Genome-wide DNA methylation profiling in mixed ancestry South Africans with diabetes or prediabetes

Abstract

Type 2 diabetes mellitus (T2DM) is a global health concern and has shown to be a rising public health crisis in Africa. South Africa, in particular, has shown the highest prevalence of the disease with a large percentage of undiagnosed individuals. There is a need for intervention strategies that include early detection of at-risk individuals and the prevention of disease progression. Although studies have shown a relationship between the occurrence of T2DM and genetic and lifestyle factors, it has been suggested that epigenetic mechanisms such as DNA methylation contribute to the pathogenesis of T2DM through its association with the transcriptional activity of genes. There are however limited studies focussing on the genome-wide DNA methylation profiling of diabetic and prediabetic subjects from Africa and in particular, South Africa. The aim of the current study therefore was to conduct genome-wide DNA methylation in South African subjects with varying glucose tolerance and investigate the relationship between the observed DMRs and cardiometabolic risk factors. A cross-sectional case-control study design was used to conduct genome-wide DNA methylation profiling and lncRNA analysis on the peripheral blood of 48 South African individuals from the Bellville South community in the Western Cape. The participants were classified according to their glucose tolerance status and comprised 12 participants with known diabetes on metformin treatment, 12 screen-detected (newly diagnosed) diabetics, 12 prediabetics and 12 participants with normal glucose tolerance (NGT). After DNA extraction, DNA immunoprecipitation sequencing was used to identify statistically significant differentially methylated regions (DMRs) and lncRNA-associated DMRs, followed by gene ontology and KEGG pathway analysis. Lastly, the significant DMRs identified were validated by performing pyrosequencing of bisulphite converted DNA. The study identified several DMRs and functional pathways affected in subjects with diabetes and prediabetes. A total of 366 DMRs and 641 lncRNA-associated DMRs were observed, of which 63% were hypermethylated and 37% hypomethylated. Gene ontology and KEGG pathway analyses identified hypermethylation in cardiovascular processes in screendetected diabetic subjects indicating that a decrease in the expression of the associated genes may be associated with these functions in diabetes. Furthermore, hypomethylation was evident in purine metabolism in the screen-detected diabetic subjects highlighting the risk of increased reactive oxygen species production, inflammation and cell damage associated with excess uric acid in diabetic individuals. The study also identified DMRs and their functional pathways which showed a possible progression from prediabetes to diabetes. Hypomethylation of LBP (Lipopolysaccharide Binding Protein) in prediabetic subjects suggested that increased expression of this DMR could potentially be used as a biomarker for the progression to diabetes due to its association with increased inflammatory cytokines, insulin resistance and beta-cell dysfunction. Furthermore, hypomethylation of the regulation of wound healing and blood coagulation and their associated DMRs, SERPINF2 and DMTN, was observed in prediabetic subjects. Elevated levels of SERPINF2 (alpha 2-antiplasmin) and DMNT (dematin) expression in hyperglycaemic subjects could be used as an indicator of the increased risk for cardiovascular disease and red blood cell stability. Metformin, which is used as the first line of treatment for T2DM, has been shown to be affected by DNA methylation patterns. Hypermethylation of the cytokine-cytokine receptor interaction and oxidative phosphorylation pathways in known diabetics on metformin treatment suggested that metformin may have an inhibitory effect on complement-mediated inflammation and mitochondrial oxidative phosphorylation in diabetic individuals. The cAMP signalling pathway was also found to be hypermethylated in metformin-treated subjects which suggest the use of DNA methylation as a potential marker to monitor the effects of metformin on glucose homeostasis. Furthermore, hypermethylation of the functional pathways and DMRs identified in metformin-treated subjects when compared to the untreated screendetected subjects were associated with Diabetic Peripheral Neuropathy (DPN). Several novel lncRNAs were observed when comparing the lncRNA-associated DMRs identified amongst individuals with varying degrees of glucose tolerance. Hypomethylation of the complement component C4 was observed in hyperglycaemic subjects suggesting a possible association with the cardiometabolic risk factors and complement-mediated inflammation associated with T2DM. The lncRNA-associated DMRs observed in metformintreated subjects included the mitochondrial ATP synthase-coupling factor 6 (ATP5J) enzyme thought to be involved in the oxidative phosphorylation pathway with associations to T2DM and hypertension. In conclusion, these findings show that DNA methylation patterns differ amongst individuals with varying degrees of glucose tolerance within a South African population. Furthermore, the study showed that DNA methylation patterns are associated with certain cardiometabolic traits and diabetic complications, and could be used as potential biomarkers for the occurrence and progression of T2DM. The expression of several DMRs and lncRNA-associated DNA methylation regions observed in metformin-treated T2DM may also be potential targets for therapeutic monitoring in patients with diabetes.