International Journal for Parasitology 39 (2009) 547–558

Contents lists available at ScienceDirect

International Journal for Parasitology

journal homepage: www.elsevier .com/locate / i jpara

Metabolic profiling of a Schistosoma mansoni infection in mouse tissues usingmagic angle spinning-nuclear magnetic resonance spectroscopy

Jia V. Li a,b, Elaine Holmes a, Jasmina Saric a,b, Jennifer Keiser c, Stephan Dirnhofer d,Jürg Utzinger b, Yulan Wang e,*a Department of Biomolecular Medicine, Division of Surgery, Oncology, Reproductive Biology and Anaesthetics (SORA), Faculty of Medicine, Imperial College, London, UKb Department of Public Health and Epidemiology, Swiss Tropical Institute, Basel, Switzerlandc Department of Medical Parasitology and Infection Biology, Swiss Tropical Institute, Basel, Switzerlandd Institute of Pathology, University Hospital Basel, Basel, Switzerlande State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Centre for Magnetic Resonance, Wuhan Institute of Physics and Mathematics,The Chinese Academy of Sciences, Wuhan 430071, People’s Republic of China

a r t i c l e i n f o a b s t r a c t

Article history:Received 9 July 2008Received in revised form 13 October 2008Accepted 14 October 2008

Keywords:ChemometricsMagic angle spinningMetabolic profilingMouseNMR spectroscopySchistosomiasisSchistosoma mansoni

0020-7519/$34.00 � 2008 Australian Society for Paradoi:10.1016/j.ijpara.2008.10.010

* Corresponding author. Address: State Key Laboraand Atomic and Molecular Physics, Wuhan Centre forInstitute of Physics and Mathematics, The Chinese A430071, People’s Republic of China. Tel.: +86 27 8718

E-mail address: yulan.wang@wipm.ac.cn (Y. Wang

In order to enhance our understanding of physiological and pathological consequences of a patent Schis-tosoma mansoni infection in the mouse, we examined the metabolic responses of different tissue samplesrecovered from the host animal using a metabolic profiling strategy. Ten female NMRI mice were infectedwith �80 S. mansoni cercariae each, and 10 uninfected age- and sex-matched animals served as controls.At day 74 post infection (p.i.), mice were killed and jejunum, ileum, colon, liver, spleen and kidney sam-ples were removed. We employed 1H magic angle spinning-nuclear magnetic resonance spectroscopy togenerate tissue-specific metabolic profiles. The spectral data were analyzed using multivariate modellingmethods including an orthogonal signal corrected-projection to latent structure analysis and hierarchicalprincipal component analysis to assess the differences and/or similarities in metabolic responsesbetween infected and non-infected control mice. Most tissues obtained from S. mansoni-infected micewere characterized by high levels of amino acids, such as leucine, isoleucine, lysine, glutamine and aspar-agine. High levels of membrane phospholipid metabolites, including glycerophosphoryl choline andphosphoryl choline were found in the ileum, colon, liver and spleen of infected mice. Additionally, lowlevels of energy-related metabolites, including lipids, glucose and glycogen were observed in ileum,spleen and liver samples of infected mice. Energy-related metabolites in the jejunum, liver and renalmedulla were found to be positively correlated with S. mansoni worm burden upon dissection. These find-ings show that a patent S. mansoni infection causes clear disruption of metabolism in a range of tissues ata molecular level, which can be interpreted in relation to the previously reported signature in a biofluid(i.e. urine), giving further evidence of the global effect of the infection.

� 2008 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Schistosomiasis is a complex of acute and chronic infectionscaused by a digenetic trematode flatworm of the genus Schistosomathat is widespread in tropical and sub-tropical environments(Gryseels et al., 2006; King and Dangerfield-Cha, 2008). An esti-mated 779 million individuals are at risk of schistosomiasis andmore than 200 million are infected, yet schistosomiasis is often ne-glected (Steinmann et al., 2006). Clinical symptoms of acute andchronic schistosomiasis include abdominal pain, anorexia, nausea

sitology Inc. Published by Elsevier

tory of Magnetic ResonanceMagnetic Resonance, Wuhancademy of Sciences, Wuhan

7143; fax: +86 27 8719 9291.).

and diarrhoea, caused by the disturbance of gastrointestinal motil-ity (Bogers et al., 2000; De Man et al., 2002). Enlargement of liverand spleen typically occurs in chronic cases (Ross et al., 2001).The pathogenesis of schistosomiasis is largely due to inflammatoryreactions caused by trapped schistosome eggs. An adult femaleSchistosoma mansoni worm, for example, produces �300 eggs perday (Moore and Sandground, 1956). Whilst about half of the eggsare excreted via faeces, the remaining eggs are trapped in the intes-tine and liver where they trigger a host immune response and sub-sequently cause granuloma formation around the eggs, whichultimately leads to fibrosis in these tissues (Boros, 1989; El-Garem,1998; De Man et al., 2002; Ross et al., 2002; Gryseels et al., 2006).

The pathology of the intestine, liver and spleen due to a Schisto-soma spp. infection, both in experimental animals and humans, hasbeen investigated by histology, ultrasonography and magnetic res-onance imaging (Freitas et al., 1999; Bogers et al., 2000; De Man

Ltd. All rights reserved.

mailto:yulan.wang@wipm.ac.cnhttp://www.sciencedirect.com/science/journal/00207519http://www.elsevier.com/locate/ijpara

548 J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558

et al., 2002; Shen et al., 2002; Bezerra et al., 2007; Lambertucciet al., 2008). Although histological techniques can offer visualiza-tion of differences between healthy and infected individuals, theyfail to provide information at the molecular level, which mightbe important for an enhanced understanding of physiological andpathological consequences of a schistosome infection.

Metabonomic strategies find increasing application in systemsbiology, as they provide insight into metabolic responses of wholeorganisms to environmental, genetic and other factors at themolecular level, such as physiological and pathological stimuli,and drug and natural product interventions (Nicholson et al.,1999, 2002; Brindle et al., 2002; Lindon et al., 2004; Wang et al.,2005b; Schlotterbeck et al., 2006). Metabonomics typically em-ploys high-resolution nuclear magnetic resonance (NMR) spectros-copy or mass spectrometry (MS), coupled with multivariate dataanalysis to explore and interpret complex spectroscopic data de-rived from biofluid or tissue samples. One of the advantages ofNMR-based metabolic profiling is that small intact tissue samplescan also be analysed with the aid of magic angle spinning (MAS).In MAS the major line broadening factors such as dipole–dipoleinteractions, NMR chemical shift anisotropy and magnetic fieldinhomogeneities are attenuated, thus producing 1H NMR spectracomparable to those observed in the solution state NMR spectros-copy. MAS can therefore provide a metabolic link between histol-ogy and metabolic profiling of biological samples, and generateuseful information on the metabolic consequences or mechanismsof external stimuli (Cheng et al., 1997; Waters et al., 2001).

A metabolic profiling strategy has been successfully applied tothe investigation of altered urinary metabolites induced by aninfection with either S. mansoni or Schistosoma japonicum in mouseand hamster respectively. The urinary metabolic fingerprint in-cluded a stimulated glycolysis process, reduced levels of the tricar-boxylic acid cycle intermediates, altered amino acid metabolismand disturbance of the population and/or activities of gut microbi-ota (Wang et al., 2004, 2006). In order to relate the metabolic sig-nature of a S. mansoni infection to functional and structuralchanges in host tissues, and to obtain a deeper metabolic charac-terization of the infection, we examined the metabolic responsesof different tissue samples obtained from mice with a patent S.mansoni infection and compared the metabolite profiles with thosefrom non-infected control mice.

2. Materials and methods

2.1. Host-parasite model

All experiments were carried out at the Swiss Tropical Institute(Basel, Switzerland), adhering to local and national guidelines andlaws of experimental work with laboratory animals (PermissionNo. 2081). A total of 20 female mice (NMRI strain), aged �3 weeks,were purchased from RCC (Itingen, Switzerland). Mice were ran-domly housed into four cages of five mice each and individuallymarked. Mice were acclimatized for 2 weeks, hence, at the onsetof the experiment mice were �5 weeks old, weighing between17.3 and 27.3 g. Mice were kept under environmentally-controlledconditions (temperature, 22 �C; relative humidity, 60–70%; light/dark cycle, 12/12 h) and were fed on rodent pellets (Nafag; Gossau,Switzerland) and water ad libitum. Half of the mice were infecteds.c. with �80 S. mansoni cercariae each. The remaining mice wereleft uninfected, and hence served as controls.

2.2. Tissue sample collection

Mice were killed by cervical dislocation on day 74 p.i. Controlanimals were killed at the same time point. Three types of intesti-nal tissues were collected, namely jejunum, ileum and colon. Small

samples (�5 mm) were obtained from the middle part of each tis-sue in order to minimise variation due to possible topographicaldifferences. Intestinal tissues were washed three times with PBSbefore freezing. Additionally, a small portion of the liver(�15 mg) was obtained from the edge of the left lobe, a smallspleen sample (�5 mm3) was cut from the distal part of the spleen,and the left kidney was removed from each mouse. Tissue sampleswere transferred into cryo-tubes, labelled with unique identifiers,immersed in liquid nitrogen and stored at �80 �C pending spectro-scopic analyses.

2.3. Histology and S. mansoni worm burden

For histological investigations, the right kidney and a small por-tion of the spleen were obtained and transferred in Eppendorftubes containing 4% formalin. Tissue samples were cut, stainedwith H & E and then examined under a light microscope. The re-sults from tissue samples obtained from S. mansoni-infected micewere compared with those recovered from non-infected controlanimals.

In the infected mice, the S. mansoni worm burden was quanti-fied upon dissection by examination of the liver and the mesentericveins surrounding the intestine, as described elsewhere (Utzingeret al., 2002).

2.4. 1H MAS-NMR spectroscopic analysis of tissue

The kidney was separated into renal cortex and medulla, andsamples of �15 mg were taken from each region. Each samplewas isolated and packed into a 4 mm outer diameter zirconia ro-tor (Bruker Analytische GmbH; Rheinstetten, Germany) with sal-ine D2O as the field lock. 1H NMR spectra were acquired on a600 MHz Bruker DRX spectrometer (Rheinstetten, Germany),operating at 600.13 MHz, equipped with a triple-resonanceMAS probe. The spin rate of the rotor was regulated at5000 Hz at 283 K. A 1H NMR spectrum was acquired from eachsample with Carr–Purcell–Meiboom–Gill (CPMG) pulse sequence[relaxation delay (RD)-90��(s�180��s)n-acquire free inductiondecay (FID)] in order to attenuate signals from macromoleculeswith a short spin–spin relaxation time. A spin relaxation delay(2 ns) of 200 ms was used. The 90� pulse length was adjustedto 10 ls. The spectral width was 20 ppm and 128 transientswere collected into 32 k data points. An exponential functionwas applied to the FID prior to Fourier transformation with aline broadening factor of 0.3 Hz.

Two-dimensional (2D) 1H–1H correlation spectroscopy (COSY)and 1H–1H total correlation spectroscopy (TOCSY) NMR spectrawere acquired from selected samples for structural assignment ofunknown metabolites. A total of 80 transients and 256 incrementswere collected into 2 k data points with a spectral width of10 ppm. COSY and TOCSY data were multiplied by an unshiftedsine-bell and a shifted sine-bell apodization function, respectively,prior to Fourier transformation.

2.5. Data reduction and multivariate data analysis

1H MAS-NMR spectra obtained from intestinal tissues weremanually phased and corrected for baseline distortions. The spec-tra were referenced using the signal from anomeric a-glucose pro-ton at d 5.223. The entire spectrum (d 0.0–10.0) was digitized into�40,000 data points with a resolution of 0.0005 ppm using an in-house developed MATLAB script (version 7.0). The region betweend 4.70 and d 5.20 was removed in order to minimize the effect ofthe imperfect baseline caused by the water suppression. In addi-tion, regions d 0.0–0.3 and d 9.0–10.0 containing only noise wereremoved.

J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558 549

Normalization to the total area of the remaining NMR spectrumwas performed prior to multivariate data analysis. A supervisedmultivariate data analysis tool, i.e. orthogonal projection to latentstructure-discriminant analysis (O-PLS-DA) (Trygg and Wold,2002), was applied to the analysis of 1H NMR spectral data in aMATLAB environment. O-PLS-DA models were constructed usingNMR data as the X matrix and the infection status (i.e. no infectionversus infection) as the dummy Y matrix. One PLS component wascalculated for each model with one orthogonal component usingmean-centred data scaled to unit variance. The validation of themodel was conducted using a sevenfold cross validation, i.e. itera-tive construction of models by repeatedly leaving out one-seventhof the samples and, subsequently, predicting them back into themodel. The total explained variation for the X matrix is indicatedby the goodness of fit (R2X) value and the corresponding cross-val-idation parameter is expressed as Q2Y.

We used a multiblock hierarchical principal component analysis(H-PCA) and a hierarchical PLS-DA (H-PLS-DA) as a method forcomparing several blocks (i.e. several PCA models from multipletissues) derived from the same ‘objects’ (e.g. host animal). Thismethod is ideally used for analysing variable-rich datasets (West-erhuis et al., 1998; Eriksson et al., 2004). Our data were dividedinto seven blocks according to the type of tissue samples (i.e. jeju-num, ileum, colon, liver, spleen, renal medulla and renal cortex)obtained from the same mouse. A PCA model was constructed forthe NMR data of each tissue type. All meaningful scores (tb) fromeach PCA model (as determined using the optimal goodness of pre-diction (Q2Y) value) were combined into a super block, T. PCA orPLS-DA was then performed on T with tb denoted as variablesand samples denoted as observations. The resulting super scores(tT) plot shows the relationship between observations, and thesuper loading plot (pT) indicates which scores (tb) are most influen-tial on the H-PCA/H-PLS-DA model, and hence facilitates visualiza-tion of differences and/or similarities in the metabolic responses ofmultiple tissues or organs.

3. Results

3.1. Gross observation, histology and S. mansoni worm burden

The S. mansoni worm burden in mice was assessed 74 days p.i.with �80 cercariae each. On average, 36 worms (SD = 7) werefound in each mouse. There was no significant difference in themean body weight between non-infected control and S. mansoni-infected mice over the course of the entire experiment. Two micein the infected group died at days 67 and 70 p.i.

Upon dissection of mice, thicker and rougher intestinal surface,enlargement of spleen and liver was observed in all S. mansoni-in-fected animals. Histological examination of spleen and kidneysamples obtained from uninfected control mice showed normalcellular architecture (Fig. 1A and B). Spleen samples obtained fromS. mansoni-infected mice showed pronounced lymphofollicularhyperplasia with markedly increased neutrophilic granulocytes(Fig. 1C), including splenic microabscesses. With the exception ofone kidney characterized by pyelonephritis (Fig. 1D), most likelydue to sepsis, the kidneys of the remaining S. mansoni-infectedmice appeared normal.

3.2. 1H MAS-NMR spectra of intact tissue samples

Representative 1H MAS-NMR CPMG spectra from tissues (jeju-num, ileum, colon, liver, spleen, renal cortex and renal medulla)from control mice are shown in Fig. 2 and provide a comprehensivedescription of tissue-specific low molecular weight componentscomprising these tissues. A number of metabolites, including arange of amino acids, acetate, creatine, glycerophosphoryl choline

(GPC), choline, uracil, glucose, fumarate, inosine, phosphoryl cho-line (PC), nicotinurate, ascorbate, lipids and scyllo-inositol wereidentified based on 2D 1H NMR spectra and literature information(Fan, 1996; Wang et al., 2005a). Our findings concur with previouspublications characterizing the metabolic profiles of the intact tis-sues investigated here. The chemical shift and peak multiplicitiesof all identified metabolites in each tissue type are listed in Table 1.

3.3. Multivariate data analysis of 1H MAS-NMR CPMG tissue spectra

Metabolic changes in each tissue type induced by a S. mansoniinfection were established using an O-PLS-DA strategy, comparing1H NMR CPMG spectra obtained from infected mice and corre-sponding controls. The total explained variation for the X matrix(R2X) and the corresponding cross-validation parameter (Q2Y) foreach model are listed in Table 2. Clear separation was achieved be-tween samples obtained from non-infected controls and S. man-soni-infected mice for all tissues examined, as evidenced by theconsistently high Q2Y values for all models.

The cross-validated scores plots depicting the distribution ofsamples based on their metabolic profiles are shown in Figs. 3and 4 (left-hand side) together with the corresponding O-PLS-DAcoefficient plots (right-hand side). These coefficient plots indicatethe key metabolites giving rise to differentiation due to an infec-tion with S. mansoni. Upward oriented peaks in the O-PLS-DA coef-ficient plots denote a relative high levels of metabolites in theinfected group, whereas downward oriented peaks indicate a lowlevels of the metabolites. The colours on the plots represent thesignificance of metabolites in differentiation between S. mansoni-infected and non-infected control mice; red indicating a strongcovariation with the infection, and blue indicating no or only littlerelation with the infection.

Table 2 summarizes those metabolites that were significantlycorrelated with a S. mansoni infection and their correspondingcoefficients. An infection with S. mansoni induced similar meta-bolic effects on the jejunum and ileum, which included high levelsof amino acids, such as glutamine, lysine, methionine, phenylala-nine and tyrosine. Differential effects between these two tissuesincluded low levels of choline, fumarate and scyllo-inositol in thejejunum and low levels of lipids together with relatively high levelsof creatine, GPC, inosine, PC and xanthine in the ileum of S. man-soni-infected mice. Marked alterations in metabolic profiles of co-lon tissue samples were also observed, including high levels of GPCand PC, and a low level of scyllo-inositol in the infected mice.

The liver and spleen of S. mansoni-infected mice manifested asimilar metabolic response, characterized by low levels of glucoseand lipids, and high levels of amino acids, including alanine, aspar-agine, creatine, glutamine, glycine, lysine, methionine, phenylala-nine and proline. Moreover, relative levels of ascorbate, fumarateand pyruvate were found to be higher in the livers of infected mice.The S. mansoni infection-induced changes in the metabolic profilesof both the renal medulla and renal cortex were characterized byhigh levels of ascorbate, asparagines, glutamine and lysine, andmarkedly low levels of choline and scyllo-inositol. Additionally,levels of ethanolamine and inosine were found to be lower in renalcortex obtained from infected mice.

3.4. Correlation of 1H NMR CPMG spectral data with S. mansoni wormburden

An O-PLS model was constructed using NMR spectra obtainedfrom tissues as X matrix and S. mansoni worm burden as the Ymatrix to enhance the recovery of metabolic features associatedwith severity of infection. One predictive component with oneorthogonal component was calculated for each model. The resultsare summarized in Fig. 5, showing that there was a significant

Fig. 1. Histology of spleen (A) and kidney (B) obtained from non-infected control mice. In mice infected with Schistosoma mansoni for 74 days, there was a markedlymphofollicular hyperplasia with prominent germinal centres with tangible body mocrophages in the spleen (C). Only one kidney from the infected mice was characterizedby pyelonephritis (D). All tissue samples were stained with H & E; and magnifications for images A, B, C and D are 100�, 40�, 100� and 400�, respectively.

550 J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558

correlation of the metabolite profiles for jejunum, liver and renalmedulla with actual worm burden. The quality of the models, asexpressed by Q2Y and R2X, are summarized in Table 3. Generally,levels of various amino acids in these tissues were positively corre-lated with worm burden, whereas levels of glucose were positivelyassociated with worm burden in renal medulla and jejunal tissues,but showed a negative association in liver tissue. In addition, gly-cogen levels in the liver and membrane component metabolitesin the renal medulla, e.g. choline and GPC, were negatively relatedto S. mansoni worm burden.

3.5. PCA and H-PCA of tissue data

In order to assess the nature of the metabolic response to infec-tion across several tissues, PCA models were constructed individu-ally for each biological matrix and the significant scores from eachmodel combined into a single new matrix to include all tissues. A to-tal of three principal components were calculated for unit variance-scaled data obtained from jejunum and ileum and two principalcomponents for the liver, renal cortex, renal medulla, spleen and co-lon. The number of components was decided by the point where Q2Yreached a maximum value before declining. For ileum, the PCA mod-el explained two-thirds of the total variance in the data matrix. Morethan 50% of the total variance was explained by PCA models for jeju-num (62.8%) and liver (52.0%). Considerably lower percentages ofthe total variance were explained by PCA models for colon, spleen,renal cortex and renal medulla; i.e. 45.7%, 44.9%, 37.6% and 31.9%,respectively, indicating a more variable or less coherent responsethan in the ileum, jejunum and liver.

The scores (tb) values for each model were combined into a‘super block’ (T) and then PCA was performed on the super blockwith unit variance-scaled data, which is termed H-PCA (Wester-huis et al., 1998). A total of two principal components were calcu-lated for the H-PCA model with 50.8% of total variance beingexplained. The H-PCA scores plot (Fig. 6A) showed a clear separa-tion between the S. mansoni-infected and the non-infected controlgroup of mice along the first principal components. To maximizethe separation, a PLS-DA model was calculated for the same unitvariance-scaled ‘super block’ dataset by using class information(i.e. infected and non-infected control) as the Y variable, withtwo PLS components. The H-PLS-DA scores plot (Fig. 6B,Q2Y = 0.94) also showed clear discrimination between the infectedand the non-infected control group along the first principal compo-nents. The corresponding loadings plot (Fig. 6C) represented asso-ciations between scores (tb) (e.g. relating to the tissues which weremost metabolically disrupted by the infection). The intestinal tis-sues, most clearly defined by the first (t[1]) and third (t[3]) compo-nents for the jejunum, the second (t[2]) component for ileum andcolon and the first (t[1]) component of renal medulla were thehighest positively weighting in the model and therefore, most dis-criminatory for the infected tissues. The first (t[1]) component ofthe liver, spleen and ileum, and the second (t[2]) component ofthe renal cortex exerted the most significant influence on the posi-tion of the control profiles in the scores plot. In order to investigatethe key metabolic changes in each tissue, the corresponding load-ing plots of these important scores (tb) were plotted (Figs. 6D–L). S.mansoni-infected mice showed higher levels of alanine, GPC, iso-leucine, leucine, PC and valine in jejunum, ileum and colon, and

6789 ppm 1.52.02.53.03.54.04.5 ppm

A

B

C

D

2630

30

34

19531

4

1822

7,11&13

18

19&31

14 12

129

4&5

1&2

28

25

24

28 2717

39

740 41

10&11 7&8

21

36

35

29 35

32 (CH=CH)

38 3842

16

15 136

37 2337

2320

20

3

43

33

14

32

δ 1H

Fig. 2. Four typical 600 MHz 1H magic angle spinning-nuclear magnetic resonance Carr–Purcell–Meiboom–Gill (MAS-NMR CPMG) spectra obtained from jejunum (A), liver(B), spleen (C) and renal medulla (D) tissues of non-infected control mice aged �15 weeks. Key: 1, leucine; 2, iso-leucine; 3, valine; 4, lactate; 5, threonine; 6, alanine; 7,lysine; 8, arginine; 9, acetate; 10, proline; 11, glutamate; 12, methionine; 13, glutamine; 14, aspartic acid; 15, asparagine; 16, ethanolamine; 17, glycine; 18, creatine; 19,GPC; 20, choline; 21, taurine; 22, serine; 23, uracil; 24, tyrosine; 25, phenylalanine; 26, a-glucose; 27, b-glucose; 28, deoxyuridine; 29, fumarate; 30, inosine; 31, PC; 32,lipids; 33, scyllo-inositol; 34, xanthine; 35, cytidine; 36, formate; 37, nicotinurate; 38, ascorbate; 39, trimethylamine-N-oxide (TMAO); 40, trimethylamine (TMA); 41,pyruvate; 42, glycerol; 43, glycogen.

J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558 551

higher levels of ascorbate and glycine in renal medulla, whereascontrol mice manifested higher levels of glucose and lipids in theileum, liver and spleen tissues, together with higher concentrationsof glycogen in the liver.

4. Discussion

We have previously characterized the urinary metabolic pro-files of mice infected with S. mansoni for 49 and 56 days. Preced-ing in vitro studies revealed that oviposition in S. mansoni occursafter 34–35 days (Clegg, 1965), and hence our previous findingspertain to metabolic changes due to a patent S. mansoni infec-

tion. In brief, characteristic metabolic changes included a stimu-lated glycolysis process, reduced levels of the tricarboxylic acidcycle intermediates, a metabolic disturbance of amino acidsand systematic variation in the level of microbiota-relatedmetabolites (Wang et al., 2004). Here, we extended our meta-bolic profiling strategy from urine to various tissue samples(i.e. liver, kidney, spleen, jejunum, ileum and colon) employingthe same host-parasite model in order to probe the aetiologyof the observed changes in urine and to relate the changes inthe urinary metabolic profile to potential mechanism of pathol-ogy. In addition, we have also employed an H-PCA analysis soas to efficiently connect information from multiple compart-

Table 1Identified metabolites in 1H magic angle spinning-nuclear magnetic resonance Carr-Purcell-Meiboom-Gill (MAS-NMR CPMG) spectra of jejunum, ileum, colon, liver, spleen andkidney obtained from mice with or without Schistosoma mansoni infection, together with the respective chemical shifts and signal multiplicities.

Metabolites (key) H group d1H (multiplicity)a Tissuesb

leucine (1) a-CH; half b-CH2; half b-CH2; c-CH; d-CH3; d0-CH3 3.72(t); 1.96(m); 1.63(m); 1.69(m); 0.91(d); 0.94(d) Aiso-leucine (2) a-CH; b-CH; half c-CH2; half c-CH2; d-CH3; d0-CH3 3.68(d); 1.93(m); 1.25(m); 1.47(m); 0.99(d); 1.02(d) Avaline (3) a-CH; b-CH; c-CH3; c0-CH3 3.62(d); 2.28(m); 0.99(d); 1.04(d) Alactate (4) a-CH; b-CH3 4.11(q); 1.32(d) Athreonine (5) a-CH; b-CH; c-CH3 3.59(d); 4.27(m); 1.32(d) L/S/Kalanine (6) a-CH; b-CH3 3.77(q); 1.47(d) Alysine (7) a-CH; b-CH2; c-CH2; d-CH2; e-CH2 3.78(t); 1.92(m); 1.72(m); 1.47(m); 3.03(t) Aarginine (8) a-CH; b-CH2; c-CH2; d-CH2 3.76(t); 1.92(m); 1.65(m); 3.24(t) Aacetate (9) CH3 1.91(s) Aproline (10) a-CH; half b-CH2; half b-CH2; c-CH2; d-CH2 4.13(t); 2.08(m); 2.37(m); 2.01(m); 3.38(t) Aglutamate (11) a-CH; b-CH2; c-CH2 3.76(m); 2.07(m); 2.34(m) Amethionine (12) a-CH; b-CH2; c-CH2; d-CH3 3.87(m); 2.14(m); 2.63(t); 2.13(s) Aglutamine (13) a-CH; b-CH2; c-CH2 3.78(m); 2.15(m); 2.45(m) Aaspartic acid (14) a-CH; b-CH2 3.89(m); 2.69(m); 2.80(m) Aasparagine (15) a-CH; b-CH2 4.01(m); 2.87(m); 2.95(m) Aethanolamine (16) NHACH2; HOACH2 3.14(t); 3.82(t) Aglycine (17) a-CH 3.55(s) Acreatine (18) NACH3; CH2 3.03(s); 3.92(s) AGPC (19) NA(CH3)3; a-CH2; b-CH2; a’-CH2; b’-CH; c0-CH2 3.22(s); 4.32(t); 3.68(t); 3.61(dd); 3.90(m); 3.72(dd) Acholine (20) NA(CH3)3; a-CH2; b-CH2 3.20(s); 4.07(t); 3.52(t) Ataurine (21) NACH2; SO3ACH2 3.43(t); 3.27(t) Aserine (22) a-CH; b-CH2 3.84(m); 3.96(m) Auracil (23) 5-CH; 6-CH 5.80(d); 7.52(d) Atyrosine (24) 2,6-CH; 3,5-CH 7.18(d); 6.88(d) Aphenylalanine

(25)2,6-CH; 3,5-CH; 4-CH; half Ar-CH2; half Ar-CH2; NCH 7.40(m); 7.33(m); 7.35(m); 3.17(dd); 3.30(dd); 3.99(dd) A

a-glucose (26) 1-CH; 2-CH; 3-CH; 4-CH; 5-CH; half 6-CH2; half 6-CH2 5.24(d); 3.56 (dd); 3.70(t); 3.40(t); 3.83(m); 3.72(dd);3.85(m)

A

b-glucose (27) 1-CH; 2-CH; 3-CH; 4-CH; 5-CH; half 6-CH2; half 6-CH2 4.65(d); 3.25 (dd); 3.47(t); 3.40(t); 3.47(ddd); 3.78(dd);3.90(dd)

A

deoxyuridine (28) 5-CH; 6-CH 5.88(d); 7.88(d) L/S/Kfumarate (29) CH 6.52(s) Ainosine (30) 2-CH; 8-CH; 2’-CH; 4’-CH; 5’-CH; CH2(i); CH2(ii) 8.34(s); 8.24(s); 6.10(d); 4.44(t); 4.28(q); 3.92(dd);

3.85(dd)A

PC (31) NA(CH3)3; NACH2; PO3ACH2 3.22(s); 3.61(m); 4.25(m) Alipids (32) ACH3; A(CH2)A; b-CH2CH2CO; a-CH2CH2CO; CH2C@CCH2; @CH-CH2-CH@;

CH@CH0.89(bs); 1.29; 1.59; 2.25; 2.03; 2.77; 5.32 A

scyllo-inositol (33) 6� CH 3.33(s) J/I/C/Kxanthine (34) CH 7.96(s) J/I/Ccytidine (35) 5-CH; 6-CH 6.09(d); 7.85(d) L/Sformate (36) CH 8.45(s) Snicotinurate (37) 2-CH; 6-CH; 4-CH; 5-CH; CH2 8.92(s); 8.70(d); 8.24(d); 7.60(dd); 3.99(s) L/Kascorbate (38) half CH2; half CH2; CHOH; a-CH 3.76(d); 3.74(d); 4.03(dd); 4.52(d) L/S/KTMAO (39) 3� CH3 3.28(s) L/STMA (40) 3� CH3 2.89(s) Lpyruvate (41) CH3 2.38(s) L/Sglycerol (42) a-CH2; c-CH2; b-CH 3.56(dd); 3.64(dd); 3.87(m) L/S/Kglycogen (43) CH 5.40(d) L

PC, phosphorycholine; TMA, trimethylamine; TMAO, trimethylamine-N-oxide.a s, singlet; d, doublets; t, triplets; m, multiplets; q, quartet; dd, double doublet; ddd, double double doublet; bs, broad singlet.b A, all; J, jejunum; I, ileum; C, colon; L, liver; S, spleen; K, kidney.

552 J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558

ments (e.g. different types of tissues). In the current study, theinfection duration was extended to 74 days in order to ensurethat overt pathology had been established based on literaturereports.

Visual inspection of intestinal tissues revealed that the surfaceof the intestine in S. mansoni-infected mice appeared to be thickerand rougher than in non-infected control mice. These observationsare consistent with previous studies in mice that noted a diffusethickening of the smooth muscle layers of the intestine and distur-bance of the architectural structure of the mesenteric plexus andthe enteric nervous system due to a patent S. mansoni infection(Balemba et al., 2000, 2001; Bogers et al., 2000; De Man et al.,2002; De Jonge et al., 2003b). Higher levels of membrane phospho-lipid metabolites, such as GPC and PC found in the ileum and colonof infected mice in the current study may also be associated withstructural changes in the intestinal architecture. Hence, high levelsof membrane phospholipid metabolites may be responsible for cel-

lular swelling, leading to the accumulation of intracytoplasmic flu-ids in cells and cellular overgrowth, as previously reported(Blennerhassett et al., 1992; Bogers et al., 2000). For example, a re-cent study showed a transient increase of mast cells in mucosa inmice, reaching a peak value at 8 weeks p.i. (De Jonge et al., 2003a).Further effects on the intestinal structure can result from S. man-soni eggs trapped in the intestine, due to the secretion of antigenswhich cause granulomous inflammation and lead to tissue fibrosis(Lukacs et al., 1993; Weinstock, 1996). Intestinal fibrosis is partic-ularly common in the distal ileum and the proximal colon (Wein-stock, 1996) and is associated with structural changes includingthickening of the ileal wall and a decrease in gastrointestinal tran-sit, which could explain the elevated levels of GPC and PC observedin the ileum and colon from S. mansoni-infected mice in the currentstudy. Biochemical changes associated with inflammation trig-gered by the S. mansoni infection appeared to be similar to thosecaused by chronic bowel diseases. For example, biopsies of colon

Table 2Metabolites contributing to the discrimination between non-infected and Schistoma mansoni-infected mice in orthogonal projection to latent structure-discriminant analysis (O-PLS-DA) models derived from jejunum, ileum, colon, liver, spleen, renal medulla and renal cortex.

Metabolites(key)

O-PLS-DA plots

JejunumQ2Y = 0.73;

R2X = 0.45

IleumQ2Y = 0.72;R2X = 0.45

ColonQ2Y = 0.81;R2X = 0.44

LiverQ2Y = 0.94;R2X = 0.56

SpleenQ2Y = 0.86;R2X = 0.45

Renal medullaQ2Y = 0.58;R2X = 0.28

Renal cortexQ2Y = 0.54;R2X = 0.33

lysine (7) +0.77 +0.84 – +0.76 +0.72 +0.77 +0.69methionine (12) +0.66 +0.83 – +0.87 +0.87 – –glutamine (13) +0.67 +0.85 – +0.90 +0.88 +0.71 –asparagine (15) +0.90 – – +0.91 +0.88 +0.71 +0.77choline (20) �0.64 – – +0.92 +0.78 �0.71 �0.80deoxyuridine (28) +0.86 +0.87 – – +0.85 – –fumarate (29) �0.89 – – +0.67 – – –tyrosine (24) +0.89 +0.88 – – +0.97 – +0.63phenylalanine (25) +0.75 +0.90 – +0.63 +0.82 – +0.72lipids: ACH3 �0.83 �0.82 �0.85A(CH2)A �0.85 �0.84 �0.84b-CH2CH2CO �0.84 �0.75 �0.83a-CH2CH2CO – �0.77 – – – – –CH2C@CCH2 �0.85 �0.75 �0.84@CHACH2ACH@ �0.75 �0.58 �0.82CH@CH �0.87 �0.75 �0.88leucine (1) +0.73 +0.88 – +0.79 +0.79 – +0.62iIsoleucine (2) +0.71 +0.68 – +0.84 +0.80 – +0.64valine (3) +0.67 +0.73 – +0.83 +0.72 – +0.63alanine (6) – +0.68 – +0.87 +0.95 – –GPC (19) – +0.76 +0.90 +0.94 +0.73 – –taurine (21) – +0.78 – +0.72 +0.73 – –glycine (17) – +0.68 – +0.94 +0.89 +0.69 –creatine (18) – +0.72 – +0.72 +0.90 – –inosine (30) – +0.86 – – – – �0.73scyllo-inositol (33) �0.76 – �0.67 – – �0.78 �0.79PC (31) – +0.73 +0.89 +0.87 +0.87 – –xanthine (34) +0.78 – – – – –TMA (40) – – – – – +0.65 –ascorbate (38) – – – +0.86 – +0.91 +0.73lactate (4) – – – +0.79 +0.80 –cytidine (35) – – – +0.87 +0.83 –uracil (23) – – – +0.70 +0.86 –nicotinurate (37) – – – +0.77 – – +0.78a-glucose (26) – – – �0.70 �0.90 –proline (10) – – – +0.93 +0.83 –pyruvate (41) – – – +0.77 – –TMAO (39) – – – +0.84 +0.69 –formate (36) – – – – +0.77 –glycogen (43) – – – �0.80 – –ethanolamine (16) +0.86 �0.82aspartic acid (14) – – – – –

Key: + (higher) and � (lower) indicate the direction of metabolite change in S. mansoni-infected versus non-infected control mice.PC, phosphorycholine; TMA, trimethylamine; TMAO, trimethylamine-N-oxide.

J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558 553

from patients with Crohn’s disease contained high levels of cho-line-containing compounds (e.g. GPC and PC) compared with nor-mal parts of colon from the same patients (Bogers et al., 2000;Bezabeh et al., 2001).

Enteric inflammation is commonly accompanied by disorders ofthe gastrointestinal tract. Khan and Collins (2006) demonstratedthe association between inflammatory response and alteration inintestinal muscle contractility in a Trichinella spiralis-rat model.Hypercontractility can physically lead to a higher level of break-down of proteins, facilitated by the assistance of enzymes frompancreas and intestinal glands, resulting in high levels of aminoacids in the small intestine (Martin et al., 2006). However, certainhelminth parasites such as Trichuris suis have been shown to ame-liorate inflammatory bowel disease (IBD) in some patients byup-regulation of Th2 and consequent suppression of Th1 type im-mune responses (Wang et al., 2008). In the present study, the mostprominent observation was high levels of amino acids in the smallintestine (jejunum and ileum) of S. mansoni-infected mice. Hyper-

contractility of intestinal muscle of infected mice requires higherenergy compared with non-infected control mice. This require-ment of energy could be met via creatine metabolism and lipid oxi-dation. Indeed, relatively high levels of creatine were observed inthe ileum of mice infected with S. mansoni after 74 days comparedwith non-infected mice. Phosphorylcreatine serves as an energystore and, together with ADP, is transferred into creatine and ATPin order to replenish the energy requirement (De Jonge et al.,2001). In the current study, lower levels of lipids were found inthe intestine of S. mansoni-infected mice, especially in the ileum.It is known that a S. mansoni infection induces free radicals inthe liver and intestine (Kazura et al., 1981; Abdallahi et al.,1999), which leads to oxidation of lipids. Alternatively, oxidationof lipids can also provide energy for intestinal motility. Our obser-vation of a significant positive relationship between worm burdenand energy-related metabolites such as glucose, lysine, glutamineand tyrosine underscores the high energy requirement of hyper-contractivity of the small intestine of S. mansoni-infected mice.

4.5 4 3.5 3 2.5 2 1.5 1 δ1H

8 7 6

8 7 6

0

0.8

0.9

0.9

0

0

0

0

0

0

60

-60

100

-100

0

0

100

-100

40-60

100

40-40

Tcv

TYosc

infected

infected

infected

non-infected

non-infected

non-infected

71213157

3320

29

242528 28

19/31

1/23

32

67

32

137

2119

3119 31

1718

25 2430

30

2828

19/31

19

3119 31

33

34

30

1/23

O-PLS coefficients (r 2)

4.5 4 3.5 3 2.5 2 1.5 1 δ1H

8 7 6

8 7 6

0

0.8

0.9

0.9

0

0

0

0

0

0

60

-60

100

-100

0

0

100

-100

40-60

100

40-40

Tcv

TYosc A

B

C infected

infected

infected

non-infected

non-infected

non-infected

71213157

3320

29

242528 28

19/31

1/23

32

67

32

137

2119

3119 31

1718

25 2430

30

2828

19/31

19

3119 31

33

34

30

1/23

O-PLS coefficients (r 2)

Fig. 3. Cross-validated score plots (left) and orthogonal projection to latent structure-discriminant analysis (O-PLS-DA) coefficient plots (right) derived from 1H NMR CPMGspectra of jejunum (A), ileum (B) and colon (C) from non-infected (blue) and Schistosoma mansoni-infected mice (red). Of note, a total of eight mice are shown in the infectedgroup since two mice died before dissection, and nine mice are shown in the control group since one animal was moribund and, therefore, killed at day 48. (For key ofmetabolites, see legend to Fig. 1 and Table 1).

554 J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558

Furthermore, taurine was present in higher levels in the ileum, li-ver and spleen of infected mice, an observation that is in contrastto the decreased levels of taurine found in the ileum of rats in otherhelminth infections such as after infection with Hymenolepis dimin-uta (Novak et al., 1993). Taurine is a multifunctional metaboliteand has been shown to have antioxidant (Das et al., 2008) andanti-inflammatory effects (Son et al., 1998) and is also known tobe involved in several Ca2+-dependent mechanisms, such as mus-cular contractility, control of cell volume and tissue osmolality(Huxtable, 1992; Martin et al., 2006). Therefore, elevated levelsof taurine in the infected mice may play a protective role againstfree radical damage and cell expansion caused by inflammation.Interestingly, taurine conjugated bile acids such as taurocholic,and tauroursodeoxycholic acid have been shown to increase eggproduction of S. mansoni in in vitro systems (Badr et al., 1999),which may relate to a host-parasite evolutionary strategy.

It is widely acknowledged that a S. mansoni infection causesinflammation of the liver and eventually leads to liver fibrosis, acrucial feature of chronic schistosomiasis in humans (Lambertucciet al., 2000). Liver fibrosis – regardless of its cause – is often asso-ciated with the deposition of extracellular matrix proteins, includ-ing collagen and a number of glycoproteins (Maher and Mcguire,1990). We found that levels of free proline, a precursor of hydroxy-proline, as well as glycine and alanine, which are major aminoacids in the synthesis of collagen, were elevated in the liver of in-fected mice, which might imply stimulated collagen synthesisactivity. Previous in vitro investigations also proved that collagen

synthesis by fibrotic liver slices grew as the concentration of freeproline was increased in the medium (Dunn et al., 1977). The highlevel of free proline could be partially attributed to the productionof S. mansoni eggs as shown in a previous study documenting a dis-tinct increase in the proline production-involved enzyme activitiesof ornithine-d-transaminase and D0-pyrroline-5-carboxylic acidreductase in S. mansoni eggs (Isseroff et al., 1983). Furthermore, ahigh level of ascorbate and a reduced level of lipid were found inthe liver of infected mice. Ascorbate is a reducing agent and playsan important role in free radical damage. The reduced levels of li-pid could be due to the oxidation damage caused by free radicalsgenerated in the infection and the oxidation of lipid could also bean important energy source when glucose and glycogen are largelydepleted as observed in the liver and spleen in the current study.Alternatively, Cho and colleagues (2001) reported that the ratioof glutamine and glutamate complex to lipid increased as theseverity of liver fibrosis progressed to the later stage, which is con-sistent with our results of higher levels of glutamine together withthe lower concentration of lipids in the liver from infected mice. Anegg granulomatous-induced liver fibrosis not only results in astructural alteration (Silva et al., 2000), but also causes functionalalteration, such as decreased oxidative deamination of amino acidsand a disturbance in ammonia metabolism (Daugherty et al.,1954). Previously, a decreased expression of enzymes related tothe citric acid cycle, the fatty acid cycle, the urea cycle and aminoacid metabolism and catabolism has been shown in mice 8 weeksp.i. with S. mansoni (Harvie et al., 2007). Here, a low level of glyco-

0

0.9

0.9

0

0

TYosc infected

non-infected

0

-60

60

32

39

20

19&31

71513 10

127

6

21

17

18

38

31420

25

37

1,2&3

32&4324

29 3523

8 67

infected

non-infected 32

39

19&2031

71513 10

127

621

17

1819 314

20

25

1,2&3

32

24 29 3528&35

8 67

197

36 28 23

140

-60

60

O-PLS coefficients (r 2)

4.5 4 3.5 3 2.5 2 1.5 1 δ1HTcv

3817

33 20

15 1340

7

16

33 20

7 1,2&3infected

non-infected

infected

non-infected

0-40 40

38

0.9

0.9

0

0

0

-100

-60

60

0

0.9

0.9

0

0

TYosc infected

non-infected

0

-60

60

A

32

39

20

19&31

71513 10

127

6

21

17

18

38

31420

25

37

1,2&3

32&4324

29 3523

8 67

infected

non-infected 32

39

19&2031

71513 10

127

621

17

1819 314

20

25

1,2&3

32

24 29 3528&35

8 67

197

36 28 23

140

-60

60

O-PLS coefficients (r 2)

4.5 4 3.5 3 2.5 2 1.5 1 δ1HTcv

3817

33 20

15 1340

7

16

33 20

7 1,2&3infected

non-infected

infected

non-infected

0-40 40

38

B

C

D

0.9

0.9

0

0

0

-100

-60

60

Fig. 4. Cross-validated score plots (left) and orthogonal projection to latent structure-discriminant analysis (O-PLS-DA) coefficient plots (right) derived from 1H NMR CPMGspectra of liver (A), spleen (B), renal cortex (C) and renal medulla (D) from non-infected (blue) and Schistosoma mansoni-infected mice (red). (For key of metabolites, seelegend to Fig. 1 and Table 1).

J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558 555

gen was found in the liver of S. mansoni-infected mice, which mightbe explained by either the inhibition of glycogen synthesis or anenhanced glycogen breakdown. It has been reported that glycogenphosphorylase for glycogenolysis was suppressed in mice due to aninfection with S. mansoni (Ahmed and Gad, 1995). Moreover, stim-ulated glycolysis in S. mansoni-infected mice has been previouslyreported (Wang et al., 2004). Therefore, it is likely that the infec-tion or infection-induced host responses lead to both the inhibitionof the glycogen synthesis and stimulation of the glycogen break-down, resulting in the depletion of the glycogen level in the liverrecovered from S. mansoni-infected mice. Independent enzymaticassay on activities of glycogen phosphorylase and UDP-glucosepyrophosphorylase would verify whether the reduction in levelsof glycogen is caused by both processes simultaneously or domi-nated by only one of these processes.

Schistosoma mansoni-induced inflammation is also reflected inan enlargement of the spleen, which was confirmed in our studyby gross observation upon dissection of mice and further under-scored by histological examination. Spleen enlargement is hypoth-esised to result from passive congestion of portal hypertensionand/or cellular hyperplasia (Ross et al., 2001). Biochemical features

associated with a S. mansoni infection included higher concentra-tions of glutamine, coupled with lower levels of lipids and glucosein the spleen from infected mice. Glutamine is an essential energysource in mitochondria and can also be utilised by lymphocytes(Klimberg and McClellan, 1996; Yaqoob and Calder, 1997). There-fore, the high level of glutamine may indicate a hyperactive im-mune system triggered by S. mansoni eggs trapped in the tissue.Glycolysis is also carried out in lymphocytes and macrophages inorder to meet the energy requirement of a stimulated immune sys-tem due to an infection. As a consequence, the low level of glucoseand the high level of lactate would be expected, consistent withour observations in the spleen of infected mice.

An epidemiological investigation carried out among schoolchil-dren in an area highly endemic for schistosomiasis found that renaldisease was uncommon (Johansen et al., 1994). On the other hand,although neither the adult worms nor the eggs of S. mansoni directlycaused a pathogenic effect in the kidney of hamsters, there was animmunological response at 6 weeks p.i., characterized by the depo-sition of immune complex and renal amyloidosis (Sobh et al., 1991).In our study, we found high levels of glutamine, GPC and choline inthe renal medulla of S. mansoni-infected mice, which were positively

0

0.9

0

0.9

0

0.9

4.5 4 3.5 3 2.5 2 1.5 1 δ1H

O-P

LS c

oeffi

cien

ts(a

.u.)

0

0.04

-0.04

0

-0.01

0.01

0.04

-0.04

0

8 7 6

2726 27 7 13 7

2426

6

2726 27

27

19

20 19&20

19&20

13

0

0.9

0

0.9

0

0.9

4.5 4 3.5 3 2.5 2 1.5 1 δ1H

O-P

LS c

oeffi

cien

ts(a

.u.)

0

0.04

-0.04

0

-0.01

0.01

0.04

-0.04

0

8 7 6 0

0.9

0

0.9

0

0.9

4.5 4 3.5 3 2.5 2 1.5 1 δ1H

O-P

LS c

oeffi

cien

ts(a

.u.)

0

0.04

-0.04

0

-0.01

0.01

0.04

-0.04

0

8 7 6

A

B

C

2726 27 7 13 7

2426

6

2726 27

27

19

20 19&20

19&20

13

Fig. 5. O-PLS plots derived from 1H magic-angle-spinning Carr–Purcell–Meiboom–Gill nuclear magnetic resonance (MAS CPMG NMR) spectra of jejunum (A), liver (B) andrenal medulla (C) from Schistosoma mansoni-infected mice using worm burden as the class descriptor (Y matrix).

Table 3Metabolites that were correlated with Schistosoma mansoni worm burden upon dissection of mice, and Q2Y and R2X of each model from jejunum, liver and renal medulla.

Metabolites (key) O-PLS plots

JejunumQ2Y = 0.61;R2X = 0.46

LiverQ2Y = 0.40;R2X = 0.41

Renal medullaQ2Y = 0.45;R2X = 0.35

b-glucose (27) +0.87 �0.54 +0.69lysine (7) +0.70 � �glutamine (13) +0.64 � +0.85a-glucose (26) +0.86 �0.67 �tyrosine (24) +0.61 � �alanine (6) � +0.70 �GPC (19) � � �0.81choline (20) � � �0.67glycogen (43) � �0.79 �

556 J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558

correlated with worm burden upon dissection of mice. Glutamineand GPC have been reported to have osmotic function in the brainand inner kidney, respectively (Miller et al., 2000). These observa-tions may indicate a counterbalance response and dysfunction ofthe kidney due to the parasitic infection. However, with the excep-tion of one mouse, the cellular architecture in the kidney of theremaining S. mansoni-infected mice appeared normal.

Metabolic effects were observed in various intestinal tissues, li-ver, spleen and kidney of mice with a patent S. mansoni infection asrevealed by a metabolic profiling strategy. Our findings indicate

both localized and remote effects of the infection. We observedevidence of inflammation and hypercontractility of the intestine,damage to the liver and kidney and enlargement of the spleen ininfected mice. These findings demonstrate that the integrated mul-tivariate strategy employed provide a holistic approach to meta-bolic effects due to a S. mansoni infection in host organs at themolecular level and indicate a coordinated response in several tis-sues in response to the infection. Further work on enzymaticactivities of glycogen phosphorylase and UDP-glucose pyrophos-phorylase as well as total collagen content in the liver of S. man-

D Med.t[1]

5.5 3.5 1.5 δ H1

G Ile.t[2]

H Col.t[2]

F Jej.t[3]

E Jej.t[1]

I Liv.t[1]

J Spl.t[1]

L Cox.t[2]

K Ile.t[1]

t[2]

t[1]

t[2]

t[1]0

4

0

2

-4

-2

42 6-4 -2-60 -4-2 -64 26

4

0

2

-4

-2

A B

Jej.t[1]

Jej.t[2]

Jej.t[3] Ile.t[1]

Ile.t[2]

Ile.t[3]

Col.t[1]Col.t[2]

Liv.t[1]

Liv.t[2]

Spl.t[1]Spl.t[2]

Cox.t[1]

Cox.t[2]

Med.t[1]

Med.t[2]

p[2]

0.4

0

0.2

-0.4

-0.2

-0.6

p[1]0 0.20.1 0.3-0.2 -0.1-0.3 0.4

C

2033

3817

1,2,3

33

712

1,2,312

1,2,3619,3119 31

21

7

19,311931

1,2,3

33

3232

32

619,312119

3232

32

171,2,3

3817

323232

2643

6

6

5.5 3.5 1.5 δ H1 5.5 3.5 1.5 δ H1

Fig. 6. Hierarchical principal component analysis (H-PCA) scores plot (A), hierarchical projection to latent structure-discriminant analysis (H-PLS-DA) scores plot (B) and H-PLS-DA loadings plot (C) obtained from scores of sub-PCA models which were separately derived from seven different tissues from S. mansoni infected (red) and non-infectedcontrol mice (blue). Sub-PCA loadings (Med.t[1], Jej.t[1], Jej.t[3], Ile.t[2], Col.t[2], Liv.t[1], Spl.t[1], Ile.t[1] and Cox.t[2]) holding significant contributions to the separation areshown in (D–L). Key: Med, renal medulla; Jej, jejunum; Ile, ileum; Col, colon; Liv, liver; Spl, spleen; Cox, renal cortex; for metabolities, refer to legend in Fig. 1 and Table 1.

J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558 557

soni-infected mice should provide in-depth understanding of theaetiology of the disease.

Acknowledgements

The authors thank Dr. Jacques Chollet for infection and dissec-tion of mice, and Dr. O. Cloarec for access to MATLAB scripts foranalysis of NMR data. This study received financial support fromthe Swiss National Science Foundation (J.V. Li, J. Saric and J. Utzin-ger, Project Nos. PPOOB–102883 and PPOOB–119129; J. Keiser,Project No. PPOOA–114941), Deputy Rector’s Award at ImperialCollege London (J.V. Li) and The Chinese Academy of Science (Y.Wang, KJXC2-YW-W11).

References

Abdallahi, O.M.S., Hanna, S., De Reggi, M., Gharib, B., 1999. Visualization of oxygenradical production in mouse liver in response to infection with Schistosomamansoni. Liver 19, 495–500.

Ahmed, S.A., Gad, M.Z., 1995. Effect of schistosomal infection and its treatment onsome key enzymes of glucose metabolism in mice livers. Arzneimittelforschung45, 1324–1328.

Badr, S.G.E., Pica-Mattoccia, L., Moroni, R., Angelico, M., Cioli, D., 1999. Effect of bilesalts on oviposition in vitro by Schistosoma mansoni. Parasitol. Res. 85, 421–423.

Balemba, O.B., Mbassa, G.K., Assey, R.J., Kahwa, C.K.B., Makundi, A.E., Hay-Schmidt,A., Dantzer, V., Semuguruka, W.D., 2000. Lesions of the enteric nervous systemand the possible role of mast cells in the pathogenic mechanisms of migrationof schistosome eggs in the small intestine of cattle during Schistosoma bovisinfection. Vet. Parasitol. 90, 57–71.

Balemba, O.B., Semuguruka, W.D., Hay-Schmidt, A., Johansen, M.V., Dantzer, V.,2001. Vasoactive intestinal peptide and substance P-like immunoreactivities inthe enteric nervous system of the pig correlate with the severity of pathologicalchanges induced by Schistosoma japonicum. Int. J. Parasitol. 31, 1503–1514.

Bezabeh, T., Somorjai, R.L., Smith, I.C.P., Nikulin, A.E., Dolenko, B., Bernstein, C.N.,2001. The use of 1H magnetic resonance spectroscopy in inflammatory boweldiseases: distinguishing ulcerative colitis from Crohn’s disease. J. Gastroenterol.96, 442–448.

Bezerra, A.S.A., D’Ippolito, G., Caldana, R.P., Cecin, A.O., Ahmed, M., Szejnfeld, J.,2007. Chronic hepatosplenic schistosomiasis mansoni: magnetic resonanceimaging and magnetic resonance angiography findings. Acta Radiol. 48, 125–134.

Blennerhassett, M.G., Vignjevic, P., Vermillion, D.L., Collins, S.M., 1992.Inflammation causes hyperplasia and hypertrophy in smooth-muscle of ratsmall-intestine. Am. J. Physiol. 262, G1041–G1046.

Bogers, J., Moreels, T., De Man, J., Vrolix, G., Jacobs, W., Pelckmans, P., Van Marck, E.,2000. Schistosoma mansoni infection causing diffuse enteric inflammation anddamage of the enteric nervous system in the mouse small intestine.Neurogastroenterol. Motil. 12, 431–440.

Boros, D.L., 1989. Immunopathology of Schistosoma mansoni infection. Clin.Microbiol. Rev. 2, 250–269.

Brindle, J.T., Antti, H., Holmes, E., Tranter, G., Nicholson, J.K., Bethell, H.W.L., Clarke,S., Schofield, P.M., McKilligin, E., Mosedale, D.E., Grainger, D.J., 2002. Rapid andnoninvasive diagnosis of the presence and severity of coronary heart diseaseusing 1H NMR-based metabonomics. Nat. Med. 8, 1439–1444.

Cheng, L.L., Ma, M.J., Becerra, L., Ptak, T., Tracey, I., Lackner, A., Gonzalez, R.G., 1997.Quantitative neuropathology by high resolution magic angle spinning protonmagnetic resonance spectroscopy. Proc. Natl. Acad. Sci. U.S.A 94, 6408–6413.

Cho, S.G., Kim, M.Y., Kim, H.J., Kim, Y.S., Choi, W., Shin, S.H., Hong, K.C., Kim, Y.B., Lee,J.H., Suh, C.H., 2001. Chronic hepatitis: In vivo proton MR spectroscopicevaluation of the liver and correlation with histopathologic findings.Radiology 221, 740–746.

Clegg, J.A., 1965. In vitro cultivation of Schistosoma mansoni. Exp. Parasitol. 16, 133–147.

Das, J., Ghosh, J., Manna, P., Sil, P.C., 2008. Taurine provides antioxidant defenseagainst NaF-induced cytotoxicity in murine hepatocytes. Pathophysiology 15,181–190.

Daugherty, J., Garson, S., Heyneman, D., 1954. The effect of Schistosoma mansoniinfections on liver function in mice. I. amino acid oxidase and ammonia. Am. J.Trop. Med. Hyg. 3, 511–517.

De Jonge, F., Van Nassauw, L., Adriaensen, D., Van Meir, F., Miller, H.R.P., Van Marck,E., Timmermans, J.P., 2003a. Effect of intestinal inflammation on capsaicin-

558 J.V. Li et al. / International Journal for Parasitology 39 (2009) 547–558

sensitive afferents in the ileum of Schistosoma mansoni-infected mice.Histochem. Cell Biol. 119, 477–484.

De Jonge, F., Van Nassauw, L., De Man, J.G., De Winter, B.Y., Van Meir, F.,Depoortere, I., Peeters, T.L., Pelckmans, P.A., Van Marck, E., Timmermans, J.P.,2003b. Effects of Schistosoma mansoni infection on somatostatin andsomatostatin receptor 2A expression in mouse ileum. Neurogastroenterol.Motil. 15, 149–159.

De Jonge, W.J., Marescau, B., D’Hooge, R., De Deyn, P.P., Hallemeesch, M.M., Deutz,N.E., Ruijter, J.M., Lamers, W.H., 2001. Overexpression of arginase alterscirculating and tissue amino acids and guanidino compounds and affectsneuromotor behavior in mice. J. Nutr. 131, 2732–2740.

De Man, J.G., Chatterjee, S., De Winter, B.Y., Vrolix, G., Van Marck, E.A., Herman, A.G.,Pelckmans, P.A., 2002. Effect of somatostatin on gastrointestinal contractility inSchistosoma mansoni infected mice. Int. J. Parasitol. 32, 1309–1320.

Dunn, M.A., Rojkind, M., Warren, K.S., Hait, P.K., Rifas, L., Seifter, S., 1977. Livercollagen-synthesis in murine schistosomiasis. J. Clin. Invest. 59, 666–674.

El-Garem, A.A., 1998. Schistosomiasis. Digestion 59, 589–605.Eriksson, L., Antti, H., Gottfries, J., Holmes, E., Johansson, E., Lindgren, F., Long, I.,

Lundstedt, T., Trygg, J., Wold, S., 2004. Using chemometrics for navigating in thelarge data sets of genomics, proteomics, and metabonomics (gpm). Anal.Bioanal. Chem. 380, 419–429.

Fan, W.M.T., 1996. Metabolite profiling by one- and two-dimensional NMR analysisof complex mixtures. Prog. Nucl. Magn. Reson. Spectrosc. 28, 161–219.

Freitas, C.R.L., Barbosa, A.A., Fernandes, A.L.M., Andrade, Z.A., 1999. Pathology of thespleen in hepatosplenic schistosomiasis. Morphometric evaluation andextracellular matrix changes.. Mem. Inst. Oswaldo Cruz 94, 815–822.

Gryseels, B., Polman, K., Clerinx, J., Kestens, L., 2006. Human schistosomiasis. Lancet368, 1106–1118.

Harvie, M., Jordan, T.W., La Flamme, A.C., 2007. Differential liver protein expressionduring schistosomiasis. Infect. Immun. 75, 736–744.

Huxtable, R.J., 1992. Physiological actions of taurine. Physiol. Rev. 72, 101–163.Isseroff, H., Bock, K., Owczarek, A., Smith, K.R., 1983. Schistosomiasis – proline

production and pelease by ova. J. Parasitol. 69, 285–289.Johansen, M.V., Simonsen, P.E., Butterworth, A.E., Ouma, J.H., Mbugua, G.G.,

Sturrock, R.F., Orinda, D.A.O., Christensen, N.O., 1994. A survey for Schistosomamansoni induced kidney-disease in children in an endemic area of Machakosdistrict, Kenya. Acta Trop. 58, 21–28.

Kazura, J.W., Fanning, M.M., Blumer, J.L., Mahmoud, A.A.F., 1981. Role of cell-generated hydrogen peroxide in granulocyte-mediated killing of schistosomulaof Schistosoma mansoni in vitro. J. Clin. Invest. 67, 93–102.

Khan, W.I., Collins, S.M., 2006. Gut motor function: immunological control in entericinfection and inflammation. Clin. Exp. Immunol. 143, 389–397.

King, C.H., Dangerfield-Cha, M., 2008. The unacknowledged impact of chronicschistosomiasis. Chronic Illn. 4, 65–79.

Klimberg, V.S., McClellan, J.L., 1996. Glutamine, cancer, and its therapy. Am. J. Surg.172, 418–424.

Lambertucci, J.R., Serufo, J.C., Gerspacher-Lara, R., Rayes, A.A.M., Teixeira, R., Nobre,V., Antunes, C.M.F., 2000. Schistosoma mansoni: assessment of morbidity beforeand after control. Acta Trop. 77, 101–109.

Lambertucci, J.R., Silva, L.C., Andrade, L.M., de Queiroz, L.C., Carvalho, V.T., Voieta, I.,Antunes, C.M., 2008. Imaging techniques in the evaluation of morbidity inschistosomiasis mansoni. Acta Trop. 108, 209–217.

Lindon, J.C., Holmes, E., Bollard, M.E., Stanley, E.G., Nicholson, J.K., 2004.Metabonomics technologies and their applications in physiologicalmonitoring, drug safety assessment and disease diagnosis. Biomarkers 9, 1–31.

Lukacs, N.W., Kunkel, S.L., Strieter, R.M., Warmington, K., Chensue, S.W., 1993.The role of macrophage inflammatory protein-1-alpha in Schistosomamansoni egg-induced granulomatous inflammation. J. Exp. Med. 177, 1551–1559.

Maher, J.J., Mcguire, R.F., 1990. Extracellular-matrix gene-expression increasespreferentially in rat lipocytes and sinusoidal endothelial-cells during hepatic-fibrosis in vivo. J. Clin. Invest. 86, 1641–1648.

Martin, F.P.J., Verdu, E.F., Wang, Y.L., Dumas, M.E., Yap, I.K.S., Cloarec, O., Bergonzelli,G.E., Corthesy-Theulaz, I., Kochhar, S., Holmes, E., Lindon, J.C., Collins, S.M.,Nicholson, J.K., 2006. Transgenomic metabolic interactions in a mouse diseasemodel: interactions of Trichinella spiralis infection with dietary Lactobacillusparacasei supplementation. J. Proteome Res. 5, 2185–2193.

Miller, T.J., Hanson, R.D., Yancey, P.H., 2000. Developmental changes in organicosmolytes in prenatal and postnatal rat tissues. Comp. Biochem. Physiol. A Mol.Integr. Physiol. 125, 45–56.

Moore, D.V., Sandground, J.H., 1956. The relative egg producing capacity of Schistosomamansoni and Schistosoma japonicum. Am. J. Trop. Med. Hyg. 5, 831–840.

Nicholson, J.K., Connelly, J., Lindon, J.C., Holmes, E., 2002. Metabonomics: a platformfor studying drug toxicity and gene function. Nat. Rev. Drug Discov. 1, 153–161.

Nicholson, J.K., Lindon, J.C., Holmes, E., 1999. ‘Metabonomics’: understanding themetabolic responses of living systems to pathophysiological stimuli viamultivariate statistical analysis of biological NMR spectroscopic data.Xenobiotica 29, 1181–1189.

Novak, M., Hudspeth, C., Blackburn, B.J., 1993. 1H NMR study of metabolicalterations in the small-intestine of rats infected with Hymenolepis-diminuta.Int. J. Biochem. 25, 1587–1591.

Ross, A.G.P., Bartley, P.B., Sleigh, A.C., Olds, G.R., Li, Y.S., Williams, G.M., McManus,D.P., 2002. Schistosomiasis. N. Engl. J. Med. 346, 1212–1220.

Ross, A.G.P., Sleigh, A.C., Li, Y.S., Davis, G.M., Williams, G.M., Jiang, Z., Feng, Z.,McManus, D.P., 2001. Schistosomiasis in the People’s Republic of China:prospects and challenges for the 21st century. Clin. Microbiol. Rev. 14, 270–295.

Shen, G.J., Jiao, M.D., Li, Q.Y., Xu, H., Han, J.M., Xin, X.M., 2002. Study on histopathology,ultrasonography and some special serum enzymes and collagens for 38 advancedpatients of schistosomiasis japonica. Acta Trop. 82, 235–246.

Schlotterbeck, G., Ross, A., Dieterle, F., Senn, H., 2006. Metabolic profilingtechnologies for biomarker discovery in biomedicine and drug development.Pharmacogenomics 7, 1055–1075.

Silva, L.M., Fernandes, A.L.M., Barbosa, A., Oliveira, I.R., Andrade, Z.A., 2000.Significance of schistosomal granuloma modulation. Mem. Inst. Oswaldo Cruz95, 353–361.

Sobh, M., Moustafa, F., Ramzy, R., Saad, M., Deelder, A., Ghoneim, M., 1991.Schistosoma mansoni nephropathy in Syrian golden hamsters: effect of dose andduration of infection. Nephron 59, 121–130.

Son, M., Ko, J.I., Kim, W.B., Kang, H.K., Kim, B.K., 1998. Taurine can ameliorateinflammatory bowel disease in rats. Adv. Exp. Med. Biol. 442, 291–298.

Steinmann, P., Keiser, J., Bos, R., Tanner, M., Utzinger, J., 2006. Schistosomiasis andwater resources development: systematic review, meta-analysis, and estimatesof people at risk. Lancet Infect. Dis. 6, 411–425.

Trygg, J., Wold, S., 2002. Orthogonal projections to latent structures (O-PLS). J.Chemometr. 16, 119–128.

Utzinger, J., Chollet, J., Tu, Z.W., Xiao, S.H., Tanner, M., 2002. Comparative study of theeffects of artemether and artesunate on juvenile and adult Schistosoma mansoni inexperimentally infected mice. Trans. R. Soc. Trop. Med. Hyg. 96, 318–323.

Wang, L.J., Cao, Y., Shi, H.N., 2008. Helminth infections and intestinal inflammation.World J. Gastroenterol. 14, 5125–5132.

Wang, Y.L., Holmes, E., Nicholson, J.K., Cloarec, O., Chollet, J., Tanner, M., Singer, B.H.,Utzinger, J., 2004. Metabonomic investigations in mice infected withSchistosoma mansoni: an approach for biomarker identification. Proc. Natl.Acad. Sci. U.S.A 101, 12676–12681.

Wang, Y.L., Tang, H.R., Holmes, E., Lindon, J.C., Turini, M.E., Sprenger, N., Bergonzelli,G., Fay, L.B., Kochhar, S., Nicholson, J.K., 2005a. Biochemical characterization ofrat intestine development using high-resolution magic-angle-spinning 1H NMRspectroscopy and multivariate data analysis. J. Proteome Res. 4, 1324–1329.

Wang, Y.L., Tang, H.R., Nicholson, J.K., Hylands, P.J., Sampson, J., Holmes, E., 2005b. Ametabonomic strategy for the detection of the metabolic effects of chamomile(Matricaria recutita L) ingestion. J. Agric. Food Chem. 53, 191–196.

Wang, Y.L., Utzinger, J., Xiao, S.H., Xue, J., Nicholson, J.K., Tanner, M., Singer, B.H.,Holmes, E., 2006. System level metabolic effects of a Schistosoma japonicuminfection in the Syrian hamster. Mol. Biochem. Parasitol. 146, 1–9.

Waters, N.J., Holmes, E., Williams, A., Waterfield, C.J., Farrant, R.D., Nicholson, J.K.,2001. NMR and pattern recognition studies on the time-related metaboliceffects of alpha-naphthylisothiocyanate on liver, urine, and plasma in the rat:an integrative metabonomic approach. Chem. Res. Toxicol. 14, 1401–1412.

Weinstock, J.V., 1996. Vasoactive intestinal peptide regulation of granulomatousinflammation in murine schistosomiasis mansoni. Adv. Neuroimmunol. 6, 95–105.

Westerhuis, J.A., Kourti, T., MacGregor, J.F., 1998. Analysis of multiblock andhierarchical PCA and PLS models. J. Chemometr. 12, 301–321.

Yaqoob, P., Calder, P.C., 1997. Glutamine requirement of proliferating Tlymphocytes. Nutrition 13, 646–651.

Metabolic profiling of a Schistosoma mansoni infection in mouse tissues using magic angle spinning-nuclear magnetic resonance spectroscopyIntroductionMaterials and methodsHost-parasite modelTissue sample collectionHistology and S. mansoni worm burden1H MAS-NMR spectroscopic analysis of tissueData reduction and multivariate data analysis

ResultsGross observation, histology and S. mansoni worm burden1H MAS-NMR spectra of intact tissue samplesMultivariate data analysis of 1H MAS-NMR CPMG tissue spectraCorrelation of 1H NMR CPMG spectral data with S. mansoni worm burdenPCA and H-PCA of tissue data

DiscussionAcknowledgementsReferences