Epigenetic regulation of core genes linking to diabetic nephropathy progression: Lesson from FinnDiane type 1 diabetes study

Reduced methylation at key transcription factor binding sites might contribute to the progression of diabetic nephropathy in type 1 diabetes through the activation of insulin signaling and other pathways.

Diabetes is prevalent worldwide, and approximately 40% of individuals with diabetes develop diabetic kidney disease (DKD), the major cause of chronic kidney disease (CKD) leading to end‐stage kidney disease (ESKD). Since diabetic kidney disease is closely associated with a poor prognosis, overcoming it is an urgent challenge. In people with diabetes, the progression of albuminuria is a vital pathogenic factor in a declining renal function. Recently, it has been proven that reduction of proteinuria by several treatments including sodium‐glucose cotransporter 2 (SGLT2) inhibitors and selective mineralocorticoid receptor antagonists can retard the progression of diabetic kidney disease. On the other hand, even with such renoprotective agents, the residual risk remains a substantial burden. Thus, developing new models to stratify patients at high risk for progressive declining renal function would help to implement personalized therapies in such patients.

Large‐scale genome‐wide association studies (GWAS) have recently been conducted to identify genetic factors affecting diabetic nephropathy (DN). One recent GWAS encompassing almost 20,000 individuals with type 1 diabetes mellitus identified 16 risk loci including a missense mutation in the collagen type IV alpha 3 chain (COL4A3) gene ¹ . However, the complete picture of the progression of diabetic nephropathy by genetic factors remains elusive. Epigenetic mechanisms influence genome function without altering the underlying sequence. Deoxyribonucleic acid (DNA) methylation (the addition of a methyl group to the 5th carbon position of cytosine) is an important epigenetic feature that plays a key role in chromosomal integrity and the regulation of gene expression. Recently, epigenetic alterations have emerged as playing a role in the development and progression of diabetic nephropathy. Patients with a progressively declining estimated glomerular filtration rate (eGFR) displayed altered whole blood DNA methylation profiles at methylation sites corresponding to the genes involved in epithelial‐mesenchymal transition (EMT), oxidative stress, and inflammation ² . The identification of new epigenetic risk markers and biological pathways that affect diabetic nephropathy will provide clues to fight this devastating disease.

Focusing on this topic, in a recent publication in Journal of Clinical Investigation, Khurana et al. ³ reported a breakthrough observation by showing that the relevance of reduced DNA methylation at key transcription factors binding sites of genes in the progression of diabetic nephropathy. The authors performed methylation sequencing (methyl‐seq) on leukocyte DNA isolated from FinnDiane type 1 diabetes mellitus participants and clustering analysis of differentially methylated regions (DMRs) in patients with diabetic nephropathy, subsequently they also confirmed their findings in three independent type 1 diabetes mellitus cohort. They confirmed that the DNA methylation status could be differentiated into non‐diabetic control, type 1 diabetes mellitus without nephropathy, or type 1 diabetes mellitus with nephropathy (dependent on albuminuria status). They further found that reduced genomic methylation was a feature at CCCTC‐binding factor (CTCF) and DNA polymerase epsilon catalytic subunit B (Pol2B) binding sites. They identified DMRs with CTCF binding sites and identified seven core genes involved in the etiology of diabetic nephropathy: mammalian target of rapamycin (MTOR), regulatory‐associated protein of mTOR (RPTOR), and insulin receptor substrate 2 (IRS2) that belong to insulin signaling pathways; thioredoxin reductase 1 (TXNRD1), lecithin‐cholesterol acyltransferase (LCAT), and sphingomyelin phosphodiesterase 3 (SMPD3) comprising lipid metabolism; and collagen type I alpha 2 chain (COL1A2) involved in fibrosis. They also found that decreased core gene methylation correlated with decreased eGFR and increased urinary albumin/creatinine ratio (UACR) in four independent type 1 diabetes mellitus cohorts, indicating the significance of the methylation score as a biomarker to predict the decline of eGFR. To investigate whether methylation regulates the diabetic nephropathy pathways, they exposed renal podocytes, proximal convoluted tubules, and M1 macrophages to clinically relevant high glucose with or without methylation inhibitors and confirmed that high glucose‐diminished DNA methylation, increased the binding of CTCF and Pol2B on DNA with the induction of core genes relevant for the progression of diabetic nephropathy. The authors further analyzed the regulation of MTOR which is one of the insulin signaling pathway genes, using primary human aortic endothelial cells (HAECs) derived from individuals with type 1 diabetes mellitus. They confirmed there was reduced methylation of MTOR gene, enhanced CTCF and Pol2B binding, and increased MTOR exon 7 mRNA levels, and the levels of mTOR phosphorylation at Ser2448 in a hyperglycemic mimicking condition.

As described above, Khurana et al. ³ provided novel findings in the pathogenesis of the progression of diabetic nephropathy by analyzing several large cohorts with type 1 diabetes mellitus with methyl‐seq, a method with a significant advantage for higher genome coverage compared with conventional methods. The information could lead us to the discovery of clues to combat diabetic nephropathy. However, at the same time, there are several questions about their findings. First, they showed differential clusters of DMRs among healthy individuals, patients with type 1 diabetes mellitus without nephropathy, or patients with type 1 diabetes mellitus with severity dependent nephropathy. In regard to this, the classical pattern of the development of nephropathy in type 1 diabetes mellitus is determined by the linear progression of urine albumin levels from normo→micro→macro. However, when looking at the cluster of DMRs between groups, most of the DMRs that each group displays are quite unique and exhibit group specific patterns but not rare overlapping phenomena. The question arises as to whether the continuity of DMRs alteration is associated with the development of nephropathy. They also showed that the magnitude of the %DNA methylation was associated with GFR or UACR. This may suggest that some individuals are prone to the development of severe disease depending on their own methylation status before the onset of the specific disease severity. Another question is how the information from this report can be applied to the worldwide pandemic of DKD, including type 2 diabetes mellitus, which accounts for the remaining 90% of diabetes. The prevalence of CKD is higher in type 2 diabetes mellitus than in type 1 diabetes mellitus. In particular, it emerges clearly that poor glycemic control itself may not be a prerequisite or the dominant factor exacerbating CKD in type 2 diabetes mellitus patients; however, factors such as abdominal obesity, insulin resistance, hyperinsulinemia, and inflammation could play an important role ⁴ . In addition, renal lesions associated with type 2 diabetes mellitus are affected not only by the classical course learned from the findings in type 1 diabetes mellitus, but also by factors such as arteriosclerosis and aging. Indeed, the CKD phenotype in type 2 diabetes mellitus is essentially altered by them. However, despite such limitations, the FinnDiane trial provides a long‐term follow‐up of type 1 diabetes mellitus, and the findings obtained in this report are very important as a phenotype of diabetes associated with pure insulin deficiency.

Importantly, hyperactivation of the mechanistic target of rapamycin complex 1 (mTORC1) has been reported to be strongly associated with both proximal tubular epithelial cell and podocyte damage in DKD ⁵ . SGLT2 inhibition contributes to renoprotection by ameliorating the hyperactivation of mTORC1 in DKD via the suppression of glycolysis ⁶ , or elevated ketone body levels ⁷ . Furthermore, an analysis of single cell ribonucleic acid (RNA) sequencing data from kidney biopsy samples from juvenile type 2 diabetes mellitus with obesity revealed that SGLT2 inhibitors suppressed glycolysis‐mTOR axis signaling, not only in the proximal tubule, but also in all segments of the nephron ⁸ . This finding may suggest that epigenetic regulation of mTOR is involved in renal protection by SGLT2 inhibitors. In Khurana's report ³ , hypomethylation is involved in aberrant activation of mTOR signaling, which is one of the key signals for DKD progression and it sheds new light on the molecular mechanisms of DKD.

In summary, the authors found that reduced DNA methylation at CTCF and Pol2B sites could represent diminished protection of core genes that may associated with the progression of diabetic nephropathy (Figure 1). Although some points require further investigation, the report by Khurana et al. ³ will lead to a better understanding of epigenetic mechanism of diabetic kidney disease progression and proposes the importance of the methylation index as a novel biomarker that may contribute to early therapeutic intervention in diabetic kidney disease.

Figure 1.

Epigenetic regulation of core genes associated with progression of diabetic nephropathy (DN) in type 1 diabetes mellitus. Reduced methylation at key transcription factor binding sites might contribute to the progression of DN through activation of insulin signaling and other pathways. COL1A2, collagen type I alpha 2 chain; IRS2, insulin receptor substrate 2; LCAT, lecithin‐cholesterol acyltransferase; MTOR, mammalian target of rapamycin; RPTOR, regulatory‐associated protein of mTOR; SMPD3, sphingomyelin phosphodiesterase 3; TXNRD1, thioredoxin reductase 1.

DISCLOSURE

KK is an Editorial Board member of Journal of Diabetes Investigation and a co‐author of this article. To minimize bias, they were excluded from all editorial decision‐making related to the acceptance of this article for publication. KK was supported by a grant from the Japan Society for the Promotion of Science to KK (22K08330), and the grant from The Ministry of Health Labour and Welfare (202112004A). KK is under a consultancy agreement with Boehringer Ingelheim. KK collaborated with Boehringer Ingelheim, Taisho Pharma and Kowa for a project not related to this manuscript. Shimane University was supported by funds from Boehringer Ingelheim, Mitsubishi Tanabe Pharma, Taisho Pharmaceutical, Ono Pharmaceutical, Kowa, Nipro, and Life Scan Japan. KK received lecture honoraria from Dainippon‐Sumitomo Pharma, Astellas, Astra Zeneca, Ono, Otsuka, Taisho, Tanabe‐Mitsubishi, Eli Lilly, Boehringer‐Ingelheim, Novo Nordisk, Sanofi, and Kowa. KK is the recipient of Japan Diabetes Society Carrier Development Award supported by Sanofi.