Regulation of osteoclast apoptosis by ubiquitylation of proapoptotic BH3-only Bcl-2 family member Bim

Abstract

Osteoclasts (OCs) undergo rapid apoptosis without trophic factors, such as macrophage colony-stimulating factor (M-CSF). Their apoptosis was associated with a rapid and sustained increase in the pro-apoptotic BH3-only Bcl-2 family member Bim. This was caused by the reduced ubiquitylation and proteasomal degradation of Bim that is mediated by c-Cbl. Although the number of OCs was increased in the skeletal tissues of bim-/- mice, the mice exhibited mild osteosclerosis due to reduced bone resorption. OCs differentiated from bone marrow cells of bim-/- animals showed a marked prolongation of survival in the absence of M-CSF, compared with bim+/+ OCs, but the bone-resorbing activity of bim-/- OCs was significantly reduced. Overexpression of a degradation-resistant lysine-free Bim mutant in bim-/- cells abrogated the anti-apoptotic effect of M-CSF, while wild-type Bim did not. These results demonstrate that ubiquitylation-dependent regulation of Bim levels is critical for controlling apoptosis and activation of OCs.

Fig. 1. Regulation of Bim expression in OCs. (A) Cytokine withdrawal caused rapid upregulation of Bim levels in OCs. OCs purified from co-cultures by removing osteoblastic cells by collagenase and dispase treatment were then maintained in the presence of M-CSF (10 ng/ml) for an additional 12 h. The expression levels of Bim and other apoptosis-regulatory proteins in OCs after M-CSF removal were analyzed by western blotting using specific antibodies. Bim levels increased within 3 h, and the upregulation was sustained at least for 12 h. This upregulation of Bim level was strongly suppressed by M-CSF, and to a lesser extent by sRANKL treatment for 12 h. (B) Intracellular signaling pathways leading to Bim downregulation. Upper panel: introduction of MEK^CA strongly suppressed the upregulation of Bim after M-CSF removal. Overexpression of myr-Akt had less effect on Bim expression in OCs. Lower panel: treating the cells with a specific inhibitor of MEK/ERK pathways, PD98059, completely abolished the suppressive effect of M-CSF on Bim expression. (C) Transcriptional regulation of bim in OCs. No significant change in the mRNA level of three isoforms of bim, i.e. bim_EL, _L and _S (upper and lower left), or the bim_EL specific mRNA level (lower right) was detected in OCs in the presence or absence of M-CSF as determined by RT–PCR (upper) or real-time PCR (lower). The y-axis indicates the relative mRNA levels. NS = not significantly different. (D) Effect of adenovirus vector-mediated overexpression of Bim_L and/or Bcl-xL on OC survival. Upper panel: TRAP staining. Lower panel: percentage of OCs surviving. Overexpression of Bim_L promoted apoptosis of OCs (AxBim_L). Not only did Bcl-xL overexpression suppress apoptosis of OCs (AxEGFP + AxBcl-xL), but co-expression of Bcl-xL together with Bim_L completely abrogated the pro-apoptotic effect of Bim_L (AxBim_L + AxBcl-xL).

Fig. 1. Regulation of Bim expression in OCs. (A) Cytokine withdrawal caused rapid upregulation of Bim levels in OCs. OCs purified from co-cultures by removing osteoblastic cells by collagenase and dispase treatment were then maintained in the presence of M-CSF (10 ng/ml) for an additional 12 h. The expression levels of Bim and other apoptosis-regulatory proteins in OCs after M-CSF removal were analyzed by western blotting using specific antibodies. Bim levels increased within 3 h, and the upregulation was sustained at least for 12 h. This upregulation of Bim level was strongly suppressed by M-CSF, and to a lesser extent by sRANKL treatment for 12 h. (B) Intracellular signaling pathways leading to Bim downregulation. Upper panel: introduction of MEK^CA strongly suppressed the upregulation of Bim after M-CSF removal. Overexpression of myr-Akt had less effect on Bim expression in OCs. Lower panel: treating the cells with a specific inhibitor of MEK/ERK pathways, PD98059, completely abolished the suppressive effect of M-CSF on Bim expression. (C) Transcriptional regulation of bim in OCs. No significant change in the mRNA level of three isoforms of bim, i.e. bim_EL, _L and _S (upper and lower left), or the bim_EL specific mRNA level (lower right) was detected in OCs in the presence or absence of M-CSF as determined by RT–PCR (upper) or real-time PCR (lower). The y-axis indicates the relative mRNA levels. NS = not significantly different. (D) Effect of adenovirus vector-mediated overexpression of Bim_L and/or Bcl-xL on OC survival. Upper panel: TRAP staining. Lower panel: percentage of OCs surviving. Overexpression of Bim_L promoted apoptosis of OCs (AxBim_L). Not only did Bcl-xL overexpression suppress apoptosis of OCs (AxEGFP + AxBcl-xL), but co-expression of Bcl-xL together with Bim_L completely abrogated the pro-apoptotic effect of Bim_L (AxBim_L + AxBcl-xL).

Fig. 1. Regulation of Bim expression in OCs. (A) Cytokine withdrawal caused rapid upregulation of Bim levels in OCs. OCs purified from co-cultures by removing osteoblastic cells by collagenase and dispase treatment were then maintained in the presence of M-CSF (10 ng/ml) for an additional 12 h. The expression levels of Bim and other apoptosis-regulatory proteins in OCs after M-CSF removal were analyzed by western blotting using specific antibodies. Bim levels increased within 3 h, and the upregulation was sustained at least for 12 h. This upregulation of Bim level was strongly suppressed by M-CSF, and to a lesser extent by sRANKL treatment for 12 h. (B) Intracellular signaling pathways leading to Bim downregulation. Upper panel: introduction of MEK^CA strongly suppressed the upregulation of Bim after M-CSF removal. Overexpression of myr-Akt had less effect on Bim expression in OCs. Lower panel: treating the cells with a specific inhibitor of MEK/ERK pathways, PD98059, completely abolished the suppressive effect of M-CSF on Bim expression. (C) Transcriptional regulation of bim in OCs. No significant change in the mRNA level of three isoforms of bim, i.e. bim_EL, _L and _S (upper and lower left), or the bim_EL specific mRNA level (lower right) was detected in OCs in the presence or absence of M-CSF as determined by RT–PCR (upper) or real-time PCR (lower). The y-axis indicates the relative mRNA levels. NS = not significantly different. (D) Effect of adenovirus vector-mediated overexpression of Bim_L and/or Bcl-xL on OC survival. Upper panel: TRAP staining. Lower panel: percentage of OCs surviving. Overexpression of Bim_L promoted apoptosis of OCs (AxBim_L). Not only did Bcl-xL overexpression suppress apoptosis of OCs (AxEGFP + AxBcl-xL), but co-expression of Bcl-xL together with Bim_L completely abrogated the pro-apoptotic effect of Bim_L (AxBim_L + AxBcl-xL).

Fig. 1. Regulation of Bim expression in OCs. (A) Cytokine withdrawal caused rapid upregulation of Bim levels in OCs. OCs purified from co-cultures by removing osteoblastic cells by collagenase and dispase treatment were then maintained in the presence of M-CSF (10 ng/ml) for an additional 12 h. The expression levels of Bim and other apoptosis-regulatory proteins in OCs after M-CSF removal were analyzed by western blotting using specific antibodies. Bim levels increased within 3 h, and the upregulation was sustained at least for 12 h. This upregulation of Bim level was strongly suppressed by M-CSF, and to a lesser extent by sRANKL treatment for 12 h. (B) Intracellular signaling pathways leading to Bim downregulation. Upper panel: introduction of MEK^CA strongly suppressed the upregulation of Bim after M-CSF removal. Overexpression of myr-Akt had less effect on Bim expression in OCs. Lower panel: treating the cells with a specific inhibitor of MEK/ERK pathways, PD98059, completely abolished the suppressive effect of M-CSF on Bim expression. (C) Transcriptional regulation of bim in OCs. No significant change in the mRNA level of three isoforms of bim, i.e. bim_EL, _L and _S (upper and lower left), or the bim_EL specific mRNA level (lower right) was detected in OCs in the presence or absence of M-CSF as determined by RT–PCR (upper) or real-time PCR (lower). The y-axis indicates the relative mRNA levels. NS = not significantly different. (D) Effect of adenovirus vector-mediated overexpression of Bim_L and/or Bcl-xL on OC survival. Upper panel: TRAP staining. Lower panel: percentage of OCs surviving. Overexpression of Bim_L promoted apoptosis of OCs (AxBim_L). Not only did Bcl-xL overexpression suppress apoptosis of OCs (AxEGFP + AxBcl-xL), but co-expression of Bcl-xL together with Bim_L completely abrogated the pro-apoptotic effect of Bim_L (AxBim_L + AxBcl-xL).

Fig. 2. In situ hybridization of the section of the metatarsal bones from a 5-week-old male wild-type mouse using digoxigenin-labeled mouse bim_L (B), procollagen type IA (C) and type II A (D) riboprobes, and TRAP enzymatic staining (A). The labeling was detected by anti-digoxigenin antibody and Alexa 488-labeled anti-rabbit IgG antibody. Note the co-localization of bim transcripts with TRAP staining (OCs) (A and B). No positive bim_L staining was co-localized with procollagen type IA staining (osteoblasts) or type IIB staining (chondrocytes). X-gal staining of the tibia from 5-week-old transgenic mice in which lacZ gene was introduced into the bim locus by homologous recombination also showed the clear positive staining in OCs but not in chondrocytes or osteoblasts (E).

Fig. 3. Bim deficiency causes mild osteosclerosis due to impaired OC activation. (A) Radiological and histological analysis of tibia from 5-week-old bim+/+ and bim–/– male mice (H&E and TRAP staining). bim–/– tibia showed expanded secondary spongiosa, exhibiting mild osteosclerosis. (B) bim–/– OCs showed smaller and shrunken morphological features compared with bim+/+ OCs in vivo as shown by TRAP staining. (C) Impaired actin ring formation in bim–/– OCs. More than 90% of bim+/+ OCs formed actin rings (left), whereas only 30% of bim–/– cells were able to do this (right). (D) Histomorphometric analysis: parameters are measured in the proximal tibia of bim+/+ and bim–/– mice. Data are expressed as means ± SD from five mice of each genotype. BV/TV, trabecular bone volume expressed as a percentage of tibia tissue volume; Tb.Th, trabecular bone thickness; Tb.N, trabecular bone number per mm; Tb.Sp, average space between neighboring trabecular bones; O.Th, osteoid thickness; OS/BS, percentage of bone surface covered by osteoid; Ob.S/BS, percentage of bone surface covered by cuboidal osteoblast; ES/BS, percentage of eroded surface; Oc.N/B.Pm, number of mature OCs per 10 mm of bone perimeter; Oc.S/BS, percentage of bone surface covered by mature OCs. bim–/– bones showed reduction of both bone formation markers (O.Th, OS/BS and Ob.S/BS) and bone resorption markers (ES/BS and Oc.S/BS), and increased Oc.N/B/Pm. *&** = significantly different. *P < 0.005, **P < 0.01. (E) Calcein double labeling of tibia trabecular bone from 13-week-old bim+/+ (left) and bim–/– mice (right). Labeling is visualized by fluorescent microscopy. (F) bim–/– OCs have a longer life span in vivo. Five-week-old bim+/+ and bim–/– mice (n = 4) were fed with water containing 1 mg/ml BrdU for 1 week (labeling period). Mice were then sacrificed either on the next day (group A) or after 6 weeks (group B) of the labeling period, and their tibias were examined by anti-BrdU immunohistochemistry. More than 100 OCs were examined by BrdU immunostaining in the serial sections of the tibia, and the number of BrdU-positive OCs was counted. Fifty percent of bim+/+ OCs and 48% of bim–/– OCs in group A were positively stained by BrdU. However, the proportion of BrdU-positive OCs was markedly reduced to <5% in group B bim+/+ mice, due to the apoptotic cell death, while >30% of group B bim–/– OCs still exhibited BrdU labeling.

Fig. 3. Bim deficiency causes mild osteosclerosis due to impaired OC activation. (A) Radiological and histological analysis of tibia from 5-week-old bim+/+ and bim–/– male mice (H&E and TRAP staining). bim–/– tibia showed expanded secondary spongiosa, exhibiting mild osteosclerosis. (B) bim–/– OCs showed smaller and shrunken morphological features compared with bim+/+ OCs in vivo as shown by TRAP staining. (C) Impaired actin ring formation in bim–/– OCs. More than 90% of bim+/+ OCs formed actin rings (left), whereas only 30% of bim–/– cells were able to do this (right). (D) Histomorphometric analysis: parameters are measured in the proximal tibia of bim+/+ and bim–/– mice. Data are expressed as means ± SD from five mice of each genotype. BV/TV, trabecular bone volume expressed as a percentage of tibia tissue volume; Tb.Th, trabecular bone thickness; Tb.N, trabecular bone number per mm; Tb.Sp, average space between neighboring trabecular bones; O.Th, osteoid thickness; OS/BS, percentage of bone surface covered by osteoid; Ob.S/BS, percentage of bone surface covered by cuboidal osteoblast; ES/BS, percentage of eroded surface; Oc.N/B.Pm, number of mature OCs per 10 mm of bone perimeter; Oc.S/BS, percentage of bone surface covered by mature OCs. bim–/– bones showed reduction of both bone formation markers (O.Th, OS/BS and Ob.S/BS) and bone resorption markers (ES/BS and Oc.S/BS), and increased Oc.N/B/Pm. *&** = significantly different. *P < 0.005, **P < 0.01. (E) Calcein double labeling of tibia trabecular bone from 13-week-old bim+/+ (left) and bim–/– mice (right). Labeling is visualized by fluorescent microscopy. (F) bim–/– OCs have a longer life span in vivo. Five-week-old bim+/+ and bim–/– mice (n = 4) were fed with water containing 1 mg/ml BrdU for 1 week (labeling period). Mice were then sacrificed either on the next day (group A) or after 6 weeks (group B) of the labeling period, and their tibias were examined by anti-BrdU immunohistochemistry. More than 100 OCs were examined by BrdU immunostaining in the serial sections of the tibia, and the number of BrdU-positive OCs was counted. Fifty percent of bim+/+ OCs and 48% of bim–/– OCs in group A were positively stained by BrdU. However, the proportion of BrdU-positive OCs was markedly reduced to <5% in group B bim+/+ mice, due to the apoptotic cell death, while >30% of group B bim–/– OCs still exhibited BrdU labeling.

Fig. 3. Bim deficiency causes mild osteosclerosis due to impaired OC activation. (A) Radiological and histological analysis of tibia from 5-week-old bim+/+ and bim–/– male mice (H&E and TRAP staining). bim–/– tibia showed expanded secondary spongiosa, exhibiting mild osteosclerosis. (B) bim–/– OCs showed smaller and shrunken morphological features compared with bim+/+ OCs in vivo as shown by TRAP staining. (C) Impaired actin ring formation in bim–/– OCs. More than 90% of bim+/+ OCs formed actin rings (left), whereas only 30% of bim–/– cells were able to do this (right). (D) Histomorphometric analysis: parameters are measured in the proximal tibia of bim+/+ and bim–/– mice. Data are expressed as means ± SD from five mice of each genotype. BV/TV, trabecular bone volume expressed as a percentage of tibia tissue volume; Tb.Th, trabecular bone thickness; Tb.N, trabecular bone number per mm; Tb.Sp, average space between neighboring trabecular bones; O.Th, osteoid thickness; OS/BS, percentage of bone surface covered by osteoid; Ob.S/BS, percentage of bone surface covered by cuboidal osteoblast; ES/BS, percentage of eroded surface; Oc.N/B.Pm, number of mature OCs per 10 mm of bone perimeter; Oc.S/BS, percentage of bone surface covered by mature OCs. bim–/– bones showed reduction of both bone formation markers (O.Th, OS/BS and Ob.S/BS) and bone resorption markers (ES/BS and Oc.S/BS), and increased Oc.N/B/Pm. *&** = significantly different. *P < 0.005, **P < 0.01. (E) Calcein double labeling of tibia trabecular bone from 13-week-old bim+/+ (left) and bim–/– mice (right). Labeling is visualized by fluorescent microscopy. (F) bim–/– OCs have a longer life span in vivo. Five-week-old bim+/+ and bim–/– mice (n = 4) were fed with water containing 1 mg/ml BrdU for 1 week (labeling period). Mice were then sacrificed either on the next day (group A) or after 6 weeks (group B) of the labeling period, and their tibias were examined by anti-BrdU immunohistochemistry. More than 100 OCs were examined by BrdU immunostaining in the serial sections of the tibia, and the number of BrdU-positive OCs was counted. Fifty percent of bim+/+ OCs and 48% of bim–/– OCs in group A were positively stained by BrdU. However, the proportion of BrdU-positive OCs was markedly reduced to <5% in group B bim+/+ mice, due to the apoptotic cell death, while >30% of group B bim–/– OCs still exhibited BrdU labeling.

Fig. 4. Effects of Bim on survival and activity of OCs. (A) Survival of bim+/+ and bim–/– OCs. More than 90% of bim–/– OCs survived 48 h after M-CSF removal, while all the bim+/+ OCs had died. (B) Adenovirus vector-mediated Bim_L expression-promoted apoptosis in bim–/– OCs as well as in bim+/+ OCs after 12 h of M-CSF deprivation. (C) Pit-forming activity. The resorption pit area formed by bim–/– OCs was significantly less than that formed by bim+/+ OCs. Overexpression of Bim_L promoted bone-resorbing activity in both bim+/+ and bim–/– OCs.

Fig. 5. Ubiquitylation-dependent degradation of Bim in OCs. (A) Western blotting. Upper panel: Bim protein level increased after 12 h of osteoblast removal, which was suppressed in the presence of M-CSF. MG132 treatment strongly upregulated Bim expression even in the presence of M-CSF. Lower panel: no obvious upregulation or downregulation of Bid, Bax, Bak, Bcl-xL and XIAP level was induced by MG132 or lactacystin. (B) Ubiquitylation of Bim in OCs. Proteins immunoprecipitated with anti-Bim antibody were immunoblotted with an anti-Bim antibody (left) or an anti-ubiquitin antibody (right). Ubiquitylated Bim was detected as upper-shifted bands in both anti-Bim and anti-ubiquitin blotting (Bim-Ub). Ubiquitylated Bim was detected when cells were treated with MG132, and marked enhancement of its ubiquitylation was induced by M-CSF treatment. (C) Association of c-Cbl with Bim in OCs. Upper panel: cell lysates of M-CSF-treated OCs were incubated with bacterially expressed GST–Bim_EL fusion protein, and association with c-Cbl was determined by western blotting. c-Cbl was associated with GST–Bim_EL but not with GST. Lower panel: Bim was immunoprecipitated from OC cell lysates, and its association with c-Cbl was examined by western blotting. Co-immunoprecipitation of c-Cbl with Bim was observed in MG132- treated OCs, which was enhanced by M-CSF treatment. (D) Double immunofluorescence of (a) Bim (green) and (b) c-Cbl (red) in authentic OCs isolated from normal mice. Co-localization of these two molecules (yellow in c and d) was observed in the presence of MG132 and M-CSF. (d) An enlargement of the rectangular area in (c). (E) Involvement of c-Cbl in the ubiquitylation of Bim in OCs. Upper panel: adenovirus vector-mediated overexpression of c-Cbl (Axc-Cbl) decreased the protein level of Bim in OCs in the absence of M-CSF (compare lanes 1 and 5). MG132 treatment increased Bim expression in c-Cbl-overexpressing cells with or without M-CSF (lanes 7 and 8). On the other hand, adenovirus vector-induced overexpression of v-Cbl (Axv-Cbl) increased Bim expression even in the absence of MG132, which was not affected by M-CSF treatment (lanes 9–12). Lower panel: Bim was immunoprecipitated and immunoblotted with anti- ubiquitin antibody. c-Cbl-overexpressing OCs showed Bim ubiquitylation in the absence of M-CSF (lane 2) to a similar extent as in EGFP virus-infected OCs treated with M-CSF (lane 1). v-Cbl overexpression suppressed Bim ubiquitylation even in the presence of MG132 and M-CSF (lane 3). (F) M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs. OCs generated from bone marrow cells of c-Cbl–/– mice (c-Cbl–/–) or their normal littermates (c-Cbl+/+) were treated or not with M-CSF. Western blotting with anti-Bim antibody demonstrates that although no difference in Bim level was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF (lanes 2 and 4, respectively), M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs (lane 3). (G) Immunofluorescence of primary OCs with anti-Bim antibody. Primary OCs isolated from c-Cbl+/+ and c-Cbl–/– mice were cultured in the presence or absence of M-CSF (10 ng/ml) for 12 h, and immunostained with anti-Bim antibody. Although no apparent difference in fluorescence intensity was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF, M-CSF-induced downregulation was scarcely detected in c-Cbl–/– OCs.

Fig. 5. Ubiquitylation-dependent degradation of Bim in OCs. (A) Western blotting. Upper panel: Bim protein level increased after 12 h of osteoblast removal, which was suppressed in the presence of M-CSF. MG132 treatment strongly upregulated Bim expression even in the presence of M-CSF. Lower panel: no obvious upregulation or downregulation of Bid, Bax, Bak, Bcl-xL and XIAP level was induced by MG132 or lactacystin. (B) Ubiquitylation of Bim in OCs. Proteins immunoprecipitated with anti-Bim antibody were immunoblotted with an anti-Bim antibody (left) or an anti-ubiquitin antibody (right). Ubiquitylated Bim was detected as upper-shifted bands in both anti-Bim and anti-ubiquitin blotting (Bim-Ub). Ubiquitylated Bim was detected when cells were treated with MG132, and marked enhancement of its ubiquitylation was induced by M-CSF treatment. (C) Association of c-Cbl with Bim in OCs. Upper panel: cell lysates of M-CSF-treated OCs were incubated with bacterially expressed GST–Bim_EL fusion protein, and association with c-Cbl was determined by western blotting. c-Cbl was associated with GST–Bim_EL but not with GST. Lower panel: Bim was immunoprecipitated from OC cell lysates, and its association with c-Cbl was examined by western blotting. Co-immunoprecipitation of c-Cbl with Bim was observed in MG132- treated OCs, which was enhanced by M-CSF treatment. (D) Double immunofluorescence of (a) Bim (green) and (b) c-Cbl (red) in authentic OCs isolated from normal mice. Co-localization of these two molecules (yellow in c and d) was observed in the presence of MG132 and M-CSF. (d) An enlargement of the rectangular area in (c). (E) Involvement of c-Cbl in the ubiquitylation of Bim in OCs. Upper panel: adenovirus vector-mediated overexpression of c-Cbl (Axc-Cbl) decreased the protein level of Bim in OCs in the absence of M-CSF (compare lanes 1 and 5). MG132 treatment increased Bim expression in c-Cbl-overexpressing cells with or without M-CSF (lanes 7 and 8). On the other hand, adenovirus vector-induced overexpression of v-Cbl (Axv-Cbl) increased Bim expression even in the absence of MG132, which was not affected by M-CSF treatment (lanes 9–12). Lower panel: Bim was immunoprecipitated and immunoblotted with anti- ubiquitin antibody. c-Cbl-overexpressing OCs showed Bim ubiquitylation in the absence of M-CSF (lane 2) to a similar extent as in EGFP virus-infected OCs treated with M-CSF (lane 1). v-Cbl overexpression suppressed Bim ubiquitylation even in the presence of MG132 and M-CSF (lane 3). (F) M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs. OCs generated from bone marrow cells of c-Cbl–/– mice (c-Cbl–/–) or their normal littermates (c-Cbl+/+) were treated or not with M-CSF. Western blotting with anti-Bim antibody demonstrates that although no difference in Bim level was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF (lanes 2 and 4, respectively), M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs (lane 3). (G) Immunofluorescence of primary OCs with anti-Bim antibody. Primary OCs isolated from c-Cbl+/+ and c-Cbl–/– mice were cultured in the presence or absence of M-CSF (10 ng/ml) for 12 h, and immunostained with anti-Bim antibody. Although no apparent difference in fluorescence intensity was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF, M-CSF-induced downregulation was scarcely detected in c-Cbl–/– OCs.

Fig. 5. Ubiquitylation-dependent degradation of Bim in OCs. (A) Western blotting. Upper panel: Bim protein level increased after 12 h of osteoblast removal, which was suppressed in the presence of M-CSF. MG132 treatment strongly upregulated Bim expression even in the presence of M-CSF. Lower panel: no obvious upregulation or downregulation of Bid, Bax, Bak, Bcl-xL and XIAP level was induced by MG132 or lactacystin. (B) Ubiquitylation of Bim in OCs. Proteins immunoprecipitated with anti-Bim antibody were immunoblotted with an anti-Bim antibody (left) or an anti-ubiquitin antibody (right). Ubiquitylated Bim was detected as upper-shifted bands in both anti-Bim and anti-ubiquitin blotting (Bim-Ub). Ubiquitylated Bim was detected when cells were treated with MG132, and marked enhancement of its ubiquitylation was induced by M-CSF treatment. (C) Association of c-Cbl with Bim in OCs. Upper panel: cell lysates of M-CSF-treated OCs were incubated with bacterially expressed GST–Bim_EL fusion protein, and association with c-Cbl was determined by western blotting. c-Cbl was associated with GST–Bim_EL but not with GST. Lower panel: Bim was immunoprecipitated from OC cell lysates, and its association with c-Cbl was examined by western blotting. Co-immunoprecipitation of c-Cbl with Bim was observed in MG132- treated OCs, which was enhanced by M-CSF treatment. (D) Double immunofluorescence of (a) Bim (green) and (b) c-Cbl (red) in authentic OCs isolated from normal mice. Co-localization of these two molecules (yellow in c and d) was observed in the presence of MG132 and M-CSF. (d) An enlargement of the rectangular area in (c). (E) Involvement of c-Cbl in the ubiquitylation of Bim in OCs. Upper panel: adenovirus vector-mediated overexpression of c-Cbl (Axc-Cbl) decreased the protein level of Bim in OCs in the absence of M-CSF (compare lanes 1 and 5). MG132 treatment increased Bim expression in c-Cbl-overexpressing cells with or without M-CSF (lanes 7 and 8). On the other hand, adenovirus vector-induced overexpression of v-Cbl (Axv-Cbl) increased Bim expression even in the absence of MG132, which was not affected by M-CSF treatment (lanes 9–12). Lower panel: Bim was immunoprecipitated and immunoblotted with anti- ubiquitin antibody. c-Cbl-overexpressing OCs showed Bim ubiquitylation in the absence of M-CSF (lane 2) to a similar extent as in EGFP virus-infected OCs treated with M-CSF (lane 1). v-Cbl overexpression suppressed Bim ubiquitylation even in the presence of MG132 and M-CSF (lane 3). (F) M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs. OCs generated from bone marrow cells of c-Cbl–/– mice (c-Cbl–/–) or their normal littermates (c-Cbl+/+) were treated or not with M-CSF. Western blotting with anti-Bim antibody demonstrates that although no difference in Bim level was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF (lanes 2 and 4, respectively), M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs (lane 3). (G) Immunofluorescence of primary OCs with anti-Bim antibody. Primary OCs isolated from c-Cbl+/+ and c-Cbl–/– mice were cultured in the presence or absence of M-CSF (10 ng/ml) for 12 h, and immunostained with anti-Bim antibody. Although no apparent difference in fluorescence intensity was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF, M-CSF-induced downregulation was scarcely detected in c-Cbl–/– OCs.

Fig. 5. Ubiquitylation-dependent degradation of Bim in OCs. (A) Western blotting. Upper panel: Bim protein level increased after 12 h of osteoblast removal, which was suppressed in the presence of M-CSF. MG132 treatment strongly upregulated Bim expression even in the presence of M-CSF. Lower panel: no obvious upregulation or downregulation of Bid, Bax, Bak, Bcl-xL and XIAP level was induced by MG132 or lactacystin. (B) Ubiquitylation of Bim in OCs. Proteins immunoprecipitated with anti-Bim antibody were immunoblotted with an anti-Bim antibody (left) or an anti-ubiquitin antibody (right). Ubiquitylated Bim was detected as upper-shifted bands in both anti-Bim and anti-ubiquitin blotting (Bim-Ub). Ubiquitylated Bim was detected when cells were treated with MG132, and marked enhancement of its ubiquitylation was induced by M-CSF treatment. (C) Association of c-Cbl with Bim in OCs. Upper panel: cell lysates of M-CSF-treated OCs were incubated with bacterially expressed GST–Bim_EL fusion protein, and association with c-Cbl was determined by western blotting. c-Cbl was associated with GST–Bim_EL but not with GST. Lower panel: Bim was immunoprecipitated from OC cell lysates, and its association with c-Cbl was examined by western blotting. Co-immunoprecipitation of c-Cbl with Bim was observed in MG132- treated OCs, which was enhanced by M-CSF treatment. (D) Double immunofluorescence of (a) Bim (green) and (b) c-Cbl (red) in authentic OCs isolated from normal mice. Co-localization of these two molecules (yellow in c and d) was observed in the presence of MG132 and M-CSF. (d) An enlargement of the rectangular area in (c). (E) Involvement of c-Cbl in the ubiquitylation of Bim in OCs. Upper panel: adenovirus vector-mediated overexpression of c-Cbl (Axc-Cbl) decreased the protein level of Bim in OCs in the absence of M-CSF (compare lanes 1 and 5). MG132 treatment increased Bim expression in c-Cbl-overexpressing cells with or without M-CSF (lanes 7 and 8). On the other hand, adenovirus vector-induced overexpression of v-Cbl (Axv-Cbl) increased Bim expression even in the absence of MG132, which was not affected by M-CSF treatment (lanes 9–12). Lower panel: Bim was immunoprecipitated and immunoblotted with anti- ubiquitin antibody. c-Cbl-overexpressing OCs showed Bim ubiquitylation in the absence of M-CSF (lane 2) to a similar extent as in EGFP virus-infected OCs treated with M-CSF (lane 1). v-Cbl overexpression suppressed Bim ubiquitylation even in the presence of MG132 and M-CSF (lane 3). (F) M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs. OCs generated from bone marrow cells of c-Cbl–/– mice (c-Cbl–/–) or their normal littermates (c-Cbl+/+) were treated or not with M-CSF. Western blotting with anti-Bim antibody demonstrates that although no difference in Bim level was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF (lanes 2 and 4, respectively), M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs (lane 3). (G) Immunofluorescence of primary OCs with anti-Bim antibody. Primary OCs isolated from c-Cbl+/+ and c-Cbl–/– mice were cultured in the presence or absence of M-CSF (10 ng/ml) for 12 h, and immunostained with anti-Bim antibody. Although no apparent difference in fluorescence intensity was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF, M-CSF-induced downregulation was scarcely detected in c-Cbl–/– OCs.

Fig. 5. Ubiquitylation-dependent degradation of Bim in OCs. (A) Western blotting. Upper panel: Bim protein level increased after 12 h of osteoblast removal, which was suppressed in the presence of M-CSF. MG132 treatment strongly upregulated Bim expression even in the presence of M-CSF. Lower panel: no obvious upregulation or downregulation of Bid, Bax, Bak, Bcl-xL and XIAP level was induced by MG132 or lactacystin. (B) Ubiquitylation of Bim in OCs. Proteins immunoprecipitated with anti-Bim antibody were immunoblotted with an anti-Bim antibody (left) or an anti-ubiquitin antibody (right). Ubiquitylated Bim was detected as upper-shifted bands in both anti-Bim and anti-ubiquitin blotting (Bim-Ub). Ubiquitylated Bim was detected when cells were treated with MG132, and marked enhancement of its ubiquitylation was induced by M-CSF treatment. (C) Association of c-Cbl with Bim in OCs. Upper panel: cell lysates of M-CSF-treated OCs were incubated with bacterially expressed GST–Bim_EL fusion protein, and association with c-Cbl was determined by western blotting. c-Cbl was associated with GST–Bim_EL but not with GST. Lower panel: Bim was immunoprecipitated from OC cell lysates, and its association with c-Cbl was examined by western blotting. Co-immunoprecipitation of c-Cbl with Bim was observed in MG132- treated OCs, which was enhanced by M-CSF treatment. (D) Double immunofluorescence of (a) Bim (green) and (b) c-Cbl (red) in authentic OCs isolated from normal mice. Co-localization of these two molecules (yellow in c and d) was observed in the presence of MG132 and M-CSF. (d) An enlargement of the rectangular area in (c). (E) Involvement of c-Cbl in the ubiquitylation of Bim in OCs. Upper panel: adenovirus vector-mediated overexpression of c-Cbl (Axc-Cbl) decreased the protein level of Bim in OCs in the absence of M-CSF (compare lanes 1 and 5). MG132 treatment increased Bim expression in c-Cbl-overexpressing cells with or without M-CSF (lanes 7 and 8). On the other hand, adenovirus vector-induced overexpression of v-Cbl (Axv-Cbl) increased Bim expression even in the absence of MG132, which was not affected by M-CSF treatment (lanes 9–12). Lower panel: Bim was immunoprecipitated and immunoblotted with anti- ubiquitin antibody. c-Cbl-overexpressing OCs showed Bim ubiquitylation in the absence of M-CSF (lane 2) to a similar extent as in EGFP virus-infected OCs treated with M-CSF (lane 1). v-Cbl overexpression suppressed Bim ubiquitylation even in the presence of MG132 and M-CSF (lane 3). (F) M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs. OCs generated from bone marrow cells of c-Cbl–/– mice (c-Cbl–/–) or their normal littermates (c-Cbl+/+) were treated or not with M-CSF. Western blotting with anti-Bim antibody demonstrates that although no difference in Bim level was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF (lanes 2 and 4, respectively), M-CSF-induced downregulation of Bim was suppressed in c-Cbl–/– OCs (lane 3). (G) Immunofluorescence of primary OCs with anti-Bim antibody. Primary OCs isolated from c-Cbl+/+ and c-Cbl–/– mice were cultured in the presence or absence of M-CSF (10 ng/ml) for 12 h, and immunostained with anti-Bim antibody. Although no apparent difference in fluorescence intensity was observed between c-Cbl+/+ and c-Cbl–/– OCs in the absence of M-CSF, M-CSF-induced downregulation was scarcely detected in c-Cbl–/– OCs.

Fig. 6. Effect of mutations in Bim that prevent ubiquitylation. (A) Phase contrast (a, c and e) and immunofluorescence microscopy: bim–/– bone marrow cells cultured in the presence of M-CSF (100 ng/ml) and zVAD-FMK (100 µM) were infected with either pMx-IRES-EGFP, pMxBim_EL-IRES-EGFP or pMxmtBim_EL-IRES-EGFP. After 7 days of the retrovirus infection, when gene expression was confirmed by EGFP fluorescence (b, d and f), cultures were deprived of zVAD-FMK. At 18 h after zVAD-FMK removal, most of pMx-IRES-EGFP- and pMxBim_EL-IRES-EGFP-infected cells survived as identified by EGFP fluorescence (h and j), compared with the survival rate of 5% in pMxmtBim_EL-IRES-EGFP virus-infected cells (l). Both pMxBim_EL-IRES-EGFP- and pMxmtBim_EL-IRES-EGFP-infected cells died when M-CSF was removed from the cultures (i and k). (B) The survival ratio of EGFP-positive cells. Almost 100% of the control cells survived even 18 h after z-VAD-FMK removal in the presence of M-CSF. At 18 h after zVAD-FMK removal, almost 100% of pMx-IRES-EGFP- and >70% of pMxBim_EL-IRES-EGFP-infected cells survived as identified by EGFP fluorescence (EGFP and WTBim_EL), compared with the survival rate of 5% in pMxmtBim_EL-IRES-EGFP virus-infected cells (MTBim_EL). (C) The proteasome inhibitor MG132 enhanced the expression level of wtBim in pMxBim_EL-IRES-EGFP-infected OC precursors (WTBim) even in the presence of M-CSF, while no obvious upregulation of MTBim was observed in pMxmtBim_EL-IRES-EGFP infected cells (MTBim). (D) WTBim or MTBim was immunoprecipitated from cell lysates of pMxBim_EL-IRES-EGFP-infected (WTBim_EL) or pMxmtBim_EL-IRES-EGFP-infected cells (MTBim_EL) using anti-Bim polyclonal antibody, and the immunoprecipitates were immunoblotted with anti-ubiquitin antibody. Treating the cells with the proteasome inhibitor MG132 strongly increased the ubiquitylation of Bim, while no ubiquitylation of mtBim was observed.

Fig. 6. Effect of mutations in Bim that prevent ubiquitylation. (A) Phase contrast (a, c and e) and immunofluorescence microscopy: bim–/– bone marrow cells cultured in the presence of M-CSF (100 ng/ml) and zVAD-FMK (100 µM) were infected with either pMx-IRES-EGFP, pMxBim_EL-IRES-EGFP or pMxmtBim_EL-IRES-EGFP. After 7 days of the retrovirus infection, when gene expression was confirmed by EGFP fluorescence (b, d and f), cultures were deprived of zVAD-FMK. At 18 h after zVAD-FMK removal, most of pMx-IRES-EGFP- and pMxBim_EL-IRES-EGFP-infected cells survived as identified by EGFP fluorescence (h and j), compared with the survival rate of 5% in pMxmtBim_EL-IRES-EGFP virus-infected cells (l). Both pMxBim_EL-IRES-EGFP- and pMxmtBim_EL-IRES-EGFP-infected cells died when M-CSF was removed from the cultures (i and k). (B) The survival ratio of EGFP-positive cells. Almost 100% of the control cells survived even 18 h after z-VAD-FMK removal in the presence of M-CSF. At 18 h after zVAD-FMK removal, almost 100% of pMx-IRES-EGFP- and >70% of pMxBim_EL-IRES-EGFP-infected cells survived as identified by EGFP fluorescence (EGFP and WTBim_EL), compared with the survival rate of 5% in pMxmtBim_EL-IRES-EGFP virus-infected cells (MTBim_EL). (C) The proteasome inhibitor MG132 enhanced the expression level of wtBim in pMxBim_EL-IRES-EGFP-infected OC precursors (WTBim) even in the presence of M-CSF, while no obvious upregulation of MTBim was observed in pMxmtBim_EL-IRES-EGFP infected cells (MTBim). (D) WTBim or MTBim was immunoprecipitated from cell lysates of pMxBim_EL-IRES-EGFP-infected (WTBim_EL) or pMxmtBim_EL-IRES-EGFP-infected cells (MTBim_EL) using anti-Bim polyclonal antibody, and the immunoprecipitates were immunoblotted with anti-ubiquitin antibody. Treating the cells with the proteasome inhibitor MG132 strongly increased the ubiquitylation of Bim, while no ubiquitylation of mtBim was observed.