Abstract NG05: Epigenetic mechanisms of endocrine therapy response in breast cancer

Abstract (ER)-positive tumors, also known as luminal breast cancers represent over 70% of breast cancers and they are treated primarily with endocrine therapy. Despite the success of endocrine therapies, resistance to these agents is a major clinical problem, and it eventually develops in all patients with metastatic disease. Therefore, the identification and understanding of the mechanisms of resistance to endocrine therapy is a major research priority. Important insights into the mechanisms of endocrine resistance have been gained with the identification of recurrent, activating mutations in ESR1 (the gene coding ER) in approximately 18 % of the tumors with acquired resistance to aromatase inhibitors. However, the mechanisms of resistance in the remaining 82 % of the cases largely unknown. We have recently genomically characterized the largest cohort of endocrine resistant metastatic breast tumors (n=1918 tumors). In addition to ESR1 mutations, we identified the presence of inactivating mutations in the ARID1A gene of the SWI/SNF complex in metastatic endocrine-resistant breast tumors. In parallel, as part of our efforts to elucidate mechanisms of resistance to endocrine therapy, we performed an epigenome CRISPR/CAS9 knockout screen that identified ARID1A as the top candidate whose loss determined resistance to the ER degrader in the clinic fulvestrant. We validated these findings showing that ARID1A knockout cells are resistant to endocrine therapy in vitro and in vivo. Hence, from both clinical observations and our unbiased CRISPR screen from our laboratory, we identified ARID1A loss as a candidate of endocrine resistance. We next performed for the first time the mechanistic characterization of ARID1A loss in ER-positive breast cancer. We determined that ARID1A loss remodels the chromatin landscape of breast cancer on a genome-wide scale, converging to a basal-like gene expression program resulting from downregulation of master transcription factors (TFs) that are required for luminal (ER+) identity. Transcriptome profiling of ARID1A knockout cell lines and patient samples with loss of ARID1A were consistent with our initial findings showing an induction of basal-like gene expression programs. Thus, we elucidated a new mechanism of endocrine resistance through lineage switching mediated by the inactivation of ARID1A. Mechanistically, we showed that ARID1A-dependent SWI/SNF complex binding was impaired at the genomic sites of the major luminal-lineage determining TFs including ER, FOXA1, and GATA3. We also determined that ARID1A regulates ER-chromatin interactions and ER-dependent transcription, the defining driver in ER+ breast cancer. Hence, tumors cells with loss of ARID1A shift from ER-dependent luminal cells to ER-independent basal-like cells and become less responsive to luminal specific anti-ER therapy. Altogether, we uncovered a critical role for ARID1A in maintaining luminal cell identity and endocrine therapeutic response in ER+ breast cancer. This work which I am the last and corresponding author of was recently published to Nature Genetics (Xu et al, Nature Genetics 2020).Interestingly, further analyses of Arid1a knockout in the mammary gland and breast organoids validated a key role of ARID1A in mammary gland morphogenesis and luminal cell differentiation. Moreover, single cell RNA-seq analyses demonstrated that ARID1A knockout breast cells did not differentiate to mature luminal cells. In the cohort of breast cancer tumors (n=1918), we also identified the presence of hotspot missense mutations in the pioneer transcription factor FOXA1 to be enriched in metastatic endocrine-resistant breast tumors. Mutations in the pioneer transcription factor FOXA1 are a hallmark of ER-positive lobular breast cancers. However, the functional consequences of FOXA1 mutations and whether they affect the therapeutic response to endocrine therapy are currently unknown. We undertook a comprehensive approach to investigate the structural consequences, including the genome-wide chromatin recruitment, chromatin accessibility, and transcriptional network in breast cancer models harboring recurrent hotspot and breast cancer specific FOXA1 mutations. By examining the landscape of FOXA1 mutations in a larger cohort (4,952 breast cancer patients), we identified several hotspot mutations in the Wing2 region, and a breast-cancer specific mutation located in the third beta strand, namely SY242CS. Interestingly, FOXA1 hotspot missense mutations were found to be more common in metastatic samples compared to primary tumors, and mutually exclusive with ESR1 mutations, suggesting a positive correlation between presence of FOXA1 mutations and resistance to endocrine therapy. Indeed, FOXA1 mutations were associated with a lower response to aromatase inhibitors in breast patient samples. Mechanistically, we uncovered that FOXA1 Wing2 mutations display increased chromatin binding affinity at ER sites upon estrogen stimulation, and enhanced ER-mediated transcription. Accordingly, 3D structural molecular simulations determined a highly organized 3D conformation conferred by FOXA1 mutations in the Wing2 loop, which may drive augmented ER recruitment or potentially allow more stable/durable interactions with ER and other cooperating factors at the chromatin. By contrast, FOXA1 SY242CS displayed a novel pioneering function over distinct genomic regions that were enriched by the presence of a new FOXA1-binding motif. This alternative binding motif was FOXA1 SY242CS-specific and common at both unique accessible and chromatin bound sites, leading to novel open chromatin regions and the induction of an alternative transcriptome regardless of estrogen stimulation, granting this variant a cellular growth advantage. Thus mechanistically, two phenotypic groups of FOXA1 mutants were established: hypermorphic Wing2 mutants that augment estrogen response, and a neomorphic SY242CS mutant that promotes an alternative pioneering, and cistromic and transcriptomic function leading to endocrine therapy resistance. This work which I am a corresponding author of was recently published to Cancer Cell Arruabarrena-Aristorena et al, Cancer Cell 2020). In summary, our data provide insights into how mutations in transcriptional regulators perturb their function to dictate breast cancer progression and therapeutic response. Our work highlights the power of examining recurrent cancer-associated mutations to advance our understanding of transcriptional regulators in breast cancer. These results also provide mechanistic insights into how FOXA1 and ARID1A mutations are associated with worse outcome to endocrine therapy and position these genetic alterations as potential biomarkers of endocrine therapy response for the treatment of metastatic ER+ breast cancer. Citation Format: Eneda Toska, Guotai Xu, Amaia Arruabarrena-Aristorena, Maurizio Scaltriti. Epigenetic mechanisms of endocrine therapy response in breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr NG05.