UCLA

Obesity is an important driver of many cardiometabolic disorders (CMDs), including type 2 diabetes (T2D) and multiple types of dyslipidemias. Genome-wide association studies (GWAS) have identified genetic variants associated with body mass index (BMI) in humans, but the genes and genomic regulatory mechanisms underlying these associations are not well-understood. Furthermore, gene-environment interactions (GxEs) likely play a role in driving the population variance in BMI, as well as downstream comorbidities. Understanding the biology underlying genetic associations requires studying the relevant cell- and tissue-types, as variants will likely only function in cell-types in which they regulate gene expression. However, determining in which cell-types and under what environmental conditions the variants function is not a trivial task. Functional fine-mapping of individual loci is laborious and time-intensive. This issue has produced a bottleneck in understanding the hundreds of loci that have been associated with CMDs. One way to prioritize strong candidates for future characterization of gene function is to first fine-map the cell- and context-specific genomic regulatory mechanisms through which variants function, thereby providing evidence of the variant having a role in a given context. Adipose tissue is the main site of fat storage in the body, and is thus highly responsive to the obesogenic environment. It must expand to accommodate excess nutrients, and it produces signaling molecules that regulate food intake and energy expenditure. Thus, genetic variants that regulate these and other important adipose tissue processes likely play a role in the etiology and pathophysiology of obesity.

The projects in this dissertation were designed with the goal of identifying genes and genomic regulatory mechanisms underlying the genetic risk for obesity and related comorbidities. I studied the key adipose tissue cell-types that are responsible for energy homeostasis: adipocytes, the energy-storing fat cells, and their progenitors, preadipocytes. In the second chapter, we fine-mapped BMI GWAS variants to those that are likely functioning in adipocytes. We performed promoter Capture Hi-C in human primary adipocytes to identify the physical interactions between gene promoters and their regulatory elements. We then linked the BMI GWAS variants to adipose gene expression by identifying cis-eQTLs in 335 subcutaneous adipose tissue biopsies from the METabolic Syndrome In Men (METSIM) cohort. By screening the adipocyte-specific promoter-interacting regions for these variants, we identified four examples of BMI GWAS variants and 38 additional candidate genes that likely function in adipocytes via promoter interactions to affect BMI in humans.

Work in the third chapter was motivated by the hypothesis that GxEs affect adipose tissue expansion and thus contribute to variation in BMI. Given the known correlation of high dietary saturated fat intake with BMI, I sought to extend this knowledge and demonstrate that causal genomic regulatory mechanisms in response to saturated fat intake in adipocytes affect variation in BMI. Genetic variants exhibiting significant interaction effects are difficult to detect in humans, for reasons including the multiple-testing burden for genome-wide GxE scans and the heterogeneity of our environments. To circumvent these issues, I first performed a controlled, in vitro treatment of human primary adipocytes with dietary saturated or monounsaturated fatty acids. I quantitatively assessed the genomic responses to this lipid challenge via changes in the accessible chromatin landscape in the adipocytes. Only the genetic variants that landed in the lipid-responsive genomic regulatory elements were selected for a GxE scan in the large UK Biobank (UKB) cohort. By prioritizing and restricting the GxE search space in this way, we identified 38 significant GxE variants that exhibit adipocyte-origin genomic regulatory mechanisms responding to dietary saturated fat to affect BMI in humans.

It is important to address not just the etiology of obesity, but its pathophysiological mechanisms as well. A pro-inflammatory environment develops in the adipose tissue in obesity, which is then thought to contribute to systemic low-grade inflammation and downstream obesity comorbidities. In chapter four, I leveraged a BMI-discordant monozygotic (MZ) twin cohort (BMI difference greater than or equal to 3 kg/m^2) to determine the contribution of preadipocyte genomic dysregulation to systemic inflammation in humans. Preadipocytes can mount an immune response to the pro-inflammatory signals from adipose tissue macrophages, at the expense of differentiating into adipocytes. However, the mechanisms underlying this dysregulation are not fully known. Furthermore, it is not clear whether the preadipocyte mechanisms are causal for inflammation, or simply reactive to the environment. By studying the chromatin accessibility and gene expression profiles in the MZ twin PAd, I showed that increased BMI alters the higher-order genomic programming of PAd. The reprogrammed regions exhibit a stronger accumulation of low p-value GxE variants that interact with BMI to affect the inflammatory marker C-reactive protein (CRP) in the UKB, thus providing evidence of PAd-origin genetic and genomic mechanisms contributing to systemic inflammation in humans.

In summary, by integrating various levels of epigenomic, transcriptomic, and genetic information across multiple cohorts with deep phenotyping, we have prioritized genes and genomic regulatory mechanisms in human adipose tissue cell-types that are important for obesity and obesity pathophysiology. The improved molecular understanding of genetic causes and GxEs underlying obesity and its comorbidities will inform prevention and treatment methods in precision medicine, toward reducing cardiovascular disease risk in humans.