MASLD has a strong genetic component with many common genetic variants modulating its onset and progression identified so far. Among these, the rs738409 in the patatinlike phospholipase domain-containing 3 (PNPLA3) [4] and the rs58542926 in the transmembrane 6 superfamily member 2 (TM6SF2) [5] genes are the strongest and the most widely validated.

PNPLA3 is a triglyceride lipase [6,7] and acyltransferase [8] primarily localized on lipid droplets, where it performs its enzymatic activity, and on endoplasmic reticulum [9]. PNPLA3 is required for triglycerides and phospholipids remodeling by modulating the transfer of very long–chain polyunsaturated fatty acids (PUFAs) from triglycerides to phospholipids [6,8]. The rs738409 variant, encoding for an isoleucine to methionine substitution at position 148 of the protein (p.I148M), results in a loss of its enzymatic activity [7]. Moreover, the PNPLA3 I148M mutant protein accumulates in the LDs causing very low-density lipoprotein (VLDL) retention [10], increases ketogenesis and decreases hepatic de novo lipogenesis and mitochondrial function [11].

M6SF2 is localized on endoplasmic reticulum, and it is required VLDL secretion and hepatic lipid droplets content [12]. The rs58542926 variant, encoding for a glutamic acid to lysine substitution at position 167 of the protein (p.E167K), results in lower TM6SF2 expression and stability5 and higher hepatic fat accumulation due to VLDL retention [13,14].

Although TM6SF2 and PNPLA3 are both localized on endoplasmic reticulum, highly expressed in hepatocytes, involved in lipid metabolism and in MASLD onset and progression, there was no evidence regarding their direct interaction.

In this edition of Clinical and Molecular Hepatology, Sun et al. [15] have now demonstrated a potential physical interaction between TM6SF2 and PNPLA3 in endoplasmic reticulum. The p.E167K variant enhances this interaction, impairs PNPLA3 migration to lipid droplets, and therefore decreases its lipase activity and the PNPLA3-mediated acyl-chain transfer of PUFAs from triglycerides to phosphatidylcholine on lipid droplets. This finally results in hepatic accumulation of PUFA-containing triglycerides content.

More specifically, authors generated a Tm6sf2^167K/K knock-in mouse model by CRISPR/Cas9 technology. The Tm6sf2^167K/K homozygous group fed a high fat diet (HFD) displayed an increased body and liver weight, increased hepatic index, and a more severe degree of steatosis compared to the control group Tm6sf2^167E/E (wild type, WT). Moreover, these mice had higher plasma alanine aminotransferase and aspartate aminotransferase levels compared to the WT group, suggesting that the TM6SF2 E167K variant may induce hepatic injury under HFD exposure. However, no differences in plasma triglycerides nor in total cholesterol levels were detected. This is in contrast with human data showing that carriers of the TM6SF2 E167K variant are characterized by lower plasma cholesterol and lipoprotein concentrations conferring protection against cardiovascular disease [16].

Then, to further investigate the impact of the TM6SF2 E167K variant on hepatic lipid composition, authors performed lipidomic analysis on liver tissue from both Tm6sf2^167K/K and WT mice. They found higher hepatic triglycerides and lower phosphatidylcholine content in the Tm6sf2^167K/K compared to the WT mice. Moreover, the Tm6sf2^167K/K mice showed higher levels of monounsaturated fatty acids- and PUFA-containing triglycerides in the liver, and lower levels of PUFA-containing phosphatidylcholine.

Authors speculated that the TM6SF2 E167K variant may affect the transfer of PUFAs [17] between triglycerides and phosphatidylcholine as PUFA-containing phosphatidylcholine/triglycerides ratio was strongly reduced in the HFD-induced Tm6sf2^167K/K compared to WT mice. Among all PUFAs, Tm6sf2^167K/K mice showed lower hepatic level of phosphatidylcholine containing alpha-linolenic acid (C18:3) suggesting a pivotal role of this fatty acid in facilitating liver fat accumulation in carriers of the E167K variant. Interestingly, authors detected lower circulating phosphatidylcholine containing C18:3 levels in MASLD individuals carrying the TM6SF2 variant compared to those homozygous for the TM6SF2 EE genotype. To test if phosphatidylcholine containing C18:3 supplementation may attenuate hepatic steatosis and injury in TM6SF2 E167K carriers, authors fed Tm6sf2^167K/K and Tm6sf2^167E/E mice a HFD supplemented with phosphatidylcholine containing C18:3. They found that this dietary addition completely abolished all liver damages induced by the E167K variant in HFD-fed mice. These results indicate that phosphatidylcholine containing C18:3 deficiency may be considered as a sole phenotype of MASLD individuals carrying the TM6SF2 E167K variant, and lifestyle modifications toward a personalized medicine, such as individualized diet interventions, may mitigate the natural history of this disorder.

To understand the mechanism behind these lipid changes, authors used string protein interaction network to screen for known and predicted protein interaction, and PNPLA3 resulted as a potential functionally related protein. In silico data were confirmed by immunofluorescence and co-immunoprecipitation analyses that revealed a co-localization of TM6SF2 and PNPLA3 proteins in mouse primary hepatocytes treated with free fatty acids. Noteworthily, the TM6SF2 E167K variant strengthened the interaction between TM6SF2 and PNPLA3 in the endoplasmic reticulum by boosting PNPLA3 distribution in this intracellular compartment and reducing its content on lipid droplets, thus strongly impacting its lipase and acyl-transferase activities [8]. Importantly, PNPLA3 148M variant did not alter the interaction or distribution of these two proteins.

Then, to test if TM6SF2 E167K variant might affect the PNPLA3-mediated transfer of PUFAs from triglycerides to phosphatidylcholine, authors measured the newly-synthesized PUFA-containing phosphatidylcholine/triglycerides ratio by using radiolabeled PUFAs. They found lower newly-synthesized PUFA-containing phosphatidylcholine/triglycerides ratio in Tm6sf2^167K/K primary hepatocytes compared to WT. These changes were reversed upon Pnpla3 knockdown, confirming that the TM6SF2 E167K variant negatively modulates PNPLA3 enzymatic activity.

Since PUFAs are susceptible to oxygen free radicals attack, authors tested the impact of the TM6SF2 E167K variant on oxidative stress in mouse primary hepatocytes. They found higher levels of oxidative stress in Tm6sf2^167K/K primary hepatocytes exposed to free fatty acids compared to WT. This suggests that TM6SF2 E167K may facilitate lipid-induced reactive oxygen species accumulation and negatively impact on cell-membrane fluidity due to lower levels of hepatic PUFA-containing phosphatidylcholine.

In conclusion, the paper from Sun et al. [15] showed that a) TM6SF2 physically interacts with PNPLA3 and the E167K increases this binding affinity. The interaction between these two proteins impairs PNPLA3 localization on lipid droplets and results in a strong reduction of the PNPLA3-mediated lipolysis and acyl-chain transfer of PUFAs from triglycerides to phosphatidylcholine, finally inducing accumulation of PUFAs-reach triglycerides, and b) dietary implementation of C18:3-containing phosphatidylcholine abolished all E167K variant-induced liver damages in HFDfed mice proposing this as a potential strategy to treat MASLD in TM6SF2 E167K carriers in a frame of precision medicine.

Considering this impressive work in mice, Sun et al. strongly contributed to elucidate the role of TM6SF2 in the pathogenesis of MASLD.

A main limitation of this work is that the data derive from mouse experiments. Indeed, authors did not find any differences in protein stability between the wild type and mutant TM6SF2 protein and, therefore, the relevance in humans should be demonstrated for example by using human liver organoids selected based on the PNPLA3 and TM6SF2 genetic alterations [18].

Moreover, it would be interesting to test if TM6SF2 E167K may interact with other validated MASLD genetic risk factors such as MBOAT7, a lysophosphatidylinositol acyltransferase (LPIAT1) localized on endoplasmic reticulum [19] and involved in the acyl chain remodeling of phospholipids and triglycerides as well as in the maintenance of cell membrane lipid composition and fluidity [19,20].