The Role And Underlying Mechanisms Of MiR-92a In The Activation Of CDX2in Gastric Intestinal Metaplasia

IntroductionGastric intestinal metaplasia(IM) is one of the most common precancerous conditionof gastric cancer. The general evolution model for gastric cancer is from the chronicatrophic gastritis, IM, to atypical hyperplasia, among which IM is very important and evenan integral part. However, the mechanism of gastric IM has not yet been elucidated. It hasbeen suspected that bile acid reflux plays a role in the pathogenesis of IM. Statistics foundthat the risk of IM in patients with reflux is11times higher than those without. Clinicalstudies also show that the bile acid reflux may be the main cause of esophageal IM.Transcription factor caudal-related homeobox2(CDX2) belongs to the Caudal hoxgene families. It can promote the transcription of a variety of intestinal genes, specifically plays an irreplaceable role in maintaining intestinal cell proliferation, development anddifferentiation. CDX2activation is considered to be the key factor of IM. Studies revealedthat IM tissues started to express CDX2, which was not expressed in normal gastricmucosa. Animal experiments found that the gastric mucosa of CDX2transgenic micecompletely transformed to IM. However, mechanisms for the abnormal CDX2expressionin IM have not yet been elucidated.MicroRNAs (miRNA) is a kind of important molecules widely expressed in theeukaryotes. MiRNAs can decrease expression of target mRNAs by binding to their3′-UTRs, leading to mRNA degradation, or translational inhibition. MiRNAs play animportant regulating role in the process of digestive tract occurrence, development anddifferentiation. For gastric IM, a pathological change closely related to cell differentiationand tissue development, is the miRNA also involved in its regulation?We used bile acids to stimulate normal gastric epithelial cells and successfullyestablished the IM cell model. The miRNA expression profile of the IM model wasdetermined using the miRNA array. Among the differentially expressed miRNAs, miR-92awas the most highly upregulated. We carried out functional studies of miR-92a and foundthat it could increase the expression of CDX2. Thus, the questions that whether miR-92a isinvolved in the regulation of CDX2in IM and what are its target genes remain to besolved.Aims1. To establish, select and verify the bile acid induced IM model.2. To screen thedifferentially expressed miRNAs between bile acids stimulated and the unstimulated cells.3. To establish the miR-92a functional gain and loss models.4. To select and verify thetarget genes of miR-92a.Methods1. Normal gastric mucosa epithelial cells were exposed to different bile acids. Then theReal Time-Polymerase Chain Reaction (qRT-PCR) was uesd to select the optimizedstimulus time and concentration. Under this condition, we carried out western blot and immunofluorescence to test the bile acid induced IM model.2. MiRNA microarray was employed to profile miRNA expressions between gastricepithelial cell GES-1and cells stimulated with bile acids.3. Functional studies were carried out by transfecting miR-92a mimics and inhibitor.4. To screen the potential target genes of miR-92a by searching multiple miRNA targetdatabase.5. Target genes of miR-92a were verified by luciferase reporter gene、qRT-PCR andwestern blot.Results1. The bile acids stimulated gastric IM cell model was successfully establishedFour bile acids: cholic acid (CA), deoxycholic acid (DCA), Cheno deoxycholic acid(CDCA) and dehydro cholic acid (DHCA) were used to stimulate the immortalized gastricmucosal epithelial cells. At different time points (0h,3h,6h,9h,12h,24h,48h) thecells were harvested and analyzed, qRT-PCR experiment showed that the mRNAexpression of intestinal epithelial specific molecules KLF4, VILLIN and MUC2weresignificantly up-regulated in the experimental group. The ratio varied from20to1000times, and has positive correlation with the stimulus concentration and time. Horizontalcomparison between the four models revealed that CDCA exposure could significantlyelevate multiple intestinal specific molecules than the others. Longitudinal comparison ofthe time and concentration in the same model revealed that intestinal specific markersexpression reached the plateau at150μM and24hours. On this basis, we chose the CDCAand24hours as the final choice of IM model for further research. We used western blotand immunofluorescence respectively to verify the expression of KLF4, VILLIN andMUC2. It was found that these molecules were significantly increased after CDCAstimulation. These results suggest that when exposed to bile acids, gastric epithelial cellshas a tendency to transform to intestinal epithelial cells through a series of genotypechange.2. miR-92a was the most highly upregulated in the miRNA microarray dataMiRNA microarray data showed that1087miRNAs were up-regulated and120 miRNAs were down-regulated, of which miR-92a was the most highly upregulated. Theresults were then validated by applying qRT–PCR to test the expression of miR-92a in theCDCA induced GES-1and BGC823cells. Then we examined the expression of miR-92ain the GES-1and multiple gastric cancer cells including AGS、SGC7901、BGC823、MKN45、MKN28、GC9811and KATOⅢ. The result revealed that miR-92a was highlyexpressed in AGS, SGC7901and BGC823cells. MKN45, GC9811and KATOⅢ cells hada relative low level of miR-92a. Whereas it can be hardly detected in the MKN28cells.3. miR-92a could enhance the expression of CDX2We further carried out functional studies of miR-92a by transfecting mimics andinhibitor in the GES-1, BGC823and AGS cells, respectively. qRT-PCR experiment foundthat transfection of GES-1and BGC-823cells with miR-92a mimics could increase themRNA and protein levels of CDX2, while inhibitor transfection decrease the CDX2levels.The downstream intestinal marker of CDX2(S-I) also showed similar protein expressionlevel in the miR-92a functional gain and loss model. These results show that miR-92a maybe playing a role in gastric IM by raising the expression of CDX2.4. miR-92a inhibit the expression of Foxd1by the binding to its3’-UTRTo study the mechanisms implied in the upregulation of CDX2caused by miR-92a,we combined using two independent database miRwalk and miRanda to predict potentialtarget genes of miR-92a. The intersection parts of the above two databases showed163kinds of potential target genes. Through further gene functional analysis, we focused ourconcentrations on the forkhead box family member Foxd1. Because its family membersplay important roles in the process of growth and development, and their abnormalexpression are closely related to the occurrence of gastric cancer. We then use luciferasereporter gene assay and confirmed that miR-92a directly bind to the3’-UTR of Foxd1. Wefurther transfected GES-1cells with miR-92a mimics and found that miR-92a can inhibitFoxd1protein expression level, but cannot influence its mRNA level. Taken together,these results suggest that miR-92a regulate the expression of Foxd1by targeting its3’-UTR. ConclusionsmiR-92a affects the gastric mucosa epithelial cells as a endogenous regulating factor.It may mediate the up-regulation of CDX2through negatively regulate the expression ofFoxd1. This will help to provide a new theoretical basis for elucidating the mechanism ofgastric IM.