MicroRNA-296-5p Promotes Gastric Cancer Growth Through Suppressing CDX1

IntroductionCaudal-related homeobox is a group of transcription factor, which have major role inthe proliferation, differentiation and development of intestinal cells. CDX1begin toexpress during the postnatal8-10week (postnatal12day in mice), which mediates thetransformation of endoderm into the simple columnar epithelium, including four cell types:Absorptive cell, Paneth cell, Goblet cell and Endocrine cell. CDX family genes areexpressed specificly in intestine and colon, while not in gastric mucosa. They include twomajor members, CDX1and CDX2. The former expressed in colon while the latter oneexpressed in intestine. Pathologically, CDX1is believed to have effects on the inductionand maintenance of gastric IM. Studies have shown that transgenic mice overexpressingCDX1in the stomach had extensive gastric IM. CDX1have been considered to play a role in tumor development. Studies have foundthat CDX1downregulation is also associated with colorectal carcinogenesis. Loss ofCDX1expression has been confirmed in both colorectal cancer cell lines and clinicalsamples. CDX1may act as an anti-proliferation factor in colorectal cancer as studies haveshown that upregulation of CDX1inhibits the colon cancer proliferation. In addition,CDX1gene expression was found to be partly diminished in several clinical GC samplescompared with gastric IM tissues. To date, the function of CDX1in GC and themechanism underlying its downregulation remains largely unknown.MicroRNAs (miRNAs) are a class of small non-coding RNA, which bond the targetmessenger RNA (mRNA) through base pairing, leading to mRNA degradation and (or)translational inhibition, eventually resulting in the suppression of target protein levelspost-transcriptionally. miRNAs have been considered to be involved in many kinds ofgastrointestinal cancers, including gastric cancer. Those include tumor suppressor oroncogenes, which acts as an inhibitor or facilitator of tumor development. Thus whetherCDX1downregulation is connected with the miRNA expression disorder and whetherCDX1function on GC development is regulated by miRNA remain unclear.In this study, we investigated function of CDX1in GC proliferation and themechanism of its downregulation directed by microRNA. We verified that CDX1wasdownregulated in GC tissues compared with adjacent IM tissues. We further found thatCDX1overexpression inhibited tumor growth in vitro. In addition, we identifiedmiR-296-5p, which is reversely correlated with CDX1in GC cell lines and tissues, as aregulating factor of CDX1by targeting the3’-UTR. Then whether miR-296-5p participatein the downregulation of CDX1? How does miR-296-5p-CDX1axis regulate GC growth?These questions are the key points of this project.Objective1. To investigate the CDX1expression and function in GC;2. To screen the upstreammiRNA of CDX1and verify the relationship of their expressions in tissues and cell lines;3. To study the mechanism of CDX1regulation by miR-296-5p;4. To validate themiR-296-5p function on GC growth through suppressing CDX1;5. To explore the signal transduction pathway through which miR-296-5p-CDX1axis regulate GC proliferation.Methods1. To detect the expression of CDX1by immunohistochemistry in normal gastricmucosa, intestinal metaplasia and gastric cancer tissue microarray;2. To build lentiviralvector encoding CDX1and transfect into gastric cancer cells to establish gain-of-functionmodels, to investigate cell proliferation, cell cycle distribution and percentage of apoptosisthrough XTT assayand flow cytometry analysis;3. To screen the potential miRNA(miR-296-5p) by utilizing multiple miRNA target databases;4. To verify the correlation betweenmiRNA and CDX1using qRT-PCR and Western Blot in gastric cancer cell lines andclinical samples;5. To validate the effects of miR-296-5p on CDX1throughfluorescentluciferase reporter gene assay and Western Blot;6. To establish gain-of-andloss-of-function model by transfecting cells with miR-296-5p mimics and inhibitors, andthen investigate cell proliferation, cell cycle distribution and percentage of apoptosisthrough XTT assayand flow cytometry analysis;7.To investigate the role of miR-296-5pon CDX1and GC growth through co-transfection of miR-296-5p mimics and lentiviralvector encoding CDX1with or without3’-UTR, and then detect the proliferation changes;8. To test the key protein changes in cell survival by western blot and use phosphataseinhibitors to investigate if the very protein mediates the function of miR-296-5p-CDX1axis.Results1. CDX1was downregulated in GC tissues and inhibits GC cell growthTo examine the CDX1expression and its significance in GC development, wemeasured the expression of CDX1in primary GC tissue arrays. Immunohistochemistryshowed that, in30group of tissues, normal gastric mucosa did not express CDX1,whereas intestinal metaplasia tissues (IM) expressed high levels of CDX1. However,CDX1expression was found to be partly or completely diminished in GC tissues ascompared with adjacent IM tissues. We further investigate if CDX1loss was involved inthe increased growth of GC cells by a gain-of-function model of CDX1. XTT assaysshowed that cell growth was significantly decreased by the induced expression of CDX1 in AGS cells as compared with Lenti-Null AGS cells. Cell cycle assays also showed thatCDX1overexpression in AGS cells increased the G0/G1population and decreased the Sand G2/M population. Moreover, apoptosis assays found that the percentage of apoptoticcells in Lenti-CDX1AGS cells increased compared with Lenti-Null AGS cells. Takentogether, these results suggest that the downregulation of CDX1is correlated with gastriccarcinogenesis and that CDX1might inhibit GC cells growth by blocking cell cycleprogression and inducing apoptosis.2. miR-296-5p and CDX1were inversely correlated in primary GC tissues and celllinesWe further used bioinformatic method to identify the potential miRNAs regulatingCDX1. Eleven miRNAs were computationally predicted using two independent miRNAdatabases: TargetScan and MicroCosm. miR-296-5p was the only miRNA which had beenreported to be involved in carcinogenesis. We measured the miR-296-5p levels and thecorresponding levels of CDX1protein in various GC cell lines. Quantitative reverse-transcriptase PCR (qRT-PCR) and western blot results indicated that the endogenous levelof CDX1expression was inversely associated with that of miR-296-5p in6GC cell lines.CDX1protein levels were relatively higher in the MKN28, SGC7901and BGC823celllines, whereas the miR-296-5p levels were relatively lower in the other3cell lines. Whilerelatively lower CDX1expression and higher miR-296-5p expression were validated inthe AGS, GC9811and MKN45cell lines.Similar results were found in16pairs of GC and adjacent IM specimens. Western blotrevealed CDX1expression was partially lost in15of16primary GC tissues, with anaverage value of0.55, as compared, with adjacent IM tissues with an average value of1.0.qRT-PCR results demonstrated that miR-296-5p was significantly upregulated in15matched pairs of GC tissues and was inversely correlated with CDX1expression. Thusthese results demonstrate that CDX1expression was progressively decreased in GCtissues, compared with precancerous lesions, and that this loss was related to theupregulation of miR-296-5p.3. miR-296-5p downregulated CDX1expression by binding its3’UTR Our further results using luciferase reporter assays showed that miR-296-5p caused asignificant decrease in relative luciferase activity when the CDX1plasmid containingwild-type3’-UTR was present, but luciferase activity was not significantly changed in the3’-UTR with mutant binding sites. Moreover, qRT-PCR showed that overexpression orknockdown of miR-296-5p had no significant effect on CDX1mRNA levels. However,western blotting showed that overexpression of miR-296-5p significantly suppressedCDX1expression in MKN28cells and that silencing of miR-296-5p increased CDX1expression in AGS cells, indicating that miR-296-5p regulates CDX1at the post-transcriptional level.4. miR-296-5p promotes GC cell growth through inhibiting CDX1We established gain-of-function and loss-of-function models through transfection oftwo concentrations of miR-296-5p inhibitor or mimics into AGS cells and MKN28cellsrespectively. XTT assays showed that after miR-296-5p downregulation, AGS cell growthwas significantly supressed while MKN28cell growth was promoted by miR-296-5pupregulation. Further results indicated that the effects of the miR-296-5p inhibitor ormimic on GC cell growth were dose dependent. In addition, apoptosis analysisdemonstrated that miR-296-5p knockdown in AGS cells dose-dependently increased thenumber of apoptotic cells, and miR-296-5p overexpression decreased MKN28apoptoticcell. We further conducted flow cytometry analysis of the cell cycle and found thatrepressing miR-296-5p expression led to a G0/G1accumulation compared with the controlcells, whereas restoring miR-296-5pexpression in MKN28cells led to the opposite results.We further transfected AGS cells with a lentiviral vector encoding the CDS region ofCDX1with or without its3’-UTR and measured CDX1expression by western blot. AftermiR-296-5p was cotransfected into AGS cells, CDX1expression was found to besignificantly reduced in cells transfected with the wild-type3’-UTR, whereas nodifference was found in cells transfected with vector lacking the3’-UTR sequence.Furthermore, XTT assays demonstrated that restoration of miR-296-5p could significantlyrescue the CDX1-induced promotion of GC cell growth. In addition, we found thatmiR-296-5p expression was able to abrogate the G0/G1accumulation induced by CDX1. Apoptosis assays also suggested that miR-296-5p upregulation was sufficient to decreasethe percentage of CDX1-induced apoptotic cells. Taken together, these results suggest thatCDX1could be a functional target of miR-296-5p.5. miR-296-5p promoted GC growth through the miR-296-5p-CDX1-ERK axisTo further investigate the mechanism underlying GC growth regulation by themiR-296-5p/CDX1axis, we examined several key molecules in important dysregulatedpathways in carcinogenesis. Western blot showed that ERK1/2phosphorylation wassignificantly inhibited after knockdown of miR-296-5p in AGS cells, whereas AKT andNFkB phosphorylation had no visible change, and no significant changes were found inERK1/2, NFkB and AKT expression. Furthermore, MAPK-related proteins and theirphosphorylation in the gain-or loss-of-function of miR-296-5p and CDX1model system.Downregulation of miR-296-5p in AGS cells lead to a significant decrease in thephosphorylation of RAF-MEK-ERK pathway-related proteins without affecting theirexpression, which was also found in AGS cells transfected with the Lenti-CDX1vector.However, miR-296-5p upregulation or CDX1knockdown in MKN28cells elicited similarresults, as the phosphorylation of RAF, MEK and ERK were decreased as compared withthe negative control. The changes likely induced by the regulation of miR-296-5p andCDX1further indicated that miR-296-5p might affect GC cell growth through thedownregulation of CDX1. More importantly, we found that restoring miR-296-5p couldreverse the inhibition of ERK phosphorylation and downstream molecular changesinduced by CDX1with a3’-UTR; however, this phenomenon was not observed in AGScells transfected with vectors coding CDX1without a3’-UTR.We further used PD98059, an inhibitor of ERK phosphorylation, for MKN28cellstransfected with miR-296-5p mimics or CDX1lentiviral siRNA vector. Western blotshowed that PD98059inhibited the phosphorylation of ERK1/2without affecting CDX1expression. XTT assay demonstrated that PD98059abrogated the promotion of GCgrowth induced by miR-296-5p upregulation or CDX1downregulation. In addition, FACSassay found that the changes of cell cycle distribution caused by miR-296-5p mimics andCDX1siRNA could be rescued by PD98059. Apoptosis assay also showed PD98059was able to increase the apoptotic cells percentage, which was decreased in cells transfectedwith miR-296-5p mimics or CDX1siRNA. Collectively, these results suggested thatmiR-296-5p regulates ERK activity and promotes GC cell growth through themiR-296-5p-CDX1-ERK1/2axis.ConclusionCDX1is downregulated in GC compared with adjacent IM tissues. miR-296-5pnegatively regulated CDX1and promotes GC growth through activation of ERK1/2; Thisnew miR-296-5p-CDX1-ERK1/2pathway help to maintain and promote the proliferationof gastric cancer cells through regulating cell cycle and apoptosis, and may shed newlights on understanding of process from IM to GC and developing new therapeuticstrategy for GC.