DNA Methylation Analysis Of Whole Genome Promoter In Clear Cell Renel Cell Carcinoma

Objectives Renal cell carcinoma (RCC) is a malignant tumor with the highest mortality rates among urinary malignant tumors. The major histological subtype is clear cell RCC (ccRCC), accounting for 80%-90% of all RCC cases, most of which are diagnosed as middle stage or advanced cancer. As one of the important epigenetic regulation modifications, the DNA methylation rapidly gained researchers’ attention, because it occurs before precancerous lesion. In this study, we analyzed promoter differential methylation and identified biomakers for ccRCC and examined methylated levels of biomarker and mRNA levels of key genes. It aimed to explore the diagnostic value of methylation in ccRCC clinically and the relationship between the ccRCC promoter methylation pattern and its development.Methods 1. Bioinformation analysis:(1) CcRCC clinical data, HumanMethylation450 level 1 raw data and RNAseqV2 level 3 data were downloaded from TCGA database; (2) Patients who have suffered from other malignancies, who have received neoadjuvant therapy or who have unknown ethnicity were removed; (3) Based on inclusive tumor and adjacent samples, differential methylation analysis was performed by RnBeads package written in R language, and promoter methylation characteristic was analyzed; (4) Optimal promoters CpGs to differentiate ccRCC tumor tissues from adjacents were identified through the PAMR; (5) Differential expression analysis for ccRCC tumor and normal tissues was performed by Deseq package written in R language; (6) The genes involving promoter differential methylated CpGs were enriched by the analysis of Gene Ontology and KEGG pathways.2. Experiment and GEO datasets validation:(1) By pyrosequencing nineteen collected ccRCCs and matched adjacents and analyzing GEO dataset, we validated biomarker CpGs; (2) Through conducting real-time PCR and analyzing GEO datasets, we further explored mRNA levels of three genes with differential promoter methylation and differential mRNA expression; Sequenom Massarray was used to detect the methylated level of gene promoter; (3) We assessed the effect of promoter methylated levels to three validated genes on their mRNA levels through Spearman’s rank correlation coefficient; (4) We estimated the relationship between promoter methylated levels to validated genes and ccRCC prognosis through Log-rank test and Cox’s proportional hazards regression model.Results 1. The TCGA methylation and transcriptome data showed that: compared to adjacent, in the ccRCC, there are 526 hypomethylated CpGs and 460 hypermethylated CpGs within promoter (FDR< 1E-10,|Delta Beta|> 0.2), 138 hypomethylated genes and 35 hypermethylated genes within promoter (FDR< 0.05,|Delta Beta]> 0.2),1798 downregulated genes and 2641 upregulated genes (FDR< 0.05,|log2 FC|> 1); 2. cg1 1201447, cg25247520, cg13309012 and cg08995609 were identified as the best promoters CpGs to differentiate ccRCCs from adjacent tissues and were hypomethylated in ccRCC, their overall error rates of distinguishing the TCGA study samples is 1.8% and the AUC value of their combined diagnosis reached 0.997; cg08995609 was significantly hypomethylated in ccRCC tissues by pyrosequencing (P= 0.700E-3), GSE61441 showed that the four CpGs were significantly hypomethylated and AUC reached 0.996; 3. The genes involving promoter hypomethylated CpGs were mainly enriched in the pathways of type I diabetes mellitus, cytokine-cytokine receptor interaction and graft-versus-host disease and the biological processes of immune response and inflammatory response, etc. The genes involving promoter hypermethylated CpGs, however, were mainly enriched in the pathway of neuroactive ligand-receptor interaction and the biological processes of cell adhesion, signaling transduction, etc.4. TCGA data showed that CYP4B1 and RAB25 are both downregulated genes with promoter hypermethylation and CA9 is upregulated genes with promoter hypomethylation, the promoter methylated levels of the three genes had significant correlations with their mRNA levels. Further, the results of real-time PCR and 10 datasets from GEO indicated that the mRNA expression of RAB25 is downregulated and CA9 is upregulated and CYP4B1 has less differences in ccRCC; The result of sequenom massarray showed that RAB25 promoter was hypermethylated in ccRCC tissue (P= 0.0003).5. The Log-rank test indicated that ccRCC patients with low level methylated CA9 promoter had higher survival rates than those with high level methylated CA9 promoter through Log-rank test. When served as an independent prognostic factor, CA9 promoter methylation didn’t show a significantly statistical difference through Cox’s proportional hazards regression model (P=0.238).Conclusions 1. In tissue level, the combination of cg11201447, cg25247520, cg13309012 and cg08995609 is a potential diagnostic biomarker for ccRCC; 2.In ccRCC, RAB25 promoter hypermethylation depresses its mRNA expression level; 3. CA9 promoter methylation is not an independent prognostic factor for ccRCC; 4. In ccRCC, DNA mathylation might regulate the expression levels of key genes involving biological processes, such as immune response, cell adhesion and so on, which thus contributes to the development of the cancer.