Reviewed by Dr. Jake Hutto
Stender S, et al. Nat Genet. 2017 June [49(6):842-847]

The rising prevalence of obesity has led to an increased prevalence of diabetes mellitus, hypertension, hyperlipidemia, and non-alcoholic fatty liver disease (NAFLD), which belongs on a spectrum of disorders ranging from excess liver fat (steatosis), inflammation (steatohepatitis), fibrosis (cirrhosis), and malignancy formation (hepatocellular carcinoma). Susceptibility to NAFLD is highly variable, as not all obese individuals develop NAFLD, and not all cases of steatosis progress further along the spectrum. As early as the mid-2000s, genome-wide association studies performed by UT Southwestern scientists have revealed insight into genetic variants that contribute to differences in hepatic triglyceride content (HTGC). It is now known that a single nucleotide polymorphism in the PNPLA3 gene that encodes the enzyme I148M (rs738409, called the “M variant”) is strongly associated with increased hepatic fat levels, increased serum ALT levels, and liver inflammation. This enzyme normally undergoes nutritional regulation, ensuring almost no amount of the enzyme exists in the fasted state, and increased levels in the carb-fed state. PNPLA3 normally has lipid acyl hydrolase activity and is catabolized through a process involving ubiquitylation and proteasome degradation, and the M variant of this protein has been found to be catalytically dead, accumulate on hepatocyte lipid droplets, and avoids ubiquitylation (thereby avoiding degradation). In this paper, researchers showed that adiposity amplified the effects of the PNPLA3 I148M variant on HTGC levels, serum ALT levels, and cirrhosis, and 2 new gene sequence variants were seen to have interactions with obesity (TM6SF2 and GCKR). Participants were included from 4 study cohorts, including the Dallas Heart Study (DHS), totaling over 100,000 included participants. Patients’ BMI, HTGC, and serum ALT levels were measured, and exome-wide genetic testing for other BMI-associated variants and other associations was performed. After stratifying DHS individuals into four groups based on BMI, the relationship between PNPLA3 genotype (wild-type II, heterozygous IM, or homozygous for mutant MM) and BMI was analyzed. Median HTGC in the leanest BMI group increased in a stepwise manner in the II, IM, and MM groups (1.8%, 2.3%, and 2.8%, respectively, P=0.0003). This effect was amplified as BMI increased, as median HTGC was 3x higher in MM than II individuals (14.2% vs 4.7%) in the highest BMI group. Serum ALT levels also increased in similar fashion for those with the M variant when stratified for BMI, but only in the 2 highest BMI groups. Two other risk variants associated with HTGC were also found to have similar effects on HTGC and ALT levels with increasing BMI: TM6SF2 rs58542926 (encoding p.3167K) and GCKR rs1260326 (encoding p.P446L). In one cohort, 384 participants had a diagnosis of cirrhosis, and the effect of the PNPLA3 I148M variant on prevalence of cirrhosis increased with increasing BMI as well. The odds ratio of cirrhosis development was 5.8 in MM homozygotes vs. II homozygotes. This study revealed that several genetic variants leading to increased susceptibility to NAFLD and the full spectrum of chronic liver disease have their effects amplified by increasing adiposity.

The results of this paper indicate that gene-adiposity interactions play a major role in the development and progression of NAFLD. The prevalence of hepatic steatosis was only 9% in the lowest BMI individuals with wild-type PNPLA3, and 84% in the highest BMI individuals homozygous for the PNPLA3 I148M variant. As similar gene-adiposity relationships were noted for two other risk genetic variants (TM6SF2 p.E167K and GCKR p.P446L), it appears obesity augments the genetic risk of NAFLD through at least 3 mechanisms, as all 3 of these proteins participate in different metabolic pathways. Due to the PNPLA3 M variant’s effect on cirrhosis development based on BMI, it is presumed that adiposity amplifies the effect of the M variant on the entire spectrum of liver disease, not just NAFLD. What makes the results of this paper, and previous related papers, important is the strength of these single gene–adiposity interactions. Hypertension and hyperglycemia both have common alleles that contribute to their development, but these alleles have much smaller phenotypic effects than the NAFLD susceptibility alleles. These disorders also both undergo homeostatic regulation, while currently there is no evidence that HTGC is subject to feedback regulation of any sort. Overall, these results show that considering adiposity levels in conjunction with genetic studies may improve predicting those at highest risk of progressing from steatosis to end-stage liver disease, that genetic screening for NAFLD risk alleles in high-BMI individuals may be valuable, and that aggressive weight-loss interventions are beneficial in those with steatosis, but particularly in the highest BMI groups.