Mo1937 MicroRNA-192 Promotes Transition of Squamous to Columnar Phenotype: Novel Insights Into the Pathogenesis of Barrett's Esophagus

Abstract

Introduction: The gastroesophageal reflux of acid and bile can cause esophagitis, and inflammation can trigger epithelial-mesenchymal transition (EMT), the process in which epithelial cells acquire phenotypic features of mesenchymal cells. EMT also occurs during carcinogenesis. Chronic exposure to acid and bile salts has been reported to cause non-neoplastic Barrett's epithelial cells in vitro to undergo malignant transformation and acquire EMT characteristics. Noting that these EMT changes occurred in cultures of initially-benign epithelial cells devoid of inflammatory cells, we hypothesized that acid and bile alone (in the absence of inflammation and neoplasia) might trigger EMT in Barrett's metaplasia. To explore this issue, we assessed features of EMT in non-neoplastic and dysplastic Barrett's epithelial cells, and we studied the effects of acid and bile salts on EMT in non-neoplastic Barrett's cells. Methods: We used: a) 2 telomerase-immortalized, non-neoplastic Barrett's epithelial cell lines (BAR-T and BAR-T10); b) BAR-T cells treated for 20 weeks with intermittent exposure to acidic bile salts (BEC-20W), and c) 3 immortalized, dysplastic Barrett's cell lines (CPB, CPC and CPD). BAR-T cells were exposed to acidic media (pH 5.5) containing conjugated bile acids (400 μM) either once for 10-minutes or for 10 minutes TID for 3 days. We evaluated features of EMT using optic morphology, Western blot and immunofluorescence (IF) for E-cadherin (CDH1) and fibronectin1 (FN1) proteins, and cell migration assays. Results: Dysplastic and BEC-20W cells had lower levels of CDH1 (an epithelial cell marker) and higher levels of FN1 (a mesenchymal cell marker) than non-neoplastic BART cells. By IF, dysplastic and BEC-20W cells demonstrated lower levels of membranous CDH1 and higher levels of cytoplasmic FN1 than BAR-T cells. Dysplastic and BEC-20W cells had a spindle-shaped appearance by optic morphology, and exhibited significantly higher migration rates than BAR-T cells (Table). In BAR-T cells, a single 10-minute exposure to acidic bile salts decreased CDH1 and increased FN1 protein levels by 6 hours. Exposure to acidic bile salts for 3 days significantly increased migration rates of BAR-T (21.7±3.3 SEM to 45.1±3.6) and BAR-T10 cells (13.2±1.8 to 22.6±2.5). Conclusion: Non-dysplastic Barrett's epithelial cells exposed to acidic bile salts for 20 weeks show evidence of EMT, as do dysplastic Barrett's epithelial cells. In non-neoplastic Barrett's cells, brief exposures to acidic bile salts induce EMT features including decreased CDH1 and increased FN1 levels, and increased cell migration. These data suggest that the gastroesophageal reflux of acid and bile might induce EMT in Barrett's metaplasia, even in the absence of inflammation and neoplasia. Conceivably, reflux-induced EMTmight contribute to remodeling and carcinogenesis in Barrett's esophagus. Migration Rates (average cells ± SEM per ≤14 fields)