Argininosuccinate synthase conditions the response to acute and chronic ethanol-induced liver injury in mice


  • Potential conflict of interest: Nothing to report.

  • Grant support: US Public Health service grants 5R01 AA017733, 5R01 AA017733-01S1, 5P20 AA017067, 5P20 AA017067-01S1 and 5P20 AA017067-03S1 from the National Institute on Alcohol Abuse and Alcoholism (to N.N.).


Argininosuccinate synthase (ASS) is the rate-limiting enzyme in both the urea and the L-citrulline/nitric oxide (NO·) cycles regulating protein catabolism, ammonia levels, and NO· generation. Because a proteomics analysis identified ASS and nitric oxide synthase-2 (NOS2) as coinduced in rat hepatocytes by chronic ethanol consumption, which also occurred in alcoholic liver disease (ALD) and in cirrhosis patients, we hypothesized that ASS could play a role in ethanol binge and chronic ethanol-induced liver damage. To investigate the contribution of ASS to the pathophysiology of ALD, wildtype (WT) and Ass+/− mice (Ass−/− are lethal due to hyperammonemia) were exposed to an ethanol binge or to chronic ethanol drinking. Compared with WT, Ass+/− mice given an ethanol binge exhibited decreased steatosis, lower NOS2 induction, and less 3-nitrotyrosine (3-NT) protein residues, indicating that reducing nitrosative stress by way of the L-citrulline/NO· pathway plays a significant role in preventing liver damage. However, chronic ethanol-treated Ass+/− mice displayed enhanced liver injury compared with WT mice. This was due to hyperammonemia, lower phosphorylated AMP-activated protein kinase alpha (pAMPKα) to total AMPKα ratio, decreased sirtuin-1 (Sirt-1) and peroxisomal proliferator-activated receptor coactivator-1α (Pgc1α) messenger RNAs (mRNAs), lower fatty acid β-oxidation due to down-regulation of carnitine palmitoyl transferase-II (CPT-II), decreased antioxidant defense, and elevated lipid peroxidation end-products in spite of comparable nitrosative stress but likely reduced NOS3. Conclusion: Partial Ass ablation protects only in acute ethanol-induced liver injury by decreasing nitrosative stress but not in a more chronic scenario where oxidative stress and impaired fatty acid β-oxidation are key events. (HEPATOLOGY 2012)

Alcoholic liver disease (ALD) is one of the leading causes of liver disease in the United States and results from ethanol binge drinking, chronic ethanol consumption, or both and progresses to fibrosis, cirrhosis, and ultimately hepatocellular carcinoma in many individuals. Steatosis and ethanol consumption are considered key hits for the development of ALD. 1-4 Mitochondrial damage, up-regulation of nitric oxide synthase-2 (NOS2) and generation of reactive oxygen and nitrogen species (ROS and RNS) condition cell viability, inflammation, and fat deposition in ALD. Thus, understanding the molecular mechanisms of pathological nitric oxide (NO·) production by NOS2 is of great relevance to prevent ethanol hepatotoxicity.

NOS2 catalyzes the nicotinamide adenine dinucleotide phosphate, reduced form (NADPH)-dependent oxygenation of L-arginine to NO· and L-citrulline. 5 Although the Nos2 gene lies quiescent under physiological conditions, cytokines, ROS, growth factors, and, most important, ethanol, initiate and sustain its activation. 2 Overexpressing NOS2 mediates mitochondrial damage as it occurs in ALD. 6 Previous work has shown that NOS2 is required for ALD due to generation of NO·-derived prooxidants. 7, 8 Indeed, ethanol hepatotoxicity was blunted in Nos2−/− mice as well as by a NOS2 inhibitor in wildtype (WT) mice. 7

The urea cycle is a metabolic pathway in which ammonia is converted to urea in the liver. The urea cycle enzymes along with the L-citrulline/NO· cycle catalyze de novo biosynthesis of L-arginine, which also serves as a substrate for NO· synthesis by NOS2. Five reactions occur within a functional complex or metabolon between the mitochondria and the cytosol (Supporting Fig. 1, green line):

  • 2ATP + HCOmath image + NHmath image → Carbamoyl phosphate + 2ADP + Pi (Carbamyl phosphate synthase-1, CPS1);

  • Carbamyl phosphate + Ornithine → Citrulline + Pi (Ornithine transcarbamylase, OTC);

  • Citrulline + Aspartate + ATP → Argininosuccinate + AMP + PPi (ASS);

  • Argininosuccinate → Arginine + Fumarate (Argininosuccinate lyase, ASL);

  • Arginine + H2O → Ornithine + Urea (Arginase-1, ARG1)

Although ASS and ASL are usually considered in the context of their contribution to the urea cycle, in conjunction with NOS2 they endow cells with a salvage pathway, the L-citrulline/NO· cycle (Supporting Fig. 1, red line) that continually generates L-arginine from L-citrulline for sustained NO· production. Physiological levels of L-arginine do not suffice to saturate NOS2 and changes in L-arginine bioavailability contribute to regulate NO· production.

Patients with type-I citrullinemia—an autosomal recessive urea cycle disorder due to Ass deficiency—develop hyperammonemia due to inefficient protein catabolism. 9, 10 Genetic disorders in the urea cycle cause steatosis and amino acid imbalance; however, the mechanism for these events is unknown. Hyperammonemia, changes in the concentration of amino acids and a decline in urea synthesis, occur in ALD patients. 11, 12 The role of the L-citrulline/NO· cycle in the liver, the potential role of ASS as an enzymatic “switch” to provide a substrate for NOS2-induced activity, and the subsequent excess of NO· biosynthesis in ALD is still to be defined. A detailed analysis of the mechanisms leading to increased nitrosative stress appears timely to advance our understanding of acute and chronic ethanol-induced liver injury.

Using a combination of proteomics and a systems biology approach to uncover mitochondrial and cytosolic proteins involved in ALD that could impact NO· synthesis, we identified ASS as up-regulated in rat hepatocytes by chronic ethanol feeding. Furthermore, livers from patients with ALD or with stage 3 hepatitis C virus (HCV)-induced cirrhosis showed correlation between the increase in ASS and NOS2, suggesting a potential link between ASS, NO· generation by NOS2, and ALD. ASS, as an enzyme shared by the urea and the L-citrulline/NO· cycles, could have a rate-limiting role for high-output NO· synthesis by way of NOS2. Virtually nothing is known of how acute or chronic ethanol consumption modulates ASS expression and how the L-arginine recycling pathway may affect NO· generation and liver injury under acute and chronic ethanol ingestion. We hypothesized that up-regulation of ASS by alcohol could increase NO· synthesis by NOS2, thus contributing to the pathophysiology of ALD.


3-NT, 3-nitrotyrosine; ACC, acetyl-CoA carboxylase; ALD, alcoholic liver disease; ALT, alanine aminotransferase; AMPK, AMP-activated protein kinase; ARG1, arginase1; ASL, argininosuccinate lyase; ASS, argininosuccinate synthase; CPS, carbamyl phosphate synthase; CPT, carnitine palmitoyl transferase; CYP2E1, cytochrome P450 2E1; ER, endoplasmic reticulum; FAS, fatty acid synthase; GCLC, glutamate-cysteine ligase catalytic subunit; GCLM, glutamate-cysteine ligase modifier subunit; GSH, glutathione; GR, glutathione reductase; GT, glutathione-S-transferase; H&E, hematoxylin and eosin; HControl, hepatocytes from rats fed the control Lieber-DeCarli diet; HEthanol, hepatocytes from rats fed the ethanol Lieber-DeCarli diet; ICAT, isotope-coded affinity tag; IHC, immunohistochemistry; NO·, nitric oxide; NOS2, nitric oxide synthase-2; OTC, ornithine transcarbamylase; PGC, peroxisomal proliferator-activated receptor coactivator; PPAR, peroxisome proliferator-activated receptor; RNS, reactive nitrogen species; ROS, reactive oxygen species; SIRT, sirtuin; SREBP, sterol regulatory element-binding protein; αMDLA, α-methyl-D,L-aspartate; WT, wildtype.

Materials and Methods

Please see Supporting Materials and Methods (online).


In Vivo Model of Ethanol Feeding for the Identification of ASS Up-regulation.

Our initial goal focused on identifying mitochondrial and/or cytosolic proteins up-regulated by ethanol to dissect how they could activate key metabolic pathways contributing to alcohol-induced liver injury. The in vivo model of ethanol feeding used in our proteomics study was rats fed the control or the ethanol Lieber-DeCarli diets for 32 weeks. 13

Livers showed minimal steatosis in control rats, whereas rats fed ethanol showed periportal and pericentral micro- and macrovesicular steatosis (Fig. 1A). Serum ammonia increased by 50% (Fig. 1B, left), whereas serum urea decreased by 20% (Fig. 1B, right) in the ethanol group compared with the control group. Ultrastructural studies using transmission electron microscopy revealed normal architecture in control livers with well-stacked endoplasmic reticulum (ER), normal mitochondria, and minimal microvesicular steatosis (Fig. 1C, left). Livers from ethanol-fed rats showed dilated ER, a large number of electron-dense mitochondria, abundant micro- and macrovesicular steatosis, and disrupted cellular membranes (Fig. 1C, right). All these parameters were good indicators of mitochondrial and ER impaired function, which play a role in the development of ALD.

Figure 1.

Chronic effect of the control and ethanol Lieber-DeCarli diets in rats and identification of ASS up-regulation by ethanol. Liver H&E staining from rats fed the control diet showed minimal steatosis (A, left), whereas rats fed the ethanol diet showed periportal and pericentral micro- (yellow arrow) and macrovesicular (green arrow) steatosis (A, right. PV: portal vein). Serum ammonia and urea. Results are average values ± SEM. n = 10/group; *P < 0.05 and ***P < 0.001 for chronic ethanol versus control (B). Transmission electron microscopy showing ultrastructural features (micro- (yellow arrow) and macrovesicular steatosis (green arrow), black arrow ER: endoplasmic reticulum, L: lysosome, M: mitochondria, N: nucleus). Livers from control rats show normal mitochondria, minimal microvesicular steatosis, and stacked ER (left), whereas livers from ethanol-fed rats show greater micro- and macrovesicular steatosis, disrupted ER, and dilated mitochondria (right) (C). Western blot analysis for CYP2E1, NOS2, DYNAMIN, HSP70, and ACTIN in HControl and HEthanol. The quantification of the intensity of the signals corrected by that of ACTIN is indicated under the blots. Results are average values ± SEM. n = 3/group; **P < 0.01 and ***P < 0.001 for HEthanol versus HControl (D). Tryptic peptides of ASS were identified by mass spectrometry as highlighted in orange in the sequence. The calculation of their abundance ratios is shown in the chromatograms reconstructed from LC/MS-MS data. Red lines are raw spectra, blue lines are smoothed spectra, green lines are the area under the curve used in the ratio calculation, and horizontal cyan lines are the background in the spectra. n = 5/group (E).

Proteomics Analysis Identified ASS as Up-regulated by Ethanol.

To identify proteins participating in ethanol hepatotoxicity and their link with signaling pathways involved in ALD, particularly NO· production, next we used a combination of proteomics along with a systems biology approach. To this end, first we used the mass spectrometry-based isotope-coded affinity tag (ICAT) proteomics technique to identify differentially expressed proteins in hepatocytes from ethanol-fed rats (HEthanol) compared with hepatocytes from control rats (HControl). Second, to dissect the differentially regulated proteins in the context of protein interaction networks we used the Institute for Systems Biology Trans-Proteomics Pipeline (Seattle) and the Gaggle 14 computer platform. Lastly, we narrowed down our search by focusing on the subproteome of mitochondrial and/or cytosolic proteins of potential significance for the development of ALD.

The ICAT labeling methodology and the proteomics analysis identified multiple differentially expressed proteins in HControl versus HEthanol with probability scores >0.5 (<5% error rate). Among them, there were several well-known alcohol-regulated proteins such as cytochrome P450 2E1 (CYP2E1) and NOS2, which were validated by western blot analysis, whereas other proteins, such as DYNAMIN and HSP70, decreased by ethanol (Fig. 1D). The acquired dataset was further analyzed on the Gene Ontology Categories, Kyoto Encyclopedia of Genes and Genomes Pathways, and Protein Interaction Networks using the Gaggle platform. 14 The systems-based quantitative proteomics analysis led us to focus on the urea and the L-citrulline/NO· cycles as likely impaired under ethanol consumption because a potential link with NO· production could be established. ASS, a novel ethanol-specific induced protein, was identified in the proteomics analysis (Fig. 1E). Hence, we explicitly selected it as a protein of interest in the follow-up analysis because it could play a role in ALD by regulating de novo biosynthesis of L-arginine from L-citrulline for high-output NO· generation by NOS2.

Next, ASS expression was validated by western blot analysis in hepatocytes. We found a 4.2-fold increase in HEthanol versus HControl (Fig. 2A). To better understand the potential physiological role of the induction of ASS, other enzymes from the urea cycle and/or the L-citrulline/NO· cycle were studied. Western blot analysis showed that ethanol induced ASL (2.2-fold), ARG1 (2.3-fold), NOS2 (2.8-fold), and 3-NT (4.1-fold), likely because of NOS2 induction and overproduction of NO· leading to nitrosative stress, whereas a decrease was observed in hepatocyte arginine residues (Fig. 2A).

Figure 2.

Chronic ethanol feeding induced the expression of enzymes from the urea and the L-citrulline/NO· cycles. The expression of ASS, ASL, ARG1, arginine residues, NOS2, 3-NT residues, and ACTIN in HControl and HEthanol was analyzed by western blot. n = 4 (A). Human liver explants from healthy (n = 5), stage-3 HCV-cirrhotic (n = 4), and ALD (n = 2) patients were evaluated for the expression of ASS, NOS2, 3-NT residues, COLLAGEN-I, and ACTIN by western blot (B). Rat hepatocytes were either treated with 5 μM citrulline, an inducer of ASS, transfected with control siRNA or Ass siRNA, or incubated with inhibitors of ASS such as 15 μM fumonisin B1, 10 μM mithramycin A, and 50 μM α-MDLA. Experiments were repeated three times and a representative blot for NOS2 and ASS is shown (C-E). The quantification of the intensity of the signal corrected by that of ACTIN is indicated under all blots. **P < 0.01 and ***P < 0.001 for treated or diseased versus control.

ALD and Cirrhosis Patients Showed Induction of ASS and NOS2.

To determine whether the results obtained in primary rat HControl and HEthanol reflected events similar to those taking place in human liver disease, we used liver samples from healthy, cirrhosis, and ALD patients. ASS, NOS2, 3-NT residues, and collagen-I increased in cirrhotic and ALD compared with control individuals (Fig. 2B). ASL and ARG1 were also elevated in cirrhosis patients (Supporting Fig. 3). These results in humans strengthen the possible link between ASS, the potential downstream events (i.e., regulation of NO· production by NOS2), ALD, and perhaps cirrhosis.

ASS Modulated NOS2 Expression.

To establish a connection between ASS and NOS2, cells were treated with inhibitors or substrates of ASS. Treatment of HControl with 5 μM citrulline for 24 hours—a substrate and inducer of ASS—elevated the expression of ASS by 3.1-fold and of NOS2 by 2.8-fold (Fig. 2C). Moreover, transfecting HControl with Ass small interfering RNA (siRNA) decreased both ASS and NOS2 proteins (Fig. 2D). Likewise, inhibiting ASS with either 15 μM fumonisin B1, 10 μM mithramycin A, or 50 μM α-methyl-D,L-aspartate (α-MDLA) for 24 hours—known inhibitors of ASS—reduced NOS2 expression in HControl (Fig. 2E). Thus, modulation of ASS expression regulates NOS2 activity and ultimately NO· production, a mechanism expected to participate in the pathophysiology of ALD.

Binge and Chronic Ethanol Feeding and the Urea Cycle Enzymes.

To determine the effects of Ass deficiency in binge and chronic ethanol drinking, mice were either gavaged twice with saline solution or ethanol or were fed with the control or ethanol Lieber-DeCarli diets for 7 weeks. Western blot analysis showed a 3-fold induction in ASS protein in both ethanol-binged and chronic ethanol-fed WT mice (Fig. 3A), yet there was only a slight increase in Ass+/− mice under chronic ethanol consumption (Fig. 3B). Chronic ethanol feeding decreased CPS1 expression by ≈20% in both WT and Ass+/− mice (Fig. 3B). The rest of the enzymes in the urea cycle remained similar under either binge or chronic ethanol feeding (Fig. 3A,B).

Figure 3.

Binge and chronic ethanol consumption impair the urea cycle more in Ass+/− than in WT mice. Western blot analysis for the urea cycle enzymes. The quantification of each signal corrected by that of CALNEXIN is shown for the ethanol binge model (A) and for the chronic ethanol feeding model (B). Liver and serum ammonia and urea (C-E). Results are average values ± SEM. n = 8/group; *P < 0.05 and **P < 0.01 for binge or chronic ethanol versus control. •P < 0.05 and ••P < 0.01 for Ass+/− versus WT.

Because defects in the urea cycle lead to hyperammonemia and hepatic encephalopathy, 7 next we analyzed ammonia and urea levels. Ass+/− mice showed higher liver ammonia but there were no changes in liver urea in either model (Fig. 3C,D). Chronic ethanol treatment increased serum ammonia (not statistically significant) (Fig. 3E, left) and reduced serum urea (Fig. 3E, right). Thus, these defects reflect functional impairment of the urea cycle by ethanol, which was more noticeable in Ass+/− than in WT mice, hence contributing to liver damage.

Ass Deficiency Reduced Hepatic Steatosis in the Ethanol Binge Model but Exacerbated It in the Chronic Ethanol-Feeding Model.

The pathology scoring from hematoxylin and eosin (H&E)-stained slides indicated minimal necrosis and inflammation in all mice but revealed the presence of lipid droplets (micro- and macrovesicular steatosis) in ethanol-binged WT but not in Ass+/− mice (Fig. 4A). This was further quantified by the steatosis grade (Fig. 4B) and by the hepatic triglycerides (Fig. 4C). Alanine aminotransferase (ALT) activity increased by 2-fold only in ethanol-binged WT mice (Fig. 4D). In contrast, chronic ethanol feeding caused greater inflammation, necrosis, and ductular reaction in Ass+/− than in WT mice (Fig. 4E,F). The steatosis grade (Fig. 4F), oil red O staining, and morphometry analysis (Supporting Fig. 4A-4B) demonstrated more neutral fat in chronic ethanol-fed Ass+/− than in WT mice, suggesting more liver injury by partial Ass ablation in the chronic ethanol feeding model.

Figure 4.

Ass deficiency protects from liver injury in binged mice but exacerbates it in chronic ethanol-fed mice. H&E staining showed increased micro- (arrow) and macrovesicular (arrow) steatosis in binged WT compared with Ass+/− mice (A). Steatosis grade (B), liver triglycerides (C), and serum ALT activity (D). The ethanol Lieber-DeCarli diet caused more inflammation (blue arrow), micro- (yellow arrow) and macrovesicular (green arrow) steatosis in Ass+/− than in WT mice as shown by the H&E staining (E). The scores for inflammation, necrosis, ductular reaction, and steatosis are shown (F). PV: portal vein. Results are average values ± SEM. n = 8/group; *P < 0.05 and ***P < 0.001 for binge or chronic ethanol versus control; •P < 0.05 and ••P < 0.01 for Ass+/− versus WT.

Ass Deficiency Reduced Nitrosative Stress in the Ethanol Binge Model but Not in the Chronic Ethanol-Feeding Model.

In order to investigate the effect of Ass deficiency on NOS2 and NO· generation, immunohistochemistry (IHC) was performed. There was more intense staining for NOS2 (5-fold) and 3-NT residues (10-fold)—the footprint for nitrosative stress—in WT given an ethanol binge compared with Ass+/− mice, which was quantified by morphometry analysis (Fig. 5A-C). Chronic ethanol feeding elevated NOS2 (2-fold, not statistically significant) and 3-NT protein adducts (3-fold) both in WT and in Ass+/− mice (Fig. 5D-F). Western blot analysis showed a 4- and a 2-fold increase in NOS2 in binged WT and Ass+/− mice, respectively (Supporting Fig. 5A, left), whereas there was only a 2-fold increase in NOS2 expression in both genotypes after chronic ethanol feeding. NOS1 and NOS3 expression remained similar with binge or chronic ethanol feeding in both WT and Ass+/− mice (Supporting Fig. 5A). However, serum nitrites plus nitrates, considered surrogate markers of NOS3 activity, remained similar in the binge model (Supporting Fig. 5B, left), but were lower in chronic ethanol-fed Ass+/− than in WT mice (Supporting Fig. 5B, right).

Figure 5.

Ass+/− mice are protected from nitrosative stress induced by an ethanol binge. IHC for NOS2 protein (arrow) (A) and 3-NT residues (arrow) (B) in binged WT and Ass+/− mice. Morphometry analysis for NOS2 and 3-NT staining (C). IHC and morphometry analysis for NOS2 and 3-NT in chronic ethanol-fed WT and Ass+/− mice (D-F). CV: central vein. Results are average values ± SEM. n = 8/group; *P < 0.05 and **P < 0.01 for binge or chronic ethanol versus control; •P < 0.05 for Ass+/− versus WT.

ROS—key players in ethanol toxicity—are generated among others by microsomal CYP2E1, which is induced by ethanol itself. 15, 16 Because alcohol intake stabilizes CYP2E1 against degradation contributing to liver injury, we examined CYP2E1 expression. Western blot analysis showed similar CYP2E1 induction by ethanol binge (Supporting Fig. 6A, left) and by chronic ethanol feeding (Supporting Fig. 6A, right) in WT and in Ass+/− mice. Lastly, IHC for 4-HNE—a lipid peroxidation end-product—was similarly increased by the ethanol binge in both groups of mice (not statistically significant) (Supporting Fig. 6B); however, the increase was much higher in chronically ethanol-fed Ass+/− than in WT mice (Supporting Fig. 6C).

Ass+/− Mice Showed Lower GSH Levels than WT Mice After Chronic Alcohol Feeding.

Glutathione (GSH) is a key endogenous antioxidant participating in detoxification reactions. 17 WT and Ass+/− mice showed similar basal GSH, whereas binge drinking reduced GSH level by 50% in both WT and Ass+/− mice (Supporting Fig. 7). Total and mitochondrial GSH were higher in Ass+/− than in WT mice in the control group chronically fed a high-fat diet (Fig. 6A). This may have served as a protective mechanism in the ethanol binge model in addition to decreased NO· generation due to impairment of the L-citrulline/NO· cycle. In the chronic ethanol-feeding model, however, both total and mitochondrial GSH decreased only in Ass+/− mice, hence promoting liver injury (Fig. 6A). A 20% increase in oxidized GSH occurred in the ethanol-fed Ass+/− compared with WT mice (not shown).

Figure 6.

Chronic ethanol feeding reduced the antioxidant defense in Ass+/− mice. Total and mitochondrial GSH levels in WT and Ass+/− mice chronically fed ethanol (A). Western blot analysis for GCLC and GCLM in WT and Ass+/− mice chronically fed ethanol. The quantification of the intensity of the signals corrected by that of ACTIN is indicated under the blots (B). Catalase (C), GT (D), and GR (E) activities. Results are average values ± SEM. n = 8/group; *P < 0.05, **P < 0.01 and ***P < 0.001 for chronic ethanol versus control; •P < 0.05 and ••P < 0.01 for Ass+/− versus WT.

The decrease in GSH possibly occurred due to a reduction in glutamate-cysteine ligase (GCLC and GCLM), the rate-limiting enzymes for GSH synthesis (Fig. 6B). Glutathione-S-transferase (GT) catalyzes the conjugation of GSH to various substrates for detoxifying endogenous compounds. Chronic ethanol feeding induced GT by 3-fold in WT mice and by 2-fold in Ass+/− mice. Furthermore, there was a 20% decrease in catalase and glutathione reductase (GR) activities in the ethanol-fed Ass+/− compared with WT mice (Fig. 6C-E). Lastly, because the urea cycle could also condition amino acid availability for GSH synthesis (i.e., methionine, glutamate, and cysteine), we analyzed amino acid content by high-performance liquid chromatography (HPLC). Chronic ethanol feeding increased glutamate and cysteine more in Ass+/− mice than in WT mice, likely affecting GSH synthesis (Supporting Table 1).

Chronic but Not Binge Ethanol Drinking Induced More Steatosis in Ass+/− than in WT Mice.

Because the data suggested that Ass+/− developed less steatosis than WT mice after ethanol binge drinking and the opposite occurred in the chronic ethanol model, we studied the expression of key proteins involved in lipolysis and lipogenesis. Peroxisome proliferator-activated receptor-γ (PPARγ) and sterol regulatory element-binding protein-1 (SREBP-1) are lipogenic transcription factors, whereas PPARα regulates lipolysis. 18, 19 Western blot analysis demonstrated greater reduction in PPARγ and SREBP1 after the ethanol binge in Ass+/− than in WT mice; however, PPARα showed similar expression in both groups (Fig. 7A, left). Hence, lipogenesis was impaired in Ass+/− mice after an ethanol binge. In contrast, PPARα, PPARγ, and SREBP-1 did not vary after chronic ethanol feeding in WT and Ass+/− mice (Fig. 7A, right).

Figure 7.

Lipogenesis and lipolysis regulators in the binge and the chronic ethanol feeding models. Western blot analysis for PPARγ, SREBP-1, PPARα, pAMPKα, AMPKα1, AMPKα2, ACTIN, and CALNEXIN in both models in WT and Ass+/− mice (A). Western blot analysis for FAS, ACC2, and CALNEXIN in both models in WT and Ass+/− mice (B). Sirt-1 (C), Pgc1-α (D), and Cpt-1 (E) mRNAs in chronic ethanol-fed WT and Ass+/− mice. Western blot analysis for CPT-II and CALNEXIN in chronically ethanol-fed WT than in Ass+/− mice (F). The quantification of the intensity of all signals corrected by that of ACTIN or CALNEXIN is indicated under the blots. Results are average values ± SEM. n = 8/group; *P < 0.05 and ***P < 0.001 for binge or chronic ethanol versus control; •P < 0.05 and ••P < 0.01 for Ass+/− versus WT.

Adenosine monophosphate (AMP)-activated protein kinase (AMPK) regulates cellular energy homeostasis and promotes fatty acid oxidation by inactivating acetyl-CoA carboxylase (ACC), 20 the rate-limiting enzyme for fatty acid synthesis and a potent inhibitor of CPT1. Ass+/− mice showed lower basal AMPKα than WT mice. Although no major difference was detected in ethanol binge drinking (not shown), the basal ratio of pAMPKα to total AMPKα was greatly reduced in Ass+/− mice compared with WT and also by chronic ethanol exposure in both genotypes (Fig. 7A, right). Fatty acid synthase (FAS) and ACC2, which provide malonyl-CoA for fatty acid biosynthesis, were analyzed. Binge drinking altered neither FAS nor ACC2 expression (Fig. 7B, left), whereas chronic ethanol feeding reduced FAS in both WT and Ass+/− mice (Fig. 7B, right). Fatty acid export into the plasma was also similar in both ethanol-fed groups (not shown).

SIRT-1 inactivates SREBP-1 by way of deacetylation. 19 Although no differences were observed in the binge model (not shown), chronic ethanol feeding increased Sirt-1 (Fig. 7C) and Pgc-1α (Fig. 7D) messenger RNAs (mRNAs) more in WT than in Ass+/− mice and no differences were observed in Acc mRNA (not shown).

Ass+/− Mice Had Higher CPT-I Activity than WT Mice.

CPT-I is the rate-limiting enzyme in fatty acid catabolism for the conversion of long-chain fatty acids into long-chain acylcarnitines, whereas CPT-II is responsible for the release of long-chain fatty acids from carnitine, inside the mitochondrial matrix, for fatty acid β-oxidation. 21 Although no changes were observed by ethanol binge drinking (not shown), Cpt1 mRNA (Fig. 7E) and CPT-II protein (Fig. 7F) were induced in chronic ethanol feeding in both WT and Ass+/− mice. The ratio of free carnitine (C0) to long-chain acylcarnitine (C16+C18) is an indicator of CPT-I activity. Ass+/− mice had higher CPT-I activity (lower C0/C16+C18 ratio) (control group: 32.7 ± 12.2; ethanol group: 31.2 ± 5.8) compared with WT mice (control group: 52.8 ± 15.6; ethanol group: 56.0 ± 19.7) but chronic ethanol feeding did not affect CPT-I activity (P < 0.05 for Ass+/− versus WT). However, CPT-II protein expression was significantly increased by ethanol feeding in WT mice compared with Ass+/− mice (Fig. 7F); hence, fatty acid β-oxidation was impaired in chronic ethanol-fed Ass+/− mice. Thus, although Ass+/− mice may have efficient fatty acid transport into the mitochondria for β-oxidation, the decrease in CPT-II under chronic ethanol drinking impaired the efficiency of this pathway, leading to fat accumulation.


Up-regulation of NOS2 along with generation of RNS plays a major role in alcohol-induced liver injury. 22 The overwhelming research on the production of NO· has been focused on the different isoforms of NOS. However, a renewal of interest in the regulation of ASS has recently emerged as a result of its possible rate-limiting role for high-output NO· synthesis. 2 Using an integrated proteomics and systems biology approach we identified NOS2 along with ASS—the rate-limiting enzyme in the urea and L-citrulline/NO· cycles—as significantly coinduced under chronic ethanol consumption in rodents, which was also validated in human samples.

In addition, ASS, ASL, ARG1, and 3-NT residues were up-regulated in both hepatocytes isolated from chronic ethanol-fed rats and in ALD and cirrhosis patients. Moreover, NOS2 was regulated by altering ASS expression in hepatocytes. Treatment with L-citrulline, an inducer of ASS, increased the expression of both ASS and NOS2, whereas downregulation of ASS by siRNA or other inhibitors significantly reduced NOS2 expression. Because the urea cycle is key for hepatic amino acid content, this result suggested that ASS may control NOS2 by modulating substrate availability in the L-citrulline/NO· cycle. Thus, the correlation between both enzymes and the induction of nitrosative stress prompted us to study the contribution of ASS to the pathogenesis ALD using in vivo models of ethanol binge and chronic ethanol drinking.

Although we initially speculated that partial deletion of Ass would confer protection from liver injury under both ethanol-drinking schemes, the in vivo data suggested that protection occurred only under an ethanol binge. This difference was likely due to a lack of ASS induction in the acute versus the chronic model of ethanol drinking and to effects on oxidative stress, lipid peroxidation, fatty acid β-oxidation, and perhaps NOS3 impairment. Therefore, ASS may have distinctive roles depending on the stage of ALD.

WT mice showed up-regulation of ASS in the binge and the chronic ethanol feeding models, whereas the rest of the urea cycle enzymes remained similar. The increased ASS expression in WT was consistent with the increased ASS expression in human cirrhosis and alcoholic liver samples. Although Ass+/− mice displayed no significant changes in any enzyme in the binge model, there was a decrease in CPS1 under chronic alcohol feeding, which could lead to hyperammonemia and thus to liver injury. Similar blood alcohol levels were found in WT and Ass+/− mice, indicating that alcohol metabolism was not affected by Ass deficiency (not shown). Although ASS increased in cirrhosis and alcoholic patients, partial Ass ablation exacerbated chronic alcohol-induced liver injury in mice. This finding suggested that elevated ASS expression during liver injury could have a protective role.

Partial Ass ablation protected from ethanol binge-induced liver injury because Ass+/− mice developed less nitrosative stress and steatosis than WT mice. This was likely due to lower NOS2, limited NO· generation, and 3-NT protein adduct formation due to lack of ASS induction by the ethanol binge. Although all three isoforms of NOS are expressed in the liver, it is likely that the extent of nitrosative stress was merely due to NOS2-derived NO· because the expression of NOS1 and NOS3 and serum nitrates plus nitrites were not altered by binge drinking. The expression of PPARγ and SREBP-1, two lipogenic transcription factors, was down-regulated by ethanol binge in Ass+/− compared with WT mice, whereas PPARα, a lipolytic factor, was unaffected, thus preventing fat deposition.

Conversely, in the chronic ethanol-feeding model, Ass+/− mice showed greater hepatic inflammation, necrosis, ductular reaction, and steatosis than WT mice. This was accompanied by high ammonia in liver and serum and by low urea. Ammonia is known for inhibiting fatty acid oxidation, thus promoting steatosis. Ethanol oxidation increases the ratio of NADH/NAD+ reducing urea synthesis by inhibiting the oxidative deamination of amino acids that precede the urea cycle. 23, 24 Increases in NADH also disrupt dehydrogenase-related reactions in the mitochondria, thereby suppressing fatty acid β-oxidation. 25 Importantly, alcohol also decreases ATP—which is required for the urea cycle 11, 26—and ATP was significantly decreased in Ass+/− mice chronically fed ethanol (9.7 ± 2.1 versus 5.7 ± 1.2 nmol/mg protein, P < 0.05). Although the expression of the urea cycle enzymes ASL, OTC, and ARG1 did not change in chronic ethanol feeding, reduced ATP levels may impair the urea cycle and lead to hyperammonia. It has been shown that hyperammonemia and reduction in urea synthesis occurs both in ALD and in cirrhosis patients. 27-29 Despite comparable nitrosative stress in the ethanol-fed mice and the perhaps metabolic induction of ASS in Ass+/− mice, the urea and L-citrulline/NO· cycles remained somewhat dysfunctional, as also demonstrated by the concentration of amino acids: increased glutamate and arginine and decreased ornithine.

NOS2 expression and 3-NT adducts were comparable between chronic ethanol-fed WT and Ass+/− mice; however, serum nitrates plus nitrites decreased in Ass+/− mice but not in WT mice. This could suggest that impaired urea and L-citrulline/NO· cycles during chronic ethanol feeding may condition NOS3 activity to enhance liver injury, as previous work has demonstrated that inhibition of NOS3 enhances alcohol-induced liver injury. 7, 30 Although there was no difference in CYP2E1 expression, higher lipid peroxidation occurred in chronic ethanol-fed Ass+/− than in WT mice, suggesting significant oxidative stress-mediated liver injury.

Impairment in cysteine metabolism is associated with fatty liver. Chronic ethanol consumption increased cysteine in Ass+/− compared with WT mice. Furthermore, there was down-regulation in the antioxidant defense in chronically ethanol-fed Ass+/− mice, because total and mitochondrial GSH, GCLC, GCLM, GT, GR, and catalase decreased compared with WT mice. In addition, the presence of LPO end-products likely contributed to liver damage in Ass+/− mice in the chronic ethanol-feeding model.

Although no major differences occurred in the lipolysis and lipogenesis factors under chronic ethanol consumption, fatty acid transport into the mitochondria for β-oxidation increased in Ass+/− fed the control diet compared with WT mice, because Ass+/− mice showed higher CPT-I activity (lower C0/C16+C18 ratio). Nevertheless, chronic ethanol feeding impaired β-oxidation because CPT-II, which transports long-chain acylcarnitine to the mitochondrial matrix for production of acetyl-CoA for β-oxidation, decreased in Ass+/− compared with WT mice. The decrease in ATP, Sirt-1, and Pgc-1α mRNAs along with the ratio of pAMPKα to total AMPKα likely facilitated fat deposition. It is also possible that generation of free radicals following ethanol exposure could modify key enzymes for fatty acid β-oxidation (e.g., acyl-CoA dehydrogenase, enoyl-CoA hydratase, and β-hydroxy-acyl-CoA dehydrogenase) leading to their inactivation and thus to fat accumulation. 31