Potential conflict of interest: Nothing to report.
The hepatoprotective mechanisms of N-acetylcysteine (NAC) in non–acetaminophen-induced liver injury have not been studied in detail. We investigated the possibility that NAC could affect key pathways of hepatocellular metabolism with or without changes in glutathione (GSH) synthesis. Hepatocellular metabolites and high-energy phosphates were quantified from mouse liver extracts by 1H- and 31P-NMR (nuclear magnetic resonance) spectroscopy. 13C-NMR-isotopomer analysis was used to measure [U-13C]glucose metabolism through pyruvate dehydrogenase (PDH) and pyruvate carboxylase (PC). NAC (150-1,200 mg/kg) increased liver concentrations of GSH from 8.60 ± 0.48 to a maximum of 12.95 ± 1.03 μmol/g ww, whereas hypotaurine (HTau) concentrations increased from 0.05 ± 0.02 to 9.95 ± 1.12 μmol/g ww. The limited capacity of NAC to increase GSH synthesis was attributed to impaired glucose metabolism through PC. However, 300 mg/kg NAC significantly increased the fractional 13C-enrichment in Glu (from 2.08% ± 0.26% to 4.00% ± 0.44%) synthesized through PDH, a key enzyme for mitochondrial energy metabolism. This effect could be uncoupled from GSH synthesis and was associated with the prevention of liver injury induced by tert-butylhydroperoxide and 3-nitropropionic acid. In conclusion, NAC (1) has a limited capacity to elevate GSH synthesis; (2) increases HTau formation linearly; and (3) improves mitochondrial tricarboxylic acid (TCA) cycle metabolism by stimulation of carbon flux through PDH. This latter effect is independent of the capacity of NAC to replete GSH stores. These metabolic actions, among other yet unknown effects, are critical for NAC's therapeutic value and should be taken into account when deciding on a wider use of NAC. (HEPATOLOGY 2006;43:454–463.)
N-acetylcysteine (NAC) is the most widely administered antidote in acetaminophen intoxication.1–4 In this setting, protection by NAC is believed to be attributable to its ability to regenerate glutathione (GSH) stores due to its capacity to provide cysteine (Cys) residues. However, NAC has also been shown to improve patient outcome after late administration,5, 6 pointing toward mechanisms independent from GSH replenishment. Furthermore, because NAC is recognized as an antioxidant, it might be useful in the setting of liver injuries different from those caused by acetaminophen intoxication.6–11
NAC has also been reported to improve oxygen delivery,4 to increase systemic oxygen consumption,12 and to have beneficial effects on liver blood flow and function.6, 8, 10 Because liver oxygen consumption is coupled with aerobic oxidation of energy substrates in the mitochondrial tricarboxylic acid (TCA) cycle, we hypothesize that NAC might improve hepatocellular energy metabolism. This effect might be salutary in the setting of non–acetaminophen-induced liver injury.
We were particularly interested in investigating how NAC is involved in hepatocellular metabolic pathways and whether these changes are associated with modulation of GSH synthesis. The liver is unique in its metabolic activity and harbors many interrelated pathways. Apart from glucose supply to other organs, the liver has to maintain its own energy metabolism. In that regard, glucose has also been recognized as an important substrate for the hepatocellular TCA cycle.13–15 We focused our experiments on 2 enzymatic pathways related to the TCA cycle: (1) pyruvate carboxylase (PC), which delivers new carbons into the TCA cycle, therefore allowing the synthesis of amino acids (including GSH); and (2) pyruvate dehydrogenase (PDH), a key enzyme for hepatocellular mitochondrial energy metabolism by acetyl-CoA supply.13–17
We investigated liver extracts by high-resolution multinuclear (1H, 31P, 13C) nuclear magnetic resonance (NMR) spectroscopy as an ideal method to identify and quantify metabolites without preselecting individual metabolites or prior separation.18–20 Furthermore, the use of 13C-labeled glucose allows the identification of several metabolic pathways simultaneously, such as PDH and PC.13–17 To identify liver metabolic alterations caused by NAC, and whether these are associated with GSH synthesis, normal mice were treated with NAC or other GSH precursors. We also investigated the association of these metabolic changes with the hepatoprotection afforded by NAC in experimental liver injuries caused by (a) tert-butylhydroperoxide (t-BHP)-induced oxidative stress, and (b) 3-nitropropionic acid (3-NPA)-induced mitochondrial impairment.
Male BALB/c mice (22-25 g) were injected intraperitoneally with NAC or other compounds. [U-13C]glucose (500 mg/kg; 2.78 mmol/kg) was injected intravenously in bolus to study metabolic changes in awake mice. Using 500 mg/kg [U-13C]glucose, plasma glucose was <10 mmol/L in all experiments. To investigate the immediate metabolic effects, the animals were killed 45 minutes after injection of the study substances and [U-13C]glucose; additional experiments were done to investigate time-dependent effects of NAC (45 minutes to 4 hours). The experiments were performed in fed mice; [U-13C]glucose was injected in the daytime between 11:00 AM and 1:00 PM. Comparison with other 13C-labeled substrates showed that [U-13C]glucose is favorably suited for the hepatocellular metabolic pathways explored in the current study.21 Additional experiments were done using [2-13C]pyruvate and [3-13C]oxaloacetate (2.78 mmol/kg) as substrates.
(1) (a) Mice were injected with increasing doses of NAC (150-1,200 mg/kg) together with [U-13C]glucose to determine its dose-dependent effects. Controls received an equivalent volume of 0.9% saline. (b) Cys and methionine (Met) (other GSH precursors) as well as hypotaurine (HTau) were injected in amounts equimolar to the 300 mg/kg NAC dose (1.47 mmol/kg). (c) To investigate whether the actions of NAC on [U-13C]glucose metabolism through PDH are independent from GSH synthesis, buthionine sulfoximine (BSO, 450 mg/kg) was used to decrease incorporation of the 13C-label into GSH.
(2) (a) To establish whether NAC exerts its hepatoprotective effect by changes in metabolic pathways independently from replenishment of the antioxidant GSH, we tested an oxidative stress-induced model of liver injury induced by t-BHP (50 mg/kg, intraperitoneally). (b) To investigate whether NAC caused similar changes after mitochondrial impairment, we used 3-NPA (50 mg/kg, intraperitoneally), which inhibits succinate dehydrogenase, a rate-limiting enzyme of the TCA cycle. Mice were killed 5 hours after administration of t-BHP or 3-NPA, whereas NAC or the other GSH precursors were injected 45 minutes before t-BHP or 3-NPA. [U-13C]glucose was injected 45 minutes before sacrifice.
The mice were killed by cervical dislocation and the livers freeze-clamped immediately. Blood was taken after section of the carotid artery and put into heparin tubes. Two hundred microliters blood was mixed with 100 μL 20% perchloric acid (PCA) and extracted for 1H-NMR analysis.22, 23 Another 200 μL blood was used for the measurement of serum alanine aminotransferase (ALT) with a biochemical multianalyzer in our Biochemistry Department. All animals received humane care according to the criteria outlined in the “Guide for the Care and Use of Laboratory Animals” prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH publication 6-23 [revised], 1985). The scientific protocol was approved by the Scientific Evaluation Committee of the Centre de Recherche, CHUM.
Tissue Extraction and NMR Spectroscopy.
Water-soluble metabolites and membrane components were extracted by a dual-extraction method with perchloric acid and methanol/chloroform.24 The lyophilized water-soluble samples were dissolved in D2O, centrifuged, and adjusted to pH 7.2. For 31P-NMR measurements, the samples were treated with 100 mmol/L ethylenediamine tetra-acetic acid and adjusted to pH 7.2. Lipid samples were dissolved in CDCl3/CD3OD (2:1). NMR spectra were recorded on DRX-600 or WB-360/NB-360 Bruker spectrometers. 1H-NMR spectra were recorded with a 5-mm H,C,N-inverse-triple-resonance probe, flip angle 40 degrees, repetition time 15 seconds, spectral width 7,183 Hz. 13C-NMR spectra were recorded with a 5-mm 1H/13C dual probe, repetition time 2.5 seconds, flip angle 27 degrees, composite pulse decoupling with WALTZ-16, spectral width 47,619 Hz. 31P-NMR spectra were recorded with a 5-mm QNP probe, flip angle 80 degrees, repetition time 3.5 seconds, spectral width 5,155 Hz, composite pulse decoupling with WALTZ-16. Metabolites were identified by gradient-selected 2-dimensional NMR homonuclear and heteronuclear correlations.
Quantification of Metabolite Concentrations.
Total metabolite concentrations were analyzed from 1H-NMR spectra of PCA extracts using (trimethylsilyl)propionic-2,2,3,3d4-acid as external standard.25 The ratio of poly-/monounsaturated fatty acids (PUFAs/MUFAs) was analyzed from 1H-NMR spectra of lipid extracts.26 Adenosine triphosphate (ATP) concentrations were quantified from31 P-NMR spectra of PCA extracts, using hexamethylphosphoric acid triamide as external standard (31.5 ppm).
Calculation of 13C-Labeled Metabolites.
Fractional 13C-enrichments (percentage 13C-labeling) of specific amino acid carbons were derived from 13C-NMR spectra.22, 23, 27 Because the use of other, unlabeled, substrates can lead to dilution of the fractional 13C-enrichment in TCA cycle–related, [U-13C]glucose-derived, compounds, the total amounts of 13C-labeled metabolites (nmol/g wet weight) were also calculated. The integrals of the signals were corrected for natural abundant 13C, nuclear Overhauser enhancement and saturation effects.27, 28
Labeling of Metabolites with 13C.
[U-13C]glucose allows determination of metabolic pathways used to synthesize downstream compounds.13–17, 27 Briefly, via glycolysis, 1 molecule [U-13C]glucose is transformed into 2 molecules [U-13C]pyruvate. [U-13C]pyruvate is transformed via PDH to [1,2-13C]acetyl-CoA, or is carboxylated via PC to [1,2,3-13C]oxaloacetate, which condenses with acetyl-CoA into citrate. Glucose flux through these 2 pathways can be quantified from 13C-NMR spectra from the 13C-labeling pattern in glutamate (Glu) synthesized from the TCA cycle intermediate α-ketoglutarate.13–17, 27 Glu is subsequently transformed to glutamine (Gln) or incorporated into the (γ-glutamyl)-residue of GSH. De novo synthesis of metabolites and associated metabolic pathways can be differentiated in a single 13C-NMR spectrum (Fig. 1 shows a small portion of a 13C-NMR spectrum demonstrating the resonances of the C-4 isotopomers synthesized through PDH). Using [2-13C]pyruvate, the 13C-label ends up in the C5 and C3 positions of Glu via PDH and PC, respectively.27 Using [3-13C]oxaloacetate, the 13C-label ends up in the C2 position of Glu.27
Each experiment was performed on at least 4 animals. The data are expressed as means ± SD. Statistical analysis was performed by 2-way ANOVA with the Tukey's method for posttest comparisons. A P value less than .05 was considered significant.
NAC Causes Selective Changes in the Concentration of GSH-Linked Amino Acids.
The total concentrations of Glu, Gln and GSH were quantified from 1H-NMR spectra of liver extracts after administration of increasing NAC doses to normal mice (Fig. 2A). NAC elevated GSH levels from 8.60 ± 0.48 μmol/g ww (controls) to a maximum of 12.95 ± 1.03 μmol/g ww at a 300 mg/kg dose (P < .05) (Fig. 3). In contrast, NAC (300 mg/kg) significantly decreased Gln from 6.98 ± 0.59 μmol/g ww to 4.83 ± 0.62 μmol/g ww (P < .05). Glu concentrations began to decrease with 300 mg/kg NAC (from 5.23 ± 0.49 μmol/g ww to 4.11 ± 0.50 μmol/g ww; P < .05) and showed no further significant decrease with increasing NAC doses (Fig. 3).
NAC Favors the Formation of HTau Instead of GSH.
NAC also induced the formation of a previously unassigned metabolite (Fig. 2A), identified as HTau through its characteristic triplets at 2.64 and 3.36 ppm and 2-dimensional NMR correlations (Fig. 2B). Like GSH, HTau can be synthesized from the Cys residue of NAC.29 HTau was found in very low concentrations in normal liver (0.05 ± 0.02 μmol/g ww), but increased considerably after administration of NAC and linearly up to 9.95 ± 1.22 μmol/g ww (Fig. 3). Total liver Tau concentrations (30.32 ± 3.27 μmol/g ww) were not affected by NAC.
NAC Inhibits Glucose Flux Through PC.
PC is a major anaplerotic enzyme that is required to replenish TCA cycle intermediates. In particular, to allow the synthesis of amino acids from TCA cycle intermediates, oxaloacetate has to be provided to condense with acetyl-CoA into citrate (Supplemental Fig. 1). 13C from [U-13C]glucose is incorporated into Glu, and thereafter into Gln or, together with the NAC-derived Cys residue, into GSH (Fig. 1). After treatment with NAC, the relative de novo synthesis of GSH via PC (fractional enrichment in [2,3-13C]GSH) decreased from 1.98% ± 0.30% to 0.87% ± 0.11% at a 300-mg/kg dose (P < .05; Fig. 4), and the total amounts of [2,3-13C]GSH declined from 169.76 ± 18.99 nmol/g ww to 112.45 ± 18.87 nmol/g ww (P < .05). This was accompanied with a diminution of the fractional 13C-enrichment in [2,3-13C]Glu (from 2.60% ± 0.19% to 1.24% ± 0.22% using 300 mg/kg NAC; P < .05).
NAC Increases Flux Through PDH.
The fractional 13C-enrichment in [4,5-13C]Glu is a measure of carbon flux through PDH (Supplemental Fig. 1).13–17, 27 NAC (300 mg/kg) caused increased flux through PDH (increased fractional 13C-enrichment in [4,5-13C]Glu from 2.08% ± 0.26% to 4.00 ± 0.44%; Fig. 5). Despite this augmented flux, the de novo synthesis of GSH via PDH only rose (but less than for Glu) with the 150 mg/kg NAC dose (increased fractional enrichment in [4,5-13C]GSH from 1.48% ± 0.14% to 1.87% ± 0.20%; P < .05), did not change significantly with 300 mg/kg NAC, and decreased at higher NAC doses to 0.61% ± 0.10% (P < .05).
NAC-Induced Changes in [2-13C]Pyruvate and [3-13C]Oxaloacetate Metabolism.
Because energy substrates may dilute the pool of 13C-labeled pyruvate derived from [U-13C]glucose, the precursor molecule for both PDH and PC, [2-13C]pyruvate, was used to investigate the effect of 300 mg/kg NAC, the administered dose showing most profound effects on liver Glu and GSH synthesis (Fig. 6A). These results clearly show that 300 mg/kg NAC decreased the fractional enrichments in [3-13C]Glu and [3-13C]GSH synthesized through PC from 2.98% ± 0.37% to 0.91% ± 1.29% and from 1.78% ± 0.21% to 0.48% ± 0.06%, respectively (P < .05), but increased PDH-mediated de novo synthesis of Glu ([5-13C]Glu) from 1.88 ± 0.14 to 3.54% ± 0.39% (P < .05). The fractional enrichments in [5-13C]GSH, however, were not different after NAC treatment (1.04% ± 0.16%) compared with controls (1.02% ± 0.13%). Using [3-13C]oxaloacetate, that does not require PC for its Krebs cycle metabolism, the fractional 13C-enrichments in GSH (0.90% ± 0.18%) and Glu (1.25% ± 0.17%) did not change significantly with NAC (Fig. 6B).
Time-Dependent Changes in Concentrations and De Novo Synthesis of GSH and Glu by NAC.
Compared with a single dose of 300 mg/kg NAC, the same dose given in 1-hour intervals caused similar depletion in liver Glu levels and slightly increased GSH levels (P < .05; Table 1). The impaired synthesis of Glu via PC also occurred after repeated NAC administration (decreased fractional 13C-enrichment in [2,3-13C]Glu from 2.60% ± 0.19% to 1.53% ± 0.23%; P < .05). PC-mediated GSH synthesis was not detectable. However, the increases in the fractional enrichments in [4,5-13C]Glu and [4,5-13C]GSH (through PDH) from 2.08% ± 0.26% to 4.88% ± 0.51% and from 1.48% ± 0.14% to 1.87% ± 0.24%, respectively, were higher compared with a single dose of 300 mg/kg NAC (P < .05). The flux through PC contributing to both Glu and GSH synthesis was slightly decreased 4 hours after NAC administration (P < .05), whereas increased fluxes through PDH were less pronounced after prolonged (2- or 4-hour) NAC exposure (P < .05 compared with its effects after 45 minutes).
Table 1. Concentrations and De Novo Synthesis of Glutathione and Glutamate After Repeated or Prolonged Treatment With NAC
Total Time of NAC Exposure
GSH Concentration (μmol/g ww)
Glu Concentration (μmol/g ww)
[2,3-13C]GSH Fractional 13C-Enrichment
[2,3-13C]Glu Fractional 13C-Enrichment
[4,5-13C]GSH Fractional 13C-Enrichment
[4,5-13C]Glu Fractional 13C-Enrichment
NOTE. The concentrations of glutathione (GSH) and glutamate (Glu) (μmol/g ww) as well as the fractional 13C-enrichments of 13C-labeled glutathione and glutamate synthesized after conversion of [U-13C]glucose via pyruvate carboxylase ([2,3-13C]isotopomers) and pyruvate dehydrogenase ([4,5-13C]isotopomers) of mouse liver extracts. The mice were treated with 300 mg/kg NAC intraperitoneally for 45 minutes, or for 3 times in 1-hour intervals with 100 mg/kg NAC, or for 2 or 4 hours with 300 mg/kg NAC. [U-13C]glucose ([U-13C]Glc was administeed 45 minutes before killing. Each experiment was performed on at least 4 animals. The data are expressed as means ± SD.
P < .05 versus saline-treated controls.
P < .05: statistically significant compared with the single administration of 300 mg/kg NAC for 45 minutes.
Effect of Other Cys-Related Compounds on GSH and HTau Levels.
To evaluate whether the effects of NAC on GSH synthesis and flux through PDH are linked, we compared the actions of other GSH precursors and of a known GSH-depleting agent, BSO. Cys and Met elevated GSH to levels at least equivalent to those observed after NAC treatment (Table 2, P < .05). HTau, conversely, did not significantly affect GSH levels. BSO decreased liver GSH from 8.60 ± 0.48 μmol/g ww to 5.18 ± 0.88 μmol/g ww (P < .05); however, it had no impact on basal HTau levels, and did not prevent the increase in HTau caused by NAC. Cys elevated HTau to levels similar to those reached with NAC (from 0.05 ± 0.02 μmol/g ww to 6.56 ± 0.62 μmol/g ww), whereas Met led to a significantly lower increase in HTau (to 2.09 ± 0.39 μmol/g ww).
Table 2. Tissue Concentrations of Amino Acids
NOTE. Tissue concentrations of glutathione and hypotaurine in mouse liver extracts (μmol/g ww), as calculated from their resonances in 1H-NMR spectra of liver extracts (Fig. 2) using an external standard. The mice were treated with 300 mg/kg NAC or other compounds related to cysteine metabolism intraperitoneally for 45 minutes, and then the freeze-clamped livers were extracted. The doses of hypotaurine, cysteine, and methionine were each 1.47 mmol/kg (corresponding to 300 mg/kg NAC). Each experiment was performed on at least 4 animals. The data are expressed as means ± SD.
Changes in GSH Synthesis and Mitochondrial Energy Metabolism (PDH) Can Be Dissociated.
We then evaluated the effect of each of the previously mentioned compounds on GSH synthesis and mitochondrial energy metabolism (Supplemental Fig. 2). Although the fractional 13C-enrichment in [4,5-13C]GSH was not affected by NAC (300 mg/kg), it increased significantly with Cys and Met from 1.48 ± 0.14 (controls) to 1.98$ ± 2.57% and to 2.26% ± 2.71%, respectively (P < .05). The rise in PDH-mediated [4,5-13C]Glu synthesis was observed only with NAC and Met (increases from 2.08% ± 0.26% to 3.99% ± 0.62% and to 3.74% ± 5.21%, respectively). These results indicate that changes in GSH synthesis and fluxes through PDH could occur independently. Furthermore, NAC could stimulate PDH-mediated Glu synthesis despite BSO-mediated decreased GSH synthesis.
NAC Is Hepatoprotective via Metabolic Changes Independent of GSH Replenishment.
To assess whether the metabolic effects of NAC, indicated to be independent of GSH synthesis, could play a role in the hepatoprotection by NAC, we tested a model of oxidative stress-induced liver injury caused by t-BHP (50 mg/kg, 5 hours; Fig. 7). Experimental procedures were chosen to obtain moderate oxidative stress and subsequent energy failure.
t-BHP caused liver oxidative stress, as evidenced by a decrease in the ratio of PUFAs/MUFA from 1.61 ± 0.22 to 0.93 ± 0.17 (P < .05) as calculated from 1H-NMR spectra of lipid extracts. t-BHP significantly reduced concentrations of GSH (from 8.42 ± 0.56 μmol/g ww to 3.95 ± 4.23 μmol/g ww) and of Tau (from 32.36 ± 4.02 μmol/g ww to 19.74 ± 22.90 μmol/g ww) (P < .05, Fig. 7). NAC attenuated the decrease in PUFA/MUFA by 27% (ratio of 1.18 ± 0.22; P < .05) and elevated liver GSH levels by 192% to 16.16 ± 23.85 μmol/g ww (P < .05) in t-BHP–treated mice. t-BHP significantly impaired PDH flux (decreased fractional enrichment in [4,5-13C]Glu from 2.17% ± 0.31% to 1.29% ± 0.16%; P < .05). Because PDH-mediated carbon entry into the TCA cycle might be coupled with energy production, we evaluated liver ATP concentrations. As expected, t-BHP reduced ATP from 3.05 ± 0.57 μmol/g ww to 1.48 ± 2.6 μmol/g ww (P < .05). NAC prevented the decrease in both PDH flux and ATP levels (P < .05 compared with t-BHP treatment alone). This was associated with partial prevention of increased serum ALT (P < .05).
Conversely, Cys alone did not prevent the impaired PDH flux or the elevated serum ALT levels. This was observed despite adequate GSH replenishment and partial prevention of the decrease in PUFA/MUFA (P < .05). Met improved t-BHP–induced ATP depletion, liver injury (P < .05), as well as the impaired PDH flux. However, the effect was less prominent than that observed with NAC (P < .05). HTau improved the decrease in PUFA/MUFA (ratio of 1.37 ± 0.18 instead of 0.93 ± 0.17; P < .05) and attenuated GSH depletion above control levels (by 23.1% to 10.37 ± 11.74 μmol/g ww; P < .05), but did not improve the decline in PDH flux or liver injury.
To further support the finding that NAC can protect the liver through its capacity to increase mitochondrial energy metabolism, similar experiments were conducted with 3-NPA (50 mg/kg, 5 hours). In contrast to t-BHP, 3-NPA induced oxidative stress and GSH depletion secondary to impaired mitochondrial metabolism.30 Like t-BHP, 3-NPA markedly decreased PDH flux (50.7% decreased fractional enrichment in [4,5-13C]Glu to 1.1% ± 0.17% (Supplemental Fig. 3). NAC significantly improved this effect (P < .05) and prevented ATP depletion (decrease from 3.05 ± 0.27 μmol/g ww to 2.50 ± 0.27 μmol/g ww vs. 1.13 ± 0.12 μmol/g ww after 3-NPA treatment without NAC) well as liver injury (P < .05). Conversely, NAC only partially prevented the decrease in GSH levels (decrease from 8.42 ± 0.56 μmol/g ww to 5.93 ± 0.66 μmol/g ww vs. 5.24 ± 0.57 μmol/g ww after 3-NPA treatment without NAC). Because oxidative stress is not the primary and immediate outcome of 3-NPA,30 we did not compare the effects of NAC with other Cys-delivering compounds.
The hepatoprotective effect of NAC is believed to be based on its ability to replenish GSH stores. However, it is also suggested that part of its actions, in particular in non–acetaminophen-induced liver injuries, might be attributable to changes in energy metabolism. Hence, we studied the immediate effect of NAC on specific hepatocellular metabolic pathways related to mitochondrial energy metabolism (PDH) and anaplerosis (PC) of the TCA cycle under normal conditions as well as in non–acetaminophen-induced experimental liver injuries in mice, and if these are associated with GSH replenishment.
Our data demonstrate that NAC has a limited capacity to synthesize GSH in the normal liver. Conversely, we observed a linear increase in liver HTau concentrations, which were barely detectable in the controls and reached levels similar to GSH. HTau can be synthesized directly from the Cys residue present in NAC. We therefore believe that the liver preferentially uses the Cys residue for HTau formation. HTau formation by NAC also occurred in the setting of liver injury induced by t-BHP. The formation of HTau has already been noted in the liver and in cultured hepatocytes.31, 32 HTau also has been shown to protect against cell injury.33, 34 We therefore assessed its capacity to protect against liver injury in the setting of oxidative stress. Despite a decreased PUFA/MUFA ratio, HTau did not prevent t-BHP liver injury, as obvious from increased serum ALT. More studies are required to reassess the hepatoprotective potential of this molecule in more detail. We cannot dismiss the possibililty that, by preferentially shunting Cys residues away from GSH formation, the liver is actually obviating part of the protective effect of NAC.
Several mechanisms may be responsible for the limited action of NAC on GSH synthesis. GSH synthesis is limited by the availability of Cys, the activity of γ-glutamylcysteine synthetase, and through feedback-inhibition.35 The formation of HTau shows that Cys was delivered to the liver in sufficient amounts. However, the levels of Glu, which is the first amino acid in the tripeptide GSH, were decreased 45 minutes after administration of 300 mg/kg NAC (Fig. 3). Furthermore, whereas the fractional 13C enrichment in Glu synthesized through PDH increased using 300 mg/kg NAC, Glu synthesized through PC from both [U-13C]glucose and [2-13C]pyruvate decreased considerably with NAC (Figs. 4–6A). Because no changes in GSH and Glu synthesis from [3-13C]oxaloacetate were observed (Fig. 6B), the limited de novo synthesis of GSH obviously occurred secondary to the reduced flux through PC rather than to the further metabolism of oxaloacetate through the Krebs cycle. Indeed, PC is the major anaplerotic enzyme required to replenish TCA cycle intermediates for amino acid biosynthesis.36 Decreased Glu also might be attributable to increased degradation or protein synthesis. However, using NMR-spectroscopic determination of fractional 13C-enrichments and total amounts of 13C in Glu, we were able to prove a reduction in the de novo formation of Glu from [U-13C]glucose through PC. Nevertheless, the inhibition of PC by NAC is surprising because acetyl-CoA, which can be formed from the acetyl group present in NAC, is an allosteric activator for PC.36 Possibly only part of the injected NAC is deacetylated before it enters the hepatocytes, or only a small part of the acetyl group of NAC is transformed to acetyl-CoA. The actual mechanism responsible for the inhibition of PC observed in our experiments remains unexplained.
Another mechanism of decreased Glu formation is through reduced Gln levels caused by NAC (Fig. 3). The liver normally synthesizes high amounts of Gln, which can be transformed to Glu via phosphate-activated glutaminase. Our observation therefore supports a previous report demonstrating that Gln is also rate-limiting for GSH replenishment.37, 38
Another striking result was that NAC increased flux through PDH, which is important for mitochondrial energy metabolism through its ability to provide acetyl-CoA for the mitochondrial TCA cycle (Supplemental Fig. 1).13–17 Although PDH has long been thought to be mainly useful for fatty acid synthesis, it is now recognized to be an important enzyme for hepatocellular mitochondrial energy metabolism.13–17 Furthermore, using [1-13C]glucose as hepatocellular substrate, 24% to 35% of total acetyl-CoA liver production was shown to occur through flux through PDH.13 The relative flux of pyruvate oxidation (which enters the TCA cycle after decarboxylation to acetyl-CoA via PDH) accounts for more than 65% of the entire hepatocellular TCA cycle flux.14 The early stimulation of PDH by NAC (within 45 minutes after administration) led us to design experiments to dissociate the effects of NAC on PDH and GSH synthesis. When Cys was delivered alone, GSH synthesis was elevated, but no increased flux through PDH was seen. Furthermore, co-treatment with BSO and NAC resulted in augmented PDH flux without heightened GSH synthesis. The mechanism by which NAC stimulates PDH-mediated energy metabolism also remains unknown. The acetyl group of NAC can be converted to acetyl-CoA. However, acetyl-CoA is a product of PDH and would inhibit this enzyme.36 At very high NAC doses (600 or 1,200 mg/kg), flux through PDH was unchanged or even decreased, presumably because of the much higher fall in PC, causing a deficiency in oxaloacetate. Thus, NAC indeed stimulates PDH, but this effect can be strongly limited by the availability of oxaloacetate (Supplementary Fig. 1) and flux through PC, which declined concentration-dependently with NAC. Considering the ability of the liver to use a variety of substrates and metabolic pathways to maintain its own energy metabolism, one could hypothesize that the deployment of other substrates to provide TCA cycle intermediates might allow for condensation with PDH-derived acetyl-CoA.
Finally, the significance of the increased flux through PDH, caused immediately after NAC administration, was demonstrated with t-BHP treatment. In particular, under oxidative stress, 45 minutes' pre-treatment with NAC efficiently protected against liver injury in parallel with its effect on PDH flux and maintenance of ATP levels, demonstrating its capacity to restore or prevent mitochondrial impairments. Despite GSH replenishment and prevention of oxidative stress, Cys did not protect against liver injury and energy failure, further confirming that the effects of NAC can occur independently from increased GSH stores (Fig. 7). Finally, NAC also prevented mitochondrial energy failure and improved liver injury, while having a less significant impact on GSH stores in mice receiving 3-NPA (Supplemental Fig. 3), a hepatotoxin known to induce rapid mitochondrial energy impairment.30
Our study shows that GSH synthesis by NAC is limited and that the hepatoprotective and immediate effects of NAC can occur independently from GSH replenishment (Fig. 8). Specifically, NAC strongly favors the formation of HTau, whereas its ability to replenish GSH is limited by the inhibition of flux through PC. Furthermore, apart from other, yet unknown effects, NAC might exert its hepatoprotective effect by improving mitochondrial energy metabolism. In light of the wide variety of pathophysiological mechanisms of liver injury, the metabolic changes caused by NAC might be either beneficial or deleterious. Although the detailed mechanisms underlying NAC's metabolic actions remain unknown, we believe that our observations could be important for the determination of NAC's therapeutic value and the decision for its wider use. Furthermore, these experiments point to the value of further investigating the hepatoprotective effects of structurally modified compounds that would have a better capacity to improve GSH synthesis and maintain mitochondrial energy metabolism.