Identification of differentially methylated genes in individuals with end-stage kidney disease attributed to diabetic kidney disease: Oral Presentation

A subset of individuals with type 1 diabetes mellitus (T1DM) are predisposed to diabetic kidney disease (DKD), which remains a leading cause of end-stage kidney disease (ESKD). Evidence suggests that epigenetic changes in DNA methylation may have a causal role in both T1DM and DKD, but robust studies in carefully phenotyped collections are unpublished. The aim of this investigation was to assess differences in blood-derived DNA methylation patterns between individuals with T1DM-ESKD (n=108) and individuals with long-duration T1DM and no evidence of renal disease upon repeat testing (n=254). We also compared DNA methylation profiles between individuals with ESKD, but no diabetes (n=52). All analyses were aligned for age, sex, ethnicity, and treatment(s). Blood samples were bisulphite treated before DNA methylation patterns from both groups were generated and analysed using Infinium MethylationEPIC BeadChip arrays for 862,927 sites. Differentially methylated cytosine-phosphate-guanine sites (dmCpGs) were identified (FDR adjusted p≤x10-8 and fold change ≥±2) by comparing methylation levels between groups with single site resolution. Genes were annotated based on Homo Sapiens hg19 genome build with gene function explored using GO and KEGG pathway enrichment analysis to identify key functions and pathways linked to the genes where the dmCpGs are located. When comparing individually matched cases who were receiving dialysis, or had a kidney transplant, against individually matched controls and subsequently against a larger control group with diabetes and no kidney disease, 41 genes were identified from both analyses. Top-ranked enrichment pathways included cardiovascular system development, focal adhesion, ErbB signalling and RAS signalling. When the case analysis was restricted to individuals with T1DM who had received a kidney transplant, CpG sites within 49 genes were identified, with enriched pathways including melanoma, Notch signalling, and Hippo signalling. When individuals with ESRD but no diabetes were directly compared to individuals with no evidence of renal disease or diabetes, dmCpGs were identified in >100 genes. Enriched pathways for these included melanoma, Alzheimer’s disease, P13K-Akt and RAS signalling pathways. Epigenetic alterations provide a dynamic link between a person’s genetic background and their environmental exposures. We have used a robust approach in carefully phenotyped individuals to identify leading DNA methylation biomarkers associated with ESKD, revealing several genes uniquely associated with ESKD in individuals with T1DM. Several top-ranked dmCpGs support previously identified kidney genes, while others represent novel genes including BMP7, CELF2, EBF4, H1FX-AS1, IFIT2, PHF15, PTPRN2, RGMA, SLC9A2, and TOM1.