Article Text
Abstract
Introduction The involvement of gastro-oesophageal-reflux-disease in the development of Barrett's metaplasia (BM) is well established. Recent evidence has begun to implicate bile acids such as the secondary bile acid deoxycholic acid (DCA) in BM and subsequent carcinogenesis. Our laboratory has identified the transcription factor retinoic acid receptor-related orphan receptor A (RORA) as a novel bile acid responsive gene with altered expression in BM through a novel integrative transcriptomic profiling approach. RORA, a putative tumour suppressor gene, has no established target hormone and is postulated to act independently from putative hormone binding. We now aim to validate RORA expression, induction and implications to oesophageal cell signalling.
Methods Oesophageal cell lines HET-1A (normal-squamous), QhTRT (metaplastic), GohTRT (dysplastic) and SKGT4(adenocarcinoma) were utilised. Expression of RORA was determined by semi-quantitative real-time RT-PCR (ABI) in BM (n=16) and control patients (n=14) and the above cell lines. RORA protein was examined by Western blotting utilising a polyclonal antibody specific to RORA (Abcam).
Results A significant reduction in RORA levels, as determined by real-time RT-PCR, was observed in a cohort of BM patients (p<0.001) consistent with previous genomic profiling. This pattern of expression was mirrored in oesophageal cell lines charting the carcinogenic sequence with HET-1A cells displaying the highest expression levels and lower levels observed in QhTRT, GohTRT and SKGT4 cells. DCA-mediated (300 um) induction of RORA expression was confirmed in all cell types by real-time RT-PCR and Western blotting. Isoform specific RT-PCR defined RORA4 as the predominant isoform (70%) expressed in oesophageal cells with small amounts of RORA1 (30%) and no detectable levels of either RORA-2 or 3. At the protein level 2 distinct bands matching both RORA4 and RORA1 were induced in oesophageal cells in agreement with isoform-specific RT-PCR data. Potential transcriptional targets of RORA were determined utilising informatic approaches generating a regulatory network of gene expression for RORA in response to DCA including: PER1, Nur77 and FOS (cell growth and circadian clock), NFkBIB, CXCL14 and COX2 (Inflammation) among others.
Conclusion This study for the first time demonstrates a loss of a critical regulator of cell growth, circadian rhythm and inflammation, RORA, in BM. Additional work described the ability of refluxed agents, such as DCA, to induce RORA in oesophageal cells. It is likely that RORA loss in BM may allow for dysregulated cellular growth, impact upon responses to nocturnal reflux through alterations in circadian rhythm and for resistance to refluxed induced damage in BM.