Super-Enhancer Redistribution as a Mechanism of Broad Gene Dysregulation in Repeatedly Drug-Treated Cancer Cells

Abstract

Summary: Cisplatin is an antineoplastic drug administered at suboptimal and intermittent doses to avoid life-threatening effects. Although this regimen shortly improves symptoms in the short term, it also leads to more malignant disease in the long term. We describe a multilayered analysis ranging from chromatin to translation—integrating chromatin immunoprecipitation sequencing (ChIP-seq), global run-on sequencing (GRO-seq), RNA sequencing (RNA-seq), and ribosome profiling—to understand how cisplatin confers (pre)malignant features by using a well-established ovarian cancer model of cisplatin exposure. This approach allows us to segregate the human transcriptome into gene modules representing distinct regulatory principles and to characterize that the most cisplatin-disrupted modules are associated with underlying events of super-enhancer plasticity. These events arise when cancer cells initiate without ultimately ending the program of drug-stimulated death. Using a PageRank-based algorithm, we predict super-enhancer regulator ISL1 as a driver of this plasticity and validate this prediction by using CRISPR/dCas9-KRAB inhibition (CRISPRi) and CRISPR/dCas9-VP64 activation (CRISPRa) tools. Together, we propose that cisplatin reprograms cancer cells when inducing them to undergo near-to-death experiences. : Ovarian tumors acquire new phenotypic features when exposed to anticancer therapy, which helps them endure the struggle for survival and increase malignancy. Using a six-layered multiomic approach, Ma et al. find that selective super-enhancer redistribution is a major driver of the acquisition of new phenotypic features. Keywords: multi-omic approach, gene dysregulation, super-enhancer, cisplatin, drug resistance, ovarian cancer, GRO-seq, ribosome profiling, Seurat, PageRank