Lipopolysaccharide-Induced Mouse Model of

Hepatology Institute, Peking University People's Hospital, Beijing 100044, China, and National Center of. Biomedical Analysis, Beijing 100039, China. ...
1 downloads 0 Views 498KB Size
Metabolic Profiling Analysis of a D-Galactosamine/ Lipopolysaccharide-Induced Mouse Model of Fulminant Hepatic Failure Bo Feng,†,# Shengming Wu,‡,# Sa Lv,† Feng Liu,† Hongsong Chen,† Xianzhong Yan,‡ Yu Li,‡ Fangting Dong,*,‡ and Lai Wei*,† Hepatology Institute, Peking University People’s Hospital, Beijing 100044, China, and National Center of Biomedical Analysis, Beijing 100039, China Received November 28, 2006

The purpose of this study was to characterize the changes in metabolic intermediates and to investigate the metabolic profile of a mouse model of fulminant hepatic failure (FHF), induced by D-galactosamine/ lipopolysaccharide (GalN/LPS). Plasma metabolite levels were detected using gas chromatography/ time-of-flight mass spectrometry, and the acquired data were transferred into Simca-P and processed using principal components analysis (PCA). In total, 45 metabolites were identified from the 267 distinct compounds found in the study. Whereas significant differences were noted in the plasma levels of the control and FHF groups, no differences in gluconeogenesis or glycolysis were noted following GalN/ LPS treatment. Our data also suggest that the production of ketone bodies, and the tricarboxylic acid and urea cycles, was inhibited. PCA data suggest that 5-hydroxyindoleacetic acid, glucose, β-hydroxybutyrate, and phosphate parameters had the highest weights on each of the principal components, and that they were the most important metabolites contributing to the separation of groups. In conclusion, this metabonomic approach can be used as a powerful tool to characterize changes in metabolic intermediates and to search for metabolic markers under certain pathophysiological conditions, such as FHF. Our data also demonstrate that a combination of 5-hydroxyindoleacetic acid, glucose, β-hydroxybutyrate, and phosphate concentrations in the plasma is a potential marker for FHF, as well as for the early prognosis of FHF. Keywords: metabonomics • GC/MS • fulminant hepatic failure

Introduction Fulminant hepatic failure (FHF), a liver disorder with often devastating consequences, is one of the most challenging gastrointestinal emergencies encountered in clinical practice. It encompasses a pattern of clinical symptoms and pathophysiological responses associated with the rapid deterioration of normal hepatic function.1 The most widely accepted definition includes evidence of coagulation abnormalities, and any degree of mental deterioration (encephalopathy) in a patient with no previous history of liver disease.2 The most prominent causes include drug-induced liver injury, viral hepatitis, autoimmune liver disease, and shock or hypoperfusion, and some cases have no discernible cause.3 Acetaminophen hepatotoxicity far exceeds other causes of FHF in the United States4,5 and often entails high morbidity and mortality. Currently, the * Corresponding authors. Lai Wei, MD, Professor, No.11 Xizhimen South Street, Beijing 100044, China. Telephone, 8610-68314422 ext. 5730; fax, 861068322662; e-mail, [email protected]. Fangting Dong, No.27 Taiping Road, Beijing 100039, China. Telephone, 8610-86538173; fax, 8610-68186281; e-mail: [email protected]. † Peking University People’s Hospital. ‡ National Center of Biomedical Analysis. # Authors contrubuted equally to this work. 10.1021/pr0606326 CCC: $37.00

 2007 American Chemical Society

overall short-term survival with transplantation is greater than 65%, compared with the lower survival rate of 15% before transplantation.2 Because it is rare, clinical investigations of FHF have been difficult to perform in depth, with very few controlled therapeutic trials performed. The liver plays a central role in the metabolism of amino acids, fatty acids, and carbohydrates, among other compounds. Metabolic changes in the liver can result from exogenous stimuli, as well as pathological states, such as FHF. Yokoyama and Arai reported that metabolic changes resulted in reduced amino acid uptake and a switch from gluconeogenesis to glycolysis in an isolated perfused liver system.6,7 Previous reports have also shown that hepatic encephalopathy, the main cause of death in FHF, is linked to a failure of the hepatic clearance of gut-derived substances (e.g., ammonia, γ-aminobutyric acid) and to changes in the amino acid metabolism that synthesizes neurotransmitters (e.g., phenylalamine, tyrosine-derived monoamines, and tryptophan-derived monoamines).8 Nutritional supplement solutions rich in branched-chain amino acids (BCAA) have been used to counteract the observed shifts in metabolite levels in the circulatory system.9 Whereas these formulas are often beneficial, some studies have failed to demonstrate positive effects.10 However, there have been no Journal of Proteome Research 2007, 6, 2161-2167

2161

Published on Web 05/12/2007

research articles previous reports, in global systems biology research, of the use of metabonomic techniques to investigate the metabolic profiles of mouse models of FHF. The systemic changes that occur in FHF reflect not only changes in liver function but also changes in whole-body metabolism. Consequently, a better understanding of the metabolic pathways directly affected during hepatic failure may provide a rational basis for the development of nutritional supports or other nonsurgical, conventional medicinal therapies. Metabonomics was originally defined by Nicholson as the quantitative measurement of the multiparametric metabolic responses of living systems to pathophysiological stimuli or genetic modifications.11 The use of metabonomics has increased in conjunction with complementary genomics and proteomics in various organisms, especially plants12,13 and microbes.14-16 Although there have been reports of the use of metabolic analysis of coronary heart disease,17 ovarian cancer,18 and other diseases,19 its use in clinical experiments and animal models of liver disease has been limited. In this study, a D-galactosamine/lipopolysaccharide (GalN/LPS)-induced FHF mouse model was established, and the plasma metabolite levels in both FHF and control mice were investigated using gas chromatography/time-of-flight mass spectrometry (GC/TOFMS). The acute hepatitis induced by GalN in mice results from the toxic effects associated with an insufficiency of uridine diphosphate (UDP)-glucose and UDP-galactose, the loss of intracellular calcium homeostasis, and the inhibition of hepatocyte energy metabolism. The activity of mitochondrial enzymes is dependent on lipoprotein interactions, and modifications induced by GalN may indirectly affect enzyme activity.20 Although GalN alone can induce FHF, the FHF mouse model generated by combining GalN with LPS is more consistent with that observed clinically in terms of the mechanisms of FHF, because LPS plays an important role in FHF patients. By characterizing the plasma metabonomics in this FHF model, we aimed (1) to demonstrate the use of GC/TOFMS for metabolic analysis of animal models and clinical diseases; (2) to investigate the metabolic markers and changes in the metabolic pathways of the FHF mice; and (3) to provide a rational basis for the development of nutritional support or other nonsurgical medical therapies for FHF.

Materials and Methods Animals. Animal studies were carried out in accordance with the Chinese National Research Council guidelines and were approved by the Subcommittee on Research Animal Care and Laboratory Animal Resources of the Peking University People’s Hospital. Male BALB/c mice (n ) 12, 18-22 g) were purchased from the Academy of Military Medical Sciences (Beijing, China). They were housed in a standard animal laboratory with a 12 h light-dark cycle and were provided with water and standard mouse chow ad libitum. The animals were randomly divided into GalN/LPS-induced FHF (n ) 6) and control (n ) 6) groups. Experimental FHF Model. The FHF model was established as described previously, with slight modifications.21 GalN and LPS (Sigma-Aldrich Chemical Co., Steinheim, Germany) were dissolved in saline before use. The mice were fasted for 12 h before treatment. The FHF group were intraperitoneally administered 0.6 g/kg GalN, followed by 8 µg/kg LPS. The control mice were administered saline twice, at the same volume as GalN and LPS. The GalN/LPS-treated mice developed severe FHF about 6 h after treatment. The mice were sacrificed for blood and liver recovery 6 h after treatment. Blood was 2162

Journal of Proteome Research • Vol. 6, No. 6, 2007

Feng et al.

collected in tubes containing 50 µL of EDTA and was immediately placed on ice, and the plasma was stored at -80 °C until analysis. The plasma was subjected to metabonomic analysis, and the liver samples were used for histopathological evaluation. Assessment of Liver Injury and Detection of Plasma Ammonia. Plasma levels of alanine aminotransferase (ALT) and T-bilirubin (TBIL), measured with a 7170A automatic analyzer (Hitachi, Japan), were used to assess the extent of liver injury. Plasma ammonia was detected with the Ammonia Checker II (Kyoto Daiichi Khgaku Co. Ltd, Kyoto, Japan). Histopathological Investigations. The liver was cut into small pieces and fixed in Bouin’s solution. Following dehydration in an ascending series of ethanol solutions (50%, 60%, 70%, and 75%), tissue samples were cleared in xylene, embedded in paraffin, and sliced into 5 mm sections, then stained with hematoxylin-eosin (H-E). Metabonomic Assays. Plasma samples (100 µL) were thawed before the immediate addition of 500 µL of methanol (100%) to stop enzymatic activity. The samples were vortexed thoroughly, and 20 µL of a ribitol stock solution (0.2 mg/mL H2O) was added as an internal reference. The mixture was placed on a shaker at 70 °C for 15 min and centrifuged at 10 000g for 10 min. The supernatant was mixed with 500 µL of pure water and 250 µL of chloroform, then centrifuged again at 4000 rpm for 15 min. The upper (polar) phase was separated and then evaporated to dryness under a stream of N2 gas in a thermostatically controlled water bath (60 °C). Metabolites in the plasma samples were derivatized prior to GC/TOFMS analysis, as previously described.22 Methoxyamine hydrochloride (20 µL, 20 mg/mL pyridine) was then added to the dried fraction of the polar phase. Following continuous shaking at 30 °C for 90 min, 40 µL of N-methyl-Ntrimethylsilyltrifluoroacetamide (MSTFA) was added, and the tube was incubated at 37 °C for 30 min. It was then kept at room temperature for 120 min, before injection (all chemicals were from Sigma-Aldrich Chemical Co. Steinheim, Germany). Solutions (0.3 µL) were injected at a split ratio of 25:1 into a GC/TOFMS system consisting of an HP 6890 gas chromatograph and a time-of-flight mass spectrometer (Waters Co., Milford, MA) on a 30 m DB-5 column (250 µm i.d., 0.25 µm film; Agilent Technologies, Palo Alto, CA). The injection temperature was 230 °C, the interface temperature was set to 290 °C, and the ion source temperature was adjusted to 220 °C, with an electron energy of 70 eV. Helium, the carrier gas, was set to a column flow rate of 1 mL/min. After a 5 min solvent delay time at 70 °C, the oven temperature was increased to 310 °C in increments of 5 °C/min, followed by a 1 min isocratic cool down to 70 °C and an additional 5 min delay. MassLynx software (Waters Co.) was used to acquire the chromatographs. NIST02 libraries with electron impact (EI) spectra were searched rigorously for all the peaks detected with the total ion current (TIC), to identify the metabolites. Compounds were also identified by comparison of their mass spectra and retention times with those of commercially available reference compounds. Data Processing and Pattern Recognition (Principal Component Analysis, PCA). All the collected data from the plasma samples were used for the analysis. Each sample was represented by a GC/TOFMS TIC chromatograph. Ribitol was added as an internal standard to correct for minor variations during sample preparation and analysis. The relative intensity of each

Metabolic Profile of a GalN/LPS-Induced Mouse Model of FHF Table 1. Assessment of Liver Injury and Detection of Plasma Ammonia parameters

FHF group

control group

P

ALT (U/L) 1879.56 ( 1020.54 45.25 ( 9.65 1.40E-04 TBIL (µmol/L) 5.00 ( 1.24 0.26 ( 0.15 2.13E-08 Ammonia (µmol/L) 137.80 ( 27.93 44.25 ( 7.63 3.58E-04

metabolite was expressed as 100 times the ratio of its peak area to that of ribitol on the same chromatograph. The data were then analyzed by principal components analysis (PCA), the most commonly used algorithm in metabonomics research.23 The raw GC/TOFMS data were initially processed using the MarkerLynx applications manager software (Waters Co.). This software incorporates a peak deconvolution package that allows the detection and retention-time alignment of the peaks eluting in each data file, across the whole data set. MarkerLynx extracts components using mass chromatograms and lists the detected peaks according to their masses and retention times, together with their associated intensities. PCA was performed after all the data were exported from MarkerLynx to SIMCA-P plus (Umetrics, Sweden). The simultaneous comparison of a large number of complex objects was facilitated by the reduction of the dimensionality of the data set with two-dimensional projection procedures. Statistical Analysis. Values are presented as means ( SD. Comparisons of the measured metabolite intensities of the control and FHF groups were made using the two-tailed Student’s t test. P < 0.05 was considered statistically significant.

Results Manifestations of the GalN/LPS-Induced Mouse Model of FHF. Almost all the BALB/c mice manifested retarded actions, slow reactions, and significantly tumescent livers, 6 h after GalN/LPS treatment. Whereas plasma levels of ALT, TBIL, and ammonia were within the normal ranges in the saline-treated control group, they were significantly increased to 40-, 20-, and 3-fold, respectively, in the GalN/LPS treatment group (Table 1). Liver inflammation, severe liver congestion, and hemorrhage were noted in the GalN/LPS-treated mice, compared with the normal liver morphologies observed in the controls (Figure 1). GC/TOFMS Analysis of Plasma Samples. The GC/TOFMS TIC chromatogram of plasma samples from the saline control and GalN/LPS treatment groups is shown in Figure 2. On the basis of the mass spectral database, 267 peaks were detected, 48 of which were confirmed to be endogenous metabolites, such as sugars, amino acids, fatty acids, and organic acids, among other compounds. Of the 48 metabolites, 45 were identified, whereas the others were unidentified metabolites in which significant differences were noted between the FHF group and the controls. Because these substances have been implicated in multiple biochemical processes, including energy and substance metabolism, the chromatograms are considered to be chemical fingerprint representations of endogenous metabolites, describing the metabolic changes induced by GalN/LPS. Significant differences between the TIC profiles of the control and treatment groups were observed (Figure 2), suggesting that endogenous metabolite levels fluctuated following GalN/LPS treatment. The metabolites were divided into the amino acid group and the nonamino-acid group for further analysis (Table 2 and Table 3). Marked changes in plasma amino acids were observed on GC/TOFMS chromatograms following GalN/LPS treatment.

research articles Whereas 16 amino acids were identified in total, amino acid concentrations were generally elevated in the FHF group, with significant differences in the levels of 15 amino acids (alanine, valine, isoleucine, glycine, serine, threonine, proline, leucine, aspartate, phenylalanine, glutamine, glutamate, ornithine, lysine, and tyrosine), with the exception of tryptophan, in which the increase was not statistically significant. The concentrations of total amino acids and gluconeogenic amino acids also increased. Furthermore, the ratio of BCAA to AAA (valine + leucine + isoleucine)/(tyrosine + phenylalanine + tryptophan) was significantly reduced in the FHF group (P < 0.05). Compared with the control group, β-hydroxybutyrate (HB), fumarate, aminomalonic acid, malate, uric acid, and naphaleneacetic acid were elevated in the treatment group. Interestingly, levels of enolpyruvate (EP), succinate, and tetradecanoic acid decreased significantly among the nonamino-organic acids identified in FHF mice. Although there was a noted decrease in galactose and D-altrose in these mice, no significant differences in other carbohydrates, including glucose, were noted. Elevated phosphate, γ-aminobutyric acid (GABA), 5-hydroxytryptamine (5-HT), and 5-hydroxyindoleacetic acid (5-HIAA) levels were also observed after GalN/LPS treatment. Levels of urea, palmic acid, and stearic acid were similar in both groups (Table 3). Effects of GalN/LPS-Induced FHF on Metabolic Pathways. Putative metabolic pathways were deduced in the FHF mice (Figure 3) based on changes in the levels of intermediates during substance metabolism, and the mechanisms underlying GalN/LPS damage to the liver were inferred. The pathways included gluconeogenesis, glycolysis, production of ketone bodies, the tricarboxylic acid (TCA) and urea cycles, among others. Whereas no significant differences between the fasted control and treated mice were noted in the gluconeogenic and glycolytic pathways, the production of ketone bodies, the TCA cycle, and the urea cycle were inhibited in the FHF mice. Although the levels of almost all amino acids were elevated in these mice, only a minority (e.g., aspartate, phenylalanine, and tyrosine) contributed to gluconeogenesis. Principal Components Analysis. PCA is a multivariate statistical method by which a data set with hundreds of variables can be represented by a few new variables (called “principal components”), which are linear combinations of the original variables, without the loss of much of the information contained in the original data set. All the GC/TOFMS data for the plasma samples of the control and FHF groups were analyzed. The first two principal components (PC1 and PC2) were calculated for the control and FHF groups. Although a greater number of components could have been used, these two were found to be sufficient to account for most of the variation in the data. A scores plot was used to represent the sample distribution in the new multivariate space, on which distinct clustering was observed between the FHF group and the control group (Figure 4A). The corresponding “loadings plot” (Figure 4B) represents the impact of each metabolite on this clustering. Analysis with SIMCA-P software indicated that the PC1 coordinate mainly comprised the concentrations of 5-HIAA, glucose, HB, phosphate, and lactate, whereas the PC2 coordinate mainly comprised the concentrations of 5-HIAA, phosphate, glucose, lactate, and HB. Our data suggest that 5-HIAA, glucose, HB, and phosphate are the parameters with the highest weights, and thus they had the greatest impact on each of the principal components in each experimental group. Journal of Proteome Research • Vol. 6, No. 6, 2007 2163

research articles

Feng et al.

Figure 1. Histopathological comparison of the FHF group and controls. Compared with the normal liver morphology in the controls (right panel), liver inflammation, severe liver congestion, and hemorrhage were observed in the GalN/LPS-treated mice (left panel). Slides were stained with hematoxylin-eosin (200×).

Figure 2. Comparison of GC/TOFMS total ion current (TIC) chromatographs of plasma from the FHF group (F) and controls (N). Each peak represents a metabolite, and the figures above the peaks represent their retention times.

Discussion Global systems biology has been receiving increasing attention in the postgenomic era. Whereas many efforts have been made to explore the relationships between genome, proteome, and phenotype in living systems, a complete understanding of the states of genes and proteins would not, in and of itself, reveal phenotype. Levels of metabolites represent integrative information on cellular function, and consequently, they define the phenotype of a cell or of an organism in response to genetic and environmental changes.24 Metabonomics was proposed as the quantitative measurement of the multiparametric metabolic responses of living systems to pathophysiological stimuli or genetic modifications. In this study, the metabolic profile of a GalN/LPS-induced mouse model of FHF was investigated using GC/TOFMS combined with multivariate statistical analysis. In metabonomic studies, GC/MS has been extensively applied to plants, despite the limitations entailed by volatile metabolites. To increase their thermal stability and to detect a greater number of metabolites, two-stage derivatization was used. First, carbonyl functional groups were converted to oximes with methoxyamine hydrochloride solutions, followed by the formation of trimethylsilyl (TMS) esters with a silylating reagent (MSTFA), to replace exchangeable protons with TMS groups.25 2164

Journal of Proteome Research • Vol. 6, No. 6, 2007

Patients with FHF typically exhibit elevated AAA and methionine plasma levels, normal or slightly increased BCAA levels, and an increase in the concentrations of a wide range of other amino acids.26 Increased severity of hepatic dysfunction also results in a decrease in the BCAA/AAA ratio, an important clinical parameter in FHF.27 In this study, similar changes in plasma amino acid levels were observed in GalN/LPS-treated mice, with significantly elevated levels of almost all the identified amino acid, including BCAA. Treatment with GalN/LPS inhibited the uptake and utilization of amino acids, especially AAA, whereas the degradation of proteins from the liver and other tissues increased.7 The inhibition of amino acid uptake, which is largely mediated by energy-dependent amino acid transporter systems, may have resulted from a decrease in hepatic adenosine triphosphate (ATP) levels in GalN-treated rats.28 Previous research has suggested a compensatory role for BCAA, particularly isoleucine and valine, in providing carbon skeletons as substrates for the TCA cycle.29 However, we believe that BCAA cannot be used in FHF because their plasma levels are elevated. In fact, there have been contradictory results regarding the use of amino acids, particularly BCAA, in FHF patients.30-33 Damage to hepatic functions can affect amino acid utilization and, in particular, protein synthesis. Systematic changes in metabolite levels reflect not only changes in liver

research articles

Metabolic Profile of a GalN/LPS-Induced Mouse Model of FHF

Table 2. Changes in Amino Acid Plasma Levels Observed on GC/TOFMS Chromatograms Following GalN/LPS Treatmenta no.

retention time

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

6.96 9.65 11.62 11.91 13.34 13.96 17.29 17.39 17.58 19.70 22.63 23.03 23.84 25.99 26.33 31.38

identified metabolites

control group

FHF group

P

alanine valine isoleucine glycine serine threonine proline leucine aspartate phenylalanine glutamate glutamine ornithine lysine tyrosine tryptophan Total amino acids Gluconeogenic amino acidsb BCAA AAA BCAA/AAA

3.3 ( 1.1 14.4 ( 10.6 11.3 ( 3.4 13.7 ( 6.4 9.8 ( 4.5 10.9 ( 6.0 14.9 ( 4.4 1.53 ( 0.82 19.9 ( 8.9 9.6 ( 4.1 1.48 ( 0.82 18.9 ( 5.2 1.4 ( 0.8 9.4 ( 3.6 5.1 ( 1.3 3.2 ( 1.8 159.48 ( 52.46 152.93 ( 49.56 35.37 ( 1.77 20.15 ( 7.65 2.25 ( 0.83

48.4 ( 14.5 33.1 ( 5.6 20.1 ( 3.0 50.2 ( 14.2 36.5 ( 8.2 37.7 ( 8.1 47.7 ( 10.2 6.51 ( 2.49 47.2 ( 7.4 26.9 ( 4.2 4.13 ( 1.39 33.7 ( 12.0 21.9 ( 11.8 46.9 ( 16.7 19.6 ( 8.1 4.4 ( 2.6 476.21 ( 111.15 442.94 ( 96.64 58.25 ( 8.14 55.57 ( 15.65 1.21 ( 0.35

1.84E-05 3.38E-03 7.22E-04 1.89E-04 3.75E-05 6.81E-05 3.11E-05 1.31E-03 1.89E-04 2.84E-05 2.44E-03 2.01E-02 1.71E-03 3.18E-04 1.49E-03 3.92E-01 1.25E-03 9.99E-04 9.57E-04 3.38E-02 3.61E-02

a Note: data are presented as means ( SD. b Gluconeogenic amino acids include aspartate, serine, glycine, glutamine, threonine, alanine, proline, lysine, tyrosine, phenylalanine, valine, and isoleucine.

Table 3. Changes in Plasma Levels of Other Metabolites on GC/TOFMS Chromatograms Following GalN/LPS Treatment no.

retention time

identified metabolites

control group

FHF group

P

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

6.04 6.39 6.62 7.69 8.31 10.75 11.18 13.17 13.80 16.05 16.55 19.04 20.40 20.59 22.75 23.94 24.55 24.74 25.44 25.79 26.17 26.68 27.16 27.35 28.48 28.95 29.09 29.44 32.05 33.51 34.05 35.11

lactate acetic acid EP ethanedioic acid β-hydroxybutyrate urea phosphate fumarate pyruvate aminomalonic acid malate succinate dodecanoic acid GABA R-phosphaglyceride citrate 1-deoxyglucose tetradecanoic acid glucose galatose glucitol D-gluconic acid D-altrose myoinositol palmic acid U42a U43a uric acid stearic acid U46a 5-HT 5-HIAA

209.2 ( 33.7 8.8 ( 5.5 2.9 ( 1.3 8.9 ( 5.8 134.8 ( 7.5 201.5 ( 53.9 228.1 ( 23.6 1.8 ( 0.8 5.4 ( 2.2 7.9 ( 4.6 5.9 ( 1.3 7.9 ( 3.0 4.3 ( 1.3 5.8 ( 2.3 4.2 ( 2.1 51.8 ( 18.7 24.6 ( 10.1 10.9 ( 3.9 131.5 ( 24.8 79.6 ( 6.56 39.1 ( 17.3 14.5 ( 5.4 119.6 ( 8.6 5.5 ( 2.1 48.5 ( 7.0 0.6 ( 0.2 39.9 ( 7.9 1.5 ( 0.8 69.5 ( 9.2 1.4 ( 0.8 8.3 ( 2.5 121.6 ( 36.2

230.0 ( 50.6 12.9 ( 4.7 1.5 ( 0.6 17.8 ( 8.1 62.4 ( 16.7 202.7 ( 85.9 366.0 ( 146.9 4.4 ( 2.4 4.8 ( 2.8 36.9 ( 3.0 16.2 ( 5.2 4.1 ( 1.7 3.9 ( 0.9 42.9 ( 16.5 35.0 ( 12.4 60.4 ( 14.4 41.1 ( 15.1 1.9 ( 1.2 155.8 ( 49.5 54.4 ( 26.1 55.8 ( 25.2 53.2 ( 10.5 84.8 ( 16.9 13.2 ( 3.5 50.1 ( 17.7 10.4 ( 2.0 72.1 ( 10.7 29.8 ( 21.0 67.2 ( 23.7 7.3 ( 2.6 51.9 ( 12.9 276.9 ( 53.2

4.22E-01 2.03E-01 3.66E-02 5.16E-02 2.12E-06 9.76E-01 4.66E-02 3.80E-02 6.90E-01 1.40E-07 9.09E-04 2.04E-02 5.50E-01 2.82E-04 1.31E-04 3.91E-01 5.11E-02 3.32E-04 3.07E-01 4.52E-02 2.10E-01 1.20E-05 1.19E-03 8.56E-04 8.39E-01 3.00E-07 1.49E-04 8.10E-03 8.32E-01 3.49E-04 1.05E-05 1.39E-02

a Represents unidentified metabolites in which significant differences were noted between the FHF group and controls.

function but also changes in the metabolism of a number of other tissues also affected by FHF. Mazziotti et al. reported that there were significant nonhepatic contributions to the increased circulating phenylalanine pool in FHF rats induced with GalN and suggested that one possibility was the muscle catabolism-dependent release of AAA in FHF.34 As reported previously,35 GABA and 5-HT, which are considered false neurotransmitters, increase significantly, and 5-HIAA derived

Figure 3. Effects of GalN/LPS-induced FHF on metabolic pathways. Putative metabolic pathways were inferred in the FHF mice from changes in the plasma levels of intermediates during substance metabolism, as was the mechanism of GalN/LPS damage to the liver. For clarity, amino acid degradation pathways are grouped according to their points of entry into the TCA cycle, gluconeogenesis, and ketogenesis pathways. Black arrows indicate the direction of the reaction assumed in the model. Red arrows indicate the reactions that are significantly elevated relative to those of the controls. Blue arrows show significantly inhibited reactions. G-3-P, glyceraldehyde 3-phosphate; PEP, phosphoenolpyruvate; EP, enolpyruvate; PG, R-phosphaglyceride; KGA, R-ketoglutarate; ACAC, acetoacetate; HB, β-hydroxybutyrate; AA, amino acids; GABA, γ-aminobutyric acid; 5-HT, 5-hydroxytryptamine. Amino acids are abbreviated using the standard three-letter convention.

from serotonin is also elevated. As one of the inhibitory neurotransmitters, 5-HIAA has been implicated in the pathogenesis of the neuropsychiatric symptoms of FHF. Journal of Proteome Research • Vol. 6, No. 6, 2007 2165

research articles

Feng et al.

Figure 4. Scores plot (A) and corresponding loadings plot (B) for PCA of GC/TOFMS data derived from all mice. The scores plot shows distinct clustering of the FHF group (F) and the controls (N). The loadings plot represents the impact of the metabolites on the clustering results. The ions most responsible for the variance in the scores plot are indicated on the loadings plot by their distance from the origin. The peaks are labeled according to their chromatographic retention times and m/z values.

It has been suggested that the arterial and hepatic venous ACAC/HB ratio decreases in FHF rats.7,8 The decreased ketone body ratio presumably reflects a decrease in the mitochondrial redox potential, or intramitochondrial NAD+/NADH, which inhibits reactions requiring NAD+ as a cofactor, such as those catalyzed by citrate synthase and R-ketoglutarate dehydrogenase (R-KGDH). This results in a decrease in the turnover rate of the TCA cycle and the entry of amino acid carbon backbones into the TCA cycle. In this study, ACAC was not identified, and thus, the ketone body ratio was not calculated. However, HB, one of the ketone bodies, was significantly decreased in the FHF group. Researchers have suggested that the injured liver acts as a major source of systemic lactate in the splanchnic circulation and that a decrease in gluconeogenesis and the eventual switch to glycolysis, concomitant with the inhibition of the TCA cycle and electron transport fluxes, are likely mechanisms for the increased hepatic lactate release during the evolution of FHF.36,37 In FHF, a greater proportion of ATP is progressively derived from glycolysis at the expense of mitochondrial oxidative phosphorylation.6 In patients with acetaminophen-induced liver failure, hyperlactatemia reflects the severity of liver injury from which the patient is unlikely to recover spontaneously.8,38 Consequently, blood lactate has been proposed as a prognostic marker and an early predictor of paracetamol-induced acute liver failure.39,40 However, in this study, the identified metabolites that are associated with glucose metabolism and the TCA cycle, fumarate and malate, were increased in the plasma. There was also a significant decrease in EP and succinate 2166

Journal of Proteome Research • Vol. 6, No. 6, 2007

plasma levels, whereas no differences were observed in glucose, pyruvate, or citrate. Lactate levels were also elevated, although not significantly. These findings may have resulted from the FHF model and the hepatotoxins used in this study, as well as from the time point chosen to kill the mice (6 h after GalN/ LPS treatment). Although there is a switch from gluconeogenesis to glycolysis as FHF progresses,6 in the earlier phase, the metabolism of the liver is likely to be in a gluconeogenic state, similar to that of the fasted control animals. This result is consistent with that of a previous study, in which FHF data collected after 1 and 4 h better fitted the gluconeogenic model in a GalN-induced FHF rat model, by perfusion of livers after the induction of FHF.6 Interestingly, plasma phosphate levels were significantly elevated in the FHF group. The FHF mice were killed at about 6 h after GalN/LPS treatment. At this time point, histopathological analysis indicated that severe liver congestion, hemorrhage, and necrosis were present in the GalN/LPS-treated mice, with no hepatocyte regeneration. Massive hepatocyte necrosis can cause an increased release of peripheral phosphate. Mice with the most severe liver injury may not have had a sufficient hepatocyte reserve to affect regeneration, and therefore, they did not consume phosphate for the formation of ATP and its high-energy intermediates. Furthermore, FHF is often concomitant with renal dysfunction, which may result in decreased urinary phosphate clearance and may contribute to the development of hyperphosphatemia. However, reports of the clinical significance of plasma phosphate levels in FHF have been contradictory.41-44

research articles

Metabolic Profile of a GalN/LPS-Induced Mouse Model of FHF

Interpretation of the metabolite level data is difficult, not only because biochemical pathways are linked and highly regulated but also because information is lost in the process of averaging. Analysis methods, such as PCA, reduce data complexity by focusing on the information content of a given data set, while maintaining most of the information in a few dimensions. In this study, a scores plot revealed distinct clustering between the FHF group and the controls. A corresponding loadings plot represented the impact of metabolites on the clustering results: the closer to 0, the smaller the influence of a linear metabolite combination. Our PCA data suggest that 5-HIAA is the most important parameter that distinguishes the control and FHF mice. Furthermore, rather than 5-HIAA alone, a combination of 5-HIAA, glucose, HB, and phosphate concentrations in the plasma could potentially be used to separate the control and FHF groups more accurately.

Conclusions A metabonomic method based on GC/TOFMS and a multivariate statistical technique has been successfully used to study FHF. Significant differences in the plasma levels of many metabolites were noted, compared with those of the controls, when a lethal model of GalN/LPS-induced FHF was used. Furthermore, the earliest metabolic perturbations detected included the inhibition of ketogenesis and the TCA and urea cycles, with no significant changes in the gluconeogenic and glycolic pathways. PCA data analysis suggests that a combination of 5-HIAA, glucose, HB, and phosphate concentrations in the plasma is a potential marker for FHF. Our results demonstrate that this metabonomic approach is a powerful tool with which to characterize changes in the levels of metabolic intermediates, and to facilitate the search for metabolic markers under certain physiopathological conditions. Markers linked to substance pathways may also be used for early prognosis of FHF. Abbreviations:FHF,fulminanthepaticfailure;BCAA,branchedchain amino acids; GalN, D-galactosamine; LPS, lipopolysaccharide; GC/TOFMS, gas chromatography/time-of-flight mass spectrometry; ALT, alanine transaminase; TBIL, T-bilirubin; MSTFA, N-methyl-N-trimethylsilyltrifluoroacetamide; PCA, principal components analysis; TIC, total ion current; AAA, aromatic amino acids; HB, β-hydroxybutyrate; EP, enolpyruvate; GABA, γ-aminobutyric acid; 5-HT, 5-hydroxytryptamine; 5-HIAA, 5-hydroxyindoleacetic acid; TCA, tricarboxylic acid; ATP, adenosine triphosphate; ACAC, acetoacetate; NADH, nicotinamide adenine dinucleotide.

Acknowledgment. Supported by the Major State Basic ResearchDevelopmentProgramofChina(GrantNo.2005CB522902) and the Beijing Science and Technology Program (Grant No. H030230280410). References (1) Sass, D. A.; Shakil, O. Liver Transpl. 2005, 11, 594-605. (2) Polson, J.; Lee, W. M. Hepatology 2005, 41, 1179-1197. (3) Ostapowicz, G. A.; Fontana, R. J.; Schiodt, F. V.; Larson, A.; Davern, T. J.; Han, S. H.; McCashland, T. M.; Shakil, A. O.; Hay, J. E.; Hynan, L.; Crippin, J. S.; Blei, A. T.; Samuel, G.; Reisch, J.; Lee, W. M. Ann. Intern. Med. 2002, 137, 947-954. (4) Liver transplantation worldwide. Liver Transpl. 2006, 12, 682686. (5) Larson, A. M.; Polson, J.; Fontana, R. J.; Davern, T. J.; Lalani, E.; Hynan, L. S.; Reisch, J. S.; Schiodt, F. V.; Ostapowicz, G.; Shakil, A. O.; Lee, W. M. Hepatology 2005, 42, 1364-1372. (6) Yokoyama, T.; Banta, S.; Berthiaume, F.; Nagrath, D.; Tompkins, R. G.; Yarmush, M. L. Metab. Eng. 2005, 7, 88-103.

(7) Arai, K.; Lee, K.; Berthiaume, F.; Tompkins, R. G.; Yarmush, M. L. Hepatology 2001, 34, 360-371. (8) Clemmesen, J. O.; Hoy, C. E.; Kondrup, J.; Ott, P. J. Hepatol. 2000, 33, 941-948. (9) Miwa, Y.; Kato, M.; Moriwaki, H.; Okuno, M.; Sugihara, J.; Ohnishi, H.; Yoshida, T.; Muto, Y.; Nakayama, M.; Morioka, Y. Hepatology 1995, 22, 291-296. (10) Mizock, B. A. Nutrition 1999, 15, 220-228. (11) Nicholson, J. K.; Lindon, J. C.; Holmes, E. Xenbiotica 1999, 11, 1181-1189. (12) Tohge, T.; Nishiyama, Y.; Hirai, M. Y.; Yano, M.; Nakajima, J.; Awazuhara, M.; Inoue, E.; Takahashi, H.; Goodenowe, D. B.; Kitayama, M.; Noji, M.; Yamazaki, M.; Saito, K. Plant J. 2005, 42, 218-235. (13) Sweetlove, L. J.; Fernie, A. R. New Phytol. 2005, 168, 9-24. (14) Villas-Boas, S. G.; Højer-Pedersen, J.; Akesson, M.; Smedsgaard, J.; Nielsen, J. Yeast 2005, 22, 1155-1169. (15) Devantier, R.; Scheithauer, B.; Villas-Boas, S. G.; Pedersen, S.; Olsson, L. Biotechnol. Bioeng. 2005, 90, 703-714. (16) Nasution, U.; van Gulik, W. M.; Kleijn, R. J.; van Winden, W. A.; Proell, A.; Heijnen, J. J. Biotechnol. Bioeng. 2006, 94, 159-166. (17) Brindle, J. T.; Antti, H.; Holmes, E.; Tranter, G.; Nicholson, J. K.; Bethell, H. W.; Clarke, S.; Schofield, P. M.; McKilligin, E.; Mosedale, D. E.; Grainger, D. J. Nat. Med. 2002, 8, 1439-1444. (18) Odunsi, K.; Wollman, R. M.; Ambrosone, C. B.; Hutson, A.; McCann, S. E.; Tammela, J.; Geisler, J. P.; Miller, G.; Sellers, T.; Cliby, W.; Qian, F.; Keitz, B.; Intengan, M.; Lele, S.; Alderfer, J. L. Int. J. Cancer 2005, 113, 782-788. (19) Shockcor, J. P.; Holmes, E. Curr. Top. Med. Chem. 2001, 14, 182191. (20) Ferencikova, R.; Cervinkova, Z.; Drahota, Z. Physiol. Res. 2003, 52, 73-78. (21) Arvelo, M. B.; Cooper, J. T.; Longo, C.; Daniel, S.; Grey, S. T.; Mahiou, J.; Czismadia, E.; Abu-Jawdeh, G.; Ferran, C. Hepatology 2002, 35, 535-543. (22) Fiehn, O.; Kopka, J.; Do¨rmann, P.; Altmann, T.; Trethewey, R. N.; Willmitzer, L. Nat. Biotech. 2000, 18, 1157-1161. (23) Anthony, M. L.; Rose, V. S.; Nicholson, J. K.; Lindon, J. C. J. Pharm. Biomed. Anal. 1995, 13, 205-211. (24) Villas-Boas, S. G.; Mas, S.; Akesson, M.; Smedsgaard, J.; Nielsen, J. Mass Spectrom. Rev. 2005, 24, 613-646. (25) Dunn, W. B.; Ellis, D. I. Trends Anal. Chem. 2005, 24, 285-294. (26) Zimmermann, C.; Ferenci, P.; Pifl, C.; Yurdaydin, C.; Ebner, J.; Lassmann, H.; Roth, E.; Hortnagl, H. Hepatology 1989, 9, 594601. (27) Katayama, K. Hepatol. Res. 2004, 30S, 71-78. (28) Maezono, K.; Mawatari, K.; Kajiwara, K.; Shinkai, A.; Maki, T. Hepatology 1996, 24, 1211-1216. (29) Ott, P.; Clemmesen, O.; Larsen, F. S. Neurochem. Int. 2005, 47, 13-18. (30) Rose, C.; Michalak, A.; Rao, K. V.; Quack, G.; Kircheis, G.; Butterworth, R. F. Hepatology 1999, 30, 636-640. (31) Maezono, K.; Kajiwara, K.; Matawatari, K.; Shinkai, A.; Torii, K.; Maki, T. Hepatology 1996, 24, 185-191. (32) Shimomura, Y.; Murakami, T.; Nagasaki, M.; Honda, T.; Goto, H.; Kotake, K.; Kurokawa, T.; Nonami, T. Hepatol. Res. 2004, 30S, 3-8. (33) Honda, T.; Fukuda, Y.; Nakano, I.; Katano, Y.; Goto, H.; Sato, Y.; Murakami, T.; Shimomura, Y. J. Hepatol. 2004, 40, 439-445. (34) Mazziotti, A.; Bernardi, M.; Antonini, L.; Dioguardi, F. S.; Bellusci, R.; Papa, V.; Tacconi, C.; Gasbarrini, G.; Cavallari, A.; Possati, L. Surgery 1981, 90, 527-534. (35) Michalak, A.; Chatauret, N.; Butterworth, R. F. Neurochem. Int. 2001, 38, 163-168. (36) Gore, D. C.; Jahoor, F.; Hibbert, J. M.; DeMaria, E. J. Ann. Surg. 1996, 224, 97-102. (37) James, J. H.; Luchette, F. A.; McCarter, F. D.; Fischer, J. E. Lancet 1999, 354, 505-508. (38) Murphy, N. D.; Kodakat, S. K.; Wendon, J. A.; Jooste, C. A.; Muiesan, P.; Rela, M.; Heaton, N. D. Crit. Care Med. 2001, 29, 2111-2118. (39) Bernal, W.; Donaldson, N.; Wyncoll, D.; Wendon, J. Lancet 2002, 359, 558-563. (40) Schmidt, L. E.; Larsen, F. S. Hepatology 2003, 37, 1199-1201. (41) Chung, P. Y.; Sitrin, M. D.; Te, H. S. Liver Transpl. 2003, 9, 248253. (42) Schmidt, L. E.; Dalhoff, K. Hepatology 2002, 36, 659-665. (43) Shakil, A. O. Liver Transpl. 2005, 11, 1028-1030. (44) MacQuillan, G. C.; Seyam, M. S.; Nightingale, P.; Neuberger, J. M.; Murphy, N. Liver Transpl. 2005, 11, 1073-1079.

PR0606326 Journal of Proteome Research • Vol. 6, No. 6, 2007 2167