EBV-Associated Gastric Cancer: From Initial Infection to Unique Therapeutic Approaches

Epstein-Barr virus (EBV) is a ubiquitous human herpesvirus that infects over 95% of the adult population. While infection is typically asymptomatic, in some individuals, particularly the immunocompromised, EBV is the causative agent of several cancers including lymphomas and epithelial cancers. Specifically, gastric carcinoma, nasopharyngeal carcinoma, and lymphoepithelioma-like carcinoma which occurs across multiple sites in the body, most notably, in the lung. EBV-associated gastric cancer (EBVaGC) is a unique subset of gastric cancer that makes up 10% of all GCs worldwide. EBVaGCs display an 80% rate of activating PIK3CA mutation and are also the most hypermethylated of any tumor type, displaying what is known as a CpG island hypermethylator phenotype (CIMP). EBV infection of B cells can be easily modeled in vitro using primary B cells and lymphoblastoid cell lines (LCLs). However, epithelial cell infection models have proven much more difficult to develop. Given these difficulties, EBV infection and outgrowth in epithelial cells is comparatively understudied and the process of tumorigenesis in vivo is poorly understood. In this dissertation, I developed methods to generate EBV infected epithelial cell lines derived from both gastric cancer and lung adenocarcinoma using diverse strains of EBV. I used these models to complete a CRISPR/Cas9 whole genome knockout screen to identify cellular restriction factors of infection and outgrowth. Together, these data will provide novel insights into the process of EBV infection and the dynamic interplay between virus and host during tumorigenesis. Furthermore, I have used these models to explore unique therapeutic approaches for EBV+ epithelial cancers. Specifically, I have characterized the lytic reactivation potential to histone deacetylase (HDAC) inhibitors and generated preclinical data supporting the use of HDAC inhibitors and the anti-viral ganciclovir for treatment of EBV+ epithelial tumors. Lastly, I have identified modulators of the response to a PI3Ka inhibitor in PIK3CA mutant gastric cancers. I found that loss of NEDD9 or inhibition of BCL-XL rendered cells hyper-sensitive to the PI3Ka inhibitor BYL719. Additionally, I found that loss of CBFB conferred resistance to BYL719 through up- regulation of the protein kinase PIM1 and defined the clinical utility of our data in the context of PI3K inhibition more broadly. The work outlined in this dissertation contributes to the study of EBV infection and tumorigenesis in the stomach as well as provides mechanistic insights into novel therapeutic approaches for EBV+ epithelial cancers and PIK3CA mutant gastric cancers.