Vulnerability of ARID1A deficient cancer cells to pyrimidine synthesis blockade

Here we report the discovery and preclinical validation of a novel precision medicine strategy for ARID1A-mutated cancer. Unbiased proteomics reveals for the first time that ARID1A protein (BAF250a) binds aspartate transcarbamoylase (ATCase), a key regulatory enzyme of the de novo pyrimidine synthesis pathway. Using isogenic paired ARID1A proficient/deficient cancer cell lines, we show that ARID1A protein deficiency (as occurs in ARID1A mutant cancers) leads to metabolic reprogramming and pyrimidine synthesis dependency. Pyrimidine synthesis blockade using the FDA-approved drug teriflunomide (a DHODH inhibitor) suppresses tumor growth and selectively induces DNA damage in ARID1A-deficient tumor models. Combining pyrimidine synthesis inhibition with DNA damage repair blockade, using teriflunomide and AZD6738 (an ATR inhibitor), achieves potent synergy and induces sustained tumor regression in ARID1A-mutant ovarian cancer patient-derived xenografts (PDX). These compelling preclinical data support the evaluation of this novel combination treatment in patients with ARID1A-mutated cancers. SIGNIFICANCEWe identified that ARID1A-deficient cells are selectively vulnerable to pyrimidine synthesis blockade. Preclinical studies demonstrate the in vivo efficacy of a synergistic drug combination that concurrently inhibits the de novo pyrimidine synthesis pathway and DNA damage repair to induce regression in patient-derived xenograft models of ARID1A-mutated cancer.