PNPLA3 I148M variant links to adverse metabolic traits in MASLD during fasting and feeding

Abstract

Background & aims: The PNPLA3 rs738409 polymorphism is the most abundant genetic risk factor associated with progression of metabolic dysfunction-associated steatotic liver disease (MASLD) to steatohepatitis (MASH) and fibrosis. Although fasting and feeding affect PNPLA3 expression, molecular insights into the pathophysiological influence remain scarce.

Methods: We analyzed 353 serum samples of patients with MASLD from two German tertiary centers using nuclear magnetic resonance (NMR)-proteometabolomics. Patients were stratified by PNPLA3 rs738409 C>G genotype: 'CC', 'CG', and 'GG'. Metabolites, lipoproteins, and glycoproteins were assessed based on fasting status.

Results: PNPLA3 GG displayed a distinct metabolic profile, with notable alterations between fasting and non-fasting states. During the latter, GG carriers showed lower levels of VLDL-1, reflecting impaired triglyceride (TG) efflux from hepatocytes. Following an overnight fast, GG carriers exhibited higher tricarboxylic acid cycle metabolites and ketone bodies, overall indicating increased β-oxidation likely attributed to lower PNPLA3 expression, facilitating unrestricted adipose triglyceride lipase activity and consecutive increased hepatic TG secretion. In addition, the ketogenic amino acid lysine, critical for mitochondrial carnitine transport, was significantly reduced (GG 0.14 $\pm$ 0.09 mM vs. CC 0.18 $\pm$ 0.08 mM, q = 0.015). Consistently, TGs were enriched in LDL and HDL particles, and an increased number of intermediate-density lipoproteins emerged as a distinct marker in fasted GG carriers (GG 202.9 $\pm$ 68.2 mg/dl vs. CC 160.8 $\pm$ 65.6 mg/dl, q = 0.007). These metabolic changes were enhanced in patients with type 2 diabetes mellitus and/or obesity.

Conclusions: Our findings suggest a dichotomous pattern of increased hepatic lipid storage during feeding and excessive lipid oxidation during fasting, which exceeds metabolic capacity, inducing cellular toxicity in PNPLA3 GG carriers. This interplay fuels a detrimental fasting/non-fasting cycle, thus pointing to the need for preventive strategies.

Impact and implications: The PNPLA3 rs738409 (p.I148M) polymorphism is the most prevalent genetic risk factor for metabolic dysfunction-associated steatotic liver disease progression and is influenced by fasting and feeding cycles. However, the pathophysiological consequences of this regulation remain poorly understood. Nuclear magnetic resonance-proteometabolomics reveals a distinct signature in homozygous PNPLA3 GG carriers that changes significantly with fasting status, providing important implications for diagnosis and preventive strategies.

Keywords: Fasting; Lipoproteins; MASLD; NMR-proteometabolomics; PNPLA3.

Graphical abstract

Fig. 1

Distinct NMR metabolite and lipoprotein profiles in PNPLA3 GG carriers. (A) Schematic overview of the study design. (B) Frequencies of PNPLA3 rs738409 C>G genotypes. (C) Distribution of PNPLA3 GG vs. CC carriers based on the O-PLS-DA of metabolites and lipoproteins using parameters with significant differences between the groups (q value <0.05), loading plot of significant parameters (red and blue indicate increase and decrease in GG carriers, respectively), and corresponding ROC curves. (D) Box plots of the three most significant lipoproteins and the most significant metabolite between groups (bars represent SD, and bold lines within the box plots represent medians). Levels of significance: ∗∗q <0.01; ∗∗∗q <0.001; ∗∗∗∗q <0.0001 (Mann–Whitney U test; FDR 5%). FDR, false discovery rate; NMR, nuclear magnetic resonance; O-PLS-DA, orthogonal partial least squares discriminant analysis; PNPLA3, patatin-like phospholipase domain-containing protein; ROC, receiver operating characteristic.

Fig. 2

Metabolite differences in the fasting and non-fasting states in PNPLA3 GG compared with CC carriers. Distribution of PNPLA3 GG vs. CC carriers based on the O-PLS-DA of metabolites in patients (A) after an overnight fast (GG n = 32/CC n = 83) and (B) in a non-fasting condition (GG n = 17/CC n = 87). (C) Volcano plot illustrating fasting metabolites of PNPLA3 GG carriers compared with CC carriers. (D) Significant fasting metabolites between groups (bars represent SD, and bold lines within the box plots represent medians). Levels of significance: ∗q <0.05; ∗∗q <0.01; ∗∗∗q <0.001 (Mann–Whitney U test; FDR = 5%). FDR, false discovery rate; O-PLS-DA, orthogonal partial least squares discriminant analysis; PNPLA3, patatin-like phospholipase domain-containing protein.

Fig. 3

Altered fasting and non-fasting lipoprotein composition in PNPLA3 GG carriers. Volcano plot showing (A) fasting and (B) non-fasting lipoproteins of PNPLA3 GG (n = 32 (A)/17 (B)) compared with CC carriers (n = 83 (A)/87 (B)). (C) Comparison of absolute concentrations of lipoprotein subgroups and their lipid composition between groups (PNPLA3 GG vs. PNPLA3 CC). Color coding represents the fold change in mean concentrations between groups. The brighter the red or blue color, the greater the increase or decrease in the absolute concentration of a lipoprotein subfraction in PNPLA3 GG compared with PNPLA3 CC, respectively, in both the fasting and non-fasting states. Significant differences in (D) fasting lipoproteins and (E) specific non-fasting lipoproteins between groups (bars represent SD, and bold lines within the box plots represent medians). Levels of significance: ∗q <0.05; ∗∗q <0.01; ∗∗∗q <0.001; ∗∗∗q <0.001 (Mann–Whitney U test; FDR = 5%). FDR, false discovery rate; PNPLA3, patatin-like phospholipase domain-containing protein.

Fig. 4

PNPLA3 GG carriers with T2DM and obesity show pronounced changes in the metabolome. (A) Distribution of PNPLA3 GG carriers with T2DM and obesity (n = 11) compared with CC carriers with T2DM and obesity (n = 46) using the O-PLS-DA of metabolites and lipoproteins, with a loading plot showing significant parameters (q value <0.05) (red and blue indicate increase and decrease in GG carriers, respectively) and corresponding ROC curves. Separation of PNPLA3 GG vs. CC carriers with T2DM and obesity based on the O-PLS-DA of (B) fasting metabolites and (C) fasting lipoproteins. (D) Comparison of absolute concentrations of fasting metabolites and lipoprotein subgroups and their lipid composition between groups (PNPLA3 GG vs. PNPLA3 CC) with T2DM, obesity, or both conditions. Color coding represents the fold change in mean concentrations between groups. The brighter the red or blue color, the greater the increase or decrease in the absolute concentration of parameters in fasting PNPLA3 GG compared with CC carriers, respectively. Levels of significance: ∗q <0.05; ∗∗q <0.01; ∗∗∗q <0.001 (Mann–Whitney U test; FDR = 5%). FDR, false discovery rate; O-PLS-DA, orthogonal partial least squares discriminant analysis; PNPLA3, patatin-like phospholipase domain-containing protein; ROC, receiver operating characteristic; T2DM, type 2 diabetes mellitus.

Fig. 5

Variations in glycoprotein and phosphatidylcholine profiles according to PNPLA3 genotype. (A) Separation of PNPLA3 GG vs. CC carriers based on the O-PLS-DA of glycoproteins and SPC signal. (B) Distribution of patients with FIB-4 <1.3 vs. >1.3 within the separation of genotypes based on glycoproteins and SPCs. Levels of significance: ∗∗q <0.01 (Mann–Whitney U test; FDR = 5%). (C) ROC curves corresponding to O-PLS-DA of glycoproteins and SPCs. (D) Box plots of specific ratios (GlycAB/SPC and GlycA/GlycB) comparing groups (bars represent SD, and bold lines within the box plots represent medians). FDR, false discovery rate; O-PLS-DA, orthogonal partial least squares discriminant analysis; PNPLA3, patatin-like phospholipase domain-containing protein; ROC, receiver operating characteristic; SPC, supramolecular phosphatidylcholine composite signal.