20), on Bcl-2 phosphorylation induced by okadaic acid in HL-60 cells. Phosphorylation of pRB was simultaneously analyzed to monitor the inhibition of CDC2 activity. Bcl-2 phosphorylation was inhibited by butyrolactone I in a dose-dependent manner (Fig. 3A) and the kinetics of inhibition was similar to that of pRB (Fig. 3B), indicating the specificity of inhibition. These results suggest that CDC2 is at least partially responsible for Bcl-2 phosphorylation in vivo.
Next, we tried to determine the sites of phosphorylation by CDC2 in the Bcl-2 sequence. Within the Bcl-2 sequence, two sites (amino acid sequences 56 -59 and 87-90) were found to match the consensus motif for CDC2 kinase; Ser/Thr-Pro-Xbasic amino acid (27, 28). We synthesized 2 peptides, designated as Bcl-2-56 and Bcl-2-87, corresponding to these putative CDC2 target sites for in vitro kinase assays (see Table I for the sequences). As controls, two peptides corresponding to residues 157-165 (Bcl-2-161) and 183-191 (Bcl-2-187) were simultaneously assayed, both of which contain serine or threonine but do not fulfill the consensus sequence for CDC2. Bcl-2-56 was shown to be phosphorylated by purified CDC2 kinase to an extent similar to SV40 large T antigen peptide, which is commonly used as a CDC2 substrate (29) (Table I). In contrast, CDC2 did not phosphorylate either Bcl-2-87 or the two control peptides, although the former contains the cognate consensus sequence.
Bcl-2 Phosphorylation Was Associated with G 2 /M Arrest but Not with Apoptosis-Next, we investigated how Bcl-2 phosphorylation is implicated in the regulation of two distinct functions of Bcl-2: anti-apoptosis and cell cycle inhibition. For this purpose, HL-60 cells were cultured with two different concentrations (500 and 10 nM) of okadaic acid to serially monitor the cell cycle profile and phosphorylation status of Bcl-2. The high dose of okadaic acid induced premature entry into mitosis, as indicated by chromatin condensation with a diploid content of DNA (data not shown and Ref. 23), along with striking Bcl-2 phosphorylation after 16 h of treatment: the ratio of phosphorylated form/unphosphorylated form was 0.72 by densitometric comparison (Fig. 4A). No significant increase in sub-G 1 fraction (apoptotic cells) was observed even after 16 h of culture. In the presence of low dose okadaic acid, HL-60 cells slowly accumulated in G 2 /M phases and underwent apoptosis after 48 h of culture (Fig. 4B). This is fully consistent with our previous observation (23). Bcl-2 phosphorylation was induced during this process, but its level was far less than that with 500 nM okadaic acid: the phosphorylated form/unphosphorylated form ratio was 0.12 at 24 h and 0.09 at 48 h. It is of note that there was no difference in the level of Bcl-2 phosphorylation between 24-and 48-h samples, indicating that Bcl-2 phosphorylation is not correlated with the extent of apoptosis. This notion is also supported by the heavy phosphorylation of Bcl-2 without significant apoptosis in cells treated with 500 nM okadaic acid for 16 h. In keeping with this view, Bcl-2 phosphorylation was not observed in anti-Fas-treated HL-60 cells in which CDC2 was aberrantly activated in association with apoptosis without inducing G 2 /M accumulation (20) (data not shown).
Bcl-2 Was Phosphorylated during G 2 /M Phases in Normal Cell Cycle-The above findings suggest that Bcl-2 phosphorylation by CDC2 is not a determinant of apoptosis but an event associated with M phase (including pseudo-M) entry. This hypothesis prompted us to investigate the cell cycle-dependent phosphorylation of Bcl-2 using normally cycling cells without any drug treatment. To this end, we separated exponentially growing HL-60 cells into fractions enriched for each phase of the cell cycle by centrifugal elutriation (21). As depicted in Fig.  5, fractions 1 and 2 almost exclusively consisted of cells in G 0 /G 1 phases, fractions 3 and 4 were enriched for cells in S phase, fraction 5 contained a significant amount of cells in G 2 /M phases, and fraction 6 was further enriched for G 2 /M. In accord with the cell cycle profile of each fraction, CDC2 was present as a highly phosphorylated, inactive form in fractions 1 to 3, dephosphorylated in fraction 4 where cells prepare to enter M phase, and remained dephosphorylated in fractions 5 and 6 (Fig. 5, upper panel). Histone H1 kinase activity was well correlated with the status of CDC2 in each fraction; it was significantly higher in fractions 4, 5, and 6 than in 1, 2, and 3. Bcl-2 phosphorylation was readily detectable in fractions 4 -6 where CDC2 was activated. The maximal phosphorylation of Bcl-2 was observed in fraction 6, which was equivalent to that in HL-60 cells cultured with 500 nM okadaic acid for 4 h. Again, apoptosis was not associated with Bcl-2 phosphorylation in these fractions. These results suggest that phosphorylation

Bcl-2 Phosphorylation by CDC2
plays a role in the cell cycle-related function of Bcl-2 rather than anti-apoptotic effect.
To confirm that threonine 56 is a relevant site of phosphorylation by CDC2 in native Bcl-2 protein in vivo, we examined whether mutation at threonine 56 affected the phosphorylation of Bcl-2 during G 2 /M phases. For this purpose, we established HL-60 sublines overexpressing wild-type Bcl-2 (Bcl-2-WT) and mutated Bcl-2 in which threonine 56 was replaced by alanine (Bcl-2-T56A). There was no difference in the amounts of overexpressed Bcl-2 protein in the established cell lines (data not shown). The phosphorylation state of Bcl-2 in these cells was analyzed by 32 P-metabolic labeling and immunoprecipitation. As shown in Fig. 6A, Bcl-2 phosphorylation at G 2 /M was severely impaired in cells carrying the mutation at threonine 56. No significant effect on Bcl-2 phosphorylation was observed when mutation was introduced at serine 87 (Bcl-2-S87A) (data not shown). To obtain conclusive evidence that threonine 56 is a target for CDC2 in the context of full-length Bcl-2, we performed in vitro kinase assay using Bcl-2-T56A. Bcl-2 protein was immunoprecipitated from Bcl-2-WT and Bcl-2-T56A cell lines, phosphorylated by purified CDC2-cyclin B, and subjected to Bcl-2 immunoblotting and autoradiography. In vitro phosphorylation of wild type Bcl-2 by CDC2 resulted in the appearance of a shifted band (Fig. 6B, left panel), which was shown to be the predominant 32 P-labeled form of Bcl-2 by autoradiography (Fig. 6B, right panel). In contrast, neither mobility shift nor 32 P incorporation was observed in Bcl-2-T56A, indicating that CDC2 phosphorylates Bcl-2 exclusively at threonine 56. Taken together, these results strongly suggest that Bcl-2 is phosphorylated by CDC2 at threonine 56 during G 2 /M phases in vivo, although the involvement of other kinases in this process cannot be ruled out.
Mutation at Threonine 56 Abrogated the Cell Cycle-inhibitory Effect but Not Anti-apoptotic Function of Bcl-2-Finally, we examined the effect of mutation at threonine 56, a putative phosphorylation site for CDC2, on the cell cycle-inhibitory func-tion of Bcl-2. When mock-transfected HL-60 cells were released from M phase synchronization by washing out nocodazole, transition from M to G 0 /G 1 phase started after 30 min, and the proportion of cells in G 0 /G 1 became dominant over that in G 2 /M after 1 h. Further progression to S phase began after 24 h of the release (Fig. 7, Mock). Overexpression of wild-type Bcl-2 delayed the transition from M phase to G 0 /G 1 (Fig. 7, Bcl-2-WT), which is compatible with previous observations in different cell types (5)(6)(7)(8). The introduction of a non-phosphorylated mutation at residue 56 almost completely abrogated the cell cycleinhibitory effect of Bcl-2 (Fig. 7, Bcl-2-T56A), whereas mutation at serine 87 did not (data not shown). In contrast, there was no significant difference in resistance to apoptosis between Bcl-2-WT, Bcl-2-T56A, and Bcl-2-S87A (Table II). Taken together, our present findings indicate that phosphorylation of Bcl-2 at threonine 56 by CDC2 is required for Bcl-2-mediated cell cycle inhibition, which may have some roles during mitosis in the normal cell cycle. DISCUSSION Although recent investigations have shown that Bcl-2 is phosphorylated on various occasions including during anticancer drug treatment, there is much controversy over the functional significance of and the kinases responsible for this phosphorylation. In the present study, we examined the possibility that Bcl-2 is phosphorylated by CDC2 kinase. When CDC2 was conditionally activated by okadaic acid, Bcl-2 phosphorylation was induced along with an accumulation of cells in the G 2 /M phases. The ability of CDC2 to phosphorylate Bcl-2 was confirmed by in vitro kinase assay using a highly purified CDC2-cyclin B complex, and by suppression of Bcl-2 phosphorylation by butyrolactone I, a cell permeable inhibitor of CDC2 kinase. These results strongly suggest that CDC2 is one of the kinases responsible for Bcl-2 phosphorylation. To our knowledge, three different kinases have been proposed as Bcl-2 kinase so far. First, Blagosklonny et al. (30) described the in- Exponentially growing HL-60 cells were separated into 6 fractions (Fractions 1 to 6) enriched for each phase of the cell cycle by centrifugal elutriation. The cell cycle profile of each fraction is shown at the bottom (the result for Fraction 2 is not displayed because it is almost the same as that for Fraction 1). Whole cell lysates were prepared from each fraction, and subjected to immunoblot analysis for Bcl-2 and CDC2, and histone H1 kinase assay. Unfractionated HL-60 cells were arrested in G 0 /G 1 phases by phorbol ester treatment and used as a CDC2-inactive control (Unfr.). HL-60 cells treated with 10 nM okadaic acid for 24 h were simultaneously assayed as a CDC2-active control (ϩOA). Data shown are representative of three independent experiments. volvement of Raf-1 kinase in Bcl-2 phosphorylation associated with anti-cancer drug-induced apoptosis. Second, protein kinase C was reported to mediate phosphorylation of Bcl-2 in a murine factor-dependent cell line treated with interleukin-3 or bryostatin-1 (15,16). Finally, c-Jun N-terminal kinase, a member of the MAP kinase family, was shown to phosphorylate Bcl-2 (31). The present study adds CDC2 to the list of Bcl-2 kinases, and is fully consistent with the recent report of Ling et al. (32), who also demonstrated that Bcl-2 can be a substrate of CDC2 when cells are treated with anti-tublin agents. It is possible that these kinases phosphorylate Bcl-2 protein at different sites and kinetics, and may have distinct effects on its function.
Then, we attempted to identify the site(s) of phosphorylation by CDC2 within the Bcl-2 sequence. Bcl-2 protein possesses two consensus sites for CDC2 kinase at amino acids 56 -59 and 87-90. Using synthetic peptides and immunoprecipitated Bcl-2 protein, we determined threonine 56 but not serine 87 as a residue of phosphorylation by CDC2 in vitro. Furthermore, in vivo phosphorylation of threonine 56 was confirmed by 32 Pmetabolic labeling of HL-60 subline carrying a substituting mutation at this site. Previous studies indicate that threonine 56 is not a target of Raf-1 and protein kinase C, which phosphorylate serine 17 and serine 70, respectively (15,30). However, threonine 56 is included in four possible phosphorylation sites in COS-7 cells transfected with activated c-Jun N-terminal kinase (31). Therefore, threonine 56 may serve as a convergent point to integrate extracellular stress-related signals (via c-Jun N-terminal kinase) and the cell cycle checkpoint pathways (via CDC2). In this regard, it is important to clarify the functional consequence of Bcl-2 phosphorylation at threonine 56.
Recent structural analysis clarified the relation between the function and domain structure of Bcl-2 (2). Bcl-2 protein possesses four conserved motifs known as Bcl-2 homology domains (BH1 to BH4). BH4 domain (the sequence between amino acids 12 and 28) is considered to be essential for anti-apoptotic function, since Bcl-2 mutant lacking BH4 was unable to block cell death (33). Further studies revealed that this region provides a direct binding site for mammalian CED-4 homologs, including Apaf-1, to prevent activation of caspases and further progression of apoptosis (34,35). On the other hand, heterodimerization among Bcl-2 family proteins is mediated through domains BH1 to BH3, and particularly, BH3 (the sequence between amino acids 97 and 104) is important for interaction with pro-apoptotic members of the family which antagonize the anti-  a The cells were seeded at an initial concentration of 5 ϫ 10 5 cells/ml, and cultured for 4 h in the absence (none) or presence of either A23187 (5 M) or cycloheximide (75 g/ml). The percentage of apoptotic cells was determined microscopically by counting more than 200 cells on Wright-Giemsa-stained cytospin slides. An apoptotic body was defined as a morphological marker of apoptosis in individual cells. Data shown are mean Ϯ S.D. of three independent experiments. apoptotic function of Bcl-2 (36 -38). The region between BH4 and BH3 (amino acids 32-80) constitutes a large unstructured loop and is rich in serine and threonine residues including threonine 56, suggesting that it is a target for post-translational modifications (39,40). Our present observation provides substantial information to this view.
There are some studies regarding the function of the loop region between BH4 and BH3. The anti-apoptotic function of Bcl-2 was improved by deletion of the entire loop region, indicating that this segment serves as a negative regulatory domain (41). It is of note that phosphorylation was no longer induced by okadaic acid in this loop deletion mutant, which is in line with the initial part of the present study. Furthermore, Uhlmann et al. (42) reported that Bcl-2 mutant lacking a large part of the loop region (amino acids 51-85) showed a reduced ability to inhibit cell cycle progression but retained anti-apoptotic activity. Similarly, selective loss of the growth inhibitory effect of Bcl-2 was obtained by the substitution of tyrosine 28 with alanine or phenylalanine (10). In this study, we also demonstrated that the mutation at threonine 56, but not serine 87, abrogated the cell cycle-regulatory effect of Bcl-2 without affecting anti-apoptotic function. These results suggest that the non-conserved loop region is involved in the regulation of the cell cycle-related function of Bcl-2 upon phosphorylation by CDC2 and/or c-Jun N-terminal kinase. In contrast, Raf-1-mediated phosphorylation of serine 17 may be related to the modulation of the anti-apoptotic function of Bcl-2, since this site is located within the BH4 domain. This hypothesis is supported by several reports which describe the association of phosphorylation and apoptosis induction by anti-cancer drugs (12,13,30,32,43,44).
What is the physiological role of the Bcl-2 phosphorylation by CDC2? We showed that Bcl-2 phosphorylation was closely associated with the accumulation of cells in G 2 /M and was not necessarily proportional to the levels of apoptosis unlike Raf-1-induced phosphorylation. Moreover, we found that Bcl-2 is phosphorylated during G 2 /M phases in normal cell cycle, suggesting a physiological role in cell cycle regulation. Given that phosphorylation at threonine 56 is essential for the cell cycle inhibitory effect, CDC2-mediated phosphorylation of Bcl-2 may be implicated in M phase-specific negative regulatory events such as a mitotic checkpoint control (45,46). Evidence supporting this view includes the observation that Bcl-2 protein is localized in the nucleus at M phase, whereas it mainly presents in the cytoplasm at interphase (47). This hypothesis can also explain why Bcl-2 protein is selectively phosphorylated by microtubule-disrupting agents such as paclitaxel and vincristine but not by anti-cancer drugs damaging DNA (43). Alternatively, phosphorylation may result in the translocation of Bcl-2 from cytoplasm to the nucleus. Investigations are currently underway in our laboratory to clarify the mechanisms of the cell cycle regulatory function of Bcl-2 and their relationship to phosphorylation.