Gene regulation of acute insulin response in human adipose tissue in obesity and after weight-loss

Abstract: The incidence and prevalence of obesity and type 2 diabetes is reaching epidemic proportions worldwide and insulin resistance, the pathophysiological link between the two conditions is consequently the focus of intense research. Although it is well-established that insulin resistance is associated with obesity, up to 1/3 of the obese subjects display insulin sensitivities comparable with non-obese individuals. Those subjects, known as insulin sensitive obese (ISO), show a different ectopic fat deposition, compared to insulin resistant obese (IRO), that can play a protective role in the metabolic consequences of obesity. In fact, ISOs are characterized by a favourable inflammatory profile that leads to a decreased cardiovascular risk. The association between obesity and insulin resistance is a cause-and-effect relationship in which weight loss/gain correlates with increasing/decreasing insulin sensitivity. The execution of biological processes requires a precise and well-orchestrated set of steps and depends on the proper spatial and temporal expression of genes. As a result, dysregulation of gene expression can lead to diseases. Over the last decades, advances in experimental and computational technologies have facilitated genomics research. The overall aim of this thesis is to study gene regulatory events in human subcutaneous adipose tissue of obese and lean subjects. In Paper I we investigated human subcutaneous adipose tissue insulin response gene expression by 5’ Cap-mRNA sequencing (CAGE) in 17 non-obese, 21 insulin sensitive obese and 30 insulin-resistant obese. We showed that insulin sensitive and insulin resistant obese share a similar transcriptional response to insulin, which differ from the NO insulin response. By identifying 231 genes altered upon hyperinsulinemia in the three groups, we were able to demonstrate that the acute transcriptional insulin response is mainly driven by obesity per se, challenging the notion of healthy obese adipose tissue. In Paper II, we used a unique loss-of-function model in human multipotent adipose-derived stem cells (hMADS) to characterize the transcriptional and epigenetic changes of the coregulator GPS2 in response to depletion during adipocyte differentiation. We show that loss of GPS2 results in reprogramming of cellular processes related to adipocyte differentiation and that GPS2-depleted adipocytes are associated with hypertrophy, triglyceride and phospholipid accumulation, and sphingomyelin depletion. Furthermore, adipocyte hypertrophy results from increased expression of ATP-Binding Cassette Subfamily G Member 1 (ABCG1), which mediates sphingomyelin efflux from adipocytes and modulates lipoprotein lipase (LPL) activity. Cellular findings obtained by gene expression analysis were validated using an obese cohort in which GPS2 is downregulated in diabetic individuals and negatively correlates with ABCG1 expression. In Paper III by characterizing the insulin transcriptional response by CAGE in 23 non-obese (NO), 23 obese (OB), and 23 post-obese (POB) women, we aimed to determine the extent of restoration of the insulin response after weight loss. We identified a common set of genes present in all three groups and characterized by genes related to lipid and cholesterol biosynthesis. An additional set of obesity-attenuated genes was identified that was related to tissue remodelling and protein translation and that was selectively regulated in the two lean states (NO and POB) and an additional, post obesity-enriched set of genes associated with genes involved in one-carbon metabolism. Overall, by identifying these three groups of genes, we have shown that human WAT exhibits a selective insulin response in obesity, with most genes normalizing after weight loss. In Paper IV by using CAGE data from Paper III, combined with chromatin conformation sequencing data (HiC) and promoter region centered HiC (PCHiC), we investigated how gene regulatory elements are involved in human subcutaneous adipose tissue insulin response in obese and in restored insulin response after weight-loss. We identified 21,632 enhancer candidates in human sWAT and 155,000 possible interactions between enhancers and genes in general and specifically 101 interactions between 130 enhancers and 47 insulin responding genes. With our analysis, we identified several enhancer-promoter interactions involved in hWAT insulin response, including well-known insulin responding genes such as AACS, LDLR and PDK4.