Identifying Cell State Vulnerabilities in Pancreatic Cancer

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy that is one of the most treatment-resistant forms of cancer. Genomic studies of PDAC have demonstrated profound transcriptional cell state heterogeneity within both primary and metastatic PDAC tumors. Moreover, preclinical experiments and recent single-cell transcriptomic studies of human PDAC samples have defined the mesenchymal cell state and the process of epithelial-to-mesenchymal transition (EMT) as an important cell state alteration that occurs in the context of chemotherapy and RAS inhibitor treatment. EMT in cancers has been associated with malignant cell invasion and metastasis. In addition, the mesenchymal state has been strongly correlated with resistance to both chemotherapy and targeted therapies across multiple cancer types. Identifying strategies to eliminate or reverse the treatment-refractory mesenchymal cell state may yield important therapeutic opportunities. To identify novel mesenchymal-selective dependencies, we performed an integrative analysis of diverse genomic, transcriptomic, functional genetic and small molecule screening datasets, using transcriptional signatures to stratify epithelial- and mesenchymal-selective phenotypes. Additionally, we probed genetic dependencies and cell surface targets using CRISPRCas screening and multi-omic data, including single-nucleus and spatial transcriptomic data from patient tumors. These studies have nominated mesenchymal state-selective vulnerabilities,including CDP-Diacylglycerol Synthase 2 (CDS2) as a novel mesenchymal-selective synthetic lethal target relationship with mesenchymal cancer cell identity. Using preclinical models, we investigated CDS2 as a paralog synthetic lethal and credentialed this protein as a new therapeutic target for the drug-resistant mesenchymal cell state. Overcoming resistance to therapies in cancer is vital to treating PDAC patients. Our lab identified that mesenchymal cell state and EMT associated with baseline resistance to KRASG12D inhibitors in PDAC cell line cultures. To uncover key drivers of resistance and discover potential combination therapy strategies, we performed genome-scale CRISPR-Cas RAS inhibitor anchor screens in the presence of or absence of RAS inhibitor and identified genes involved in sensitivity and resistance to RAS inhibition in mesenchymal PDAC cell lines. We identify previously observed pathways such as RTK signaling members and CRK-related components as well as the novel multi-subunit C-terminal to LisH (CTLH) complex across the sensitizers. Moreover, we demonstrate proof-of-concept data that GNB1 suppression conferred sensitivity to RAS inhibition, in the context of both pan-RAS (RMC-7977) and KRASG12D mutant-selective (RMC-9945) inhibition. Thus, these data suggest GNB1 as a potential candidate for combination therapy with RAS inhibition.