A stapled BIM peptide overcomes apoptotic resistance in hematologic cancers

Abstract

Cancer cells subvert the natural balance between cellular life and death, achieving immortality through pathologic enforcement of survival pathways and blockade of cell death mechanisms. Pro-apoptotic BCL-2 family proteins are frequently disarmed in relapsed and refractory cancer through genetic deletion or interaction-based neutralization by overexpressed antiapoptotic proteins, resulting in resistance to chemotherapy and radiation treatments. New pharmacologic strategies are urgently needed to overcome these formidable apoptotic blockades. We harnessed the natural killing activity of BCL-2–interacting mediator of cell death (BIM), which contains one of the most potent BH3 death domains of the BCL-2 protein family, to restore BH3-dependent cell death in resistant hematologic cancers. A hydrocarbon-stapled peptide modeled after the BIM BH3 helix broadly targeted BCL-2 family proteins with high affinity, blocked inhibitory antiapoptotic interactions, directly triggered proapoptotic activity, and induced dose-responsive and BH3 sequence–specific cell death of hematologic cancer cells. The therapeutic potential of stapled BIM BH3 was highlighted by the selective activation of cell death in the aberrant lymphoid infiltrates of mice reconstituted with BIM-deficient bone marrow and in a human AML xenograft model. Thus, we found that broad and multimodal targeting of the BCL-2 family pathway can overcome pathologic barriers to cell death.

Introduction

The BCL-2 family interaction network lies at the crossroads of the cell’s life-death decision (1, 2). Among the oncogenic signals that drive the development and maintenance of cancer, the pathologic reprogramming of the BCL-2 family interaction network has emerged as a formidable barrier to modern day anticancer treatment. Antiapoptotic members of the BCL-2 family contain a surface hydrophobic groove that can bind, sequester, and neutralize the critical BH3 death domains of proapoptotic members (3). Small molecules, such as ABT-737 and ABT-263, that selectively inhibit the antiapoptotic proteins BCL-2 and BCL-X_L induce apoptosis of tumors that are especially dependent upon this subset of survival proteins (4, 5). However, the compounds are rendered ineffective by cellular expression of alternate antiapoptotic proteins, such as MCL-1 and BFL1/A1, which lie outside the molecules’ range of binding specificity or by the relative or absolute absence of proapoptotic effectors, such as BAX and BAK (6, 7). Thus, a pharmacologic quest is underway to develop next-generation agents that simulate broader spectrum BH3-dependent killing activities and effectively deactivate the deflector shields of relapsed and refractory cancers.

Chief among the killer BH3-only proteins, Bcl-2–interacting mediator of cell death (BIM) exhibits the most broad-ranging BCL-2 protein interactions, engaging all of the antiapoptotic proteins with high affinity (8, 9) and also directly triggering death effectors, such as BAX (10, 11). A recent genetic analysis elegantly demonstrated that only BIM BH3, but not other BH3 sequences inserted into the BIM protein, could rescue the leukocyte accumulation and autoimmune phenotype of Bim^–/– mice (12). These data emphasize the unique killing potency of BIM BH3, which may derive from its higher-affinity interactions with antiapoptotics, dual capacity to block antiapoptotics and directly activate proapoptotics, and/or engagement of yet uncharacterized proteins that contribute to induction of cell death. From a clinical translational standpoint, BIM induction appears essential for cell death responses to a host of chemotherapeutic agents, and BIM suppression by genetic deletion, miRNA-based suppression, or protein interaction-based neutralization or degradation accounts for the emergence of drug resistance (13–15). Therefore, harnessing the BIM BH3 helix to promote a death response and, in particular, chemical replacement of this death functionality, in the context of drug resistance, are logical pharmacologic goals (16–18).

We previously developed stabilized α helix of BCL-2 domains (SAHBs), which are structurally stable, protease-resistant, and cell-permeable compounds capable of targeting BCL-2 family proteins with high affinity (19, 20). Here, we examined the pharmacologic potential of BIM SAHB_A, a stapled BIM BH3 helix that contains an i, i+4 all-hydrocarbon crosslink spanning positions 154 and 158 (designated “A”), to restore cell death in the context of formidable and clinically relevant apoptotic resistance. We found that BIM SAHB_A reactivates cell death in a BH3 sequence–dependent fashion in vitro and in vivo and, importantly, preferentially affects cells driven by a deregulated apoptotic pathway.

Results

BIM SAHB_A binds with high affinity to a broad range of BCL-2 family proteins.

To explore the breadth of BIM BH3 activity, we generated a hydrocarbon-stapled BIM BH3 helix (amino acids 146–166) (19) and a negative control peptide based on an R153D reverse polarity mutation within the core BH3 sequence (ref. 10, Figure 1A, and Supplemental Table 1; supplemental material available online with this article; doi: 10.1172/JCI46231DS1). Circular dichroism analysis demonstrated that BIM SAHB_A and BIM SAHB_A(R153D) exhibit enhanced α-helical structure in solution compared with that of the corresponding unmodified BIM BH3 peptide (ref. 19 and Supplemental Figure 1), effectively recapitulating the native structure of this domain when bound to apoptotic targets (21–23). Equilibrium fluorescence polarization (FP) binding analysis using FITC-βAla-BIM SAHB_A peptides and recombinant and tagless BCL-X_LΔC, MCL-1ΔNΔC, BCL-wΔC, and BFL1/A1ΔC revealed 1–11 nM K_d values for BIM SAHB_A across antiapoptotic targets, whereas R153D mutagenesis consistently reduced binding activity (Figure 1, B–E). We previously demonstrated, both by FP (19) and NMR (10), that BIM SAHB_A and an analog optimized for structural analysis (amino acids 145–164), respectively, also bind to proapoptotic BAX. In contrast to the static interaction between BH3 domains and their canonical binding pocket on antiapoptotic proteins, the BIM BH3 helix engages a geographically distinct BH3-binding site on BAX (10), triggering a major structural reorganization that underlies the dynamic BAX activation process (24). Indeed, combining BIM SAHB_A with recombinant BAX triggered dose-responsive BAX-mediated liposomal release, linking the direct binding interaction to BAX’s transformation from an inactive monomer to a functional pore (Figure 1F). Importantly, the single R153D point mutation at the binding interface abrogated the BAX-triggering activity of BIM SAHB_A in this assay (Figure 1G). As a further measure of specificity, BIM SAHB_A–triggered and BAX-mediated liposomal release was blocked in the presence of antiapoptotic proteins BCL-X_LΔC, MCL-1ΔNΔC, BCL-wΔC, and BFL-1/A1ΔC (Supplemental Figure 2). Thus, BIM SAHB_A engages a diversity of multidomain BCL-2 family proteins, consistent with the established broad-spectrum interaction profile of native BIM protein (8–10, 19).

Figure 1. Broad-spectrum BCL-2 family binding activity of a stapled BIM BH3 helix.

(A) Sequence composition of BIM SAHB_A and its R153D mutant. “X” denotes the position of the nonnatural amino acid within each peptide. (B–E) BIM SAHB_A binds to a broad range of antiapoptotic targets with high affinity. R153D mutagenesis shifts the isotherms to the right, reflecting the impairment of binding activity. Data are mean ± SEM for experiments performed at least in triplicate. mP, units of millipolarization. (F and G) BIM SAHB_A triggers dose-responsive BAX-mediated liposomal release, whereas the R153D mutant has no effect. The liposomal release assays were performed in triplicate with similar results.

BIM SAHB_A blocks antiapoptotic sequestration of the BAX/BAK BH3 helices and triggers BAX/BAK-dependent mitochondrial cytochrome c release.

An established paradigm for overcoming the apoptotic blockade of refractory cancer involves pharmacologic derepression of sequestered proapoptotic BCL-2 family members by competitive binding to antiapoptotic targets (8, 25–27). To explore the capacity of BIM SAHB_A to function as a derepressor compound, we performed competition assays using FITC-βAla-BAX SAHB_A/BCL-X_LΔC (K_d, 58.6 nM) and FITC-βAla-BAK SAHB_A/MCL-1ΔNΔC (K_d, 12.3 nM) (Figure 2, A and B). In each case, BIM SAHB_A dose responsively competed with the death effector BH3 helix for antiapoptotic binding, whereas BIM SAHB_A(R153D) was a relatively weak competitor (Figure 2, C and D).

Figure 2. BIM SAHB_A dissociates inhibitory antiapoptotic interactions and directly triggers BAX/BAK-dependent mitochondrial cytochrome c release.

(A and B) The antiapoptotic proteins BCL-X_LΔC and MCL-1ΔNΔC, respectively, sequester the proapoptotic BH3 helices of BAX and BAK in high-affinity complexes. (C and D) BIM SAHB_A effectively dissociates the inhibitory complexes, whereas R153D mutagenesis markedly impairs competitive binding. (E) BAX/BAK-deficient MLM (mouse liver mitochondria) undergo cytochrome c release in response to combination treatment with recombinant BAX (50 nM) and increasing doses of BIM SAHB_A. The specificity of action is highlighted by (a) blockade of the effect by added recombinant BCL-X_LΔC (2 μM), (b) loss of activity upon R153D mutagenesis, and (c) the inability of BAX and BIM SAHB_As to influence cytochrome c release when administered as single agents. (F) To monitor BAK-dependent cytochrome c release, WT MLM that contain native BAK were exposed to BIM SAHB_A and its R153D mutant. BIM SAHB_A triggered BH3 sequence–dependent and BAK-dependent cytochrome c release, with little to no effect on the corresponding BAX/BAK-deficient MLM in the 0–200 nM dosing range. Data are mean ± SEM for experiments performed at least in triplicate.

To examine the functional impact of the broad targeting activity of BIM SAHB_A in the mitochondrial context, we performed cytochrome c release assays in the presence of BAX (Alb-cre^posBax^fl/–Bak^–/– mouse liver mitochondria [MLM] with added recombinant BAX) or BAK (WT MLM) or without either effector protein (Alb-cre^posBax^fl/–Bak^–/– MLM). Applying a 0–200 nM dosing range, BIM SAHB_A selectively triggered recombinant BAX- and native BAK-dependent cytochrome c release, with little to no release observed with BIM SAHB_A(R153D) or in the absence of BAX and BAK (Figure 2, E and F). Thus, we demonstrate that low nanomolar BIM SAHB_A has the capacity to broadly engage BCL-2 family targets to selectively trigger BH3 sequence–dependent and BAX/BAK-dependent mitochondrial apoptosis.

BIM SAHB_A activates caspase-3/7 and induces cell death in resistant hematologic cancer cells.

We examined the utility of BIM SAHB_A in overcoming the apoptotic blockades of refractory cancer by testing a series of hematologic cancer cell lines that are relatively resistant to ABT-737 (5, 6, 28, 29), a selective BCL-2/BCL-X_L inhibitor whose activity is dampened by cellular expression of antiapoptotic proteins that lie outside its binding spectrum (4, 6, 7, 30). We first documented that BIM SAHB_A and its mutant derivative exhibit dose-equivalent uptake by the panel of cancer cell lines, which included OCI-AML3 (acute myeloid leukemia [AML]), U937 (monocytic leukemia/lymphoma), OPM-2 (multiple myeloma), Pfeiffer (diffuse large B cell lymphoma), and K562 (chronic myelogenous leukemia) cells (Figure 3A). Cultured cells were treated with vehicle or FITC-derivatized BH3 peptides (1 μM) in serum-free media for 2 hours, followed by centrifugation, trypsinization (to digest surface proteins and thereby eliminate potential nonspecific surface binding), washing, lysate preparation, electrophoresis, and fluorescence scan detection. Importantly, FITC-BIM SAHB_A and FITC-BIM SAHB_A(R153D) were present at similar levels within the cellular lysates, whereas the vehicle- and unmodified FITC-BIM BH3–treated cells showed no signal (Figure 3A). Live cell confocal microscopy of FITC-SAHB–treated cells further confirmed cellular uptake of the stapled peptides (Figure 3B). Whereas FITC-BIM SAHB_A largely colocalized with mitochondria, the predominant site of BCL-2 family targets (31), the R153D point mutant that exhibits substantially impaired binding activity showed little to no such colocalization. We further confirmed that at the micromolar dosing levels used for cellular studies, BIM SAHB_A peptides were solubilized as uniform, monomeric species, as assessed by gel filtration chromatography (Supplemental Figure 3).

Figure 3. Cellular uptake and localization of BIM SAHB_As.

(A) Cultured cancer cells were exposed to FITC derivatives of BIM BH3 peptides (1 μM) for 2 hours, followed by centrifugation, trypsinization, washing, lysate preparation, electrophoresis, and fluorescence detection. FITC-BIM SAHB_A and its R153D point mutant were present at similar levels in the cellular lysates, whereas no uptake was observed for the unmodified FITC-BIM BH3 peptide. (B) Live cell confocal microscopy of U937 cancer cells treated with FITC-BIM SAHB_As (1 μM, 2 hours) demonstrated the uptake and intracellular distribution of the peptides. FITC-BIM SAHB_A, but not the impaired R153D mutant, largely colocalizes with mitochondria, the site of BCL-2 family targets. Nuclei (Hoechst), SAHBs (FITC), mitochondria (MitoTracker), and the BIM SAHB_A/mitochondria colocalization (FITC/MitoTracker) are shown in blue, green, red, and yellow, respectively. An example of the BIM SAHB_A/mitochondria colocalization is highlighted by the white arrow. Original magnification, ×100.

Cancer cells were then exposed to serial dilutions of vehicle, BIM SAHB_A, and BIM SAHB_A(R153D), and the effect of treatment was monitored by cell viability assay, annexin V binding FACS analysis, and caspase-3/7 activation assay (Figure 4). In each case, BIM SAHB_A impaired cancer cell viability in a dose-responsive and sequence-dependent manner (IC₅₀, 1.8–3.4 μM), an effect that correlated with both annexin V positivity and caspase-3/7 activation. When applied to nonmalignant cells that also demonstrated efficient uptake of BIM SAHB_A (Supplemental Figure 4), such as the human fibroblast cell line WS1 and WT and Bax^–/–Bak^–/– (DKO) mouse embryo fibroblasts (MEFs), neither BIM SAHB_A nor BIM SAHB_A(R153D) had any cytotoxic or caspase-3/7 activation effects in the same 0–10 μM dosing range (Figure 5). Only when increasing the dose to 20 μM do we begin to observe a decrease in WS1 cell viability, which coincides with caspase-3/7 activation (Supplemental Figure 5A); in adherent MEFs, this dose escalation does not impact cell viability, but a mild elevation in caspase-3/7 activation is observed in WT MEFs but not DKO MEFs, consistent with a BAX/BAK-dependent proapoptotic mechanism (Supplemental Figure 5, B and C). To confirm that the relative lack of toxicity of human BIM SAHB_A in murine fibroblasts did not derive from cross-species variation, we repeated the experiment with the corresponding murine BIM SAHB_As (Supplemental Table 1) and obtained similar results (data not shown). Thus, these data indicate that at low micromolar dosing, BIM SAHB_A activates the death pathway in a variety of hematologic cancer cells that are otherwise resistant to selective BCL-2/BCL-X_L inhibition (6, 7). Importantly, the relative resistance of nonmalignant fibroblasts to BIM SAHB_A treatment suggested that a therapeutic window may exist for BIM BH3 replacement in vivo, with preferential toxicity to cells driven by deregulated apoptosis.

Figure 4. Sequence-dependent cytotoxicity of BIM SAHB_A in hematologic cancer cells resistant to selective BCL-2/BCL-X_L inhibition.

BIM SAHB_A dose responsively reduces the viability of (A) OCI-AML3, (B) U937, (C) OPM-2, (D) Pfeiffer, and (E) K562 cancer cells, as measured at 24 hours by CellTiter-Glo assay (left column). R153D mutagenesis markedly impairs the cytotoxic activity of BIM SAHB_A. Cells exposed to BIM SAHB_A, but not its R153D mutant, exhibit dose-dependent annexin V binding at 6 hours after treatment, as monitored by FACS analysis (middle column). The negative effect of BIM SAHB_A on cancer cell viability correlates with dose-responsive activation of caspase-3/7, as assessed by monitoring the cleavage of proluminescent caspase-3/7 substrate at 6 hours after treatment (right column). Data are mean ± SEM for experiments performed at least in triplicate. RLU, relative luminescence units.

Figure 5. Viability of human and mouse fibroblasts exposed to BIM SAHB_A peptides.

WS1 human fibroblasts showed no response to BIM SAHB_A peptides (0–10 μM), as assessed by (A) viability and (B) caspase-3/7 activation assays performed at 24 and 6 hours after treatment, respectively. Staurosporine (100 nM), which impaired viability and induced caspase-3/7 activation, served as a positive control. Adherent WT and DKO MEFs likewise showed no response to BIM SAHB_A treatment in this dose range, as assessed by (C and E) viability and (D and F) caspase-3/7 activation assays. As a positive control, (C and D) staurosporine (100 nM) triggered caspase-3/7 activation and reduced viability of WT MEFs, (E and F) whereas DKO MEFs exhibited relative resistance to staurosporine with no caspase-3/7 activation, as reported previously (50). Data are mean ± SEM for experiments performed at least in triplicate.

Mechanism of BIM SAHB_A–induced cytotoxicity.

To further probe the protein interaction-based mechanism of BIM SAHB_A activity, we treated OCI-AML3 cells with BIM SAHBs (0–10 μM) and monitored for disruption of the inhibitory MCL-1/BAK complex (32) and conversion of BAX to its conformationally activated form (10) by immunoprecipitation with antibodies to MCL-1 and the BAX activation epitope (6A7) (33), respectively. Cellular treatment with BIM SAHB_A, using the same dose range that induced cytotoxicity, annexin V binding, and caspase-3/7 activation (Figure 4), triggered both dose-responsive dissociation of MCL-1/BAK (Figure 6A) and exposure of the BAX activation epitope (Figure 6B), consistent with pharmacologic induction of the mitochondrial apoptotic pathway. Importantly, BIM SAHB_A(R153D) showed no such effects. In accordance with the observed BIM SAHB_A–induced biochemical changes, OCI-AML3 cells likewise manifested the morphologic features of apoptosis, as reflected by cellular contraction, chromatin condensation and marginalization, and nuclear fragmentation of Wright-Giemsa–stained cytospun cells (Figure 6C).

Figure 6. BIM SAHB_A treatment induces biochemical and morphological changes that reflect a proapoptotic mechanism of action.

(A) OCI-AML3 cells treated with BIM SAHB_A (0–10 μM) manifest dose-responsive dissociation of the native BAK/MCL-1 complex, as assessed by MCL-1 immunoprecipitation and anti-BAK Western analysis. No effect is observed upon treatment with BIM SAHB_A(R153D). (B) BIM SAHB_A (0–10 μM) likewise triggers BAX activation, as monitored by BAX immunoprecipitation using the 6A7 antibody, which selectively recognizes BAX’s N-terminal activation epitope. R153D mutagenesis abrogates the effect. (C) A Wright-Giemsa stain of BIM SAHB_A–treated OCI-AML3 cells (5 μM) revealed cellular shrinkage, chromatin condensation and marginalization, and nuclear fragmentation, all hallmark morphological features of apoptosis induction (arrowheads). Vehicle- and BIM SAHB_A(R153D)–treated cells showed no such effects. Original magnification, ×100; third column, additional ×15 zoom. (D) BIM SAHB_A (25 mg/kg/d i.v.) suppressed tumor growth compared with vehicle (n = 7 per arm; 2-way ANOVA, P = 0.018), as assessed by in vivo bioluminescence imaging of a subcutaneous OCI-AML3 tumor xenograft model. Data are mean ± SEM. *P < 0.05.

Tumor suppressive activity of BIM SAHB_A as a single agent and in combination.

To extend these findings to an in vivo context, we tested the efficacy of BIM SAHB_A in an OCI-AML3 human xenograft model (34). OCI-AML3-LucNeo cells (5 × 10⁶ cells) were implanted subcutaneously with 30% Matrigel into the right hind flank of 6-week-old female NOD scid gamma (NSG) mice. Mice with equivalent disease, as determined by bioluminescent imaging on day 8 after inoculation, were divided into 2 groups (n = 7 per arm), treated daily with either i.v. vehicle (5% DMSO/D5W) or BIM SAHB_A (25 mg/kg), and imaged on days 0, 4, 7, and 10. Throughout the treatment period, mice receiving BIM SAHB_A manifested a statistically significant suppression of tumor growth compared with that of vehicle-treated animals, as reflected by an approximately 50% decrease in relative bioluminescence (2-way ANOVA, P = 0.018) (Figure 6D).

We next explored whether the single agent activity of BIM SAHB_A could be combined with other proapoptotic agents to enhance the antitumor response. BIM SAHB_A was combined with either TRAIL, an extrinsic apoptotic pathway stimulus, or ABT-737, an intrinsic apoptotic pathway stimulus, in synergy analyses (35) performed on OCI-AML3 cells. The cancer cells were treated with serial dilutions of BIM SAHB_A, TRAIL, or ABT-737 or the combination of BIM SAHB_A and TRAIL or BIM SAHB_A and ABT-737, followed by viability assay at 24 hours (Figure 7, A and B). In each case, synergy was observed, as reflected by combination indices of less than 1 at the 50%, 75%, and 90% effective doses. To determine whether a drug combination could enhance the antitumor activity of BIM SAHB_A in vivo, we repeated the OCI-AML3 xenograft study (n = 8 per arm), examining the impact of adding the orally bioavailable BCL-2/BCL-X_L inhibitor ABT-263 (5, 36) (100 mg/kg/d per orem [PO]) to either BIM SAHB_A or its point-mutant control BIM SAHB_A(R153D) (25 mg/kg/d i.v.). Mice were treated for 15 days with vehicle, the single agent SAHB, or the combination of SAHB with ABT-263, and bioluminescence was measured on days 0, 3, 7, 10, 13, and 16. As previously observed (Figure 6D), single agent BIM SAHB_A suppressed tumor growth throughout the treatment period (P = 0.0299) (Figure 7C). Single R153D point mutagenesis abrogated this effect, with BIM SAHB_A(R153D) treatment showing no statistically significant tumor suppressive activity (P = 0.3934). The addition of ABT-263 to BIM SAHB_A yielded the most significant antitumor response (P = 0.0009), whereas combining ABT-263 with BIM SAHB_A(R153D) did not achieve a significant antitumor effect (P = 0.1982) (Figure 7C). Importantly, neither BIM SAHB_A nor BIM SAHB_A/ABT-263 combination treatment adversely affected peripheral white blood cell counts (Supplemental Figure 6), and no weight loss was observed among the treatment groups (Figure 7D). Taken together, the in vivo efficacy data are consistent with BIM BH3 sequence–dependent antitumor activity, both as a single agent and in combination with ABT-263, in an otherwise aggressive and relatively resistant xenograft model of human AML.

Figure 7. Antitumor activity of combination treatment with BIM SAHB_A.

Treatment of OCI-AML3 cells with the combination of (A) BIM SAHB_A/TRAIL or (B) BIM SAHB_A/ABT-737 caused a synergistic decrease in cell viability, as reflected by CalcuSyn analysis and the plotted dose-effect curves (combination index [CI] <1). EC₅₀, 50% effective concentration; ED₇₅, 75% effective dose; ED₉₀, 90% effective dose. (C) The combination of BIM SAHB_A (25 mg/kg/d i.v.) and ABT-263 (100 mg/kg/d PO) caused the most significant suppression of OCI-AML3 tumor growth in vivo, as assessed by in vivo bioluminescence imaging. Two-way ANOVA, overall P = 0.0052. Vehicle versus: BIM SAHB_A, P = 0.0299; BIM SAHB_A(R153D), P = 0.3934; BIM SAHB_A and ABT-263, P = 0.0009; and BIM SAHB_A(R153D) and ABT-263, P = 0.1982. Data are mean ± SEM (n = 8 per arm). *P < 0.05; **P < 0.01 (D) No weight loss manifested among the 5 treatment cohorts over the 16-day observation period. Data are mean ± SEM (n = 8 per arm).

Therapeutic potential of BIM BH3 replacement.

To explore the in vivo capacity of BIM SAHB_A to selectively reactivate the death pathway in cells driven by a formidable genetic apoptotic blockade, we transplanted the bone marrow (2 × 10⁶ cells) from Bim^–/– mice (37) into Rag2^–/–gc^–/– recipients, yielding uniform cohorts of mice manifesting the immune cell accumulation of native Bim^–/– mice (38), as evidenced by organ infiltrates (Figure 8A), hypergammaglobulinemia (Supplemental Figure 7A), and splenomegaly (Supplemental Figure 7B) evaluated at 19, 15, and 12 weeks post transplant, respectively. Histologic analysis of solid organs, such as the kidney and liver by H&E stain and anti-B220 immunohistochemistry, documented the presence of pathologic B cell infiltrates throughout the tissues (Figure 8A). Post transplant day–matched mouse cohorts (n = 4) were treated with daily i.v. injections of vehicle (5% DMSO/D5W) or 10 mg/kg of murine BIM SAHB_As for 5 days, followed by necropsy and histologic analysis. The relative degree of apoptosis induction was measured by TUNEL staining, which revealed a statistically significant increase in TUNEL positivity in the liver and kidney infiltrates of BIM SAHB_A–treated mice compared with that in vehicle- and BIM SAHB_A(R153D)–treated animals (Figure 8, B–D). Importantly, no increase in TUNEL staining was observed in the surrounding normal liver and kidney parenchyma of BIM SAHB_A–treated mice (Figure 8, B, C, and E). Of note, Bim^–/– bone marrow–reconstituted Rag2^–/–gc^–/– mice treated with BIM SAHB_A for 5 days also manifested a trend toward lower spleen weight and spleen weight per body weight compared with that of vehicle- and BIM SAHB_A(R153D)–treated animals (Supplemental Figure 8, A and B). An analogous study performed with WT B6/129 mice showed equivalent spleen weights across the vehicle, BIM SAHB_A, and BIM SAHB_A(R153D) treatment groups (Supplemental Figure 8, C and D). In addition, no suppressive effects on peripheral white blood cell counts or peripheral or splenic lymphocyte subsets were detected in the WT mice (Supplemental Figure 9). Thus, we find that BIM SAHB_A triggered cell death of deregulated Bim^–/– immune cells in a BH3 sequence–specific manner and spared normal host tissues, highlighting the potential of this BIM BH3 replacement strategy to surmount pathologic apoptotic blockades in vivo.

Figure 8. Pharmacologic replacement of BIM BH3 restores the death pathway in Bim^–/– immune cell infiltrates.

(A) The infiltrative B cell disease of Bim^–/– mice was recapitulated in irradiated Rag2^–/–gc^–/– mice transplanted with the bone marrow of Bim^–/– mice. Histologic analysis by H&E stain and anti-B220 immunohistochemistry at 19 weeks post transplant revealed extensive B cell infiltration of solid organs. Original magnification: ×2 (whole organs); ×40 (H&E), ×20 (B220 immunohistochemistry). (B and C) Increased TUNEL staining was evident in the aberrant infiltrates of BIM SAHB_A–treated mice as compared with that in vehicle- and BIM SAHB_A(R153D)–treated animals. Importantly, the observed increase in TUNEL positivity in BIM SAHB_A–treated mice appeared restricted to the deregulated B cells and was not evident in the normal parenchymal tissue. Original magnification, ×100. (D) The percentage TUNEL positivity within the immune cell infiltrates was calculated by counting approximately 6,000 cells per 5-μm tissue section from each experimental animal (n = 4 per cohort) in a blinded fashion. A statistically significant increase in percentage of TUNEL positivity was observed in the kidney (P = 0.007) and liver (P = 0.0007) infiltrates of BIM SAHB_A–treated mice, an effect that was abrogated by R153D point mutagenesis. Data are mean ± SD. **P < 0.01. (E) Importantly, no increase in the number of TUNEL-positive cells was found in kidney and liver parenchymal tissue devoid of the infiltrates. Data are mean ± SD.

Discussion

To achieve a survival advantage, cancer cells overexpress antiapoptotic BCL-2 family proteins (39) and neutralize the proapoptotic members by interaction-mediated sequestration (40, 41) and/or genetic deletion (14, 15, 42, 43). Targeted pharmacologic strategies to reactivate apoptosis in cancer have focused on inhibiting individual (32) or subsets (4, 5, 44, 45) of BCL-2 family antiapoptotic proteins. Relapsed and refractory cancer, however, is often characterized by the emergence of apoptotic blockade reinforcements that derive from (a) a surge in expression of alternate antiapoptotic proteins or (b) genetic silencing/deletion of proapoptotic proteins, such as the BH3-only subclass. In this context, multimodal targeting of the BCL-2 family pathway with agents designed to block all antiapoptotic proteins, directly trigger the multidomain death proteins, and even pharmacologically replace the death functionality of deleted BH3-only proteins, may be ideally suited to confront chemoresistant disease.

The BH3-only protein BIM has emerged as the most potent death sentinel, engaging all antiapoptotic targets (8, 9, 12) and directly triggering the multidomain proapoptotic proteins (10, 11, 19, 46). Thus, we generated a structurally reinforced and cell permeable stapled BIM BH3 helix to maximally reactivate the BCL-2 family death pathway in resistant cancer. We found that BIM SAHB_A binds to a diversity of antiapoptotic and proapoptotic targets with high affinity and potently triggers BAX- and BAK-dependent mitochondrial apoptosis in vitro. In a cellular context, BIM SAHB_A dissociates inhibitory heterodimeric interactions between antiapoptotic and proapoptotic proteins, such as MCL-1/BAK, and triggers conformational activation of the proapoptotic effector protein BAX. Consistent with these biochemical changes induced by low micromolar dosing, BIM SAHB_A dose responsively impairs the viability of hematologic cancer cells, promoting annexin V binding, caspase-3/7 activation, and the morphologic features of apoptosis. Combining BIM SAHB_A with an extrinsic or intrinsic apoptotic pathway stimulant caused synergistic cancer cell death, highlighting the potential utility of BIM SAHB_A in lowering the apoptotic threshold to enhance the cytotoxicity of chemotherapeutic agents. Importantly, BIM SAHB_A imposes preferential toxicity upon cancer cells, as no significant changes in viability were observed for human and murine fibroblasts subjected to BIM SAHB_A treatment at the same micromolar dosing levels.

In the in vivo context, as a single agent and in combination with ABT-263, BIM SAHB_A suppressed tumor growth in a xenograft model of human AML. In addition, BIM SAHB_A selectively restored the death pathway in tissue-infiltrating Bim^–/– immune cells. Single point mutagenesis of the BIM SAHB_A helical binding surface abrogated death-inducing activity, and no acute toxicity was observed in normal tissues, highlighting the specificity of action of BIM SAHB_A. Interestingly, the therapeutic window observed for this and other BCL-2 family modulators (4, 20, 44) appears to derive from the antiapoptotic reserve of unstressed cells, which sustain targeted activation of the BCL-2 family pathway by deploying their natural reservoir of antiapoptotics to bind and sequester activated proapoptotic proteins. In contrast, diseased cells whose antiapoptotic reserve is taxed by tonic death signaling are unable to withstand further BAX/BAK-mediated mitochondrial assault, making them preferentially susceptible to the apoptotic trigger (8). Thus, multimodal targeting of the BCL-2 family pathway through combined antiapoptotic inhibition, direct proapoptotic activation, and pharmacologic replacement of genetically deleted BH3 functionalities may unmask a critical opportunity to overturn the apoptotic blockades of relapsed and refractory cancer and achieve a durable clinical remission.

Methods

SAHB synthesis and characterization.

SAHBs were synthesized, derivatized, purified, stored, reconstituted, and characterized by circular dichroism as described in detail previously (20, 47). All peptides were purified by liquid chromatography/mass spectrometry to more than 95% purity and quantified by amino acid analysis. The primary sequence compositions of all SAHBs used in this study are listed in Supplemental Table 1. To examine cellular uptake, nonadherent hematologic cancer cells (2 × 10⁶ cells) and adherent fibroblasts (4 × 10⁵ fibroblasts plated 48 hours prior to testing) were incubated with vehicle or BIM BH3 peptides (1 μM) in Opti-MEM (Invitrogen) at 37°C for 2 hours in the dark. Cells were washed with PBS (Invitrogen) and then incubated with 0.25% trypsin (Invitrogen) for 5 minutes at 37°C, washed again with PBS, and incubated on ice in lysis buffer (50 mM Tris [pH 7.4], 150 mM NaCl, 1 mM EDTA, 1 mM DTT, 1% [v/v] Triton X-100, complete protease inhibitor tablet [Roche]). Cellular debris was pelleted at 14,000 g for 10 minutes at 4°C, and the supernatant was collected, electrophoresed, and subjected to fluorescence imaging using a Typhoon 9400 (GE Healthcare Life Sciences). To monitor the solution state of BIM SAHBs, lyophilized peptides were resuspended in aqueous buffer containing 50 mM Tris and 100 mM NaCl (pH 8) to generate a 200 μM solution, which was subjected to FPLC analysis using a Superdex-75 column (Amersham Pharmacia) and molecular weight standards (GE Healthcare).

Protein production.

Recombinant and tagless BCL-X_LΔC, MCL-1ΔNΔC, BCL-wΔC, and BFL-1/A1ΔC were produced and purified as described previously (19, 32). Briefly, glutathione-S-transferase fusion proteins were expressed in Escherichia coli BL21 (DE3) using pGEX2T (Pharmacia Biotech) constructs. Bacterial cells were cultured in ampicillin-containing Luria Broth, and protein expression was induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside. The bacterial pellet was resuspended in PBS containing 1 mg/ml lysozyme, complete protease inhibitor tablet, and 1% (v/v) Triton X-100 and sonicated, and, after centrifugation at 45,000 g for 45 minutes, the supernatant was applied to a glutathione-agarose (Sigma-Aldrich) column and washed with PBS. Tagless protein was obtained by overnight on-bead digestion with thrombin (50 units) (GE Healthcare) in PBS (3 ml) at room temperature, and the cleaved proteins were purified by size exclusion chromatography using 150 mM NaCl and 50 mM Tris (pH 7.4) buffer conditions. Full-length recombinant BAX protein was expressed as previously described (10), and pure monomeric protein was obtained by size exclusion chromatography using a Superdex-75 column (GE Healthcare) and 20 mM HEPES (pH 7.2) and 150 mM KCl buffer conditions.

FP binding assay.

FP assays were performed as previously described (19, 48). FITC-BIM SAHB_A and FITC-BIM SAHB_A(R153D) (25 nM) were incubated with serial dilutions of recombinant protein in binding buffer (50 mM Tris, 100 mM NaCl [pH 8.0]) until equilibrium was reached. For competition assays, FITC-BAX SAHB_A and FITC-BAK SAHB_A (25 nM) were incubated with BCL-X_LΔC and MCL-1ΔNΔC (100 nM), respectively, in binding buffer for 10 minutes at room temperature. Serial dilutions of acetylated BIM SAHB_A peptides were then added and incubated in the dark at room temperature until equilibrium was reached. FP was measured using a SpectraMax M5 microplate reader (Molecular Devices). For direct binding experiments, dissociation constants (K_d) were calculated by nonlinear regression analysis of dose-response curves using Prism software (GraphPad) (48). For competition experiments, IC₅₀ values were determined by nonlinear regression analysis of dose-response curves using a one-site competition model (Graphpad).

Liposomal release assay.

Large unilamellar vesicles (LUVs) were generated from a lipid mixture containing 48% phosphatidylcholine, 28% phosphatidylethanolamine, 10% phosphatidylinositol, 10% dioleoyl phosphatidylserine, and 4% tetraoleoyl cardiolipin (Avanti Polar Lipids), as described previously (49). Briefly, the lipid mixture (1 mg aliquots) was prepared from chloroform stocks and dried as a thin film in glass test tubes under nitrogen gas and then under vacuum for 15 hours. Lipids were reconstituted with 45 mM DPX (Invitrogen) and 12.5 mM ANTS (Invitrogen) in 10 mM HEPES (pH 7.0), 200 mM KCl, and 1 mM MgCl₂. The lipid suspensions were freeze-thawed 10 times and extruded 11 times through a polycarbonate membrane (100-nm pore size; Avestin) using an extruder (Avanti Polar Lipids), resulting in the generation of LUVs. LUVs were isolated from free DPX/ANTS by gel filtration chromatography using a 10-ml Sepharose CL2B column (GE Healthcare). Liposomal release was quantified based on the increase in fluorescence that occurs when the ANTS fluorophore is separated from the DPX quencher upon release from the liposomes into the supernatant (excitation 355 nm; emission 520 nm; 20-nm bandwidths). Monomeric full-length human BAX protein was purified as described above. LUVs (10 μl) were treated with BAX (400 nM) and SAHBs (200 nM, 400 nM), singly and in combination, in 96-well format (Corning) in a total reaction volume of 100 μl. To evaluate the inhibitory capacity of antiapoptotic BCL-2 family proteins, recombinant BCL-X_LΔC, MCL-1ΔNΔC, BCL-wΔC, and BFL-1/A1ΔC (400 nM) were preincubated with BIM SAHB_A (400 nM) for 20 minutes and then treated with BAX (400 nM) and LUVs (10 μl) in 96-well format in a total reaction volume of 100 μl. Fluorescence (F) was measured from t = 0 (F0) to t = 2 hours, and then Triton X-100 was added to a final concentration of 0.2% (v/v), followed by measurement of fluorescence for an additional 10 minutes to determine maximal release (F100). The percentage ANTS/DPX release was calculated with the following formula: ([F – F0]/[F100 – F0]) × 100.

Cytochrome c release assay.

MLM (0.5 mg/ml) were isolated from WT and Alb-cre^posBax^fl/–Bak^–/– mice as described previously (19, 48) and incubated with the indicated BCL-2 family proteins and SAHB ligands at 37°C for 40 minutes. The pellet and supernatant fractions were isolated by centrifugation, and cytochrome c was quantitated using a colorimetric ELISA assay (R&D Systems). The percentage cytochrome c released into the supernatant from releasable mitochondrial pools (%cytoc) was calculated according to the following equation: %cytoc = [(cytoc_sup – cytoc_backgr)/(cytoc_total – cytoc_backgr)] × 100, where cytoc_sup, cytoc_backgr, and cytoc_total represent cytochrome c detected in the supernatants of SAHB-, vehicle-, and 1% Triton-X 100–treated samples, respectively.

Cell culture.

OCI-AML3, U937, OPM-2, Pfeiffer, and K562 cells were maintained in RPMI 1640 media (Invitrogen) supplemented with 10% FBS, 100 U ml^–1 penicillin/streptomycin, 2 mM l-glutamine, 50 mM HEPES, 0.1 mM MEM nonessential amino acids, and 50 μM β-mercaptoethanol. WT and DKO MEFs were maintained in DMEM (Invitrogen) supplemented with 10% FBS, 100 U ml^–1 penicillin/streptomycin, 2 mM l-glutamine, 0.1 mM MEM nonessential amino acids, and 50 μM β-mercaptoethanol. WS1 fibroblasts were maintained in EMEM (ATCC) supplemented with 10% FBS, 100 U ml^–1 penicillin/streptomycin, 2 mM l-glutamine, 1 mM sodium pyruvate, and 0.1 mM MEM nonessential amino acids.

Live cell confocal microscopy.

Cells were incubated with 1 μM FITC-BIM SAHBs for 2 hours in Opti-MEM media lacking phenol red (Invitrogen), followed by 30 minutes of incubation in prewarmed Opti-MEM solution containing 50 nM MitoTracker Red (Invitrogen) and 5 μg/ml Hoechst 33342 (Invitrogen). The cells were then washed twice and plated on glass bottom poly-d-lysine–coated culture dishes (MatTek Corporation) in Opti-MEM media lacking phenol red (Invitrogen). Confocal images were collected on a Yokogawa CSU-X1 spinning disk confocal system (Andor Technology) attached to a Nikon Ti-E inverted microscope (Nikon Instruments). Excitation of the 3 fluorophores was performed sequentially using the 405-nm, 488-nm, and 561-nm lasers. Images were acquired using a ×100 Plan Apo objective lens with a Hamamatsu OrcaER camera (Hamamatsu Photonics). Acquisition parameters, shutters, filter positions, and focus were controlled by Andor iQ software (Andor Technology).

Cell viability assay.

Cancer cells (1 × 10⁴ cells, 50 μl) were aliquoted in 96-well opaque plates and treated with serial dilutions of vehicle (0.2% DMSO), BIM SAHB_A, or BIM SAHB_A(R153D) in serum-free RPMI media for 2 hours, followed by serum replacement with 20% FBS-containing RPMI media (50 μl) for a final volume of 100 μl containing 10% FBS. Cell viability was assayed at 24 hours by addition of CellTiter-Glo chemiluminescence reagent (Promega), and luminescence measured by a SpectraMax M5 microplate reader (Molecular Devices). For adherent fibroblasts, WS1 (2.5 × 10³) and WT (2.5 × 10³) and DKO (1.5 × 10³) cells were plated in their respective culture media (see above), and, 24 hours later (~75%–90% cellular confluence), media was removed and the cells were washed with the corresponding serum-free media. The indicated doses of SAHB were then added in serum-free media (50 μl), serum was replaced after 2 hours (20% FBS-containing media, 50 μl), and cell viability was measured at 24 hours as above. For synergy analyses, OCI-AML3 cells (1 × 10⁴ cells) were treated with serial dilutions of BIM SAHB_A (0.01–20 μM), human TRAIL (0.2–200 ng/ml; R&D systems), ABT-737 (0.01–20 μM; Selleck), or the BIM SAHB_A/TRAIL and BIM SAHB_A/ABT-737 combinations. Cell viability was measured at 24 hours by CellTiter-Glo, according to the manufacturer’s protocol. Synergy analyses were then performed using CalcuSyn software (35).

Annexin V binding.

Cancer cells (5 × 10⁴ cells) were treated with SAHBs as above, washed with PBS and annexin V binding buffer after 6 hours, and then incubated with FITC–annexin V (BD Pharmingen) for 20 minutes at room temperature in the dark, as described previously (48). Cells were then washed with binding buffer and analyzed by flow cytometry on a FACSCalibur (Becton Dickinson), and the percentage annexin V positivity was calculated using FlowJo software.

Caspase-3/7 activation assay.

Cancer cells and fibroblasts were treated with SAHBs as described above for cell viability assays, and caspase-3/7 activation was measured at 6 hours by addition of the Caspase-Glo 3/7 chemiluminescence reagent in accordance with the manufacturer’s protocol (Promega). Luminescence was detected by a SpectraMax M5 microplate reader (Molecular Devices).

Immunoprecipitation.

To monitor the MCL-1/BAK complex, OCI-AML3 cells (5 × 10⁶ cells) were incubated with vehicle, BIM SAHB_A, or BIM SAHB_A(R153D) at the indicated concentrations in Opti-MEM media at 37°C for 2 hours, followed by PBS wash at 4°C, and incubation on ice in 200 μl lysis buffer (50 mM Tris [pH 7.4], 150 mM NaCl, 1 mM EDTA, 1 mM DTT, 0.5% [v/v] NP-40, complete protease inhibitor tablet). After 15 minutes, cellular debris was pelleted at 14,000 g for 10 minutes at 4°C, and the supernatant was collected and incubated with pre-equilibrated protein A/G sepharose beads (Santa Cruz Biotechnology Inc.). The precleared supernatant was then incubated with anti–MCL-1 antibody (4 μg/ml) (S-19, Santa Cruz Biotechnology Inc.) overnight at 4°C, followed by the addition of pre-equilibrated protein A/G sepharose beads for 1 hour. The beads were pelleted, washed with lysis buffer 3 times at 4°C, and protein eluted by heating the beads at 90°C for 10 minutes in LDS/DTT loading buffer. The immunoprecipitates were subjected to electrophoresis and Western blot analysis using anti-BAK (1:500) (NT, CalBiochem) and anti–MCL-1 (1:200) (S-19, Santa Cruz Biotechnology Inc.) antibodies. For detection of activated BAX, OCI-AML3 cells (2.5 × 10⁶ cells) were treated as above, except that a CHAPS-based buffer was used for cellular lysis (10 mM HEPES [pH 7.4], 150 mM NaCl, 1% [w/v] CHAPS, complete protease inhibitor tablet). Lysates were precleared as above, protein concentrations were determined by Bradford assay (Bio-Rad Protein Assay), and 500 μg of precleared lysate was incubated overnight with the 6A7 anti-BAX antibody (6 μg/ml) (Santa Cruz Biotechnology Inc.) at 4°C, followed by addition of pre-equilibrated protein A/G sepharose beads for 2 hours. The beads were pelleted, washed 5 times in CHAPS lysis buffer at 4°C, and protein eluted by heating at 90°C for 10 minutes in LDS/DTT loading buffer. The immunoprecipitates were subjected to electrophoresis and Western blot analysis using anti-BAX antibody (1:2,000) (N20, Santa Cruz Biotechnology Inc.).

Wright-Giemsa stain.

OCI-AML3 cells (2.5 × 10⁶ cells) were treated with vehicle or BIM SAHBs at the indicated concentrations in Opti-MEM media for 4 hours at 37°C. Cells were collected and washed twice with PBS containing 2% FBS and resuspended at 2.5 × 10⁶ cells/ml for mounting of the cellular suspension (100 μl) onto glass slides by cytospin (Cytospin3, Shandon). The slides were air dried and then stained with Accustain Modified Wright-Giemsa according to the manufacturer’s protocol (Sigma-Aldrich).

OCI-AML3 xenograft and in vivo efficacy study.

Six-week-old female NSG mice were purchased from The Jackson Laboratory and acclimated for 1 week prior to tumor cell inoculation. OCI-AML3-LucNeo cells (5 × 10⁶ cells) were injected subcutaneously with 30% matrigel into the right hind flank. In vivo bioluminescence imaging was conducted after injection of 75 mg/kg of d-luciferin (Promega) using a Xenogen IVIS Spectrum (Caliper Life Sciences). Eight days after inoculation, mice with equivalent disease, as determined by bioluminescence imaging, were divided into 2 groups (n = 8 per arm) and treated with the indicated agent, route, and duration. The diluent was 5% DMSO/D5W for BIM SAHB and 60% Phosal/30% PEG-400/10% ethanol for ABT-263. Tumor burden was monitored by serial bioluminescence imaging and quantified using the Living Images software package (Caliper Life Sciences).

Bim^–/– bone marrow transplantation and in vivo efficacy study.

Rag2^–/–gc^–/– mice (20 weeks old; Taconic) were subjected to irradiation (900 cGy; Gammacell-40 Exactor irradiator, Theratronics) and tail vein infusion of 2 × 10⁶ bone marrow cells harvested from the femurs and tibias of 16-week-old Bim^+/+ or Bim^–/– mice (37). The spleens of Bim^+/+ and Bim^–/– mice (18–20 weeks old) and of mock, Bim^+/+ , and Bim^–/– bone marrow–reconstituted Rag2^–/–gc^–/– mice (12 weeks post transplant) were imaged by micro-ultrasound (Vevo 770 system, VisualSonics) to evaluate splenomegaly. Mice were injected intraperitoneally with Avertin (10 g tribromoethanol and 10 ml tertiary amyl alcohol [Sigma-Aldrich]) at a dose of 0.015 ml/g body weight, followed by removal of abdominal hair with electric clippers and depilatory cream (Nair). Medical ultrasound acoustic gel was applied to the abdomen and ultrasound imaging was performed, followed by calculation of spleen volumes using the Vevo 770 imaging software. To assess for hypergammaglobulinemia, blood was withdrawn from Bim^+/+ and Bim^–/– mice (18–20 weeks old) and Bim^–/– bone marrow–reconstituted Rag2^–/–gc^–/– mice (15 weeks post transplant) by retro-orbital bleed, and serum Ig titers were quantified using the mouse immunoglobin isotyping ELISA system (BD Pharmingen) according to the manufacturer’s protocol. Serum samples were diluted by 1:10,000–1:100,000 depending on the isotype, such that absorbance readings were measured in the linear range of Ig controls. For in vivo efficacy analysis, 3 cohorts (n = 4) of post transplant day–matched, Bim^–/– bone marrow–reconstituted, Rag2^–/–gc^–/– mice were treated with once daily tail vein injections of vehicle (5% DMSO/D5W), BIM SAHB_A (10 mg/kg), or BIM SAHB_A(R153D) (10 mg/kg) for 5 days. Twenty-four hours after the final injection, the mice were euthanized by CO₂ asphyxiation. Animals and excised spleens were weighed, and tissues were harvested for fixation in 10% buffered formalin (Fisher Scientific). Paraffin-embedded sections (5 μm) were stained with H&E or subjected to immunohistochemistry (Specialized Histopathology Service, Brigham and Women’s Hospital). For immunohistochemistry, slides were soaked in xylene, passed through graded alcohols, immersed in distilled water, and treated with either EDTA or 10 mM citrate (pH 6.0) (Zymed Laboratories) in a steam pressure cooker (Decloaking Chamber, BioCare Medical) according to the manufacturer’s instructions and then washed in distilled water. All subsequent steps were performed at room temperature in a hydrated chamber. Slides were treated with Peroxidase Block (DakoCytomation) for 5 minutes to quench endogenous peroxidase activity and then primary monoclonal B220 antibody (1:200) (DakoCytomation) applied in DakoCytomation diluent for 1 hour. Slides were washed in 50 mM Tris-Cl (pH 7.4), and a rabbit anti-rat antibody was applied for 30 minutes. To develop the slides, an anti-rabbit HRP-conjugated antibody (Envision Detection Kit, DakoCytomation) was applied for 30 minutes, followed by washing, immunoperoxidase staining with the Diaminobenzidine Chromogen Kit (DakoCytomation), and hematoxylin counterstaining. TUNEL staining was performed using the ApopTag Peroxidase ISOL Apoptosis Detection Kit (Millipore) in accordance with the manufacturer’s protocol. Slides were analyzed using an Olympus BX50 microscope equipped with an Olympus Qcolor5 digital photo camera and images downloaded to Adobe Photoshop software. The percentage of TUNEL-positive lymphocytes within the infiltrates was calculated by counting approximately 6,000 lymphocytes per 5-μm section per animal in blinded fashion using pictures of random ×40 fields ([TUNEL-positive cells/total cells counted] × 100). The number of TUNEL-positive cells per ×20 field of infiltrate-free tissue (~40 fields of kidney tissue and ~60 fields of liver tissue within a 5-μm section per animal) was counted to determine background TUNEL positivity within the normal parenchyma. Statistical significance was calculated using the 2-tailed Student’s t test. For comparative cytotoxicity analysis, 3 cohorts (n = 4 per arm) of 15- to 16-week-old WT B6/129 mice (Taconic), reflecting the same mixed background upon which the Bim^–/– mice were generated (37), were treated with vehicle, BIM SAHB_A, or BIM SAHB_A(R153D) as above. Prior to the study and 24 hours after the last injection, blood was collected by retro-orbital bleed and subjected to cell counting (ADVIA 2120i, Siemens Healthcare Diagnostics Inc.). In addition, peripheral blood and splenocytes were analyzed by flow cytometry (FACSCalibur, Becton Dickinson) using the following monoclonal antibodies (BD Pharmingen) conjugated to FITC, PE, or APC: GK1.5 (anti-CD4), 53-6.7 (anti-CD8), and RA3-6B2 (anti-B220). FACS data were analyzed using FlowJo software (Tree Star Inc.).

Statistics.

Statistical significance was determined by 2-tailed Student’s t test for pair-wise comparison of groups. Two-way repeated-measures ANOVA with Bonferroni correction was used for comparisons of 2 or more groups over a time course. P values of less than 0.05 were considered significant.

Study approval.

All animal experiments were approved by and performed in accordance with the guidelines and regulations set forth by the Institutional Animal Care and Use Committee of the Dana-Farber Cancer Institute.