ARID1A and the BAF complex are determinants of breast cancer treatment response [CRISPR gRNA-seq]. ARID1A and the BAF complex are determinants of breast cancer treatment response [CRISPR gRNA-seq]

Global CRISPR screens provide an unparalleled, longer-term experimental approach for the identification of essential genes in drug resistance. We used an ~18,000 gene deletion screen to discover ARID1A and other BAF complex components as the most critical factors required for response to two classes of Estrogen Receptor (ER) antagonists, namely ER degraders and Selective Estrogen Receptor Modulators (SERMs). Unexpectedly, ARID1A was also the top candidate for response to the BET inhibitor JQ1, but in the opposite direction, where loss of ARID1A sensitised breast cancer cells to BET inhibition. We show that ARIDA binds chromatin at ER cis-regulatory elements and can physically associate with ER in model systems and primary tumour samples. ARID1A binding to ER enhancer elements, can occur in the absence of ER, suggesting that its repressive activity occurs in an enhancer-specific, but ER-independent manner. Specific targeting of ARID1A validated the CRISPR screen and shows that depletion of BAF activity, does not result in redundancy from P-BAF, the other ATP-dependent chromatin remodelling complex, but instead results in loss of HDAC1 binding, increased Histone 4 lysine acetylation and subsequent BRD4-driven growth. ARID1A and the BAF complex therefore function as a critical mechanism of antiestrogen activity and mutation or depletion in BAF activity drives a BRD4-mediated proliferative program that is refractory to ER targeted agents. Since ARID1A is mutated in a subset of treatment-resistant disease, these findings provide mechanistic insight and treatment strategies for patients, based on BAF complex fidelity status. Overall design: CRISPR gRNA sequencing of Cas9 expressing MCF7 cells infected with pooled gRNA libraries with different days of infection and with drug treatments (JQ1, 4-hydroxytamoxifen and Fulvestrant).

全基因组CRISPR筛选为鉴定耐药相关必需基因提供了前所未有的长期实验手段。本研究借助约18000个基因的缺失筛选体系，鉴定出ARID1A与其他BAF复合物组分为响应两类雌激素受体（Estrogen Receptor, ER）拮抗剂的最关键调控因子，这两类拮抗剂分别为ER降解剂与选择性雌激素受体调节剂（Selective Estrogen Receptor Modulators, SERMs）。出乎意料的是，ARID1A同时也是响应BET抑制剂JQ1的顶级候选靶点，但其作用方向截然相反：ARID1A的缺失会使乳腺癌细胞对BET抑制作用更为敏感。本研究证实，ARID1A可结合ER顺式调控元件区域的染色质，并在模型系统与原发性肿瘤样本中与ER发生物理相互作用。ARID1A可在ER缺失的情况下结合ER增强子元件，这提示其抑制活性以增强子特异性、但ER非依赖的方式发挥功能。通过特异性靶向ARID1A验证了本次CRISPR筛选的结果，同时证实BAF活性缺失并不会被另一ATP依赖的染色质重塑复合物P-BAF所代偿，反而会导致组蛋白去乙酰化酶1（Histone Deacetylase 1, HDAC1）结合丢失、组蛋白H4赖氨酸乙酰化水平升高，进而触发溴结构域蛋白4（BRD4）介导的细胞增殖。综上，ARID1A与BAF复合物是抗雌激素活性的关键调控机制；BAF活性的突变或缺失会驱动BRD4介导的增殖程序，使肿瘤对ER靶向治疗药物产生耐药性。鉴于ARID1A在部分耐药性疾病中存在突变，本研究结果为基于BAF复合物保真度状态的患者分层与治疗策略提供了机制层面的阐释。
实验设计：对表达Cas9的MCF7细胞开展CRISPR向导RNA（guide RNA, gRNA）测序，该细胞被感染带有不同感染天数的混合gRNA文库，并分别接受JQ1、4-羟基他莫昔芬与氟维司群的药物处理。