Deregulated cell death and lymphocyte homeostasis cause premature lethality in mice lacking the BH3-only proteins Bim and Bmf

Abstract

BH3 domain-only proteins (BH3-only) proteins are members of the Bcl-2 family that play crucial roles in embryogenesis and the maintenance of tissue homeostasis by triggering apoptotic cell death. The BH3-only protein Bim is critical for developmental apoptosis of lymphocytes, securing establishment of tolerance and for the termination of immune responses. Bim is believed to act in concert with other BH3-only proteins or members of the tumor necrosis factor receptor family in getting rid of unwanted cells. Bmf, a related BH3-only protein, was shown to play a role in B-cell homeostasis and to mediate cell death in response to certain apoptotic triggers, including glucocorticoid, histone deacetylase inhibitors, and overexpression of the c-Myc proto-oncogene. Here we show that Bim and Bmf have overlapping functions during mouse development and coregulate lymphocyte homeostasis and apoptosis in a nonredundant manner. Double deficiency of Bim and Bmf caused more B lymphadenopathy than loss of either BH3-only protein alone, and this was associated with autoimmune glomerulonephritis and a range of malignancies in aged mice. Thus, our results demonstrate that Bim and Bmf act in concert to prevent autoimmunity and malignant disease, strengthening the rational for the development of BH3-only protein mimicking therapeutics for the treatment of such disorders.

Figure 1

Impaired developmentally programmed apoptosis in mice lacking Bim and Bmf. Bim and Bmf mediate the removal of interdigitating mesenchymal cells in (A) WT, (B) Bim^−/−, and (C) Bmf^−/− mice, whereas (D-E) Bim^−/−Bmf^−/− mice present with soft tissue syndactyly, (G) vaginal aplasia, and (I) malocclusion of the incisors, (A,F,H) phenotypes that are not or only rarely observed in single knockout mice or WT control animals.

Figure 2

Bim and Bmf coregulate thymocyte apoptosis in response to stress, but only Bim appears critical for the developmentally programmed death of thymocytes. (A) Representative dot plots from flow cytometric analysis of single cell suspensions of thymocytes from 7- to 9-week-old mice of the indicated genotypes stained with CD4- and CD8-specific antibodies. (B) Sorted CD4⁺CD8⁺ thymocytes from 7- to 9-week-old mice of the indicated genotypes were placed in culture and incubated in the absence or presence of the indicated cell death inducers. Cell viability was assessed over time by AnnexinV/propidium iodide (PI) staining and flow cytometric analysis (Annexin V⁻/PI⁻ cells were considered alive). Data are presented as means ± SEM of ≥4 independent experiments and 4 to 8 animals per genotype. Significant differences by ANOVA (P < .05) at individual time points ^&between WT and Bim^−/−, ^#between WT and DKO, and ^§between Bim^−/− and DKO cells. Due to differences in spontaneous cell death in culture, specific drug-induced killing was calculated using the formula (induced apoptosis − spontaneous cell death)/(100 − spontaneous cell death) (%).

Figure 3

Overlapping roles of Bim and Bmf in the control of pre-B-cell survival. (A) Pre-B cells (B220⁺CD43⁻IgM⁻) were isolated from the bone marrow of 7- to 9-week-old mice of the indicated genotypes and placed in culture (left) without additional treatment or in the presence of 10 ng/mL recombinant IL-7. (Right) Relative survival advantages induced by IL-7 were assessed over time. (B) B220⁺IgM⁻ pro/pre-B cells were sorted from the bone marrow of 7- to 9-week-old mice of the indicated genotypes and cultured for 96 hours in the presence or absence of IL-7. sIgM expression was monitored by flow cytometric analysis. Representative dot plots from 1 of 3 independent experiments are shown. (C) Fluorescence-activated cell sorter-sorted pre-B cells from the bone marrow of 7- to 9-week-old mice of the indicated genotypes were treated in culture with the HDAC inhibitor SAHA or the glucocorticoid dexamethasone, and cell survival was assessed over time by flow cytometric analysis. Data are presented as means ± SEM of ≥3 independent experiments and 3 to 6 animals per genotype. *Significant differences by ANOVA (P < .05) at individual time points between WT and Bmf^−/− derived cells; ^&between WT and Bim^−/−, ^βbetween Bim^−/− and Bmf^−/−, ^#between WT and DKO, and ^§between Bim^−/− and DKO cells. Due to differences in spontaneous cell death in culture, specific drug-induced killing was calculated using the formula (induced apoptosis − spontaneous cell death)/(100 − spontaneous cell death) (%).

Figure 4

Loss of Bmf enhances the B cell-restricted lymphadenopathy caused by loss of Bim. (A-H) Spleens were harvested from 7- to 9-week-old mice of the indicated genotypes and single cell suspensions were counted to assess total leukocyte cellularity. Cell suspensions were stained using different combinations or fluorochrome-labeled antibodies for identification of total splenic T and B cells, as well as T1, T2, MZ, and FO B cells and PC/PBs. Data are presented as means ± SEM of ≥4 independent experiments and 4 to 8 animals per genotype. Significant differences by unpaired Student t test (P < .05) are indicated *between WT and Bmf^−/−, **between WT and Bim^−/−, ^#between WT and DKO, and ^§between Bim^−/− and Bim/Bmf DKO mice.

Figure 5

Combined loss of Bim and Bmf renders B cells independent of survival-promoting cytokines and refractory to BCR ligation-induced apoptosis. Transitional T1 (sIgM^highCD21^low) or T2 (sIgM^highCD21^high) B cells were sorted from spleens of 8- to 10-week-old mice of the indicated genotypes and placed in culture. Cells were either (A,C) left untreated or (B,D) stimulated with plate-bound anti-IgM F(ab)₂ fragments for 48 hours. (E-F) Follicular (FO) B cells (CD21⁺CD23⁺) or plasma cells/plasmablasts (PC/PB) (B220^+/lowCD138⁺) were sorted from spleens of mice of the indicated genotypes and placed untreated in culture. Cell viability was assessed after 24 and/or 48 hours by flow cytometric analysis. Data are presented as means ± SEM of ≥3 independent experiments and 3 to 5 animals per genotype. Significant differences (P < .05) are indicated *between WT and Bmf^−/−, **between WT and Bim^−/−, ^#between WT and DKO, and ^§between Bim^−/− and Bim/Bmf DKO mice.

Figure 6

Hypergammaglobulinemia in Bim^−/−Bmf^−/− mice is associated with increased levels of λ light chain-containing antibodies. (A) Sera were collected from 8- to 10-week-old mice of the indicated genotypes and total Ig, Igλ, and Igκ titers were quantified by ELISA. Overall quantities and λ/κ ratio are depicted based on OD₄₅₀ values. Data from OD values are represented as box plots (n = 5-6 animals per genotype). Box length equals interquartile range. Circles represent minimal and maximal values. Significant differences (P < .05) are indicated *between WT and Bmf^−/−, **between WT and Bim^−/−, ^#between WT and Bim^−/−Bmf^−/−, ^§between Bim^−/− and Bim^−/−Bmf^−/−, and ^&between WT and Bim^−/−Puma^−/− mice (n = 4). (B) Autoantibodies to dsDNA were quantified by ELISA using calf thymus dsDNA for coating of plates and serum samples from 8- to 10-week-old mice of the indicated genotypes. (C) Immunofluorescence staining of paraffin-embedded kidney sections reveals the presence of IgG deposits in glomeruli of female morbid tumor-free Bim^−/−, DKO, and Vav-Bcl-2 transgenic mice (age range, 8-14 months). Healthy Bmf^−/− mice, 18 months of age, were included as control. Magnification, ×400.

Figure 7

Combined loss of Bim and Bmf causes premature lethality. Cohorts of mice of the indicated genotypes were followed over 18 months. Mice developing overt signs of illness (cachexia, short breath, fuzzy fur, and enlarged lymph nodes or spleen) were killed, and organs were subjected to histopathological assessment in a blinded fashion revealing evidence for autoimmunity and tumor formation in a significant portion of double mutant mice. (A) Kaplan-Meier analysis of disease-free survival. Significant differences of interest (log-rank Mantle-Cox): WT vs Bim^−/− mice (P = .036); WT vs Bim^+/−Bmf^−/− (P = .0074); WT vs Bim^−/−Bmf^+/− (P = .0001); WT vs DKO (P < .0001); WT vs Vav-Bcl2 (P < .0001); Bmf^−/− vs DKO (P = .0009); Bim^−/− vs Bim^−/−Bmf^+/− (P = .0065); Bim^−/− vs DKO (P = .0055); and DKO vs Vav-Bcl2 (P = .032). The relative frequency of autoimmune pathologies and different types of malignancies observed are shown in B and C, respectively. (D) Kaplan-Meier analysis of disease-free survival of C57BL/6 Ly5.1⁺ coisogenic recipients reconstituted with 2 × 10⁶ bone marrow cells of mice of the indicated genotypes (all C57BL/6 Ly5.2⁺) after a single dose of γ-irradiation (9.5 Gy). Significant differences (log-rank Mantle-Cox) were observed between WT vs Bim^−/− (P = .022); WT vs DKO (P < .0001); Bmf^−/− vs Bim^−/− (P = .0262); Bmf^−/− vs DKO (P = .0003); and Bim^−/− vs DKO (P = .0423).