Activating ESR1 Mutations Differentially Affect the Efficacy of ER Antagonists

Abstract

Recent studies have identified somatic ESR1 mutations in patients with metastatic breast cancer and found some of them to promote estrogen-independent activation of the receptor. The degree to which all recurrent mutants can drive estrogen-independent activities and reduced sensitivity to ER antagonists like fulvestrant is not established. In this report, we characterize the spectrum of ESR1 mutations from more than 900 patients. ESR1 mutations were detected in 10%, with D538G being the most frequent (36%), followed by Y537S (14%). Several novel, activating mutations were also detected (e.g., L469V, V422del, and Y537D). Although many mutations lead to constitutive activity and reduced sensitivity to ER antagonists, only select mutants such as Y537S caused a magnitude of change associated with fulvestrant resistance in vivo Correspondingly, tumors driven by Y537S, but not D5358G, E380Q, or S463P, were less effectively inhibited by fulvestrant than more potent and bioavailable antagonists, including AZD9496. These data point to a need for antagonists with optimal pharmacokinetic properties to realize clinical efficacy against certain ESR1 mutants.Significance: A diversity of activating ESR1 mutations exist, only some of which confer resistance to existing ER antagonists that might be overcome by next-generation inhibitors such as AZD9496. Cancer Discov; 7(3); 277-87. ©2016 AACR.This article is highlighted in the In This Issue feature, p. 235.

Figure 1. Newly detected ESR1 mutations exhibit a range of estrogen-independent activities

(A) Diagram of ESR1 Ligand Binding Domain with somatic mutations identified from 929 breast tumors analyzed. Height of the circles correlates to the number of cases with that specific mutation. The color codes of the circles are as follow: green for missense mutations, red for truncating mutations (Nonsense, Nonstop, Frameshift deletion, Frameshift insertion, Splice site) and black for in frame mutations. (B) Activation of ER reporter gene. ER+ MCF7 cells were transfected with empty vector, HA-ERα wild type (WT) or indicated ESR1 mutation, ERE-luciferase and Renilla luciferase reporter constructs in hormone-depleted medium with 10 nM of E2 added for 24 hours where indicated. Firefly luciferase activity shows increased activity in absence of E2 or presence of E2 for certain mutations. Graphs were plotted with the mean ± SD of three biological replicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. (C) Activation of ER target genes. MCF7 cells were transfected with empty vector, HA-ERα WT or mutant in hormone-depleted medium and harvested 48 hours post-transfection for qRT-PCR analysis. Bars represent mean ± SD of three technical replicates normalized to actin (ACTB) expression. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. (D) Activation of ER phosphorylation in MCF7 cells. Expression level of the mutant HA-tagged ERs and their relative phosphorylation status at Serine118 and Serine 167, treated with or without 10 nM E2 for 24 hours by immunoblot analysis with specific antibodies as indicated. (E) Activation of hormone independent cell proliferation. Doxycycline inducible ER mutant receptors (E380Q, S463P, L536R and Y537S) expressing MCF7 cells were seeded in 96-well plates in hormone-depleted medium with or without the addition of doxycycline and proliferation was assayed using resazurin regeant. Data show sufficiency of these 4 mutants to promote cell growth in the absence of estradiol. Each point in the graph represented mean ± SD of 6 technical replicates. (F) Binding of the SRC3 NRD to Y537S, D538G, E380Q or S463P ERα LBD in the absence or presence of E2. SRC3 was titrated into a fixed amount of ER-LBD-biotin and time-resolved Förster resonance energy transfer (tr-FRET) indicated that only Y537S and D538G were able to recruit SRC3 in the absence of E2 but not E380Q and S463P. LBD, ligand-binding domain.

Figure 2. Efficacy of SERDs against ER mutants

Inhibition of WT and mutant driven ERE-luciferase (A–B) and proliferation (C–D) by fulvestrant (ICI). Doxycycline-induced WT and mutant ER expressing MCF7 cells were treated with various doses of each antagonist in regular medium, demonstrating that the more active mutants required higher level of antagonists for complete inhibition. Graphs were plotted with the mean ± SD of 2 technical or 6 biological replicates respectively. (Notes: SD values for Figure 3A–B are too low for error bars to be visible in the graphs) (E) Detection of PR levels in MCF7 Y537S CRISPR knock-in cell by Western blot showed elevated PR levels in the Y537S knock-in cells in comparison to the parental line, indicating the expression of Y537S ER mutant receptors. (F) Proliferation assays of parental and Y537S CRISPR knock-in cell lines performed in hormone-depleted or regular media whereby Y537S knock-in cells showed estrogen-independent cell proliferation. (G–H) Proliferation assays of parental MCF7 and MCF7 Y537S CRISPR knock-in cells treated with various doses of fulvestrant, AZD9496 or 4OHT, demonstrated significantly higher doses are required to cause growth inhibition of Y537S expressing cells by ER antagonists. Cell confluency was measured using the IncuCyte Zoom standard software over a few days. Graphs were plotted with the mean ± SD of three biological replicates.

Figure 3. AZD9496 demonstrates superior anti-tumor effects on ER mutant expressing xenografts

Mice bearing MCF7 inducible HA-ER WT (A), Y537S (B), D538G (C), E380Q (D) or S463P (E) tumors were randomly assigned to treatment groups of either 15 mg/kg of AZD9496, daily orally or 200 mg/kg of Fulvestrant twice weekly, s.c. Tumors treated with AZD9496 showed greater growth inhibition as compared to those treated with Fulvestrant. The result was presented as average tumor volume measured for each group ± SD (n = 10 mice/group). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. (F–G) Scatter plots of the volumes of tumors expressing WT, Y537S or D538G treated with either fulvestrant (ICI) or AZD9496 taken at the end of the xenograft studies shown in A–C. T-tests comparing the volumes of mutant tumors to that of the WT indicated significant resistance of Y537S mutant tumors to fulvestrant treatment. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. (H) Patient-derived xenograft (PDX), CTC-714, with D538G mutation, showed greater growth arrest with 5mg/kg of AZD9496. The result was presented as average tumor volume measured for each group ± SEM. (I) Quantitative PCR detection of various ER target genes of AZD9496 treated CTC-714, the D538G PDX model, showed significant reduction in the transcript levels of GREB1, STC2, TFF1 and PGR when compared to the vehicle, indicating inhibition of ER signaling pathway by AZD9496. Graphs were plotted with the mean ± SD of three technical replicates.