High Levels of FOXA1 Triggers Super-Enhancer-Engaged HIF2α/EPAS1 to Promote Endocrine-Resistance and Metastasis in Estrogen Receptor-Positive Breast Cancer

We recently reported that acquired endocrine resistance (Endo-R) in multiple estrogen receptor-positive (ER+) breast cancer cell models is driven by high levels of FOXA1 (H-FOXA1), via gene amplification and/or overexpression (OE), leading to coordinated reprogramming of the FOXA1 genome binding (cistrome), and transcriptome. Forced FOXA1 OE in parental (P) cells induced similar transcriptional reprogramming leading to Endo-R and metastasis. Recent clinical data reveal an enrichment of FOXA1 amplification in ER+ metastases, further supporting the clinical importance of our findings. However, the molecular components, including key druggable targets, and the mechanism of H-FOXA1-induced transcriptional reprogramming in Endo-R and metastasis remain elusive.

Using an integrative multi-OMICS approach, we first analyzed the transcriptome (RNA-seq), FOXA1 cistrome, and active enhancer mark histone H3K27 acetylation (ac) (ChIP-seq) in our H-FOXA1-expressing MCF7 tamoxifen-resistant (TamR) and P/FOXA1-OE cells. We found that FOXA1 OE reprogrammed the FOXA1 cistrome in P cells to resemble that of the TamR cells. The FOXA1 cistrome significantly correlated with the deposition of H3K27ac in TamR vs. P cells. Intersection of H-FOXA1-induced transcriptome and distinct FOXA1 cistrome-predicted genes defined a H-FOXA1 core gene signature that was (i) highly represented in the Endo-R transcriptome across multiple cell models, (ii) linked to multiple metastasis-related GO terms including “hypoxia response”, and (iii) enriched for the cancer secretome gene set. Integrative analysis of H3K27ac-defined super-enhancers (SEs) and altered cistrome/transcriptome upon H-FOXA1 nominated HIF2α (EPAS1), a hypoxia-inducible transcription factor, as the top candidate of SE-activated transcription factors amplifying the H-FOXA1 signaling. HIF2α blockade, either genetically by siRNA or pharmacologically by its selective small molecule inhibitor, currently in clinical trial for renal cell carcinoma, markedly repressed the secretome gene expression (e.g., S100P, TGFB1, and LOX), cell migration, and invasion in TamR cells, and the growth of tumor organoids derived from the high HIF2α expressing PDX tumor BCM4888. Finally, we found that primary ER+ breast tumors (TCGA) with high HIF2α are enriched for a cancer secretome gene set, and that high HIF2α predicts poor distant metastasis-free survival in ER+ breast cancer patients treated with endocrine therapy.

Overall, our findings suggest that H-FOXA1 induces transcriptional reprogramming by coordinating the histone enhancer marks to activate HIF2α via SE-mediated mechanism, which in turn amplifies transcriptional reprogramming, partly by inducing a pro-metastatic secretome, to promote Endo-R and metastasis. Targeting the High-FOXA1/ HIF2α axis to block transcriptional reprogramming may offer a new therapeutic strategy to prevent and treat Endo-R metastatic ER+ breast cancer.