Inactivation of oxidized and S-nitrosylated mitochondrial proteins in alcoholic fatty liver of rats


  • Kwan-Hoon Moon,

    1. Laboratory of Membrane Biochemistry and Biophysics, National Institute on Alcohol Abuse and Alcoholism, Bethesda, MD
    Search for more papers by this author
    • Kwan-Hoon Moon and Brian L. Hood equally contributed to this work.

  • Brian L. Hood,

    1. Laboratory of Proteomics and Analytical Technologies, SAIC-Frederick, Inc., Frederick, MD
    Search for more papers by this author
    • Kwan-Hoon Moon and Brian L. Hood equally contributed to this work.

  • Bong-Jo Kim,

    1. Laboratory of Membrane Biochemistry and Biophysics, National Institute on Alcohol Abuse and Alcoholism, Bethesda, MD
    Search for more papers by this author
  • James P. Hardwick,

    1. Department of Microbiology, Immunology, and Biochemistry, Northeastern Ohio University College of Medicine, Rootstown, OH
    Search for more papers by this author
  • Thomas P. Conrads,

    1. Laboratory of Proteomics and Analytical Technologies, SAIC-Frederick, Inc., Frederick, MD
    Search for more papers by this author
  • Timothy D. Veenstra,

    1. Laboratory of Proteomics and Analytical Technologies, SAIC-Frederick, Inc., Frederick, MD
    Search for more papers by this author
  • Byoung J. Song

    Corresponding author
    1. Laboratory of Membrane Biochemistry and Biophysics, National Institute on Alcohol Abuse and Alcoholism, Bethesda, MD
    • Laboratory of Membrane Biochemistry and Biophysics, National Institute on Alcohol Abuse and Alcoholism, 9000 Rockville Pike, Bethesda, MD 20892-9410
    Search for more papers by this author
    • fax: 301-594-3113

  • Potential conflict of interest: Nothing to report.


Increased oxidative/nitrosative stress is a major contributing factor to alcohol-mediated mitochondrial dysfunction. However, which mitochondrial proteins are oxidatively modified under alcohol-induced oxidative/nitrosative stress is poorly understood. The aim of this study was to systematically investigate oxidized and/or S-nitrosylated mitochondrial proteins and to use a biotin-N-maleimide probe to evaluate their inactivation in alcoholic fatty livers of rats. Binge or chronic alcohol exposure significantly elevated nitric oxide, inducible nitric oxide synthase, and ethanol-inducible CYP2E1. The biotin-N-maleimide-labeled oxidized and/or S-nitrosylated mitochondrial proteins from pair-fed controls or alcohol-fed rat livers were subsequently purified with streptavidin-agarose. The overall patterns of oxidized and/or S-nitrosylated proteins resolved by 2-dimensional polyacrylamide gel electrophoresis were very similar in the chronic and binge alcohol treatment groups. Seventy-nine proteins that displayed differential spot intensities from those of control rats were identified by mass spectrometry. These include mitochondrial aldehyde dehydrogenase 2 (ALDH2), ATP synthase, acyl-CoA dehydrogenase, 3-ketoacyl-CoA thiolase, and many proteins involved in chaperone activity, mitochondrial electron transfer, and ion transport. The activity of 3-ketoacyl-CoA thiolase involved in mitochondrial β-oxidation of fatty acids was significantly inhibited in alcohol-exposed rat livers, consistent with hepatic fat accumulation, as determined by biochemical and histological analyses. Measurement of activity and immunoblot results showed that ALDH2 and ATP synthase were also inhibited through oxidative modification of their cysteine or tyrosine residues in alcoholic fatty livers of rats. In conclusion, our results help to explain the underlying mechanism for mitochondrial dysfunction and increased susceptibility to alcohol-mediated liver damage. (HEPATOLOGY 2006;44:1218–1230.)

Excessive alcohol intake for short (binge exposure) or extended (chronic exposure) durations can cause cell/organ damage, partly through enhanced production of reactive oxygen and nitrogen species (ROS/RNS) and mitochondrial dysfunction,1–3 the severity of which can be alleviated by simultaneous treatment with various antioxidants, especially in the early stages.4 The molecular mechanisms contributing to alcohol-mediated tissue damage have been investigated using several animal models including genetically selected modified mouse5, 6 and cultured cell models such as ethanol-sensitive E47 HepG2 human hepatoma cells transduced with ethanol-inducible cytochrome P450 2E1 (CYP2E1).7 Chronic alcohol exposure is known to induce CYP2E1 and to activate other enzymes such as xanthine oxidase and NADPH oxidase,1–3 both of which can produce ROS. Alcohol is also known to directly suppress mitochondrial function, especially at the levels of complexes I and III,8 leading to greater amounts of ROS generated through aberrant action of the electron transport chain.9 ROS can be produced during CYP2E1-mediated metabolism of the numerous potentially toxic substrates10–12 as well as in the absence of its substrates.13 Furthermore, alcohol elevates expression of the inducible isoform of nitric oxide synthase (iNOS) in the liver and other tissues.14, 15 Nitric oxide, produced through induction of iNOS, may contribute to increased nitrosative stress through production of the toxic peroxynitrite anion (ONOO) in the presence of superoxide.16 Peroxynitrite, an extremely powerful oxidizing agent, readily reacts with DNA and proteins and is known to inhibit their cellular functions, contributing to increased sensitivity to cell damage.16 In fact, many cellular proteins have been shown to be modified by peroxynitrite, leading to often deleterious alterations in their physiological functions.17, 18 Chronic alcohol exposure reduces the levels of oxidation scavengers such as glutathione and other antioxidants1–3 and the activity of certain protective enzymes such as isocitrate dehydrogenase14 and peroxiredoxin.19 Our recent study showed that the activity of mitochondrial aldehyde dehydrogenase 2 (ALDH2) was reversibly inhibited through S-nitrosylation of its cysteine residues in the presence of various nitric oxide (NO) donors.20 Alcohol-related increases of nitrite and iNOS14, 15 may lead to inhibition of ALDH2 and other mitochondrial proteins through potential oxidation and/or S-nitrosylation, although this possibility has not been systematically studied. Despite the well-established effect of alcohol on ROS/RNS production and mitochondrial dysfunction, it is poorly understood which proteins are S-nitrosylated or oxidized and whether their biological functions are altered. To address these problems, we used recently developed methods19, 21 to identify oxidized mitochondrial proteins and compared them with S-nitrosylated mitochondrial proteins in alcohol-exposed rat livers, using sodium ascorbate (Asc) to selectively convert S-nitrosylated cysteines to free cysteine residues.22, 23 In addition, we investigated the potential mechanisms of inhibition of select oxidatively modified mitochondrial proteins that might account for alcohol-induced mitochondrial dysfunction.


CYP2E1, ethanol-inducible cytochrome P450 2E1; iNOS, inducible nitric oxide synthase; ALDH2, mitochondrial aldehyde dehydrogenase 2; Asc, sodium ascorbate; biotin-NM, biotin-N-maleimide; DTT, dithiothreitol; 3-NT, 3-nitrotyrosine; β-ATP synthase, ATP synthase β subunit.

Materials and Methods


Biotin-conjugated N-maleimide (biotin-NM), anti-S-nitrosocysteine (S-NO-Cys) antibody, the specific antibody to ATP synthase β subunit; (β-ATP synthase), propionyl aldehyde, pyrazole, dithiothreitol (DTT), sodium dithionite, and L-sodium ascorbate were obtained from Sigma Chemical (St. Louis, MO) in the highest purity. The specific antibodies to iNOS and 3-nitrotyrosine (3-NT) were purchased from Santa Cruz Biotechnologies and Upstate Biotechnologies (Waltham, MA), respectively. The specific antibodies to ATP synthase α subunit (α-ATP synthase) and prohibitin were purchased from Invitrogen (Carlsbad, CA) and Oncogene Science (Cambridge, MA), respectively.

Animal Maintenance and Pair-Feeding Treatment.

Young male Sprague-Dawley rats (n ≥ 10; from Taconic Farms, Rockville, MD), maintained in accordance with National Institutes of Health guidelines, were acutely treated with alcohol for 4 consecutive days (the binge treatment), as previously described,4 or with an isocaloric dextrose-containing diet (the dextrose control). Another group of young adult rats (n = 6 per group) was individually housed and fed a Lieber-DeCarli alcohol liquid diet (with 35% of the daily calories derived from ethanol) or an isocaloric dextrose control diet for 4 weeks, as previously described.19

Identification of S-Nitrosylated and Oxidized Proteins by Mass Spectrometry.

Mitochondrial fractions were prepared from pooled rat livers (n ≥ 6 per group) freshly obtained from the differently treated groups by a previously described method.20, 21 The oxidized proteins were labeled with biotin-NM according to a method (Fig. 1) previously detailed,19, 21 with Asc (10 mM final concentration) substituted for DTT in order to identify S-nitrosylated proteins.22, 23 Purified biotin-NM-labeled S-nitrosylated and oxidized proteins bound to the streptavidin-agarose beads were washed twice before they were resolved by 2-dimensional polyacrylamide gel electrophoresis (2-D PAGE), stained with silver, scanned, and analyzed. In-gel digestion of protein gel spots, nanoflow reversed-phase liquid chromatography (nano-RPLC)–tandem mass spectrometry (MS/MS), and bioinformatic analyses were performed as recently described.19, 21

Figure 1.

Schematic diagram identifying oxidatively modified Cys residues. Oxidized and/or S-nitrosylated Cys residues of target proteins in control or alcohol-exposed animals were labeled with NM-biotin and detected or purified, as detailed previously.19, 21

Immunoaffinity Purification or Immunoprecipitation and Immunoblot Analyses.

Immunoaffinity purification of ALDH2 proteins from alcohol- or pair-fed rats was performed as previously described.20 A separate aliquot of mitochondrial proteins was incubated with 5 μg of the specific antibody to α- or β-ATP synthase for 2 hours with constant agitation, followed by the addition of protein G–agarose for an additional hour. Proteins bound to the protein G–agarose were washed 3 times with 1× phosphate-buffered saline with 1% CHAPS in order to remove nonspecific protein interactions. After centrifugation, bound proteins were dissolved in Laemmli buffer for immunoblot analysis using the specific antibodies against each target protein.

Determination of Nitrite, Triglyceride, and Cholesterol Concentrations.

Total nitrite concentration was determined as recently described.20 The supernatant fraction of liver homogenates (10 μL/sample) was used to determine the levels of triglycerides or cholesterol using a QVET® kit according to the manufacturer's protocol (DREW Scientific, Oxford, CT) and normalized for the protein concentration. For analysis of fat accumulation, small pieces of liver from differently treated rats were fixed in 10% formalin, and 10-μm-thick frozen sections prepared using a cryostat were stained with Oil Red O (Vector Laboratories), counterstained with Mayer's hematoxylin, and analyzed by light microscopy.24

Activity Measurements of Various Mitochondrial Enzymes.

ALDH2 activity was measured by increased production of NADH at 340 nm, as described.20 The NOS activity was measured using 0.4 mg of protein and the fluorescence indicator 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate with excitation at 490 nm and emission at 510 nm.25 The activity of 3-ketoacyl-CoA thiolase was determined by the absorbance change at 303 nm following disappearance of the Mg2+-enolate complex of acetoacetyl-CoA (ε303= 16.5 mM−1 cm−1) as previously described.26 ATP synthase activity was determined using an ATP bioluminescence assay kit (Sigma) according to the manufacturer's protocol. One unit of ATP synthesis activity represents 1 nM ATP produced per hour per milligram of protein at ambient temperature.

Data Processing and Statistical Analysis.

All data in this report represent results from at least 3 separate experiments, unless stated otherwise. Statistical analyses were performed using the Student t test, and differences were considered statistically significant at P < .05. Other materials and methods not described here were performed as previously described.19–21


Increased Nitrite Concentration, NOS Activity, CYP2E1, and iNOS in Alcohol-Treated Rat Livers.

To demonstrate increased nitrosative stress in alcohol-exposed rat livers, we determined the nitrite and NOS activity levels. Nitrite levels in the mitochondria of rat livers were significantly elevated, from approximately 1.1 ± 0.05 μM to 2.8 ± 0.63 and 5.2 ± 0.25 μM, after the chronic and binge ethanol treatments, respectively (Fig. 2A). In parallel, the activity of NOS was increased approximately 2.6- and 5.0-fold in chronic- and binge-alcohol-treated rat livers compared to that in the corresponding pair-fed and dextrose-treated controls, respectively (Fig. 2B). Microsomal CYP2E1 (Fig. 2C, top panel), often recognized as a surrogate marker for increased oxidative stress, was greater in alcohol-exposed rat livers than in the corresponding controls. The level of mitochondrial iNOS (middle panel), a source of increased nitrosative stress, was very low in the dextrose-treated control animals but significantly elevated in rat livers following the binge ethanol and chronic ethanol treatments. In contrast, a similar amount of mitochondrial prohibitin, used as a loading control, was found in all samples examined (bottom panel).

Figure 2.

Increased mitochondrial nitrite concentration, NOS activity, CYP2E1, and iNOS in chronic- and binge-alcohol-exposed rat livers compared to dextrose-fed control rat livers. (A) Mitochondrial protein (0.4 mg/assay) was used to determine the nitrite concentration. Each point represents the average ± SD of 3 determinations. (B) Mitochondrial protein (0.4 mg/assay) was used to determine NOS activity. Each point represents the average ± SD of 3 determinations. (C) Equal amounts of mitochondrial protein (10 μg/well) from binge- or chronically ethanol-fed rat livers were separated on 12% SDS-PAGE, transferred to PVDF-immobilon membranes, and subjected to immunoblot analysis using anti-CYP2E1, iNOS, and prohibitin antibody. A typical result from 2 separate experiments is shown (*significantly different from the control samples at P < .05; **significantly different from the controls at P < .001; and ***significantly different from the controls at P < .01).

Increased Levels of Oxidatively Modified Protein in Alcohol-Fed Rats.

Because of the elevated levels of nitrite, NOS activity, and CYP2E1, we hypothesized that more proteins were S-nitrosylated and/or oxidized in the livers of alcohol-exposed rats than in the untreated controls. Figure 3A (left panel) shows a pattern of Coomassie blue staining that indicates similar levels of proteins were analyzed for all samples. Nitrosylated protein levels (Asc treatment) were markedly higher in alcohol-exposed rats than in their pair-fed counterparts (right panel) when their relative levels were determined by immunoblot analyses using the monoclonal antibody (MAb)-biotin-horseradish peroxidase (HRP) as a detection probe. Similar results showing increased numbers of oxidized and/or S-nitrosylated proteins in alcohol-exposed rats were also observed with the streptavidin-HRP probe (Fig. 3B, right panel).

Figure 3.

Increased levels of oxidized and/or S-nitrosylated mitochondrial protein in chronically alcohol-fed rats compared to dextrose-fed control rats. Typical results from 2 separate experiments are shown. Oxidatively modified mitochondrial proteins in the control or ethanol-exposed rat liver mitochondria were labeled with biotin-NM in the presence of DTT or Asc, as indicated, and purified with streptavidin-agarose beads. Purified biotin-NM-labeled mitochondrial proteins (20 μg/well) were then separated on 12% SDS-PAGE, transferred to PVDF-Immobilon membranes, and stained with Coomassie blue (A and B, left panels) or subjected to immunoblot analysis using MAb-biotin-HRP (A, right panel) or streptavidin-HRP (B, right panel) antibodies..

Summary of Protein Sequencing Analyses of Oxidized and/or S-Nitrosylated Mitochondrial Proteins.

Because of elevated levels of oxidized and/or S-nitrosylated mitochondrial proteins in alcohol-treated rats, we sought to identify each oxidatively modified protein through labeling with biotin-NM.19, 21 Only a few biotin-NM-labeled oxidized and S-nitrosylated proteins were detected in the pair-fed control rat livers (Supplementary Fig. 1A; Supplementary material available at: http://, whereas the numbers and intensities of oxidized (Supplementary Fig. 1B) and S-nitrosylated proteins (Supplementary Fig. 1C) were markedly increased in the chronically alcohol-fed groups. Similar results were also observed in rats that received the binge alcohol treatment: exposure to a high dose of ethanol for 4 consecutive days. For comparison, the intensity of the darkest spot (spot 14, designated with a square) in the dextrose-treated control samples (Fig. 4A) was matched with those of the same protein in the alcohol-exposed samples. The numbers and spot intensities of oxidized (Fig. 4B) and/or S-nitrosylated mitochondrial proteins (Fig. 4C) were markedly increased after binge alcohol treatment compared to those in dextrose-treated controls. Seventy-nine proteins with increased spot intensity in rat liver mitochondria after binge alcohol treatment (Fig. 4) were identified by MS analysis; the data are summarized in Table 1. Protein sequence analysis revealed that many mitochondrial proteins were concurrently oxidized and S-nitrosylated after alcohol exposure. These proteins include metabolic enzymes (ALDH2, α- and β-ATP synthase, glutamate dehydrogenase, methylmalonate-semialdehyde dehydrogenase, short-chain acyl-CoA dehydrogenase, 3-hydroxyacyl-CoA dehydrogenase, etc), chaperone proteins (Hsp60, glucose-regulated proteins [GRP78, GRP75]), and mitochondrial electron-transport and ion-transport proteins including electron transfer flavoprotein α and β subunits, voltage-dependent anion channels, and prohibitin. In contrast, certain proteins such as 3-ketoacyl-CoA-thiolase (spots 15-24) were found to be only oxidized, whereas malate dehydrogenase (MDH; spots 55-67) was only S-nitrosylated (as detected in DTT- and Asc-treated samples, respectively). Interestingly, several spots (13, 14, 40, and 50) were identified as β subunits of ATP synthase, which does not contain Cys residues that can be modified under the influence of alcohol. Therefore, it is likely that β-ATP synthase must coresolve with the oxidatively modified α subunit.

Figure 4.

Comparison of S-nitrosylation and oxidized mitochondrial proteins by 2-D PAGE in binge-alcohol-treated rat livers. Oxidized mitochondrial proteins (10 mg/sample) from (A) dextrose-treated control and (B and C) binge alcohol-treated rat livers were labeled with biotin-NM in the presence of (B) DTT or (C) Asc and purified with streptavidin-agarose. Purified biotin-NM-labeled proteins (0.25 mg/sample) were resolved by 2-D PAGE, and stained with silver. Individual protein spots (spots 1-79) with differential intensities were marked with different numbers, excised from that particular gel (pH range 3-10), and subjected to MS analysis following in-gel trypsin digestion. Spot 14, designated with a square, was used as an internal standard for a comparison of the 3 gels for dextrose-fed control and binge-ethanol-treated rat livers.

Table 1. Summary of LC-MS/MS Peptide Sequence Analyses of Oxidized and/or S-Nitrosylated Proteins in Alcohol-Treated Rat Liver Mitochondria
Spot No.Protein IdentifiedAccession No.No. of Peptides IdentifiedModif
  1. NOTE. Biotin-NM-labeled oxidized and/or S-nitrosylated mitochondrial proteins were isolated with streptavidin-agarose, washed, resolved on 2-D PAGE, and stained with silver. Each spot as indicated was picked up with a razor blade, and subjected to mass spectrometric analysis with rat protein database, as described in the Materials and Methods section. Certain proteins might have been missed because they were in very low abundance in the 2-D PAGE and because of contamination by other proteins present in large quantities. Despite the small number of peptides identified for some mitochondrial proteins, we are confident the peptides identified by MS analysis represent the peptide sequences of listed proteins (O, oxidized proteins; N, S-nitrosylated proteins; O/N, both oxidized and S-nitrosylated proteins).

1NADH-dehydrogenase (ubiquinone) (75 kDa)Q66HF111O
2Stress-70 protein, mitochondrial (GRP75)P4872134O/N
 78-kDa Glucose-regulated protein (GRP78)P0676129 
 Succinate dehydrogenaseQ920L22 
360 kDa Heat shock protein (Hsp60)P6303967O/N
 Stress-70 protein, mitochondrial (GRP75)P4872110 
 Protein disulfide isomerase A3 precursorP115985 
 78-kDa Glucose-regulated protein (GRP78)P067613 
 Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P118842 
4Protein disulfide isomerase A3 precursorP1159814O/N
 Methylmalonate-semialdehyde dehydrogenaseQ0225310 
 60-kDa Heat shock protein (Hsp60)P630394 
5Sterol carrier protein 2, mitochondrialP119154O/N
6Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P1188411O/N
 ATP synthase α chain, mitochondrialQ6P7533 
 60-kDa Heat shock protein (Hsp60)P630392 
 Retinal dehydrogenase 1 (RALDH1)P516472 
7Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P1188412O/N
 ATP synthase α chain, mitochondrialQ6P7535 
 60-kDa Heat shock protein (Hsp60)P630394 
8Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P1188416O/N
 ATP synthase α chain, mitochondrialQ6P7538 
 Retinal dehydrogenase 1 (RALDH1)P516472 
9ATP synthase α chain, mitochondrialQ6P75313O/N
 Methylmalonate-semialdehyde dehydrogenaseQ0225313 
 Glutamate dehydrogenase 1, mitochondrial (GDH)P108604 
10ATP synthase α chain, mitochondrialQ6P75316O/N
 Methylmalonate-semialdehyde dehydrogenaseQ0225314 
11ATP synthase α chain, mitochondrialQ6P75333O/N
 Glutamate dehydrogenase 1, mitochondrial (GDH)P108606 
 Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P118844 
 Methylmalonate-semialdehyde dehydrogenaseQ022532 
12ATP synthase α chain, mitochondrialQ6P75325O/N
 Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P118842 
 Glutamate dehydrogenase 1, mitochondrial (GDH)P108602 
13ATP synthase β chain, mitochondrialP1071954O/N
 ATP synthase α chain, mitochondrialQ6P7534 
 Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P118844 
 Protein disulfide isomerase A6 precursorQ630813 
14ATP synthase β chain, mitochondrialP107199O/N
 60-kDa heat shock protein, mitochondrial (Hsp60)P630392 
153-Ketoacyl-CoA thiolase, mitochondrialP134379O
163-Ketoacyl-CoA thiolase, mitochondrialP1343712O
173-Ketoacyl-CoA thiolase, mitochondrialP134379O
183-Ketoacyl-CoA thiolase, mitochondrialP1343711O
193-Ketoacyl-CoA thiolase, mitochondrialP1343719O
203-Ketoacyl-CoA thiolase, mitochondrialP1343715O
 Acetoacetyl-CoA thiolase, mitochondrialP177642 
21Acetoacetyl-CoA thiolase, mitochondrialP177647O
 3-Ketoacyl-CoA thiolase, mitochondrialP134376 
 Acyl-CoA dehydrogenase, medium-chain, mitochondrialP085032 
223-Ketoacyl-CoA thiolase, mitochondrialP1343718O
 Acetoacetyl-CoA thiolase, mitochondrialP177647 
 Acyl-CoA dehydrogenase, medium-chain, mitochondrialP085033 
233-ketoacyl-CoA thiolase, mitochondrialP134378O
 Acetyl-CoA thiolase, mitochondrialP177644 
24Acetyl-CoA thiolase, mitochondrialP177649O
 3-Ketoacyl-CoA thiolase, mitochondrialP134375 
2560-kDa Heat shock protein, mitochondrial (Hsp60)P630392O/N
26Electron-transfer flavoprotein α subunit (α-ETF)P138035O/N
 3-Hydroxyisobutyrate dehydrogenase, mitochondrialP292662 
273-Hydroxyisobutyrate dehydrogenase, mitochondrialP292664O/N
28Electron-transfer flavoprotein α subunit (α-ETF)P138037O/N
 3-Hydroxyisobutyrate dehydrogenase, mitochondrialP292664 
293-Hydroxyisobutyrate dehydrogenase, mitochondrialP292663O/N
30Electron-transfer flavoprotein α subunit (α-ETF)P138038O/N
 3-Hydroxyisobutyrate dehydrogenase, mitochondrialP292666 
31Not identified   
32Electron-transfer flavoprotein α subunit (α-ETF)P138038O/N
33Electron-transfer flavoprotein α subunit (α-ETF)P1380312O/N
 Voltage-dependent anion-selective channel protein 1Q9Z2L06 
 Short-chain 3-hydroxyacyl-CoA dehydrogenase, mitoQ9WVK73 
 2,4-Dienoyl-CoA reductase, mitochondrialQ645912 
34Electron-transfer flavoprotein α subunit (α-ETF)P1380315O/N
 Voltage-dependent anion-selective channel protein 1Q9Z2L04 
 Short chain 3-hydroxyacyl-CoA dehydrogenase, mitoQ9WVK74 
35Voltage-dependent anion-selective channel protein 1Q9Z2L013O/N
 ATP synthase α chain, mitochondrialQ6P7536 
 Electron-transfer flavoprotein α subunit (α-ETF)P138036 
 Short chain 3-hydroxyacyl-CoA dehydrogenase, mitoQ9WVK75 
 Malate dehydrogenase, mitochondrialP046363 
 3-Ketoacyl-CoA thiolase, mitochondrialP134372 
36Short chain 3-hydroxyacyl-CoA dehydrogenase, mitoQ9WVK74O/N
37Voltage-dependent anion-selective channel protein 1Q9Z2L09O/N
 ATP synthase α chain, mitochondrialQ6P7533 
3860-kDa Heat shock protein, mitochondrial (Hsp60)P630392O/N
 Electron-transfer flavoprotein β subunit (β-ETF)Q68FU38 
 Enoyl-CoA hydratase, mitochondrialP146043 
40NADH-dehydrogenase (Ubiquinone) flavoprotein 2Q6PDU96O
 ATP synthase β chain, mitochondrialP107194 
41Electron-transfer flavoprotein β subunit (β-ETF)Q68FU39O/N
42Electron-transfer flavoprotein β subunit (β-ETF)Q68FU33O/N
433-Hydroxyacyl-CoA dehydrogenase type IIO703518O/N
443-Hydroxyacyl-CoA dehydrogenase type IIO7035110O/N
45ATP synthase D chain, mitochondrialP313999O
 ATP synthase α chain, mitochondrialQ6P7532 
46ATP synthase D chain, mitochondrialP3139914O
47ATP synthase D chain, mitochondrialP3139916O
 Casein kinase I, α-like isoformP976332 
48Electron-transfer flavoprotein β subunit (β-ETF)Q68FU37O/N
49Not identified   
50Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P1188413O/N
 ATP synthase β chain, mitochondrialP107196 
 ATP synthase α chain, mitochondrialQ6P7534 
 60-kDa Heat shock protein, mitochondrial(Hsp60)P630393 
 78-kDa Glucose-regulated protein precursor (GRP 78)P067612 
51Glutamate dehydrogenase 1, mitochondrial (GDH)P108605N
 Methylmalonate-semialdehyde dehydrogenaseQ022532 
52Methylmalonate-semialdehyde dehydrogenaseQ022533N
 Glutamate dehydrogenase 1, mitochondrial (GDH)P108602 
53Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P1188412O/N
 ATP synthase α chain, mitochondrialQ6P7538 
 Glutamate dehydrogenase 1, mitochondrial (GDH)P108605 
54ATP synthase α chain, mitochondrialQ6P75314O/N
 Glutamate dehydrogenase 1, mitochondrial (GDH)P1086011 
 Aldehyde dehydrogenase 2, mitochondrial (ALDH2)P1188410 
 Methylmalonate-semialdehyde dehydrogenaseQ022532 
55Liver-type arginaseP078242N
56Liver-type arginaseP078244N
57Acyl-CoA dehydrogenase, short-chain specific, mitoQ6IMX32N
 Liver-type arginaseP078242 
58Liver-type arginaseP078242N
 Liver-type aldolaseP008842 
59Malate dehydrogenase, mitochondrialP046363N
60Malate dehydrogenase, mitochondrialP046362N
61Malate dehydrogenase, mitochondrialP046365N
62Malate dehydrogenase, mitochondrialP046366N
 Ornithine carbamoyltransferase, mitochondrialP004812 
63Malate dehydrogenase, mitochondrialP046367N
64Malate dehydrogenase, mitochondrialP046365N
65Malate dehydrogenase, mitochondrialP046367N
66Malate dehydrogenase, mitochondrialP046366N
67Malate dehydrogenase, mitochondrialP0463611N
68Electron-transfer flavoprotein α subunit, (α-ETF)P1380313O/N
69Malate dehydrogenase, mitochondrialP046366N
70Collagen alpha 1Q630792N
71Electron-transfer flavoprotein β subunit (β-ETF)Q68FU34N
 Enoyl-CoA hydratase, mitochondrialP146042 
72Electron-transfer flavoprotein β subunit (β-ETF)Q68FU37N
 Enoyl-CoA hydratase, mitochondrialP146042 
73Electron-transfer flavoprotein β subunit (β-ETF)Q68FU312O/N
 Enoyl-CoA hydratase, mitochondrialP146042 
74Enoyl-CoA hydratase, mitochondrialP300847O/N
75Electron-transfer flavoprotein β subunit (β-ETF)Q68FU311O/N
763-Hydroxyacyl-CoA dehydrogenase type IIO7035110O/N
773-Hydroxyacyl-CoA dehydrogenase type IIO703519O/N
 Stress-70 protein, mitochondrial (GRP75)P487212 
783-Hydroxyacyl-CoA dehydrogenase type IIO703519O/N
79ATP synthase α chain, mitochondrialQ6P75311O/N

Many proteins were identified at apparently lower molecular weights (e.g., spots 35, 37, and 79 for the ATP synthase α subunit) and likely represent peptide fragments of the original proteins (e.g., 58 kDa for α-ATP synthase). Increased degradation of oxidized proteins was also observed with Grp78, Hsp60, β-ATP synthase, and the electron-transfer flavoprotein β subunit. The detection of smaller protein fragments may be related to the increased susceptibility of oxidized mitochondrial proteins to ubiquitin-dependent proteasomal activity, which is elevated under oxidative/nitrosative stress or spontaneous fragmentation of oxidized proteins, as observed with cytosolic proteins in alcohol-exposed mouse livers.19 Furthermore, increased degradation of oxidatively modified proteins may contribute to the reduced levels of several hepatic mitochondrial proteins such as ATP synthase and 3-ketoacyl-CoA thiolase in alcohol-exposed rats, as previously reported.27, 28

Our 2-D PAGE and MS analyses also revealed that many protein spots, though identified as being the same protein, exhibited different pI values that possessed the same or similar apparent masses. For example, several proteins with apparently different pI values were each identified as ALDH2 (spots 3,6-8,11,13,53,54), ATP synthase (spots 6-13), 3-ketoacyl-CoA thiolase (spots 15-24), electron-transfer flavoprotein β subunit (spots 26-30 and 32-34), and MDH (spots 55-67). These observations strongly indicate possible posttranslational modifications of these proteins, such as phosphorylation and oxidation. Because ethanol is known to activate various protein kinases, including c-Jun N-terminal protein kinase and p38 kinase (according to Pastorino et al.29 and our unpublished results), the observed pI shift of certain proteins identified in the current study may have been a result of phosphorylation. Alternatively, these spots may represent different degrees of oxidized cysteines as hyperoxidized proteins often exhibit lower pI values than their nonoxidized counterparts, as exemplified with hyperoxidized Cys (sulfinic and sulfonic acids) of peroxiredoxin.30

Inactivation of Oxidized and/or S-Nitrosylated Mitochondrial Enzymes in Alcohol-Exposed Rat Livers.

We hypothesized that the biological functions of certain oxidized proteins might be inhibited through S-nitrosylation and oxidative modifications of Cys and other oxidation-sensitive amino acid residues such as His, Met, and Tyr.31 Indeed, several proteins contain Cys in their catalytic sites such as ALDH220 and 3-ketoacyl-CoA thiolase.26 Therefore, we further evaluated whether the catalytic activities of ALDH2, 3-ketoacyl-CoA thiolase, and ATP synthase were altered in alcohol-exposed rat livers. ALDH2 activity was 0.28 ± 0.10 and 0.37 ± 0.09 units after chronic and binge alcohol exposure, respectively, representing a reduction of ALDH2 activity of approximately 80% and 59% in chronic- and binge-ethanol-treated rats, respectively, compared to the corresponding dextrose-treated controls (Fig. 5A). To further understand the mechanism underlying alcohol-mediated inhibition, ALDH2 was purified from the mitochondria of pair-fed control and ethanol-fed rats by immunoaffinity chromatography.20 Similar levels of immunopurified ALDH2 were present in both pair-fed controls and alcohol-fed rats (Fig. 5B, left panel). Immunoblot analysis with the anti-S-NO-Cys antibody showed the presence of S-nitrosylated Cys in ALDH2 purified from the alcohol-fed rats but not from the pair-fed control rats (right panel) and correlated with the inhibition of ALDH2 activity. Furthermore, the S-nitrosylated ALDH2 band disappeared, and ALDH2 activity was completely recovered in the presence of DTT, suggesting reversible S-nitrosylation and inactivation of ALDH2 (Fig. 5B, right panel).

Figure 5.

Inactivation of ALDH2 in alcohol-exposed rat livers and immunoblot analysis of immunopurified ALDH2 with the specific anti-ALDH2 or S-NO-Cys antibodies. (A) Mitochondrial ALDH2 activity in pair-fed control and chronic- or binge-alcohol-treated rat livers in the absence and the presence of 15 mM DTT. One unit of ALDH2 activity represents a reduction of 1 μmol NAD+/min/mg protein. (B) Mitochondrial ALDH2 proteins (1 mg/sample) from the pair-fed or chronically alcohol-fed rat livers in the absence and the presence of 15 mM DTT were purified by immunoaffinity chromatography, separated on 12% SDS-PAGE, and subjected to immunoblot analysis using, left, the anti-ALDH2 antibody or, right, the anti-S-NO-Cys antibody (*significantly different from the controls at P < .001; **significantly different from the controls at P < .0025).

Two Cys residues (Cys92 and Cys382) and one His (His352) are essential for the activity of 3-ketoacyl-CoA thiolase26 and are likely targets of oxidative/nitrosative modification followed by inactivation. Indeed, mitochondrial 3-ketoacyl-CoA thiolase activity was inhibited by approximately 38% and 51% after chronic and binge alcohol treatments, respectively. The level of 3-ketoacyl-CoA thiolase activity was 4.70 ± 1.88 and 2.91 ± 0.65 units in pair-fed controls and chronically ethanol-fed rats, respectively, whereas it was 5.45 ± 2.24 and 2.68 ± 0.05 units in dextrose-treated controls and binge-ethanol-treated rats, respectively (Fig. 6A). To further evaluate the functional implication of alcohol-mediated inhibition of this enzyme in mitochondrial β-oxidation of fatty acids, hepatic triglyceride and cholesterol levels were measured. The amount of hepatic triglycerides increased 2.4- and 2.2-fold with chronic and binge alcohol treatment, respectively, whereas the corresponding level of hepatic cholesterol was similarly elevated, by 2.9- and 2.1-fold, respectively (Fig. 6B). These results are consistent with an earlier report32 of alcohol exposure significantly elevating triglyceride and cholesterol levels. Histological analysis with Oil Red O staining confirmed the marked accumulation of fat in the alcohol-exposed rat livers compared to in the dextrose-treated controls (Fig. 6C).

Figure 6.

Inactivation of 3-ketoacyl-CoA thiolase in alcohol-exposed rat liver mitochondria and biochemical or histological analysis of fat accumulation. (A) Mitochondrial 3-ketoacyl-CoA thiolase activity in pair-fed control and chronic- or binge-alcohol-treated rat livers. One unit of thiolase was defined as the amount of thiolase that catalyzed the cleavage of 1 μmol acetoacetyl-CoA to acetyl-CoA per minute per gram of protein at room temperature. (B) Hepatic levels of total triglycerides and cholesterol in pair-fed control and chronic- or binge-alcohol-treated rat livers, with n = 4-5 per group (*significantly different from the controls at P < .05; **significantly different from the controls at P < .005). (C) Fat accumulation is visualized by Oil Red O staining of livers (n = 4 per group) from dextrose-fed control (left), chronically ethanol-fed (middle), and binge-ethanol-treated (right) rats. Representative livers are shown (×200).

It is possible that critical Tyr residues33 at the active site of ATP synthase may undergo nitration by elevated levels of nitrite and/or peroxynitrite in alcohol-exposed rat livers. Therefore, we measured ATP synthase activity and determined 3-nitrotyrosine levels in the immunoprecipitated ATP synthase from pair-fed controls and alcohol-exposed rats. Mitochondrial ATP synthase activity was also significantly inhibited by alcohol treatment. ATP synthase activity was 644.3 ± 95.3 and 107.4 ± 12.8 units in pair-fed control and ethanol-fed rats, respectively, whereas it was 894.2 ± 41.6 and 46.4 ± 0.9 units in dextrose-treated controls and binge ethanol-treated rats, respectively (Fig. 7A). To further elucidate the mechanism for alcohol-mediated inhibition of ATP synthase activity, the β-ATP synthase protein was immunoprecipitated from the mitochondria of pair-fed controls and chronically ethanol-fed rats, respectively. Immunoblot analysis with an anti-β-ATP synthase antibody showed similar levels of β-ATP synthase in pair-fed controls and alcohol-fed rats (Fig. 7B, left panel). Analysis with the anti-3-NT antibody clearly detected an immunoreactive 3-NT-labeled band in the immunoprecipitated β-ATP synthase from the alcohol-fed rat livers but not in that from the pair-fed control rats (right panel). One 3-NT-labeled band was observed only in alcohol-fed rat livers when the specific antibody to α-ATP synthase was used in the immunoprecipitation followed by immunoblot with the anti-3-NT antibody (data not shown). Although we can not determine whether this 3-NT immunoreactive band represents α- or β-ATP synthase, it is likely that this alteration accounts for the reduction in ATP synthesis. Furthermore, the 3-nitroTyr band disappeared in the presence of dithionite, possibly after its conversion to 3-aminoTyr (Fig. 7B, right panel). Consistent with the immunoblot results, MS analysis confirmed tyrosine nitrosylation in the β subunit of ATP synthase in the alcohol-exposed rat liver mitochondria (Supplementary Fig. 2).

Figure 7.

Inactivation of ATP synthesis in alcohol-exposed rat livers and immunoblot analysis of immunoprecipitated β-ATP with the antibody specific to β-ATP or 3-NT. (A) Mitochondrial ATP synthase activity of pair-fed control and chronic- or binge-alcohol-treated rat livers. One unit of ATP synthase represents the nanomoles of ATP produced per hour per milligram of protein at room temperature. (B) Mitochondrial β-ATP synthase from the chronically alcohol-fed and fair-fed control rat livers in the absence and the presence of 0.5 M sodium dithionite were immunoprecipitated with the specific antibody against β-ATP, separated on 12% SDS-PAGE, and subjected to immunoblot analysis using, left, the anti-β-ATP synthase antibody or, right, the anti-3-NT antibody (*significantly different from the controls at P < .025; **significantly different from the controls at P < .001).


It is well established that chronic ethanol exposure can damage tissues from the liver, brain, and many other organs. In the liver, chronic alcohol consumption can lead to alcoholic fatty liver, hepatitis, fibrosis, cirrhosis, and carcinogenesis. The detrimental effects of ethanol are partly mediated through increased oxidative/nitrosative stress resulting from the enhanced activity of a variety of proteins including CYP2E1, NADPH oxidase, xanthine oxidase, and inducible NOS, with reduced levels of GSH and other antioxidants.1–3 Increased nitric oxide and superoxide, which can be also produced by resident macrophages (Kupffer cells) and infiltrating leukocytes (neutrophils) in alcohol-exposed animals,2, 3 can lead to the elevated production of peroxynitrite, which may cause cellular damage.16, 34–36 In fact, the immunoblot results showed increased CD16, a marker protein for neutrophils, in the alcohol-exposed rats (Supplementary Fig. 3), suggesting that neutrophil-derived oxidants may also contribute to oxidative inactivation of mitochondrial enzymes/proteins, leading to mitochondrial dysfunction and cellular damage. Despite these reports, it is still poorly understood how increased oxidative/nitrosative stress directly affects mitochondrial proteins and their function. In fact, relatively little is known about which mitochondrial proteins are S-nitrosylated after alcohol treatment and how their functions are altered. Therefore, we sought to identify S-nitrosylated mitochondrial proteins in alcohol-exposed rat livers using Asc to selectively convert S-nitrosylated cysteines to unmodified cysteine residues22, 23 in a manner similar to that recently developed to identify oxidized proteins using DTT.19, 21 In addition to cysteine oxidation, a variety of amino acids (ie. Met, His, Trp, and Tyr) may also undergo modification under oxidative/nitrosative stress, potentially leading to inactivation and mitochondrial dysfunction (e.g., through reduced levels of ATP or accumulation of toxic aldehydes through inactivation of ATP synthase and ALDH2, respectively).16, 31, 34 In the absence of proper management, these changes may ultimately contribute to alcohol-related cell/tissue damage.

In a previous study,21 we used ethanol-sensitive E47 HepG2 cells to demonstrate that certain mitochondrial proteins such as ALDH2, Grp78, Hsp60, and prohibitin were oxidized following alcohol treatment. Our current results with alcohol-exposed rat livers not only confirm these earlier results but also suggest that many more mitochondrial proteins, such as α- and β-ATP synthase, GRP75, glutamate dehydrogenase, electron-transfer flavoprotein α and β subunits, and mitochondrial peroxiredoxin (Table 1), are both S-nitrosylated and oxidized after exposure to alcohol. The overall 2-D PAGE patterns of oxidatively modified mitochondrial proteins were very similar in the chronic and binge ethanol treatment groups. Because DTT is considered a stronger reducing agent than Asc,22, 23 we expected that DTT treatment would effectively reduce disulfides, sulfenic acids, and S-nitrosylated and S-glutathionylated proteins37 whereas milder reducing agents, such as Asc and glutathione, would only reduce S-nitrosylated or S-glutathionylated Cys residues.22, 23 With this assumption, a greater number of oxidized proteins (detected after treatment with DTT) than of S-nitrosylated proteins (after treatment with Asc) were likely to be produced. Contrary to this notion, the number of S-nitrosylated proteins was similar to that of oxidized proteins when only a small number of mitochondrial proteins were either S-nitrosylated (e.g., MDH) or oxidized (e.g., 3-ketoacyl-CoA thiolase). In fact, our results with both the binge and chronic ethanol treatments consistently show that multiple spots representing MDH (spots 55-67) and other proteins (Fig. 3) were only S-nitrosylated or S-glutathionylated, as detected after treatment with Asc but not DTT. These results suggest that DTT does not always reduce all S-nitrosylated cysteine residues as expected and that conversion of S-nitrosylated cysteines to free sulfhydryls may take place more readily than reduction of oxidized cysteines in the presence of Asc. In contrast, 3-ketoacyl-CoA thiolase (spots 15-24) was only oxidized and thus reduced only in the presence of DTT. This protein did not appear to be S-nitrosylated despite elevated levels of nitrite and peroxynitrite. The reason for this selective labeling or discrepancy needs to be further clarified.

Physiological levels of NO below the micromolar range are beneficial to the maintenance of normal signal transduction in various tissues and to endothelial vascular function.38, 39 However, large amounts of NO, produced through induction and/or activation of iNOS,40, 41 may cause nitrosative stress accompanied by cellular toxicity through modification of critically important Tyr and Cys residues of mitochondrial proteins. Recent studies have shown that alcohol-dependent hepatotoxicity and mitochondrial dysfunction are significantly reduced in iNOS-deficient mice compared to wild-type mice,15, 35, 36 suggesting an important role for peroxynitrite in alcohol-mediated cellular toxicity. Peroxynitrite could inactivate a variety of mitochondrial proteins, including NADH-ubiquinone oxidoreductase (complex I) and ATP synthase (complex V),41 through modification of active-site tyrosine residues, leading to mitochondrial dysfunction and cellular damage.42 Our data provide direct evidence of peroxynitrite-mediated modification of Tyr residues in the α and β subunits of ATP synthase and inhibition of its activity. Alternatively, oxidative modification of other amino acids in addition to the critical Tyr residues might also contribute to the reduced activity of ATP synthase. Our results showing reduced ATP synthase in alcohol-exposed rat livers are consistent with earlier reports27, 28, 43 and explain the molecular mechanism for its inhibition. Furthermore, increased levels of nitrite and peroxynitrite may inhibit other mitochondrial enzymes such as ALDH2 and 3-ketoacyl-CoA thiolase through oxidative modification of their critical Cys residues, as demonstrated with ALDH2 in cultured hepatoma cells20 and in alcohol-exposed rats.27, 28 Likewise, the cellular functions of other oxidatively modified mitochondrial proteins might be altered by similar mechanisms, although in the current study we did not investigate the functional changes of other proteins.

Despite the many reports of manifestations of alcohol-mediated fatty liver, the molecular mechanisms leading to increased fat accumulation in the liver are still unknown and under active investigation. Ethanol seems to inhibit the peroxisome proliferator-activated receptor alpha (PPARα)–related signaling pathway and PPARα-regulated gene expression.44, 45 Consistently, PPARα-knockout mice are more susceptible to alcohol-mediated liver injury (hepatomegaly, inflammation, toxicity, fibrosis, and mitochondrial swelling), suggesting a role for the peroxisome proliferator-activated receptor (PPARα) pathway in fat accumulation in alcoholic liver diseases. Alcohol is also known to activate acetyl-CoA carboxylase and fatty acid synthase46 through activation of a transcription factor sterol regulatory element binding protein 1 in the liver, leading to increased fat biosynthesis. In addition to increased fat biosynthesis, our current results suggest that the potential inhibition of mitochondrial fatty acid β-oxidation may also contribute to elevated fat accumulation in alcohol-exposed rats, as all 4 enzymes involved in mitochondrial β-oxidation of fat (Acyl-CoA dehydrogenase, β-hydroxyacyl-CoA dehydrogenase, enoyl-CoA hydratase, and 3-ketoacyl-CoA thiolase) were oxidized, and the activity of 3-ketoacyl-CoA thiolase (Fig. 6A) was inhibited following alcohol treatment. The peroxisomal β-oxidation of fatty acids may also have been inhibited based on the similar enzyme system in peroxisomes. Our results are consistent with an earlier study47 that found alcohol consumption to cause reduced fatty acid oxidation. Oxidation and/or S-nitrosylation of critical Cys residues in these proteins may lead to inhibition of the enzymes involved in β-oxidation of fatty acids and thus may contribute to the alcoholic fatty liver observed in animal models.32 It would be of interest whether these proteins are also oxidized and their activity reduced in cases of nonalcoholic steatohepatitis.

Mitochondrial ALDH2 is involved in the metabolism of acetaldehyde and lipid aldehydes such as 4-hydroxynonenal.48, 49 Therefore, inhibition of ALDH2 may cause accumulation of toxic acetaldehyde and lipid aldehyde such as the 4-hydroxynonenal produced after exposure to alcohol and other toxic compounds.50 Likewise, inhibition of ATP synthase leads to reduced levels of ATP,43 which may shift cell death mechanisms to necrosis from apoptosis, which requires ATP as an energy source. Inhibition of many mitochondrial proteins involved in the electron transport chain such as NADH-ubiquinone oxidoreductase (complex I) would leak more reactive oxygen species, profoundly affecting mitochondrial dysfunction.8, 9, 43 Our current data clearly underscore the important role of oxidative and/or S-nitrosylated modifications of key mitochondrial proteins or enzymes involved in mitochondrial β-oxidation, electron transfer, and energy production, which may be the underlying reason for alcohol-mediated fat accumulation, mitochondrial dysfunction, and increased susceptibility to tissue damage.

Using a biotin-NM-targeted proteomic approach, we have unambiguously identified many mitochondrial proteins oxidized and/or S-nitrosylated after chronic or binge exposure to alcohol. Repeated detection of smaller fragments of many mitochondrial proteins also was evidence that these modified proteins are easily degraded, either spontaneously or through increased susceptibility to proteolytic enzymes, similar to what was observed for oxidized cytosolic proteins. In addition, observing mitochondrial proteins at multiple pI values suggests significant posttranslational modification of oxidatively modified proteins. Finally, this approach can be used to detect oxidized and/or S-nitrosylated proteins in different subcellular fractions of tissues from various animal models of disease states and human tissues of clinical symptoms in which oxidative/nitrosative stress is implicated.


We thank Dr. Norman Salem Jr. and Dr. Robert Eskay for support and providing fresh liver tissues of dextrose-treated rats and binge-ethanol-treated rats, respectively.