FOXA1 upregulation promotes enhancer and transcriptional reprogramming in endocrine-resistant breast cancer

Xiaoyong Fua,b,c,1, Resel Pereiraa,b,c, Carmine De Angelisa,b,d, Jamunarani Veeraraghavana,b,d, Sarmistha Nandaa,b,d, Lanfang Qina,b,d, Maria L. Cataldoa,b,d, Vidyalakshmi Sethunatha,b,d, Sepideh Mehravaranb,e, Carolina Gutierrezb,e, Gary C. Chamnessa,b,d, Qin Fengf, Bert W. O’Malleyb,c, Pier Selenicag, Britta Weigeltg, Jorge S. Reis-Filhog, Ofir Cohenh,i,j, Nikhil Wagleh,i,j, Agostina Nardonek, Rinath Jeselsohnh,k, Myles Brownh,k,1, Mothaffar F. Rimawia,b,d, C. Kent Osbornea,b,c,d, and Rachel Schiffa,b,c,d,1

aLester and Sue Smith Breast Center, Baylor College of Medicine, Houston, TX 77030; bDan L. Duncan Comprehensive Cancer Center, Baylor College of Medicine, Houston, TX 77030; cDepartment of Molecular and Cellular Biology, Baylor College of Medicine, Houston, TX 77030; dDepartment of Medicine, Baylor College of Medicine, Houston, TX 77030; eDepartment of Pathology, Baylor College of Medicine, Houston, TX 77030; fDepartment of Biology and Biochemistry, University of Houston, Houston, TX 77204; gDepartment of Pathology, Memorial Sloan Kettering Cancer Center, New York, NY 10065; hDepartment of Medical Oncology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02210; iCenter for Cancer Precision Medicine, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02210; jBroad Institute of MIT and Harvard, Cambridge, MA 02142; and kCenter for Functional Cancer Epigenetics, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02210

Contributed by Myles Brown, November 6, 2019 (sent for review July 9, 2019; reviewed by Douglas Yee and Wilbert Zwart)

Forkhead box A1 (FOXA1) is a pioneer factor that facilitates chromatin binding and function of lineage-specific and oncogenic Significance transcription factors. Hyperactive FOXA1 signaling due to gene am- plification or overexpression has been reported in estrogen receptor- FOXA1 augmentation, including by genetic aberrations, drives + + positive (ER ) endocrine-resistant metastatic breast cancer. However, aggressive phenotypes of estrogen receptor-positive (ER ) the molecular mechanisms by which FOXA1 up-regulation promotes breast cancer (BC). Here, we show that FOXA1 upregulation in-

these processes and the key downstream targets of the FOXA1 on- duces genome-wide enhancer reprogramming and adopts a MEDICAL SCIENCES cogenic network remain elusive. Here, we demonstrate that FOXA1 superenhancer mechanism to activate the master transcription + overexpression in ER breast cancer cells drives genome-wide en- factor HIF-2α and a prometastatic transcriptional program. The hancer reprogramming to activate prometastatic transcriptional hyperactive FOXA1/HIF-2α transcriptional axis is observed to be + programs. Up-regulated FOXA1 employs superenhancers (SEs) to largely nonconcurrent with the ESR1 mutations in clinical ER / − synchronize transcriptional reprogramming in endocrine-resistant HER2 metastatic BC datasets, suggesting different mechanisms breast cancer cells, reflecting an early embryonic development pro- of resistance. Furthermore, a selective HIF-2α inhibitor, currently cess. We identify the hypoxia-inducible transcription factor hypoxia- in clinical trials for advanced renal cell carcinoma and glioblas- inducible factor-2α (HIF-2α) as the top high FOXA1-induced SE target, toma, inhibits the clonogenicity, migration, and invasion of mediating the impact of high FOXA1 in activating prometastatic endocrine-resistant BC cells. These findings demonstrate the role gene sets and pathways associated with poor clinical outcome. Using of FOXA1 upregulation in enhancer reprogramming and a clinical ER+/HER2− metastatic breast cancer datasets, we show that therapeutic approach of targeting deregulated transcriptional the aberrant FOXA1/HIF-2α transcriptional axis is largely nonconcur- programs to circumvent endocrine-resistant metastatic BC. rent with the ESR1 mutations, suggesting different mechanisms of endocrine resistance and treatment strategies. We further demon- Author contributions: X.F., M.B., and R.S. designed research; X.F., R.P., C.D.A., J.V., S.N., α L.Q., M.L.C., V.S., Q.F., O.C., N.W., A.N., and R.J. performed research; X.F. and R.S. super- strate the selective efficacy of an HIF-2 antagonist, currently in vised the study; B.W.O. and M.F.R. contributed new reagents/analytic tools; X.F., S.M., clinical trials for advanced kidney cancer and recurrent glioblastoma, C.G., P.S., B.W., J.S.R.-F., O.C., N.W., and R.S. analyzed data; and X.F., G.C.C., R.J., M.B., in reducing the clonogenicity, migration, and invasion of endocrine- C.K.O., and R.S. wrote the paper. resistant breast cancer cells expressing high FOXA1. Our study has Reviewers: D.Y., University of Minnesota Medical Center; and W.Z., The Netherlands uncovered high FOXA1-induced enhancer reprogramming and HIF- Cancer Institute. 2α–dependent transcriptional programs as vulnerable targets for Competing interest statement: C.K.O. is a consultant/advisory board member for AstraZeneca, treating endocrine-resistant and metastatic breast cancer. GlaxoSmithKline, Pfizer, Puma Biotechnologies, and Tolmar, and on the Data Monitor- ing Committee for Eli Lilly. R.S. has received research support from AstraZeneca, GlaxoSmithKline, Gilead, and Puma Biotechnology, served as a consultant to Eli Lilly, and is breast cancer | endocrine resistance | FOXA1 | enhancer/transcriptional a consultant/advisory board member for MacroGenics. J.S.R.-F. has received personal/consul- reprogramming | metastasis tancy fees from Goldman Sachs, VolitionRx, Page.AI, Grail, Roche, Invicro, and Ventana Medical Systems, outside the submitted work. N.W. has received research support from Novartis and Puma Biotechnology, consults with Novartis, consults with and holds stock from esistance to endocrine therapy in estrogen receptor-positive Foundation Medicine, and is a consultant/advisor of Eli Lilly. R.J. has received research fund- + R(ER ) breast cancer (BC) is common, and leads to poor ing from Pfizer. M.B. receives sponsored research support from Novartis, serves on the Sci- entific Advisory Board of Kronos Bio and is a consultant to H3 Biomedicine. M.F.R. has clinical outcome (1). It has been shown that when ER is inhibited, received research support from GlaxoSmithKline and Pfizer, and consults with Genentech, tumors may activate growth factor receptor (GFR)-related path- Novartis, Daiichi, and MacroGenics. The remaining authors declare that they have no ways to drive endocrine resistance (2–4). However, with the competing interests. exception of the mammalian target of rapamycin inhibitor Published under the PNAS license. everolimus (5), CDK4/6 inhibitors (6), and the PI3K-α isoform- Data deposition: The data reported in this paper have been deposited in the Gene Ex- specific inhibitor alpelisib (7), results of clinical trials using kinase pression Omnibus (GEO) database, https://www.ncbi.nlm.nih.gov/geo (accession no. GSE124656). inhibitors targeting the GFR-related pathways, especially as single 1To whom correspondence may be addressed. Email: [email protected], myles_brown@ agents, are mostly disappointing. Recurrent ESR1 mutations, ob- dfci.harvard.edu, or [email protected]. ∼ + served in 30% of ER metastatic BCs (MBCs), especially those This article contains supporting information online at https://www.pnas.org/lookup/suppl/ treated with aromatase inhibitors (AIs), are recognized as an doi:10.1073/pnas.1911584116/-/DCSupplemental. important mechanism of endocrine resistance, but only in a subset

www.pnas.org/cgi/doi/10.1073/pnas.1911584116 PNAS Latest Articles | 1of12 Downloaded by guest on September 24, 2021 + of ER tumors (8). Other molecular mechanisms underlying endo- thereby promoting tumor progression. Upon FOXA1 induction, crine resistance in the metastatic disease are still poorly understood. FOXA1 chromatin-immunoprecipitation and sequencing (ChIP- FOXA1 is a transcription factor (TF) of the Forkhead box seq) revealed a substantial increase of 60,543 FOXA1 binding (FOX) protein family. It functions as a pioneer factor that binds sites, while 12,721 preexisting sites remained but 15,210 were lost to condensed chromatin to facilitate subsequent binding of ER (Fig. 1A). The increased FOXA1 binding mainly occurred at (9) and other lineage-specific TFs. By characterizing multiple intronic and intergenic regions (SI Appendix,Fig.S1B), similar to endocrine-resistant preclinical BC cell models, we have recently that observed for endogenous FOXA1 binding in both MCF7-P shown that high FOXA1 (H-FOXA1), via gene amplifica- and TamR cells (10, 17, 22). tion and overexpression (OE), plays a key role in promoting To determine how the enhancer landscape evolves upon endocrine-resistant cell growth and invasiveness by reprogram- H-FOXA1 induction, we next performed ChIP-seq of the two ming the ER-dependent transcriptome (10). In addition, several + enhancer marks, H3K27 acetylation (ac) and H3K4me1. Dif- clinical sequencing studies of ER disease reported that about ferential peak analysis identified more regions with increased 6% of primary and 10% of metastatic tumors harbor FOXA1 (GAIN) than with decreased (LOSS) H3K27ac (5,010 vs. 1,722) genetic aberrations, including gene amplification and missense (Fig. 1B). Additionally, substantially more GAIN than LOSS mutations associated with FOXA1 activation (11, 12). A recent regions of H3K4me1 (7,516 vs. 68) were identified in +Dox vs. deep-sequencing study of the regulatory regions in primary breast −Dox cells (Fig. 1C). These findings suggest that ectopic FOXA1 + tumors further revealed recurrent mutations at the FOXA1 pro- OE in ER BC cells promotes an overall chromatin state ame- moter, resulting in high binding affinity for the E2F TF and in- nable for gene transcription. To further assess the occupancy creased FOXA1 gene transcription (13). FOXA1 up-regulation, of H-FOXA1 at enhancers, we overlaid the FOXA1 cistrome due to increased expression and activity, has been reported in signal with the altered enhancer histone marks. This analysis — other metastatic tumors including esophagus, lung, thyroid, showed concordant increase and decrease in FOXA1 binding at – — and prostate (14 16) suggesting potential shared transcrip- the H3K27ac GAIN and LOSS regions, respectively (Fig. 1D). tional programs imposed by H-FOXA1 during malignant disease H3K4me1 GAIN regions were also enriched for increased progression. FOXA1 binding (Fig. 1E). Reciprocally, substantially increased FOXA1 binds to enhancers enriched in histone H3 lysine 4 enhancer marks were observed at the FOXA1 GAIN regions, mono/di-methylation (H3K4me1/me2) (17), where FOXA1 can especially those with high-intensity FOXA1 peaks (SI Appendix, further recruit histone methyltransferase (18) and enhance the + Fig. S2A), of which 15% and 33% were marked with H3K27ac hormone-driven ER activity in ER BC cells (19). This chro- and H3K4me1, respectively (SI Appendix,Fig.S2B and C). matin remodeling activity of FOXA1 is reminiscent of its role in These data suggest that in this inducible FOXA1 OE cell model, inducing tissue-specific gene expression during the development only a subset of the gained FOXA1 binding sites is fully engaged and differentiation of liver, lung, kidney, pancreas, prostate, and + in enhancer activation, which could possibly be attributed to the mammary gland (20). A recent study of the ER BC epigenome insufficiency in additional essential epigenetic modulators, revealed that the enhancer expansion in metastatic tumors is beyond FOXA1. linked to FOXA1 and its network activation (21). However, the mechanism by which H-FOXA1 impacts global enhancers to H-FOXA1–Induced Enhancer Reprogramming Coordinates Gene- promote an endocrine-resistant and metastatic phenotype is not Expression Profiles Enriched for Developmental and Prometastatic clear. Furthermore, the key downstream mediators of H-FOXA1 Processes. We focused our subsequent analysis on enhancer signaling that could serve as therapeutic targets remain to be GAIN regions, which represent the majority of the events during identified. enhancer reprogramming induced by H-FOXA1. More than In this study, we integrated FOXA1 cistrome, epigenetic his- + 90% of GAIN regions of H3K27ac and H3K4me1 were outside tone marks, and transcriptomic data from ER BC cell models promoter regions (Fig. 2 A and B). It has been shown that the expressing H-FOXA1 at the time of acquired or induced endo- crine resistance, and characterized its role in driving BC en- chromatin state at enhancers, rather than at promoters, strongly correlates with cell-type–specific gene-expression programs (23). hancer reprogramming to activate prometastatic transcriptional – programs. We identified the hypoxia-inducible transcription To assess the impact of H-FOXA1 induced enhancer repro- factor-2α (HIF-2α) as the top H-FOXA1–driven superenhancer gramming on gene expression, we integrated these histone ChIP- (SE) target mediating H-FOXA1–induced transcriptional reprog- seq data with our previously obtained RNA sequencing (RNA- ramming in endocrine-resistant BC models. Importantly, we show seq) data in this cell model (10). The overall density of H3K27ac, that targeting HIF-2α by a selective small-molecule inhibitor, over 2 to 10 kb from a transcription start site (TSS) per gene, was currently in clinical trial for advanced renal cell carcinoma, leads coordinately increased or decreased in the up-regulated (UP) or to significant reduction in clonogenicity, migration, and invasion down-regulated (DN) gene sets induced by FOXA1 OE, re- of endocrine-resistant cells. Our study proposes a therapeutic spectively (Fig. 2C). A similar correlation was also observed for approach, via blockade of SE-targeted TFs and aberrant tran- H3K4me1 around the differentially expressed genes (Fig. 2D). scriptional programs, to circumvent endocrine-resistant meta- The partial overlap of these 2 histone marks at GAIN enhancer static BC and possibly other types of aggressive cancers expressing regions upon FOXA1 OE (Fig. 2E) recalls the fact that, in de- H-FOXA1. velopmental model systems, the most open enhancer regions are marked by H3K4me1, but only active enhancers carry both Results H3K4me1 and H3K27ac (24). We therefore annotated the dif- + FOXA1 OE Induces Enhancer Reprogramming in ER BC Cells. To ferentially expressed genes upon FOXA1 OE with the distribu- elucidate the impact of H-FOXA1 on genome-wide enhancers in tion of the enhancers’ marks. As expected, the UP genes were + ER BC, we first used a doxycycline (Dox)-inducible OE system enriched for GAIN enhancers, especially those marked with both in parental (P) endocrine-sensitive MCF7L cells, to achieve H3K27ac and H3K4me1 (Fig. 2F). In contrast, compared to the FOXA1 OE comparable to the high levels found in the MCF7L genes with no change (NC) in expression, the DN genes upon tamoxifen-resistant (TamR) derivative due to endogenous FOXA1 FOXA1 OE were less enriched for the enhancers marked with gene amplification (SI Appendix,Fig.S1A). We hypothesized that H3K27ac. Enhancers marked with H3K4me1 alone were ectopic FOXA1 OE in P cells alters the enhancer landscape to equally enriched in the DN and NC genes. It is possible that for activate transcriptional programs associated with multiple onco- those NC genes harboring H3K4me1-only enhancers, addi- genic pathways, such as GFR and integrin-related signaling (10), tional TFs are needed to fully activate the so-called poised or

2of12 | www.pnas.org/cgi/doi/10.1073/pnas.1911584116 Fu et al. Downloaded by guest on September 24, 2021 ABFOXA1 ChIP-seq Average signal of H3K27ac ChIP-seq -Dox +Dox.1 +Dox.2 FOXA1 binding )012,51( 70 -Dox GAIN +Dox.1 Lost n +Dox.2 ( =5,010)

35 -value) )127,21( LOSS P (

10 (n=1,722) 0 Shared -2 0 2 kb -log 70

35 log2(fold-change) H3K4me1 ChIP-seq

)345,06( C 0 -2 0 2 kb

Gained 70 GAIN -value) Fig. 1. H-FOXA1 reprograms genome-wide enhancers

35 P + ( (n=7,516) in ER BC cells. (A) Heatmap representation of FOXA1 10 LOSS binding intensity based on ChIP-seq reads in MCF7L-P 0 -log (n=68) + -5 5 kb -2 0 2 kb cells with FOXA1 OE ( Dox, duplicates) vs. non-OE (−Dox). Signals within 5 kb around the center of bind- 0 60 0 60 ing peaks are displayed in a descending order for each log (fold-change) 2 clustered FOXA1 binding event (i.e., lost, shared, and + − DE-Dox +Dox.1 +Dox.2 Average signal of -Dox +Dox.1 +Dox.2 Average signal of gained upon Dox vs. Dox). (Right) Plots of average FOXA1 binding FOXA1 binding signal of FOXA1 binding at the clustered regions. (B) Volcano plot of H3K27ac ChIP-seq of MCF7L-P cells with 30 -Dox 30 -Dox +Dox.1 +Dox.1 or without FOXA1 OE. The orange and green dots

(5,010) +Dox.2 +Dox.2 correspond to the regions with GAIN and LOSS H3K27ac + − 15 (7,516) 15 in Dox vs. Dox cells, respectively. (C) Volcano plot of

GAIN H3K4me1 ChIP-seq of MCF7L-P cells with or without ca7

0 GAIN 0 FOXA1 OE. The purple and cyan dots correspond to the MEDICAL SCIENCES 2 + − K3H -2 0 2 kb -2 0 2 kb regions with GAIN and LOSS H3K4me1 in Dox vs. Dox cells, respectively. The threshold calling for GAIN/LOSS is > < -Dox -Dox set as fold-change 4andP 1e-10. (D)Heatmapof

30 H3K4me1 30 +Dox.1 +Dox.1 FOXA1 binding signal over 5 kb around the center of +Dox.2 +Dox.2 H3K27ac GAIN and LOSS regions in +Dox vs. −Dox cells. (1,722)

15 (68) 15 (Right) Plots of average signal intensity of FOXA1

LOSS binding at these regions. (E)Samedatarepresentation

-5 5 kb LOSS -5 5 kb 0 0 for FOXA1 binding at H3K4me1 GAIN and LOSS regions -2 0 2 kb -2 0 2 kb 0 30 0 30 0 30 0 30 as in D.

predetermined enhancers (24), similar to what is seen in embryonic metastatic pancreatic cancer cell model expressing H-FOXA1 (27) gene transcription. (SI Appendix,Fig.S4A and B). In addition, we analyzed the se- Gene ontology (GO) analysis (25) showed that the genes quencing data from a large clinical MBC cohort (28) focusing + harboring the GAIN enhancers with both H3K27ac and H3K4me1 mainly on the ER subtype. We found that the FOXA1 and ESR1 (n = 1,024) were enriched in proproliferation, antiapoptosis, and genetic aberrations (amplification and active mutations) are largely + − developmental signaling, such as “utero embryonic development” nonconcurrent in ER /HER2 metastatic tumors (n = 781) (SI (P = 0.002), “TGF-β receptor signaling” (P = 0.003), and “neuron Appendix,Fig.S4C), suggesting a mechanism of FOXA1 UP, dif- + migration” (P = 0.024) (Fig. 2G). Most of the LOSS enhancers ferent from that of the ESR1 mutations (29), in promoting ER BC were marked with H3K27ac before FOXA1 OE (SI Appendix,Fig. endocrine resistance and metastasis. Overall, these findings support S3A). The top GO term enriched for the genes harboring the the role of H-FOXA1 and its induced enhancer reprogramming in LOSS enhancers was “regulation of Rho protein signal trans- activating prometastatic transcriptional programs associated with + duction” (SI Appendix, Fig. S3B). Interestingly, one of these GO ER disease progression. term genes that was DN upon FOXA1 OE was the gene encoding DLC1 (SI Appendix,Fig.S3C), a Rho-GAP protein that was re- Endocrine Resistance Driven by FOXA1 Amplification Is Associated cently reported to play an estrogen-induced tumor-suppressor role with a FOXA1 Cistrome That Correlates with Active Enhancers (26). This finding is also in line with the reduced classic estrogen- and Transcriptome. The observation that ectopic H-FOXA1 in regulated ER signaling in H-FOXA1–expressing cells, as we have MCF7L-P cells induces enhancer reprogramming to activate previously reported (10). Focusing on the GAIN enhancer genes prometastatic transcriptional programs suggests that H-FOXA1 UP by FOXA1 OE (n = 253), GO analysis uncovered the en- might be the driving force for acquired endocrine resistance in + richment of prometastatic processes, including “regulation of cell ER disease progression. We therefore used our MCF7L-TamR migration” (P = 0.002) and “membrane invagination” (P = 0.003) cell model, harboring endogenous FOXA1 amplification and (Fig. 2H), which share the same characteristics of cellular mor- expressing H-FOXA1 (10), to further elucidate the correlation phogenesis during embryonic development. A recent study has between FOXA1 cistrome, active enhancers, and associated shown that H-FOXA1 drives aberrant enhancers with increased gene expression. We performed H3K27ac ChIP-seq and identi- H3K27ac to promote pancreatic cancer metastasis by activating an fied twice as many H3K27ac gained (6,043) vs. lost (2,658) re- embryonic endoderm transcriptional program (27). Interestingly, gions in TamR vs. P cells (Fig. 3A). Our previous FOXA1 the 253 H-FOXA1–induced genes harboring GAIN enhancers in cistrome data revealed an increased number of gained (31,766) vs. our model were also highly enriched in the gene-expression lost (15,778) binding sites in MCF7L-TamR vs. P cells (10). profiles of embryonic foregut endoderm as well as the reported Plots of the H3K27ac signal over the 5-kb regions around the

Fu et al. PNAS Latest Articles | 3of12 Downloaded by guest on September 24, 2021 ACP=6.0e-12 EFOXA1+Dox vs. -Dox H3K27ac GAIN (5,010) P=7.2e-14 5% 2% 9 K4me1 GAIN K27ac GAIN Promoter 6 Fig. 2. H-FOXA1–induced enhancer reprogramming 30% Intergenic 3 coordinates gene expression toward developmental 63% Intron 3,388 1,604 5,897 and prometastatic transcriptional programs. (A)Pie Other

(2-10 kb from TSS)from(2-10 kb 0 –

Log2 H3K27ac H3K27ac Log2 signal chart showing the genomic annotations of H-FOXA1 -Dox +Dox -Dox +Dox induced H3K27ac GAIN regions according to the loca- FOXA1-UP FOXA1-DN tion of a given peak. (B) The same pie chart as in A for H3K4me1 GAIN regions. (C) Violin plots showing the distribution of UP and DN genes (jG-foldj > 2) upon BDP=6.7e-03 F FOXA1-UP FOXA1-DN FOXA1-NC H3K4me1 GAIN (7,516) P=1.9e-03 H-FOXA1 with average H3K27ac signal over 2 to 10 kb 16 P=2.0e-6 P=4.4e-60 5% 1% P=3.1e-38 from the TSS of each gene in +Dox vs. −Dox cells. The

5 GAIN 12 H3K27ac signal was calculated over a 2-kb range cen- Promoter P=1.4e-36 2.5 tered on each peak and normalized by a total of 10 34% Intergenic 8 * *** million reads. (D) Same violin plots as in C for the al-

Intron enhancers 60% 0 4 tered genes with differential H3K4me1. P value was Other *** (2-10 kb from TSS)from(2-10 kb % genes with 0 calculated by a paired Student’s t test. (E)Venndia- Log2 H3K4me1 H3K4me1 Log2 signal -Dox +Dox -Dox +Dox K27ac + - + + gram representing overlap between H-FOXA1–induced K4me1 - + ++ + FOXA1-UP FOXA1-DN Overlaid yes no H3K27ac GAIN and H3K4me1 GAIN regions in Dox vs. −Dox cells. (F) Bar charts showing the proportion of genes within each of the categories of H-FOXA1– K27ac+/K4me1+ GAIN FOXA1-UP & K27ac+/K4me1+ G -log10(P-value)H -log10(P-value) induced altered (UP or DN) and NC genes that over- predicted genes (1,024) 0 2 4 GAIN predicted genes (253) 0 2 4 lap with the genes near enhancers (within ±100 kb Posive regulaon of cell division Muscle cell differenaon from TSS) demarcated with GAIN H3K27ac and/or Negave regulaon of apoptosis Membrane organizaon H3K4me1. P value was calculated by Bonferroni- In utero embryonic development Regulaon of cell migraon corrected multiple pairwise comparisons with a Fisher’s Posive regulaon of GTPase acvity Purine nucleoside biosynthesis exact test. *P < 0.05, ***P < 0.001. (G) Bar charts TGF beta receptor signaling pathway Membrane invaginaon showing enriched GO terms for the genes predicted Endoplasmic reculum stress Endocytosis from the GAIN enhancers marked by both H3K27ac DNA methylaon Regulaon of locomoon and H3K4me1 (n = 1,024). (H) Bar charts showing Posive regulaon of lipophagy Regulaon of cell moon enriched GO terms for the FOXA1-induced UP genes Histone H3-K9 demethylaon Cholesterol biosynthec process with GAIN enhancers marked by both H3K27ac and Neuron migraon Cytoskeleton intracellular transport H3K4me1 (n = 253). Dashed gray line indicates P = 0.05.

centers of the top 1,000 FOXA1 unique peaks in TamR vs. top enriched motif at the FOXA1/H3K27ac-gained enhancers, P cells showed a coordinated deposition of differential H3K27ac irrespective of ER binding (Fig. 3 F and G). Instead of the es- at these FOXA1 binding sites (Fig. 3B), in accordance with the trogen response elements (ERE), other motifs (e.g., SIX1, ELF3, overall enrichment of gained H3K27ac at the unique FOXA1 and GRHL1) were enriched at the FOXA1/H3K27ac-gained en- occupancy sites in TamR vs. P cells (SI Appendix,Fig.S5A). hancers without ER occupancy, suggesting an interplay between Conversely, FOXA1 unique binding in TamR and P cells was H-FOXA1 and AP-1–involved TF complexes, likely via an ER- positively correlated with the TamR-gained and -lost H3K27ac, independent mechanism, to promote enhancer reprogramming in respectively (SI Appendix,Fig.S5B). endocrine resistance. Of the total gained FOXA1 binding sites in TamR vs. P cells, 29% carry gained and 10% carry lost/NC H3K27ac (Fig. 3C). Of Activation of SE Formation Is Correlated with Reprogrammed FOXA1 note, the unique FOXA1 binding sites marked with H3K27ac Cistrome in Endocrine Resistance. Recent studies have shown that upon FOXA1 OE in P cells shared more common regions with aberrant SEs, a cluster of cis-regulatory elements controlling the unique FOXA1 binding in TamR vs. P cells (23% vs. 4%, mammalian cell identity by activating lineage-specific transcrip- respectively) (SI Appendix, Fig. S5C). Compared to the NC tion (34), underlie the deregulated gene network in several types genes, the UP and DN genes in TamR vs. P cells were signifi- of cancer (35). To further dissect the enhancer reprogramming cantly enriched for the gained FOXA1-binding in TamR and P in endocrine resistance, we identified SEs based on H3K27ac cells, respectively (Fig. 3D). The great enrichment observed by and annotated the TFs nearby these SEs in MCF7L-P and TamR intersecting the altered transcriptome with the genes associated cells (Fig. 4 A and B). More than 50% of these SEs in TamR cells with FOXA1-bound active enhancers (Fig. 3 D, Right) further were demarcated with gained H3K27ac (Fig. 4C). Importantly, supports the enhancer-level activation of FOXA1 in regulating 88% (575 of 653) of these SEs with gained H3K27ac in TamR gene expression in endocrine resistance. We and others have cells were not shared by P cells (Fig. 4D), suggesting that the previously shown that the AP-1 TF, mediating the alteration of enhanced H3K27ac observed in TamR cells is associated with the ER transcriptome, is a key determinant of endocrine re- global SE reprogramming. Similarly, we also observed that the sistance (30–32). We found that the AP-1 binding motif resides majority of the SEs with gained H3K27ac in the MCF7L-P cells at a higher frequency at the FOXA1-bound enhancers with with ectopic inducible FOXA1 OE were not shared by the P cells gained H3K27ac than those with lost/NC H3K27ac (Fig. 3E). without FOXA1 OE (SI Appendix, Fig. S6 A–C). The top Similar analysis of the unique FOXA1 binding upon FOXA1 OE enriched motifs at the TamR H3K27ac-gained SEs include those in P cells also showed that the AP-1 motif was enriched at those bound by AP-1, the embryonic TFs (e.g., STAT5, SOX9, and regions overlapping with the unique FOXA1 binding in TamR vs. SMAD3), and FOXA1 (Fig. 4 C, Right and Dataset S1). These Pcells(SI Appendix,Fig.S5D). Since ER remains expressed findings are of particular interest because several AP-1 and SOX and functional in the MCF7L-TamR cells (10), we further in- family members also harbor SEs at their own gene loci in TamR tegrated these FOXA1 enhancer-binding sites with our pre- cells (Fig. 4B), potentially forming self-regulatory circuits as viously reported ER cistrome in the same model (33). Motif shown in embryonic stem cells (ESCs) (34). In agreement with enrichment analysis showed that the AP-1 binding motif was the our previous reports (30, 31), the essentiality of AP-1 activity at

4of12 | www.pnas.org/cgi/doi/10.1073/pnas.1911584116 Fu et al. Downloaded by guest on September 24, 2021 TamR vs. P H3K27ac P TamR AB P TamR FOXA1 H3K27ac FOXA1 H3K27ac Gained P FOXA1 top binding Lost (n= 6,043) 1.4 P= 3.4e-76 (n= 2,658) -value)

P 1.0 ( 10 0.6 -log

H3K27ac signal 0.2 -5 0 5 P FOXA1 top P FOXA1 top 1,000 Distance (kb) log2(fold-change) C Gained TamR FOXA1 TamR FOXA1 top binding Non-H3K27ac 1.4 Gained H3K27ac 1.0 Lost/NC H3K27ac 0.6 61% P= 9.6e-50 H3K27ac signal 0.2 TamR FOXA1 top 1,000 FOXA1 top TamR -5 5 kb -5 0 5 Distance (kb) 10% 29% 0 30 0 5+ 0 30 0 5+

D TamR-UP TamR-DN NC E 0.015 1.8 fold P= 3.6e-4 Gained

h 80 P=1.0e-90 80 f per f t Lost/NC i 1.5 fold 2.4 fold 0.01 Ɵ ws 60 P=7.2e-25 60 2.3 fold P=1.5e-103 eneg 40 40 P=7.0e-44 0.005

20 20 bp per peak AP-1 mo % FOXA1 binding 0 0 0 P only TamR only P only TamR only -500 0 500 Overlay with H3K27ac Distance from center (bp) MEDICAL SCIENCES F Gained FOXA1_H3K27ac/ P value G Gained FOXA1_H3K27ac/ P value ka TamR-ER TamR non-ER

epr AP-1 1e-85 FOXA1 1e-891 0.005 FOXA1 0.005 FOXA1

eppbr 0.004 AP-1 ERE 1e-50 0.004 AP-1 AP-1 1e-622 ERE ERE 0.003 FOXA1 1e-46 0.003 SIX1 1e-83

epsfi 0.002 0.002 HIC1 1e-14 ELF3 1e-68 0.001 0.001 fs per bp per peak bp per fs per t SMAD3 1e-13 Ɵ GRHL1 1e-65 oM 0 0 -500 0 500 SIX2 1e-13 Mo -500 0 500 NFIL3 1e-54 Distance from center (bp) Distance from center (bp)

Fig. 3. H-FOXA1 in acquired TamR cells driven by FOXA1 amplification coordinates active enhancers and an altered transcriptome. (A) Scatter plots showing relative enrichment of H3K27ac ChIP-seq signal in MCF7L-P vs. TamR cells. Peaks differentially enriched between P and TamR cells are shown in green and orange (fold-change > 4andP < 1e-10). (B) Heatmap of FOXA1 binding and H3K27ac signal over 5 kb around the center of the top 1,000 FOXA1 binding sites with unique peaks observed in P or TamR cells. (Right) Plots of average H3K27ac signal at these regions. P value was calculated by the Wilcoxon matched-pairs signed rank test. (C) Pie chart showing the proportion of TamR-only FOXA1-binding sites without H3K27ac (gray), or with gained (orange) or lost/NC (green) H3K27ac signal. (D) Bar charts showing the proportion of genes within the altered (UP or DN) and NC groups (TamR vs. P) that overlap with the predicted genes with P-only or TamR-only FOXA1 binding, without (Left) or with overlaid H3K27ac (Right). P value was calculated by Bonferroni-corrected multiple pairwise comparisons with a Fisher’s exact test. (E) Enrichment for the AP-1 binding motif in the center (±500 bp) of TamR-only FOXA1 binding sites with gained vs. lost/NC H3K27ac. P value was calculated by the Wilcoxon matched-pairs signed rank test. (F and G) Enrichment for the FOXA1, AP-1, and ERE binding motifs around the center (±500 bp) of FOXA1/H3K27ac-gained enhancers with or without ER binding in TamR cells. Right panels show the top six enriched motifs at these regions. The occurrence of the motifs was normalized to the length and number of peaks. P value was calculated based on the cumulative distribution function of the hypergeometric distribution.

SEs for endocrine-resistant cell growth was further supported Fig. S6E), signifying the role of H-FOXA1 in promoting SE by the significantly higher sensitivity to gene knockdown (KD) of formation. two AP-1 key components (JUN and FOS) in TamR vs. P cells (Fig. 4E). HIF-2α Is the Top TF Target of H-FOXA1–Engaged SEs in TamR Cells. We further examined the correlation of FOXA1 binding and To further determine the impact of H-FOXA1 on SE-induced these SEs using the distinct FOXA1 cistrome identified in the gene transcription, we overlapped the TamR H3K27ac-gained same model. FOXA1 unique binding in TamR vs. P cells was SEs and the FOXA1 cistrome and identified the potential SE significantly enriched at the SEs of TamR vs. P cells (Fig. 4F). target genes (Fig. 5A). The majority of these genes harbor gained However, the observed elevated proportion of the SEs occu- FOXA1 binding in TamR vs. P cells, which is in line with the fact pied by FOXA1 in TamR vs. P cells (22% vs. 6%) (Fig. 4G) that 74% of these FOXA1-engaged SE target genes were UP in suggests the contribution of H-FOXA1 in the endocrine TamR vs. P cells (SI Appendix,Fig.S7A). Importantly, changes resistance-associated SE formation. Similarly, gained FOXA1 in expression of these FOXA1-engaged SE target genes are binding was significantly enriched at the SEs in P cells with highly concordant between the TamR cells with endogenous ectopic FOXA1 OE (SI Appendix,Fig.S6D). Notably, about H-FOXA1 and MCF7L-P cells upon ectopic FOXA1 OE (Fig. 5B 50% of the SEs identified in the P cells with ectopic FOXA1 and SI Appendix,Fig.S7B), supporting the role of SEs in pro- OE overlapped with the SEs in TamR cells, and shared several moting transcriptional reprogramming even in P cells with ectopic enriched motifs, including SMAD3 and STAT5 (SI Appendix, FOXA1 OE.

Fu et al. PNAS Latest Articles | 5of12 Downloaded by guest on September 24, 2021 A MCF7L-P MCF7L-TamR langis B 6,000 HOXC10 (5) 3,500 ELF3 (75) TEAD3 (43) EPAS1 (83) ZFHX3 (82) 3,000 SOX4 (92) ca 5,000 ELF3 (110) ESR1 (111) 72K3Hde 4,000 ERF (114) 2,500 ERF (131) ATF4 (185) 2,000 ZNF281 (147) 3,000 SOX4 (192) ATF4 (160) SMAD7 (229) 1,500 FOSL1 (202)

z RUNX1 (288) 1,144 SEs ilamr GRHL1 (257)

2,000 1,036 SEs IRF3 (306) 1,000 NR4A2 (261) 1,000 YY1 (311) 500 NFIC (297) o N 0 Normalized H3K27ac signal 0 0 3,000 6,000 9,000 12,000 0 3,000 6,000 9,000 12,000 Enhancer ranked by H3K27ac Enhancer ranked by H3K27ac

C P value D TamR P SEs H3K27ac gained H3K27ac- AP-1 1e-11 gained SEs STAT5 1e-11 575 78 958 16,612 SOX9 1e-10 SMAD3 1e-10 E 1.2 YY1 1e-8 P 653 ** TamR SIX1 1e-8 491 SEs TamR-gained 0.7 TamR SE FOXA1 1e-7 0.2 * Cell growth (%) growth Cell -0.3 si-NS si-c-Jun si-c-Fos F G P-SE TamR-SE

Pearson residuals FOXA1_P_only 6% gn P = 5.1e idnib1 FOXA1_shared - 13 A XO FOXA1_TamR_only

F 22%

0 10 20 30 Pearson’s χ2 test | P = 8.9e-13 % Overlap with FOXA1 sites

Fig. 4. SEs are engaged in H-FOXA1-induced enhancer reprogramming in endocrine resistance. (A and B) Hockey-stick plots showing ranked enhancers based on H3K27ac of MCF7L-P and TamR cells. A total of 1,036 and 1,144 SEs were annotated for the enhancers with a slope value of >1 in P and TamR cells, respectively. SE-targeted TFs are indicated with rank orders. (C) Venn diagram showing the overlap between gained (fold-change > 2) H3K27ac- demarcated enhancers and the SEs annotated in TamR cells. Motif enrichment analysis reveals the top enriched motifs within the overlapping regions of TamR H3K27ac-gained SEs (n = 653). (D) Venn diagram showing the overlap between TamR H3K27ac-gained SEs and the SEs annotated in P cells. (E)Bar charts showing the normalized cell growth rate of P and TamR cells with control (nonspecific, NS), c-Jun, or c-Fos siRNA KD. P value was calculated by an unpaired Student’s t test. *P < 0.05, **P < 0.01. (F)Aχ2 test of the contingency table counting the overlapping regions between the FOXA1 cistrome (unique and shared binding sites in TamR vs. P cells) and the SEs annotated in P and TamR cells. Pearson standardized residues for each intersection are color-scaled and the size of the circle is proportional to the amount of the contribution to the total χ2 score. (G) Bar charts showing the proportion of the SEs that overlap with the unique and shared FOXA1 binding sites in TamR vs. P cells. P value was calculated by Bonferroni-corrected multiple pairwise comparisons with a Fisher’s exact test.

We next focused on the EPAS1 gene as the top H-FOXA1– (Fig. 5 D and E and SI Appendix,Fig.S8A–C). HIF-2α was slightly engaged SE target upon endocrine resistance. This gene encodes UP upon short-term tamoxifen treatment in MCF7L-P cells, but HIF-2α, and fits our selection criteria for TFs based on the facts was robustly induced in TamR cells (SI Appendix,Fig.S8D). HIF- that it was: 1) The top SE-targeted TF identified in TamR vs. 2α, but not HIF-1α, was also markedly UP in our previously P cells; 2) the top SE target carrying gained FOXA1 binding in reported MCF7L endocrine-resistant xenograft tumors, especially TamR vs. P cells; 3) the most up-regulated SE-targeted gene in TamR and fulvestrant-resistant (FulR) (38) (SI Appendix, Fig. TamR vs. P cells; and 4) one of the top genes highly induced by S8E). Kaplan–Meier meta-analysis (39) showed that high mRNA FOXA1 OE in P cells (Fig. 5 A and B). Notably, the concordant levels of HIF-2α, but not HIF-1α, predict poor distant metastasis- + EPAS1/HIF-2α induction upon FOXA1 OE in P cells, and re- free survival of patients with ER tumors treated with endocrine duction upon FOXA1 KD in TamR cells, were not observed for therapy (SI Appendix,Fig.S9A and B). In contrast, both HIF-1α HIF1A, encoding another HIF family member, HIF-1α (SI Ap- and HIF-2α predict poor distant metastasis-free survival in pa- − pendix,Fig.S7C–F). In addition, compared to the P cells, HIF-2α tients with ER BC (SI Appendix,Fig.S9C and D). was UP in other MCF7L endocrine-resistant derivatives (10, 36), We then asked whether the associated SE is responsible for in company with H-FOXA1 (Fig. 5C). Induction of HIF-2α,but activating HIF-2α gene (EPAS1) transcription. Using CRISPR/ not HIF-1α, was also observed in additional T47D and 600MPE Cas9-mediated gene editing, we deleted fragments of the SE endocrine-resistant cell models (10, 36, 37) expressing H-FOXA1 harboring gained FOXA1 binding sites within the first intron of

6of12 | www.pnas.org/cgi/doi/10.1073/pnas.1911584116 Fu et al. Downloaded by guest on September 24, 2021 the EPAS1 gene in TamR cells (Fig. 5F). EPAS1 transcription was reduced by 60% and 30% in the cells with deleted proximal and distal intronic regions, respectively, adjacent to the EPAS1 AB4 TamR H3K27ac-gained SE- TamR H3K27ac-gained SE genes TSS (Fig. 5G). Conversely, EPAS1/HIF-2α expression was highly predicted genes (334) EPAS1 2 UP upon FOXA1 OE in P cells, in line with the fact that 0 H3K27ac was increased at 2 of the 3 EPAS1 SE regions (Fig. R) TamR vs. P 2 FOXA1 tags 5H). These data support the engagement of H-FOXA1-induced -2 (log SE in activating EPAS1/HIF-2α gene transcription. To determine the correlation of FOXA1 and HIF-2α expres- P = 4.6e-22 sion in human samples, we performed immunohistochemistry for TamR vs. P (log2R of fpkm) -11 8 R2 = 0.245 HIF-2α using tissue microarrays that we had previously stained for ∼ + FOXA1+Dox vs. -Dox -4 5 FOXA1 (10). We found that 78% of the ER tumors expressed (log2R of fpkm) HIF-2α (Fig. 5I). Importantly, the concordance between nuclear + staining of FOXA1 and HIF-2α was high in ER but not in − CDMCF7L T47D E600MPE ER tumors (Fig. 5 J and K). The fact that HIF-2α expression also *** − FOXA1 ** els *** occurred in ∼26% of ER tumors with absent or low FOXA1 400 *** v 50 *** 15 HIF-2 *** suggests additional regulatory mechanisms of HIF-2α activation, 200 *** *** ** 25 NA levels 10 3 9 ** *** possibly via high GFR-related signaling (40), as also shown in ** *** 2 6 mR e 5 neuroblastoma (41). ve mRNA le v ** i 1 3 i * lat lative m RN A levels

e 0 0 0 Relat Re R High HIF-2α Signaling Is Associated with Metastatic Traits and Poor P lR P R R P R mR + EDR Fu EDR Ful amR ED Ta Tam T Clinical Outcome of ER BC. We have previously shown that ec- topic FOXA1 OE elicits a wide spectrum of activated GFR- 7L-P/FOXA1-Dox +

FGHSE: ~ 45 kb α

H3K27ac FOXA1 H3K27ac 7L-P/FOXA1+Dox related signaling in multiple ER BC cell lines (10). We asked 15 chr2: 46,523,219-46,568,173 P 1.2 ** * 0.06 EPAS1/HIF-2α whether HIF-2α, as the top H-FOXA1–induced SE target, me- 15 TamR 0.8 60 30 diates the impact of H-FOXA1 on cellular signaling pathways. We

0.04 mRNA 100 P EPAS1/HIF-2 0 1 2 3 0.4 Dox: - + first used reverse-phase protein arrays (RPPA) on MCF7L-P cells expression ve ChIP signal ChIP 100 TamR * α 0 0.02 ** with Dox-inducible HIF-2 OE to identify differentially expressed Rela proteins (Dataset S2). Intersection of the UP proteins upon HIF- MEDICAL SCIENCES EPAS1 0 α H3K27ac signal signal of (% H3K27ac input) 1 2 3 2 OE and FOXA1 OE (10) revealed a significant overlap of 29 proteins corresponding to >70% of the UP proteins upon HIF- IJER+ tumor K2α OE in P cells (Fig. 6A). Kyoto Encyclopedia of Genes and rho = 0.34 rho = 0.01 Genomes (KEGG) (42)-defined signaling analysis identified P = 0.001 P = 0.952 FOXA1 the top enriched oncogenic pathways of ERBB2, focal adhe- sion, and insulin (Fig. 6B), which noticeably were also the top + enriched pathways upon FOXA1 OE in multiple ER BC cell Allred score Allred score α α models (10). ER+ (n = 89) ER- (n = 31) We next performed RNA-seq on MCF7L-TamR cells with HIF-2 HIF-2α HIF-2 HIF-2α KD using two validated small-interfering RNA (siRNA) α FOXA1 Allred score FOXA1 Allred score sequences, showing no interference on HIF-1 gene expression (Fig. 6C). Overall, 917 and 1,107 genes were commonly DN and Fig. 5. HIF-2α is the top SE-targeted TF induced by H-FOXA1 in TamR cells UP (false-discovery rate [FDR] < 0.05), respectively (Fig. 6D). + and correlated with FOXA1 expression in ER tumors. (A) Waterfall plot Focusing on the top-ranked DN genes (n = 138), we found that showing the genes predicted from the TamR H3K27ac-gained SEs ranked by the the enriched GO terms were predominantly linked to the af- differential binding signal of FOXA1 at these gained SEs in TamR vs. P cells. The fected signaling pathways in P cells upon HIF-2α OE as shown lower panel shows the top genes with high FOXA1 occupancy that are differ- above, including those associated with tumor metastatic traits, entiallyexpressedinTamRvs.P(redvs.blue),andinMCF7L-Pcellswith(+Dox) − such as “extracellular matrix” (i.e., TNC, SERPINE1,andFN1), vs. without ( Dox) FOXA1 OE (orange vs. green). (B)Scatterplotshowingthe “ ” “ correlation of expression changes of the TamR-gained SE genes from A in TamR focal adhesion (i.e., CD44, ANXA1,andPLAUR), and angiogen- vs. P cells and in P cells with (+Dox) vs. without (−Dox) FOXA1 OE. The extent of esis” (i.e., VEGFA, EPAS1,andKLF5)(Fig.6E). Gene set enrichment gained FOXA1 binding at each gene in TamR vs. P cells was color-scaled. P value analysis (GSEA) further showed that this HIF-2α–dependent gene was calculated using the Pearson correlation coefficient method. (C–E) qRT-PCR set was significantly enriched in the transcriptomes of The Can- + + of HIF-2α and FOXA1 in three ER BC cell models and their endocrine-resistant cer Genome Atlas (TCGA) ER breast tumors (11) (Fig. 6F) derivatives made resistant to tamoxifen (TamR), estrogen-deprivation (EDR), or expressing high HIF-2α, but not high HIF-1α (SI Appendix, Fig. ± < < < + fulvestrant (FulR). Data represent mean SEM. *P 0.05, **P 0.01, ***P S10A). Importantly, the transcriptional profiles UP in ER tumors 0.001, two-tailed t test. (F) H3K27ac and FOXA1 ChIP-seq profiles showing the expressing high HIF-2α were not enriched for the reported HIF- gained SE overlapping with the gained FOXA1 binding at the first intron of the α 1α–dependent gene signature (GS) (43) (SI Appendix,Fig.S10B), EPAS1 locus (encoding HIF-2 ) in TamR vs. P cells. Three guide RNA target sites α– for the CRISPR/Cas9-based editing are indicated by pointers. (G) Bar plots of suggesting that our HIF-2 dependent gene set was not associ- EPAS1/HIF-2α mRNA levels (normalized to GAPDH mRNA) in the MCF7L- ated with HIF-1α signaling. We further defined the clinically rel- TamR cells with CRISPR/Cas9-mediated intronic deletion (del.) at sites in- evant HIF-2α GS by selecting the top-ranked 40 genes with dicated in F. Cells edited by scrambled guide RNA sequences were used as significant GSEA scores (Dataset S3). Oncomine Concepts Map the negative control. (H) ChIP-qPCR analysis of H3K27ac (shown as the per- analysis (44) revealed that this HIF-2α GS was significantly cor- centage of input) in MCF7L-P/FOXA1 ±Dox cells at the three intronic sites related with the gene sets overrepresented in breast tumors at indicated in F. Inset shows the EPAS1/HIF-2α mRNA levels upon FOXA1 OE. ± < < advanced stages, with high GFR/ERBB2 signaling, or with me- Data represent mean SEM. *P 0.05, **P 0.01, two-tailed t test. (I) G Representative immunohistochemistry images of FOXA1 and HIF-2α in one tastasis (Fig. 6 and Dataset S4). + μ To further address the relevance of our HIF-2α GS especially of the ER tumors integrated in a tissue microarray. (Scale bar, 20 m.) (J and K) + in ER metastatic tumors, we evaluated the HIF-2α GS in an Scatter plots showing the correlation between the Allred scores of FOXA1 and + HIF-2α nuclear staining in ER+ (n = 89) and ER− (n = 31) breast tumors. P value ongoing study of ER MBC with available transcriptomic profiles. was calculated using a Spearman’s rank-order correlation test. Specifically, we utilized the currently available transcriptome of 70

Fu et al. PNAS Latest Articles | 7of12 Downloaded by guest on September 24, 2021 ABMCF7L-P UP proteins 60 14/16 6 -log FOXA1 OE -OE HIF-2α HIF-2α OE 12/34 40 11/14 GAPDH 4 10 6/8 ( P /FOXA1 6/13 4/6 -value) 5/145/7 3/8 20 4/14 2/14 2 9 29 32 1/10 HIF-2α

% of UP proteins of % 0 0 upon P = 7.4e-4

si-NS CE -log (P-value) si-HIF-2α_1 GO terms (138 DN genes, log R < -1) 10 180 si-HIF-2α_2 2 01 2 34 extracellular matrix

(fpkm) 120 focal adhesion

seq extracellular exosome - 60 G2/M transion of mitoc cell cycle

RNA 0 response to pepde hormone HIF-2α HIF-1α angiogenesis endothelial cell migraon D G Advanced stage High GFR -log10(P-value) signaling 15 0

917 DN HIF-2α GS (40) -1.5 Z-score high of Number genes (FDR < 0.05) Common genes Common 1,107 UP 1.5 low HIF-2α ErbB2 Mets < 5 yrs GAPDH tumors

FHTCGA ER+ (n = 594) MBC ER+ (n = 70) IMBC ER+ (n = 70) e rocst 0.6 NES = 1.77 0.6 NES = 1.93 0 NES = -1.52 P = 0.02 P = 0.00 P = 0.04 0.3 0.3 -0.3 n e 0 m 138 DN genes 0 HIF-2α GS (40) -0.6 HIF-2α GS (40) h c irn Enrichmentscore Enrichmentscore E HIF-2α high low HIF-2α high low ESR1 mut wt

Fig. 6. High HIF-2α signaling is associated with prometastatic traits and poor clinical outcome. (A) Venn diagram showing the overlap of UP proteins (measured by RPPA) between MCF7L-P cells with HIF-2α OE and P cells with FOXA1 OE. (B) Enriched KEGG pathways represented by the commonly UP proteins upon HIF-2α/FOXA1 OE. Significance in enrichment was calculated by a Fisher’s exact test (P = 0.05 is indicated by a dashed line). Inset showing the Western blot of HIF-2α upon ectopic OE. (C)ReadsofHIF-2α and HIF-1α mRNA (fpkm) from RNA-seq of MCF7L-TamR cells with HIF-2α KD using two different siRNA sequences. (D) Hierarchical clustering of the differentially expressed genes (FDR < 0.05) in TamR cells with HIF-2α KD compared to the

nonspecific (NS) KD control. (E) Bar charts showing enriched GO terms for the 138 DN genes (log2R < −1) upon HIF-2α KD. Dashed gray line indicates P = + 0.05. (F) Normalized enrichment score (NES) of the DN gene set from E in the TCGA ER breast tumors (n = 594) ranked by high vs. low HIF-2α mRNA levels. (G) Oncomine Concepts Map analysis was used to compare the high HIF-2α GS against the published BC GSs. Circles represent GSs with size proportional to number of genes and color based on enrichment significance (P < 1e-3, odds ratio > 2). Straight lines connect GSs with the line weight proportional to the degree of overlap. GSs sharing similar concepts but from independent studies are grouped together by dashed rectangles. (H)NESoftheHIF-2α GS from G in the liver metastases of ER+ MBC (n = 70) expressing high vs. low HIF-2α mRNA levels. (I)NESoftheHIF-2α GS in the same cohort of ER+ MBC stratified by the ESR1 mutant (mut, n = 22) vs. wild-type (wt, n = 48) status.

biopsies exclusively obtained from liver metastases to minimize the cancer lesions from pancreatic metastases and prostate bone divergence in gene expression of normal tissue from the site of the metastases, compared to their levels in the primary tumors (SI metastatic lesion (29). GSEA showed that our HIF-2α but not Appendix,Fig.S11). These findings suggest that high FOXA1/ + HIF-1α–dependent GS was significantly enriched in the ER liver HIF-2α signaling identified in endocrine-resistant MBC is a com- metastases expressing high HIF-2α (Fig. 6H and SI Appendix, Fig. mon mechanism of disease progression, at least in a subset of tu- S10C). In line with the largely nonconcurrent FOXA1 and ESR1 mors in other cancer types. genetic aberrations in MBC (SI Appendix,Fig.S4C), GSEA revealed an inverse correlation of the HIF-2α but not HIF-1α GS HIF-2α Inhibition Suppresses Aggressive Phenotypes of Endocrine- + with the ESR1 mutations in these ER liver metastases (Fig. 6I Resistant BC Cells. To test whether HIF-2α is a viable druggable and SI Appendix,Fig.S10D), suggesting potential different sig- target to treat endocrine-resistant BC, we used a specific HIF-2α naling exerted by high FOXA1/HIF-2α and ESR1 mutations in antagonist, the small-molecule inhibitor PT2385, with proven promoting endocrine-resistant MBC. Using additional clinical efficacy and a favorable safety profile, as shown in a recent phase I gene-expression profiling datasets (45, 46), we found that both trial for advanced kidney cancer (47). PT2385 is a first-in-class FOXA1 and HIF-2α mRNA levels were significantly increased in antagonist that allosterically blocks the heterodimerization of

8of12 | www.pnas.org/cgi/doi/10.1073/pnas.1911584116 Fu et al. Downloaded by guest on September 24, 2021 HIF-2α, but not HIF-1α, with ARNT/HIF-1β and inhibits HIF-2α PT2385 under different endocrine regimens. Addition of PT2385 to target gene expression (48). Similar to what has been reported in estrogen, ED, or Ful, ±Tam,ledtosignificantreductioninclonal renal cell carcinoma cells (49), short-term treatment with PT2385 formation of the MCF7L-TamR cells (Fig. 7E). Of note, in the using the lower range of the dose tested in renal cell carcinoma endocrine-sensitive MCF7L-P cells, the addition of PT2385 further models (48) significantly, although only minimally, reduced the increased the antitumor effect in the presence of endocrine treat- growth of both MCF7L-TamR and T47D-TamR cell models, with ments (SI Appendix,Fig.S12C). Similarly, combination of PT2385 no effect on P cell growth (SI Appendix,Fig.S12A and B). In with Ful reduced clonogenicity of the MCF7L-EDR cells (Fig. 7F), contrast, PT2385 led to robust reduction in clonogenicity of both in which the classic estrogen-dependent genes (e.g., PGR and MCF7L and T47D TamR cell models, while no effect was ob- BCL2) still operate (36), suggesting the necessity of cotargeting the served on P cells expressing lower HIF-2α (Fig. 7 A and B). In HIF-2α and ER pathways in this model. Taken together, these data addition, pharmacologic HIF-2α inhibition also reduced clonoge- suggest the model-specific response to the selective HIF-2α in- nicity of the MCF7L-FulR, T47D- estrogen deprivation-resistant hibitor and signify the importance of combinatory therapy cotar- (EDR), and T47D-FulR models, but not the MCF7L-EDR geting the likely nonredundant HIF-2α and ER pathways in endocrine model (Fig. 7 C and D), suggesting that the antitumor activity of resistance. the HIF-2α inhibitor is not specific for TamR. Since our MCF7L- Since FOXA1 KD reduced the TamR cell migration (10), we TamR cells are growth-dependent on Tam (33) and responsive to asked whether targeting HIF-2α could achieve a similar effect. estrogen treatment (Fig. 7E), we further tested the efficacy of Indeed, PT2385 showed marked efficacy in reducing cell migration

T47D-TamR MCF7L-FulR ABMCF7L-TamR T47D-P C DT47D-EDR MCF7L-EDR ) *** ** *** ) *** K 30 3 N.S 10 * K 30 1 ** * 1 ***

. × 8 MCF7L-P N.S. ty (×

20 2 i 20 T47D-FulR sity (× 1K) 6 N.S n ** 4 10 . 1 10 * 2 ea × Dens rea × De r rea × Density ( A A Area × (× Density 1K) 0 0 0 A 0 0 1 5 0 1 5 0 1 5 0 1 5 0 1 5 0 1 5 0 1 5 0 1 5 PT2385 (μM) PT2385 (μM) PT2385 (μM) PT2385 (μM) MEDICAL SCIENCES

EFMCF7L-TamR MCF7L-EDR GMCF7L-TamR DMSO 40 30 DMSO 400 DMSO

PT2385 5 M 1K) *** PT2385 1 M × PT2385 5 M *** 30 *** 300 PT2385 5 M ty (

i 20

*** ells /field 20 c 200 *** *** *** 10 ** 10 100 a×Dens

*** e Counted Area × (× Density 1K) 0 Ar 0 0 l D m D n E2 E Fu E ion Ta +Ful Ful s ratio 2+Tam nva E Tam I Mig HI60 2.5 -log 2 10 ( 40 1.5 P -value) 20 1 0.5 upon PT2385 0 0 pAKT-S473 % ofproteins DN 2 -0.5

de pAKT-T308 re fold tne 2 0.5 pp42/44 C-log EGFR J Endocrine-sensive classic N FOXA1 ER D ER-dependent si-NS log Enh. ERE programs

FOXA1 Endocrine-resistant -/+ HIF-2α pro-metastac programs

-FOXA1 vs. FOXA1FOXA1 ER si PU SEs TFs HIF-2α MCF7L-TamR -3 0 1 Enh. TFs

Fig. 7. HIF-2α inhibition suppresses aggressive phenotypes of endocrine-resistant BC cells. (A and B) Quantification of colony formation of MCF7L-P/TamR andT47D-P/TamRcellstreatedwiththeHIF-2α antagonist PT2385 at 1 or 5 μM for 14 d. (C and D) Quantification of colony formation of MCF7L-EDR/FulR and T47D-EDR/FulR cells treated with the HIF-2α antagonist PT2385 at 1 or 5 μMfor14d.(E) Quantification of colony formation of MCF7L-TamR cells treated with estrogen (E2), estrogen-deprivation (ED), or fulvestrant (Ful), each ±Tam and ±PT2385 (5 μM). (F) Quantification of colony formation of MCF7L-EDR cells treated with ED or Ful, each ±PT2385 (5 μM). (G) Quantification of cell migration and invasion of MCF7L-TamR cells treated with PT2385. Data represent mean ± SEM. N.S., nonsignificant, *P < 0.05, **P < 0.01, ***P < 0.001, two-tailed t test. (H) Hierarchical clustering of RPPA data for the altered expression (two-way ANOVA, Tukey’s multiple comparisons, P < 0.05) of proteins (26 DN and 3 UP) in MCF7L-TamR cells treated with PT2385.

(Right) Bar charts of altered expression (log2 ratio) of the corresponding proteins in TamR cells upon FOXA1 KD. (I) Enriched KEGG pathways represented by the DN proteins upon PT2385 treatment. Significance in enrichment was calculated by a Fisher’s exact test (P = 0.05 is indicated by a dashed gray line). (J) A schematic diagram of H-FOXA1–induced enhancer and transcriptional reprogramming. FOXA1 augmentation leads to establishment of SEs and rearrangement of TFs to activate HIF-2α transcription, which in turn induces a prometastatic transcriptional program to mediate the impact of H-FOXA1 to promote ER+ BC progression from endocrine-sensitive to endocrine-resistant and metastatic disease. Targeting HIF-2α represents a strategy targeting deregulated transcriptional programs, resulting in blockade of H-FOXA1–induced enhancer (Enh.) reprogramming to circumvent endocrine-resistant and metastatic disease progression. ERE, estrogen response elements.

Fu et al. PNAS Latest Articles | 9of12 Downloaded by guest on September 24, 2021 of both MCF7L and T47D TamR models in a dose-dependent the traits of metastatic cancer cells, possibly explaining the biologic manner (Fig. 7G and SI Appendix, Fig. S13A), but had no effect consequences of H-FOXA1–induced enhancer reprogramming by + on their less migratory P cell counterparts (SI Appendix, Fig. S13 hacking into developmental programs to promote ER disease B and C). The suppression of cell migration in these two TamR progression and metastasis. cell models treated with PT2385 was recapitulated by HIF-2α The role of H-FOXA1 in enhancer reprogramming, especially KD using two siRNA sequences (SI Appendix,Fig.S13D and via SEs in endocrine resistance, further supports the link be- E), supporting the on-target efficacy of the HIF-2α antagonist. tween embryonic development and tumorigenesis. Originally Furthermore, the invasiveness of MCF7L-TamR cells was drasti- identified as large docking sites for lineage-specific TFs in plu- cally repressed upon either PT2385 treatment or HIF-2α KD (Fig. ripotent ESCs, SEs constitute a cluster of stitched enhancers 7G and SI Appendix,Fig.S13F). controlling gene expression to define mammalian cell identity To further explore the alteration in cellular signaling pathways (34). It has been shown that estrogen induces SE formation in underlying these phenotypic changes, we performed RPPA on MCF7-P cells around the core of the ERE (54). Activation of + MCF7L-TamR cells treated with two doses of PT2385. Pairwise SEs has also been shown in ER BC cell models made resistant and clustering analysis revealed a dose-dependent effect of to AIs (55). Our study using the FOXA1-amplifed TamR cell PT2385 on proteins that were DN (n = 26) in treated vs. un- model reveals a reconfiguration of SEs, highly associated with treated TamR cells (Fig. 7H and Dataset S5). The majority FOXA1 binding and altered transcriptome, as one of the (>50%) of these DN proteins were also DN upon FOXA1 KD in mechanisms to promote endocrine-resistant and metastatic TamR cells (10), including those belonging to GFR-related sig- phenotypes. The engagement of SEs in such epigenetic and naling, such as EGFR and downstream phosphorylated AKT and transcriptomic reprogramming involves multiple key TFs. Spe- MAPK (Fig. 7 H, Right). Western blots further confirmed the cifically, we found the highly enriched binding motifs of AP-1 DN of these proteins in TamR cells upon HIF-2α KD (SI Ap- and other developmental TFs (e.g., STAT5 and SOX9) at newly pendix, Fig. S14A), in line with UP in P cells upon HIF-2α OE (SI established enhancers and SEs in TamR cells. We have recently Appendix, Fig. S14B), supporting the on-target effect of PT2385. shown that transcription of the embryonic factor SOX9 is stim- Finally, KEGG pathway analysis showed the significant enrich- ulated by the ER–RUNX2 complex in endocrine resistance ment of VEGF, ERBB2, and focal adhesion signaling in the (33). Interestingly, several master TFs in ESCs are often driven proteins DN upon PT2385 treatment (Fig. 7I). Importantly, by and also bound to SEs, thus forming a self-regulatory circuit these same pathways were also enriched in the proteins UP upon to amplify transcriptional programs (34). This raises the pos- HIF-2α or FOXA1 OE in P cells (Fig. 6B) and in the gene set sibility that H-FOXA1 adopts a similar mechanism, as shown in DN upon HIF-2α KD in TamR cells (Fig. 6E), supporting the embryonic development, to establish SEs engaged by multiple notion that high HIF-2α and FOXA1 at least partly converge on self-reinforcing TFs and to activate transcriptional reprog- the same signaling that can be impeded by the specific HIF-2α ramming to promote endocrine resistance and metastasis. Fu- antagonist. ture studies are needed to further advance our understanding of the involvement of 3D chromatin architecture (56) in en- Discussion hancer reprogramming upon endocrine resistance driven by In this study, we characterized the H-FOXA1–induced epi- H-FOXA1 signaling, including by endogenous or exogenous genomic and transcriptional alterations in endocrine resistance FOXA1 OE and amplification. using BC cell models with endogenous or exogenous FOXA1 FOXA1 up-regulation has been documented in immature/ OE. We showed the coordination of H-FOXA1–induced progenitor luminal BC stem cells (57) and in murine transgenic enhancer and transcriptional reprogramming, resulting in gene- luminal tumors enriched for a luminal progenitor gene signature expression profiles associated with embryonic development and (58). In metastatic pancreatic cancer, the role of FOXA1 up- prometastatic processes. We identified HIF-2α as the top SE- regulation in promoting enhancer reprogramming is associated targeted TF induced by H-FOXA1, its role in mediating H- with the progenitor subtype of pancreatic cancer, whereas FOXA1 signaling, and its therapeutic potential in treating FOXA1 expression is low in the squamous subtype (27). In- endocrine-resistant and metastatic BC (Fig. 7J). The effective- creased luminal progenitor/stem cell activity has been docu- ness of a specific HIF-2α antagonist in treating our H-FOXA1– mented in endocrine-resistant BC (59). Regardless of the ER expressing endocrine-resistant models, and our findings of high binding, the AP-1 binding motif was mostly enriched at the HIF-2α signaling in endocrine-resistant metastatic BC, highlight the gained FOXA1 enhancers/SEs in TamR cells, in line with our importance of developing HIF-2α–targeted therapy to improve previous findings for the strong dependency of endocrine re- patient outcome. sistance on AP-1 (31) and the fact that the AP-1 motif was also Recent clinical and preclinical studies by us and others suggest enriched at the gained enhancers in metastatic pancreatic tumors that FOXA1 augmentation, via gene amplification and OE, is a expressing H-FOXA1 (27). In addition, other TFs, such as SIX1, + driver for endocrine-resistant and metastatic ER BC (10, 28). with an active role in mammary luminal progenitor cells and Substantial enrichment of activating FOXA1 genomic aberrations metastatic tumors (60), were identified as potential components has been recently reported in advanced hormone-resistant prostate residing at the gained FOXA1 enhancers without ER binding cancer using large primary and metastatic cohorts (50, 51). FOXA1 (Fig. 3G), suggesting ER-independent events of H-FOXA1– assists and orchestrates lineage-specific TF binding and action induced enhancer reprogramming in endocrine resistance. It in both embryonic development and tumorigenesis (17, 52), the has recently been reported that up-regulated GATA5 cooperates two biological events possibly sharing traits induced by H-FOXA1 with H-FOXA1 to activate enhancer reprogramming in meta- signaling. Interestingly, our study shows that H-FOXA1 induces static models of pancreatic cancer (27). Interestingly, we have genome-wide enhancer reprogramming in endocrine-resistant BC previously reported that the expression of GATA3, a lineage- to promote transcriptional programs associated with early de- specific TF in luminal-subtype BC, is substantially decreased in + velopmental processes. Our findings are in line with what has been ER P cells upon FOXA1 OE and in the FOXA1-amplified shown in metastatic pancreatic cancer, where H-FOXA1 drives a MCF7-TamR cells (10, 61). The decreasing GATA3, via ge- retrograde developmental transition for disease progression and netic and epigenetic mechanisms, may also explain our observed metastasis (27). Both mammary gland development and embryonic loss of a subset of FOXA1 binding sites, as shown in the endoderm differentiation involve the branching morphogenesis of H-FOXA1–induced FOXA1 cistrome, where GATA3 and FOXA1 rapidly proliferating epithelial cells that invade into mesenchymal conjointly dictate the luminal-lineage gene expression (62, 63). tissues (53). Such cellular characteristics share commonality with Whether other GATA family members or other lineage-specific

10 of 12 | www.pnas.org/cgi/doi/10.1073/pnas.1911584116 Fu et al. Downloaded by guest on September 24, 2021 TFs are engaged in H-FOXA1–induced enhancer reprogram- and showed that a specific HIF-2α antagonist, currently in clinical ming is yet to be determined. The interplay between H-FOXA1, trial for advanced kidney cancer, suppressed the endocrine- AP-1, and other key TFs—in both an ER-dependent and -in- resistant metastatic cell propensities, including clonogenicity (es- dependent manner, as we have shown at the enhancer level, pecially in the presence of Ful), as well as cell migration and in- to drive a luminal progenitor transcriptional program and vasion. By characterizing H-FOXA1–induced reprogramming and thereby contribute to endocrine resistance—remains to be further its top downstream target, our findings support a highly effective investigated. and less toxic therapeutic strategy of targeting HIF-2α and its Further dissection of H-FOXA1–engaged SEs revealed HIF- prometastatic transcriptional program, to circumvent endocrine- 2α as the top TF of SE targets to mediate H-FOXA1 signaling in resistant metastatic BC with H-FOXA1 signaling. endocrine resistance. Recent studies have illustrated SEs as hot spots of cis-regulatory elements driving the transcription of key Materials and Methods oncogenes, such as MYC (64), in tumorigenesis. A link between Cell Lines. Endocrine-resistant cell models (MCF7L, T47D, and 600MPE) were HIF-2α expression and SE activation was unexpected, as the HIF established as described previously (10, 36, 37). Cells were maintained in family member HIF-1α is mainly induced by hypoxia at the phenol red-free medium supplemented with 1% penicillin/streptomycin/ protein level via the oxygen-dependent hydroxylase pathway glutamine (Gibco), 10% charcoal-stripped FBS (for EDR), and 100 nM 4-OH- α tamoxifen (Sigma) or 100 nM fulvestrant (Sigma) for TamR and FulR cells, (65). However, we show that HIF-2 up-regulation is concordant respectively. The isogenic MCF7L lines expressing Dox-inducible HIF-2α or with H-FOXA1 across our multiple endocrine-resistant cell FOXA1 were constructed as described previously (10). All of the cells passed models grown under nonhypoxic conditions. Noticeably, high authentication before the experiments. GFR-related signaling, such as EGFR, has been shown to induce HIF-2α expression in endocrine-resistant BC cells (40). There- ChIP-Seq. Cells were grown to 85% confluence in 15-cm plates and cross- fore, both H-FOXA1–induced SEs and high GFR-related sig- linked with 1% formaldehyde (10 min) followed by quenching (125 mM naling (10), which also activates FOXA1, may converge to activate glycine). Cells were washed in cold PBS and harvested by a cell scraper in HIF-2α transcription and signaling in endocrine-resistant BC. cold PBS with protease inhibitors (Roche). Cell pellets were resuspended in In addition, we found that perturbation of either HIF-2α or cytosolic and then nuclear lysis buffer, and sheared at 4 °C using a Bioruptor FOXA1 in our endocrine-resistant cell models led to a sub- (Diagenode) for 20 min at high output (30-s on, 30-s off). Sonicated lysates were cleared by centrifuging at 20,000 × g for 10 min and diluted 4 times stantial overlapping of concordantly altered GFR-related sig- before preincubation with protein-A/G beads (Santa Cruz) for 30 min at naling, such as ERBB2, and prometastatic VEGF and focal α 4 °C. ChIP was performed by overnight incubation at 4 °C with antibody adhesion pathways. This suggests that HIF-2 is a key mediator against human FOXA1 (Abcam, ab23738), H3K27ac (Active Motif, #39134), of H-FOXA1 signaling, potentially forming a vicious cycle involving or H3K4me1 (Abcam, ab176877), followed by an additional 1-h incubation MEDICAL SCIENCES high GFR-related signaling to further accelerate H-FOXA1– with protein-A/G beads. For FOXA1 ChIP-seq, we added spike-in Dro- induced enhancer reprogramming. Interestingly, in the endocrine- sophila melanogaster chromatin along with the antibody against histone sensitive MCF7L-P cells, PT2385 further enhanced the efficacy of variant H2Av (Active Motif, # 61752). Beads were washed consecutively antiestrogen treatments (SI Appendix,Fig.S12C), possibly by with low- and high-salt wash buffer, once with LiCl wash buffer (20 mM counteracting the early partial induction of HIF-2α upon short- Tris pH 8.0, 1 mM EDTA, 250 mM LiCl, 1% Nonidet P-40, 1% sodium deoxycholate), and once with TE buffer. DNA was eluted in elution buffer term endocrine therapy (SI Appendix,Fig.S8D), as an early (50 mM NaHCO3 and 1% SDS) and then supplemented with 300 mM NaCl. adaptive mechanism of resistance. These data are also sup- Cross-links were reversed overnight at 67 °C. RNA was digested at 37 °C ported by previous findings showing HIF-2α reduction in + with 0.1 mg/mL RNase A for 30 min. DNA was purified with a PCR purifi- ER BC cells upon estrogen stimulation (66). Nevertheless, cation kit (Qiagen). at the time of acquired resistance the induction in HIF-2α Indexed libraries were prepared from ChIP DNA using the KAPA Hyper levels is much greater, partly due to the H-FOXA1–induced Library Preparation Kit (Kapa Biosystems). Libraries were amplified by 12 SE activation. cycles of PCR, and then assessed for size distribution using the 4200 TapeStation We found the HIF-2α–dependent GS to be highly associated High Sensitivity D1000 ScreenTape (Agilent Technologies) and quantified + with prometastatic processes, with poor outcome of ER tumors using the Qubit dsDNA HS Assay Kit (ThermoFisher). The indexed libraries were multiplexed, 10 libraries per pool. The pool was quantified by qPCR treated with endocrine therapy, and with endocrine-resistant + using the KAPA Library Quantification Kit (KAPA Biosystems) and then se- metastatic ER tumors harboring wild-type but not mutant quenced on the Illumina NextSeq500 using the high-output 75-bp single-read ESR1. This raises the hypothesis that the ESR1 mutations, configuration. mainly found in metastatic post-AI treatment tumors, and the Additional information on ChIP-seq data analyses, ChIP-qPCR, RNA-seq activation of H-FOXA1/HIF-2α signaling represent two different and analyses, GSEA, CRISPR/Cas9-mediated enhancer disruption, qRT-PCR, + mechanisms in driving ER disease progression. Interestingly, we immunohistochemistry, RNA interference, Kaplan–Meier curves, clono- found that the HIF-2α antagonist PT2385 inhibited clonogenicity genic assay, cell migration and invasion assays, RPPA assay, Western of MCF7L-TamR and FulR cells, but not the EDR cells, which blotting, and statistical analyses is provided in SI Appendix, Materials have been shown to be hypersensitive to residual low levels of and Methods. estrogen (67, 68). In combination with Ful, the efficacy of PT2385 ACKNOWLEDGMENTS. We thank Rena Mao and Joy Guo for performing in reducing clonogenicity was enhanced in MCF7L-TamR cells, immunohistochemistry; and Fuli Jia, Myra Costello, and Dr. Kimberley and was also observed in MCF7L-EDR cells (Fig. 7 E and F). + Holloway for performing the reverse-phase protein array assays. This work These findings suggest that cotargeting the ER pathway in ER - was supported by the Department of Defense Breakthrough Award resistant tumors can enhance the efficacy of anti–HIF-2α or at W81XWH-14-1-0326 (to X.F.); the Breast Cancer Research Foundation BCRF least sensitize tumors to such treatment. Continuous accrual of 17-143 and 18-145 (to R.S. and C.K.O.); Stand Up to Cancer Translational + Grant SU2C-AACR-DT0409 (to R.S. and C.K.O.); National Institutes of clinical cohorts of advanced ER tumors with sequencing profiles Health (NIH) Breast Cancer Specialized Programs of Research Excellence and therapeutic records is therefore essential to test our hypoth- Grants P50-CA058183 and CA186784 (to C.K.O., R.S., and M.F.R.), and P50- esis and to guide future HIF-2α–targeted therapy, especially in CA168504 (to N.W.); NIH/National Cancer Institute Cancer Center Support Grants P30-CA125123 and CA008748 (to C.K.O.), and P30-CA008748 (to combination with Ful or new selective ER degraders (36), to J.S.R.-F. and B.W.); NIH Grant R61AI133697 (to Q.F.); Susan G. Komen Grant prevent or treat endocrine-resistant and metastatic disease. Pref- CCR15333343 (to N.W.); The V Foundation (N.W.); The Breast Cancer erentially targeting SEs or SE-targeted genes, instead of targeting Alliance (N.W.); and The Cancer Couch Foundation (N.W.). This work was one gene at a time, has been proven to be a powerful way to treat also supported by the Cancer Prevention & Research Institute of Texas (CPRIT) Grant RP190398 (to R.S. and X.F.); CPRIT Proteomics & Metabo- aggressive tumors driven by aberrant transcriptional programs lomics Core Facility Support Award RP170005; and National Cancer In- (64). Although FOXA1 is currently undruggable, we have identi- stitute Cancer Center Support Grant to Antibody-based Proteomics Core/ fied HIF-2α as a top H-FOXA1 downstream signaling component, Shared Resource P30CA125123.

Fu et al. PNAS Latest Articles | 11 of 12 Downloaded by guest on September 24, 2021 1. C. K. Osborne, R. Schiff, Mechanisms of endocrine resistance in breast cancer. Annu. 37. G. Morrison et al., Therapeutic potential of the dual EGFR/HER2 inhibitor AZD8931 in Rev. Med. 62, 233–247 (2011). circumventing endocrine resistance. Breast Cancer Res. Treat. 144, 263–272 (2014). 2. C. J. Creighton et al., Proteomic and transcriptomic profiling reveals a link between 38. C. J. Creighton et al., Development of resistance to targeted therapies transforms the the PI3K pathway and lower estrogen-receptor (ER) levels and activity in ER+ breast clinically associated molecular profile subtype of breast tumor xenografts. Cancer Res. cancer. Breast Cancer Res. 12, R40 (2010). 68, 7493–7501 (2008). 3. S. Massarweh et al., Tamoxifen resistance in breast tumors is driven by growth factor 39. B. Györffy et al., An online survival analysis tool to rapidly assess the effect of 22,277 receptor signaling with repression of classic estrogen receptor genomic function. genes on breast cancer prognosis using microarray data of 1,809 patients. Breast Cancer Res. 68, 826–833 (2008). Cancer Res. Treat. 123, 725–731 (2010). 4. L. Formisano et al., Association of FGFR1 with ERalpha maintains ligand-independent 40. M. W. Alam et al., HIF2α contributes to antiestrogen resistance via positive bilateral ER transcription and mediates resistance to estrogen deprivation in ER(+) breast crosstalk with EGFR in breast cancer cells. Oncotarget 7, 11238–11250 (2016). cancer. Clin. Cancer Res. 23, 6138–6150 (2017). Erratum in: Clin. Cancer Res. 25, 1433 41. S. Mohlin et al., PI3K-mTORC2 but not PI3K-mTORC1 regulates transcription of HIF2A/ (2019). EPAS1 and vascularization in neuroblastoma. Cancer Res. 75, 4617–4628 (2015). 5. N. Kornblum et al., Randomized phase II trial of fulvestrant plus everolimus or pla- 42. M. Kanehisa, S. Goto, KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids cebo in postmenopausal women with hormone receptor-positive, human epidermal – growth factor receptor 2-negative metastatic breast cancer resistant to aromatase Res. 28,27 30 (2000). α inhibitor therapy: Results of PrE0102. J. Clin. Oncol. 36, 1556–1563 (2018). 43. J. Yang et al., Estrogen receptor- directly regulates the hypoxia-inducible factor 1 6. M. Preusser et al., CDK4/6 inhibitors in the treatment of patients with breast cancer: pathway associated with antiestrogen response in breast cancer. Proc. Natl. Acad. Sci. Summary of a multidisciplinary round-table discussion. ESMO Open 3, e000368 (2018). U.S.A. 112, 15172–15177 (2015). 7. F. André et al.; SOLAR-1 Study Group, Alpelisib for PIK3CA-mutated, hormone 44. D. R. Rhodes et al., ONCOMINE: A cancer microarray database and integrated data- receptor-positive advanced breast cancer. N. Engl. J. Med. 380, 1929–1940 (2019). mining platform. Neoplasia 6,1–6 (2004). 8. R. Jeselsohn, C. De Angelis, M. Brown, R. Schiff, The evolving role of the estrogen 45. R. A. Moffitt et al., Virtual microdissection identifies distinct tumor- and stroma- receptor mutations in endocrine therapy-resistant breast cancer. Curr. Oncol. Rep. 19, specific subtypes of pancreatic ductal adenocarcinoma. Nat. Genet. 47, 1168–1178 35 (2017). (2015). 9. J. S. Carroll et al., Genome-wide analysis of estrogen receptor binding sites. Nat. 46. A. Kumar et al., Substantial interindividual and limited intraindividual genomic di- Genet. 38, 1289–1297 (2006). versity among tumors from men with metastatic prostate cancer. Nat. Med. 22, 369– 10. X. Fu et al., FOXA1 overexpression mediates endocrine resistance by altering the ER 378 (2016). transcriptome and IL-8 expression in ER-positive breast cancer. Proc. Natl. Acad. Sci. 47. K. D. Courtney et al., Phase I dose-escalation trial of PT2385, a first-in-class hypoxia- – U.S.A. 113, E6600 E6609 (2016). inducible factor-2alpha antagonist in patients with previously treated advanced clear 11. Cancer Genome Atlas Network; Comprehensive molecular portraits of human breast cell renal cell carcinoma. J. Clin. Oncol. 36, 867–874 (2018). tumours. Nature 490,61–70 (2012). 48. E. M. Wallace et al., A small-molecule antagonist of HIF2alpha is efficacious in pre- 12. G. Ciriello et al.; TCGA Research Network, Comprehensive molecular portraits of in- clinical models of renal cell carcinoma. Cancer Res. 76, 5491–5500 (2016). vasive lobular breast cancer. Cell 163, 506–519 (2015). 49. H. Cho et al., On-target efficacy of a HIF-2α antagonist in preclinical kidney cancer 13. E. Rheinbay et al., Recurrent and functional regulatory mutations in breast cancer. models. Nature 539, 107–111 (2016). Nature 547,55–60 (2017). 50. E. J. Adams et al., FOXA1 mutations alter pioneering activity, differentiation and 14. C. S. Grasso et al., The mutational landscape of lethal castration-resistant prostate – cancer. Nature 487, 239–243 (2012). prostate cancer phenotypes. Nature 571, 408 412 (2019). 15. L. Lin et al., The hepatocyte nuclear factor 3 alpha gene, HNF3alpha (FOXA1), on 51. A. Parolia et al., Distinct structural classes of activating FOXA1 alterations in advanced – chromosome band 14q13 is amplified and overexpressed in esophageal and lung prostate cancer. Nature 571, 413 418 (2019). adenocarcinomas. Cancer Res. 62, 5273–5279 (2002). 52. K. S. Zaret, J. Lerner, M. Iwafuchi-Doi, Chromatin scanning by dynamic binding of 16. C. Nucera et al., FOXA1 is a potential oncogene in anaplastic thyroid carcinoma. Clin. pioneer factors. Mol. Cell 62, 665–667 (2016). Cancer Res. 15, 3680–3689 (2009). 53. G. M. Bernardo et al., FOXA1 is an essential determinant of ERalpha expression and 17. M. Lupien et al., FoxA1 translates epigenetic signatures into enhancer-driven lineage- mammary ductal morphogenesis. Development 137, 2045–2054 (2010). specific transcription. Cell 132, 958–970 (2008). 54. D. Bojcsuk, G. Nagy, B. L. Balint, Inducible super-enhancers are organized based on 18. K. M. Jozwik, I. Chernukhin, A. A. Serandour, S. Nagarajan, J. S. Carroll, FOXA1 directs canonical signal-specific transcription factor binding elements. Nucleic Acids Res. 45, H3K4 monomethylation at enhancers via recruitment of the methyltransferase MLL3. 3693–3706 (2017). Cell Rep. 17, 2715–2723 (2016). 55. V. T. Nguyen et al., Differential epigenetic reprogramming in response to specific 19. K. Gala et al., KMT2C mediates the estrogen dependence of breast cancer through endocrine therapies promotes cholesterol biosynthesis and cellular invasion. Nat. α – regulation of ER enhancer function. Oncogene 37, 4692 4710 (2018). Commun. 6, 10044 (2015). 20. J. R. Friedman, K. H. Kaestner, The Foxa family of transcription factors in development 56. Y. Zhou et al., Temporal dynamic reorganization of 3D chromatin architecture in – and metabolism. Cell. Mol. Life Sci. 63, 2317 2328 (2006). hormone-induced breast cancer and endocrine resistance. Nat. Commun. 10, 1522 21. D. K. Patten et al., Enhancer mapping uncovers phenotypic heterogeneity and evo- (2019). lution in patients with luminal breast cancer. Nat. Med. 24, 1469–1480 (2018). 57. K. Tachi et al., FOXA1 expression affects the proliferation activity of luminal breast 22. A. Hurtado, K. A. Holmes, C. S. Ross-Innes, D. Schmidt, J. S. Carroll, FOXA1 is a key cancer stem cell populations. Cancer Sci. 107, 281–289 (2016). determinant of estrogen receptor function and endocrine response. Nat. Genet. 43, 58. F. Chen et al., Tip30 controls differentiation of murine mammary luminal progenitor 27–33 (2011). to estrogen receptor-positive luminal cell through regulating FoxA1 expression. Cell 23. N. D. Heintzman et al., Histone modifications at human enhancers reflect global cell- type-specific gene expression. Nature 459, 108–112 (2009). Death Dis. 5, e1242 (2014). 24. M. P. Creyghton et al., Histone H3K27ac separates active from poised enhancers and 59. B. M. Simões et al., Anti-estrogen resistance in human breast tumors is driven by – predicts developmental state. Proc. Natl. Acad. Sci. U.S.A. 107, 21931–21936 (2010). JAG1-NOTCH4-dependent cancer stem cell activity. Cell Rep. 12, 1968 1977 (2015). 25. W. Huang, B. T. Sherman, R. A. Lempicki, Systematic and integrative analysis of large 60. R. Iwanaga et al., Expression of Six1 in luminal breast cancers predicts poor prognosis gene lists using DAVID bioinformatics resources. Nat. Protoc. 4,44–57 (2009). and promotes increases in tumor initiating cells by activation of extracellular signal- 26. D. Chi et al., Estrogen receptor signaling is reprogrammed during breast tumori- regulated kinase and transforming growth factor-beta signaling pathways. Breast genesis. Proc. Natl. Acad. Sci. U.S.A. 116, 11437–11443 (2019). Cancer Res. 14, R100 (2012). 27. J. S. Roe et al., Enhancer reprogramming promotes pancreatic cancer metastasis. Cell 61. Q. Feng et al., An epigenomic approach to therapy for tamoxifen-resistant breast 170, 875–888.e20 (2017). cancer. Cell Res. 24, 809–819 (2014). 28. P. Razavi et al., The genomic landscape of endocrine-resistant advanced breast can- 62. S. L. Kong, G. Li, S. L. Loh, W. K. Sung, E. T. Liu, Cellular reprogramming by the cers. Cancer Cell 34, 427–438.e6 (2018). conjoint action of ERα, FOXA1, and GATA3 to a ligand-inducible growth state. Mol. 29. R. Jeselsohn et al., Allele-specific chromatin recruitment and therapeutic vulnerabil- Syst. Biol. 7, 526 (2011). – ities of ESR1 activating mutations. Cancer Cell 33, 173 186.e5 (2018). 63. J. Eeckhoute et al., Positive cross-regulatory loop ties GATA-3 to estrogen receptor 30. R. Schiff et al., Oxidative stress and AP-1 activity in tamoxifen-resistant breast tumors alpha expression in breast cancer. Cancer Res. 67, 6477–6483 (2007). – in vivo. J. Natl. Cancer Inst. 92, 1926 1934 (2000). 64. J. Lovén et al., Selective inhibition of tumor oncogenes by disruption of super- 31. L. Malorni et al., Blockade of AP-1 potentiates endocrine therapy and overcomes enhancers. Cell 153,320–334 (2013). resistance. Mol. Cancer Res. 14, 470–481 (2016). 65. G. L. Semenza, Hypoxia-inducible factors in physiology and medicine. Cell 148, 399– 32. M. Lupien et al., Growth factor stimulation induces a distinct ER(alpha) cistrome 408 (2012). underlying breast cancer endocrine resistance. Genes Dev. 24, 2219–2227 (2010). 66. J. H. Fuady et al., Estrogen-dependent downregulation of hypoxia-inducible factor 33. R. Jeselsohn et al., Embryonic transcription factor SOX9 drives breast cancer endocrine (HIF)-2α in invasive breast cancer cells. Oncotarget 7, 31153–31165 (2016). resistance. Proc. Natl. Acad. Sci. U.S.A. 114, E4482–E4491 (2017). 67. C. M. Chan, L. A. Martin, S. R. Johnston, S. Ali, M. Dowsett, Molecular changes asso- 34. W. A. Whyte et al., Master transcription factors and mediator establish super- enhancers at key cell identity genes. Cell 153,307–319 (2013). ciated with the acquisition of oestrogen hypersensitivity in MCF-7 breast cancer cells – 35. S. Sengupta, R. E. George, Super-enhancer-driven transcriptional dependencies in on long-term oestrogen deprivation. J. Steroid Biochem. Mol. Biol. 81, 333 341 cancer. Trends Cancer 3, 269–281 (2017). (2002). 36. A. Nardone et al., The oral selective oestrogen receptor degrader (SERD) AZD9496 68. R. Santen et al., Adaptive hypersensitivity to estradiol: Potential mechanism for sec- is comparable to fulvestrant in antagonising ER and circumventing endocrine ondary hormonal responses in breast cancer patients. J. Steroid Biochem. Mol. Biol. resistance. Br.J.Cancer120, 331–339 (2019). 79 , 115–125 (2001).

12 of 12 | www.pnas.org/cgi/doi/10.1073/pnas.1911584116 Fu et al. Downloaded by guest on September 24, 2021