B-1 lymphocytes represent a distinct B cell subset with unusual mitogenic responses. PMA alone promotes proliferation in B-1 cells, but not in splenic B-2 cells. Although cyclin D2-cyclin-dependent kinase 4 (cdk4) complexes mediate early retinoblastoma gene product (pRb) phosphorylation in B-1 cells, the transient nature of their accumulation cannot account for the continued increase in pRb phosphorylation, which is maximal at 24 h. We show herein that PMA promotes the accumulation of functional cyclin D3-cdk4 complexes in B-1 cells following loss of cyclin D2. PMA also induces accumulation of cyclin D3-cdk4 complexes in B-2 cells; however, these complexes do not phosphorylate pRb. Thus, PMA is sufficient to induce synthesis and assembly of cyclin D3-cdk4 complexes in B-1 and B-2 cells; however, PMA triggers cyclin D3-cdk4 activation only in B-1 cells. These results reveal a novel regulatory step that controls activation of cyclin D3-cdk4 complexes whose function segregates differentially in B cell subsets.
B-1 cells constitute a unique subset of B lymphocytes, distinguished from conventional B lymphocytes (B-2) by numerous phenotypic and functional characteristics (reviewed in Refs. 1, 2, 3). As an example, B-1 cells localize primarily to the peritoneum, whereas B-2 cells predominate in spleen and lymph nodes. B-1 cells contribute substantial proportions of nonimmune (resting) IgM and IgA that is repertoire restricted. Whether B-1 cells represent a developmental lineage distinct from B-2 cells or, alternatively, derive from a single B cell lineage in which B-2 cells differentiate to B-1 cells in response to B cell Ag receptor (BCR)6-derived signals remains a matter of controversy (4, 5, 6). Although B-1 cells resemble activated B-2 cells in terms of surface expression of IL-5R, CD44, and nuclear, activated STAT3 (7, 8), many additional molecular and transcriptional markers associated with B-2 cell activation are absent in B-1 cells (9, 10).
B-1 cells differ significantly from B-2 cells in the signals required to induce proliferation. B-1 cells fail to enter S phase in response to anti-Ig, whereas B-2 cells are mitogenically stimulated by BCR cross-linking (11, 12). Treatment with phorbol ester alone is sufficient to stimulate B-1 cells to enter S phase, whereas B-2 cells exit quiescence, but subsequently arrest in G1 phase of the cell cycle (progression to S phase requires a second signal provided by calcium ionophore) (11, 12). The molecular basis underlying the unique proliferative response of B-1 cells to phorbol esters is not completely understood.
It is generally considered that growth signals regulate mammalian cell cycle entry by stimulating the accumulation of D-type cyclins (cyclins D1, D2, and D3) that function to activate a subset of cyclin-dependent kinases (cdks) (reviewed in Ref. 13). The retinoblastoma gene product (pRb) is a target of cdks and acts to suppress G1-to-S phase progression (14, 15, 16). pRb is presently the most plausible candidate for regulating progression through the restriction (R) point (14, 16). A current model holds that pRb suppression is alleviated through hyperphosphorylation that is mediated by both cyclin E and D-type cyclin kinase complexes (15, 16). The proper timing and extent of cdk activation is controlled by dephosphorylation of inhibitory sites, phosphorylation of activating sites, the action of two families of cdk inhibitors (Ink4 and Cip/Kip family), and by cyclin binding (17). For example, D-type cyclins and cyclin E function as positive regulatory subunits for cdk4/6 and cdk2, respectively (13). The requirement of mammalian cyclins D1 and D2 in G1 phase progression has been definitively established (18, 19). In keeping with this, distinct phenotypes have been reported in cyclin D1−/− and cyclin D2−/− mice (20, 21). Recent studies suggest that cyclin D3 may function to limit the rate of G1 phase progression (18, 22). Interestingly, in cyclin D2−/− mice, B-2 cells appear to remain responsive to mitogenic signals due to a compensatory induction of cyclin D3 levels (23).
Cell cycle progression to S phase in normal B-2 cells requires the accumulation of cyclin D2 and cdk4 and to a lesser extent cdk6 (24, 25, 26). B-2 cells from xid mice, which exhibit aborted activation in response to BCR cross-linking, do not up-regulate cyclin D2 protein, suggesting that accumulation of cyclin D2 in B-2 cells may be linked to passage through the R point (26). Interestingly, Solvason et al. (27) recently demonstrated a requirement for cyclin D2 expression in CD5 B cell development. We have previously demonstrated that phorbol ester induces unusually early and transient expression of cyclin D2 in B-1, but not in B-2 cells (28). As such, we proposed that the early induction of cyclin D2 may account for the rapid entry of B-1 cells into the cell cycle following phorbol ester stimulation. However, the transient nature of cyclin D2 expression in phorbol ester-stimulated B-1 cells suggests that it is unlikely to be responsible for progression through the G1/S transition. Herein, we report experiments demonstrating that in phorbol ester-stimulated B-1 cells, cyclin D3-cdk4 complexes assemble and are active pRb protein kinases in late G1 phase of the cell cycle. By contrast, phorbol ester stimulation of splenic B-2 cells results in the assembly of cyclin D3-cdk4 complexes; however, these complexes fail to phosphorylate pRb in vitro. These findings constitute the first demonstration in a primary mammalian cell that cyclin D-cdk complex assembly and activation can be dissociated and regulated by different signals.
Materials and Methods
Male BALB/cByJ mice at 8–14 wk of age were obtained from The Jackson Laboratory (Bar Harbor, ME). Mice were cared for and handled at all times in accordance with National Institutes of Health and institutional guidelines.
B cell purification
B-1 and B-2 lymphocytes were prepared by negative selection from peritoneal washout cells and from spleen cell suspensions, respectively, as described (9). The recovered B-1 cell population was 90–96% sIgM+, CD5/Mac-1+ by flow cytometric analysis. The B cells were cultured at 37°C with 5% CO2 in RPMI 1640 medium (BioWhittaker, Walkersville, MD) supplemented with 10% heat-inactivated FBS (Sigma, St. Louis, MO), 10 mM HEPES (pH 7.2), 50 μM 2-ME, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin.
B cells were solubilized in 1 ml Nonidet P-40 buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 20 mM EDTA, 0.5% Nonidet P-40, 1 mM PMSF, 2.5 μg/ml leupeptin/aprotinin, 1 mM Na3VO4, and 10 mM β-glycerophosphate) for 30 min (4°C) (28). Insoluble debris was removed by centrifugation at 15,000 × g for 15 min (4°C). The detergent-soluble cell lysates were incubated for 3 h with 1.5 μg nonimmune IgG or 1.5 μg anti-cdk4 Ab, or 1.5 μg anti-cdk6 Ab, followed by the addition of 50 μl of a 1:1 slurry of protein G-agarose. After 90 min, the immune complexes were collected, washed several times in Nonidet P-40 buffer, and separated by electrophoresis through a 10% polyacrylamide SDS gel. The proteins were transferred to Immobilon-P membrane (Millipore, Bedford, MA) and immunoblotted with an anti-cyclin D3 mAb (1:500 dilution in TBST) as described below.
Immune complex kinase assays
B cells were sonicated (4°C) in 1 ml Rb buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM EDTA, 2.5 mM EGTA, 1 mM DTT, 0.1% Tween20, 10% glycerol, 0.1 mM PMSF, 1 μg/ml leupeptin/aprotinin, 10 mMβ-glycerophosphate, 1 mM NaF, and 0.1 mM Na3VO4) (29). Insoluble material was removed by centrifugation, and the supernatant was incubated with 1.5 μg nonimmune rabbit IgG, 1.5 μg of anti-cdk4, or 1.5 μg of anti-cdk6 Abs. After 3 h, 50 μl of a 1:1 slurry of protein G-agarose was added and incubated for 1 h. The immune complexes were recovered by centrifugation and washed six times with Rb buffer and then three times in a buffer of 50 mM HEPES, pH 7.4, and 1 mM DTT. The immune complexes were resuspended in 30 μl of Rb kinase buffer (50 mM HEPES, pH 7.5, 10 mM MgCl2, 5 mM MnCl2, 1 mM DTT, 2.5 mM EGTA, 10 mM β-glycerophosphate, 0.1 mM Na3VO4, and 10 μCi γ[32P]ATP at 6000 Ci/mmol) in the presence of 1 μg of a truncated Rb protein substrate (p56Rb). After 15 min at 30°C, the reactions were terminated by the addition of 2× SDS sample buffer, and the reaction products were separated through a 10% polyacrylamide SDS gel. Phosphorylated Rb was detected by autoradiography of the dried gel.
For the detection of cyclins D2 and D3, B lymphocytes were solubilized in 100 μl Triton X-100 buffer (50 mM HEPES, pH 7.4, 15 mM EGTA, 137 mM NaCl, 15 mM MgCl2, 0.1% Triton X-100, 10 mM β-glycerophosphate, 1 mM Na3VO4, 1 mM PMSF, and 1 μg/ml aprotinin/leupeptin); for detection of endogenous pRb, B cells were solubilized in 100 μl Nonidet P-40 buffer containing 20 mM NaF (28). Insoluble debris was removed by centrifugation at 15,000 × g (15 min), and 10–20 μg of total protein was separated by polyacrylamide SDS gel electrophoresis and transferred to Immobilon-P membrane. Immune detection was conducted as previously described (24).
F(ab′)2 of goat anti-mouse IgM was obtained from Jackson ImmunoResearch (West Grove, PA). PMA and the calcium ionophore, ionomycin, were obtained from Sigma. Human pRb mAb (clone G3-245) was obtained from PharMingen (San Diego, CA). Anti-rabbit and anti-mouse IgG-conjugated HRP Abs and anti-cdk4 Ab (sc-260) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-cdk6 Ab (13446E) was obtained from PharMingen. Protein G-agarose was obtained from Life Technologies (Gaithersburg, MD). Mouse anti-cyclin D2 Ab (DCS-3 and DCS-5) and anti-cyclin D3 Ab (DCS-22) were a gift from Jiri Bartek (Division of Cancer Biology, Danish Cancer Society, Copenhagen, Denmark) (19). The truncated Rb substrate protein (p56Rb) was obtained from QED Advanced Research Technologies (San Diego, CA). Enhanced chemiluminescence reagents were obtained from Kirkegaard & Perry Laboratories (Gaithersburg, MD).
Results and Discussion
We previously demonstrated that B-1 cell stimulation with the phorbol ester, PMA, produced cyclin D2 accumulation in a rapid and transient manner (28). The initial onset of Cdk4-mediated pRb phosphorylation on Ser780 correlated with the accumulation and assembly of cyclin D2 into higher order complexes containing cdk4. However, we noted that phosphorylation of pRb continued to increase as B-1 cells progressed through the G1-to-S phase transition. Interestingly, the maximal level of pRb phosphorylation occurred near the G1/S transition at a time when cyclin D2 was not expressed. These findings suggested that PMA stimulation of B-1 cells promotes the accumulation of additional G1-cyclin complexes capable of phosphorylating pRb. To investigate the nature of G1-cyclin complexes that might contribute to the phosphorylation of endogenous pRb in late G1 phase, we evaluated the expression of cyclin D3 in B-1 cells, noting that cyclin D1 is not expressed in murine B lymphocytes (24, 25, 26). B-1 cells were cultured in medium alone or stimulated with PMA for various times; cells were then collected and detergent-solubilized proteins were separated by SDS-PAGE and immunoblotted with an anti-cyclin D3 mAb (Fig. 1). Cyclin D3 was not detected in B cells cultured in medium alone. Stimulation of B-1 cells with PMA induced the accumulation of cyclin D3 at the 17- and 24-h time points. No cyclin D3 was expressed after PMA stimulation for 4 h, the point at which cyclin D2 expression peaked (28), indicating that PMA-stimulated accumulation of cyclin D3 and cyclin D2 are distinct. The accumulation of cyclin D3 paralleled peak endogenous pRb phosphorylation in PMA-stimulated B-1 cells (Fig. 2). PMA also induced the accumulation of cyclin D3 in B-2 cells; however, in contrast to B-1 cells, PMA stimulation of B-2 cells at parallel time points (i.e., 4, 16, 24 h) did not lead to increased pRb phosphorylation. As a positive control for these experiments, pRb phosphorylation was induced in B-2 cells stimulated with several different mitogens, including 25 μg/ml LPS, 10 μg/ml anti-Ig and the combination of PMA (300 ng/ml) plus ionomycin (400 ng/ml) for 24 and 36 h (Fig. 2).
The timing of cyclin D3 accumulation in B-1 cells suggests that it may contribute to the phosphorylation of endogenous pRb. To test this further, we sought to determine whether PMA promotes the assembly and activation of cyclin D3-cdk complexes at a time commensurate with peak PMA-induced phosphorylation of endogenous pRb. B-1 and B-2 cells were cultured in medium alone or were stimulated with PMA for 4 and 24 h, at which times cells were detergent-solubilized and immunoprecipitated with rabbit anti-mouse cdk4 or anti-mouse cdk6 Abs (Fig. 3). The immune complexes were separated by SDS-PAGE and immunoblotted with anti-cyclin D3 mAb. PMA stimulation of B-1 cells induced the assembly of cyclin D3-cdk4/6 complexes, as evidenced by the appearance of cyclin D3 in cdk4 and cdk6 immunoprecipitates at 24 h. Surprisingly, the same was true of PMA-stimulated B-2 cells. No detectable cyclin D3 was present in cdk4/6-immune complexes isolated from unstimulated B-1 or B-2 cells, and in parallel experiments, cyclin D3 was not detected in any sample after immunoprecipitation with nonimmune serum (data not shown). Importantly, cyclin D3-cdk4/6 holoenzyme complexes were not detected in B-1 cells stimulated with PMA for 4 h, which is consistent with the lack of detectable cyclin D3 expression at this time (see Fig. 1).
To determine whether the PMA-induced cdk4- and cdk6-containing complexes were functional, Rb protein phosphorylation activity in cdk4 and cdk6 immunoprecipitates was analyzed using a recombinant COOH-terminal-truncated Rb protein as substrate (29). In extracts prepared from B-1 cells stimulated with PMA for 24 h, Rb phosphorylation was produced by cdk4 immune complexes and to a much lesser extent by cdk6 immune complexes (Fig. 4). Nonimmune complexes were devoid of Rb kinase activity (data not shown). In parallel experiments, Rb phosphorylation was not stimulated above control by cdk4 or cdk6 immune complexes isolated from PMA-stimulated B-2 cells. As a positive control, cdk4 and to a lesser extent cdk6 immune complexes recovered from B-2 cells stimulated with the combination of PMA plus calcium ionophore exhibited inducible Rb kinase activity (Fig. 4, lane P/I).
We previously reported that cyclin D2 accumulates rapidly and in a transient manner following stimulation of B-1 cells with PMA alone (28). Commensurate with cyclin D2 accumulation, B-1 cells express phosphorylated pRb on Ser780, which increases in a time-dependent manner. These results suggest that cyclin D2 plays an important role in early G1 phase progression in B-1 cells during PMA-mediated proliferation. Although our previous findings in B-1 cells are consistent with the emerging view that cyclin D2 is the primary G1 cyclin involved in passage through the R point during BCR-mediated B-2 cell proliferation (23, 24, 25, 26), we note that the majority of pRb phosphorylation occurs in mid-to-late G1 phase and at a time during which cyclin D2 is not detected in B-1 cells by Western blot analysis (28). This observation points to the presence of additional G1 cyclins that contribute to pRb phosphorylation following loss of detectable cyclin D2. The results presented herein indicate that cyclin D3 accumulates in response to PMA stimulation of B-1 cells. Furthermore, the timing of cyclin D3 accumulation, which was distinct from that of cyclin D2, suggests the involvement of cyclin D3 in PMA-induced mid-to-late G1 phase progression. Consistent with this conclusion, cyclin D3 assembled into higher order complexes containing both cdk4 and cdk6 in response to PMA in B-1 cells. Furthermore, cdk4 and to a lesser extent cdk6 immune complexes from PMA-stimulated B-1 cells were capable of phosphorylating Rb in vitro at levels that were substantially greater than parallel immune complexes recovered from control cells. Taken together, these results suggest a role for cyclin D3-cdk4 holoenzyme in passage through the G1/S phase transition during PMA-mediated B-1 cell proliferation. In keeping with our findings herein, mounting evidence in several cell types supports a role for cyclin D3 function in the control of mammalian cell G1/S transition (19, 22, 30).
The accumulation and assembly of substantial levels of cyclin D3-containing cdk complexes in B-2 cells following PMA stimulation was unexpected. The absence of pRb kinase activity associated with these complexes in B-2 cells suggests that an additional signal(s) is (are) required for function and points to the existence of a hitherto unknown regulatory step in B cells that controls the activation of preformed cyclin D3-cdk complexes. It is well established that formation of cyclin D holoenzyme complexes relies upon growth factor signals that act both transcriptionally, to induce accumulation of D-type cyclin and cdk, and posttranslationally, to promote cyclin D-cdk assembly (17, 30). For example, ectopically expressed cyclin D1 and cdk4 subunits are not active in NIH 3T3 cells in the absence of serum because they fail to assemble in the absence of mitogenic signals. The mitogenic signal for assembly can be provided by activation of the Ras/Raf-1/Erk pathway or by ectopic expression of MEK1 (30). Thus, growth factor signals not only function to induce cyclin D1 transcription, but also to promote assembly of cyclin D1 into cdk4-containing catalytically active complexes. Our findings in B-2 cells suggest that accumulation and assembly of cyclin D3-dependent kinases is not sufficient to induce kinase activation. To our knowledge, this constitutes the first demonstration in a primary (not genetically engineered) mammalian cell that cyclin D-cdk accumulation/assembly and D-type cyclin-cdk activation can be dissociated and are regulated by different signals.
At present the nature of the regulatory step necessary to promote activation of assembled cyclin D3-cdk complexes in B-2 cells is unknown. Given the presence of assembled D-type cyclin-cdk complexes, it is unlikely that formation of inactive binary cdk/Ink4 complexes or decreased expression of D-type cyclin or cdk accounts for the observed inhibition of cyclin D3-cdk complex activity in B-2 cells (13, 17). PMA treatment of B-2 cells does not alter the relative amount of D-type cyclin-associated p21Cip1 and p27Kip1 proteins in comparison to control cells (data not shown). Thus, regulation of cyclin D3-cdk complex activity might depend on a previously uncharacterized enhancing protein, a previously uncharacterized inhibitory protein, or a posttranslational modification of the D-type cyclin and/or cdk4/6 subunits (e.g., phosphorylation on an as yet unmapped amino acid residue) induced by PMA in one B cell population and not the other. Studies are presently underway to further understand the molecular step(s) that regulate cyclin D3-cdk complex activation in B lymphocytes.
We thank Dr. Jiri Bartek (Division of Cancer Biology, Danish Cancer Society, Copenhagen, Denmark) for the murine D-type cyclin mAbs.
This work was supported by Grant MCB-9603784 awarded by the National Science Foundation (to T.C.C.) and U.S. Public Health Service Grant AI29690 awarded by the National Institutes of Health (to T.L.R.).
Abbreviations used in this paper: BCR, B cell Ag receptor; pRb, retinoblastoma gene product; cdk, cyclin-dependent kinase; R point, restriction point.