ESR1 mutations affect anti-proliferative responses to tamoxifen through enhanced cross-talk with IGF signaling

Abstract

The purpose of this study was to address the role of ESR1 hormone-binding mutations in breast cancer. Soft agar anchorage-independent growth assay, Western blot, ERE reporter transactivation assay, proximity ligation assay (PLA), coimmunoprecipitation assay, silencing assay, digital droplet PCR (ddPCR), Kaplan-Meier analysis, and statistical analysis. It is now generally accepted that estrogen receptor (ESR1) mutations occur frequently in metastatic breast cancers; however, we do not yet know how to best treat these patients. We have modeled the three most frequent hormone-binding ESR1 (HBD-ESR1) mutations (Y537N, Y537S, and D538G) using stable lentiviral transduction in human breast cancer cell lines. Effects on growth were examined in response to hormonal and targeted agents, and mutation-specific changes were studied using microarray and Western blot analysis. We determined that the HBD-ESR1 mutations alter anti-proliferative effects to tamoxifen (Tam), due to cell-intrinsic changes in activation of the insulin-like growth factor receptor (IGF1R) signaling pathway and levels of PIK3R1/PIK3R3. The selective estrogen receptor degrader, fulvestrant, significantly reduced the anchorage-independent growth of ESR1 mutant-expressing cells, while combination treatments with the mTOR inhibitor everolimus, or an inhibitor blocking IGF1R, and the insulin receptor significantly enhanced anti-proliferative responses. Using digital drop (dd) PCR, we identified mutations at high frequencies ranging from 12 % for Y537N, 5 % for Y537S, and 2 % for D538G in archived primary breast tumors from women treated with adjuvant mono-tamoxifen therapy. The HBD-ESR1 mutations were not associated with recurrence-free or overall survival in response in this patient cohort and suggest that knowledge of other cell-intrinsic factors in combination with ESR1 mutation status will be needed determine anti-proliferative responses to Tam.

Keywords: Estrogen Receptor; Everolimus; Fulvestrant; IGF-1; Mutations.

Figure 1. HBD-ESR1 mutations influence Tam response

a, immunoblot analysis of detect ER expression in MCF-7 stable clones. GAPDH was used as loading control. b, Representative photographs of three different well plates captured from soft agar assays showing higher number of large colonies (200μm) in mutant-expressing cells. Experiments were performed in triplicate and error bars indicate SD. *P<0.05; ***P<0.001. d, Cells were plated for soft agar assays, and then treated with vehicle or two doses of Tam (100nM or 1000nM). Percentage of growth reduction with Tam treatment is shown. Experiments were performed in triplicate and error bars indicate SD. **P<0.01; ***P<0.001.

Figure 2. IGF-1 signaling pathway activation in HBD-ESR1 mutants

a, Ingenuity Pathway Analysis (IPA) to identify activation of signaling pathways in mutant MCF-7 vs. ZR-75B. b, Total cellular extracts were analyzed for phosphorylation and expression of ER, IGF1Rβ, IRS-1, and mTOR; GAPDH was used as a loading control. Immunoblots show a representative example of three experiments. c, Total cellular extracts were analyzed for phosphorylation and expression of ERα, IGF1Rβ, IRS-1, and mTOR; GAPDH was used as a loading control. Immunoblots show a representative example of three experiments.

Figure 3. Specific blockade of the IGF1R pathway restores anti-proliferative effects of Tam in mutant-expressing cells

a, ERE-luc assays in MCF-7 cells transiently transfected with WT-, Y537N-, Y537S-, or D538G-ER plasmids. Experiments were performed in triplicate and error bars indicate SD *P<0.05; **P<0.01; ***P<0.001. b, Experiments were performed in triplicate and error bars indicate SD. ns, not significant. *P<0.05; **P<0.01; ***P<0.001. c and d, Cells were transfected for 24 hours with two different IGF1Rβ siRNAs (100nM) and were maintained for 14 days in culture. 10μg of protein were loaded per well and the blot was probed with antibodies detecting IGF1Rβ; GAPDH was used as a loading control (c), Same cells were plated for soft agar assay and then treated with vehicle (-) or with Tam (100nM). Experiments were performed in triplicate and error bars indicate SD. ns, not significant, *P<0.5.

Figure 4. MCF-7 Y537S cells show the lowest anti-proliferative response to Tam through increased interaction between IGF1R/ER

a, b, and c, Immunoblot analysis to detect expression of IGF1Rβ, pIRS-1, IRS-1 and ER in input extracts; GAPDH was used as a loading control (a). Coimmunoprecipitation assays of IGF1R/ER interactions (b and c). Duolink staining (red) demonstrated binding between IGF1R and ER in MCF-7(d) and ZR-75B cells (e). Images are representative of three different experiments.

Figure 5. Fulvestrant in combination with everolimus or GSK enhances anti-proliferative effects in mutant-expressing cells

a, Immunoblot analysis to detect IGF1Rβ and pS6 levels; GAPDH was used a loading control. Immunoblots show a representative example of three different experiments. b, Cells were plated in soft agar assay and then treated with vehicle (-), Tam (100nM), or everolimus (1nM). Experiments were performed in triplicate and error bars indicate SD. ns, not significant. *P<0.05; **P<0.01; ***P<0.001. c, Cells were plated in soft agar assay and then treated with vehicle (-) or with fulvestrant (100nM) and everolimus (1nM). Experiments were performed in triplicate and error bars indicate SD. *P<0.05. d. Cells were plated in soft agar assay and treated with vehicle (-), fulvestrant (100nM) and/or GSK (1μM). Experiments were performed in triplicate and error bars indicate SD. *P<0.05, ***P<0.001.

Figure 6. ESR1 mutation do not predict outcomes in patient treated with adjuvant Tam

a, ddPCR analysis showing allele frequencies of three HBD-ESR1 mutations in primary breast cancers. b and c, Kaplan–Meier analysis of RFS (b), and OS (c). (d) Heatmaps of mRNA levels of genes involved in IGF-1 signaling pathway comparing MCF-7 and ZR-75B stable. Immunoblot analysis to detect phosphorylation and total protein expression of IGF1Rβ, pIRS-1, p85 and p55; GAPDH was used as a loading control (e). Immunoblot of cells transduced with siRNAs for p55 and p85. (f). Cells were also plated in soft agar assays and treated with vehicle (-) or with Tam (10 and 100nM). Experiments were performed in triplicate and error bars indicate SD. ns, not significant. *P<0.05; ***P<0.001.