Covalent ERα Antagonist H3B-6545 Demonstrates Encouraging Preclinical Activity in Therapy-Resistant Breast Cancer

Abstract

Nearly 30% of patients with relapsed breast cancer present activating mutations in estrogen receptor alpha (ERα) that confer partial resistance to existing endocrine-based therapies. We previously reported the development of H3B-5942, a covalent ERα antagonist that engages cysteine-530 (C530) to achieve potency against both wild-type (ERαWT) and mutant ERα (ERαMUT). Anticipating that the emergence of C530 mutations could promote resistance to H3B-5942, we applied structure-based drug design to improve the potency of the core scaffold to further enhance the antagonistic activity in addition to covalent engagement. This effort led to the development of the clinical candidate H3B-6545, a covalent antagonist that is potent against both ERαWT/MUT, and maintains potency even in the context of ERα C530 mutations. H3B-6545 demonstrates significant activity and superiority over standard-of-care fulvestrant across a panel of ERαWT and ERαMUT palbociclib sensitive and resistant models. In summary, the compelling preclinical activity of H3B-6545 supports its further development for the potential treatment of endocrine therapy-resistant ERα+ breast cancer harboring wild-type or mutant ESR1, as demonstrated by the ongoing clinical trials (NCT03250676, NCT04568902, NCT04288089).

Summary: H3B-6545 is an ERα covalent antagonist that exhibits encouraging preclinical activity against CDK4/6i naïve and resistant ERαWT and ERαMUT tumors.

©2022 The Authors; Published by the American Association for Cancer Research.

Figures

Figure 1.

Identification of H3B-6545, a covalent antagonist with improved potency. A, Structural comparison between H3B-5942 and H3B-6545. B, X-ray crystal structure of H3B-5942 bound in the ligand-binding domain of ERαY537S. H3B-5942 is rendered as a stick model. The hydrophobic pocket is indicated by the semitransparent surface. H-bonding interactions are rendered as a green dashed line. Other key residues are labeled. C, X-ray crystal structure of H3B-6545 bound in the ligand-binding domain of ERαY537S. The protein is rendered as a cartoon model in gray while the ligand is depicted in yellow as a stick model. The 2Fo-Fc electron density map is illustrated as a blue mesh and contoured at 1 sigma. The co-complex structure was crystallized as a dimer in the asymmetric unit. The protein and ligand conformations are identical in chain A and B with the exception of the tail acrylamide of the ligand, which is more flexible. D, Comparison of the cocrystal structures of H3B-5942 and H3B-6545 in complex with ERα. The structure of H3B-5942 and H3B-6545 in the cocrystal structure are shown as yellow and magenta stick models, respectively. ERα conformation in chain B of the H3B-6545 cocrystal structure is shown in cyan cartoon model and ERα conformation in chain B of H3B-5942 cocrystal structure shown in gray cartoon model.

Figure 2.

H3B-6545 is not solely dependent on covalency for anti-ERα activity. A, Structures of H3B-9224 and H3B-9709, the saturated analogs of H3B-5942 and H3B-6545, respectively. B, Jump dilution experiment to evaluate reversibility of binding of E2 and various inhibitors to ERαWT (yellow bars) and ERαY537S (green bars). SoC agents 4-OHT and fulvestrant, covalent antagonists H3B-5942 and H3B-6545, and saturated analogs H3B-9224 and H3B-9709 were tested in the assay. Left, ERα was incubated with saturating concentrations of binders followed by 20-fold dilution. A 15-fold final molar excess (7.5 nmol/L) of 3H-E2:binder was then used to titrate the number of ERα sites that were no longer bound to the initial binder after dilution. The assay was performed as described previously (13). Right, The assay was performed similarly except that saturating concentrations of 3H-H3B-9709 were used followed by dilution and incubation with an excess of unlabeled E2 to titrate unbound ERα. C, Kinetics of covalent engagement monitored by 3H-protein labeling for ERαWT and ERαY537S. ERα was incubated with 3H-H3B-6545 and 3H-H3B-9709 at room temperature for various times followed by protein denaturation and precipitation. The protein pellet was resolubilized and 3H-labeling was quantified by liquid scintillation counting. kobs values of (1.2 ± 0.2) × 10−3 second−1 and (1.7 ± 0.3) × 10−3 second−1 were obtained for the labeling of ERαWT and ERαY537S by 3H-H3B-6545, respectively. D, Western blot analysis of ERα expression following a 24-hour treatment with the indicated compounds at 100 nmol/L in the MCF7 parental line (MCF7-Parental) and the ST941 PDX model derived cell line (PDX-ERαY537S/WT). GAPDH was used as loading control. E, Representative plots showing dose-dependent suppression in GREB1 expression in ERαWT (left) and ERαY537S (right) expressing MCF7 cells following a 6-day treatment with H3B-6545 (green solid) and H3B-9709 (green dotted). F, Representative graphs showing a dose-dependent decrease in proliferation of ERαWT and ERαY537S MCF7 overexpressing lines with or without C530S mutation in ERα following 6-day treatment with H3B-6545 (left) and H3B-9709 (right).

Figure 3.

H3B-6545 demonstrates potent antagonist activity across a panel of ERαWT/MUT cell lines and SERM activity in bone and uterine tissues. A, IncuCyte-based confluency assessment of ST941 cells following treatment with H3B-6545, 4-OHT, fulvestrant, and GDC-0810 at the indicated concentrations. Data represent the mean confluence ± SD. B, Western blot analysis for ERα confirming H524 L overexpression in MCF7 lines overexpressing ERαWT or ERαY537S with α-tubulin as a loading control (left) and table showing antiproliferative activity (GI50) for indicated compounds in ERαWT and ERαY537S overexpressing MCF7 cells with or without H524 L ESR1 mutation (right). C, X-ray cocrystal structures of 4-OHT (pdb: 3ERT) in complex with ERα, H3B-6545 (pdb: 6OWC) in complex with ERαY537S, H3B-6545 (homology model) in complex with ERαH524L, fulvestrant analog ICI164384 (pdb: 1HJ1) and raloxifene (pdb: 1ERR) (right) in complex with ERαWT. D, Left, femur global BMD measured by DXA in rats that underwent sham or OVX surgery followed by daily treatment with 17β-estradiol (E2, 0.01 mg/kg/day), tamoxifen (TAM, 1 mg/kg/day), and H3B-6545 (3, 10, and 30 mg/kg/day) for 6 weeks. Data represent the percentage of change at week 5/6 after treatment compared with pretreatment (mean ± SEM, N = 20). *, P < 0.05; **, P < 0.01 versus OVX control. D, (Right), the relative uterus weight (vs BW) in rats that underwent sham or OVX surgery followed by daily treatment with E2, TAM, and H3B-6545 for 6 weeks. Data represent the mean ± SEM (N = 20). *, P < 0.05; **, P < 0.01 versus OVX control.

Figure 4.

H3B-6545 demonstrates single-agent antitumor activity in ERαWT and ERαMUT xenograft models. A, Antitumor activity of H3B-6545 compared with fulvestrant in the ERαWT ST986 PDX model. H3B-6545 was administered orally once daily at 100 mg/kg in mice bearing xenograft tumors representing ERαWT breast cancer. Fulvestrant was dosed subcutaneous once weekly at 5 mg/mouse. Data represent the mean tumor volume ± SEM (N = 6 for H3B-6545 treatment arm, N = 6 for all other treatments). Statistical significance assessed on day 60. B, Antitumor activity of H3B-6545 compared to fulvestrant in the ERαWT MCF7 CDX model. H3B-6545 was administered orally once daily at 1, 3, 10, and 30 mg/kg whereas fulvestrant was dosed subcutaneously once weekly at 5 mg/mouse. Data represent the mean tumor volume ± SEM (N = 6). Statistical significance assessed on day 17. C, Antitumor activity of H3B-6545 compared with tamoxifen and fulvestrant in the ERαWT ST1799 PDX model. H3B-6545 was administered orally once daily at 100 mg/kg in mice bearing xenograft tumors representing ERαWT breast cancer. Tamoxifen was dosed subcutaneously three times a week at 1 mg/mouse whereas fulvestrant was dosed subcutaneously once weekly at 5 mg/mouse. Data represent the mean tumor volume ± SEM (N = 5 for H3B-6545 treatment arm, N = 6 for all other treatments). Statistical significance assessed on day 31. D, Antitumor activity of H3B-6545 compared with tamoxifen and fulvestrant in the ERαY537S/WT ST941 PDX model. H3B-6545 was administered orally once daily at 3, 10, 30, and 100 mg/kg in mice bearing xenograft tumors representing ERαY537S/WT breast cancer. Tamoxifen was dosed subcutaneously three times a week at 1 mg/mouse whereas fulvestrant was dosed subcutaneously once weekly at 5 mg/mouse. Data represent the mean tumor volume ± SEM (N = 8 for vehicle, tamoxifen and fulvestrant arms, N = 6 for all H3B-6545 treatments). Statistical significance assessed on day 28. E, Antitumor activity of H3B-6545 compared with fulvestrant in the ERαY537S ST2177 PDX model. H3B-6545 was administered orally once daily at 3, 10, and 30 mg/kg in mice bearing xenograft tumors. Fulvestrant was dosed subcutaneously once weekly at 5 mg/mouse. Data represent the mean tumor volume ± SEM (N = 6 for fulvestrant, N = 7 for all other treatments). Statistical significance assessed on day 56. F, Antitumor activity of H3B-6545 compared with fulvestrant in the ERαY537S ST2056 PDX model. H3B-6545 was administered orally once daily at 100 mg/kg whereas fulvestrant was dosed subcutaneously once weekly at 5 mg/mouse in mice bearing xenograft tumors. Data represent the mean tumor volume ± SEM (N = 8). Statistical significance assessed on day 22. G, Antitumor activity of H3B-6545 in the palbociclib-resistant ST1799-PBR model. H3B-6545 was administered orally once daily at 3, 10, and 30 mg/kg and palbociclib was administered orally once daily at 75 mg/kg. Data represent the mean tumor volume ± SEM (N = 7 for vehicle arm, N = 8 for all other treatments). Statistical significance assessed on day 46. H, Antitumor activity of H3B-6545 in the palbociclib-resistant ST941-PBR model. H3B-6545 was administered orally once daily at 30 mg/kg, palbociclib was administered orally once daily at 50 mg/kg, and fulvestrant was administered subcutaneously once weekly at 5 mg/mouse. Data represent the mean tumor volume ± SEM (N = 6). Statistical significance assessed on day 42. *, P < 0.05 versus vehicle control (multiple unpaired t tests with significance determined using the Holm-Sidak method). All doses and regimens were well tolerated.