MEK/ERK-Mediated Phosphorylation of Bim Is Required to Ensure Survival of T and B Lymphocytes during Mitogenic Stimulation

The rates of cell proliferation, cell death, and cell differentiation are coordinated with respect to each other to ensure proper development and functioning of an organism (1). During differentiation in primary lymphoid organs and after activation in the periphery, B and T lymphocytes undergo successive rounds of cell division and growth arrest followed by the death of cells that are useless or potentially dangerous (2). Whether a cell lives or dies is regulated to a large extent by the balance between pro- and antiapoptotic members of the Bcl-2 protein family. The antiapoptotic members Bcl-2, Bcl-x_L, Bcl-w, A1, and Mcl-1 share up to four regions of homology (Bcl-2 homology or BH regions), and gene targeting experiments have shown that they are critical for cell survival and function in a cell type-specific manner (3). Bax, Bak, and Bok/Mtd have three BH regions and despite extensive structural similarity to their antiapoptotic relatives, they are essential for activation of the downstream apoptosis effector processes, including mitochondrial outer membrane permeabilization and unleashing of the caspase cascade (4). The BH3-only proteins (Bad, Bik/Blk/Nbk, Hrk/DP5, Bid, Bim/Bod, Bmf, Noxa, Puma/Bbc3) contain only the BH3 domain, and they are required to initiate apoptosis signaling (5). Biochemical and genetic experiments have indicated that BH3-only proteins trigger apoptosis by binding to prosurvival Bcl-2 family members thereby unleashing the Bax/Bak-like proteins (6), but direct activation of Bax/Bak has also been proposed (4).

Bim, a proapoptotic BH3-only Bcl-2 family member (7), is expressed in a broad range of tissues, including hemopoietic, epithelial, neuronal, and germ cells (8). Alternative splicing generates several Bim isoforms (including Bim_EL, Bim_L, and Bim_S; Ref. 7) of which Bim_EL is the most abundantly expressed (8). Studies with gene-targeted mice have shown that Bim is required for developmentally programmed cell death as well as cytokine deprivation or stress-induced apoptosis of T as well as B lymphocytes, myeloid cells, neurons, and epithelial cells (9, 10, 11, 12, 13).

The proapoptotic activity of Bim can be regulated by several transcriptional as well as posttranslational mechanisms (14). For example, cytokine withdrawal causes shutdown of the PI3K/AKT signaling pathway, resulting in the activation of the transcription factor FOXO3A, which induces Bim mRNA synthesis (15). In response to cytokine withdrawal, the levels and proapoptotic activity of Bim can also be controlled by posttranslational mechanisms (14). In growth factor-stimulated cells, Bim is phosphorylated by ERK1/2 on multiple sites, which is thought to reduce its binding to Mcl-1 and Bcl-x_L (16) and was reported to also target it for ubiquitination and proteasomal degradation (17, 18).

We investigated the control of Bim in mouse T and B cells during transition from the resting to the cycling state after mitogenic stimulation. We found that in naive, resting T and B cells Bim exists in a hypophosphorylated form but is expressed at relatively low levels. Upon mitogenic stimulation, Bim is rapidly phosphorylated in a MEK/ERK-dependent manner and subsequently declines in level. Studies using pharmacological inhibitors and gene-targeted mice showed that MEK/ERK-mediated phosphorylation of Bim is required for survival of mitogen-stimulated B and T cells but that cell cycling proceeds by a MEK/ERK-dependent mechanism that is independent of Bim inactivation.

All experiments with animals were performed according to the guidelines of The Walter and Eliza Hall Institute Animal Ethics Committee. Wistar rats and C57BL/6 mice were obtained from our Institute’s breeding facility at Kew, Australia. The Bim^−/− (9), vav-bcl-2-transgenic (19), Bad^−/− (20), Bid^−/− (21), Bim^−/−Bad^−/− (22), and Bim^−/−Bid^−/− (6) mice have all been previously described. The Bim^−/− and Bad^−/− mice were both originally produced on a mixed C57BL/6 × 129SV background using 129SV-derived ES cells but have been backcrossed for >10 generations with C57BL/6 mice before they were used for these studies. The vav-bcl-2 transgenic mice, expressing constitutively high levels of (human) Bcl-2 in all hemopoietic cell types, were generated on an inbred C57BL/6 background. The Bid^−/− mice were produced on an inbred C57BL/6 background using C57BL/6-derived ES cells.

B lymphocytes were purified from spleen and lymph nodes of mice by negative sorting after staining of unwanted cell types (T cells, macrophages, granulocytes, and erythroid cells) with FITC-conjugated surface marker-specific rat mAbs (RB6-8C5, anti-Gr-1; F4/80, anti-macrophage marker; MI/70, anti-Mac-1; H129.1, anti-CD4; YTS169, anti-CD8; T24.3.21, anti-Thy-1; and Ter119, anti-erythroid marker). Viable B cells that were not stained with FITC-labeled mAbs or the vital dye propidium iodide (PI⁴; Sigma-Aldrich) at 2 μg/ml (FITC⁻PI⁻) were purified on a MoFlo (Cytomation) or the FACStar⁺ (BD Biosciences) high-speed cell sorter. Staining with Abs to CD19, CD45R-B220, IgM, and/or IgD followed by FACS analysis revealed that B cell purity was typically >95%.

T lymphocytes were purified from lymph nodes by depletion of all other cell types (B cells, macrophages, granulocytes, and erythroid cells) by staining with cell surface marker-specific rat mAbs (RA3-6B2, anti-CD45R-B220; 5.1, anti-IgM; 11-26C, anti-IgD; RB6-8C5, anti-Gr-1; M1/70, anti-Mac-1; Ter119, anti-erythroid marker) and magnetic beads coupled with goat anti-rat IgG Abs (Qiagen). Staining with Abs to Thy-1, CD4, and CD8 revealed that T cell purity was typically >95%.

Purified B lymphocytes from spleen and lymph nodes were stimulated for 24–48 h in culture with 20 μg/ml F(ab′)₂ goat anti-mouse IgM Ab fragments (Jackson Immunoresearch Laboratories) plus saturating concentrations of recombinant mouse ILs (IL-2, −4, and −5), produced by X63/0 hybridoma cells transfected with the relevant expression vectors (23).

Purified T lymphocytes from lymph nodes were stimulated for 24–48 h in culture with 2 ng/ml PMA plus 0.1 μg/ml ionomycin (both from Sigma-Aldrich) and IL-2 or with plate-bound hamster mAbs to CD3 (145-2C11) and CD28 (37N51), both at 10 μg/ml in the coating solution (PBS) plus IL-2 or with 0.1–1.0 μg/ml ionomycin alone. Mitogen-stimulated B and/or T cells were also treated with either JNK inhibitor 1 (l stereoisomer; Axxora) at 1–25 μg/ml, U0126 MEK1/2 inhibitor (Cell Signaling) at 10 or 20 μg/ml (with drugs replenished after 24 h), the proteasomal inhibitors PS341/Velcade (Janssen-Cilag) at 10–50 μg/ml, or MG132 (Calbiochem) at 10–50 μg/ml, both added 2 h before and during mitogen stimulation.

Thymocytes, spleen cells, purified mature T cells, or thymocyte subpopulations were fixed for 10 min at room temperature with PBS containing 1% paraformaldehyde (BDH), permeabilized, and stained for 30 min at 4°C in PBS/0.3% saponin (Sigma-Aldrich) plus 2% FCS containing rat anti-Bim mAbs (3C5 or 10B12 at 5 μg/ml; Alexis) or an isotype-matched control mAb (rat IgG2a; BD Pharmingen), followed by secondary staining with biotin-conjugated mouse anti-rat IgG2a Ab (BD Pharmingen), and detected with streptavidin-FITC (Caltag) as detailed (24) and analyzed in a FACScan (BD Biosciences).

For cell cycle analysis, cells were fixed (>8 h at 4°C) in 70% ethanol and then treated for 20 min at 37°C with 0.5 μg/ml DNase-free RNase A (Promega) with 69 mM PI (Sigma-Aldrich) in 0.1 M sodium citrate (pH 7.4). Flow cytometric analysis (10,000 cells per sample) was performed on a FACScan and cell cycle distribution determined by manual gating. Statistical analysis was performed using a two-sided Student t test with equal variance.

Cell lysates were prepared, and proteins were size-fractionated on polyacrylamide gels (Novex) and transferred to nitrocellulose membranes (Amersham Pharmacia) as described previously (8). Nonspecific binding of Abs to membranes was blocked by overnight incubation in PBS, 5% skim milk, 1% casein, 0.05% Tween 20. Membranes were then probed with the following Abs: rat anti-Bim mAbs (clones 3C5 or 10B12, Alexis); rat isotype-matched control Ab (IgG2a/κ; BD Pharmingen); rabbit polyclonal anti-Bim Ab (Stressgen); hamster anti-mouse Bcl-2 mAb (3F11); rabbit polyclonal anti-Mcl-1 Ab (Rockland); or rabbit anti-phospho-p44/42 MAP kinase (ERK1/2) Ab (Cell Signaling). To control for the concentration and integrity of proteins in the tissue lysates, blots were probed with mouse anti- heat shock protein 70 (HSP70) mAb N6 (gift from Drs. R. Anderson, Peter MacCallum Cancer Research Institute, Melbourne, Australia, and W. Welch, University of California, San Francisco, CA) or a mouse anti-actin Ab (Sigma-Aldrich). Bound Abs were visualized with goat anti-rat IgG (Southern Biotechnology), sheep anti-rabbit Ig (Chemicon), or sheep anti-mouse IgG Abs (Chemicon), all conjugated to HRP, followed by ECL (Amersham Pharmacia). Quantitative Western blotting was performed by comparing the strength of the Bim signal to that of the HSP70 or actin signal after staining with secondary goat anti-rat IgG or goat anti-rabbit IgG Abs conjugated to IR dye 800 (Rockland) or Alexa Fluor 680 or 800 dye (Molecular Probes), using the Odyssey infrared imaging system according to the manufacturer’s instructions (Li-Cor Biosciences).

Cell lysates (prepared as above without NaF or Na₃VO₄) were incubated for 1 h at 30°C with λ-PPase (NEB; 15 μg/200 U) according to the manufacturer’s recommendations. λ-PPase activity was inhibited by the inclusion of NaF (5 mM) and Na₃VO₄ (2 mM).

Several studies have indicated that Bim is expressed at low levels in resting lymphoid cells (7, 8). We confirmed that thymocytes and spleen cells express only low levels of Bim by staining such cells from wild-type (wt) and, as a negative control, those from Bim^−/− mice with novel anti-Bim mAbs (3C5 or 10B12) that we developed (supplemental Fig. 1)⁵ followed by flow cytometric analysis (Fig. 1,A). As a further control, we stained these cells with an Ig isotype-matched control mAb. Upon exposure to a Bim-dependent apoptotic stimulus, such as treatment with the calcium ionophore ionomycin (9), both wt thymocytes and splenic T cells showed a marked increase in anti-Bim Ab staining intensity in comparison with cells from Bim^−/− mice or staining with an isotype-matched control Ab (Fig. 1,A). Quantitative Western blot analysis using the Odyssey infrared imaging system revealed a 3.8-fold elevation in Bim_EL levels in thymocytes after 24 h of exposure to ionomycin (Fig. 1,B). In contrast, treatment with the protein kinase C activator, PMA, did not cause an increase in Bim protein levels in thymocytes (Fig. 1, B and C). This is consistent with the observation that PMA kills thymocytes by a mechanism that is independent of Bim (9) but instead requires the BH3-only protein Puma (25).

Next, we performed a time course analysis to examine the differing effects of PMA or ionomycin on Bim expression levels in both thymocytes and mature T cells. Treatment with ionomycin caused a gradual (up to ∼3.5-fold) increase in Bim_EL levels peaking at 24 h. In contrast, within 1–6 h of treatment with PMA we observed an upward electrophoretic mobility shift in both Bim_EL and Bim_L (with Bim_EL being the predominant isoform; Fig. 1,C), suggestive of increased phosphorylation. This was accompanied by a gradual decline in the levels of Bim (Fig. 1,C). The levels of Bcl-2 remained relatively constant after exposure to PMA or ionomycin, whereas the levels of Mcl-1 had declined significantly by 24 h (Fig. 1,C). Activation of p44 and p42 MAP kinases (ERK1 and ERK2), revealed by staining with phospho-ERK-specific Abs, was evident within 0.5 h of treatment with PMA (Fig. 1 C).

Bim has been shown to be required for cytokine deprivation induced apoptosis of activated T lymphocytes (9). Accordingly, we found that withdrawal of IL-2 from Con A-activated splenic T cell blasts resulted in a ∼3- to 4-fold increase in Bim_EL expression (Fig. 1 D), but no upward mobility shift on SDS-PAGE was evident. Growth factor deprivation did not cause an obvious change in the levels of Bcl-2 or Mcl-1 during the period of observation. Collectively, these results show that some but not all apoptotic stimuli cause an increase in Bim (particularly Bim_EL) expression in immature as well as mature T lymphoid cells.

During the induction of an immune response, the control of cell death and cell cycling must be coordinated (26). We therefore investigated whether Bim was modified in mitogenically activated mouse T cells during the transition from the resting (G₀) to the cycling state. A time course analysis of T cells stimulated with PMA plus ionomycin plus IL-2, or anti-CD3 plus anti-CD28 mAbs plus IL-2, showed an upward electrophoretic mobility shift of Bim_EL, detectable from 1 h (Fig. 2,A). The slower migrating band detected with the anti-Bim Abs clearly represents a modified form of Bim and not a cross-reacting protein in that it was not observed in lysates from resting or mitogenically stimulated Bim^−/− T cells (Fig. 2,B). Beyond 6 h of stimulation, there was a gradual decline in Bim to levels below those found in resting T cells (Fig. 2,A). Activation of ERK1/2, as visualized by staining with phospho-ERK specific Abs (Fig. 2,A), was evident from 1 h and was maximal at 6 h, paralleling the upward electrophoretic mobility shift in Bim_EL. All of these changes, particularly ERK1/2 activation, were more pronounced for stimulation with PMA plus ionomycin compared with treatment with Abs to CD3 plus CD28, consistent with a previous report (27). Mitogenic stimulation also resulted in a gradual decline in Bcl-2 levels over 24 h, while the levels of Mcl-1 remained constant (Fig. 2 A).

To examine whether the modification of Bim (upward electrophoretic mobility shift) was due to phosphorylation, lysates from mitogen-activated (wt) T cells were incubated with λ-PPase, which dephosphorylates modified serine, threonine, and tyrosine residues. This resulted in the disappearance of the slower migrating form of Bim and increased abundance of the nonmodified form (Fig. 2,C). Addition of the λ-PPase inhibitors NaF or Na₃VO₄ prevented the appearance of the more rapidly migrating (dephosphorylated) form of Bim (Fig. 2 C). Collectively, these results show that Bim is phosphorylated during mitogenic activation of T lymphocytes.

The MEK/ERK pathway has been shown to be critical for cell cycle progression and survival (28) and is known to inhibit the proapoptotic activity of Bim (18). Therefore, and because ERK activation paralleled the changes in Bim mobility on SDS-PAGE (Figs. 1 and 2), we examined the impact of the MEK/ERK pathway on the phosphorylation of Bim during mitogenic stimulation of T cells. Treatment with U0126, an inhibitor of MEK1/2, the upstream activators of ERK1/2, abrogated the Bim phosphorylation that was observed in mitogenically stimulated T cells (Fig. 3,A, lanes 2 and 6), so that only the nonphosphorylated form of Bim_EL could be detected (Fig. 3,A, lanes 3 and 7). As a control, treatment with DMSO (vehicle control) had no effect on the migration of Bim on SDS-PAGE (Fig. 3 A, lanes 4 and 8).

It has been reported that Bim can also be phosphorylated by JNK on Thr¹¹² (29), which is thought to activate the proapoptotic of Bim (30 , 49) as does ERK-mediated phosphorylation on Thr¹¹², whereas ERK-mediated phosphorylation on Ser^55/65/73 has been reported to inhibit Bim proapoptotic activity and promote cell survival (18). Regardless, addition of the JNK inhibitor (cell permeable JNK inhibitor 1 specific for JNK1, JNK2, and JNK3) did not diminish Bim phosphorylation, but rather resulted in an accumulation of the phosphorylated form of Bim with increasing dosage of the JNK inhibitor (Fig. 3 B). These results show that the MEK/ERK signaling pathway is required for the changes in Bim phosphorylation during mitogenic stimulation of T lymphocytes.

Protein phosphorylation was reported to prime Bim for ubiquitination and proteasomal degradation (18). We therefore investigated whether the phosphorylation of Bim seen in mitogenically stimulated T cells primes this BH3-only protein for ubiquitination and proteasomal degradation. Addition of the proteasome inhibitors MG132 (Fig. 3,A, lane 9) or PS341 (bortezomide) did not block phosphorylation of Bim in mitogen-activated T cells but inhibited the decline in Bim levels, resulting in an accumulation of the phosphorylated form (Fig. 3, C and D). This protein band was not seen in lysates of similarly treated Bim^−/− T cells (not shown), demonstrating that it represents a posttranslationally modified form of Bim_EL. These results demonstrate that ERK1/2-mediated phosphorylation promotes a decline in Bim levels in mitogen-activated T lymphocytes that involves proteasomal degradation.

The MEK/ERK pathway is required for both survival and entry into the cell cycle during mitogenic stimulation (28). We examined the role of ERK1/2-mediated inactivation of Bim in the control of survival of mitogen-activated T cells by using the MEK1/2-specific inhibitor U0126 at a dose determined to be suitable in exploratory studies (Fig. 3,E and supplemental Fig. 2). Purified T cells from wt or Bim^−/− mice were stimulated with mitogens (anti-CD3/anti-CD28 mAbs plus IL-2 or PMA plus ionomycin plus IL-2) with further addition of the MEK1/2 inhibitor U0126 or DMSO (vehicle control), and cell viability was determined after 24 or 48 h by flow cytometric analysis. Fig. 4,A shows that treatment with the MEK inhibitor significantly increased apoptosis of mitogen stimulated wt T cells (24 h, 20 ± 3% (UO126) vs 9 ± 6% (DMSO); 48 h, 32 ± 6% (UO126) vs 5 ± 1% (DMSO)). In contrast, treatment with UO126 did not cause a significant increase in apoptosis of mitogen-stimulated Bim^−/− T cells (24 h, 8 ± 5% (UO126) vs 6 ± 3% (DMSO); 48 h, 11 ± 6% (UO126) vs 6 ± 4% (DMSO)). Treatment with the MEK inhibitor reduced the fraction of cycling (percent cells residing in S phase) cells in mitogen-stimulated T lymphocytes from wt (48 h, 11 ± 3% (UO126) vs 25 ± 2% (DMSO)) or Bim^−/− mice (48 h, 8 ± 1% (UO126) vs 19 ± 5.0% (DMSO)) to a similar extent (Fig. 4, B and C). Collectively, these results show that ERK1/2-mediated inactivation of Bim is critical for the survival of mitogen-stimulated T lymphocytes but does not play a role in controlling cell cycle entry.

Phosphorylation of Bim_EL has previously been observed in B lymphoma-derived cell lines stimulated with PMA or BCR cross-linking (31, 32). Therefore, we investigated whether the Bim phosphorylation that we observed in mitogen-stimulated T cells during the transition from the resting (G₀) to the cycling state did also occur when B cells are stimulated through their BCR. A time course analysis revealed a readily detectable increase in the electrophoretic mobility shift on SDS-PAGE in both Bim_EL and Bim_L in B cells within 1–3 h of stimulation with F(ab′)₂ goat anti-mouse IgM Ab fragments plus ILs (IL-2, IL-4, and IL-5; Fig. 5,A). Activation of ERK1/2, visualized by staining with phospho-ERK specific Abs, was evident within 1 h of mitogenic activation (Fig. 5,A), paralleling the electrophoretic mobility shift in Bim_EL and Bim_L. Treatment with the MEK inhibitor UO126 abrogated the BCR cross-linking-induced mobility shift in Bim_EL and Bim_L (Fig. 5,A), reminiscent of previous observations in B lymphoma-derived cell lines (31, 32). Addition of the proteasome inhibitors MG132 (Fig. 5,B) or PS341 (bortezomide) did not block phosphorylation of Bim in mitogen-activated B cells but inhibited the decline in Bim levels, resulting in an accumulation of the phosphorylated form (Fig. 5 B). These results show that in B cells, as in T cells, mitogenic stimulation causes ERK1/2-mediated phosphorylation of Bim_EL and Bim_L.

Because ERK1/2-mediated phosphorylation of Bim is required for survival of mitogen stimulated T cells, we examined the role of Bim in the survival of activated B cells. Purified B cells from wt or Bim^−/− mice were stimulated with F(ab′)₂ goat anti-mouse IgM Ab fragments plus ILs (IL-2, IL-4, and IL-5) with addition of the MEK1/2 inhibitor U0126 or DMSO (vehicle control) and measured apoptosis and cell cycle distribution by flow cytometric analysis. Fig. 6,A shows that MEK1/2 inhibition caused substantial apoptosis in mitogen stimulated wt B cells (48 h, 32 ± 2% (UO126) vs 16 ± 6% (DMSO)). Mitogen-stimulated B cells from Bim^−/− mice were less sensitive to MEK1/2 inhibition (48 h, 22 ± 3% (UO126) vs 5 ± 1% (DMSO); p = 0.01 wt vs bim^−/− 48 h with MEK inhibition), but protection was not complete and, in fact, was less prominent than the protection seen in mitogenically activated T cells. Flow cytometric analysis revealed that MEK1/2 inhibition reduced cycling (percent of cells in S phase) of mitogen-stimulated wt (48 h, 3 ± 1% (UO126) vs 18 ± 4% (DMSO)) and Bim^−/− (48 h, 2 ± 0.5% (UO126) vs 12 ± 3% (DMSO)) B cells to a similar extent (Fig. 6, B and C). Collectively, these results show that Bim contributes to MEK inhibition-induced killing of mitogen-stimulated B lymphocytes but appears to play no role in cell cycle entry.

Since loss of Bim afforded mitogen-stimulated B cells only with partial protection against MEK/ERK inhibition, whereas Bcl-2 overexpression inhibited this apoptosis completely (24 h, 42 ± 24.% (wt) vs 3 ± 1% (vav-bcl-2 tg), p < 0.05; 48 h, 54 ± 15% (wt) vs 6 ± 1% (vav-bcl-2 tg), p < 0.05; Fig. 7), we reasoned that one or more additional BH3-only proteins may contribute to this B cell death. The MEK/ERK signaling pathway has been reported to also promote cell survival through the activation of RSK-mediated phosphorylation of Bad (33), which inhibits the proapoptotic activity of this BH3-only protein by sequestration from Bcl-2 prosurvival family members. We therefore investigated whether mitogen-stimulated B cells lacking both Bim and Bad are more resistant to MEK/ERK inhibition than those lacking only Bim. However, cell survival analysis revealed that loss of Bad did not augment resistance of anti-IgM Ab stimulated Bim-deficient B cells to treatment with UO126 (Fig. 8,A) and also had no impact on the rate of cycling of these cells (Fig. 8,B). These results indicate that Bad does not play a major role in MEK/ERK inhibition-induced apoptosis of mitogen-stimulated B cells. In addition, we found that Bim^−/−Bid^−/− B cells were similarly resistant to MEK inhibition as Bim^−/− B cells and that Bid^−/− B cells were normally sensitive to this treatment (Figs. 8, C and D). Collectively, these results indicate that BH3-only proteins other than Bad and Bid are likely to cooperate with Bim in MEK/ERK inhibition-induced killing of mitogen-stimulated B cells.

Connections between the pathways that control cell survival and cell cycling are critical to ensure normal development and function of a multicellular organism (1). In this study, we have investigated the roles of MEK/ERK signaling and the proapoptotic BH3-only proteins Bim, Bad, and Bid in the regulation of cell survival and proliferation in mitogen-stimulated T and B lymphocytes.

Many apoptotic stimuli have been shown to cause an increase in Bim levels in a broad range of cell types, such as nerve growth factor withdrawal from sympathetic neurons (34), serum deprivation from fibroblasts (35), cAMP-induced apoptosis in S49 T lymphoma cells (36), or AgR stimulation in T (11, 37) or B (12, 31, 32) lymphoid cells. We found that treatment with the calcium ionophore ionomycin, which kills thymocytes via a Bim-dependent process (9), increases Bim_EL levels by ∼3- to 4-fold. Increased expression of Bim_EL was also observed in T cell blasts after withdrawal of IL-2, and it has previously been reported that de-phosphorylation of Bim_EL correlates with apoptosis induction in this setting (38) and that Bim is critical for this pathway to cell death (9).

Because Bim plays a critical role in controlling survival of T (39, 40) and B (41) lymphocytes during shutdown of immune responses, we examined the regulation of Bim expression after mitogenic stimulation. We found that Bim is rapidly phosphorylated in mitogen-stimulated T and B lymphocytes. Experiments with pharmacological inhibitors indicated that this phosphorylation was mediated by a MEK/ERK-dependent process but did not require JNK, another kinase implicated in the phosphorylation and regulation of Bim (18). After phosphorylation, Bim levels declined in mitogen-stimulated B and T cells, and treatment with proteasome inhibitors prevented this decline. This is consistent with previous observations that phosphorylation by ERK targets Bim for ubiquitination and proteasomal degradation (17, 42). ERK-mediated phosphorylation has also been shown to inhibit the pro-apoptotic activity of Bim by reducing its binding to the anti-apoptotic Bcl-2 family members Mcl-1 and Bcl-x_L (16). The relative contributions of these two processes to the prosurvival effects of MEK/ERK signaling were examined by the generation of gene-targeted mice that carry mutant alleles expressing phosphorylation-defective Bim proteins. Mutation of phosphorylation site Thr¹¹² causes decreased binding of Bim to Bcl-2 and thus increased survival, whereas in contrast mutation of Ser^55/65/73 caused increased apoptosis, due to reduced proteasomal degradation (29).

The ERK signaling pathway is critical for ensuring cell survival and proliferation during mitogen-induced T cell proliferation (28). Our experiments using pharmacological inhibitors (UO126) and lymphocytes from wt or Bim^−/− mice indicate that phosphorylation of Bim and its subsequent reduction in level are critical to allow the survival of mitogen-stimulated T and B cells. Presumably, the stresses associated with transition from the quiescent (G₀) state into the cell cycle augment Bim synthesis, and this must be counteracted by TCR or BCR stimulation-induced ERK activation. MEK/ERK-mediated phosphorylation of Bim_EL has recently been observed in fibroblast growth factor-stimulated fibroblasts and macrophages undergoing mitosis (43, 44), indicating that Bim phosphorylation may also play a critical role in inhibiting apoptosis at that stage of the cell cycle. Although loss of Bim greatly reduced MEK/ERK inhibition-induced apoptosis of mitogen-stimulated T and B cells, it had no effect on the reduction in cell cycling. This is consistent with the view that MEK/ERK signaling promotes survival and cell proliferation through distinct effector pathways.

Loss of Bim protected mitogen-activated T cells better against treatment with the MEK inhibitor than activated B cells. The observation that Bcl-2 overexpression could completely block MEK inhibitor-induced killing of mitogen-stimulated B cells indicates that BH3-only proteins in addition to Bim must be inactivated to allow their optimal survival. Our experiments indicate that Bad, which like Bim can also be regulated by phosphorylation (14, 45), and Bid are unlikely to perform this function. With respect to other BH3-only proteins, Puma is an attractive candidate for several reasons: 1) it is expressed in both T and B cells; 2) its loss has been shown to increase resistance of both T and B lymphocytes to a range of apoptotic stimuli (25, 46, 47); and 3) Bim and Puma have been shown to have overlapping function in apoptosis initiation (48).

We are grateful to Drs. S. Cory, J. M. Adams, N. Danial, and A. Ranger for gifts of gene-targeted mice and reagents; Dr. M. Cragg for advice; K. Vella, G. Siciliano, A. Naughton, K. Pioch, N. Iannarella, and J. Allen for expert animal care; K. Scalzo for technical assistance; B. Helbert and M. Robati for genotyping; Dr. F. Battye, V. Milovac, C. Tarlinton, C. Young, and J. Garbe for cell sorting.

The authors have no financial conflict of interest.