Identification and functional impact of NAFLD associated genetic variants (GOLD functional)

a MiCDA Research Project Description

Investigators: Elizabeth Speliotes, Jiandie Lin, Charles Burant, Patricia A. Peyser, Sharon L. R. Kardia, Lawrence F. Bielak, Michael Lee Boehnke, Goncalo Abecasis

Funding: National Institute of Diabetes and Digestive and Kidney Diseases, 2015-2020 (1 R01 DK 106621 01)

Nononalcoholic fatty liver disease (NAFLD) affects up to 29 million adults in the U.S. and will become the number one cause of liver disease worldwide by 20201. There are few effective ways to prevent or treat this disease making it one of the biggest unmet medical needs of our time. Better understanding of its etiology is needed to improve diagnosis, management, treatment, and ultimately prevention of NAFLD. We found that NAFLD is heritable/genetic and identified single nucleotide polymorphisms (SNPs) that associate with hepatic steatosis in individuals of European ancestry2. We found that even if disease associated variants are not themselves coding, they often lie in or near genes that (1) contain coding functional disease promoting coding variants and (2) when perturbed in cell/animal models phenocopy the disease genetically and biochemically4. Coding variants can be used to help classify individuals into disease risk sub categories. Further, by determining the genes and genetic/biochemical mechanism by which functional variants cause disease we can use this information to develop future therapeutics/ biomarkers for NAFLD. Recent whole genome and exome sequencing studies identified many new coding variants but many are low frequency (1-5%) or rare (< 1%) and neutral as opposed to damaging5. Thus we will need large sample sizes and conditional/burden analyses to statistically distinguish functional variants/genes from neutral ones5. Towards this end we (1) used the Illumina HumanExome BeadChip that contains coding variants identified through large exome sequencing studies to economically genotype the largest (>17K) most ancestrally diverse collection of population based samples to date assayed for NAFLD (2) implemented new analysis methods8 (Raremetalworker and Raremetal) that reduce the analytic burden on individual cohorts and give us the ability to carry out single variant, gene based, and conditional analyses centrally using the covariance matrices derived from individual cohorts and (3) created a cell based model for NAFLD where we can test genes and variants for effects on hepatic steatosis. We hypothesize that NAFLD affecting coding variants can be (1) identified through powered exome chip analyses across ancestries and (2) will change lipid content after overexpressing/knocking down genes they affect in liver cell lines. To test these aims we will harmonize the fatty liver phenotype and genotypes in our cohorts and carry out single variant, gene burden, and conditional meta analyses across groups to narrow down NAFLD causal variants using Raremetalworker and Raremetal6. We will follow up top associating variants from AIM 1A in >4500 samples of histology measured NAFLD to replicate our findings. We will overexpress/knock down wildtype/mutant versions of three NAFLD candidate genes previously implicated from GWAS analyses using lentiviral expression vectors in human liver cell lines and assess for lipid accumulation using a colorimetric triglyceride assay/neutral lipid stains. If changes are observed, we will assess the knockdown/overexpression phenotypes can carry out metabolomic analyses to determine the gene?s genetic/lipidomic mechanism of action.

Research Signature Theme:

Aging, Genetics, and Social Science