Genetics of diabetic nephropathy

There is a genetic predisposition to diabetic nephropathy

Diabetic nephropathy clusters in families, which suggests a shared genetic or environmental contribution to this disease. Only a subset of individuals living with diabetes (~40%) develop diabetic nephropathy, and studies have shown that this is not just due to poor control of blood glucose levels. The prevalence of diabetic nephropathy varies worldwide and that also suggests a genetic source; for example, three in every five patients beginning renal replacement therapy in Singapore have end stage renal disease due to diabetes, while Denmark has a much lower incidence.[1]

Figure 1: Timeline illustrating the pace of significant developments in genetic studies for diabetic nephropathy. The DNA helix (www.jameshedberg.com/scienceGraphics.php?sort=all&id=dsDNA_doubleHelix_threeStrands ) used to create this timeline was created by James Hedberg under a Creative Commons License.
Figure 1: Timeline illustrating the pace of significant developments in genetic studies for diabetic nephropathy. The DNA helix (www.jameshedberg.com/scienceGraphics.php?sort=all&id=dsDNA_doubleHelix_threeStrands ) used to create this timeline was created by James Hedberg under a Creative Commons License.
Diabetic nephropathy is a multifactorial disease with complex inheritance mechanisms. A range of approaches have been employed to investigate genetic risk factors for diabetic nephropathy. Historically, linkage studies were conducted in multigenerational families (or using pairs of siblings who were concordant or discordant for diabetic nephropathy) to track how particular regions of chromosomes were transmitted to affected individuals through families. In 1996, a landmark paper[2] suggested that association studies offered more statistical power to identify multiple genetic risk factors, with moderate effect sizes, that influence complex diseases such as diabetic nephropathy. Several family-based studies have been conducted, but the majority of association studies compare affected individuals (cases) to individuals with diabetes who are unaffected by nephropathy (controls). Genetic association studies developed from single variant, to single gene, targeted genetic regions, genome-wide, and now meta-analysis of genome-wide association studies for millions of single nucleotide polymorphisms (SNPs). Ongoing projects include next generation sequencing approaches to elucidate genetic risk factors for diabetic nephropathy.

What do we know?

Despite some initially very promising findings, the majority of genetic variants suggested to influence diabetic nephropathy have not been sustained. However, advances in genotyping technologies, improved understanding of genetic variation in humans, and the creation of international consortia are now able to identify genetic risk factors that are robustly replicated and supported by functional studies. Several consortia have been created to investigate genetic risk factors for diabetic nephropathy including GoKinD (Genetics of Kidneys in Diabetes) UK and USA, Warren 3 collection, Euragedic, FinnDiane, FIND (Family Investigation of Nephropathy and Diabetes) and SUMMIT. Extended collaborations such as these facilitate a multidisciplinary approach with shared expertise, enable the consistent use of harmonised inclusion / exclusion phenotype criteria for case-control studies, and share complementary resources. Genetic risk factors for diabetic nephropathy appear to be different between individuals who have type 1 diabetes compared to those with type 2 diabetes, although some overlap has been reported.

Genetic risk factors for Type 1 diabetes and nephropathy (T1DN)

Many genetic factors have been linked with diabetic nephropathy, primarily from small-scale studies in terms of the sample size and number of variants examined. Several linkage studies have been performed, but did not reveal very encouraging results. Association studies are more likely to reveal genetic risk factors for common disease, so the first GWAS was conducted using microsatellites in 2006, followed by two SNP-based GWAS for T1DN in 2009. Large-scale studies are subject to multiple testing issues, so ideally a p value <10-8 is required for genome-wide significant association between a variant and disease. These studies did not reveal strong, genome-wide significant associations with T1DN, suggesting that no major genetic risk alleles influence T1DN with strong effect sizes. Meta-analyses of previously reported associations has revealed conflicting results depending on the number of individuals and mix of ethnicities studied.[3][4]

To help elucidate genetic risk factors for T1DN, the GENIE (GEnetics of Nephropathy, an International Effort) consortium was created, bringing together international collaborators with sufficient sample sizes to have adequate power to identify risk alleles with moderate effects. GENIE conducted a meta-analysis of three independent GWAS with 6,691 individuals, followed by subsequent replication using all available collections of DNA worldwide (n=5,873 individuals).[5] This revealed an intronic SNP in the ERBB4 gene (p=2.1x10-7) suggestively associated with T1DN, supported by differential expression for diabetic nephropathy in persons with type 2 diabetes. The strongest finding was for the more extreme phenotype of end stage renal disease, where two SNPs (rs7583877 in the AFF3 gene, p=1.2x10-8; rs12437854 between RGMA and MCTP2 genes, p=2x10-9) were identified. Functional evidence supports AFF3 as a novel risk factor for end stage renal disease in T1D, perhaps influencing renal tubule fibrosis via the TGFβ pathway.

There are sex-specific differences influencing diabetic nephropathy, so studies have investigated independent genetic variants for association in men and women separately; rs4972593 is physically located near to the SP3 gene and was identified as a sex-specific SNP associated with protection against end stage renal disease in women with type 1 diabetes (p<5x10-8, OR 1.81, 95% CI: 1.47-2.24) following meta-analysis of three cohorts and significant changes in SP3 glomerular gene expression.[6]

Recent evidence suggests that the phenotype under investigation (persistent proteinuria, declining renal function, or end stage renal disease) may have different genetic risk profiles for individuals with T1DN.[7][8] This has important consequences for future studies and ongoing work is evaluating genetic risk factors in larger case-control and longitudinal cohorts for a variety of renal phenotypes.

Genetic risk factors for Type 2 diabetes and nephropathy (T2DN)

Identifying robust genetic risk factors for T2DN is complicated by a complex phenotype including independent risks for renal disease unrelated to diabetes. A shared genetic risk has been suggested for T2D and kidney complications in affected individuals, including WFS1, FTO, KCNJ11 and TCF7L2 genes,[9] but this is not necessarily classical ‘diabetic nephropathy’, as individuals with declining glomerular filtration rate were evaluated, including those without albuminuria.

Similar to T1DN approaches, multi-centre consortia are pooling resources and conducting meta-analyses on large numbers of SNPs. As the number of individuals with type 2 diabetes continue to rise, larger research studies are now possible, thus increasing the likelihood that true risk variants will be identified for T2DN. Phenotypes examined for T2DN include albumin : creatinine ratio, estimated glomerular filtration rate, proteinuria, creatinine clearance, and end stage renal disease. Many SNPs have been reported as associated with T2DN, but few have been robustly replicated in subsequent studies with strong significance values. An early (2005) larger-scale study examining 80,000 gene-based SNPs in Japanese individuals revealed association with the ELMO1 gene and T2DN;[10] this has been replicated by multiple groups across several ethnicities with supporting evidence from sequencing and gene expression experiments. FRMD3, MYH9-APOL1 and ACACB genes have also been identified across multiple studies and with supporting functional evidence that SNPs in these genes are important risk factors for T2DN.[11][12]

Where next?

The genetic architecture of DN is being elucidated, but there is still much work to be done. One option to identify genetic risk factors for DN is to use extreme phenotypes for ‘enriched’ case-control studies, where cases have early onset DN or rapidly declining renal function, and are compared to controls with a long duration of diabetes with no evidence of kidney disease. As clarity is gained on the natural history of DN, more precise phenotypes can be used to minimise genetic heterogeneity that is unrelated to DN between individuals with diabetes. Case-control, family-based and longitudinal studies are all facilitating the identification of genetic risks for DN.

The majority of studies to date have focused on common SNPs that may provide risk or protection for DN. Higher density chips now contain more SNPs and thus allow more coverage of the human genome, whether these are focused on the exome (protein coding) or entire genome (five million SNPs). It is possible that more rare variants (<5% minor allele frequency) contribute to a genetic risk profile for DNA. Identifying association with rare variants and DN is now possible from the increased power of merged datasets and large collaborative studies. Ongoing studies also include analysis of X, Y and mitochondrial chromosomes, which are typically excluded from genome-wide association studies.

More comprehensive analyses of the human genome are also possible through advances in next generation sequencing technologies and analysis options. Community-based initiatives such as the English 100K genome project, which plan to link detailed medical records with the individual genomes of 100,000 individuals by 2017, offer further potential for in-depth analysis at a population level. In the USA, the Genetic Epidemiology Research on Aging project has made publicly available genetic data linked to medical information on ~78,000 individuals. Next generation sequencing provides rich, complementary information on common SNPs, rare SNPs and copy number variation when using DNA-seq, gene expression when using RNA-seq as well as providing information on epigenetic modifications when using bisulfite-treated DNA-seq or chip-seq. However, the use of such large datasets means that stringent statistical thresholds, quality control, replication and meta-analysis are very important for all studies. Incorporating genetic, epigenetic, transcriptomic, and proteomic datasets alongside detailed clinical information are helping to explain how genes and pathways influence diabetic nephropathy.

References

  1. ^ U.S. Renal Data System, Chapter 12, International comparisons in USRDS 2013 Annual Data Report: Atlas of End-Stage Renal Disease in the United States, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2009. http://www.usrds.org/2013/view/v2_12.aspx

  2. ^ Risch N, Merikangas K. The future of genetic studies of complex human diseases. Science 1996;273(5281):1516-7.

  3. ^ Mooyart AL, Valk EJ, van Es LA, Bruijn JA et al. Genetic associations in diabetic nephropathy: a meta-analysiss. Diabetologia 2011;54(3):544-53.

  4. ^ Williams WW*, Salem RM*, McKnight AJ*, Sandholm N* et al., * denotes joint 1st author. Association testing of previously reported variants in a large case-control meta-analysis of diabetic nephropathy. Diabetes 2012;61:2187-94.

  5. ^ Sandholm N*, Salem RM*, McKnight AJ*, Brennan EP* et al., * denotes joint 1st author. New susceptibility Loci associated with kidney disease in type 1 diabetes. PLoS Genetics 2012;8(9):e1002921.

  6. ^ Sandholm N, McKnight AJ, Salem RM, Brennan EP, et al. Chromosome 2q31.1 Associates with ESRD in Women with Type 1 Diabetes. Journal of the American Society of Nephrology 2013, 24(10): 1537-43

  7. ^ Sandholm N, Forsblom C, Mäkinen VP, McKnight AJ, et al. Genome-wide association study of urinary albumin excretion rate in patients with type 1 diabetes. Diabetologia 2014, in press.

  8. ^ Chan Y, Lim ET, Sandholm N, Wang SR et al. An excess of risk-increasing low frequency variants can be a signal of polygenic inheritance in complex diseases. American Journal of Human Genetics 2014, in press.

  9. ^ Franceschini N, Shara NM, Wang H, Voruganti VS. The association of genetic variants of type 2 diabetes with kidney function. Kidney International 2012:82(2):220-5.

  10. ^ Shimazaki A, Kawamura Y, Kanazawa A, Sekine A et al. Genetic variations in the gene encoding ELMO1 are associated with susceptibility to diabetic nephropathy. Diabetes 2005:54(4):1171-8.

  11. ^ Freedman BI, Langefeld CD, Lu L, Divers J et al. Differential effects of MYH9 and APOL1 risk variants on FRMD3 association with diabetic ESRD in African Americans. PLoS Genetics 2011:7(6):e1002150.

  12. ^ Maeda S, Kobayashi MA, Araki S, Babazono T et al. A single nucleotide polymorphism with the acetyl-coenzyme A carboxylase beta gene is associated with proteinuria in patients with type 2 diabetes. PLoS Genetics 2010:6(2):e1000842.

Comments

Nobody has commented on this article

Commenting is only available for registered Diapedia users. Please log in or register first.