Construction Of Prediction Model And Screening And Validation Of Related Long-chain Non-coding RNAs For Oral Cancer

ObjectiveTo evaluate the environmental factors,and develop the predictive model for the risk of oral cancer;to screen for differentially expressed long-chain non-coding RNAs in oral and paracancerous tissues,and validate the results with a larger sample size;to perform cellular functional assays for the pathogenesis of oral cancer in vitro.Methods1.Using a case-control study design,1012 oral cancer patients and 3087healthy controls were collected through face-to-face interviews.Unconditional logistic regression model was utilized to calculate odds ratios(ORs)and corresponding 95%confidence intervals(95%CI)for environmental factors.2.Based on the above independent factors,three types of predictive model were developed according to the ORs,?coefficients and nomogram scores,respectively.The predictive ability of these models was assessed by receiver operating characteristic curve.3.The differential expression of lncRNAs in three pairs of oral cancer tissue and paracancerous tissue were screened by Agilent Human lncRNA 4×180K microarray.Five lncRNAs were verified by qRT-PCR in 81 matched oral cancer tissues and paracancerous tissues.t-test or Wilcoxon rank-sum test was used to compare the differences between the two groups.The correlation between lncRNAs and clinicopathological features were analysed by?~2 test.4.Construction of overexpression recombinant plasmid of lncRNA ENST00000470447.1 and blank control plasmid.The effects of overexpression of lncRNA ENST00000470447.1 on Tca-8113 cell proliferation,migration invasion and apoptosis were evaluated by CCK-8,Transewell,cell scratch and flow cytometry.Results1.Smoking,alcohol drinking,cooking oil fume exposure(moderate or heavy),high oral hygienic index(poor oral hygiene)were risk factors of oral cancer.Tea consumption,high frequency intake of seafood,fish,fresh vegetables and fruits and eggs were significantly associated with decreased risk of oral cancer.2.The prediction model based on nomogram shows the best predictive ability(AUC=0.721,95%CI:0.702-0.739).The higher the score of nomogram model,the higher the risk of oral cancer.There was a multiplicative and additive interaction between nomogram scores and family history of cancer(OR multiplication=1.81,95%CI:1.58-2.08).3.There were 2617 differentially expressed lncRNAs(accounting for 8.97%(2617/29187)of the total lncRNAs),of which 280 were up-regulated and 2337down-regulated.4.TherelativeexpressionlevelsofENST00000470447.1,ENST00000412353.1 and ENST00000506106.1 in oral cancer tissues were lower than those in adjacent normal tissues(P<0.001).The verification results were consistent with the above results.5.The results of?~2 test showed that the expression level of ENST00000470447.1 was correlated with the pathological types and differentiation of patients with oral cancer(P<0.05).The expression of ENST00000506106.1 was correlated with pathological types and lymph node metastasis(P<0.05).6.Compared with the control plasmid,ENST00000470447.1 overexpression plasmid could reduce the survival rate,mobility rate and invasion rate of Tca-8113cells,and the number of early apoptotic cells increased significantly.Conclusions1.Smoking,alcohol consumption,cooking oil fume exposure and poor oral hygiene are risk factors for oral cancer.Drinking tea,high-frequency intake of seafood,fish,fresh vegetables and fruits and eggs are the protective factors of oral cancer.2.The prediction model for oral cancer risk based on Nomogram has the best predictive ability and application value.3.ENST00000470447.1?ENST00000412353.1?ENST00000506106.1were under-regulated in oral cancer patients.4.ENST00000470447.1 and ENST00000506106.1 were associated with the pathological types,differentiation and lymphatic metastasis in patients with oral cancer.5.Overexpression of ENST00000470447.1 can inhibit the proliferation,migration and invasion of Tca-8113 cells and promote its apoptosis,which play an important role in the carcinogenesis of oral cancer.