| Objective:Hepatocellular carcinoma(HCC)is the main pathological type of primary liver cancer in China,and it is common in patients with liver disease background.The onset of hepatocellular carcinoma is hidden,and most patients are diagnosed with advanced tumors at the time of diagnosis,with a poor prognosis.Early diagnosis of disease and early prognosis prediction can improve the effect of clinical intervention,thereby improving the prognosis of patients.Early screening of hepatocellular carcinoma mainly depends on abdominal color Doppler ultrasound and serum alpha-fetoprotein(AFP)examination.The prognosis prediction is mainly based on the pathological stage and clinical stage of the tumor.However,some patients have normal serum alpha-fetoprotein,and the complete pathological diagnosis results need to be obtained after surgery,which will affect the timing of clinical intervention.At present,DNA methylation has been deeply studied in the formation,development and metastasis of various tumor diseases,but the markers of single DNA methylation still lack high sensitivity and specificity in tumor diagnosis and prognosis prediction.Therefore,the purpose of this study is to screening and verify DNA methylation markers specific for hepatocellular carcinoma and establish a prognostic model for hepatocellular carcinoma based on N6-methyladenine(m6A)in order to improve hepatocellular carcinoma early diagnosis rate and early prognosis prediction ability.Methods:1.Twenty patients with pre-operative serum alpha-fetoprotein<400?g/L were selected,postoperative pathologically confirmed hepatocellular carcinoma,and retain liver cancer tissue and adjacent tissues for high-throughput DNA methylation chip sequencing.Screen and verify differential methylation sites.2.Analyze the differential expression of m6A gene through the transcriptome data of hepatocellular carcinoma transcriptome downloaded from TCGA.To identify the relationship between the expression level of these genes and the prognosis of hepatocellular carcinoma,Lasso regression analysis was used to construct a prognostic model,and univariate and multivariate Cox regression analysis was used to confirm whether this model can be used as an independent prognostic factor.Finally,polymerase chain reaction(PCR)was used to verify the expression level of genes included in the model.Results:1.Methylation chip sequencing screened a total of about 865,000 methylation sites differentially expressed in liver cancer and adjacent tissues.According to P<0.05,|??|>0.1,there were 26301 significant hypermethylation sites and 219853 significant hypomethylation sites in liver cancer tissues.Among hypermethylated sites,81 candidate sites were screened based on ??>0.3 and paracancerous ?<0.1.Among the hypomethylated sites,66 candidate sites were selected based on ??<-0.35 and paracancerous ?>0.9.Based on the number of sites between cancer and paracancerous differences,9 sites were selected in hypermethylated sites.Bisulfite sequencing PCR and Sanger sequencing verified that the methylation markers of LDHB,CELF2 and MTHFD2 genes were found in most liver cancers.There is a significant difference between tissues adjacent to cancer.2.We confirmed 13 m6A genes.There are 50 normal liver tissue samples and 374 liver cancer samples in TCGA's hepatocellular carcinoma data;according to P value<0.05,11 of 13 genes are differentially expressed,and all the 11 m6A genes are highly expressed in liver cancer.According to the gene expression level,the samples were analyzed by consensus cluster plus analysis,and 374 liver cancer samples were divided into two groups;survival analysis found that the m6A related gene high expression group had a worse prognosis(P=6.197 × 10-4).Univariate survival analysis showed that 9 m6A genes were associated with prognosis of HCC patients.LASSO regression analysis included the five genes YTHDF2,YTHDF1,METTL3,KIAA1429,and ZC3H13 in the risk model.According to the risk score,the liver cancer samples were divided into high-risk and low-risk groups,and the prognosis of the high-risk group was found to be worse(P=1.72 × 10-4).The prognostic model was analyzed by ROC curve and AUC=0.617.Univariate(P<0.001,1.213(1.136-1.295))and multivariate Cox regression analysis(P<0.001,1.198(1.115-1.288))results showed that this model can be used as an independent prognostic factor for HCC patients.Conclusions:The detection of LDHB,CELF2 and MTHFD2 hypermethylated gene is expected to be an effective indicator for early screening of hepatocellular carcinoma patients with serum alpha-fetoprotein<400 ?g/L.m6A genes have a good predictive role in the prognosis of HCC.However,the specific molecular mechanism and the role in the occurrence,progression and metastasis of HCC will be important research directions in the future. |
PDF Full Text Request