Search

Selenium (Se) is an essential trace element incorporated into selenoproteins (SEPs), which provide a broad scope of antioxidant functions. Se status inversely correlates with inflammatory bowel disease (IBD) severity, and animal models have shown that Se deficiency worsens colitis and colitis-associated cancer (CAC). One of the most highly expressed and functionally critical SEPs is Selenoprotein P (SEPP1/SELENOP). SELENOP serves two known roles: 1) distributing Se throughout the organism, and 2) acting as a lipid peroxidase. SELENOP is also markedly reduced in IBD. Global SELENOP reduction augmented experimental CAC. SELENOP is primarily produced in the liver for tissue distribution through the plasma; however, liver-specific SELENOP deletion failed to modify CAC, thus implicating epithelial SELENOP as a primary disease modifier. As the roles of intestinal epithelial SELENOP in colitis and CAC remain unknown, we investigated these roles using the Vil-CreERT2 driver to inducibly delete intestinal epithelial SELENOP in AOM/DSS inflammatory carcinogenesis models. First, Selenop deletion was confirmed in intestinal epithelial cells of tamoxifen-induced Vil-CreERT2; Selenopfl/fl mice (ΔIEC). Epithelial-specific SELENOP loss promoted tumor initiation, as evidenced by greater tumor multiplicity visible by colonoscopy in ΔIEC mice. Upon sacrifice, ΔIEC mice demonstrated increased colon tumor number and size. Additionally, ΔIEC colon tumors displayed greater apoptosis (cleaved caspase-3), DNA damage (pγH2AX), and dysplasia severity. Overall, these results are consistent with our model that epithelial SELENOP loss promotes tumor formation through increased intratumoral oxidative stress and resultant genomic instability. To further investigate SELENOP’s contributions in pure epithelial cultures, we determined that both CACO2 cells and human colonoids produce SELENOP in a sodium selenite-dependent fashion; furthermore, polarized CACO2 cells preferentially secrete SELENOP basolaterally, suggesting a local extracellular role for SELENOP in the immediate colonic epithelial microenvironment. Also, SELENOP knockdown in human intestinal organoids derived from ulcerative colitis (UC) patients increased oxidative stress, increased sensitivity to H2O2 treatment, and reduced organoid growth, implicating SELENOP in modulating the epithelial ROS-protective response. RNA-seq analysis from the PROTECT cohort revealed reduced SELENOP expression in pediatric UC biopsies that inversely correlated with disease activity and S100A8 (calprotectin) levels. Lastly, SELENOP RNA in situ hybridization of adult UC/Crohn’s Disease and CAC TMAs demonstrated decreased epithelial transcript in IBD and stage-dependent reductions in CAC. These data collectively indicate critical roles for SELENOP in epithelial and colonic microenvironment antioxidant defenses in IBD.

Patients with inflammatory bowel disease demonstrate selenium nutritional deficiencies and are at increased risk for colon cancer. Previously, we determined that global loss of the selenium-containing antioxidant protein, Selenoprotein P (SELENOP), exacerbates experimental colitis-associated cancer (CAC). Using a variety of tissue-specific knockout mouse lines, we further determined that epithelial-derived SELENOP served as the primary mediator of disease severity, and Villin-Cre; Selenop f/f (Selenop ΔIEC) mice have increased tumor incidence, number, size, and dysplasia as compared to WT cohorts. We next aimed to establish the mechanism by which epithelial-derived SELENOP modifies inflammatory tumorigenesis. We first investigated SELENOP’s effect on tumor promotion, and although Selenop ΔIEC tumors were larger, tumor cell proliferation was unchanged. Because global loss of SELENOP modified immune cell recruitment, we next characterized tumoral immunoinfiltrates by immunohistochemistry and seq-ImmuCC analysis of RNA-sequencing results. Neither metric identified significant changes in immune cell populations in either tumor or adjacent normal tissues. Further, histological endpoint analysis noted no change in degree of inflammation in Selenop ΔIEC colons at the end of the azoxymethane/dextran sodium sulfate protocol. Thus, epithelial-derived SELENOP likely does not augment inflammatory tumorigenesis via tumor promotion or inflammatory means by the measured parameters. We next hypothesized that loss of epithelial SELENOP modifies tumor initiation, and in support of this, analysis of endoscopy videos identified increased tumor numbers at early stages of experimental carcinogenesis. Tumors from Selenop ΔIEC mice also showed increased staining for markers of apoptosis and DNA damage, suggesting earlier tumor initiation may be due to increased oxidative stress. Further, myeloid-specific Selenop knockout, which did not alter inflammatory tumorigenesis, upregulated other antioxidant selenoproteins which may compensate for SELENOP loss, but this increase was not observed in the Selenop ΔIEC model. Interestingly, query of the Predicting Response to Standardized Pediatric Colitis Therapy (PROTECT) cohort indicated that SELENOP is downregulated in UC tissues, with greater downregulation in severe disease, but levels of other key antioxidant selenoproteins were again unchanged. Finally, to more clearly investigate SELENOP loss in epithelial cells, SELENOP was knocked down in human adult ulcerative colitis organoids. Here, knockdown induced a pronounced growth defect, increased apoptosis, and increased oxidative stress as measured by CellRox staining intensity. Taken together, these results suggest a model in which epithelial SELENOP protects from CAC by buffering oxidative stress in colonic epithelial cells in the setting of an inflammatory microenvironment.