Identification of synthetic lethal compounds targeting cancers with cohesin mutations

Abstract

The cohesin complex, comprising four core subunits, RAD21, SMC3, SMC1A, and STAG1/2, is important for the regulation of many biological processes, including sister chromatid cohesion, regulation of gene expression, chromatin architecture, DNA damage repair, and ribosome biogenesis. Germline mutations of the cohesin complex and its regulators cause developmental disorders, including Roberts syndrome (RBS) and Cornelia de Lange syndrome (CdLS). Whole-genome sequencing identified somatic mutations in cohesin genes in a wide range of human cancers, including acute myeloid leukemia (AML). Low expression of mutant cohesin transcsripts and protein are correlated with poor survival in AML patients. Despite this correlation, currently, there is no targeted strategy for cancer patients with cohesin mutations.

Synthetic lethality has emerged as a promising approach for targeted therapy. While synthetic lethal interactions of cohesin subunit, STAG2 have been demonstrated with its paralog, STAG1, and with drugs targeting the DNA damage repair pathway, synthetic lethal effects of other pathways via chemical perturbations have not been explored. In this project, I aim to identify synthetic lethal drugs that preferentially inhibit cells with cohesin deficiency.

Using CRISPR-Cas9, I generated three isogenic cohesin-deficient MCF10A cell lines, RAD21+/-, SMC3+/-, and STAG2-/-. Functional characterization of these cell lines showed that they are similar to the MCF10A parental line concerning growth rate, morphology, and chromosome stability. Analysis of nucleolar morphology of cohesin-deficient MCF10A cells using nucleolar markers, fibrillarin, and nucleolin, showed altered nucleolar phenotypes, resembling the nucleolar alterations seen in Roberts Syndrome. In support of this observation, RNA-sequencing analysis revealed downregulation of genes involved in RNA transcription and processing, in addition to dysregulation of cell cycle, DNA damage repair, signaling pathways, chromosome organization, and development. Using MCF10A parental cells and isogenic cohesin-deficient MCF10A cell lines, we performed a high-throughput drug screen with compound libraries of FDA-approved drugs, kinase, and epigenetics inhibitors. Synthetic lethal hits were ranked based on the differential area over the curve between MCF10A parental and cohesin-deficient MCF10A cells. Selected hits were then re-screened in a secondary screen to confirm their synthetic lethal effects in cohesin-deficient MCF10A. Inhibitors targeting PI3K/AKT/mTOR, Wnt signaling pathway, and epigenetic regulators are among the top hits identified in our synthetic lethal screen, demonstrating that cohesin deficiency creates druggable vulnerabilities in cancer cells. In summary, we have identified potential synthetic lethal compounds that may provide a new strategy for targeted therapies of cancers with cohesin deficiency or mutations.