Abstract 5132: Pooled shRNA screens conducted in vitro and in vivo identify genes that are essential for the growth and/or survival of human tumor cells

Abstract

Abstract Identification of novel drug targets is a major challenge in cancer research. While numerous studies have identified genes that are mutated and/or amplified in cancers, the functional significance of most of these genetic alterations remains unknown. To directly identify genes that are required for the growth of human tumor cells, and that could potentially be targets for therapeutic monoclonal antibodies, we employed a custom shRNA library containing ∼13,000 shRNAs targeting 3,210 genes (∼4 shRNAs per gene) that encode cell surface or secreted proteins. Cultured tumor cells were infected with the virally-packaged shRNA library such that each shRNA (and an associated molecular barcode) was integrated into ∼200 cells. Cells that expressed shRNAs that target essential genes were progressively depleted from the population upon in vitro passage or following implantation into mice, resulting in decreased representation of those shRNAs/barcodes versus reference samples. In vitro screens were conducted in 3 colorectal cancer cell lines (HCT116, DLD1 and Colo320DM) and in NCI-H1975 lung cancer cells. About 1-3% of the genes in the library were identified as essential for tumor cell growth and/or survival in vitro. Importantly, these genes included positive controls such as PCNA that were included in the library to validate the screening procedure, as well as components of pathways that are known to control cell growth such as IGF-1R, Wnt9a, sonic hedgehog and RON. In vivo screens were performed with HCT116 and NCI-H1975 cells by implanting cells into mice immediately following shRNA integration. Representation of shRNAs versus the reference sample was assessed in small tumors (∼100 mm3) and in large tumors (1000 mm3). While we observed very few hits in small tumors, the number of hits in large tumors was greater than that observed in vitro (e.g., ∼3-fold higher for HCT116), suggesting that tumor growth in vivo is dependent on a larger set of genes than is cell growth in vitro. Interestingly, the genes that were uniquely essential in tumors versus in vitro included several that are involved in cell/cell adhesion (e.g., cadherin 24, claudin 3) and cell signaling (epidermal growth factor receptor). In summary, our shRNA screens focusing on secreted and cell surface targets have identified genes that may be essential for tumor growth, including many genes that have not previously been implicated in this process. Upon further validation, these genes could potentially serve as novel targets for therapeutic antibodies in cancer. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 5132. doi:1538-7445.AM2012-5132