Pan-cancer DNA Methylation Patterns And TP53 Mutation Related Epigenetic And Transcriptional Integration Analysis Based On TCGA Database

Cancer is a complex disease caused by abnormal cell proliferation,with biological characteristics such as abnormal cell differentiation,persistent cell proliferation signals,resist apoptosis,activation of invasion and metastasis.However,the mechanism of its occurrence and development is still unclear due to the highly heterogeneity among samples.The development of cancer is often accompanied by changes in genetic and epigenetic modification such as mutations in the driver genes,differential expression of genes,and aberrant DNA methylation.Explaining the regulatory relationships among genomes,epigenetic modifications,and transcriptomes in cancer patients is critical to our understanding of how cancer develops and it can also lay a theoretical foundation for future precision cancer therapy.This study aims to construct a regulatory relationship between driver gene mutations,DNA methylation,and gene expression in pan-cancer landscape,and to reveal the underlying mechanisms by which driver gene mutations affect cancer progression.This study is based on the DNA methylation array,WES and RNA-Seq data of eight cancer types in the TCGA database.We used the bioinformatics methods to conducted hierarchical clustering for samples in each cancer types,and performed the correlation analysis between the results of subtypes and clinical phenotype as well as molecular characteristicc.Then we screened out the driver gene mutations that affecting the global DNA methylation,and indentified the differentially methylated probes and differentially expressed genes which are associated with TP53 mutation.Finally,we constructed a regulatory relationship between the change of DNA methylation level and differentially expressed genes which are associated with TP53 mutation.The results are as follows:1.We obtained DNA methylation array data of 8 cancers from the TCGA database,and performed hierarchical clustering and divided patients into 2-4 cancer subtypes.Survival analysis indicated that the patients belongs to subtypes of LUAD and UCEC have a significant difference in prognosis.Chi-square test shows there are differences between patients in different subtypes with their clinical stages in BRCA,COAD,KIRC and THCA,the tumor mutation burden and tumor infiltrating lymphocyte are also significant different between subtypes in COAD,LUAD and UCEC.These results indicate that heterogeneity of DNA methylation is commonly between patients in each cancer types and may associated with clinical phenotypes and molecular characteristic.In addition,we use the Mutsig to screened out the driven genes in each cancer type,and performed Chi-square test with the subtyping results,we found there were significant differences in the mutation frequencies of some driver genes between different cancer subtypes,for example,the THCA driver gene BRAF has a mutation frequency of 82.5%(198 in 240 patients with the BRAF mutation)in THCA subtype 2 and only 7.35%(10 in 136 patients with BRAF mutation)in THCA subtype 1.These results indicate that mutations in driver genes were associated with patients ’ global DNA methylation levels.We eventually screened TP53 for downstream analysis because TP53 mutations can affect patients’ DNA methylation in five cancers.2.We divided cancer patients into mutation groups and non-mutation groups based on whether they have TP53 mutations.Then we performed differentially DNA methylation probes and differentially expressed genes analyses and obtained differential DNA methylation probes as well as differentially expressed genes that induced by TP53 mutation.We find that cg11804789,cg24205914,and cg26383138,as well as DCAF4L2,PAGE1,MAGEC2,WIF1,SPATA18,and PNMA5 that differed in a variety of cancers and showed consistent trends.KEGG pathway enrichment analysis shows that differentially expressed genes related to TP53 mutation are mainly enriched in Cell Cycle、DNA Replication and Neuroactive ligand-receptor interaction pathways.Among the differentially expressed genes,some genes are confirmed as target genes regulated by TP53,such as PTCHD4 in breast cancer and PGC in lung adenocarcinoma.At the same time,some differentially expressed genes are significantly correlated with TP53,such as ARHGAP26 in HNSC and DCDC2 in LIHC,which may be potential target genes of TP53.Finally,some potential tumor prognostic markers were obtained,such as the probe cg08698854 and gene ZNF541.3.We performed integrated analysis and finally constructed a regulatory relationship that between the change of DNA methylation level and differentially expressed genes which are associated with TP53 mutation.Some key probe-gene regulatory pairs were screened out,for example,the probe cg19003337 was demethylated in the TP53 mutated group in LIHC,which may cause the gene KRT17 up-regulated.Survival analysis results showed that the demethylation of the probe cg19003337 and the up-regulated of the gene KRT17 were significantly related to the poor prognosis of patients in LIHC,which further confirmed that the cg19003337 demethylated-KRT17 up-regulated relationship plays an important role in the cancer progression of LIHC.Conclusion: We found that there has strong heterogeneity among different samples in each cancer types,and the heterogeneity may be related to the mutation of driver genes.At the same time,we identified the differential DNA methylation probes and differentially expressed genes associated with TP53 mutations,and established a regulatory relationship between them,revealing the potential relationship between driving mutations and cancer progression.help us to further understand the occurrence and development of cancer,at the same time,potential tumor prognostic markers were screened to provide candidate markers for tumor diagnosis and prognosis.