Identification and Characterization of a Small Molecule Bcl-2 Functional Converter

Bcl-2-family proteins are evolutionarily conserved regulators of apoptosis. All members possess at least one of the four conserved motifs called Bcl-2 homology (BH) domains. The anti-apoptotic Bcl-2 family members consist of Bcl-2, Bcl-xL, Mcl-1, AI-1, Bcl-B and contain BH domains 1–4. There are two groups of pro-apoptotic Bcl-2 members; the BH1–3 containing group such as Bax and Bak, or the BH3 only group such as Bid, Bim, and Puma (1–6). The BH3 only pro-apoptotic members function by activating multi-domain pro-apoptotic proteins like Bax and Bak (7–9). The anti-apoptotic Bcl-2 members use the hydrophobic surface binding pocket formed by the BH 1–3 domains to neutralize these pro-apoptotic proteins. It is the ratio of the levels of these anti-apoptotic and pro-apoptotic proteins that cancer cells exploit for their survival advantage (6). Specifically, Bcl-2 overexpression leads to cancer progression and resistance to cancer therapies (10–13).

The Bcl-2 family of proteins is being pursued with multiple approaches to achieve therapeutic outcomes (14–19). Small molecules like ABT-737 or ABT-199 are being developed to either inhibit multiple anti-apoptotic members or a single anti-apoptotic member (20–22). These inhibitors are designed to bind to the hydrophobic groove formed by the BH1–3 domains of anti-apoptotic members (23–25). NuBCP-9, a 9 amino acid peptide derived from nuclear receptor Nur77 binds the loop domain in Bcl-2, resulting in exposure of its BH3 domain and conversion of Bcl-2 into a pro-death protein (26–41). We hypothesized that small molecules targeting the Bcl-2 that switch its function to a pro-death protein can be developed to effectively treat cancers that express Bcl-2.

To test this hypothesis, we screened the Chembridge DIVERset® library using MDA-MB-231 cells stably expressing Bcl-2 along with their empty vector controls that have very low expression of Bcl-2 in a viability assay. Compounds were shortlisted on the basis of their ability to inhibit viability of Bcl-2–expressing cancer cells, compared with their control cells. These shortlisted compounds were further probed for their ability to induce a conformational change in Bcl-2 and promote Bcl-2–dependent apoptosis. One such compound, termed here as Bcl-2 functional converter (BFC)1108 (5-chloro-N-(2-ethoxyphenyl)-2-[(4-methoxybenzyol)amino]benzamide), induced a conformational change in Bcl-2, resulting in Bcl-2–dependent cell death. The anticancer effects of BFC1108 were further tested in a breast cancer xenograft mouse model using MDA-MB-231 cells expressing Bcl-2 and in a breast cancer lung metastatic model.

MDA-MB-231 (ATCC, catalog no. CRM-HTB-26, RRID:CVCL_0062), MCF-7 (ATCC, catalog no. HTB-22, RRID:CVCL_0031), HepG2 (ATCC, catalog no. HB-8065, RRID:CVCL_0027), ZR-75-1 (ATCC, catalog no. CRL-1500, RRID:CVCL_0588), A375 (ATCC, catalog no. CRL-1619, RRID:CVCL_0132), mouse embryonic fibroblasts (MEF; ATCC, catalog no. CRL-2991, RRID:CVCL_U630), and H460 (ATCC, catalog no. HTB-177, RRID:CVCL_0459) were cultured in DMEM with l-glutamine (Corning, catalog no. 10-013-CV) supplemented with 10% FBS (tissue culture biologicals, catalog no. 101 HI) with 10,000 U/mL penicillin streptomycin (Corning, catalog no. 30-002-Cl) in a humidified chamber with 5% CO₂. LNCaP cells (ATCC, catalog no. CRL-1740, RRID:CVCL_0395) were cultured in RPMI1640 (Corning, catalog no. 10-040-CV). MCF-10A cells (ATCC, catalog no. CRL-10317, RRID:CVCL_0598) were cultured in DMEM/F-12 (Gibco, catalog no. 11-320-033) with appropriate supplements. All cells were passaged once every 3 days.

MDA-MB-231 and MCF-7 breast cancer cell lines stably overexpressing Bcl-2 were generated via electroporation and subsequent clonal expansion in G418-containing media. MDA-MB-231 cell clones with sufficiently high expression of Bcl-2 (MDA-MB-231/Bcl-2) were determined by Western blotting. These were utilized for chemical library screening along with their vector controls (MDA-MB-231/Vector). Similarly, MCF-7 cell clones were derived and were utilized for characterization of lead molecules.

The DIVERset® chemical library was purchased from ChemBridge Corporation. All compound dilutions were made in DMSO. BFC1108 (5-chloro-N-(2-ethoxyphenyl)-2-[(4-methoxybenzyol)amino]benzamide) was custom synthesized by ChemBridge Corporation.

CellTiter-Glo assay (Promega, catalog no. G7573) was used for assessing viability (42, 43). Annexin V (eBioscience, catalog no. 88-8008) was used to assess apoptosis (44, 45).

Bcl-2 BH3 antibody (catalog no. AP1303a) was obtained from Abgent. Bcl-2 antibody (catalog no. 13-8800) was obtained from InvitrogenCA. Ki-67 and hematoxylin and eosin (H&E) was obtained from Cell Signaling Technology. JC-1 dye (catalog no. T3168) was obtained from Roche. Western Blotting, immunofluorescence, and IHC experiments were done as described in (refs. 26–28, 35).

For Bcl-2 knockdown experiments, MDA-MB-468 cells were seeded at 4,000 cells per well in a 96-well black plate. The day after seeding, cells were transfected using Dharmafect reagent (catalog no. T-2001-03) with Bcl-2 siRNA (25 ng; Dharmacon Research) or 25 ng control luciferase siRNA (catalog no. D-002050-01, Dharmacon Research). After 72 hours of transfection, cells were treated with indicated compounds for 48 hours and viability was assayed with CellTiter-Glo Luminescent Cell Viability Assay (Promega, catalog no. G7573).

GST-Bcl-2 and GST only fragments (60 µg) were incubated with 0.1% DMSO or BFC1108 (50 µmol/L) for 1 hour and proteolyzed by 3 µg/mL of trypsin for 1, 2, 4, and 8 minutes. Reactions were quenched by the addition of Laemmli buffer. The resulting protease cleavage fragments were resolved on a 12% SDS-PAGE gel and visualized by Coomassie blue staining.

Bcl-2 protein was mixed with elution buffer and SYPRO Orange dye and then mixed with BFC1108 at varying concentrations. Samples were loaded into PCR plate and covered. A 7500 Fast PCR system (Applied Biosystems) was used with detection set to ROX and initial 2-minute incubation at 25°C followed by step and hold 0.5°C for 30 seconds up until 95°C. Melting temperature was calculated using Applied Biosystems software (46).

All animal experiments were performed as per approved protocol no. 4452 by the Institutional Animal Care and Use Committee (IACUC) at Oregon State University (Corvallis, OR). For the mammary fat pad xenograft study, 5-week-old NOD.SCID (RRID:IMSR_ JAX:001303) mice were procured from Jackson labs. About 10⁶ cells were injected into both the lower flanks of mammary fat pad in PBS. Palpable tumors were detected in 4 weeks after which mice were treated with BFC1108 (100 mg/kg) twice a week till end of the study (N = 8 for vehicle and BFC1108 treatments). Tumor mass was measured using digital calipers twice a week along with body weight. At the end of the study, mice were euthanized and tumors were collected for IHC analysis.

For the lung metastatic study, 5-week-old Nude mice (RRID:IMSR_JAX:002019) were procured from Jackson Labs. About 200,000 LMD231-Luc2 cells in PBS were injected into the tail vein. Lung metastasis was detected by 2 weeks. Treatments with BFC1108 at 100 mg/kg, four times a week, were started a day after the cells were injected (N = 8 for vehicle and N = 9 for BFC1108 treatment). Body weight was measured twice a week. Bioluminescence was measured once a week while the mice are under anesthesia. At the end of the study, mice were euthanized and lung tissues were dissected and fixed for IHC analysis.

The statistical significance of difference between groups was analyzed by two-sided unpaired nonparametric Student t test with Mann–Whitney test using PRISM software version 9. Results were considered significant at P < 0.05.

The data generated in this study are available within the article and its Supplementary Data.

To identify small molecules that induce apoptosis in Bcl-2–expressing cancer cells, MDA-MB-231 triple-negative breast cancer (TNBC) cells stably expressing Bcl-2 (MDA-MB-231/Bcl-2) along with its vector controls (MDA-MB-231/Vector) were generated. These cells were used in a viability assay to screen ChemBridge DIVERset® library of compounds. Promising hits were shortlisted and retested for Bcl-2–dependent growth inhibitory effects. A lead compound, termed here as BFC1108 (Fig. 1A) was identified, which dose dependently reduced the viability of Bcl-2–expressing cells (MDA-MB-231/Bcl-2) compared with their vector controls (Fig. 1B). We tested to confirm whether this phenomenon is also observed in other cancer cell types. MCF-7 breast cancer cells with high Bcl-2 expression were more sensitive to BFC1108 compared with their vector controls (Fig. 1C). Similarly, Jurkat T-cell lymphocytes with high Bcl-2 expression were more responsive to BFC1108 with significant reduction in viability compared with their vector controls (Fig. 1D). We next tested the Bcl-2–dependent effect of BFC1108 in a second TNBC MDA-MB-468 cell line. Bcl-2 expression was suppressed in MDA-MB-468 cells by siRNA. BFC1108 reduced the viability of control siRNA-transfected cells in a dose-dependent manner. In contrast, Bcl-2 knocked down MDA-MB-468 cells were resistant to BFC1108 (Fig. 1E). Thus, the effects of BFC1108 are dependent on Bcl-2 expression in distinct cancer cells.

We next determined the effect of BFC1108 on clonogenic survival of MDA-MB-231 cells. Upon exposure to BFC1108, MDA-MB-231/Bcl-2 cells formed fewer colonies compared with their vector controls (Fig. 2A and B). Treatment of MDA-MB-231/Bcl-2 cells with 10 µmol/L BFC1108 resulted in a 50% reduction in their colony-forming ability. This effect was not observed in MDA-MB-231/Vector control cells with low Bcl-2 expression. We next tested whether the effect of BFC1108 on cell viability was due to apoptosis. For this, MDA-MB-231/Bcl-2 and their vector control cells were exposed to 10 µmol/L BFC1108 for 48 hours and subsequently stained for annexin V to detect early stages of apoptosis. Similar to the results obtained in viability assays, there was a significant increase in apoptosis of MDA-MB-231/Bcl-2 cells compared with control cells, indicating BFC1108-induced apoptosis depends on Bcl-2 expression (Fig. 2C; Supplementary Fig. S1). The reduction in viability induced by BFC1108 correlated with increased Bcl-2 expression in breast cancer cells (Fig. 2D). Specifically, BT474 cells with higher levels of Bcl-2 expression exhibited a more pronounced response compared with HCC1395 cells, which have lower Bcl-2 expression. Similarly, BT549 and HCC1806, two TNBC cell lines expressing Bcl-2, were responsive to BFC1108 (Fig. 2D). BFC1108 was also effective against a variety of cancer cell types, including human ZR-75-1 hormone receptor–positive breast cancer cells, LNCaP prostate cancer cells, HepG2 hepatocellular carcinoma cells, H460 lung cancer cells, and A375 melanoma cells (Fig. 2E). Notably, BFC1108 exhibited minimal effects on MCF10A nontransformed breast epithelial cells (Fig. 2E). We next tested the effect of BFC1108 on colony-forming ability of transformed MEFs. Wild-type (WT) MEFs formed fewer colonies upon exposure to BFC1108 (10 µmol/L) compared with their vehicle controls (Fig. 3A and B). Bcl-2^−/− MEF cells did not respond to BFC1108, indicating a role for Bcl-2 in the reduction of colony-forming units in WT MEFs.

To test whether BFC1108 triggers an intrinsic apoptotic pathway, we determined the requirement of Bax and/or Bak in MEFs. BFC1108 reduced the viability of WT (Bax^+/+ and Bak^+/+), Bax knockout (Bax^−/− Bak^+/+), and Bak knockout (Bak^−/− Bax^+/+) MEFs. Notably, Bax and Bak double knockout (Bax^−/− and Bak^−/−) MEFs were unaffected by BFC1108 treatment (Fig. 3C). This observation suggests that Bax and/or Bak are essential for the Bcl-2–dependent induction of apoptosis by BFC1108. Furthermore, H460 lung cancer cells which have high levels of endogenous Bcl-2 expression, when treated with BFC1108, exhibited a decrease in mitochondrial membrane potential upon staining with JC-1 dye, indicating the collapse of the mitochondrial outer membrane (Fig. 3D). Thus, the anticancer effects induced by BFC1108 are mediated through an intrinsic mitochondrial cell death pathway.

Nuclear receptor Nur77/TR3 upon translocation from the nucleus or NuBCP-9, a 9 amino acid peptide derived from Nur77, interacts with the loop domain of Bcl-2 and induces Bcl-2–dependent cell death (27, 28, 33, 35). To assess the potential interaction of BFC1108 with Bcl-2, we examined its impact on trypsin-mediated proteolysis of Bcl-2. Our results revealed that BFC1108 increased the proteolysis of the Bcl-2 loop domain compared with the vehicle (Fig. 4A), implying an interaction with the loop domain. To further assess the interaction between Bcl-2 and BFC1108, we purified the Bcl-2 protein and conducted a thermal shift assay. The Bcl-2 full-length protein was incubated with increasing concentrations of BFC1108 or vehicle control and its melting temperature was calculated (46). Bcl-2 full-length protein in the presence of vehicle (DMSO) had a melting point of 73°C and this was stabilized to above 82°C in the presence of BFC1108 (Fig. 4B). To rule out an interaction with the fused Z-tag used for Bcl-2 protein purification, a Z-Tag-only control was employed, showing no change in the melting point temperature (Tm) in the presence of BFC1108. These findings suggest an interaction between BFC1108 and Bcl-2. To explore whether this interaction is specifically with the Bcl-2 loop, the loop domain was expressed as a fusion with the Z-tag, purified, and then subjected to a thermal shift assay after cleaving the Z-tag. The Bcl-2 loop domain was incubated with increasing concentrations of BFC1108, which increased the Bcl-2 loop melting point from 29°C to 78°C, at 1 µmol/L, suggesting an interaction. This melting point remained over 78°C with increased BFC1108 concentrations (Fig. 4C).

NuBCP-9 binding leads to a conformational change in Bcl-2, exposing its BH3 domain and leading to the conversion of Bcl-2 from an anti-apoptotic to a pro-apoptotic protein. This change in conformation of Bcl-2 can be detected by a Bcl-2-BH3 domain antibody. Upon treatment with BFC1108, MDA-MB-231/Bcl-2 cells exhibited the exposure of BH3 domain in Bcl-2, as detected by flow cytometry (Fig. 4D). Similarly, H460 lung cancer cells with high endogenous Bcl-2 expression also showed BH3 domain exposure following treatment with BFC1108 (Fig. 4D) and exhibited responsiveness to BFC1108 (Fig. 2E). Taken together, these findings indicate that BFC1108 likely interacts with the Bcl-2 loop domain, inducing a Bcl-2 conformation change and promoting cell death in Bcl-2–expressing cancer cells.

We compared the anticancer effects of Bcl-2 functional conversion by BFC1108 to Bcl-2 inhibition through the use of a BH3 mimetic, ABT199 (47). TNBC cell lines that express Bcl-2 (HCC1187, BT549, MDA-MB-157, MDA-MB-468) were treated for 72 hours with BFC1108 or ABT199 and cell death was assessed (Fig. 5). Notably, BFC1108 induced a more substantial level of cell death in HCC1187, MDA-MB-157, and MDA-MB-468 compared with ABT199 (Fig. 5). These findings suggest that Bcl-2 functional conversion may represent a preferable therapeutic strategy in Bcl-2–expressing cancers.

To evaluate the in vivo effectiveness of BFC1108, mice bearing orthotopic breast cancer xenografts originating from high Bcl-2–expressing MDA-MB-231 cells (MDA-MB-231/Bcl-2) underwent a 4-week treatment administered via the intraperitoneal route. The administration of BFC1108 led to a significant reduction in the growth of MDA-MB-231/Bcl-2 xenograft tumors (Fig. 6A), without causing a substantial change in body weight of mice (Fig. 6B). We examined tumor sections for apoptosis by staining for TUNEL and active caspase-3. The detection of TUNEL positive cells provided confirmation of BFC1108-induced apoptosis in vivo (Fig. 6C). In addition, we investigated these tumor sections to detect changes in Bcl-2 conformation (Fig. 6D). In the tumor sections from mice treated with BFC1108, we observed an exposure of the Bcl-2 BH3 domain (Fig. 6D). To establish a connection between the altered Bcl-2 conformation, the activation of Bax, and the induction of apoptosis, we probed the tumor sections with antibodies that recognize activated Bax and cleaved (active) caspase-3 (Fig. 6E and F). Bax activation was detected in tumor sections from mice treated with BFC1108, along with the presence of activated caspase-3. Thus, there is a correlation between exposure of BH3 domain in Bcl-2, activation of Bax, and the induction of apoptosis in vivo.

We next tested the efficacy of BFC1108 in suppressing breast cancer lung metastasis. LMD231 cells are lung metastatic cells derived from TNBC MDA-MB-231 cells (48, 49). LMD231 cells were engineered to stably express luciferase (LMD231-Luc) to aid in the detection of metastasis. Following injection through the tail vein, LMD231-Luc cells formed lung metastases within 2 weeks. Mice were treated with 100 mg/kg of BFC1108, four times a week via intraperitoneal injections. After 7 weeks, BFC1108 significantly suppressed the growth of lung metastasis (Fig. 7A and B; Supplementary Fig. S2). There was no significant change in body weight of mice during the course of the treatment (Fig. 7C). The reduction in lung metastasis was independently confirmed through histologic analysis (Fig. 7D). IHC of the lung tissues from vehicle-treated animals revealed Ki-67 positivity, indicating presence of actively proliferating tumor cells. In contrast, Ki-67 staining was dramatically reduced in lung tissues from BFC1108-treated mice (Fig. 7E).

In this study, we identified BFC1108, a small molecule Bcl-2 functional converter, that suppresses primary and metastatic breast cancers in mouse xenograft models. Bcl-2 and its family members are prime molecular targets for developing new cancer therapeutics (3, 6, 14, 17, 18). There have been several efforts in development of Bcl-2 inhibitors for solid cancers, but not with very encouraging results in clinical trials (18, 22, 24). Our study provides a proof of concept for the therapeutic targeting of Bcl-2 through the use of small molecules that induce the Bcl-2 killer conformation.

BFC1108 induces apoptosis in cancer cells expressing Bcl-2. There is also a correlation between the extent of Bcl-2 expression and responsiveness of cancer cells to BFC1108. This may be due to the interaction of BFC1108 with the loop domain of Bcl-2 resulting in exposure of its BH3 domain and conversion into a pro-apoptotic protein. The interaction between BFC1108 and Bcl-2 was demonstrated through thermal shift and limited proteolysis assays. Similar to NuBCP-9, BFC1108-induced apoptosis is also dependent on Bax and/or Bak. More importantly, BFC1108 suppressed xenograft growth of human breast cancer cells with high Bcl-2 expression in mice, without any systemic toxicity. The apoptotic effect of BFC1108 is not inhibited, but rather potentiated, by Bcl-2 expression in multiple cancer cell lines. BFC1108 also suppressed the growth of breast cancer lung metastasis, underscoring its potential for treating metastatic disease. Thus, BFC1108 is a promising clinical candidate for treating both primary and metastatic cancers expressing Bcl-2. Notably, BFC1108 exhibits significantly better efficacy in TNBC cells expressing Bcl-2 compared with a Bcl-2 inhibitor, ABT199.

The ability to target and kill Bcl-2–expressing cells is exceptionally promising, as Bcl-2 is overexpressed in several cancers including TNBC. Selective targeting of Bcl-2 through an unexplored region and elimination of cancer cells expressing Bcl-2 open up new therapeutic possibilities. “Bcl-2 functional converters” are envisioned to function as targeted cancer therapeutics, exploiting the cancer-specific overexpression of Bcl-2.

M.C. Pearce reports a patent. C.V. Löhr reports grants from US Army Medical Research and Material Command, US Army Medical Research and Material Command, and NIH during the conduct of the study. E.F. O'Donnell reports grants from Oregon State University during the conduct of the study. S.K. Kolluri reports grants from US Army Medical Research and Material Command, NCI, and National Institute of Food and Agriculture during the conduct of the study; in addition, S.K. Kolluri has a patent to Bcl-2 functional converters. pending and issued. No disclosures were reported by the other authors.

P.R. Kopparapu: Data curation, formal analysis, validation, investigation, visualization, methodology, writing-original draft. M.C. Pearce: Resources, validation, methodology. C.V. Löhr: Resources, data curation, formal analysis, investigation, methodology. C. Duong: Investigation. H.S. Jang: Investigation, methodology. S. Tyavanagimatt: Resources, investigation, methodology. E.F. O'Donnell 3rd: Investigation, methodology. H. Nakshatri: Resources, investigation, visualization, methodology. S.K. Kolluri: Conceptualization, resources, funding acquisition, investigation, methodology, writing-review and editing.

Sincere thanks to Drs. Xiao-kun Zhang and Arnold Satterthwait for numerous discussions that formed the basis for this research and providing material support. We express our gratitude to current and former members of the Kolluri laboratory for substantial technical, and experimental support, as well as material contributions. Special thanks to Rhand S. Wood, Jacob Kreitzer, Amanda Galbreath, and Kyla M. Guertin for providing excellent laboratory assistance, Pranav J. Kolluri for article editing, Dr. Sarah Clark for substantial help with bacterial expression and purification of Bcl-2 proteins and Dr. Joseph G. Christison, Oregon State University Advantage for help with collaborative translational research agreements. Our appreciation extends to Dr. Helen Diggs, the former director of Laboratory Animal Resource Center (LARC), the dedicated technical and support staff at LARC, Dr. Chrissa Kioussi, and the IACUC review committee. This work was supported by grants to S.K. Kolluri from The U.S. Department of Agriculture's (USDA) National Institute of Food and Agriculture (NIFA) Multistate project ORE00108E, the U.S. Army Medical Research and Material Command (W81XWH-08-1-0600 and W81XWH-12-1-0069) and National Cancer Institute of the National Institutes of Health under award number CA249627. The sponsors have no role in study design, conclusions, or submission of this article.