Integrated Analysis of Blood DNA Methylation, Genetic Variation, Circulating Proteins, microRNAs, and Kidney Failure in Type 1 Diabetes

Research Background

Diabetic Kidney Disease (DKD) is one of the major complications of Type 1 Diabetes (T1D). Approximately 40% of T1D patients develop DKD, and 10% to 15% eventually progress to Kidney Failure (KF), requiring dialysis or kidney transplantation. Current clinical indicators are insufficient for predicting the onset of kidney failure, prompting the need for further research to uncover the underlying mechanisms, develop targeted therapies, and identify early warning biomarkers for timely intervention.

DNA methylation (DNAm) is one of the most stable epigenetic modifications, typically occurring at cytosine-guanine dinucleotides (CpG sites). Previous studies have shown that changes in DNA methylation are related to diabetes and its complications. However, the specific mechanisms linking these DNA methylation changes measured in blood cells to the risk of kidney failure are still unclear. This uncertainty hinders the potential for personalized medical applications based on blood cell DNA methylation.

To address these issues, the authors conducted a comprehensive study integrating multiple omics data, such as DNA methylation, genetic variation, circulating proteins, and microRNAs, to investigate their associations with kidney failure.

Authors and Sources

The paper was authored by Zhuo Chen, Eiichiro Satake, Marcus G. Pezzolesi, and others, from institutions including City of Hope, Joslin Diabetes Center, University of Utah, NIH, Folkhälsan Research Center, Helsinki University, Monash University, and University of Virginia. The research findings were published in the May 22, 2024 issue of Science Translational Medicine.

Research Methods and Processes

Research Design

The study included 277 T1D patients with diabetic kidney disease at the start of the study, with 142 (51.3%) progressing to kidney failure during a follow-up period of 7 to 20 years. The researchers measured DNA methylation levels in blood cells using Illumina’s Epic chip and conducted a genome-wide analysis using Epigenome-Wide Association Studies (EWAS). Additionally, they analyzed the integration of DNA methylation with genetic variation, circulating proteins, microRNAs, and the risk of kidney failure.

Research Process

1. Sample Collection and DNA Methylation Measurement: DNA was extracted from whole blood samples of 277 DKD patients, and DNA methylation levels were measured using the Illumina Infinium MethylationEpic BeadChip.

2. Epigenome-Wide Association Studies (EWAS): A univariate Cox proportional hazards model was used to analyze the association between methylation levels of 846,816 CpG sites and the risk of kidney failure, identifying potential candidate methylation sites. Multivariate Cox models were then used to further verify the independent association of these candidate sites with kidney failure risk.

3. Multi-Omics Integration Analysis: The integration of DNA methylation, genetic variation, circulating proteins, and microRNAs was analyzed to understand the mechanisms of interaction among genes, proteins, and microRNAs under methylation regulation. The temporal stability of DNA methylation levels was also assessed to verify its feasibility as a predictive biomarker.

4. Predictive Model Construction: Based on the results, a new predictive model combining DNA methylation data was constructed and compared with traditional models containing only clinical variables to evaluate their predictive performance.

Study Results

1. Association of DNA Methylation with Kidney Failure Risk: EWAS analysis identified 17 DNA methylation CpG sites associated with kidney failure risk. The methylation levels of these sites remained highly stable over time, suggesting their potential as stable predictive biomarkers.

2. Stability of DNA Methylation at CpG Sites: In early and late blood samples from 68 patients (half of whom progressed to kidney failure during follow-up), no significant differences were found in the methylation levels of the 17 kidney failure-associated CpG sites, further validating their stability.

3. Association of Genetic Variation with DNA Methylation: Integration of genetic variation data with DNA methylation data revealed that the methylation states of seven CpG sites were significantly influenced by genetic variation. Some of these variation sites have been reported in other studies to be associated with changes in kidney function, indicating that these variations may affect the risk of kidney failure by regulating methylation levels.

4. Mediating Role of Circulating Proteins and microRNAs: Linear regression and mediation analysis revealed that DNA methylation indirectly affects the occurrence of kidney failure by regulating the expression of circulating proteins and microRNAs. For example, the methylation of the cg12075771 site may increase the risk of kidney failure by regulating the expression of various circulating proteins such as KIM1 and DLL1.

5. Optimization of Predictive Models: The new predictive model incorporating DNA methylation data showed higher accuracy in predicting kidney failure risk (C-statistic=0.93) compared to the traditional clinical predictive model (C-statistic=0.85).

Study Conclusions

This study elucidates the mechanisms by which DNA methylation influences the occurrence of kidney failure in T1D patients through genetic variation, circulating proteins, and microRNAs via integrated multi-omics analysis. It extends the understanding of DKD progression mechanisms and demonstrates the potential application of blood cell DNA methylation as a predictor of kidney failure risk. Future research can further validate these findings and develop personalized intervention strategies based on these biomarkers.

Study Highlights

• Multi-Omics Integration: This study is the first to combine DNA methylation, genetic variation, circulating proteins, and microRNAs data to systematically investigate their relationships with kidney failure risk.
• Optimization of Predictive Models: The introduction of DNA methylation data significantly improved the accuracy of kidney failure risk prediction models, offering new possibilities for early clinical warning and intervention.
• Stability of Biomarkers: The study validated the temporal stability of methylation levels at 17 kidney failure-associated CpG sites, enhancing their potential application as long-term predictive markers.

This research provides new insights and tools for early warning and prevention of kidney failure in T1D patients, holding significant implications for personalized treatment of diabetes and its complications.