Expressions Of Long Non-coding RNAs In Gastric Cancer And Its Clinicopathological Significance

ObjectiveLong non-coding RNA (lncRNA) is a new class of regulative non-coding RNA, whichlength is larger than200nucleotides. Recent studies found that there are close relations betweenlncRNAs and tumors. To screen the existence of lncRNAs which are associated with gastriccancer, we investigate the expression of lncRNAs from gastric cancer tissues and their matchedadjacent non-tumorous tissues (ANT). Combined with clinicopathological factors of patientswith gastric cancer, we explore the clinicopathological significance of these lncRNAs, which areexpected to become new candidate tumor-markers of gastric cancer.Methods1. To screen the lncRNAs which were up-regulated and down-regulated over four-fold,Gene microarray was used to detect the difference of expression profile of lncRNAs betweengastric cancer tissues and ANT.2. Bioinformatics analysis was employed to predict the target genes of these lncRNAs.3. Real-time quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) wasused to verify the expression levels of lncRNAs in61gastric cancer tissues and their matchedANT. The difference of expression levels was analyzed.4. Combined with clinicopathological data of patients with gastric cancer, statistic analysiswas made. The correlations between the expression levels of lncRNAs and clinicopathologicalfactors such as tumor size, depth of invasion, lymphatic metastasis, TNM stage, Histologic grade,levels of serum CA72-4, CEA and CA19-9were assessed. The clinicopathological significanceof lncRNAs associated with gastric cancer was clarified.Results1. Human lncRNA microarray was adopted to detect the difference of expression profile oflncRNAs between gastric cancer tissues and their matched ANT. Results showed that lncRNAswhich up-regulated and down-regulated over four-fold were ten and nine, respectively.2. Bioinformatic technique was used to forecast the target genes of19kinds of lncRNAs.We found that the target gene of AK054978was RP1-177G6.2. ERGIC3and SPAG4were thetarget genes of FER1L4. Studies showed that RP1-177G6.2, ERGIC3and SPAG4wereassociated with a variety of tumors. 3. Using a real-time qRT-PCR method, the expression levels of FER1L4and AK054978were detected in61pairs of gastric cancer tissues and ANT. Compared to ANT,56of61(91.80%) cases revealed a decrease in the FER1L4expression level in gastric cancer tissues (P＜0.0001). On the contrary, increased expression level of AK054978in gastric cancer tissueswas presented in50of61(81.97%) cases (P＜0.0001).4. In this study, both AK054978and FER1L4were associated with tumor size, Histologicgrade, depth of invasion, lymphatic metastasis, TNM stage and vessel or nerve invasion (P＜0.05). Whereas, the factors such as gender, age, tumor location, levels of serum CA72-4, CEAand CA19-9had no significant correlation with AK054978and FER1L4(P＞0.05).ConclusionsUsing gene microarray analysis and real-time qRT-PCR method, we find the lncRNA-AK054978and FER1L4which are associated with gastric cancer. Compared to ANT, theexpression levels of AK054978significantly increase in gastric cancer tissues. On the contrary, theexpression levels of FER1L4significantly decrease. Combined with clinicopathological factors, thepatients who had higher expression of AK054978and lower expression of FER1L4tended to havelarger tumor size, poorer histologic grade, more invasive and metastatic ability, later TNM stage.We suspect that AK054978and FER1L4might have close association with gastric carcinogenesis.They are expected to become new tumor-marker and molecular target of gastric cancer.