Fatty liver disease (FLD) is becoming the leading cause of chronic liver disease worldwide. It is well established that genetic background is a major driver of FLD and its progression to liver inflammation and fibrosis.1

A large body of evidence demonstrates the Patatin‐like phospholipase domain‐containing 3 (PNPLA3 ) rs738409, resulting in an isoleucine to methionine substitution at position 148 (I148M) of the protein, exerts the most robust genetic effect on FLD progression. PNPLA3 is a hepatic lipase with a triglyceride hydrolase activity and its 148M variant results in a loss of this function impeding the remodelling of mono and polyunsaturated fatty acid and favouring their retention in liver.2, 3 Additionally: in hepatocytes, the 148M loss of function variant leads to an increased lipid accumulation by reducing VLDL secretion4; in stellate cells, the 148M loss of function variant leads to a retinol retention with an increase in secretion of pro‐fibrogenic and inflammatory.5, 6

PNPLA3 was firstly identified in 2001, way before the genetic discovery, in mice as an adipocyte specific nutritionally regulated transmembrane protein called Adiponutrin (Adpn).7

While the role of PNPLA3 in liver tissue has been extensively investigated, its role and the effect of the 148M variant in adipose tissue (AT) were poorly investigated so far. This is probably due to the fact that in humans, and contrarily to what detected in mice and rats, the PNPLA3 expression is higher in liver rather than in AT.8, 9 Recently, it has been shown that, in severely obese patients, PNPLA3 mRNA was strongly expressed in the liver and clearly detectable in subcutaneous adipose tissue.10 In these individuals, weight loss after bariatric surgery increased the PNPLA3 expression in adipose tissue with no changes in the hepatic expression level.10 However, differences in PNPLA3 protein synthesis/level between these two tissues as well as whether the lipidomic signature in the AT was different in carriers of the PNPLA3 148M variant have never been examined before.

Qadri et al shed some lights on this in their recent work on liver international.11 The authors performed an exhaustive lipidomic analysis with mass spectrometry and gas chromatography on subcutaneous abdominal AT biopsy of 125 patients who underwent laparoscopic bariatric surgery procedure, stratified by PNPLA3  148M genotype. In these individuals, Qadri et al additionally measured the PNPLA3 amount by western blotting. The main findings of this work are that: (a) PNPLA3 is highly synthesized in the adipose tissue although its higher mRNA levels in adipose tissue,12, 13 and (b) consistently with an appreciable synthesis of the protein in AT, carriers of the 148M genetic variant have a specific lipidomic fingerprint resembling the one observed in the liver. Interestingly, carriage of the PNPLA3 I148M variant did not affect the insulin suppression of glycerol rate of appearance, the AT lipolysis, nor levels of serum non‐esterified fatty acids (NEFAs). These results may be interpreted as a potential mechanism leading to liver disease. Indeed, in carriers of the mutant PNPLA3 protein the beneficial mono and polyunsaturated fatty acids deriving from food intake remain trapped in the AT while saturated fatty acids are preferably redirected to the liver.

In morbidly obese individuals, carriage of the PNPLA3 148M variant results in a higher insulin resistance and diabetes,14 while a recent study in mice shows how overexpression of the PNPLA3 148M resulted in higher visceral fat accumulation and more insulin resistance as compared to the wild type protein.15 One would be tempted to link these findings to the data from Qadri et al although one should refrain from doing so because Qadri et al did not examined visceral adipose tissue in their study.

The downregulation by antisense oligonucleotide (ASO) of the mutant Pnpla3 in mice has given promising results in reducing liver fat, inflammation and fibrosis.16 In this work, although the ASO had a high specificity for the liver, it was also up taken by the AT.16 If these results are confirmed, a potential treatment using PNPLA3 ASO will also have a direct effect on the adipose tissue.

This study from Qadri et al, although very elegant, has three main limitations: (a) the results may not be applicable to the normal weight individuals because the cohort examined was composed only by severe obese individuals, (b) authors did not consider the impact of the PNPLA3  148M variant on the adipokines secretion which might be involved in mediating the susceptibility to metabolic‐associate FLD17, and (c) the absence of an independent cohort warrants a replication of these data.

If confirmed by further studies, the study from Qadri et al will lead to a new definition of the PNPLA3 role, no longer only in the liver but also in the adipose tissue, helping understanding the cross talk between adipose and liver tissue in FLD pathogenesis.