Mutant PNPLA3 I148M protein as pharmacological target for liver disease

Potential conflict of interest: Nothing to report. See Article On Page 1111 Following identification of the I148M protein variant of patatin‐like phospholipase domain‐containing 3 (PNPLA3) as a dominant genetic determinant of hepatic fat content,1 robust evidence has accumulated that the mutant (I148M) variant protein is a major determinant of the full spectrum of nonalcoholic fatty liver disease,2 as well as of steatohepatitis of diverse etiologies and of the progression toward cirrhosis and liver cancer.3 However, the precise molecular mechanisms underlying these associations remain to be defined. Indeed, while the I148M variant is associated with loss of enzymatic activity,4 genetic deletion of PNPLA3 in knockout mice confers no obvious hepatic lipid phenotype, while predisposition to steatosis seems to depend upon the accumulation of PNPLA3 on the surface of lipid droplets in mice overexpressing the mutant protein.5 In this issue, BasuRay and colleagues examined the mechanism linking the I148M variant with hepatic fat accumulation in Pnpla3I148M knock‐in mice compared to wild‐type controls.6 Starting from their previous discoveries, the authors confirm that triglyceride synthesis, oxidation, and secretion are unaltered in high sucrose–fed Pnlpa3I148M knock‐in mice, despite the observation that these mice develop hepatic steatosis. These findings are consistent with a direct effect of the mutant 148M protein in the induction of lipid retention by altering triglyceride remodeling in the lipid droplet.7 Among the key findings are that accumulation of the mutant 148M protein precedes hepatic fat accumulation and is dependent on reduced ubiquitylation and impaired proteasomal degradation of I148M PNPLA3. The ligation of ubiquitin molecules to Lys (K) or Met (M) residues is a highly regulated process responsible for posttranslational regulation of protein activity, localization, and for targeting proteins for proteasomal degradation. In the present study, in vivo inhibition of proteasomal activity, but not macroautophagy, markedly delayed wild‐type PNPLA3 degradation during fasting, revealing that the protein undergoes polyubiquitylation. However, proteasome inhibition did not further increase the high baseline levels of PNPLA3 148M protein, which was also shown to be less ubiquitylated. This phenomenon was confirmed in mice overexpressing either human wild‐type or mutant (I148M) PNPLA3, which is an important control, because human PNPLA3 has a longer protein sequence and higher baseline hepatic expression than the mouse ortholog.9 Finally, mice carrying another artificial mutant form of PNPLA3 (S47A) that disrupts its catalytic domain displayed a phenotype similar to that observed in Pnpla3I148M knock‐in mice. This last finding suggests that it is the lack of enzymatic activity that determines PNPLA3 entrapment and its association with the lipid droplet surface, where it is likely less accessible to ubiquitin ligases. The next steps will be to find out how this loss of enzymatic activity impairs PNPLA3's trafficking across intracellular membrane compartments, to elucidate the mechanisms and sites of PNPLA3 ubiquitylation, and to determine how the mutant PNPLA3 148M functions to inhibit the activity of other lipases on the lipid droplet (mainly PNPLA2/adipose triglyceride lipase), allowing fat accumulation. In a broader context, human genetic studies also support the hypothesis that accumulation of mutant PNPLA3 148M protein is necessary to induce liver damage.10 Along these lines another naturally occurring PNPLA3 polymorphism, rs2294918 encoding the E434K protein variant, was recently linked to reduced hepatic PNPLA3 protein abundance.10 Whether the 434K residue is ubiquitylated and contributes to the decrease PNPLA3 protein levels is presently not known, although predictive algorithms suggest that this residue is a candidate site in human PNPLA3. In contrast with the more common PNPLA3 combined 148M‐434E risk variant, carriage of the 148M‐434K version of the protein, which is expressed at low levels, did not predispose to liver damage. However, in the context of preserved enzymatic activity, the 434K variant slightly reduced the protective impact of wild‐type 148I PNPLA3.10 Taken together, these data suggest that accumulation of the 148M, but not wild‐type, PNPLA3 protein drives hepatic fat accumulation and that it may be possible to mitigate the effects of the 148M mutant by preventing its accumulation. Furthermore, because of the association in patients with nonalcoholic fatty liver disease with the 148M variant, we demonstrated that PNPLA3 expression in hepatic stellate cells is increased in the liver of carriers of the mutation, which is important because stellate cells are involved in the regulation of fibrogenesis and carcinogenesis.11 Paralleling what happens in hepatocytes, overexpression of the mutant I148M PNPLA3 protein in stellate cells in vitro impaired the release of retinol and lipids that results from dynamic remodeling of intracellular lipid droplets, resulting in a more fibrogenic phenotype.11 This conceptual framework opens up new avenues for the treatment of liver disease. Indeed, the PNPLA3 148M variant is highly and dose‐dependently enriched in individuals with progressive liver disease due to dietary factors/obesity/insulin resistance, alcohol abuse, and other hepatotoxic insults, with most at‐risk individuals carrying at least one copy of the variant allele.1 Therefore, agents that specifically decrease the abundance of the mutant PNPLA3 148M protein, e.g., by interfering with its translation, would be predicted to decrease liver damage by reducing the impairment in lipid droplet remodeling in hepatocytes and stellate cells. Conversely, down‐regulation of wild‐type PNPLA3 might increase liver damage. Finally, as discussed by the authors of the current report, although modulation of protein degradation would not be feasible, due to anticipated off‐target effects, the findings raise the possibility for a targeted, precision medicine approach to treat a large number of patients affected by a very common condition. A schematic of the findings of BasuRay and colleagues6 and of the possible therapeutic implications is presented in Fig. 1.Figure 1: (A) Wild‐type PNPLA3 is induced concomitantly with increased free fatty acid availability during hyperinsulinemia and expressed in the endoplasmic reticulum and on the surface of lipid droplets, where it mediates triglyceride turnover. However, it is rapidly ubiquitylated and degraded upon fasting. (B) PNPLA3 148M localizes on the lipid droplets, thereby escaping degradation; and its accumulation favors lipid retention and steatohepatitis. (C) Reduced expression of PNPLA3 148M due to the copresence of the 434K variant seems to prevent liver damage expression. This genetic model suggests that pharmacological approaches aimed at reducing PNPLA3 148M protein synthesis would have a beneficial effect on liver disease by restoring triglyceride remodeling and dismissal. Wild‐type PNPLA3 (148I) with intact protective enzymatic activity is shown as green curves and PNPLA3 148M causing lipid droplet formation as red curves. The number is proportional to protein levels. Red arrows indicate damaging pathways; green arrows protective pathways. Dashed lines denote suppressed pathways. Abbreviations: TAG, triglyceride; Ub, ubiquitin. Author names in bold designate shared co‐first authorship.