Epigenetics of diabetic nephropathy
Diabetes is responsible for a large proportion of chronic kidney disease and end-stage renal disease worldwide. Careful monitoring and balanced control of blood glucose for individuals with type 1 diabetes can delay or prevent the onset of micro- and macro-vascular complications, however hyperglycaemia does not explain all of the risk for these diabetic complications. Genetic risk factors for diabetic nephropathy are being identified through international collaborations. However despite these advances a significant proportion of susceptibility remains unexplained, the so-called ‘missing heritability’.
Epigenetics is a connection between our genes and the environment that is now being explored. Epigenetic features act on nucleic acids to affect gene expression, rather than directly changing the DNA sequence. These epigenetic features may be inherited or they can be changed by the environment in response to illness, diet, drugs, and exercise. Importantly, epigenetic changes caused by the environment may be passed down through multiple generations. Epigenetics is emerging as an important risk factor for complex diseases such as diabetic nephropathy. As epigenetic features may be dynamic, they are
Figure 1: Evolution of epigenetics for diabetic nephropathy highlighting some key events and the increase in publications from PubMed 2007-2014.potentially modifiable and drugs that target the epigenome are already on the market for several diseases.
Complementary to genome-wide association studies (GWAS), it is now feasible and cost-effective to perform population-based studies of the epigenome, known as ‘epigenome-wide association studies’ (EWAS). Major epigenetic modifications include histone modification, RNA interference, and DNA methylation as discussed briefly below.
Histone proteins help to protect and compress DNA, by forming a tightly packaged DNA-protein complex ultimately known as chromatin. When tightly packaged, this compact unit is not accessible to the cell’s transcriptional machinery so gene expression cannot easily occur. Gene expression occurs where this chromatin package is more relaxed. Histones may have several covalent modifications such as methylation, acetylation, phosphorylation and sumoylation, which influence gene expression; for example, acetylation typically results in transcriptional activation.
Many studies have shown that histone modifications influence renal scarring and fibrotic pathways in animal models and human populations. Drugs that target histone modifications exist, but there has been limited investigation for diabetic nephropathy. In 2008, treatment with curcumin involved changes in post-translational modifications to protect against the development of diabetic nephropathy in an animal model. In 2014, treatment with losartan was shown to ameliorate diabetic nephropathy and reverse epigenetic changes in diabetic mice. Histone modifications have been reported in response to high blood glucose, with a 2014 report describing hyperacetylated promoters in genes related to diabetic complications, which are primarily induced by high glucose. This data from the Epidemiology of Diabetes Interventions and Complications (EDIC) open follow-up study of the Diabetes Control and Complications Trial (DCCT), demonstrated increased acetylation for eight genes in the intensively controlled group (controls) compared to those assigned to conventional therapy (cases); higher acetylation levels were observed in 37 genes in cases compared to controls.
A class of non-protein-coding RNAs known as microRNAs (miRNAs, ~ 22 nucleotides long) are being increasingly recognised as important regulators of gene expression which influence complex diseases. Many miRNAs have been associated with diabetic nephropathy across a variety of models and human studies; the majority studied in depth to date are involved in the TGFβ pathway. miRNAs are promising biomarkers, which may have therapeutic potential for complex diseases such as diabetic nephropathy.
Long, non-coding RNA (lncRNA) is also being shown to have regulatory functions including influencing gene expression, splicing, structural components for protein complexes, and activity or localisation of proteins. To date, many of those lncRNAs reported as affecting kidney disease are also involved in the TGFβ pathway. One function of lncRNA is inactivation of X chromosomes and skewing of female X chromosome inactivation has been associated with renal outcomes.
DNA methylation is one of the best studied epigenetic features and commercial arrays are now available that cover individual methylomes with single site resolution. DNA methylation is an important mechanism to maintain chromosomal integrity and enhance or repress gene expression with refined control. Promoters of genes that are methylated are typically silenced and not expressed. However, DNA methylation may be ‘activating’ for a protein-coding gene if it prevents binding of transcription factors or miRNA that would otherwise repress the protein-coding gene.
Specific DNA methylation profiles have been associated with diabetic nephropathy. One of the early, larger scale studies examined 192 genes using an established cell model for diabetic nephropathy, where cells were exposed to high glucose for a short time. This failed to identify strongly associated genes, which may be due to the fact that the genes under investigation in this study were truly not involved in disease, or that a longer exposure to stimuli is required to see epigenetic changes.
Bell and colleagues studied 14,495 genes in 192 individuals with and without diabetic nephropathy, identifying 19 differentially methylated regions, including the UNC13B gene. A later study of 14,000 genes focusing on diabetic patients with end stage renal disease revealed differences in 187 genes, of whom 39 were previously reported to affect kidney disease. Examining 485,577 unique sites in 255 individuals with CKD (cases, of whom 113 had diabetic nephropathy) and 152 individuals without evidence of renal disease (controls) revealed statistically significant differences in methylation of CUX1, ELMO1, FKBP5, INHBA-AS1, PTPRN2, and PRKAG2 genes, with several being supported by gene expression changes. Encouragingly, 2/3 of the top-ranked genes were also differentially regulated in kidney tubular epithelial cells when comparing 12 cases with chronic kidney disease to 14 healthy controls.
Differentially regulated regions have been observed in human kidney samples (primarily proximal tubule) assessed for genome-wide methylation. This revealed that differentially methylated regions for chronic kidney disease (including diabetic nephropathy) were more likely to be located in putative enhancer regions and affect transcription factor binding rather than gene promoters to modify gene expression.
There are multiple complications associated with epigenetic studies and this complex disease field is developing very fast. Some common issues to consider are described below:
Type of biological sample: Each individual’s inherited genome is fixed throughout life and consistent in every tissue. However, genes are typically expressed in different tissues and at different times throughout life, so each individual ‘epigenome’ is more variable. This raises questions as to the best cells for experiments to study the epigenome:
Blood is a good source of high quality DNA, may be obtained from a simple venepuncture, and may be already available from established collections and historical genetic studies. However, whole blood comprises many different types of white blood cells and each has their own ‘epigenome’.
Kidney tissue may be used, but kidney biopsies are invasive tests that carry a surgical risk to patients so this is not suited for routine biomarker testing. Kidneys also comprise multiple cell types with several epigenomes, even if using very small samples from laser capture microdissection.
Some established DNA collections for diabetic nephropathy use DNA from Epstein-Barr Virus transformed lymphoblastoid cell lines (LCL), however these should be used for epigenetic studies with caution. Approximately eight percent of genes in a small study (n=318 genes) revealed different methylation profiles in DNA from LCLs compared to those in DNA derived from peripheral blood in the same individuals.
Encouraging for larger-scale population-based studies, a recent paper observed that the majority of blood-derived DNA methylation signatures were also observed in kidney tissue, as well as reflecting changes in gene expression.
Storage of material: Genetic studies have been ongoing for many years and it is possible that DNA from established collections may be used for epigenetic investigations for diabetic nephropathy. However, the epigenome is affected by the method of DNA extraction and storage so it is critical that all studies use consistent extraction and storage protocols.
Type of study: Similar to genetic studies, epigenetic studies may be cross-sectional or longitudinal. Ideally epigenetic studies would investigate changes in a person’s epigenome over time, particularly before and after the development of diabetic nephropathy, but practically that is challenging given the relatively new feasibility of larger epigenetic investigations.
Significance, validation and replication: With increasing numbers of markers being assessed in a single study, the p value at which statistical significance is reported must take account of the multiple comparisons performed. Reporting of epigenetic studies are not yet standardised, but validation of results using a second technology and replication in an independent collection are considered good practice.
Epigenetic studies in the kidney are helping to reveal mechanistic changes associated with normal development and disease. The identification of a blood-based epigenetic signature may lead to earlier diagnosis through the development of a ‘risk’ profile and / or offer improved clinical evaluation and targeted treatment options.
^ Grossniklaur U, Kelly WG, Ferguson-Smith et al. Transgenerational epigenetic inheritance: how important is it? Nature Reviews Genetics 2013;14:228-35
^ Tikoo K, Meena RL, Kabra DG et al. Change in post-translational modifications of histone H3, heat shock protein-27 and MAP kinase p38 expression by curcumin in streptozotocin-induced type 1 diabetic nephropathy. British Journal of Pharmacology 2008;153(6):1225-31
^ Reddy MA, Sumanth P, Lanting L et al. Losartan reverses permissive epigenetic changes in renal glomeruli of diabetic db/db mice. Kidney International 2014;85(5):362-73.
^ Miao F, Chen Z, Genuth S et al. Evaluating the role of epigenetic histone modifications in the metabolic memory to type 1 diabetes. Diabetes 2014; Epub ahead of print.
^ Kato M, Natarajan R. MicroRNA circuits in transforming growth factor-β actions and diabetic nephropathy. Seminars in Nephrology 2012;32(3):253-60.
^ Alvarez ML, DiStefano JK. Towards microRNA-based therapeutics for diabetic nephropathy. Diabetologia 2013;56(3):444-56.
^ Zhou Q, Chung AC, Huang XR et al. Identification of novel long noncoding RNAs associated with TGFb-SMAD3-mediated renal inflammation and fibrosis by RNA sequencing.
^ Brennan EP, Ehrich M, O’Donovan H, et al. DNA methylation profiling in cell models of diabetic nephropathy. Epigenetics 2010;5(5):396-401.
^ Bell CG, Teschendorff AE, Rakyan VK et al. Genome-wide DNA methylation analysis for diabetic nephropathy in type 1 diabetes mellitus. BMC Med Genomics 2010;3:33
^ Sapienza C, Lee J, Powell J et al. DNA methylation profiling identifies epigenetic differences between diabetes patients with ESRD and diabetes patients without nephropathy. Epigenetics 2011; 6(1):20-8.
^ Smyth LJ, McKay GJ, Maxwell AP et al. DNA hypermethylation and DNA hypomethylation is present at different loci in chronic kidney disease. Epigenetics 2013;9(3): Epub ahead of print.
^ Ko YA, Mohtat D, Suzuki M, et al. Cytosine methylation changes in enhancer regions of core pro-fibrotic genes characterise kidney fibrosis development. Genome Biology 2013;14:R108
^ Brennan EP, Ehrich M, Brazil DP et al. Comparative analysis of DNA methylation profiles in peripheral leukocytes versus lymphoblastoid cell lines. Epigenetics 2009;4(3):157-64.