Ashani Lecamwasam, Boris Novakovic1,2, Braydon Meyer1,2, Elif I Ekinci3,4, Karen M Dwyer5, Richard Saffery1,2. 1. Epigenetics Research, Murdoch Children's Research Institute, Melbourne, Victoria, Australia. 2. Department of Paediatrics, University of Melbourne, Melbourne, Victoria, Australia. 3. Department of Endocrinology, Austin Health, Melbourne, Victoria, Australia. 4. Department of Medicine, University of Melbourne, Melbourne, Victoria, Australia. 5. School of Medicine, Faculty of Health, Deakin University, Melbourne, Victoria, Australia.
Abstract
BACKGROUND: We investigated a cross-sectional epigenome-wide association study of patients with early and late diabetes-associated chronic kidney disease (CKD) to identify possible epigenetic differences between the two groups as well as changes in methylation across all stages of diabetic CKD. We also evaluated the potential of using a panel of identified 5'-C-phosphate-G-3' (CpG) sites from this cohort to predict the progression of diabetic CKD. METHODS: This cross-sectional study recruited 119 adults. DNA was extracted from blood using the Qiagen QIAampDNA Mini Spin Kit. Genome-wide methylation analysis was performed using Illumina Infinium MethylationEPIC BeadChips (HM850K). Intensity data files were processed and analysed using the minfi and MissMethyl packages for R. We examined the degree of methylation of CpG sites in early versus late diabetic CKD patients for CpG sites with an unadjusted P-value <0.01 and an absolute change in methylation of 5% (n = 239 CpG sites). RESULTS: Hierarchical clustering of the 239 CpG sites largely separated the two groups. A heat map for all 239 CpG sites demonstrated distinct methylation patterns in the early versus late groups, with CpG sites showing evidence of progressive change. Based on our differentially methylated region (DMR) analysis of the 239 CpG sites, we highlighted two DMRs, namely the cysteine-rich secretory protein 2 (CRISP2) and piwi-like RNA-mediated gene silencing 1 (PIWIL1) genes. The best predictability for the two groups involved a receiver operating characteristics curve of eight CpG sites alone and achieved an area under the curve of 0.976. CONCLUSIONS: We have identified distinct DNA methylation patterns between early and late diabetic CKD patients as well as demonstrated novel findings of potential progressive methylation changes across all stages (1-5) of diabetic CKD at specific CpG sites. We have also identified associated genes CRISP2 and PIWIL1, which may have the potential to act as stage-specific diabetes-associated CKD markers, and showed that the use of a panel of eight identified CpG sites alone helps to increase the predictability for the two groups.
BACKGROUND: We investigated a cross-sectional epigenome-wide association study of patients with early and late diabetes-associated chronic kidney disease (CKD) to identify possible epigenetic differences between the two groups as well as changes in methylation across all stages of diabetic CKD. We also evaluated the potential of using a panel of identified 5'-C-phosphate-G-3' (CpG) sites from this cohort to predict the progression of diabetic CKD. METHODS: This cross-sectional study recruited 119 adults. DNA was extracted from blood using the Qiagen QIAampDNA Mini Spin Kit. Genome-wide methylation analysis was performed using Illumina Infinium MethylationEPIC BeadChips (HM850K). Intensity data files were processed and analysed using the minfi and MissMethyl packages for R. We examined the degree of methylation of CpG sites in early versus late diabetic CKD patients for CpG sites with an unadjusted P-value <0.01 and an absolute change in methylation of 5% (n = 239 CpG sites). RESULTS: Hierarchical clustering of the 239 CpG sites largely separated the two groups. A heat map for all 239 CpG sites demonstrated distinct methylation patterns in the early versus late groups, with CpG sites showing evidence of progressive change. Based on our differentially methylated region (DMR) analysis of the 239 CpG sites, we highlighted two DMRs, namely the cysteine-rich secretory protein 2 (CRISP2) and piwi-like RNA-mediated gene silencing 1 (PIWIL1) genes. The best predictability for the two groups involved a receiver operating characteristics curve of eight CpG sites alone and achieved an area under the curve of 0.976. CONCLUSIONS: We have identified distinct DNA methylation patterns between early and late diabetic CKD patients as well as demonstrated novel findings of potential progressive methylation changes across all stages (1-5) of diabetic CKD at specific CpG sites. We have also identified associated genes CRISP2 and PIWIL1, which may have the potential to act as stage-specific diabetes-associated CKD markers, and showed that the use of a panel of eight identified CpG sites alone helps to increase the predictability for the two groups.