Dietary and protein modifiers of colitis-associated carcinoma: selenium, selenoproteins, and myeloid translocation genes and their impacts on inflammation, stem cell properties, and oxidative stress

Abstract

Inflammatory bowel disease (IBD), which affects 1 in 600 Americans, is characterized by severe and chronic inflammation, a known contributor to cancer. As such, the risk for cancer is increased in patients with IBD compared with the general population, and cancer is the most significant cause of mortality in IBD. This thesis will identify modifiers of colitis-associated cancer (CAC) utilizing a murine CAC model applied to models of dietary and genetic deficiencies. First, this thesis will focus on dietary selenium and two plasma selenoproteins as suppressors of tumorigenesis. Selenium deficiency results in increased inflammation in response to an inflammatory model utilizing dextran sodium sulfate (DSS) which, when coupled with the use of an initiator, azoxymethane (AOM/DSS inflammatory carcinogenesis model), leads to an increase in tumorigenesis concomitant with an increase in DNA damage. Knockout of the plasma selenoprotein glutathione peroxidase-3 (Gpx3) also results in increased tumor number as well as a higher degree of dysplasia. Finally, I demonstrate that selenoprotein P (Sepp1), which is considered to have both selenium transport and antioxidant properties, is a bimodal modifier of CAC. Complete loss of Sepp1 leads to decreased tumorigenesis as the result of clearance of initiated cells while haploinsufficiency leads to an increase in tumorigenesis which is mimicked by loss of either the putative selenium transport or antioxidant domains of Sepp1. A second set of CAC modifiers identified in this thesis are the myeloid translocation genes (MTGs). The MTGs are a family of transcriptional corepressors that were originally identified as targets of chromosomal translocations in acute myeloid leukemia. This thesis will first identify MTG, related-1 (MTGR1) as a modifier of inflammation which, upon its loss, leads to a decrease in inflammatory carcinogenesis. Moreover, I identify an interaction between MTG16 and the transcriptional repressor Kaiso. Knockout of MTG16 alone leads to increased tumorigenesis while knockout of Kaiso leads to unaltered tumorigenesis. Interestingly, absence of both proteins leads to rescue of the MTG16 phenotype, suggesting that Kaiso is epistatic to MTG16.