Toxicology and applied pharmacology
- 格式：pdf
- 大小：1.30 MB
- 文档页数：10
A metabonomic evaluation of the monocrotaline-induced sinusoidal obstruction syndrome (SOS)in ratsR.Conotte 1,J.-M.Colet ⁎University of Mons —UMONS,Department of Human Biology &Toxicology,20Place du Parc,7000Mons,Belgiuma b s t r a c ta r t i c l e i n f o Article history:Received 14November 2013Revised 31January 2014Accepted 12February 2014Available online 25February 2014Keywords:Sinusoidal obstruction syndrome Metabonomic Monocrotaline 1H-NMRMultivariate data analysis HepatotoxicityThe main curative treatment of colorectal cancer remains the surgery.However,when metastases are suspected,surgery is followed by a preventive chemotherapy using oxaliplatin which,unfortunately,may cause liver sinu-soidal obstruction syndrome (SOS).Such hepatic damage is barely detected during or after chemotherapy due to a lack of effective diagnostic procedures,but liver biopsy.The primary objective of the present study was to iden-tify potential early diagnosis biomarkers of SOS using a metabonomic approach.SOS was induced in rats by monocrotaline,a prototypical toxic substance.1H NMR spectroscopy analysis of urine samples collected from rats treated with monocrotaline showed signi ficant metabolic changes as compared to controls.During a first phase,cellular protective mechanisms such as an increased synthesis of GSH (reduced taurine)and the recruit-ment of cell osmolytes in the liver (betaine)were seen.In the second phase,the disturbance of the urea cycle (increased ornithine and urea reduction)leading to the depletion of NO,the alteration in the GSH synthesis (increased creatine and GSH precursors (glutamate,dimethylglycine and sarcosine)),and the liver necrosis (decrease taurine and increase creatine)all indicate the development of SOS.©2014Elsevier Inc.All rights reserved.IntroductionColorectal cancer (CRC)is the second most frequently diagnosed cancer in Europe (Konopke et al.,2012).In the case of localized CRC,surgery remains the main curative treatment.However,in patients pre-senting aggressive tumors and risks of metastases,surgery is not suf fi-cient and must be followed by adjuvant chemotherapy to increase remission chances (Nelson et al.,2001).Metastases are the main cause of death in CRC patients,with liver as the main target organ of metasta-tic diseases.Approximately 20%of CRC patients will be diagnosed with metastases in the liver and 20to 30%will develop metastasis after primary tumor resection (Kawada et al.,2011).In the presence of re-sectable liver metastases,surgery is the therapy of choice.For patients presenting unresectable liver metastases though,neoadjuvant chemo-therapy and ablative techniques (i.e.,radio frequency)are used to reduce the size and number of hepatic metastases at levels whereresection becomes possible.This procedure signi ficantly increases the survival rate without relapse (Brouquet et al.,2011;Nordlinger et al.2008).However,these neoadjuvant chemotherapies can themselves lead to hepatic toxicity.For example,oxaliplatin is associated with im-portant hepatic sinusoidal obstruction syndrome (SOS)presenting sinu-soidal fibrosis,veno-occlusive lesions and occasionally with nodular regenerative hyperplasia in patients with a metastatic colorectal cancer (Rubbia-Brandt et al.,2004).This potential hepatotoxicity can lead to a higher risk of post-surgery complication and decreased liver regenera-tion after partial hepatectomy (Schiffer et al.,2009).This toxicity can limit the use and the duration of neoadjuvant chemotherapy and thus negatively impact the overall patient survival rate.Diagnosis of patients with SOS is de fined by the clinical syndrome of hepatomegaly,ascites and unexplained weight gain and elevated serum bilirubin (Chen and Huo,2010;DeLeve et al.,2002;Helmy,2006).How-ever,SOS diagnosis is dif ficult to establish because either clinical data can suggest a different liver pathology,or the timing of events can be un-usual,or the patient presents only one of the above-mentioned criteria (Helmy,2006).Various changes in the clinical chemistry parameters have been reported in SOS patients:increases in serum bilirubin,aspar-tate amino transferase (AST),von Willebrand factor or thrombomodulin levels;an increase in coagulation activation marker and procoagulants and a decrease in natural concentration of anticoagulant and in the von Willebrand factor protease activity (DeLeve et al.,2002;Helmy,2006).However,the use of those clinical parameters has proved unsuit-able as diagnostic or prognostic tools.The histopathological examination of liver biopsy remains the gold standard diagnostic tool of SOS (ChenToxicology and Applied Pharmacology 276(2014)147–156Abbreviations:CRC,Colorectal cancer;SOS,Sinusoidal obstruction syndrome;AST,Aspartate amino transferase;MCT,Monocrotaline;PA,Pyrrozolidine alkaloide;GSH,Glutathione;SEC,sinusoidal endothelial cell;1H NMR,Proton nuclear magnetic resonance;PCA,Principal component analysis;BW,Body weight;TSP,3-(Trimethylsilyl)propionic-2,2,3,3-d4acid;FID,Free induction decays;PLS-DA,Partial least square discriminant anal-ysis;TCA,Tricarboxylic acid cycle;AA,Amino acid;DMG,Dimethylglycine;KC,Kupffer cell;AST,Aspartate aminotransferase;MMP-9,Matrix metallopeptidase 9;NO,Nitric oxide.⁎Corresponding author at:University of Mons -UMONS Departement of Human Biology and Toxicology 20,Place du Parc 7000Mons,Belgium.Fax:+3265373526.E-mail address:jean-marie.colet@umons.ac.be (J.-M.Colet).1Fax:+3265373526./10.1016/j.taap.2014.02.0080041-008X/©2014Elsevier Inc.All rightsreserved.Contents lists available at ScienceDirectToxicology and Applied Pharmacologyj o u r na l ho m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /y t a a pand Huo,2010).In experimental research,SOS toxicity can be induced by the administration of monocrotaline(MCT),a pyrrozolidine alkaloide (PA)plant toxin.PA is constitutively expressed in several plant families such as Crotalaria genus which produce MCT and are toxic for the liver (Chojkier,2003;Copple et al.,2002).Humans may be exposed to MCT or other PA by consumption of herbal remedies,herbal teas of food grain product contaminated by PA plants(Chen and Huo,2010; Chojkier,2003;Copple et al.,2002).To induce hepatotoxicity,MCT requires bioactivation in the liver by cytochrome P450-3A(CYP3A)in dehydromonochrotalin,a pyrrolic derivative(Chen and Huo,2010; Copple et al.,2002).This pyrrole is an alkylating compound that is highly reactive to proteins and nucleic acids.When adducts of pyrrole to pro-teins or nucleic acids are formed,they can persist in tissues and induce chronic injury(Chen and Huo,2010).In SOS toxicity,damages to sinusoidal endothelial cell(SEC)are observed.This cellular type is more sensitive to MCT toxicity than hepatocytes due to lower levels of glutathione(GSH).GSH is indeed mandatory for MCT detoxification through the synthesis of a less toxic conjugate that can be excreted into the bile(Chen and Huo,2010;Copple et al.,2002).Acute toxicity is also accompanied with centrilobular hepatocellular necrosis,dilated and congested sinusoids and bleeding(Helmy,2006;Yee et al.,2003; DeLeve et al.,2002).Recently proposed as a potential new tool in preclinical toxicology studies,metabonomics has proved to be one of the most powerful predictive techniques for investigating the response of organisms to xe-nobiotics(Shockcor and Holmes,2002).Early work primarily focused on renal and hepatic toxicities for which characteristic metabolic signa-tures were reported(Nicholson et al.,2002;Robertson,2005;Shockcor and Holmes,2002).This technique uses spectroscopic analysis of biofluids such as urine,blood or saliva to determine alterations of the metabolome in the studied organisms.The metabolome is constituted by the entire set of small molecules(typically b1000Da)involved in the intermediate metabolism and called a“metabolite”.Metabonomics is defined as“the quantitative measurement of the time related multi-parametric metabolic response of living systems to pathophysiological stimuli or genetic modification”(Lindon et al.,2006;Nicholson and Lindon,2008).The spectroscopic methods include proton nuclear mag-netic resonance(1H NMR)and mass spectrometry(Beckonert et al., 2007;Lindon and Nicholson,2008).1H NMR spectroscopy is a non-destructive and fast technique with a minimal sample preparation to obtain rich molecular information.Such spectral profiles contain a bench of signals resulting from hundreds of endogenous metabolites in-volved in dozens of biochemical pathways(Hasim et al.,2012;Shockcor and Holmes,2002).The modulations in the1H NMR spectrum of extra-cellularfluids can be seen as a reflection of the metabolic changes occur-ring in cells exposed to a chemical substance or due to the development of a disease(Shockcor and Holmes,2002).The huge amount of data generated during the metabonomic studies requires multivariate analy-sis to extract pertinent information.Among these,principal component analysis(PCA)is applied to reduce a large number of spectral variables to a few new variables(principal components).After identification of variables responsible for the discrimination between groups,metabo-lites are identified according to their chemical shift and multiplicity (Robertson,2005).The main objective of the present study was to identify a metabonomic signature characteristic of MCT-induced SOS in rats and to isolate a series of potential early diagnostic biomarkers which could be further developed in clinics for the evaluation of SOS during chemotherapy.MethodsMale“Wistar Han”rats(Janvier Elevage,Le Genest-Saint-Isle, France)of7to9weeks old were individually housed in metabolism cages during the duration of the experiment for urine collection.They were maintained on a12h light/darkness cycle at controlled room temperature(21±2°C)and humidity(between40and60%).Ani-mals received30to35g/day of food and were allowed free access to water.Rats were assigned randomly into2groups of four animals each,in-cluding a control group(receiving the saline only)and one treated group(receiving monocrotaline in saline).The treated group received during4days a daily i.p.injection of monocrotaline at a dose of 100mg/kg in2ml of sterile saline vehicle.An equivalent volume of saline was given daily to control animals.Drug administration and blood sampling were performed on anes-thetized animals placed in induction chamberfilled with4.0%isoflurane at aflow rate of1.0l/min.During the i.p.injection and blood sampling, anesthesia was maintained by isoflurane(1.5at2.0%at aflow rate of 0.5l/min).One day before dosing,pre-test urine samples were collected. After drug administration,urines were collected daily from0to8 h and from8to24h during7days in refrigerated tubes containing 1ml of sodium azide1.0%per24-hour period.Urine samples were cen-trifuged at1600g for5min at4°C.Finally,urine samples were stored at −80°C.Before dosing and on Days4and7post-dosing,600μl of blood samples was collected from the tail vein for clinical chemistry.Blood samples were centrifuged at3000g for15min at4°C.Finally,serum was transferred into Eppendorf tubes and stored at−80°C.Livers were dissected at necropsy for further histopathological evaluation.400μl of urine samples was mixed with200μl of phosphate buff-er(0.2M Na2HPO4/0.04M NaH2PO4,pH7.4)prepared in a mixture H2O/D2O(80:20;v:v)to minimize pH variation.Then the samples were centrifuged at13,000g for10min.Supernatants were transferred into Eppendorf tubes,in which50μl of a12mM3-(Trimethylsilyl) propionic-2,2,3,3-d4acid(TSP)solution prepared in100%deuterium oxide was added.550μl of samples was transferred into5-mm NMR tube and NMR spectrum was recorded at302°K using a Bruker Avance400spectrom-eter at400MHz for proton observation(Pharmaceutical Chemistry and Radiopharmacy department at the Catholic University of Louvain; UCL–CMFA;Louvain-La-Neuve,Belgium).One-dimensional spectrum was acquired using a NOESYPRESAT-1d pulse sequence.For each sam-ple,64free induction decays(FIDs)with32,768data point per FID were collected using a spectral width of6410.2Hz,an acquisition time of2.5s and a pulse recycle delay of2s.The FID was Fourier transformed and a line broadening of0.3Hz was applied.The spectra were automat-ically phase-and baseline-corrected using MestReNova5.2.0software (Mestrelab Research,Santiago de Compostela,Spain).Spectra were calibrated against TSP placed at0.00ppm.For multivariate analysis purposes,the spectral region from0.08to 10.00ppm was automatically reduced to248integrated regions (buckets)with0.04ppm width each.The regions from 4.50to 5.00ppm and from5.50to6.00ppm were excluded from the analysis to remove the residual water signal and the daily urea variation.Each integrated region was normalized to the total spectrum area.Thefinal data set that consisted of222integrated reduced variables was imported into SIMCA-P+12.0software(Umetrics,Umeå,Sweden). After mean-centering of the data without scaling,multivariate analysis was conducted.Principal component analysis(PCA)and Partial least square discriminant analysis(PLS-DA)were carried out to discriminate the metabolic patterns between MCT treated rats and control.The quality and reliability of the models were described by R2and Q2param-eters.The goodness offit parameter R2represents the explained varia-tion in the data and the goodness of prediction parameter Q2uses cross-validation to estimate the predictive ability of the model.Clinical chemistry analysis of serum samples was carried out with an automatic biochemical analyzer(Spotchem EZ SP-4430,Menarini Diagnostics).SPOTCHEM II liver-1kit was used to measure lactate dehy-drogenase,alanine aminotransferase,aspartate aminotransferase(AST),148R.Conotte,J.-M.Colet/Toxicology and Applied Pharmacology276(2014)147–156albumin,total protein and total bilirubin.Differences observed between groups were evaluated using Wilcoxon tests in R software version 2.15.0.A p-value b 0.05was considered statistically signi ficant.ResultsDaily exposure to MCT (100mg/kg bw)resulted in a statistically sig-ni ficant decrease in body weight (body weight gain:12.3±6.7and −43.3±15.3g for controls and MCT-exposed animals,respectively at Day 7compared to Day 1)with a concomitant statistically signi ficant in-crease in relative liver weight (Fig.1)as compared to controls (3.9±0.3and 5.4±0.9%for controls and MCT-exposed animals,respectively).Two unexpected deaths occurred after drug administration in the group receiving MCT.One rat died on Day 5and one on Day 7.Statisti-cally signi ficant increases in AST were also noticed in animals treated with MCT as compared to controls (Table 1).The presence of a fluid in the intraperitoneal cavity was seen in two rats from the MCT group.SOS modelFig.2shows representative 400MHz 1D 1H NMR spectra of urine samples from a control rat and from a rat treated with MCT.These spec-tra were annotated with databases such as Human Metabolome Data-Base (HMDB)and according to reference tables developed in-house allowing the identi fication of the urine metabolites based on the chem-ical shift and the multiplicity of the resonances.Visual examination of 1H NMR spectra already revealed major differences in the urine compo-sition of controls and treated animals.To further investigate these differences in all animals and at all experimental time points,multivar-iate data analysis models were built using 1H NMR data.At first,only MCT-exposed animals were analyzed to build the predictive model of SOS.A principal component analysis (PCA)was performed using urine samples collected from MCT or control rats on pre-test to Day 7.Despite the unexpected death of a second animal on Day 7,we did obtain a com-plete urinary pro file for 3MCT-treated rats.The corresponding score plot showed a separate time evolution of the MCT data as compared to controls.In addition,a clear break in the metabolic path was observed between early and late time points following exposure to MCT (R 2X =0.803;Q 2=0.467;Fig.3).The first direction was formed by the urine samples collected during the first 3days (early stage),while the second direction corresponded to urines from days 4to 7(late stage).In order to better characterize the metabolic changes occurring during the early and the late phases,those data were reanalyzed separately.Metabolic changes during the early stageUnsupervised multivariate data analysis of control urines and urines collected on Days 1to 3of exposure was performed.1H NMR data not only allowed to discriminate urines of MCT-exposed animals from controls but,in the drug-exposed group,they also showed a clear time separation between the fractions collected during the periods from 0to 8h and from 8to 24h (data not shown).The PCA performed on the MCT data only (R 2X =0.850;Q 2=0.467)showed a clear separa-tion between the 0–8h and the 8–24h periods during the first 3days and this effect was mostly seen along the first principal component (Fig.4).The major spectral sub-regions responsible for that intra-group separation corresponded to a doublet at 1.20ppm and two sin-glets resonating at 1.37ppm and 1.48ppm,respectively.A comparison of those discriminating peaks with the 1H NMR spectrum of MCT published by Yang et al.,(2011),showed an obvious matching.Those urine changes were thus related to the urine excretion of MCT.Conse-quently,they were excluded for the subsequent analysis that still revealed (Fig.5)a clear separation between controls and rats receiving 100mg/kg bw of MCT (R 2X =0.897;Q 2=0.437).The corresponding loading plot showed an increase in the level of acetate (1.92ppm,singlet),trimethylamine (2.88ppm,singlet),ornithine (3.05ppm,triplet)and four unidenti fied metabolites (Table 2)in the urine of rats exposed to MCT as compared to controls.In parallel,decreases in the level of alpha-ketoglutarate (2.45ppm,triplet;3.01ppm,triplet),citrate (2.54ppm,doublet;2.67ppm,doublet),succinate (2.40ppm,singlet),hippurate (3.97ppm,doublet;7.55ppm,triplet;7.64ppm,triplet;7.84ppm,doublet),taurine (3.25ppm,triplet;3.43ppm,triplet),betaine (3.27ppm,singlet;3.90ppm,singlet),trimethylamine-N-oxide (3.27ppm,singlet),trans-aconitate (6.60ppm,singlet),aminohippurate (6.86ppm,doublet),choline (3.21ppm,singlet;4.02ppm,multiplet),and glucose (5.24ppm,doublet)were also observed (Table 2).Metabolic changes during late stageThe urine samples collected during the late period (study days 4to 7)from MCT-exposed and control rats were also submitted to principal component analysis.The corresponding score plot (Fig.6)showed a clear separation between both groups of rats (R 2X =0.935;Q 2=0.693).The main urine changes seen in MCT-exposed animals were increased levels of allantoate (5.26ppm,singlet),alanine (1.46ppm,doublet),acetate (1.92ppm,singlet),dimethylamine (2.72ppm,singlet)or sarcosine (2.72ppm,singlet),trimethylamine (2.88ppm,singlet),dimethylglycine (2.93ppm,singlet),ornithine (3.05ppm,triplet),n-acetylglutamate (2.04ppm,singlet),creatinine (3.05ppm,singlet;4.06ppm,singlet),creatine (3.04ppm,singlet;3.93ppm,singlet),leucine (0.96ppm,triplet),isoleucine (0.94ppm,triplet),glutamate (2.32ppm,multiplet)and decreased excretion of alpha-ketoglutarate (2.45ppm,triplet;3.01ppm,triplet),citrate (2.54ppm,doublet;2.67ppm,doublet),succinate (2.40ppm,singlet),hippurate (3.97ppm,doublet;7.55ppm,triplet;7.64ppm,triplet;7.84ppm,Fig.1.Relative weights of livers at necropsy (mean ±standard deviation).Blue =control animals receiving vehicle and green =animals treated with 100mg/kg bw of monocrotaline.Table 1Serum levels of aspartate aminotransferase (AST)in IU/L.AnimalControlMonocrotaline 96h168h 152152044482531092–3652756–4Not sampled 558501Mean56.71481.5a Standard deviation 7.2936.4–Animal death.aSigni ficantly different as compared to control rats (p b 0.05).149R.Conotte,J.-M.Colet /Toxicology and Applied Pharmacology 276(2014)147–156Fig.2.400MHz 1H NMR urine spectra (A —0.5–4.3ppm and B —5.1–9.0ppm)from controls and MCT-treated animals on Day 7.a —acetate;b —succinate;c —dimethylglycine;d —citrate;e —trimethylamine;f —alpha-ketoglutarate;g —dimethylamine/sarcosine;h —creatine;i —creatinine;j —ornithine;k —taurine/trimethylamine-N-oxide/betaine;l —taurine;m —betaine;n —alanine;o —glucose;p —allantoin;q —urea;r —trans aconitate;s —hippurate.150R.Conotte,J.-M.Colet /Toxicology and Applied Pharmacology 276(2014)147–156doublet),taurine (3.25ppm,triplet;3.43ppm,triplet),betaine (3.27ppm,singlet;3.90ppm,singlet),trimethylamine-N-oxide (3.27ppm,singlet),trans-aconitate (6.60ppm,singlet),glucose (5.24ppm,doublet),allantoin (5.39ppm,singlet)and urea (5.78ppm,singlet)(Table 2).MCT induce SOS —predictive modelThe metabonomic model of hepatic sinusoidal obstruction syndrome was based on a partial least square discriminant analysis (PLS-DA)per-formed on urines from MCT-exposed animals and control rats.In this model,the animals were divided into three classes.Class 1consisted in control rat urines while classes 2and 3corresponded to early (days 1to 3)and late (days 4to 7)stages of the SOS model,respectively.The score plot corresponding to that PLS-DA analysis showed a good separa-tion between the 3groups (R 2X =0.717;R 2Y =0.770;Q 2=0.691;Fig.7A)and revealed a metabonomic trajectory re flecting timely urinemetabolic changes.To further validate the established model,a random permutation test with 50permutations was performed with PLS-DA.The plot presented in Fig.7B strongly indicates the validity of our PLS-DA model because the permuted R 2(in green)and Q 2(in blue)values located in the left side of the graph were lower than the original points to the right and Q 2regression lines (in blue)have negative inter-cepts (R 2:[0.0,0.116];Q 2:[0.0;−0.242]).Additionally,CV-ANOVA test (cross-validation analysis of variance test)was performed to evaluate the statistical signi ficance of our PLS-DA model.This resulted in a score of p =2.21×10−29,indicating that the difference between the three groups within the model was highly signi ficant at a level of signif-icance α=0.05.DiscussionChemotherapy is part of a routine treatment applied in the fight against many cancer types.In conjunction with other cancertreatments,Fig.3.PCA (scores plot)of 1H NMR spectra of urine samples from rats repeatedly exposed to doses of monocrotaline (red square)and control (black square).Red arrow represents the direction of early phase urine sampling (Days 1to 3)and in blue the direction of the late phase (Days 4to 7).The model parameters were:R 2Xcum =0.803;Q 2cum =0.467;and Hotelling's T2:0.95.Fig.4.PCA (scores plot)of 1H NMR spectra of urine samples collected during the first 3days from rats repeatedly exposed to doses of monocrotaline.Red squares represent urine samples collected during the period 0–8h after monocrotaline administration and blue squares the period 8–24h after treatment.The model parameters were:R 2Xcum =0.850;Q 2cum =0.467;and Hotelling's T2:0.95.151R.Conotte,J.-M.Colet /Toxicology and Applied Pharmacology 276(2014)147–156such as radiation therapy or surgery,it is used with a curative intent or,in many cases,it helps prolonging life.During chemotherapy,the patient is exposed to drugs which have been developed to weaken and destroy cancer cells in the body.It is a systemic therapy,which means that it may affect the whole body due to its distribution through the bloodstream.Hence,chemotherapeutic agents may also induce adverse effects in patients and negatively impact survival.Among the most often encountered adverse effects associated with chemotherapy,hepatotoxicity has been reported during CRC treatments with irinotecan and oxaliplatin.In the case of oxaliplatin induced SOS,this potential hep-atotoxicity can lead to a higher risk of post-surgery complication and de-creased liver regeneration after partial hepatectomy (Schiffer et al.,2009).This toxicity can limit the use and the duration of neoadjuvant chemotherapy and thus negatively impact the overall patient survival rate.Currently,there is no simple,predictive and noninvasive diagnostic test,but a biopsy.In recent years,the metabonomics approach has been successfully used to identify characteristic metabonomic signatures of some diseases and adverse effects,from which sets of biomarkers were validated for diagnostic and prognostic purposes.In the present study,we used a NMR-based metabonomics strategy and MCT as a prototypical molecule to build a predictive model of drug-induced SOS in rats.The earliest changes after MCT ingestion lead to the loss of the SEC fenestra-tions and the appearance of gaps in the lining.The sinusoidal damage can cause obstruction resulting in focal ischemia and progressive microvas-cular,parenchymal,and Kupffer cell phagocytic alterations in the liver.Those severe histopathological changes usually lead to death.A strong fibrotic reaction in the sinusoids characterizes the later stages of SOS and leads to obliteration of the venules (Kumar et al.,2003).Many fac-tors such as SEC glutathione depletion,nitric oxide (NO)depletion and increased matrix metalloproteins and vascular endothelial growth factor are involved in the SOS pathogenesis (Helmy,2006).During the study,a signi ficant weight loss was observed in animals treated with MCT.This weight loss did not represent a loss of 20%of the initial weight,animals have therefore not been euthanized.Under our experimental conditions,the clinical data collected in MCT-exposed animals showed an increased relative liver weight,evidencing hepatomegaly.In addition,ascites were seen in two animals in the group treated with MCT.Both hepatomegaly and ascites have been reported in many cases of SOS in humans (DeLeve et al.,2002;Helmy,2006).Another clinical sign occurring in humans and observed in our experimental study was the increased ALT serum levels in the course of SOS.This effect most likely re flects the ischemic hepatocyte necrosis (DeLeve et al.,2002).The metabonomic evaluation of the MCT-induced model of SOS also revealed multiple metabolic changes which are gathered in functional clusters and discussed hereafter.Dysfunction of energy metabolismThe urine levels of several intermediate products of Tricarboxylic acid (TCA)cycle (citrate,succinate,alpha-ketoglutarate)were de-creased in the MCT group as compared to controls.A concurrent decrease in the level of trans -aconitate,spontaneously formed from cis -aconitate,was also seen.TCA cycle is the central metabolic energy pathway that leads in the presence of oxygen to the production of ATP in mitochondria.The metabonomics concept portends that a change in the intracellular levels of those metabolites able to cross cell membranes will spontaneously be re flected in extracellular fluids to maintain homeostasis.So,the decrease in urine levels of TCA cycle inter-mediates observed after exposure to MCT most likely re flects a lowering of the mitochondrial respiratory functions at the cellular level.This downregulation of TCA cycle could induce organ failure (Liu et al.,2010;Lu et al.,2010;Tang et al.,2012;Zhang et al.,2011).Other meta-bolic changes supporting a disorder in energy metabolism were also seen such as increased urine level of acetate which is the end product of fat metabolism (Liu et al.,2010).It is known that acetyl-CoA produced by the beta -oxidation of fatty acids can be metabolized to acetate con-comitantly with the formation of ketones.Acetate synthesis is catalyzed by the liver mitochondrial enzyme acetyl-CoA hydrolase (Yamashita et al.,2006).Hippurate formation is a commonly used clinical marker of liver function and also provides indication of the energy level status.Hippurate is an acyl glycine formed by the conjugation of benzoic acid with glycine in the liver.The metabolism of benzoate depends on an ap-propriate supply of ATP via oxidative phosphorylation.A decrease in hippurate level can be explained by a lower ATP production caused by a default in the oxidative phosphorylation (Lu et al.,2010).The last change that indicates energy metabolism disorder is a decrease in urinary glucose that could be explained by the reduction in food consumption.Glucose is the primary energy source for the body and is metabolized to pyruvate during glycolysis.In aerobic conditions,pyru-vate is metabolized into acetyl-CoA which can enter the TCA cycle.A decrease in the contribution of glucose by food consumption should di-rectly impact the ATP synthesis in the body.To reduce the impact of this intermittent contribution,the body first depletes liver glycogen before gluconeogenesis to maintain blood glucose concentration and satisfy energy requirement of thebody.Fig.5.PCA (scores plot)of 1H NMR spectra of urine samples collected during the first 3days from rats repeatedly exposed to doses of monocrotaline (red squares)and control rats (black squares).The model parameters were:R 2Xcum =0.897;Q 2cum =0.437;and Hotelling's T2:0.95.152R.Conotte,J.-M.Colet /Toxicology and Applied Pharmacology 276(2014)147–156。
