Genetics of diabetes-associated microvascular complications

Abstract

Diabetes is associated with excess morbidity and mortality due to both micro- and macrovascular complications, as well as a range of non-classical comorbidities. Diabetes-associated microvascular complications are those considered most closely related to hyperglycaemia in a causal manner. However, some individuals with hyperglycaemia (even those with severe hyperglycaemia) do not develop microvascular diseases, which, together with evidence of co-occurrence of microvascular diseases in families, suggests a role for genetics. While genome-wide association studies (GWASs) produced firm evidence of multiple genetic variants underlying differential susceptibility to type 1 and type 2 diabetes, genetic determinants of microvascular complications are mostly suggestive. Identified susceptibility variants of diabetic kidney disease (DKD) in type 2 diabetes mirror variants underlying chronic kidney disease (CKD) in individuals without diabetes. As for retinopathy and neuropathy, reported risk variants currently lack large-scale replication. The reported associations between type 2 diabetes risk variants and microvascular complications may be explained by hyperglycaemia. More extensive phenotyping, along with adjustments for unmeasured confounding, including both early (fetal) and late-life (hyperglycaemia, hypertension, etc.) environmental factors, are urgently needed to understand the genetics of microvascular complications. Finally, genetic variants associated with reduced glycolysis, mitochondrial dysfunction and DNA damage and sustained cell regeneration may protect against microvascular complications, illustrating the utility of studies in individuals who have escaped these complications.

Keywords: DNA; Diabetes; Genetics; Intrauterine programming; Metabolism; Microvascular diseases; Nephropathy; Neuropathy; Retinopathy; Review.

Fig. 1

The natural history of the development of DKD, retinopathy and peripheral neuropathy complications in individuals with type 2 diabetes may exhibit different trajectories depending on the time of diabetes onset. Temporal risk factors that may act as triggers of disease (red circles), modifying risk factors (pink circles) or both (half red/half pink circles) are shown, ordered depending on their approximate supposed appearance before or after diabetes onset. (a) The different stages of kidney disease (stage 1, normal; stage 2, moderate; stage 3, severe; stage 4; and end-stage renal disease) are defined based on eGFR. The natural history of kidney disease may commence at different blood glucose levels, including both prior to diabetes onset, where blood glucose levels may be normal or elevated, and after diabetes diagnosis (time of diabetes diagnosis indicated by time 0 on the graph). Individuals may experience different kidney function trajectories that either include a stage of hyperfiltration preceding a relatively faster decline in eGFR (dark green line; rapid decline in eGFR illustrated by hatched dark grey area), or a trajectory without hyperfiltration preceding a more moderate–slow pattern of eGFR decline (purple lines; slow decline in eGFR illustrated by hatched light grey area). A decline in kidney function may occur some years before diabetes onset (see green lines) owing to glucose-unrelated aetiologies. In individuals with preceding stage 2 (moderate) CKD, or stage 3 (severe) CKD (defined by a decline in eGFR <60 ml/min per 1.73 m²), progression to stage 3 DKD and beyond may occur much faster after the onset of diabetes (green lines) as compared with individuals without a precedent decline in eGFR (purple lines). At the GWAS significance level, 878 genetic loci have been reported as being associated with eGFR in a combined meta-analyses of individuals without and with diabetes [45], while 29 have been found to be associated with eGFR in those with diabetes [38]. Genetic markers for eGFR overlap in individuals with and without type 2 diabetes at different levels of significance, with the same directionality but with different magnitudes of effect. However, NEUROD2 has been associated with eGFR specifically in individuals without diabetes, while CSRNP1 has been reported as a specific risk variant in individuals with diabetes. The combination of CKD due to hypertension, obesity and/or other glucose-unrelated factors in individuals with type 2 diabetes may hinder identification of genetic loci specific to DKD. ESRD, end-stage renal disease; T2D, type 2 diabetes. (b) The natural history of diabetic retinopathy involves both neuronal (blue line) and vascular (orange line) lesions. Currently, progression to severe pre-proliferative and proliferative stages of diabetic retinopathy are defined only by vascular lesions. However, even prior to diabetes diagnosis (prior to time 0 on the graph), early neurodegenerative processes (blue line) are seen twice as frequently as vascular alterations in individuals without visual signs of diabetic retinopathy (dashed lines). To date, no genetic variants at any appropriate GWAS significance level have been associated with diabetic retinopathy; however, candidate gene studies for severe diabetic retinopathy have validated variants in the VEGF and TCF7L2 genes as being associated with this complication. Notably, the effects of the TCF7L2 variants are likely to be mediated by subtle or overt hyperglycaemia occurring even prior to diabetes diagnosis. DR, diabetic retinopathy; PDR, proliferative diabetic retinopathy; T2D, type 2 diabetes. (c) The natural history of peripheral neuropathy (blue line) and neuropathic pain (pink line) may commence at the normoglycaemic stage, some years prior to diabetes onset (prior to time 0 on the graph, dashed lines). The severity of peripheral neuropathy is often linked with the development of severe neuropathic pain. Most genes associated with these conditions have been identified through candidate gene studies, via combined analysis in individuals with and without diabetes with neuropathy/neuropathic pain. T2D, type 2 diabetes. This figure is available as a downloadable slide