Mass Spectrometry-Based Metabolic Profiling of Rat Urine Associated

Jan 19, 2008 - Enzyme beats catalysts at oxidation game. Oxidizing the end carbon atom of a terminal alkene to form an aldehyde, a process called anti...
0 downloads 11 Views 370KB Size
Download PDF
288

Chem. Res. Toxicol. 2008, 21, 288–294

Chemical Profile Mass Spectrometry-Based Metabolic Profiling of Rat Urine Associated with General Toxicity Induced by the Multiglycoside of Tripterygium wilfordii Hook. f. Minjun Chen,†,⊥ Yan Ni,†,⊥ Hongquan Duan,‡,⊥ Yunping Qiu,† Congying Guo,† Yang Jiao,† Huijuan Shi,§ Mingming Su,† and Wei Jia*,† School of Pharmacy, Shanghai Jiao Tong UniVersity, Shanghai 200240, People’s Republic of China, School of Pharmacy, Tianjin Medical UniVersity, Tianjin, People’s Republic of China, and Shanghai Institute of Planned Parenthood Research, Shanghai, People’s Republic of China ReceiVed August 16, 2007

We propose here a combined gas chromatography/mass spectrometry (GC/MS) and liquid chromatography/mass spectrometry (LC/MS) metabolic profiling strategy to elucidate the toxicity in rats induced by orally administered multiglycosides of Tripterygium wilfordii Hook. f. (GTW) in multiple organs including the kidney, liver, and testis. Overnight 12-h urine samples were collected from Sprague–Dawley male rats exposed to GTW (100 mg/kg/day, n ) 6) and healthy controls (n ) 6) at predose and at the 1st, 3rd, 6th, 10th, and 14th day postdose for both GC/MS and LC/MS analyses. The integrated urinary MS data were analyzed via multivariate statistical techniques such as principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) to identify the differential metabolites and pertinent altered biological pathways in response to the herbal toxin. The liver, kidney, and testis were also assessed using conventional histopathological examinations at the end point of the experiment. This work indicates that GTW caused a time-dependent toxic effect at a high dose as revealed by the perturbed metabolic regulatory network involving disorders in energy metabolism, elevated amino acid and choline metabolism pathways, as well as altered structure of gut flora. This integrated MS-based metabolic profiling approach has been able to capture and probe the metabolic alterations associated with the onset and progression of multiorgan toxicity induced by GTW, thereby permitting a comprehensive understanding of systemic toxicity for phytochemicals and other types of xenobiotic agents. Introduction Tripterygium wilfordii Hook. F. (GTW), a woody vine also called Thunder God Vine native to Eastern and Southern China, Korea, Japan, and Burma, has been widely used as a herbderived remedy for rheumatoid arthritis (RA), inflammation, cancer, chronic nephritis, hepatitis, systemic lupus erythematosus, ankylosing spondylitis, and a variety of skin conditions (1, 2). A prospective, double-blind, placebo-controlled study on patients with long-standing RA who failed to respond to conventional therapy indicated that ethylacetate extracts of GTW exhibited a considerable effect at a dose of 360 mg/day (3). The US Food and Drug Administration has recently approved a clinical study to evaluate the benefits of certain Tripterygium extracts in the treatment of RA (4). Despite its potential clinical utility, many side effects, such as gastrointestinal upset, hair loss, infertility, and suppression of lymphocyte proliferation, and renal failure, have been reported (3, 5, 6). The antifertility activity of GTW also suggested reproductive toxicity as * To whom correspondence should be addressed. Tel: 86-21-6293-2292. Fax: 86-21-6293-2292. E-mail: [email protected]. † Shanghai Jiao Tong University. ‡ Tianjin Medical University. § Shanghai Institute of Planned Parenthood Research. ⊥ These authors are contributed equally to this work.

evidenced by the dramatic decrease in the viability of epididymal spermatozoa and sperm density in male rats (7, 8). These studies have raised considerable concerns on the adverse effects associated with the long-term use of GTW. Nevertheless, to our knowledge, research on GTW-induced toxicity was scarce, presumably due to the multiple targeting sites involved in its in ViVo toxic effect as well as the complexity of multiple chemical ingredients, which may vary with geographical locations and the methods of processing of such a medicinal herb (1). Conventional biochemical studies on drug toxicity have been challenged for many reasons. First, only a small part of endogenous metabolites and key enzymes are routinely detected and assessed in biofluids such as urine and plasma. Second, many of these endogenous metabolites and enzymes assessed by conventional biochemistry are extremely sensitive to time or different analytical techniques and interindividual variation. Hence, the results may tend to be misleading or even contradictory without histopathological examination (9). Fortunately, aiming to monitoring as many endogenous low-molecularweight metabolites (generally less than 1000 Da.) as possible, metabolic profiling strategy enables biologists to identify the consistently varying metabolites and elucidate changes in the relevant metabolic regulatory network using high-field nuclear

10.1021/tx7002905 CCC: $40.75  2008 American Chemical Society Published on Web 01/19/2008

Metabolic Profiling of Rat Urine

Figure 1. Graph of body weight changes in the two groups with time change (*, p < 0.05, significant difference with the CG; **, p < 0.01, significant difference with the CG; each time point presents mean ( S.D., n ) 6 per group).

magnetic resonance (NMR), gas chromatography/ mass spectrometry (GC/MS), liquid chromatography/mass spectrometry (LC/MS), and capillary electrochromatography/mass spectrometry (CE/MS) in combination with multivariate statistical techniques (10). 1H NMR spectroscopy has been extensively utilized to identify a number of toxicity-related biomarkers, such as taurine, in urine, plasma, or organs as a biochemical indicator of hepatotoxicity, serine and creatinine as markers of nephrotoxicity (11–15). Alongside these, mass spectrometry-based metabolic profiling strategy has been commonly used to seek for candidate biomarkers for diverse diseases, probe dynamic progression of certain morbidity (16), and unravel potential therapeutic mechanisms of drugs (17–20). In reality, no single analytical instrument is powerful enough to capture the entire endogenous metabolites. Thus, the parallel use of multiple analytical methods will be advantageous in identifying a broader spectrum of metabolites relevant to physiopathological alteration (21). We conceive that the use of combined MS-based analytical methods is able to gain more latent information from metabolic profiles of biological samples. The application of combined GC/ MS and LC/MS metabolic profiling has been recently reported to investigate aristolochic acid-induced acute nephrotoxicity in rats by our group (21). In this study, we also adopted a combined MS-based metabolic profiling approach to understanding the toxicological outcomes of GTW in multiple organs through monitoring urinary low-molecular-weight metabolites, which potentially offers a comprehensive understanding of the systemic toxicity of phytochemicals and other types of xenobiotic agents.

Materials and Methods Animal Handling Procedure and Tissue Preparation. All animal studies were in compliance with national legislations of China and local guidelines, and were performed at the Centre of Laboratory Animals, Shanghai Jiao Tong University, Shanghai, People’s Republic of China. Oral Tripterygium wilfordii multiglycoside (10 mg/tablet, lot no. 20060413) was purchased from Shanghai Fudan Forward Pharmaceutical Co., Ltd. A total of 12 nine-week-old male Sprague–Dawley (S.D.) rats (180 ( 10 g) were purchased from Shanghai Laboratory Animal Co. Ltd. (SLAC, Shanghai, China), individually housed in stainless steel wire-mesh cages, and fed with a certified standard rat chow (lot. no. M02; SLAC, Shanghai, China) and tap water ad libitum. Room temperature and humidity were regulated at 24 ( 1 °C and 45 ( 15%, respectively. A light cycle of 12 h on and 12 h off was set, with lights on at 08:00 a.m. After two weeks of acclimatization, rats

Chem. Res. Toxicol., Vol. 21, No. 2, 2008 289 were randomly divided into two groups as follows: the dosed group (DG, n ) 6), received GTW by oral gavage at a daily dose of 100 mg/kg body weight from day 1 to day 14; the control group (CG, n ) 6), received the same volume of vehicle (0.9% saline solution) daily. Drug or vehicle was administrated between 8:30 and 9:30 a.m. to minimize any effect of circadian rhythm. Overnight 12-h urine samples from each rat were collected at predose and at the 1st, 3rd, 6th, 10th, and 14th day postdose, and directly centrifuged at 5,000g for 10 min at room temperature to remove particle contaminants. The resultant supernatants were stored at -80 °C pending LC/MS and GC/MS analyses. The urine volume and body weight of each rat were recorded across the experiment course. Histopathological Examination. All animals were sacrificed on the 15th day for histopathological assessment of the liver, kidney, and testis, all of which were dissected free of fat and adjacent tissue, weighed, and processed for the extent of tissue lesion. Representative samples for histology were fixed in 10% formalin for 12 h, embedded in paraffin wax, cut into 5-µm sections, stained with hematoxylin-eosin (HE), and finally examined by light microscopy. GC/MS Spectrum Acquisition. The urine samples were prepared for GC/MS analysis, and relevant spectral acquisition was obtained according to our previously developed methods with minor modifications (18). Briefly, rat urine samples were prepared using a 600-µL aliquot of diluted urine sample (urine/water ) 1:1, v/v) for ethyl chloroformate (ECF) derivatization, with L-2-chlorophenylalanine employed as an internal standard to monitor the batch reproducibility in parallel. A 1-µL aliquot of analyte was injected into a DB-5MS capillary column coated with 5% diphenyl crosslinked 95% dimethylpolysiloxane (30 m × 250 µm i.d., 0.25-µm film thickness; Agilent J&W Scientific, Folsom, USA) and conducted on a hyphenated Perkin Elmer gas chromatograph and TurboMass-Autosystem XL mass spectrometer (Perkin Elmer Inc., USA). All of the samples were analyzed at a random sequence to eliminate any systemic bias. LC/MS Spectrum Acquisition. Details of this protocol were provided in our previous paper (19). In brief, urinary samples were centrifuged, and the resultant supernatant was transferred into a 2-mL autosampler vial. Each 20-µL aliquot of urine sample was injected into a 4.6 mm × 15 cm Extend-C18 5-µm column using an Agilent 1100 Series for LC/MS (Agilent Tech., USA). The positive ion data were collected in full scan mode (m/z 100–1000). Samples were analyzed randomly for unbiased measurement with tune mixture solution (m/z ) 118.09, 622.05, or 922.02, Agilent Tech.) employed as internal standards for quality control. Data Reduction and Statistical Analysis. Nonprocessed GC/ MS and LC/MS data files were converted into NetCDF format via DataBridge (Perkin Elmer Inc., USA) and Agilent LC/MSD ChemStation (Rev.B.01.03, Agilent Tech., USA), respectively. Each file was extracted subsequently using our custom scripts in MATLAB 7.0 (The MathWorks, Inc. USA), where baseline correction, peak deconvolution and alignment, internal standard exclusion, and normalization to the total sum of the chromatogram were carried out. The resultant three-dimensional matrix encompassing compound indices (retention time-m/z pairs), sample names (observations), and normalized peak areas (variables) were imported into the SIMCA-P 11.0 software package (Umetrics, Umeå, Sweden) and preprocessed using mean-centering and pareto-scaling prior to multivariate statistical analysis. Mean centering involves calculating the average spectrum of the data set and subtracting that average from each spectrum. Pareto-scaling weights each variable by the square root of its standard deviation, which amplifies the contribution of lower concentration metabolites but not to such an extent where noise produces a large contribution (22). Principal component analysis (PCA) was initially used to visualize general clustering, trends, and outliers among the observations. For further identifying the differentially expressed metabolites accountable for the separation between the dosed and control group, a more sophisticated partial least-squares-discriminant analysis (PLS-DA) was carried out on the combined MS data set. Additionally, on the basis of the threshold of p values and fold change values from the nonparametric Wilcoxon-Mann–Whitney test implemented in

290 Chem. Res. Toxicol., Vol. 21, No. 2, 2008

Chen et al.

Figure 2. Typical GC/MS spectra and LC/MS spectra of urine samples derived from different groups (CG, A and C; DG, B and D) at the end point of the experiment course. The key of the GC/MS spectra is listed in Table 1. IS, internal standard.

Table 1. List of Most Profound Metabolites Relevant to GTW-Induced Toxicity DG/CG a

b

DG/CG

no.

metabolites (GC/MS)

corr. coeffs.

P

fold changes

no.

metabolites (LC/MS)

corr. coeffs.

P

fold changes

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

phenylacetatec alanine succinate leucinec glucosec malate aconitate asparate glutarate hippurate citrate m-hydroxy-phenylpropionate glutamine hexadecanoic acid lysine tyrosine 5-hydroxy-indol-3-acetate

0.73(v) 0.78(v) -0.88(V) 0.71(v) -0.86(V) -0.85(V) -0.91(V) 0.91(v) 0.88(v) 0.89(v) -0.88(V) -0.71(v) 0.70(v) 0.70(v) 0.61(v) 0.60(v) -0.67(V)

0.0222 0.0066 0.0027 0.0222 0.0027 0.0042 0.0027 0.0027 0.0027 0.0027 0.0027 0.0222 0.0042 0.0321 0.0864 0.0455 0.0222

2.1 2.4 -2.6 2.1 -2.6 -2.5 -2.6 2.6 2.6 2.5 -2.6 -2.1 2.5 2.0 1.7 1.9 -2.1

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

acetylcarnitine phenylacetatec sarcosine pyruvate acetoacetate acetyl-lysine creatinine fumarate gluconic acid leucinec nicotinamide choline serine glucosec linolenic acid phenylacetylglycine palmitoleic acid

0.80(v) 0.83(v) 0.78(v) 0.62(v) 0.65(v) 0.60(v) 0.90(v) -0.89(V) 0.62(v) 0.77(v) -0.88(V) 0.62(v) 0.61(v) -0.84(V) 0.91(v) 0.60(v) 0.89(v)

0.0027 0.0132 0.0027 0.0046 0.0321 0.0455 0.0027 0.0027 0.0222 0.0027 0.0027 0.0321 0.0455 0.0027 0.0027 0.0222 0.0027

2.9 3.0 2.9 2.5 2.6 2.9 2.6 -2.6 2.1 2.9 -2.6 2.0 1.9 -2.6 2.6 2.1 2.6

a The correlation coefficients shown were obtained from a cross-validated PLSDA of the integrated MS data set; the positive value indicates an increase of the metabolites in the DG (denoted by an up-arrow). b The critical value was set at 0.05. c The metabolites were detected by both GC/MS and LC/MS analyses.

the Matlab statistical toolbox, the differential metabolites derived from the correlation coefficients (22) of a cross-validated PLS-DA model were validated at a univariate level. In the SIMCA-P package, a typical cross-validation procedure was conducted by leaving 1/7th samples out in each round so as to validate the PLS-DA model against overfitting. The critical p value of the test was set at 0.05 throughout this study.

Results General Observations. As compared with the control group, the rats in the dosed group exhibited general variations, for example, smaller body weight gain (Figure 1), less physical activity, higher volume of water consumption, and the distinctly altered urine excretion volume (decreased at day 1 and 3, and increased at the time points shown in Table S1, Supporting Information). The significance of the body weight changes was evaluated by a one-way ANOVA test implemented in the Matlab software. All of these observed significant variations in the dosed group are associated with GTW exposure.

Figure 3. PCA metabolic trajectory plot (PC1 vs PC2) of mean scores from a combined GC/MS- and LC/MS-based metabolic profiling of urine samples from the two groups (DG vs CG) at predose (0) and at 1st, 3rd, 6th, 10th, and 14th day following GTW administration. Each point denotes six subjects. The error bar represents the standard deviation for each time point obtained by the first principal component.

Metabolic Profiling of Rat Urine

Figure 4. Three-dimensional-PCA scores plot derived from the integrated GC/MS and LC/MS data of urine samples from the two groups (DG, 9; CG, b) at day 14. Each point denotes a subject.

GTW-Induced Histopathological Examination. The weight variation between the DG and CG was observed in the liver (8.79 ( 1.18 g vs 8.12 ( 2.11 g, p > 0.05, ANOVA), the combined kidney (2.10 ( 0.11 g vs 2.93 ( 0.22 g, p < 0.05), and the combined testis (1.73 ( 0.31 g vs 2.68 ( 0.17 g, p < 0.01). Comparison of HE sections of liver samples indicated slight centrilobular inflammation and multifocal single cell necrosis in GTW-induced rats (Supporting Information, Figure S1), whereas no obvious signs were noted in CG rats (Figure S2). Moderate degeneration of the distal tubules and extensive tubular necrosis were seen in the kidney of DG rats (Figure S4) as compared with the CG (Figure S3). The testis section of CG rats showed normal testicular structure with orderly arrangement of germinal cells (Figure S5), whereas GTW administration induced apparent testicular atrophy with extensive degeneration and exfoliation of germ cells in seminiferous tubules (Figure S6). Visual Examination of MS Spectra. Typical GC/MS and LC/MS total ion current (TIC) chromatograms of urine samples derived from the dosed and control groups at the 14th day postdose are illustrated in Figure 2. Using our GC/MS and LC/ MS analytical protocols coupled with a software-based peak deconvolution procedure, a total of 261 individual metabolites were consistently detected in at least 95% of the urine samples. Compound identification was carried out by comparing mass spectra and retention time with those of commercially available authentic standards. On the basis of this approach, we were able to identify 87 (33.3%) of the 261 metabolites, most of which were organic acids, amines, amino acids, and amino-alcohols. Additionally, we have used the Kyoto Encyclopedia of Genes and Genomes (KEGG) Database (23) to connect these altered metabolites to different pathways and key enzymes. Visual inspection of the TIC chromatograms from GC/MS (Figure 2A and B) and LC/MS (Figure 2C and D) analyses revealed the general fluctuation of the urinary metabolite composition between groups. For example, as compared to the control group, lower signals of certain metabolites including succinate, aconitate, citrate, and glucose were observed in the DG while elevated intensities of alanine, glutamine, lysine, and tyrosine were observed from GC/MS spectra. Analogously, from LC/MS spectra, the signal ranging from 2.8 to 3.0 min (retention time) significantly increased in the DG versus the CG, and in such a region, sarcosine, pyruvate, and creatinine were identified by authentic standards. Because of the complexity of TIC chromatograms for biological samples, sophisticated peak decon-

Chem. Res. Toxicol., Vol. 21, No. 2, 2008 291

volution procedures were utilized to resolve many coeluting peaks present in the chromatograms. Multivariate statistical techniques such as principal component analysis (PCA) were further employed to envisage altered metabolic patterns in association with GTW intervention. Detailed descriptions are given in our previous paper (19, 21). Time-Dependent Metabolic Effects. To investigate the metabolic variation in rats following GTW administration, MS spectra of all of the urine samples collected in this study were integrated and analyzed using PCA with the first three PCs describing 81% of the total variance. Metabolic trajectories from mean values of PC1 and PC2 scores clearly indicated a timedependent development of GTW-induced toxicity (Figure 3). As compared to the metabolic profiles in the CG, which largely reflect physiological variations, the DG rats exhibited a progressive shift from the predose state, reaching maximum deviation from the predose state at the 10th and 14th day. This timedependent tendency is consistent with the general observations of dosed rats throughout the experiment as well as with the histopathological examination at the end point of the experiment. Identification of the Differential Metabolites. To further assess the ultimate toxin-induced biochemical changes, PCA analysis of urinary MS data sets from the two groups at the end point of the experiment (day 14) was performed. The 3DPCA scores plot shows a clear separation between the DG and CG in the PC1 dimension (Figure 4). To maximize such variation and identify the differential metabolites accountable for the separation between groups at day 14, a two-component cross-validated PLS-DA model was obtained with satisfactory predictive accuracy (Q2Ycum ) 0.90) (24). The corresponding correlation coefficients are listed in Table 1 in which a positive value means a significant elevation in the DG, whereas a negative value represents a significant decrease in this group. The p values and fold change values (expressed as relative concentration) for these metabolites were obtained using a nonparametric Wilcoxon-Mann–Whitney test. The higher values in the fold change together with statistically significant p values demonstrated a severe disturbance of the endogenous metabolic network in the DG rats as compared to that in the CG. In this way, the differentially expressed metabolites from multivariate statistical analyses were validated at a univariate level.

Discussion Energy Metabolism. A group of metabolites involved in energy generation and consumption were significantly perturbed as a combined result of GTW administration and altered dietary status (25) in response to GTW exposure (Figure 5A). A significantly decreased concentration of glucose was detected in DG urine samples. Alteration of glucose is considered to be involved in liver and renal toxicity (9, 15), suggesting an underlying impairment in liver and kidney function of GTWexposed rats. This is evidenced by the moderate and severe hepatocyte necrosis and extensive epithelial cell debris observed in our histological examination. Meanwhile, the high levels of acetoacetate, pyruvate, and acetylcarnitine detected in urinary samples indicate the increased metabolism of fatty acids toward ketone body formulation and utilization. The disturbed fatty acid metabolism may contribute to the smaller body weight gain of GTW-dosed rats. Incompatible with such enhanced upstream biochemical reaction, the decreased urinary excretion of TCA intermediates including citrate, aconitate, succinate, and malate were observed from GC/MS analysis, and fumarate was detected by LC/MS. This phenomenon may be attributed to the decreased

292 Chem. Res. Toxicol., Vol. 21, No. 2, 2008

Chen et al.

Figure 5. Perturbed metabolic regulatory network in response to GTW exposure. (A) GTW-induced disorders of energy metabolism. (B) Perturbed choline-related metabolic pathway. The black square denotes an elevated urinary level of metabolites in DG rats, whereas the gray square means a reduced level of metabolites in this group as compared to that in the CG. The solid line represents the direct reaction, while the dashed line indicates the omission of several intermediate processes.

activity of citrate synthase in mitochondria, where TCA metabolism takes place. A similar observation on the decreased activity of citrate synthase was reported in the lung, liver, kidney, and testis of male rats 48 h postexposure to cadmium at a dose of 850 µg Cd/m3 (26). Hence, the inconsistent response of up and downstream energy metabolism induced by GTW appeared to be indicative of the impaired mitochondrial system in which citrate synthase was disturbed, leading to the accumulation of metabolites in the upstream pathway as well as the enervation of downstream TCA cycle. In addition, the decreased nicotinamide, an important precursor of the coenzymes NADH and NADPH, which are indispensable electron transporters involved in TCA cycle (27), also indicated the impaired energy metabolism. Amino Acid Metabolism. Correlation coefficients from a cross-validated PLS-DA model suggested the significantly elevated levels of urinary amino acids, such as leucine, lycine, acetyl-lycine, tyrosine, and glutamine. It has been reported that the tyrosine aminotransferase activity was inhibited and the corresponding biotransformation from tyrosine to tyramine could be affected in the benzene intoxicated rat kidney and liver (28). In this case, the increased level of tyrosine detected in urine may be an indication of GTW-induced damage to animal kidney and liver. Additionally, the elevated levels of other amino acids, serine in particular, metabolized through amino acid oxidase located in the peroxisomes of renal tubular epithelial cells, suggested the impairment of the kidney, where reabsorption and reutilization of amino acids

occurred (19). In general, GTW administration at a toxic dose would induce the necrosis of liver parenchyma and disturb muscle proteolysis, accompanied by abnormalities in the metabolism of these amino acids in the mean time. Choline Metabolism. The raised levels of choline, sarcosine, creatinine, and serine were observed in urine via LC/MS analysis. All of these compounds of interest were critical intermediates of choline metabolism (Figure 5B) (29). Choline and phosphcholine, products of lipid catabolism mediated by enzymes such as phosphopase C and D, play an important role in protecting the liver from lipidosis. The upregulated choline metabolism occurring in DG rats may be a metabolic regulatory response to the administration of GTW xenobiotic agents. Additionally, the altered levels of urinary unsaturated fatty acids such as linolenic acid and palmitoleic acid also suggested the disrupted pathway of choline metabolism. We thus conceive that the alteration of choline metabolism is a direct outcome of GTW-induced toxicity in the liver and kidney, as a selfcomplementary mechanism, consistent with the hepatic and renal lipidosis observed in this study. Gut Microbiota-Related Metabolites. The prominently differential urinary excretion of hippurate, phenylacetate, 5-hydroxy-indol-3-acetate, and m-hydroxy- phenylpropionate was identified in the DG using the GC/MS technique (30, 31). Thus, the significant alteration of these metabolites may reflect the altered structure and activity of symbiotic gut microbiota. Therefore, this finding suggests that GTW may alter the

Metabolic Profiling of Rat Urine

metabolic profiles through multitargeted interactions in ViVo with the host as well as with symbiotic gut microbiota. This is in good agreement with widely reported toxin-induced variation in gut microbiota (15). Merits of the Metabolic Profiling Approach for General Toxicity Evaluation. Recent innovations in analytical instruments such as 1H NMR, FTMS, Q-TOFMS, and GC-TOFMS have enabled the high-throughput detection of global changes in endogenous metabolites. Monitoring the entire set of metabolites (metabolome) present in the urine using modern sophisticated analytical instruments coupled with multivariate statistical techniques, researchers are able to probe the systemic and dynamic variations in the metabolic profiles and provide a complementary or alternative way to evaluate the toxicity induced by xenobiotics. The metabolic profiling strategy we adopted through integrating data from GC/MS and LC/MS analyses greatly widens the metabolic window for understanding the underlying events of xenobiotic intervention. Most of the differentially expressed metabolites detected from the two analytical methods are complementary to each other, while only three compounds, leucine, phenylacetate, and glucose, were overlapped. Furthermore, while only several metabolites are specific to certain impaired organ functions, most of these metabolites are generally involved in the metabolic regulatory pathways, reflecting the xenobiotic induced global biochemical changes and physiological consequences. Further studies should be directed toward assessing the endogenous metabolites of other types of biological samples such as serum and tissues of specific organs from the same groups and correlating the data to each other in order to elucidate the complex mechanisms of in ViVo toxicity induced by this herbal agent. Acknowledgment. This study was financially supported by research grants 06DZ05906 and 064307067 from Shanghai Science and Technology Commission. Supporting Information Available: Urine volume varying with time and histopathological examination (×250) of rat liver, kidney, and testis. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Brinker, A. M., Ma, J., Lipsky, P. E., and Raskin, I. (2007) Medicinal chemistry and pharmacology of genus Tripterygium (Celastraceae). Phytochemistry 68, 732–766. (2) Shamon, L. A., Pezzuto, J. M., Graves, J. M., Mehta, R. R., Wangcharoentrakul, S., Sangsuwan, R., Chaichana, S., Tuchinda, P., Cleason, P., and Reutrakul, V. (1997) Evaluation of the mutagenic, cytotoxic, and antitumor potential of triptolide, a highly oxygenated diterpene isolated from Tripterygium wilfordii. Cancer Lett. 112, 113–117. (3) Tao, X., Younger, J., Fan, F. Z., Wang, B., and Lipsky, P. E. (2002) Benefit of an extract of Tripterygium Wilfordii Hook F in patients with rheumatoid arthritis: a double-blind, placebo-controlled study. Arthritis Rheum. 46, 1735–1743. (4) Pyatt, D. W., Yang, Y., Mehos, B., Le, A., Stillman, W., and Irons, R. D. (2000) Hematotoxicity of the chinese herbal medicine Tripterygium wilfordii hook f in CD34-positive human bone marrow cells. Mol. Pharmacol. 57, 512–518. (5) Gu, J., Zhu, C., Wang, W., and Wang, L. (2001) The effects of Lei Gong Teng on reproductive hormones. J. Tradit. Chin. Med. 21, 50–51. (6) Tao, X., and Lipsky, P. E. (2000) The Chinese anti-inflammatory and immunosuppressive herbal remedy Tripterygium wilfordii Hook F. Rheum. Dis. Clin. North Am. 26, 29–50. (7) Chan, W. Y., and Ng, T. B. (1995) Adverse effect of Tripterygium wilfordii extract on mouse embryonic development. Contraception 51, 65–71. (8) Sharma, M., Li, J. Z., Sharma, R., Artero, M., Ge, X., McCarthy, E. T., Wang, H. Y., and Savin, V. (1997) Inhibitory effect of Tripterygium wilfordii multiglycoside on increased glomerular albumin permeability in vitro. Nephrol., Dial., Transplant 12, 2064–2068.

Chem. Res. Toxicol., Vol. 21, No. 2, 2008 293 (9) Nicholls, A. W., Holmes, E., Lindon, J. C., Shockcor, J. P., Farrant, R. D., Haselden, J. N., Damment, S. J., Waterfield, C. J., and Nicholson, J. K. (2001) Metabonomic investigations into hydrazine toxicity in the rat. Chem. Res. Toxicol. 14, 975–987. (10) Nicholson, J. K. (2006) Global systems biology, personalized medicine and molecular epidemiology. Mol. Syst. Biol. 2, 52. (11) Beckwith-Hall, B. M., Holmes, E., Lindon, J. C., Gounarides, J., Vickers, A., Shapiro, M., and Nicholson, J. K. (2002) NMR-based metabonomic studies on the biochemical effects of commonly used drug carrier vehicles in the rat. Chem. Res. Toxicol. 15, 1136–1141. (12) Bollard, M. E., Xu, J., Purcell, W., Griffin, J. L., Quirk, C., Holmes, E., and Nicholson, J. K. (2002) Metabolic profiling of the effects of D-galactosamine in liver spheroids using (1)H NMR and MAS-NMR spectroscopy. Chem. Res. Toxicol. 15, 1351–1359. (13) Corcoran, O., Mortensen, R. W., Hansen, S. H., Troke, J., and Nicholson, J. K. (2001) HPLC/1H NMR spectroscopic studies of the reactive alpha-1-O-acyl isomer formed during acyl migration of S-naproxen beta-1-O-acyl glucuronide. Chem. Res. Toxicol. 14, 1363– 1370. (14) Mortishire-Smith, R. J., Skiles, G. L., Lawrence, J. W., Spence, S., Nicholls, A. W., Johnson, B. A., and Nicholson, J. K. (2004) Use of metabonomics to identify impaired fatty acid metabolism as the mechanism of a drug-induced toxicity. Chem. Res. Toxicol. 17, 165– 173. (15) Waters, N. J., Waterfield, C. J., Farrant, R. D., Holmes, E., and Nicholson, J. K. (2005) Metabonomic deconvolution of embedded toxicity: application to thioacetamide hepato- and nephrotoxicity. Chem. Res. Toxicol. 18, 639–654. (16) Denkert, C., Budczies, J., Kind, T., Weichert, W., Tablack, P., Sehouli, J., Niesporek, S., Konsgen, D., Dietel, M., and Fiehn, O. (2006) Mass spectrometry-based metabolic profiling reveals different metabolite patterns in invasive ovarian carcinomas and ovarian borderline tumors. Cancer Res. 66, 10795–10804. (17) Jia, W., Liu, P., Jiang, J., Chen, M. J., Zhao, L. P., Zhou, M. M., Yang, L. P., Wang, M. Q., Qiu, M. F., and Zhang, Y. Y. (2006) Application of metabonomics in complicated theory system research of traditional Chinese medicine. Zhongguo Zhong Yao Za Zhi 31, 621– 624. (18) Qiu, Y., Su, M., Liu, Y., Chen, M., Gu, J., Zhang, J., and Jia, W. (2007) Application of ethyl chloroformate derivatization for gas chromatography-mass spectrometry based metabonomic profiling. Anal. Chim. Acta 583, 277–283. (19) Chen, M., Su, M., Zhao, L., Jiang, J., Liu, P., Cheng, J., Lai, Y., Liu, Y., and Jia, W. (2006) Metabonomic study of aristolochic acid-induced nephrotoxicity in rats. J. Proteome Res. 5, 995–1002. (20) Chen, M., Zhao, L., and Jia, W. (2005) Metabonomic study on the biochemical profiles of a hydrocortisone-induced animal model. J. Proteome Res. 4, 2391–2396. (21) Ni, Y., Su, M., Qiu, Y., Chen, M., Liu, Y., Zhao, A., and Jia, W. (2007) Metabolic profiling using combined GC-MS and LC-MS provides a systems understanding of aristolochic acid-induced nephrotoxicity in rat. FEBS Lett. 581, 707–711. (22) Cloarec, O., Dumas, M. E., Trygg, J., Craig, A., Barton, R. H., Lindon, J. C., Nicholson, J. K., and Holmes, E. (2005) Evaluation of the orthogonal projection on latent structure model limitations caused by chemical shift variability and improved visualization of biomarker changes in1H NMR spectroscopic metabonomic studies. Anal. Chem. 77, 517–526. (23) Kyoto Encyclopedia of Genes and Genomes (KEGG) Database. http:// www.genome.ad.jp/kegg/. (24) Trygg, J., Holmes, E., and Lundstedt, T. (2007) Chemometrics in metabonomics. J. Proteome Res. 6, 469–479. (25) Connor, S. C., Wu, W., Sweatman, B. C., Manini, J., Haselden, J. N., Crowther, D. J., and Waterfield, C. J. (2004) Effects of feeding and body weight loss on the 1H-NMR-based urine metabolic profiles of male Wistar Han rats: implications for biomarker discovery. Biomarkers 9, 156–179. (26) Prasada Rao, P. V., and Gardner, D. E. (1986) Effects of cadmium inhalation on mitochondrial enzymes in rat tissues. J. Toxicol. EnViron. Health 17, 191–199. (27) Fedyk, M., Velykyi, M. M., Zababurina, M. L., and Oliiarnyk, O. D. (1996) The dynamic biosynthesis of nicotinamide coenzymes from nicotinamide and nicotinic acid in rat tissues. Ukr. Biokhim. Zh. 68, 29–33. (28) Bong, M., Michalska, A., Laskowska-Klita, T., and Szymczyk, T. (1985) Tyrosine aminotransferase activity in the benzene intoxicated rat. Bull. EnViron. Contam. Toxicol. 34, 36–44. (29) Garrod, S., Bollard, M. E., Nicholls, A. W., Connor, S. C., Connelly, J., Nicholson, J. K., and Holmes, E. (2005) Integrated metabonomic analysis of the multiorgan effects of hydrazine toxicity in the rat. Chem. Res. Toxicol. 18, 115–122. (30) Li, H., Ni, Y., Su, M., Qiu, Y., Zhou, M., Qiu, M., Zhao, A., Zhao, L., and Jia, W. (2007) Pharmacometabonomic phenotyping reveals

294 Chem. Res. Toxicol., Vol. 21, No. 2, 2008 different responses to xenobiotic intervention in rats. J. Proteome Res. 6, 1364–1370. (31) Dumas, M. E., Wilder, S. P., Bihoreau, M. T., Barton, R. H., Fearnside, J. F., Argoud, K., D’Amato, L., Wallis, R. H., Blancher, C., Keun, H. C., Baunsgaard, D., Scott, J., Sidelmann, U. G., Nicholson, J. K.,

Chen et al. and Gauguier, D. (2007) Direct quantitative trait locus mapping of mammalian metabolic phenotypes in diabetic and normoglycemic rat models. Nat. Genet. 39, 666–672.

TX7002905