Epigenetic analysis of type 2 diabetes and measures of glucose metabolism.

Diabetes mellitus is one of the most widespread diseases worldwide . About 381.8 million people have the disease, and it is estimated that just as many have undetected diabetes or are at high risk (pre - diabetes). In order to better understand the disease and its associated burden it is important to understand the molecular disease mechanisms. The genetic aspects of t ype 2 diabetes (T2D) , which impact the majority of diabetes cases , h as been comprehensively studied in genome - w ide genetic variation studies. Thus, today we know about a large number of genes involved in disease developm ent. However, to my knowledge, there are no systematic genome - wide studies on epigenetic regulation mechanisms in whole blood and participants with European ancestry . Therefore , the objectives of this thesis are to identify an association between DNA methylation and i) incident T2D , and ii) measures of glucose metabolism in whole blood using samples from the Cooperative Health Research in the Region of Augsburg (KORA) study. In the first part of this thesis the methylation degree s of six CpG sites w ere genome - wide significant ly associated with incident T2D [annotated to CASZ1 , TMEM57 , VIM , C14orf182 , AKT2 , and one unknown (adjusted p - value s = 6.2x10 - 3 to 1.9x10 - 2 )]. Three CpG sites flanking the leading CpG site annotated to AKT2 and one to C14orf182 were found to be significantly associated with the phenotype in the replication. These associations seem to be driven by body mass index (BMI) . Further mor e , performing pathway analysis with DNA methylation , data pathways such as “ N erve G rowth F actor S ignaling”, “ I ntegrin S ignaling” or “G - protein gamma/beta S ignaling ” were detected, each of which can be linked to diabetes . Using data from KORA for replicatio n of the results from the London Life Science Prospective Population (LOLIPOP) study , which consists of Indian Asians and European individuals , it was possible to replicate five of the seven significant CpG sites (annotated to TXNIP , PHOSPHO1 , SOCS3 , SREBF 1 , and ABCG1 ). The difference in findings can be explained by differing study design s , especially concerning different ethnicities , matching criteria for cases and controls , and sample sizes . In the second part of this thesis association s between the methy lation degree s of 15 CpG sites and measures of glucose metabolism were detected [four Summary VII for fasting glucose, seven for fasting insulin, ten for homeostasis model assessment - insulin resistance (HOMA - IR); adjusted p - value s = 6.8x10 - 5 to 0.043]. In addition, DN A methylation at cg06500161 ( ABCG1 ) was significantly associated with fasting insulin, fasting glucose, 2 - hour glucose , and HOMA - IR (adjusted p - value s = 8.1x10 - 4 , 6.8x10 - 5 , 1.3x10 - 3 , and 7.5x10 - 5 , respectively). BMI explains around 30% of these association s. Further more , a significant association between the CpG sites found for the phenotypes named above and 2 - hour insulin was observed in a subset of samples . These were also in part significant after additional adjustment for BMI. A nalyzing association s bet ween DNA methylation and gene expression in another subset, CpG site cg06500161 ( ABCG1 ) showed an association with ABCG1 gene expression level (adjusted p - value = 1.1x10 - 9 ). Additionally, an accumulation of the top 1,000 CpG sites in pathways such as “ Lept in Signaling in Obesity”, “Ephrin A/B Signaling”, “Netrin S ignaling” , and “Phospholipase C Signaling” , which can be linked to diabetes, was found. In conclusion , this thesis indicates an association between DNA methylation , incident T2D / measures of glucose metabolism in whole blood which seem s to be BMI mediated. These results can help to better understand the underlying pathogenesis of T2D and measures of glucose metabolism and can thus be used to potentially develop biomarkers and new therapies for T2D