Integrative analysis of patient-derived tumoroids and ex vivo organoid modeling of ARID1A loss in bladder cancer reveals therapeutic molecular targets

Somatic mutations in ARID1A (AT-rich interactive domain-containing protein 1A) are present in approximately 25% of bladder cancers (BC) and are associated with poor prognosis. With a view to discover effective treatment options for ARID1A-deficient BC patients, we set out to identify targetable effectors dysregulated consequent to ARID1A deficiency. Integrative analyses of ARID1A depletion in normal organoids and data mining in publicly available datasets revealed upregulation of DNA repair and cell cycle-associated genes consequent to loss of ARID1A and identified CHEK1 (Checkpoint kinase 1) and chromosomal passenger complex member BIRC5 (Baculoviral IAP Repeat Containing 5) as therapeutically drug-able candidate molecular effectors. Ex vivo treatment of patient-derived BC tumoroids with clinically advanced small molecule inhibitors targeting CHEK1 or BIRC5 was associated with increased DNA damage signalling and apoptosis, and selectively induced cell death in tumoroids lacking ARID1A protein expression. Thus, integrating public datasets with patient-derived organoid modelling and ex-vivo drug testing can uncover key molecular effectors and mechanisms of oncogenic transformation, potentially leading to novel therapeutic strategies. Our data point to ARID1A protein expression as a suitable candidate biomarker for the selection of BC patients responsive to therapies targeting BIRC5 and CHEK1. Overall design: Transcriptomic profiles of patient-derived bladder cancer tumoroid (PDT) and normal bladder organoid (NO) lines were characterised by quant-seq on total RNA. Additionally, the effect of ARID1A-depletion was investigated by comparing ARID1A-depleted normal organoids vs. parental or shRNA controls. Chromatin accessibility profiling by ATAC-seq was performed on a patient-derived tumoroid line depleted for ARID1A or E2F8 vs. shRNA control.