Methionine Adenosyltransferase α1 Is Targeted to the Mitochondrial Matrix and Interacts with Cytochrome P450 2E1 to Lower Its Expression

Abstract

Methionine adenosyltransferase α1 (MATα1, encoded by MAT1A) is responsible for hepatic biosynthesis of S-adenosyl methionine, the principal methyl donor. MATα1 also act as a transcriptional cofactor by interacting and influencing the activity of several transcription factors. Mat1a knockout (KO) mice have increased levels of cytochrome P450 2E1 (CYP2E1), but the underlying mechanisms are unknown. The aims of the current study were to identify binding partners of MATα1 and elucidate how MATα1 regulates CYP2E1 expression. We identified binding partners of MATα1 by coimmunoprecipitation (co-IP) and mass spectrometry. Interacting proteins were confirmed using co-IP using recombinant proteins, liver lysates, and mitochondria. Alcoholic liver disease (ALD) samples were used to confirm relevance of our findings. We found that MATα1 negatively regulates CYP2E1 at mRNA and protein levels, with the latter being the dominant mechanism. MATα1 interacts with many proteins but with a predominance of mitochondrial proteins including CYP2E1. We found that MATα1 is present in the mitochondrial matrix of hepatocytes using immunogold electron microscopy. Mat1a KO hepatocytes had reduced mitochondrial membrane potential and higher mitochondrial reactive oxygen species, both of which were normalized when MAT1A was overexpressed. In addition, KO hepatocytes were sensitized to ethanol and tumor necrosis factor α-induced mitochondrial dysfunction. Interaction of MATα1 with CYP2E1 was direct, and this facilitated CYP2E1 methylation at R379, leading to its degradation through the proteasomal pathway. Mat1a KO livers have a reduced methylated/total CYP2E1 ratio. MATα1's influence on mitochondrial function is largely mediated by its effect on CYP2E1 expression. Patients with ALD have reduced MATα1 levels and a decrease in methylated/total CYP2E1 ratio. Conclusion: Our findings highlight a critical role of MATα1 in regulating mitochondrial function by suppressing CYP2E1 expression at multiple levels.

Figure 1. MATα1 is localized to the mitochondrial matrix in liver

A. Western blot analysis of MATα1 in whole mouse liver showing both cytosolic and mitochondrial fractions. Mitochondrial (COX IV) and cytosolic (Tubulin) specific markers were immunoblotted to demonstrate fraction purity (n=3). B. Subfractionation of mouse liver (n=3) treated with Proteinase K (PK), Digitonin (Dig) and/or TritonX-100 (TTX). For the different mitochondrial compartments VDAC (outer membrane), COX IV (inner membrane), Cytochrome C (intermembrane space and inner membrane), SDHA (matrix). C. Immunogold-EM of WT mouse liver (n=2) mitochondria showing presence of MATα1 in the mitochondrial matrix (left panel, distinct concentric circles) as compared to the absence of a specific signal in the no Ab negative control (right panel). Black arrows indicate MATα1. Images are 80,000X magnification.

Figure 2. MATα1 interacts with heatshock proteins and requires HSP70A8 to gain entry into the mitochondria

A. IP of WT mouse liver lysate (700 μg, n=3) with anti-MATα1 or anti-HSP70A8 Ab followed by Western blot analysis with HSP70A8 or MATα1. B. IP of WT mouse liver lysate (700 μg, n=3) with anti-MATα1 or anti-HSP90β Ab followed by Western blot analysis with HSP90β or MATα1. Blots were re-probed with Ab used for IP. Normal Ig was used as a negative control, 10 µg of lysate used in the IP were loaded as input control. C. Western blot analysis of cytosolic and mitochondria fraction of HepG2 cells (n=3) after silencing of HSPA8 (100 nM) for 72 hours. Tubulin was used a cytosolic marker, while COX IV was used as a mitochondrial marker. Actin was used as housekeeping control to normalize protein expression. D. Graph summarizes the densitometric values expressed as % of scramble control (SC) from three independent experiments, *P < 0.05 versus SC. E. As in C but after silencing of HSP90AB1 (30 nM). F. As in D.

Figure 3. Mat1a KO mice mitochondria have elevated mROS production and reduced membrane potential

A. Membrane potential and mitochondrial ROS (mROS) of isolated wild type (WT), Mat1a knockout (KO) hepatocytes and Mat1a KO hepatocytes overexpressing MAT1A (48 hours, N=3). *P < 0.05 versus WT, ^tP < 0.05 versus KO. B. Membrane potential and mROS production were measured in WT mouse hepatocytes where MAT1A was either over-expressed (48 hours) or knocked down (66 hours, 50 nM) (n=3). *P < 0.05 versus empty vector (EV), ^tP < 0.05 versus scramble control (SC). C. Membrane potential and mROS levels in HepG2 cells with MAT1A knock down (66 hours, 25 nM) (n=3). *P < 0.05 versus SC. D. Mitochondrial respiration in isolated liver mitochondria from 7 month old female WT and Mat1a KO mice. Values are represented as mean ± SEM (n=6 replicates from three independent experiments). *P < 0.05 Mat1a KO versus WT. E. Oxygen consumption rate (OCR) measured in HepG2 cells with MAT1A knockdown. Oligomycin (Oligo) inhibits ATP synthesis, FCCP uncouples oxidative phosphorylation, rotenone (Rot) inhibits mitochondrial complex I and antimycin A (AA) inhibits complex III in the respiratory chain. OCR profile was normalized to protein level. Basal and ATP-linked OCR, maximal, and reserve respiratory capacity were calculated. Values are represented as mean ± SEM (n=16 replicates from four independent experiments). *P < 0.05 versus SiMAT1A.

Figure 4. MATα1 and CYP2E1 directly interact

In vitro pull-down assay using A. immobilized recombinant MATα1 with recombinant CYP2E1 or B. immobilized recombinant CYP2E1 with recombinant MATα1 and IB for MATα1 and CYP2E1 (n=3). C. IP of total mouse liver lysate (700 μg, n=3) with anti-CYP2E1 or anti-MATα1 Ab followed by Immunoblot (IB) analysis for MATα1 or CYP2E1. D. IP of isolated WT mouse liver mitochondria (300 μg, n=3) with anti-CYP2E1 or anti-MATα1 Ab followed by IB analysis for MATα1 or CYP2E1. Normal Ig was used as a negative control and 10 μg of lysate was used as input control. Blots were re-probed with Ab used for IP.

Figure 5. CYP2E1 expression is negatively regulated by MATα1 and MAT1A knockdown-mediated mitochondrial dysfunction is dependent on CYP2E1

A. Relative mRNA expression of 3-month old male Mat1a and Cyp2e1 in WT and Mat1a KO liver. Results represent a total of at least three independent experiments done in duplicate. *P < 0.05 versus WT. B. Protein expression of CYP2E1 in both WT and Mat1a KO mouse livers, *P < 0.05 versus WT (n=3). C. Expression of CYP2E1 in isolated mouse liver mitochondria in both WT and Mat1a KO. *P < 0.05 versus WT (n=5). Succinate dehydrogenase (SDHA) and COXIV were used as mitochondrial specific markers, and tubulin was used as a cytosolic marker to demonstrate purity. SDHA was used as a housekeeping control to normalize protein expression. D. Mitochondrial CYP2E1 protein quantification of WT and Mat1a KO mouse liver mitochondria. Peptide to protein rollup was done using MSSTATS. Error bars represent standard deviation from biological replicates. *P < 0.05 versus WT. E. Western blot analysis of CYP2E1 and MATα1 in WT mouse hepatocytes, when either Mat1a (66 hours, 50 nM) and/or Cyp2e1 (66 hours, 30 nM) were silenced (n=3). GAPDH was used as housekeeping control to normalize protein expression. *P < 0.05 versus double scramble control (SC+SC). F. Membrane potential and mitochondrial ROS (mROS) levels from experiment described in E. *P < 0.05 versus double scramble control (SC+SC) (n=3).

Figure 6. CYP2E1 is a methylated protein and level of me-CYP2E1 is reduced in Mat1a KO liver mitochondria

A. IP of 5-month old male mouse liver mitochondria lysate from WT or Mat1a KO mice (n=3) with anti-mono/di methyl arginine Ab followed by immunoblot (IB) analysis for CYP2E1. Amount of methylated CYP2E1 was determined first by normalizing to Ig for each IP and then comparing that value to total levels of CYP2E1 that had been normalized to SDHA. *P < 0.05 versus WT. B. Intensity of the extracted precursor isotopic envelope (M, M+1, M+2) of a representative CYP2E1 peptide in WT (left panel) and Mat1a KO (right panel). C. CYP2E1 protein quantification from a mitochondrial methyl-immunoprecipitation using WT and Mat1a KO 5-month old male mouse liver mitochondria. Peptide to protein rollup was done using MSSTATS. Error bars represent standard error from biological replicates (n=4). D. Western blot analysis of HepG2 cells that were over-expressed with wither WT CYP2E1, R100N CYP2E1, R379N CYP2E1, or R100N/R379N CYP2E1 mutants for 48 hours in the presence or absence of MAT1A (48 hours) (n=3). GAPDH was used as a housekeeping control to normalize protein expression , *P < 0.05 versus specific CYP2E1+EV control.

Figure 7. Overexpression of MAT1A enhanced CYP2E1 degradation via the proteasomal pathway

A. Western blot analysis of HepG2 cells that were co-overexpressed with CYP2E1 and MAT1A for 48 hours before CHX addition for 5 hours in serum free media (n=3). GAPDH was used as housekeeping control to normalize protein expression. *P < 0.05 versus CYP2E1+empty vector (EV) control B. Graph shows the t_½ for CYP2E1 as compared to EV at time 0. All densitometric values were normalized to GADPH and expressed as % of CYP2E1+EV at time 0, average of 3 independent experiments are shown. Regression analysis using the best fit line was done to calculate the half-lives of CYP2E1 with or without MAT1A over-expression. *P < 0.05 versus CYP2E1+ EV. C. Western blot analysis of HepG2 cells that were over-expressed with or without MAT1A in addition to CYP2E1 for 42 hours before the addition of MG132 for 6 hours (n=3). GAPDH was used as housekeeping control to normalize protein expression. *P < 0.05 versus CYP2E1 control. D. Western blot analysis of MATα1 using whole mouse liver showing isolated microsomal fraction. Endoplasmic recticulum marker (Calnexin) and cytosolic marker (Tubulin) were immunoblotted to demonstrate fraction purity (n=3). E. IP of isolated WT mouse liver microsome (300 μg, n=3) with anti-CYP2E1 Ab followed by Western blot analysis for MATα1 or F. IP with anti-MATα1 Ab followed by Western blot analysis for CYP2E1. Normal Ig was used as a negative control and 10 μg of lystate was used as input control. Blots were re-probed with Ab used for IP.

Figure 8. MATα1 interacts with CYP2E1 in human liver and methylated CYP2E1 is reduced in human ALD

A. Western blot analysis of MATα1 in whole human liver (n=3) showing both cytosolic and mitochondrial fractions. Mitochondrial (SDHA) and cytosolic (Tubulin) specific markers were immunoblotted to demonstrate fraction purity. B. IP of total human liver lysate (700 μg, n=3) with anti-MATα1 or anti-CYP2E1 Ab followed by IB analysis for either CYP2E1 or MATα1. Blots were re-probed with Ab used for IP. Normal Ig was used as a negative control and 10 µg of lysate used as input control. C. Control (n=4) and ALD (n=5) total liver lysate (lys) IB for MATα1 and densitometric values were normalized to GADPH and expressed as % over control. D. IP with anti-mono/di methyl arginine Ab followed by IB analysis for CYP2E1. Amount of methylated CYP2E1 was determined first by normalizing to Ig for each IP and then compared to total levels of CYP2E1 that had been normalized to actin. *P < 0.05 versus control. E. IP of 3.5-month old female mouse liver lysate from pair-fed or NIAAA mice (n=5 and 6 respectively). IP of total mouse liver lystate (700μg) with anti-mono/di methyl arginine Ab followed by IB analysis for CYP2E1. Amount of methylated CYP2E1 was determined first by normalizing to Ig for each IP and then compared to total levels of CYP2E1 that had been normalized to GAPDH. *P < 0.05 versus control.