Search

Synthetic lethality occurs when a single gene alteration is compatible with cell viability, but an additional co-occurring genetic alteration leads to cell death. In the context of cancer therapy, synthetic lethality can occur through the inhibition of a target that is selectively essential to tumors harboring a specific genetic alteration. Gene paralog pairs represent one promising class of synthetic lethal cancer targets, wherein the function of one paralog is lost in tumor cells, rendering them dependent on the remaining paralog to carry out an essential cellular process. To identify essential gene paralog pairs as starting points for drug discovery programs, we mined publicly available CRISPR genetic loss-of-function data and associated molecular datasets collected across a diverse panel of cancer cell lines. We first identified pairs of gene paralogs where one paralog was essential in a subset of cell lines, and then filtered these genes based on function, known literature, enrichment in specific lineages and integration of external datasets. These efforts identified VPS4A as a synthetic lethal target in cancers harboring copy number loss of VPS4B. VPS4A and VPS4B are highly homologous AAA ATPases that carry out multiple essential cellular processes including nuclear membrane remodeling and endosomal membrane biogenesis. VPS4B loss occurs as a passenger deletion during loss of the tumor suppressors SMAD2 and SMAD4. Loss of VPS4B creates a genetic dependency on VPS4A to drive essential VPS4-dependent processes. VPS4B deletion occurs at a frequency of up to 3% in multiple solid tumor types including esophageal, head and neck, pancreatic and colorectal cancers. To further explore the potential of VPS4A as a therapeutic target in VPS4B-deleted tumors, we first validated the synthetic lethal relationship between VPS4A/B using isogenic cell line pairs. HCT116 cells with an engineered homozygous loss of VPS4B, but not wild-type HCT116 cells, showed profound cell kill in response to genetic silencing of VPS4A. Moreover, simultaneous siRNA-mediated knockdown of VPS4A and VPS4B resulted in cell death across a panel of cancer cell lines (e.g. H1975, Panc0403), while knockdown of either gene alone was compatible with cell viability. Encouraged by these results, we profiled several previously reported small-molecule inhibitors of VPS4A (e.g. DBeQ and MSC1094308) in a suite of biochemical assays. Notably, these molecules were inactive against VPS4A. We have discovered a novel series of VPS4A inhibitors and are advancing this inhibitor series through lead optimization. Potent, selective, and pharmacologically active VPS4A inhibitors are expected to be well tolerated and have strong single-agent activity in tumors bearing VPS4B homozygous deletions. Citation Format: Meredith Kuo, Jason Chen, Sacha Holland, Eugene Lurie, An-Angela Ngoc Van, Francesco Parlati, Tayna Santos, Eric Sjogren, Natalija Sotirovska, Susanne Steggerda, Andrew MacKinnon. Identification of novel VPS4A inhibitors for the treatment of VPS4B-deleted cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1816.