PNPLA3 (I148M), TM6SF2, and MBOAT7 in MASLD and Cirrhosis

PNPLA3 (I148M), TM6SF2, and MBOAT7 form the primary oligogenic determinant of metabolic dysfunction-associated steatotic liver disease (MASLD, formerly termed nonalcoholic fatty liver disease, NAFLD), its progressive necroinflammatory form metabolic dysfunction-associated steatohepatitis (MASH), bridging cirrhosis, and hepatocellular carcinoma (HCC). Across global genome-wide and exome-wide association cohorts, common coding and regulatory alleles at three hepatocyte-expressed loci—patatin-like phospholipase domain-containing protein 3 (PNPLA3 on chromosome 22q13.31), transmembrane 6 superfamily member 2 (TM6SF2 on Chr19p13.11), and membrane-bound O-acyltransferase domain-containing 7 (MBOAT7 on Chr19q13.42)—control the partitioning of neutral triacylglycerols (TAG), retinyl esters, and arachidonoyl-containing phospholipids across the endoplasmic reticulum (ER) and cytoplasmic lipid droplets (LDs).

Whereas simple hepatic triglyceride accumulation (>5% liver weight) is frequently benign in isolation, the synergistic convergence of PNPLA3 Ile148Met (rs738409), TM6SF2 Glu167Lys (rs58542926), and MBOAT7 rs641738 impairs hepatic lipid turnover, traps polyunsaturated fatty acids (PUFAs) inside enlarged droplets, and triggers lipotoxic organelle stress that activates resident Kupffer cells and collagen-depositing hepatic stellate cells (HSCs). Conversely, human population genomics has revealed protective loss-of-function epistasis mediated by HSD17B13 (rs72613567:TA), MARC1 (p.Ala165Thr), and PSD3 (p.Leu186Thr), establishing validated genetic targets for N-acetylgalactosamine (GalNAc)-conjugated small interfering RNA (siRNA) and antisense oligonucleotide (ASO) liver therapeutics.

Patatin Catalytic Dyad and Ubiquitylation Evasion at Residue 148

Encoded by the PNPLA3 gene on human chromosome 22q13.31 (9 exons, 481 amino acids, ~53 kDa), PNPLA3 (also known as adiponutrin) is a membrane-anchored lipid-droplet enzyme sharing high N-terminal sequence homology (~37% identity across the patatin domain, residues 10–179) with adipose triglyceride lipase (ATGL / PNPLA2). Wild-type PNPLA3 functions as a triacylglycerol lipase and acylglycerol transacylase in hepatocytes and retinol-storing hepatic stellate cells through a catalytic Ser47–Asp166 dyad embedded within a canonical Gly-X-Ser-X-Gly (GXSXG) nucleophilic elbow (residues 45–49). Under fasting conditions, PNPLA3 transcription is minimal; following postprandial insulin and glucose stimulation, carbohydrate-responsive element-binding protein (ChREBP) and sterol regulatory element-binding protein 1c (SREBP-1c) upregulate PNPLA3 expression by more than 10- to 25-fold.

The common cytosine-to-guanine transversion rs738409 (c.444C>G, p.Ile148Met or I148M) substitutes a bulky methionine side chain for isoleucine at position 148 inside the hydrophobic substrate-binding groove adjacent to the Ser47–Asp166 catalytic dyad. This substitution reduces intrinsic triglyceride hydrolase and retinyl-palmitate lipase V_max by >80% without altering substrate binding affinity (K_m). However, simple genetic deletion of Pnpla3 in mice (Pnpla3^−/−) fails to cause hepatic steatosis, proving that I148M acts via a toxic gain-of-function (neomorphic) mechanism on the lipid droplet surface. Specifically, wild-type PNPLA3 undergoes rapid turnover (t_1/2 ≈ 3–4 hours) via ubiquitylation at Lys434/Lys437 and degradation by the 26S proteasome and chaperone-mediated autophagy. The I148M substitution alters the C-terminal conformation so that mutant PNPLA3-I148M evades ubiquitylation (BASP1/ubiquitin E3 ligase escape), accumulating to 10- to 40-fold higher protein density on the phospholipid monolayer of hepatocyte lipid droplets.

Sequestration of ABHD5 (CGI-58) and Inhibition of ATGL Lipolysis

The pathogenic consequence of PNPLA3-I148M accumulation on lipid droplets was elucidated by Hobbs, Cohen, and colleagues: hyper-abundant PNPLA3-I148M binds with high affinity (K_d in the nanomolar range) to α/β-hydrolase domain-containing protein 5 (ABHD5, also designated CGI-58), the obligatory rate-limiting co-activator required for ATGL (PNPLA2) catalytic activity. By competitively sequestering ABHD5/CGI-58 into catalytically dead PNPLA3-I148M:CGI-58 complexes on the droplet surface, PNPLA3-I148M starves endogenous ATGL of its co-activator, stalling basal and β-adrenergic-stimulated mobilization of hepatic triglycerides. Simultaneously, within hepatic stellate cells, PNPLA3-I148M blocks the hydrolysis of intracellular retinyl-ester stores (retinyl palmitate), triggering stellate cell transdifferentiation into myofibroblasts that secrete transforming growth factor-β1 (TGF-β1), tissue inhibitor of metalloproteinases-1 (TIMP-1), and fibrillar Type I and Type III collagens (COL1A1, COL3A1).

TM6SF2 Glu167Lys and the Hepatic-Cardiovascular Dissociation Paradox

Located on chromosome 19p13.11, TM6SF2 encodes a 377-amino-acid, 10-transmembrane-domain polytopic protein localized to the smooth endoplasmic reticulum (ER) and the ER-Golgi intermediate compartment (ERGIC) of hepatocytes and enterocytes. In the two-step assembly of very-low-density lipoproteins (VLDL), microsomal triglyceride transfer protein (MTTP) first cotranslationally lipidates apolipoprotein B-100 (ApoB-100) in the rough ER lumen to form a dense primordial particle; in the second step, TM6SF2 complexes with ApoB-100 and ER lipid-droplet-associated proteins (ERLIN1/ERLIN2) to catalyze the bulk fusion of luminal neutral triglycerides (TAG) and cholesteryl esters onto the maturing VLDL1 particle prior to Golgi secretion.

The missense variant rs58542926 (c.499G>A, p.Glu167Lys or E167K) replaces a negatively charged glutamate with a positively charged lysine within an exposed cytoplasmic loop, destabilizing the TM6SF2 protein fold and reducing steady-state hepatic TM6SF2 protein levels by ~50% to 70%. Without sufficient TM6SF2, hepatocytes secrete triglyceride-depleted, smaller VLDL2 particles and retain excess unexported triacylglycerols within cytosolic lipid droplets. This creates a striking hepatocyte-vascular dissociation: carriers of TM6SF2 E167K suffer a 1.9-fold higher risk of MASH fibrosis, cirrhosis, and HCC, yet exhibit 10% to 15% lower circulating plasma triglycerides and low-density lipoprotein cholesterol (LDL-C), conferring a ~12% to 18% reduction in myocardial infarction (MI) and coronary artery disease risk—the mirror image of PCSK9 or ANGPTL3 inhibition.

MBOAT7 (LPIAT1) and Phosphatidylinositol Lands Cycle Remodeling

The third major susceptibility locus, rs641738 C>T at the TMC4–MBOAT7 boundary on chromosome 19q13.42, acts by downregulating hepatic mRNA and protein levels of MBOAT7 (also termed lysophosphatidylinositol acyltransferase 1, LPIAT1) by ~35% to 45%. An integral 6-transmembrane ER and mitochondrial-associated membrane (MAM) enzyme containing a catalytic His356–Asn321 dyad, MBOAT7 operates within the Lands phospholipid remodeling cycle, specifically transferring an polyunsaturated arachidonoyl (20:4 n-6) acyl chain from arachidonoyl-CoA onto the sn-2 position of lysophosphatidylinositol (LPI) to synthesize 38:4 phosphatidylinositol (1-stearoyl-2-arachidonoyl-sn-glycero-3-phosphoinositol). When MBOAT7 activity declines in rs641738-T carriers, hepatic membranes become depleted of arachidonoyl-PI and accumulate proinflammatory 18:0 and 18:1 lysophosphatidylinositol (LPI) species. Even in the absence of SREBP-1c activation, impaired phosphatidylinositol turnover accelerates phosphoinositide phospholipase C (PLC) cleavage of PI into diacylglycerol (DAG), driving non-canonical DAG acyltransferase-2 (DGAT2)-mediated triglyceride synthesis and directly triggering hepatic stellate cell collagen deposition.

Polygenic Risk Amplification and Protective HSD17B13 / MARC1 Rescue

In clinical hepatology, the risk of progressing from asymptomatic steatosis to decompensated cirrhosis and HCC scales super-additively across the PNPLA3–TM6SF2–MBOAT7–HSD17B13 genetic axis. In the UK Biobank and FinnGen cohorts, individuals carrying 4 or more risk alleles across PNPLA3 (I148M), TM6SF2 (E167K), and MBOAT7 (rs641738-T) coupled with concurrent obesity (BMI > 30 kg/m²) or type 2 diabetes exhibit a 12- to 18-fold increase in cirrhosis prevalence and a >20-fold increase in HCC incidence compared to lean non-carriers. Conversely, co-inheritance of loss-of-function variants in HSD17B13 (rs72613567:TA), MARC1 (mitochondrial amidoxime-reducing component 1, p.Ala165Thr rs2642438, which lowers nitric oxide and lipid peroxide generation on outer mitochondrial membranes), or PSD3 (p.Leu186Thr) epistatically blunts PNPLA3-I148M-driven lobular necroinflammation and fibrosis.

GalNAc-siRNA and ASO Silencing of PNPLA3 I148M

Because PNPLA3-I148M causes steatohepatitis via toxic protein accumulation and CGI-58 sequestration rather than simple loss of enzymatic function, therapeutic intervention requires depleting mutant PNPLA3 mRNA and protein in hepatocytes rather than overexpressing wild-type PNPLA3. Triantennary N-acetylgalactosamine (GalNAc)-conjugated siRNAs (JNJ-75220795, AMG-609, ALN-PNP) and antisense oligonucleotides (AZD2693) selectively enter hepatocytes via the asialoglycoprotein receptor (ASGPR1) and trigger RNA-induced silencing complex (RISC) or RNase H1 cleavage of PNPLA3 transcripts. In Phase 1 and Phase 2 clinical trials specifically recruiting PNPLA3 148M/148M homozygous MASH patients, subcutaneous GalNAc-siRNA dosing frees CGI-58 to restore endogenous ATGL lipolysis, achieving a 45% to 70% relative reduction in magnetic resonance imaging proton density fat fraction (MRI-PDFF) and sustained normalization of serum alanine aminotransferase (ALT).

Why does knocking out the Pnpla3 gene cure fatty liver in I148M mice, whereas overexpressing PNPLA3-I148M causes severe steatosis?

The PNPLA3 I148M variant is a neomorphic (toxic gain-of-function) allele rather than a simple null mutation. Mutant PNPLA3-I148M protein evades ubiquitylation and accumulates up to 40-fold on lipid droplets, where it binds and sequesters ABHD5 (CGI-58), the obligatory co-activator of adipose triglyceride lipase (ATGL / PNPLA2). Removing PNPLA3-I148M via genetic knockout or GalNAc-siRNA releases CGI-58 back to ATGL, restoring normal hepatocyte triglyceride lipolysis.

How does the TM6SF2 E167K variant simultaneously cause liver cirrhosis and protect against heart attacks?

TM6SF2 facilitates the bulk addition of neutral triglycerides to nascent ApoB-100 VLDL particles in the hepatocyte endoplasmic reticulum. The destabilizing E167K (rs58542926) variant reduces hepatic TM6SF2 levels, trapping triglycerides inside the liver (raising MASH cirrhosis and HCC risk by ~1.9-fold) while reducing the secretion of atherogenic VLDL and lowering plasma LDL-C and triglycerides, which decreases myocardial infarction risk by ~12% to 18%.

What is the mechanism linking MBOAT7 (rs641738) deficiency to non-canonical hepatic lipogenesis?

MBOAT7 (LPIAT1) catalyzes the insertion of arachidonic acid (20:4 n-6) into lysophosphatidylinositol (LPI) within the Lands phospholipid remodeling cycle. The common rs641738-T allele lowers hepatic MBOAT7 expression by ~40%, depleting 38:4 phosphatidylinositol (PI) in endoplasmic reticulum membranes. To replenish PI, hepatocytes accelerate phosphatidylinositol synthesis and turnover, which generates excess diacylglycerol (DAG) that feeds directly into DGAT2-driven triglyceride synthesis and promotes stellate cell fibrosis independently of insulin or SREBP-1c.

1. Romeo, S., Kozlitina, J., Xing, C., et al. (2008). "Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease." Nature Genetics, 40(12), 1461–1465. https://doi.org/10.1038/ng.257
2. Wang, Y., Kinch, L. N., Mitsche, M. A., et al. (2019). "Ainya-like mechanism: PNPLA3(I148M) promotes hepatic steatosis by interfering with ATGL-mediated lipolysis through CGI-58 sequestration." Cell Metabolism, 29(5), 1187–1200. https://doi.org/10.1016/j.cmet.2019.01.018
3. Kozlitina, J., Smagris, E., Stender, S., et al. (2014). "Exome-wide association study identifies a TM6SF2 variant that confers susceptibility to nonalcoholic fatty liver disease." Nature Genetics, 46(4), 352–356. https://doi.org/10.1038/ng.2901
4. Tanaka, Y., Shimanaka, Y., Caddeo, A., et al. (2021). "LPIAT1/MBOAT7 depletion increases triglyceride synthesis fueled by high phosphatidylinositol turnover." Gut, 70(1), 180–193. https://doi.org/10.1136/gutjnl-2020-320646