Exome-Wide Association Study on Alanine Aminotransferase Identifies Sequence Variants in the GPAM and APOE Associated With Fatty Liver Disease

Abstract

BACKGROUND & AIMS: Fatty liver disease (FLD) is a growing epidemic that is expected to be the leading cause of end-stage liver disease within the next decade. Both environmental and genetic factors contribute to the susceptibility of FLD. Several genetic variants contributing to FLD have been identified in exome-wide association studies. However, there is still a missing hereditability indicating that other genetic variants are yet to be discovered. METHODS: To find genes involved in FLD, we first examined the association of missense and nonsense variants with alanine aminotransferase at an exome-wide level in 425,671 participants from the UK Biobank. We then validated genetic variants with liver fat content in 8930 participants in whom liver fat measurement was available, and replicated 2 genetic variants in 3 independent cohorts comprising 2621 individuals with available liver biopsy. RESULTS: We identified 190 genetic variants independently associated with alanine aminotransferase after correcting for multiple testing with Bonferroni method. The majority of these variants were not previously associated with this trait. Among those associated, there was a striking enrichment of genetic variants influencing lipid metabolism. We identified the variants rs2792751 in GPAM/GPAT1, the gene encoding glycerol-3phosphate acyltransferase, mitochondrial, and rs429358 in APOE, the gene encoding apolipoprotein E, as robustly associated with liver fat content and liver disease after adjusting for multiple testing. Both genes affect lipid metabolism in the liver. CONCLUSIONS: We identified 2 novel genetic variants in GPAM and APOE that are robustly associated with steatosis and liver damage. These findings may help to better elucidate the genetic susceptibility to FLD onset and progression.

Somatically mutated genes in fatty liver disease have minimal influence on germline risk

“…We hypothesised that germline variation providing weaker modulation of disease phenotypes would exist and might improve the performance of polygenic risk scores being developed for prognostication in NAFLD and ARLD. Amongst these genes, germline coding variants GPAM have been associated with serum ALT levels in an exome-wide association study of the UK Biobank cohort [16]. Further study found that these variants were also associated with hepatic fat content and histological markers of liver damage in independent NAFLD cohorts.…”

Insights into Nonalcoholic Fatty-Liver Disease Heterogeneity

“…[30][31][32][33] We are therefore presently moving to a model where, across the continuum of the distribution of the inherited predisposition to NAFLD in the population, the development of polygenic risk scores (PRS) can stratify the risk of this condition and of liver-related complications. 30,31,[34][35][36] For example, "high" PRS can predict the evolution to cirrhosis and HCC independently of classical risk factors in individuals with dysmetabolism and/or NAFLD, potentially being able to aid in the clinical management and in the design of referral pathways. 35 On the other hand, it has become clear that the main common risk variant and PRS cannot discriminate accurately between "metabolic" and "genetic" NAFLD.…”

“… 14 17 28 29 Due to the increased power of studies conducted in large population-based cohorts, the number of NAFLD risk variants is now increasing by the day. 30 31 32 33 We are therefore presently moving to a model where, across the continuum of the distribution of the inherited predisposition to NAFLD in the population, the development of polygenic risk scores (PRS) can stratify the risk of this condition and of liver-related complications. 30 31 34 35 36 For example, “high” PRS can predict the evolution to cirrhosis and HCC independently of classical risk factors in individuals with dysmetabolism and/or NAFLD, potentially being able to aid in the clinical management and in the design of referral pathways.…”

Untitled

“…14,17,28,29 Due to the increased power of studies conducted in large population-based cohorts, the number of NAFLD risk variants is now increasing by the day. [30][31][32][33] We are therefore presently moving to a model where, across the continuum of the distribution of the inherited predisposition to NAFLD in the population, the development of polygenic risk scores (PRS) can stratify the risk of this condition and of liver-related complications. 30,31,[34][35][36] For example, "high" PRS can predict the evolution to cirrhosis and HCC independently of classical risk factors in individuals with dysmetabolism and/or NAFLD, potentially being able to aid in the clinical management and in the design of referral pathways.…”

“…[30][31][32][33] We are therefore presently moving to a model where, across the continuum of the distribution of the inherited predisposition to NAFLD in the population, the development of polygenic risk scores (PRS) can stratify the risk of this condition and of liver-related complications. 30,31,[34][35][36] For example, "high" PRS can predict the evolution to cirrhosis and HCC independently of classical risk factors in individuals with dysmetabolism and/or NAFLD, potentially being able to aid in the clinical management and in the design of referral pathways. 35 On the other hand, it has become clear that the main common risk variant and PRS cannot discriminate accurately between "metabolic" and "genetic" NAFLD.…”