Discovery and functional validation of hepatic eQTLs in African Americans.

Currently, many genome-wide association studies (GWAS) show that single nucleotide polymorphisms (SNPs) associated to drug response and drug metabolism are in non-coding regions of the genome. Assignment of function to these SNPs requires multiple levels of data. Multi-omic data exist for populations of European ancestry and have been used extensively to elucidate the target genes of associated SNPs through studies of expression quantitative trait loci (eQTLs - SNPs associated with gene expression). However, population diversity within publicly available multi-omic databases is sorely lacking. Importantly, the liver is key in drug detoxification and multi-omic data specific to this tissue would benefit the translation of pharmacogenomic findings. Populations of African ancestry have greater genetic diversity than any other world population. Thus, African Americans (AAs) may differ in allele frequency, allele distribution, and combination of variants when compared to European populations. These unique aspects of the AA genome were leveraged in our previous work which discovered AA-specific eQTLs, and gene expression that was significantly associated with ancestry proportions. We will continue this important work by adding needed functional information to a larger dataset of African American multi-omic data. One recuring problem in eQTL mapping is the ability to identify causal regulatory variants, as eQTL studies only show association between SNPs and gene expression and require functional validation. Massive Parallel Report Assays (MPRAs) have been used to help pinpoint SNPs that have causal effect on gene expression[MOU1] . This becomes even more important in studies on non-European subjects, as few public datasets of any kind exist in which findings from GWAS or eQTLs can be validated. This has led to several genetic associations in African Americans for which the mechanism of effect is unknown. By first mapping regulatory genetic variants in open chromatin and then validating their effect in a cellular assay, we will narrow down our eQTL findings to those that have validated functional effects. Lastly, true translation of genetic findings into clinical care require that the causal SNPs have an effect that is large enough to be clinically meaningful. This has been a difficult hurdle for most non-coding genetic findings. However, in pharmacogenomics, which is known for its large effect sizes, the possibility of single variants resulting in an implementable genetic test, exist. As such, we will further evaluate eQTLs for drug metabolizing enzymes (DMEs) in a pharmacokinetic assay to determine the effect of these variants on enzyme activity. Thus, we hypothesize that using our prioritization scheme and functional assays, we will identify genetic regulatory variants for genes in the liver as well as in DMEs that are clinically relevant for African Americans. The need for more predictive biomarker is clear in this populations, and our work will fill this gap. AIM1: Expansion of hepatocyte eQTL data in African Americans. To improve our ability to find functional elements that affect gene expression, we will perform eQTL mapping in 120 hepatocyte cultures derived from African Americans. These eQTLs will be colocalized with GWAS findings in relevant phenotypes (e.g. lipid traits, liver enzymes), fine-mapped to find potential causal variants for each eGene as well as profiled for transcriptional factor binding changes. AIM2: Determining function of eQTLs via MPRA in HepG2. Determining which eQTL are truly regulatory as opposed to just associated to gene expression due to linkage disequilibrium (LD), requires additional validation. As such we will create a MPRA for all eQTLs. We will use HepG2s (immortalized hepatocyte cell line) to determine the effect of each prioritized eQTL. This will allow us to break the LD structure within these regions and narrow down eQTL signal to causal variants. Our preliminary data show how this method was used to identify potential causal regulatory variants in factor 5 (F5), a gene important to thrombosis. AIM3: Investigate the effect of eQTLs on enzyme activity genes encoding DMEs via genome editing. As the data created is in hepatocytes, which are critical in drug metabolism, we will investigate the effect of eQTL identified in DMEs genes for changes in enzyme function. We will prime edit HepaRG cells to carry eQTLs link to Phase I DMEs and measure the parent disappearance and metabolite formation of known probe substrates. This will allow us to test how changes in gene expression result in changes to drug metabolism. We have already identified an eQTLs within CYP3A5. We will evaluate the role of this eQTL in CYP3A5 activity and use similar methods for other DME eQTL identified.