Depletion of mitochondrial methionine adenosyltransferase α1 triggers mitochondrial dysfunction in alcohol-associated liver disease

Abstract

MATα1 catalyzes the synthesis of S-adenosylmethionine, the principal biological methyl donor. Lower MATα1 activity and mitochondrial dysfunction occur in alcohol-associated liver disease. Besides cytosol and nucleus, MATα1 also targets the mitochondria of hepatocytes to regulate their function. Here, we show that mitochondrial MATα1 is selectively depleted in alcohol-associated liver disease through a mechanism that involves the isomerase PIN1 and the kinase CK2. Alcohol activates CK2, which phosphorylates MATα1 at Ser114 facilitating interaction with PIN1, thereby inhibiting its mitochondrial localization. Blocking PIN1-MATα1 interaction increased mitochondrial MATα1 levels and protected against alcohol-induced mitochondrial dysfunction and fat accumulation. Normally, MATα1 interacts with mitochondrial proteins involved in TCA cycle, oxidative phosphorylation, and fatty acid β-oxidation. Preserving mitochondrial MATα1 content correlates with higher methylation and expression of mitochondrial proteins. Our study demonstrates a role of CK2 and PIN1 in reducing mitochondrial MATα1 content leading to mitochondrial dysfunction in alcohol-associated liver disease.

Fig. 1. Mitochondrial MATα1 level is selectively reduced in alcohol-associated liver disease.

Western blots (a) and densitometry analyses (b) of total, cytosolic and mitochondrial MATα1 in human normal and AH livers (n = 5 normal and n = 6 AH independent patients. p = 0.027 for total and p = 0.0006 for mitochondrial MATα1). Western blots (c) and densitometry analyses (d) of total, cytosolic and mitochondrial MATα1 in pair-fed and ethanol-fed mouse livers (n = 5 pair-fed and n = 6 ethanol-fed independent animals. p = 0.049 for total and p = 0.0004 for mitochondrial MATα1). Western blots (e, f) and densitometry analyses (g) of total, cytosolic and mitochondrial MATα1 in AML-12 cells treated with ethanol (n = 3 independent experiments. p = 0.018 for total and p = 0.024 for mitochondrial MATα1). Data are presented as mean values ± SEM. *p < 0.05. Statistical significance was determined by using two-tailed, one-sample t-test. Source data are provided as a Source data file. Total lysate is shown in white, cytosolic fraction in blue and mitochondrial fraction in red. AH alcoholic hepatitis.

Fig. 2. Alcohol enhances PIN1 and MATα1 interaction in the liver.

a Human and mouse MATα1 protein sequences showing that PIN1 binding motif is conserved between species. b IP analysis of human hepatocytes lysate using anti-PIN1 antibody followed by western blot analysis against MATα1. c IP analysis of WT mouse liver hepatocytes lysate with anti-PIN1 (left panel) or anti-MATα1 (right panel) antibody followed by western blot analysis against MATα1 (left panel) or PIN1 (right panel). d IP analysis using PIN1 and MATα1 recombinant proteins and beads conjugated to anti-PIN1 and anti-MATα1 antibodies. IgG was used as a negative control. IP analysis of human normal and end-stage cirrhotic ALD liver lysates (e) (n = 3 normal and n = 5 ALD independent patients. p = 0.044), pair-fed and ethanol-fed mouse livers whole lysates (f, n = 4 pair-fed and n = 4 ethanol-fed independent animals. p = 0.002) and (g, n = 3 pair-fed and n = 3 ethanol-fed independent animals p = 0.027) and cytosolic fractions (h, n = 3 pair-fed and n = 3 ethanol-fed independent animals. p = 0.038), and AML-12 cells treated with ethanol (i and j, n = 3 independent experiments. p = 0.005) using anti-PIN1 or anti-MATα1 antibodies followed by western blot analysis against MATα1 or PIN1. Densitometry analysis are shown; ethanol-fed vs pair-fed; and ethanol vs control. Results are shown as mean ± SEM. *p < 0.05. Statistical significance was determined by using two-tailed, one-sample t-tests. Source data are provided as a Source data file. ALD alcohol-associated liver disease, IP immunoprecipitation, r recombinant.

Fig. 3. PIN1 negatively regulates MATα1 mitochondrial targeting.

Western blots of MATα1 and PIN1 in AML-12 and HepG2 cells after PIN1 silencing (a) and overexpression (b). Western blots and densitometry analyses of cytosolic and mitochondrial MATα1 and PIN1 in AML-12 and HepG2 cells after PIN1 silencing (c and e, AML-12: n = 4 independent experiments, p = 0.033 and HepG2: n = 3 independent experiments, p = 0.022) and overexpression (d and f, AML-12: n = 3 independent experiments, p = 0.003 and HepG2: n = 3 independent experiments, p = 0.009). AML-12 are shown in dark blue and HepG2 in light blue. g Western blot and h densitometry analyses of total and mitochondrial MATα1 in AML-12 cells after PIN1 silencing and ethanol treatment (n = 3 independent experiments, p = 0.006 for total and p = 0.005 for mitochondrial MATα1 SC ethanol vs SC control; p = 0.047 for total and p = 0.046 for mitochondrial MATα1 siPin1 ethanol vs SC ethanol). SC is shown in blue and siPin1 in red. i ATP levels in AML-12 cells after Pin1 silencing (n = 3 independent experiments, p = 0.034). j Triglycerides levels (n = 3 independent experiments, p = 0.004 SC ethanol vs control; p = 0.048 siPin1 control vs SC control; and p = 0.049 siPin1 ethanol vs SC ethanol) and k Oil red O staining in AML-12 after Pin1 silencing and ethanol treatment. SC is shown in blue and siPin1 in red. l Western blot of total and mitochondrial MATα1 in HepG2 cells after PIN1 WT or R68A/R69A catalytic mutant overexpression. *p < 0.05. Statistical significance was determined by using two-tailed, one-sample t-test for treatment comparisons and ANOVA test for group comparisons. Results are shown as mean ± SEM. Source data are provided as a Source data file. DDK DYKDDDDK-Tag, EV empty vector, Mut mutant, OE overexpression, SC scramble, si silencing.

Fig. 4. Alcohol-induced MATα1 phosphorylation at Ser114 is required for PIN1 binding.

a IP analysis of human normal and AH liver lysates, pair-fed and ethanol-fed mouse livers, and AML-12 cells treated with ethanol using anti-phosphoserine antibody followed by western blot analysis against MATα1 (n = 3 independent samples/experiments, p = 0.001 for AH vs normal; p = 0.021 ethanol-fed vs pair-fed; and p = 0.008 ethanol vs control AML-12). b Phos-Tag gels of phosphorylated MATα1 in cytosolic and mitochondrial fractions of AML-12 cells treated with ethanol. Unphosphorylated and phosphorylated recombinant human MATα1 protein is shown as a control (left). c Area under the curve of the phosphorylated peptide which corresponds to MATα1 Ser114 and its unmodified counterpart in WT mouse liver. d Quantitation of the MATα1 Ser114 in normal (n = 4) and AH (n = 5) human livers (p = 0.03). The center line, bounds of box, and whiskers represent mean, 25th to 75th percentile range, and minimum to maximum range. Normal liver is shown in blue and AH in red. e IP analysis of AML-12 and f HepG2 overexpressing MATα1 WT or S114A cell lysates using anti-His-Tag antibody followed by western blot analysis against PIN1. *p < 0.05. Statistical significance was determined by using two-tailed, one-sample t-test for treatment comparisons and ANOVA test for group comparisons. Results are shown as mean ± SEM. Source data are provided as a Source data file. AH alcoholic hepatitis, EV empty vector, IP immunoprecipitation, pSer phosphor-serine, WT wild type.

Fig. 5. Blocking the interaction with PIN1 protects MATα1 against alcohol-induced downregulation.

a MAT enzymatic activity measured as methionine consumption and SAMe production using recombinant MATα1 WT and S114A proteins. b MAT1A mRNA levels (n = 3 independent experiments, p = 0.007 for WT ethanol vs control and p = 0.0002 S114A ethanol vs control), c western blot, and d corresponding densitometry analysis of AML-12 cells lysates after transfection with EV, MATα1 WT or S114A vectors and ethanol treatment using anti-His-Tag antibody (n = 3 independent experiments, p = 0.003 for WT ethanol vs control and p = 0.005 S114A ethanol vs WT ethanol). e Western blot analysis of AML-12 cells expressing MATα1 WT or S114A and treated with ethanol before CHX addition and graph on the right showing the t_1/2 for MATα1 WT and S114A with and without ethanol treatment. f Western blot and g densitometry analysis of the mitochondrial fraction of AML-12 cells expressing MATα1 WT or S114A after ethanol treatment (n = 3 independent experiments, p = 0.003 for WT ethanol vs control and p = 0.016 S114A ethanol vs WT ethanol). h His-Tag (red) and TOM70 (green) immunofluorescence and mitochondrial. *p < 0.05. Statistical significance was determined by using two-tailed, one-sample t-test for treatment comparisons and ANOVA test for group comparisons. Results are shown as mean ± SEM. Source data are provided as a Source data file. WT is shown in blue and S114A is shown in red. CHX cycloheximide, EV empty vector, WT wild type.

Fig. 6. Blocking PIN1-MATα1 interaction protects against alcohol-induced mitochondrial injury by increasing MATα1 mitochondrial content.

a Mitotracker staining (red) and b quantification (n = 10 pictures/3 independent experiments, p = 0.002 for WT ethanol vs control; p = 0.047 for S114A ethanol vs control; p = 0.013 S114A ethanol vs WT ethanol), c mitochondrial membrane potential (n = 3 independent experiments, p = 0.045 for WT ethanol vs control; p = 0.022 S114A ethanol vs WT ethanol), d mitochondrial respiration and basal respiration (p = 0.015 for WT ethanol vs control; p = 0.049 S114A ethanol vs control; p = 0.035 WT ethanol vs S114A ethanol), ATP production (p = 0.022 for WT ethanol vs control; p = 0.041 S114A ethanol vs control; p = 0.041 WT ethanol vs S114A ethanol) and maximal respiratory capacity (p = 0.005 for WT ethanol vs control; p = 0.040 S114A ethanol vs control; p = 0.004 WT ethanol vs S114A ethanol from 3 independent experiments), e ATP (n = 3 independent experiments, p = 0.035 for WT ethanol vs control; p = 0.022 S114A ethanol vs control; p = 0.041 WT ethanol vs S114A ethanol), f mROS levels (n = 3 independent experiments, p = 0.013 for WT ethanol vs control; p = 0.04 WT ethanol vs S114A ethanol), g triglycerides content (n = 3 independent experiments, p = 0.017 for WT ethanol vs control; p = 0.008 WT ethanol vs S114A ethanol), and h Oil red O staining in AML-12 cells expressing MATα1 WT or S114A after ethanol treatment. *p < 0.05. Statistical significance was determined by using two-tailed, one-sample t-test for treatment comparisons and ANOVA test for group comparisons. Results are shown as mean ± SEM. Source data are provided as a Source data file. WT is shown in blue and S114A is shown in red. AA antimycin A, FCCP Carbonyl cyanide-p-trifluoromethoxyphenylhydrazone, mROS mitochondrial reactive oxygen species, OCR oxygen consumption rate, Oligom oligomycin, Rot rotenone, WT wild type.

Fig. 7. CK2 phosphorylates MATα1 at Ser114 to inhibit its mitochondrial localization in ALD.

Western blots and densitometry analyses of CK2α in a human normal (n = 5) and AH livers (n = 6) (p = 0.017), b pair-fed (n = 5) and ethanol-fed (n = 6) mouse livers (p = 0.026), and c AML-12 cells treated with ethanol (p = 0.005). d CK2 activity in AML-12 cells treated with ethanol (n = 5 independent experiments, p = 0.014). In vitro kinase assay of CK2 on e recombinant human MATα1, and f recombinant MATα1 WT and S114A proteins. g Western blot and h densitometry analyses of total and mitochondrial MATα1 in AML-12 cells after Csnk2a1 silencing and ethanol treatment (n = 3 independent experiments. For mitochondrial MATα1 p = 0.00002 SC ethanol vs control; p = 0.007 SC ethanol vs siCsnk2a1 ethanol. For total MATα1 p = 0.002 SC ethanol vs control; p = 0.007 SC ethanol vs siCsnk2a1 ethanol). i Mitochondrial respiration and basal respiration (p = 0.004 for SC ethanol vs control; p = 0.002 siCsnk2a1 ethanol vs control; p = 0.03 SC ethanol vs siCsnk2a1 ethanol), ATP production (p = 0.0003 for SC ethanol vs control; p = 0.001 siCsnk2a1 ethanol vs control; p = 0.002 SC ethanol vs siCsnk2a1 ethanol) and maximal respiratory capacity (p = 0.0009 SC ethanol vs control; p = 0.0002 siCsnk2a1 ethanol vs control; p = 0.003 SC ethanol vs siCsnk2a1 ethanol), j ATP levels (p = 0.047 WT ethanol vs control; p = 0.013 SC ethanol vs control; p = 0.041 SC ethanol vs siCsnk2a1 ethanol), k mROS levels (p = 0.0017 WT ethanol vs control; p = 0.013 SC ethanol vs control; p = 0.047 SC ethanol vs siCsnk2a1 ethanol), and l triglycerides content (n = 4 independent experiments, p = 0.006 WT ethanol vs control; p = 0.005 SC ethanol vs control; p = 0.041 SC ethanol vs siCsnk2a1 ethanol) in AML-12 Csnk2a1 silenced cells after ethanol treatment. n = 3 independent experiments unless specified. *p < 0.05. Statistical significance was determined by using two-tailed, one-sample t-test for treatment comparisons and ANOVA test for group comparisons. Results are shown as mean ± SEM. Source data are provided as a Source data file. SC is shown in blue and siCsnk2a1 is shown in red. AA antimycin A, FCCP Carbonyl cyanide-p-trifluoromethoxyphenylhydrazone, IP immunoprecipitation, mROS mitochondrial reactive oxygen species, OCR oxygen consumption rate, Oligom oligomycin, pSer phosphoserine, r recombinant, Rot rotenone, SC scramble, si silencing, WT wild type.

Fig. 8. Mitochondrial MATα1 protects mitochondrial proteome against ethanol-induced degradation.

a Overlap of mitochondrial and total proteins identified by MS in human livers after MATα1 co-IP and list of mitochondrial proteins involved in TCA cycle, OXPHOS, and fatty acid β-oxidation found to interact with MATα1 in human liver. b IP analysis of human normal liver lysates using anti-MATα1 antibody followed by western blot analysis against CPT1α, MCAD, and SDHα. c Western blot analysis using an OXPHOS antibody cocktail which detects ATP5A (Complex V), UQCRC2 (Complex III), SDHB (Complex II), and NDUFB8 (complex I) in lysates of AML-12 cells expressing MATα1 WT or S114A after ethanol treatment (n = 3 independent experiments). d Protein abundance heatmap for MitoPlex proteins detected in AML-12 cells expressing MATα1 WT or S114A after ethanol treatment. Expression is displayed as LogFC of ethanol versus control, ranging from downregulated (red) to upregulated (green) (n = 4 independent experiments). e Western blot analysis using a methyl-lysine antibody in mitochondrial fractions of AML-12 cells transfected with empty vector (EV), MATα1 WT or S114A after ethanol treatment (n = 3 independent experiments). Results are shown as mean ± SEM. Source data are provided as a Source data file. IP immunoprecipitation, LogFC Log2 fold change, MS mass spectrometry, OXPHOS oxidative phosphorylation, TCA tricarboxylic, WT wild type.

Fig. 9. PIN1 impairs MATα1 mitochondrial targeting iEn alcohol-associated liver disease.

Graphical summary of the study. Alcohol promotes CK2 phosphorylation of MATα1 at Ser114 and interaction with PIN1, thereby inhibiting MATα1 mitochondrial targeting. Alcohol-induced mitochondrial MATα1 depletion contributes to mitochondrial dysfunction and the pathogenesis of ALD.