Gene by environment interactions in the regulation of gene expression across primary tissues and their application to obesity and related traits

Gene expression is an important cellular phenotype under both genetic and environmental control. Changes in gene expression appear to underlie the majority of genetic associations identified in genome-wide association studies of common disease. Studying the genetic regulation of gene expression has thus been a fruitful strategy to unraveling the underlying mechanisms of common disease, both in identifying the genes and pathways that mediate the effects of disease and in characterizing the molecular mechanisms and tissues by which disease-associated genetic variants act.

Previous research into the genetics of gene expression has focused on identifying genetic variants that regulate the total expression level of a gene. This study will expand on this approach by incorporating the effects of the environment and its interactions with the genome. In the context of a cellular trait like gene expression the environment can be interpreted as exposures that operate on the organism (eg diet, medication, physical activity, or smoking) or on the physiology of tissues and cells (including obesity, insulin and cholesterol levels). This study will investigate the joint effects of the environment (lifestyle factors and biomedical measurements) and genetics on the regulation of gene expression in a deeply phenotyped, 'omics level dataset. Importantly, this study includes multiple disease-relevant tissues as both gene regulation and disease manifestation is tissue specific.

Obesity is a worldwide epidemic and is robustly associated with many co-morbidities including type 2 diabetes, cardiovascular disease, osteoarthritis, certain cancers and increased mortality. However, not all obese individuals develop the same co-morbidities and some obese individuals even appear to be metabolically healthy. Incomplete knowledge of the processes involved in the development of obesity-associated co-morbidities has limited the development of treatments or prevention strategies. While the genes driving increased body mass index (BMI) appear to act primarily in the Central Nervous System, genes underlying body fat distribution and co-morbidities such as insulin resistance and Type 2 Diabetes are active in fat tissue. Obesity has a dramatic effect on fat tissue, including cell composition, metabolism and gene expression. Obesity can thus mask, confound or modify associations between genetic variants and gene expression in fat tissue. This study will consider obesity (measured both as total adiposity and body fat distribution) as an environment in order to identify genetic variants that interact with obesity to regulate gene expression. This will allow an expanded understanding of the molecular effects of obesity on fat cells and identify genotype-dependent differences in the cellular response to obesity that could be linked to differential development of obesity co-morbidities.

This study will identify gene by environment interactions (GEI) influencing gene expression. GEI effects have been difficult to identify at the disease and trait level owing to the large sample sizes needed to detect an interaction in a genome-wide search and the required careful measurement of environment for all subjects. Studying GEI effects on an expression level has many advantages. The effect size of genetic regulation of gene expression is quite large, reducing the need for huge and costly sample sizes. As each gene has a known physical position in the genome the statistical burden of multiple testing can be greatly reduced by limiting the search space to the local region around each gene. Environmental measures that operate on the organismal or cellular level can be assayed, including quantitative biomedical phenotypes.

This study will utilize gene expression data measured with next-generation sequencing technology (RNAseq) in four primary tissues (adipose, skin, whole blood, muscle) and one cell line (transformed lymphocytes) from ~800 individuals. RNAseq data allows interrogation of effects on both total gene expression and alternative splicing. A wide array of environmental exposures will be investigated, including organism level exposures such as diet, smoking and medication, and biomedical measures such as cholesterol and insulin that influence the cellular environment.

Particular attention will be paid to obesity-related exposures and their effects on adipose expression. Obesity has a dramatic impact on the physiology of adipose tissue. In this study measures of adiposity will be utilized to identify GEI regulatory effects, where adiposity is the exposure and expression the outcome variable. This will identify downstream molecular consequences of obesity, the genetic variants that mediate them, and potentially why different individuals develop different obesity associated co-morbidities.

The principal beneficiaries of the research will be:
1) Academics in the fields outlined in the "academic beneficiaries" section
2) Clinicians
3) Pharmaceutical industry
4) The wider public if the translational potential is realized and results in better therapies and/or counseling for treatment or prevention of obesity related co-morbidities

The major impacts will include.

1) Improved ability to stratify risk and response to treatment and interventions.

2) Targeted treatments could help target individuals responsive to diet or physical activity intervention

3) As not all obese individuals develop the same downstream morbidities, identification of BMI-based GEI interactions could help predict which co-morbidity an individual is more likely to develop based on their genotype.

4) Identifying the early consequences of obesity will allow for early monitoring, its use in clinical trials and potentially interventional strategies.

5) Pharma are particularly interested in the downstream affects of obesity and how the adverse consequences could be prevented therapeutically. Data on gene expression and networks will be of great value to this effort.

6) The catalogue of genes and genetic variants with regulatory GEI effects will be useful to study designs, particularly interventions, which wish to target genes without GEI effects to ensure a more homogenous response.

7) GWAS genes frequently are related to therapeutic targets and this study will improve the functional understanding of those targets