Experimental Parasitology 124 (2010) 409–420

Contents lists available at ScienceDirect

Experimental Parasitology

journal homepage: www.elsevier .com/locate /yexpr

Schistosoma mansoni: Egg-induced downregulation of hepatic stellate cellactivation and fibrogenesis

Barrie Anthony a, William Mathieson b,1, William de Castro-Borges b,2, Jeremy Allen a,*

a Centre for Parasitology and Disease, Biomedical Sciences Research Institute, University of Salford, Manchester M5 4WT, UKb Department of Biology, University of York, PO Box 373, York YO10 5W, UK

a r t i c l e i n f o

Article history:Received 25 June 2009Received in revised form 30 November 2009Accepted 21 December 2009Available online 4 January 2010

Keywords:SchistosomesLiver fibrosisRelative gene expressionImmunocytochemistryTrematodes

0014-4894/$ - see front matter � 2010 Elsevier Inc. Adoi:10.1016/j.exppara.2009.12.009

* Corresponding author. Address: Biomedical Scienof Environment & Life Sciences, University of Salford, M+44 (0)161 295 5015.

E-mail address: j.t.allen@salford.ac.uk (J. Allen).1 Present address: Department of Histopathology, Im

Road, London W12 0NN, UK.2 Present address: Instituto de Ciências Exatas e

Ciências Biológicas, Universidade Federal de Ouro PreBrasil.

a b s t r a c t

Eggs of Schistosoma mansoni trapped in human liver can lead to fibrosis. Since liver fibrosis requires acti-vation of hepatic stellate cells (HSC) from a quiescent to a myofibroblastic phenotype, we investigated theeffects of S. mansoni eggs on this process using in vitro co-cultures with human HSC and evaluated estab-lished biomarkers for activation and fibrosis. HSC demonstrate significantly reduced expression ofa-smooth muscle actin (p < 0.001), connective tissue growth factor (p < 0.01) and type I collagen(p < 0.001) but significantly increased expression of peroxisome proliferator-activated receptor-c(p < 0.01). Morphologically, HSC exhibited elongated fine cellular processes and reduced size, increasedaccumulation of lipid droplets and reduced expression and organization of a-smooth muscle actin andF-actin stress fibres. Additionally, schistosome eggs prevented the HSC fibrogenic response to exogenoustransforming growth factor-b. In summary, schistosome eggs blocked fibrogenesis in HSC, a findingwhich may have implications for our understanding of the fibrotic pathology in S. mansoni infections.

� 2010 Elsevier Inc. All rights reserved.

1. Introduction sites which can result in fibrosis and portal hypertension (Booth

Schistosomiasis is an important helminth infection of man. It isestimated to infect 200 million people in the world (Crompton,1999; Chitsulo et al., 2000), with 120 million symptomatic (Chitsuloet al., 2000), 20 million people with severe morbidity (Crompton,1999) and 20,000 deaths a year (Crompton, 1999) resulting in lossof 4.5 million disability adjusted life years (DALYs) (WHO, 2002).The two most important species with regards to human liver diseaseare Schistosoma mansoni and Schistosoma japonicum. S mansoni isestimated to infect 83 million people in Africa, Caribbean and eastMediterranean regions (Crompton, 1999). Adult pairs of S. mansoniworms reside within the mesenteric veins where females releaseon average 340 eggs per female per day with rates ranging between190 and 658 depending upon strain and experimental host used (Lo-ker, 1983). More than 50% of eggs are carried to the liver by the portalcirculation where they become trapped in the liver sinusoids (Wynnet al., 2004). The host immune response to the eggs in the liver is welldocumented and typically involves granuloma formation at the egg

ll rights reserved.

ces Research Institute, Schoolanchester M5 4WT, UK. Fax:

perial College London, DuCane

Biológicas, Departamento deto, Ouro Preto, Minas Gerais,

et al., 2004; de Jesus et al., 2004; Wynn et al., 2004).In the liver, the cells primarily responsible for fibrogenesis are

the hepatic stellate cells (HSC). The HSC is located in the space ofDisse in the liver sinusoid and is responsible for maintenance ofthe extracellular matrix (ECM) (Sato et al., 1998), storage of vita-min A (Blomhoff and Wake, 1991), and has also been demonstratedto have a possible role in controlling blood flow through the liver(Reynaert et al., 2002). In response to liver injury, normally quies-cent HSC are activated and undergo a process of transdifferentia-tion towards a myofibroblastic phenotype (Friedman, 2000).Adoption of this phenotype is associated with transforming growthfactor (TGF)-b (Dooley et al., 2001a,b; Leask and Abraham, 2004),most notably with isoform 1 (Wickert et al., 2002). The fibrotic re-sponse of HSC to TGF-b1 stimulation is well documented and hasbeen widely used as a model for in vitro HSC transdifferentiationstudies (Dooley et al., 2001a,b). Myofibroblasts are characterizedby their increased ECM production, contractility, expression ofaSMA and loss of vitamin A storage (Friedman, 2000, 2003; Dooleyet al., 2001a,b). The wound healing response is terminated whenmyofibroblasts are removed from the area at the end of the insulteither by apoptosis or reversion back to a quiescent phenotype,however, during fibrogenesis this process fails (Friedman, 2003).A number of studies report that agonists of the peroxisome prolif-erator-activated receptor (PPAR)c can block and reverse theprocess of transdifferentiation in HSC (She et al., 2005; Tsukamoto,2005; Zhao et al., 2006), by a mechanism that appears to preventTGF-b1-mediated signalling (Zhao et al., 2006). Key features of

410 B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420

the PPARc-mediated quiescence of the cells are lipid retention anda reduction of collagen production (She et al., 2005; Zhao et al.,2006). In addition, treatment with adipogenic factors has also beenshown to restore the quiescent phenotype in primary rat cells (Sheet al., 2005) and the human LX-2 cell line (Zhao et al., 2006). TheLX-2 cell line is a recently developed immortalised human HSC cellline (Xu et al., 2005) that retains the key features of primary HSC.As these cells are readily activated by culture on plastic they can bemanipulated between quiescent and activated states and thereforeprovide a useful model for investigating HSC transdifferentiation.Human HSC have not been used widely in research as it is difficultto isolate them from tissue due to the infrequent availability of tis-sue that is suitable for cell isolation (Xu et al., 2005).

To date only a few studies have investigated the interactionsbetween schistosome eggs and host cells. S. mansoni eggs or theircrude soluble extracts have been found to stimulate hepatic endo-thelial cell proliferation (Freedman and Ottesen, 1988), migrationand angiogenesis (Kanse et al., 2005) and to induce fibroblast pro-liferation (Wyler and Tracy, 1982) and collagen synthesis (Borosand Lande, 1983). Even less well understood is the role of HSC inschistosomiasis. It is thought that the HSC is one of the mainsources of collagen deposition and ECM remodelling in schistoso-miasis (Booth et al., 2004) and most recently, HSC have been iden-tified in the periphery of egg granulomas in human S. japonicuminfections and myofibroblasts were observed in end-stage humandisease associated with fibrosis (Bartley et al., 2006). However, de-spite these findings and the in vivo HSC and egg co-localisation,there are no prior studies in which the direct effects of parasiteeggs on these cells are considered.

This study was designed to investigate the nature of S. mansoniegg activity on human HSC using an LX-2 cell model. Specifically,morphological and transcriptional biomarkers of HSC activationand fibrogenesis were examined with HSC cultured on plastic inthe presence of adipogenic factors or transforming growth factor(TGF)-b.

2. Materials and methods

2.1. Egg isolation

The parasite material was kindly provided by Prof. Alan Wilson,University of York, UK. Male C57BL/6xCBA mice were each infectedwith 180 S. mansoni cercariae via the shaved abdomen (Smithersand Terry, 1965). Seven weeks later the livers were removed,homogenised and digested with trypsin (Sigma–Aldrich, Poole,UK) for 3 h at 37 �C. The homogenate, containing the eggs waspassed through a series of sieves (300–180 lm), the eggs collectedby sedimentation then cleaned by washing six times in sterilephosphate-buffered saline. The clean eggs were then counted andresuspended in DMEM (Lonza, Wokingham, UK) cell culture med-ium. Eggs were stored in DMEM with 10% fetal bovine serum (FBS,Lonza) overnight before use the next day. Non-viable intact eggswere obtained by freezing them in PBS at �80 �C. Viability was as-sessed by monitoring the ability of eggs to hatch in fresh water.HSC were then co-cultured for up to 3 days in the presence of via-ble eggs at concentrations of 500, 1000 or 1500 eggs/ml. At theseconcentrations previous studies have elicited host cell responsesfrom whole eggs, or from the equivalent concentrations of ex-tracted soluble egg antigens (SEA). Accordingly, approximately10 lg/ml SEA corresponds with 1000 eggs.

2.2. Cell cultures

LX-2 cells were a kind gift of Prof. Friedman, Mount Sinai Schoolof Medicine, New York, USA. Cultures were maintained in DMEM

containing 2% FBS plus antibiotics (complete medium) at 37 �Cand with 5% CO2. Medium was changed after 48 h. Cells wereseeded in 24 well plates (Greiner bio-one, Stonehouse, UK) at adensity of 4.3 � 103 cells per cm2. For collagen gel experiments areconstitution buffer was made with 0.2 ml 1N NaOH, 0.5 ml7.5% NaHCO3 and 0.3 ml of dH2O. One hundred and twenty millili-ters of this buffer was mixed with 120 ll of 10� DMEM and 960 llof type 1 rat tail collagen to make the gel mixture. In a 35 mm plate1 ml of gel was used and in a 24 well plate 300 ll was used. The gelwas allowed to set at 37 �C before use. For co-culture experimentswith parasite eggs, cells were treated for up to 3 days with eithercomplete medium alone or complete medium + S. mansoni eggs(at concentrations of 500, 1000 and 1500 eggs/ml). For adipogenicdifferentiation cells were either treated with complete mediumalone or complete medium + adipogenic factors, comprised of0.5 mM isobutylmethylxanthine, 1 lM dexamethasone, 1 lMinsulin (Sigma–Aldrich) and compared to complete medium +S. mansoni eggs (1000 eggs/ml). For experiments with recombinantTGF-b1 (Peprotech EC, London, UK), cells were treated with eithercomplete medium + TGF-b1 (at a concentration of 5 ng/ml) or withcomplete medium + S. mansoni eggs (1000 eggs/ml), + TGF-b1(5 ng/ml).

2.3. Immunofluorescence and phase-contrast microscopy

Cells were fixed in methanol for immunofluorescence thenwashed three times in PBS. A mouse monoclonal anti-humanaSMA antibody (Sigma–Aldrich, clone 1A4) and Texas-Red conju-gated anti-mouse IgG antibody (Vector Laboratories, Peterborough,UK) were employed for detection of aSMA expression. For this thefixed cells were washed with PBS before a blocking solution of 5%bovine serum albumin (BSA, Sigma–Aldrich) was added to thecells. This was removed and the cells incubated at room tempera-ture with the primary antibody (1:800 in 5% BSA/PBS) for 1 h. Thiswas then removed and the cells washed in PBS before the cellswere then incubated at room temperature in the dark with the sec-ondary antibody (1:300 in 5% BSA/PBS) for 45 min. Cells were thenwashed in PBS and nuclei were counterstained with 6-diamidino-2-phenylindole (DAPI, Vector Laboratories). For staining of F-actin,cells were fixed in 3.7% formaldehyde solution in PBS after a pre-wash with PBS. The cells were then washed with PBS and left at�20 �C in acetone to permeabilise. The cells were again washedin PBS before a staining solution containing FITC-conjugated phal-loidin (Invitrogen, Paisley, UK) diluted 1:50 in 1% BSA was appliedfor 20 min. Cells were counterstained with DAPI as before. Phase-contrast imaging was used for general cell morphology and fordetection of cellular lipids. Cells for lipid droplet staining werefixed for 30 min in 2% paraformaldehyde (Sigma–Aldrich) andwashed in dH2O before incubation with 99% (v/v) propane 1,2-diol(Sigma–Aldrich) for 5 min. This was removed and Oil-Red O (0.5%w/v) in propane 1,2-diol was added to the cells for 30 min. Oil-Red O solution was removed and the cells washed for 5 min with85% (v/v) propane 1,2-diol. This was removed and the cells washedwith dH2O. Cells were then counterstained with Haematoxylin(Vector Laboratories). Phase-contrast and immunofluorescenceimages were obtained with a Nikon TE2000 Eclipse inverted fluo-rescence microscope with a cooled Hamamatsu Orca camera sys-tem and merged using Image-Pro Lab v3.7 image analysissoftware (Nikon UK Ltd., Kingston upon Thames, UK). Images ofOil-Red O stained lipid droplets were obtained using a Nikon Cool-pix 4500 camera attached to the Nikon TE2000 microscope.

2.4. Realtime PCR

Primers and probes targeting expression of type I collagen(Col1A1), connective tissue growth factor (CTGF), and peroxisome

B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420 411

proliferator-activated receptor-c (PPARc) were obtained as Per-fectProbe gene detection kits from PrimerDesign Ltd. (Southamp-ton, UK). a-Smooth muscle actin (aSMA) was obtained as aTaqman gene expression assay from Applied Biosystems (Warring-ton, UK). Sequences of primers are given in Table 1. The most stableendogenous control genes for normalization in this experimentalmodel were determined, from a panel of 12 using the geNorm kitand software (PrimerDesign), to be ATP5B and YWHAZ. As therewas little difference found in using one or both of these referencegenes together for normalization of target gene expression (datanot shown), ATP5B was selected for normalization of results andused in all realtime analyses. Cells were seeded into six well plateswith 105 cells per well in complete medium. After the cells reached70% confluence, complete medium was replaced with DMEM con-taining 0.1% FBS and incubated overnight before treatment in thefollowing morning. Treatments were then performed as describedin cell cultures above. After 1 or 3 days cells were lysed in situ usingcell protect reagent (Qiagen) and stored at �80 �C prior to RNAextraction. RNA was extracted, and genomic DNA removed, byspin-column purification according to the suppliers recommenda-tions (RNeasy Mini Plus kit, Qiagen). RNA yield and purity wereevaluated spectrophotometrically. Absorbance was measured at260 and 280 nm and the purity was determined from the absor-bance ratio (A260/A280). Total RNA (1 lg) was used in a reverse tran-scription reaction and cDNA synthesis performed using a precisionreverse transcription kit according to the suppliers instructions(PrimerDesign). The cDNA (2.5 ll), diluted 1 in 10, was used ineach PCR reaction, performed in triplicate on an Applied Biosys-tems 7500 Fast system. All PCR reactions were prepared using Pre-cision Mastermix with low ROX (PrimerDesign), according tosuppliers recommendations. PCR conditions were as follows, forall genes: 95 �C for 10 min, then 50 cycles of 95 �C for 15 s, 50 �Cfor 30 s, and 72 �C for 15 s.

2.5. Statistical analysis

Cell culture experiments were performed three times indepen-dently and representative images used. Gene experiments wereperformed twice independently. Realtime PCR data were comparedusing the Relative Expression Software Tool (REST) 2008 software(Pfaffl et al., 2002). This determines whether there are significantdifferences between samples and controls, while taking intoaccount issues of reaction efficiency and reference gene normaliza-tion. Statistical validation was performed within REST 2008 using

Table 1Primer and probe sequences for realtime PCR analysis.

Genesymbol

Primers/probe

Sequences or source if unavailable

aSMA Forward Applied Biosystems Taqman gene expressionassay: Hs00909449_m1Reverse

CTGF Forward 50-CCCAGACCCAACTATGATTAGAG-30

Probe 50-CCCATCCCACAGGTCTTGGAACAGGCG-30

Reverse 50-AGGCGTTGTCATTGGTAACC-30

PPARc Forward 50-GAATAAAGATGGGGTTCTCATATCC-30

Probe 50-AAGTCACCAAAAGGCTTTCGCAGGCTCTTT-30

Reverse 50-AACTTCACAGCAAACTCAAACTT-30

Col1A1 Forward 50-AGACAGTGATTGAATACAAAACCA-30

Probe 50-CAAGACCTCCCGCCTGCCCATCATCGT-30

Reverse 50-GGAGTTTACAGGAAGCAGACA-30

ATP5B Forward PrimerDesign PerfectProbe house-keeping geneassay for ATP5BReverse

aSMA, a-smooth muscle actin; CTGF, connective tissue growth factor; Col1A1, typeI collagen; PPARc, peroxisome proliferator-activated receptor-c; ATP5B, mito-chondrial ATP synthase.

the Pair Wise Fixed Reallocation Randomisation Test and 10,000randomisations were performed. Statistical significance wasaccepted when p < 0.05.

3. Results

3.1. S. mansoni eggs induce a quiescent transcriptional profile in LX-2

Preliminary findings (data not shown) indicated that there wasa quiescent effect of parasite material on HSC and therefore exper-iments were conducted using pre-activated HSC, induced by cul-ture on plastic and treatment with TGF-b1. Several key targetgenes were analyzed based on their known association with a qui-escent or myofibroblastic HSC phenotype. The selected genes maybe described as broadly pro-fibrogenic or anti-fibrogenic. Thus,quiescent cells demonstrate increased expression of PPARc andshow reduced expression of aSMA, whereas activated HSC wouldbe expected to show a reversal of this profile. In addition, Col1A1and CTGF were evaluated as markers of increased fibrogenesis. Via-bility of the cells was checked using trypan blue staining and wasobserved not to be decreased by egg treatments compared to con-trols (data not shown).

aSMA expression (Fig. 1A and B): in response to treatment withTGF-b (5 ng/ml) LX-2 cells significantly (p < 0.001) increased theirexpression of aSMA approximately 3-fold after 24 h or 2-fold after72 h, compared to untreated controls (shown in 1A and B for com-parison). Treatment with viable schistosome eggs at a range of con-centrations (500–1500 eggs/ml) caused a significant (p < 0.001)downregulation of aSMA by 2–5-fold after 24 h, compared to un-treated controls. A more dramatic effect on aSMA expression wasobserved after 72 h with egg treatments causing a significant(p < 0.001) 11–14-fold reduction in aSMA expression, comparedto controls (1A). Similar data were obtained with non-viable eggs(1B). There were no significant differences in the effects inducedbetween the egg concentrations employed with 500 eggs/mlevoking a maximal reduction in expression. Furthermore, whenschistosome eggs were present they were able to prevent theTGF-b-mediated increase in aSMA expression observed with TGF-b alone. Viable eggs (1A) caused 4- and 6-fold significant downre-gulations (p < 0.001) whereas non-viable eggs (1B) caused 4- and17-fold decreases (p < 0.01), after 24 and 72 h, respectively, com-pared to control values. In response to adipogenic factors (1B),aSMA expression did not alter significantly at either timepoint,whereas co-treatment with TGF-b and adipogenic factors produced2–3-fold increases (p < 0.01) in aSMA expression, similar to thatproduced by TGF-b alone. This suggests adipogenic factors do notappear to significantly regulate aSMA expression in LX-2 cells.

CTGF expression (Fig. 1C and D): in response to treatment withTGF-b (5 ng/ml) LX-2 cells significantly (p < 0.01) increased theirexpression of CTGF approximately 6–8-fold after 24 or 72 h, com-pared to untreated controls (shown in 1C and D for comparison).Treatment with viable schistosome eggs also induced significantincreases (p < 0.01) in CTGF expression approximately 4–7-foldafter 24 h. CTGF expression was suppressed by 72 h in the presenceof the eggs at 1000 or 1500 eggs/ml but it was still significantly ele-vated (p < 0.01) by 3-fold, compared to control at an egg dose of500/ml (1C). Non-viable eggs produced similar results althoughCTGF remained slightly elevated at 72 h compared to controls(p < 0.01) (1D). The presence of viable or non-viable eggs failedto diminish a large induction (10–14-fold) of CTGF expression(p < 0.01) by treatment with TGF-b by 24 h, but by 72 h this in-crease had been completely ablated by viable eggs, compared toTGF-b alone, and CTGF expression was reduced to a level compara-ble with untreated cells (1C). Non-viable eggs also suppressed theeffect of TGF-b on CTGF but were less effective and its expression

Fig. 1. Realtime PCR expression analysis demonstrates Schistosoma mansoni eggs induce a quiescent phenotype. HSC were cultured for 24 or 72 h, total RNA was preparedfrom triplicate wells and aSMA (A and B), CTGF (C and D), collagen (E and F), and PPARc (G and H) gene expression was assessed as follows: in the upper graphs (A, C, E, andG), HSC were cultured in the presence of viable schistosome eggs at different concentrations (500, 1000, 1500 eggs/ml), or with TGF-b (5 ng/ml), or with TGF-b andschistosome eggs (500 eggs/ml) together; in the lower graphs (B, D, F, and H), HSC were cultured in the presence of non-viable (NV) schistosome eggs at differentconcentrations (500, 1000 eggs/ml), or with TGF-b (5 ng/ml), or with TGF-b and schistosome eggs (500 eggs/ml) together. In addition, the effects of adipogenic factors (Adipo)alone, or with TGF-b (1 ng and 5 ng/ml) are reported. Results were normalized against ATP5B and expressed as the mean fold change in normalized expression of the targetgene compared to the untreated control. *p < 0.05, **p < 0.01, ***p < 0.001 compared to control.

412 B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420

remained slightly but significantly (p < 0.01) elevated, compared tocontrol (1D). In response to adipogenic factors, CTGF expressionwas elevated marginally (p < 0.01) at both timepoints (1D). Inter-estingly, co-treatment with TGF-b (1 ng and 5 ng/ml) and adipo-genic factors resulted in dose and time-dependent increases inCTGF expression (p < 0.01), compared to TGF-b alone.

Col1A1 expression (Fig. 1E and F): in response to treatment withTGF-b (5 ng/ml) LX-2 cells significantly (p < 0.001) increased theirexpression of Col1A1 approximately 2–3-fold after 24 or 72 h,compared to untreated controls (shown in 1E and F for compari-son). Treatment with viable schistosome eggs induced significant2–3-fold downregulation of Col1A1 at concentrations of 500(p < 0.001) or 1000 (p < 0.05) eggs/ml after 24 h. At 1500 eggs/mlCol1A1 was also downregulated but this was of marginal statisticalsignificance. However, by 72 h all the egg treatments had inducedhighly significant (p < 0.001) downregulation of Col1A1 expressionby 9–15-fold, compared to controls (1E). Non-viable eggs were notas effective as viable eggs (1F); after 24 h there were marginal in-creases in collagen expression (p < 0.01) and although these hadbeen reversed by 72 h (p < 0.01) the magnitude of the reductionin collagen expression (4–5-fold) was less than for viable eggs

(8–10-fold). In the presence of eggs, treatment with TGF-b failedto upregulate Col1A1, instead there were large time-dependentreductions (p < 0.001) in Col1A1 expression; a 3-fold decrease at24 h and a 11-fold decrease at 72 h, compared to controls. How-ever, non-viable eggs were less effective, producing only a 2-foldreduction (p < 0.05) in collagen expression after 72 h, comparedto control (1F). In response to adipogenic factors, collagen expres-sion was marginally but significantly reduced (p < 0.01) at bothtimepoints (1F). Co-treatment with TGF-b (1 ng and 5 ng/ml) andadipogenic factors reversed this and produced dose-dependentincreases in collagen expression (p < 0.01), compared to control.

PPARc expression (Fig. 1G and H): in response to treatment withTGF-b (5 ng/ml) LX-2 cells demonstrated small, but not statisticallysignificant, reductions in their expression of PPARc approximately2-fold after 24 or 72 h, compared to untreated controls (shown in1G and H for comparison). Treatment with viable schistosome eggsinduced marginal increases in PPARc expression, approximately2-fold, after 24 h. However, by 72 h PPARc expression was highlyinduced by eggs up to approximately 7-fold (p < 0.01), comparedto controls (1G). Non-viable eggs produced similar results and alsoinduced PPARc after 72 h (p < 0.01), although the magnitude of

Fig. 1 (continued)

B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420 413

induction (3–4-fold) was less compared to viable eggs (1H). In thepresence of eggs, TGF-b treatment significantly inhibited PPARcexpression at 24 h (p < 0.05), but by 72 h PPARc expression was in-creased by eggs back to control levels (1G). A similar effect wasachieved with non-viable eggs (1H). In response to adipogenic fac-tors, PPARc expression was significantly increased (p < 0.01) in atime-dependent manner (1H). However, co-treatment with TGF-b(1 ng and 5 ng/ml) and adipogenic factors blocked this effect andPPARc expression remained similar to, or marginally inhibited(p < 0.05), compared to control.

3.2. LX-2 morphology and phenotype

Untreated LX-2 cells display time-dependent activation whencultured on plastic. By 24 h, cells already appear partially activatedshowing a flattened morphology with few extended cellular pro-cesses evident and numerous mitotic figures observed underphase-contrast (Fig. 2A). Observed with immunofluorescencemicroscopy, some cells express the myofibroblastic marker aSMA(Fig. 2E) and stain heavily for F-actin cytoskeletal elements(Fig. 2I). By 72 h all of these features are present within more ofthe cell population, indicating more cells have adopted an acti-vated myofibroblastic phenotype. Prominent stress fibre stainingpatterns are associated with aSMA and F-Actin (Fig. 2C, G, and K).

3.3. S. mansoni eggs induce a quiescent LX-2 morphology andphenotype

LX-2 cells were cultured in the presence of a range of viableschistosome egg concentrations (500–1500 eggs/ml), consistentwith egg numbers employed in previous studies. Data are shownfor 500 eggs/ml. Findings were similar for the higher egg concen-trations. Under phase-contrast, by 24 h a proportion of cells hadadopted a different morphology compared to control cells, havinga more slender appearance with extended cellular processes a fea-ture of most cells (Fig. 2B). By 72 h these features were present inalmost all cells and by this time cells had become quite slenderwith a reduced cytoplasmic area and had extended more cellularprocesses that often seemed to be linked, forming a mesh like pat-tern (Fig. 2D). This morphology resembled that induced by adipo-genic growth factors (Fig. 4C), or that produced when cells weregrown on a collagen extracellular matrix (data not shown). Itwas strongly suggestive of a more quiescent phenotype and thesefindings were supported when immunofluorescence was per-formed with aSMA and F-Actin. At the 24 h time point (Fig. 2F),but more extensively by 72 h (Fig. 2H), a marked reduction in thenumber of cells expressing aSMA had occurred. A few cells ap-peared to remain activated, displaying stress fibres but these wereonly a small proportion of the cell population, compared to con-

Fig. 1 (continued)

414 B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420

trols. A dose–response effect was also noticed, with higher num-bers of eggs appearing to reduce the numbers of SMA positive cells(data not shown). For F-actin expression, rather than a reduction inthe overall number of cells expressing F-actin, it was observed thatthere were reductions in the organization of F-actin into stressfibres, which were much less apparent compared to untreated con-trols at both timepoints, but especially after 72 h (Fig. 2I–L). Thisappears to reflect the altered morphology of the HSC. Parallel stud-ies were then performed with non-viable eggs (500–1500 eggs/ml)and similar effects were noted (data not shown).

3.4. S. mansoni eggs induce LX-2 lipid retention

Since the HSC quiescent phenotype is associated with cytoplas-mic storage of lipid droplets, we examined the effect of eggs on thisprocess using Oil-Red O staining. At 24 or 72 h timepoints, un-treated cells showed little Oil-Red O staining indicating few lipiddroplets were present and suggestive of a (partially) activated cellphenotype (Fig. 3A and B). However, in the presence of 500 or 1000viable eggs/ml, there was considerable prominent lipid dropletstaining observed, consistent with quiescence of the cells. This al-tered lipid accumulation was apparent after 24 h (Fig. 3C and E)but there appeared to be a time-dependent increase in stainingwith the greatest number of perinuclear and cytoplasmic lipiddroplets observed after 72 h with either 500 or 1000 eggs/ml (Figs.3D and 4F). A qualitative comparison of lipid staining suggested

that viable eggs produced a greater effect than non-viable eggs(data not shown).

3.5. S. mansoni eggs reverse TGF-b-mediated myofibroblastic LX-2morphology and phenotype

HSC treated with pro-fibrogenic TGF-b are induced to adopt amyofibroblastic activated phenotype, typified by high expressionof aSMA, loss of lipid droplets and a large flat morphology. To dem-onstrate the reversibility of the process in LX-2 cells these were co-treated with adipogenic factors and TGF-b. LX-2 cells treated with1 ng/ml (data not shown) or 5 ng/ml TGF-b alone displayed a flat,spread myofibroblast phenotype (Fig. 4A) in which aSMA stainingassociated with prominent stress fibres was observed (Fig. 4E) andvery few lipid droplets were detectable (Fig. 4I). Treatment withadipogenic factors reduced the cell size producing a more slendercell with reduced cytoplasmic area and elongation of cellularprocesses (Fig. 4C). Adipogenic treatment did not prevent aSMAexpression, although unlike TGF treatment this was not associatedwith prominent stress fibre patterns (Fig. 4G). However, numerouslipid droplets were observed in the cells by Oil-Red O staining(Fig. 4K). Adipogenic factors co-administered with TGF-b producedsimilar patterns of staining to those observed with adipogenic fac-tors alone (Fig. 4D, H, and L). Thus, adipogenic treatment appearedto suppress the myofibroblastic phenotype induced by TGF-b. Wenext investigated whether eggs had any similar effects on the

Fig. 1 (continued)

B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420 415

TGF-b-mediated fibrogenic process in LX-2 cells. Co-treatment ofLX-2 cells with TGF-b and viable eggs (500 eggs/ml) produced cellswith a reduced cytoplasmic area, resembling untreated cells(Fig. 4B) but lacking the elongated cellular processes that eithereggs alone (Fig. 2D), or adipogenic factors (Fig. 4C), produced. How-ever, in contrast to treatment with TGF-b alone, co-treatment witheggs completely inhibited aSMA expression (Fig. 4F) and extensivelipid droplets were observed (Fig. 4J), similar to the effect of adipo-genic treatment (Fig. 4K). Thus, co-treatment of HSC with TGF-band eggs appears to prevent acquisition of the myofibroblasticattributes due to TGF-b treatment alone and at least in regard toinhibiting aSMA expression, appears more effective than adipo-genic factors.

4. Discussion

Activated HSC are considered to be the main effector cell typefor hepatic fibrosis in a number of human diseases and in animalmodels (Friedman, 2003) and have been implicated in the fibro-genesis observed in schistosomiasis (Booth et al., 2004). There isgood evidence that activated HSC are present in patients withfibrosis and contribute to disease caused by S. japonicum (Bartleyet al., 2006). In livers of patients infected with S. mansoni, typicalmyofibroblasts have been demonstrated in scar tissue (Grimaud

and Borojevic, 1986), although it has also been suggested thatmyofibroblasts appear as transient cells which focally accumulatearound damaged portal veins and that the cells tended to disap-pear with time (Andrade et al., 1999). Another study of S. mansoniinfection conducted on liver biopsies has reported increased num-bers of cells expressing the HSC and myofibroblast markers glialfibrillary acid protein and aSMA (Chang et al., 2006). The originsof such cells are unclear and apart from resident HSC, myofibro-blast-like cells may be derived from portal fibroblasts (Guyotet al., 2006), circulating fibrocytes (Forbes et al., 2004) or from anepithelial to mesenchymal transition process (EMT) (Ikegamiet al., 2007).

In this study it was shown that the presence of viable and non-viable schistosome eggs induces the increased number and size ofHSC cytoplasmic lipid droplets and a morphological profile ofsmaller cell size and the development of extended cellular pro-cesses. These data were confirmed with crude SEA, or with the eggsphysically separated from HSC by a cell culture insert (data notshown). Moreover, HSC cultured with eggs showed growth pat-terns similar to that observed when HSC are cultured either onor within a collagen gel substrate. All of these characteristics aretypically associated with the in vivo appearance of quiescent HSC(Senoo et al., 1996; Sato et al., 1998). Quiescent HSC are primarilylipid (vitamin A)-storing cells and were known as ‘fat-storing cells’

416 B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420

because of their lipid content. Maintenance of this phenotype isanalogous to adipocyte differentiation, requiring adipogenic tran-scriptional regulation (Miyahara et al., 2000; She et al., 2005). In

this study, it was found that activated HSC treated with an ‘adipo-genic differentiation mixture’ can be reverted back to quiescence,consistent with previous data (She et al., 2005), Importantly this

Fig. 3. Schistosoma mansoni eggs induce HSC lipid accumulation. HSC were cultured alone on plastic for 24 or 72 h (A and B) or in the presence of viable eggs (C–F) at differentconcentrations (500 or 1000 eggs/ml). At each timepoint cells were fixed and stained with Oil-Red O to detect intracellular lipids (red stain). Untreated HSC show someresidual lipids at 24 h (A) but this disappears by 72 h (B). In the presence of viable eggs there is accumulation of lipids at 24 h (C and E) and at 72 h (D and F), with evidence ofincreased lipid retention at the higher egg concentration, as shown in image (F), compared to image (J), as indicated by stain intensity and lipid distribution. Data obtainedwith non-viable eggs also showed increased lipid staining compared to controls, but was reduced compared to viable eggs (data not shown). Image magnification 400�. (Forinterpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420 417

treatment also induces expression of the adipogenic transcrip-tional regulator PPARc which is required for maintenance of thedifferentiated HSC phenotype. Depletion of PPARc is associatedwith HSC activation whereas increasing PPARc expression inducesHSC quiescence and has been found to inhibit activation markerssuch as aSMA and collagen (Hazra et al., 2004). In this study, theeffects of schistosome eggs on HSC were comparable to those ob-served with adipogenic factors; namely, increased PPARc expres-sion and lipid droplet accumulation concomitant with decreasedexpression of type I collagen. However, whereas treatment withadipogenic factors failed to reduce TGF-b-mediated aSMA expres-sion, eggs inhibited aSMA expression and organization. Moreover,the egg-mediated increase in PPARc expression was maintainedeven in the presence of TGF-b. Results obtained with adipogenic

Fig. 2. Schistosoma mansoni viable eggs induce a quiescent HSC morphology and phenotyviable schistosome eggs at a concentration of 500 eggs/ml. HSC were examined by (1) phand D). Presence of eggs leads to cells becoming smaller, slender and more elongated,detection of aSMA expression with Texas-Red labelling and DAPI nuclear counterstaininpositive cells and aSMA organization into stress fibres; (3) immunofluorescence detectionand J) and 72 h (K and L). As cell morphology is altered in the presence of eggs, F-actinpatterns. Similar data were obtained with egg concentrations of 1000 or 1500 eggs/ml (

3

factors and eggs in the presence of TGF-b appear more impressivewhen the high dosage of recombinant TGF-b (5 lM) is considered.Experiments with lower doses of recombinant TGF-b were notundertaken in this study, but for comparison Zhao et al. (2006)showed PPARc agonists inhibited fibrogenic responses in LX-2 cellsinduced by only 100 pM TGF-b (Zhao et al., 2006). CTGF, an ‘imme-diate-early’ gene, is typically rapidly induced by TGF-b via aresponsive element in its promoter and in the liver CTGF is pro-duced by numerous cell types and has been implicated in HSC acti-vation and liver fibrogenesis (Rachfal and Brigstock, 2003). Recentfindings from in vitro and animal models suggest that CTGF mayhave a significant role in hepatocyte EMT, either directly or actingas a downstream mediator of TGF-b (Meindl-Beinker and Dooley,2008). In this study it was found that by 3 days, the effect of eggs

pe. HSC were cultured for 24 or 72 h as follows; HSC alone or HSC in the presence ofase-contrast microscopy for morphological appearance at 24 h (A and B) and 72 h (Cwith cells extending cellular processes by 72 h; (2) indirect immunofluorescenceg at 24 h (E and F) and 72 h (G and H). Presence of eggs reduces number of aSMA-of FITC-phalloidin (F-actin) expression with DAPI nuclear counterstaining at 24 h (Iexpression is maintained but becomes noticeably less organized into stress fibre

data not shown). Scale bars are 100 lm.

Fig. 4. TGF-b-mediated activation of HSC is opposed by Schistosoma mansoni eggs and adipogenic factors. HSC were cultured for 72 h as follows; with TGF-b (5 ng/ml) alone,with adipogenic factors alone, with TGF-b + viable eggs (500 eggs/ml), with TGF-b + adipogenic factors. HSC were examined (1) by phase-contrast microscopy for morphology(A–D); (2) by indirect immunofluorescence detection of aSMA expression with Texas-Red labelling and DAPI nuclear counterstaining (E–H); (3) by Oil-Red O staining for lipidaccumulation (I–L). TGF-b induces a larger, flat cell phenotype with no cellular processes (A), many aSMA-positive cells with organized stress fibres (E), and few lipid droplets(I). Adipogenic factors promote a smaller, slender elongated phenotype with extended cellular processes (C), aSMA is expressed but not organized into stress fibres (G), andthere are abundant lipids (K). This phenotype is also induced when HSC are co-treated with adipogenic factors and TGF-b (D, H, and L). Presence of eggs leads to more cellsbecoming smaller and elongated, compared to TGF-b alone, with some cells extending cellular processes (B), very few aSMA-positive cells (F) and lipid accumulation evident(J). Similar data were obtained with non-viable eggs at the same concentration (data not shown). Scale bars are 100 lm, other images are 400� magnification.

418 B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420

reduced CTGF expression to levels comparable with controls andfurthermore that co-treatment of HSC with eggs and TGF-b inhib-ited CTGF induction, from a 6-fold increase with TGF-b alone, backto control levels. These data suggest that eggs can block TGF-b-dependent CTGF induction. Downregulation of CTGF in the pres-ence of eggs may be an important factor influencing HSC differen-tiation and quiescence as CTGF has been recently shown to blockdevelopment of the adipogenic phenotype, including suppressionof lipid accumulation and PPARc expression (Tan et al., 2008).

Gene expression of the myofibroblast marker aSMA was signif-icantly reduced in HSC treated with eggs and/or TGF-b, indicatingthat myofibroblast transdifferentiation induced by TGF-b isblocked by treatment with schistosome eggs. In support, aSMAprotein expression was dramatically inhibited by the presence ofeggs and the overall pattern of expression was changed. In re-sponse to TGF-b, HSC myofibroblasts express abundant aSMAwhich is highly organized in stress fibres (Uemura et al., 2005).In contrast, data presented here showed eggs reduced both thenumber of aSMA-positive cells, their organization of aSMA instress fibres, and blocked the TGF-b-mediated upregulation.

F-actin expression in response to eggs was not lost, but showedevidence of re-organization with staining of cellular processesapparent. This is consistent with reported F-actin data from HSCcultured on extracellular matrix collagen (Sato et al., 1998).Transcriptional activation of type I collagen is a key feature ofHSC activation and transdifferentiation (Friedman, 2000) and inthis study, schistosome eggs downregulated type I collagen expres-sion in HSC and were also able to block its induction in them byTGF-b.

The host response to chronic helminth infections is consideredto be host-protective and to feature a type 2-regulatory response;a modified type 2 characteristically exhibiting reduced IL-5 andIL-13 secretion and increased anti-inflammatory secretion, e.g., ofIL-10 (Diaz and Allen, 2007). Increased numbers of regulatoryCD4+ T cells (Tregs) in the liver control the magnitude of the im-mune response by suppressing activation of autoreactive cells(Wilson et al., 2007). Acute Schistosomiasis begins after 3–9 weeks(Lambertucci, 1993) and elicits a pro-inflammatory immune-med-iated hypersensitivity reaction manifest as Katayama syndrome(Burke et al., 2009). As the disease develops beyond 5 weeks, egg

B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420 419

deposition induces a switch away from typical type 1 cytokines toa response typified by increased IL-4 and crucially regulated by IL-10 (Caldas et al., 2008). At this time the size of the granulomas isobserved to reduce which itself will provide relief to vascularobstruction (Lambertucci, 1993). IL-10-deficient mice fail to down-regulate their granulomatous response (Sadler et al., 2003). IL-10acts as a regulatory cytokine, suppressing the production of type1 cytokines but additionally prevents the resultant TH2 dominatedresponse from becoming excessive (Burke et al., 2009). Failure tomodulate this response leads to a situation in which the hostresponds vigorously throughout infection leading to severe pathol-ogy and fibrosis (Lambertucci, 1993). In this regard, severity ofhepatic fibrosis is associated with increased IL-13 expression(Caldas et al., 2008). In response to IL-13, macrophages becomealternatively activated to hydrolyse intracellular L-arginine, pro-moting collagen production and thus fibrogenesis (Wilson et al.,2007). High arginase activity is associated with large granulomas,whereas classically activated macrophages that produce NOS-2are associated with smaller granulomas and less severe fibrosis(Wynn et al., 2004). Viable, non-viable eggs or SEA induce a type2 response from the human host (Faveeuw et al., 2003; Schrammet al., 2003; Everts et al., 2009) and SEA-derived fractions areimmunogenic. For example, IPSE is secreted from the egg subshellenvelope and induces human basophils to produce IL-4 and IL-13(Schramm et al., 2003), Much work has focused on the interactionof eggs or SEA with dendritic cells (DC), reviewed in Perona-Wrightet al. (2006). Exposure to SEA induces only limited activation of hu-man DC in vitro, compared to murine DC and in both cases a mutedpro-inflammatory response results with a notable lack of IL-10.However, these ‘activated’ DC are capable, in vitro and in vivo, ofdriving T cell development and a type 2 polarisation. Several eggcomponents, for example omega-1, a3-fucosyltransferase and coreg 2-xylosyltransferase glycoconjugates seem to play a role in thistype 2 polarisation (Everts et al., 2009; Faveeuw et al., 2003).

The observations from this study clearly show a suppressive ef-fect of schistosome eggs on the activation of HSC. This appears tobe at odds with the fibrosis that is a clinical feature in some pa-tients with schistosomiasis. To account for this discrepancy aworking hypothesis is proposed based on events localised at thesite of the granuloma. Early granuloma formation in response toegg deposition is dominated by the type 2 response with alterna-tively activated macrophages contributing to collagen deposition.HSC, recruited to, or resident at, the granuloma site may becomeactivated and will also contribute to fibrogenesis. However, HSCin the immediate vicinity of the eggs themselves will, accordingto data presented here, remain quiescent and will also be non-responsive to factors such as TGF-b, from other cells. This is consis-tent with reports that in granulomas of mice and humans infectedwith S. japonicum (Bartley et al., 2006) or with S. mansoni (Burkeet al., 2009), activated HSC are to be found at the periphery ofthe granulomas and not in the immediate vicinity of the eggs. Thislocalisation of HSC is coincident with collagen deposition and des-min and aSMA immunoreactivity (Bartley et al., 2006). This sug-gests a concentration gradient of egg-secreted soluble factorsexists, decreasing towards the periphery of the granuloma. HSCactivation would therefore take place in areas of lower concentra-tions of egg factors, i.e., at the granuloma periphery. There is someevidence to support a concentration-dependent effect from preli-minary studies with SEA (data not shown) and from the dose–re-sponse gene expression studies with eggs. This concentrationgradient may be in part due to the nature of the egg’s secretions.In a recent study IPSE was found to make up 83% of the egg secre-tions and due to its ‘‘sticky” nature it may sequester other proteinsin the vicinity of the egg (Mathieson and Wilson, 2009). The dose–response relationship may be complicated by the fact that eggs re-lease different antigen mixes when they are alive or dead (Mathie-

son and Wilson, 2009). The use of separate antigen preparationsmade from egg secretions (representing the live egg), and thenhatch fluid and miracidial proteins (representing the dead/disinte-grating egg) may provide a clearer picture. Eggs that are trapped inthe liver will die within approximately 20 days but total destruc-tion of the egg takes weeks or even months (Lambertucci, 1993).Consistent with this timescale, HSC activation was detectable after6 weeks coinciding with the TH2 shift and downmodulation ofgranuloma size (Bartley et al., 2006). Data presented in this studysuggests proteins leaking from dead eggs will continue to influencethe local microenvironment. After the eggs have been destroyedthe granuloma itself can disappear but localised fibrosis can persistand contribute to host pathology (Lambertucci, 1993; Burke et al.,2009). In the absence of inhibitory egg factors local HSC may sub-sequently become activated and contribute to this persistencewhich may be sustained from the more generalised immuneresponse.

In conclusion, the data suggest that schistosome eggs exert sig-nificant influence on HSC in their immediate proximity, causinginhibition of fibrogenesis and the adoption of a more quiescentHSC phenotype. These data may have significance for our under-standing of hepatic fibrogenesis in schistosomiasis. Further workis necessary, to clarify the nature of these effects in terms of immu-nopathogenesis and to try and identify agents responsible.

Acknowledgments

The authors gratefully acknowledge the support of University ofSalford research infrastructure fund for financial support (J.A.) andfor the award of a research studentship to B.A. The authors thankProf. S. Friedman, Mt. Sinai School of Medicine, New York, for LX-2 hepatic stellate cells and Prof. A. Wilson, York University, UKfor the supply of S. mansoni infected livers. The authors would alsolike to acknowledge the technical assistance of Mr. S. Bennett dur-ing preliminary experiments.

References

Andrade, Z.A., Guerret, S., et al., 1999. Myofibroblasts in schistosomal portal fibrosisof man. Memórias do Instituto Oswaldo Cruz 94, 87–93.

Bartley, P.B., Ramm, G.A., et al., 2006. A contributory role for activated hepaticstellate cells in the dynamics of Schistosoma japonicum egg-induced fibrosis.International Journal of Parasitology 36, 993–1001.

Blomhoff, R., Wake, K., 1991. Perisinusoidal stellate cells of the liver: importantroles in retinol metabolism and fibrosis. FASEB Journal 5, 271–277.

Booth, M., Mwatha, J.K., et al., 2004. Periportal fibrosis in human Schistosomamansoni infection is associated with low IL-10, low IFN-gamma, high TNF-alpha,or low RANTES, depending on age and gender. Journal of Immunology 172,1295–1303.

Boros, D.L., Lande, M.A., 1983. Induction of collagen synthesis in cultured humanfibroblasts by live Schistosoma mansoni eggs and soluble egg antigens (SEA).American Journal of Tropical Medicine and Hygiene 32, 78–82.

Burke, M.L., Jones, M.K., et al., 2009. Immunopathogenesis of humanschistosomiasis. Parasite Immunology 31, 163–176.

Caldas, I.R., Campi-Azevedo, A.C., et al., 2008. Human schistosomiasis mansoni:immune responses during acute and chronic phases of the infection. ActaTropica 108, 109–117.

Chang, D., Ramalho, L.N., et al., 2006. Hepatic stellate cells in human Schistosomiasismansoni: a comparative immunohistochemical study with liver cirrhosis. ActaTropica 97, 318–323.

Chitsulo, L., Engels, D., et al., 2000. The global status of schistosomiasis and itscontrol. Acta Tropica 77, 41–51.

Crompton, D.W., 1999. How much human helminthiasis is there in the world?Journal of Parasitology 85, 397–403.

de Jesus, A.R., Magalhaes, A., et al., 2004. Association of type 2 cytokines withhepatic fibrosis in human Schistosoma mansoni infection. Infection andImmunity 72, 3391–3397.

Diaz, A., Allen, J.E., 2007. Mapping immune response profiles: the emerging scenariofrom helminth immunology. European Journal of Immunology 37, 3319–3326.

Dooley, S., Delvoux, B., et al., 2001a. Transforming growth factor beta signaltransduction in hepatic stellate cells via Smad2/3 phosphorylation, a pathwaythat is abrogated during in vitro progression to myofibroblasts. TGFbeta signaltransduction during transdifferentiation of hepatic stellate cells. FEBS Letters502, 4–10.

420 B. Anthony et al. / Experimental Parasitology 124 (2010) 409–420

Dooley, S., Streckert, M., et al., 2001b. Expression of Smads during in vitrotransdifferentiation of hepatic stellate cells to myofibroblasts. Biochemical andBiophysical Research Communications 283, 554–562.

Everts, B., Perona-Wright, G., et al., 2009. Omega-1, a glycoprotein secreted bySchistosoma mansoni eggs, drives Th2 responses. Journal of ExperimentalMedicine 206, 1673–1680.

Faveeuw, C., Mallevaey, T., et al., 2003. Schistosome N-glycans containing core alpha3-fucose and core beta 2-xylose epitopes are strong inducers of Th2 responsesin mice. European Journal of Immunology 33, 1271–1281.

Forbes, S.J., Russo, F.P., et al., 2004. A significant proportion of myofibroblasts are ofbone marrow origin in human liver fibrosis. Gastroenterology 126, 955–963.

Freedman, D.O., Ottesen, E.A., 1988. Eggs of Schistosoma mansoni stimulateendothelial cell proliferation in vitro. Journal of Infectious Disease 158, 556–562.

Friedman, S.L., 2000. Molecular regulation of hepatic fibrosis, an integrated cellularresponse to tissue injury. Journal of Biological Chemistry 275, 2247–2250.

Friedman, S.L., 2003. Liver fibrosis – from bench to bedside. Journal of Hepatology38 (Suppl. 1), S38–S53.

Grimaud, J.A., Borojevic, R., 1986. Portal fibrosis: intrahepatic portal vein pathologyin chronic human Schistosomiasis mansoni. Journal of Submicroscopic Cytology18, 783–793.

Guyot, C., Lepreux, S., et al., 2006. Hepatic fibrosis and cirrhosis: the(myo)fibroblastic cell subpopulations involved. International Journal ofBiochemistry and Cell Biology 38, 135–151.

Hazra, S., Xiong, S., et al., 2004. Peroxisome proliferator-activated receptor gammainduces a phenotypic switch from activated to quiescent hepatic stellate cells.Journal of Biological Chemistry 279, 11392–11401.

Ikegami, T., Zhang, Y., et al., 2007. Liver fibrosis: possible involvement of EMT. CellsTissues Organs 185, 213–221.

Kanse, S.M., Liang, O., et al., 2005. Characterisation and partial purification ofSchistosoma mansoni egg-derived pro-angiogenic factor. Molecular andBiochemical Parasitology 144, 76–85.

Lambertucci, J.R., 1993. Schistosoma mansoni: pathological and clinical aspects. In:Jordan, P., Webbe, G., Sturrock, R.F. (Eds.), Human Schistosomiasis. CabInternational, Cambridge, pp. 195–235.

Leask, A., Abraham, D.J., 2004. TGF-beta signaling and the fibrotic response. FASEBJournal 18, 816–827.

Loker, E.S., 1983. A comparative study of the life-histories of mammalianschistosomes. Parasitology 87, 343–369.

Mathieson, W., Wilson, R.A., 2009. A comparative proteomic study of theundeveloped and developed Schistosoma mansoni egg and its contents: themiracidium, hatch fluid and secretions. International Journal for Parasitology,doi:10.1016/j.ijpara.2009.10.014.

Meindl-Beinker, N.M., Dooley, S., 2008. Transforming growth factor-beta andhepatocyte transdifferentiation in liver fibrogenesis. Journal ofGastroenterology and Hepatology 23 (Suppl. 1), S122–S127.

Miyahara, T., Schrum, L., et al., 2000. Peroxisome proliferator-activated receptorsand hepatic stellate cell activation. Journal of Biological Chemistry 275, 35715–35722.

Perona-Wright, G., Jenkins, S.J., et al., 2006. Dendritic cell activation and function inresponse to Schistosoma mansoni. International Journal of Parasitology 36, 711–721.

Pfaffl, M.W., Horgan, G.W., et al., 2002. Relative expression software tool (REST) forgroup-wise comparison and statistical analysis of relative expression results inreal-time PCR. Nucleic Acids Research 30, e36.

Rachfal, A.W., Brigstock, D.R., 2003. Connective tissue growth factor (CTGF/CCN2) inhepatic fibrosis. Hepatology Research 26, 1–9.

Reynaert, H., Thompson, M.G., et al., 2002. Hepatic stellate cells: role inmicrocirculation and pathophysiology of portal hypertension. Gut 50, 571–581.

Sadler, C.H., Rutitzky, L.I., et al., 2003. IL-10 is crucial for the transition from acute tochronic disease state during infection of mice with Schistosoma mansoni.European Journal of Immunology 33, 880–888.

Sato, M., Kojima, N., et al., 1998. Induction of cellular processes containingcollagenase and retinoid by integrin-binding to interstitial collagen in hepaticstellate cell culture. Cell Biology International 22, 115–125.

Schramm, G., Falcone, F.H., et al., 2003. Molecular characterization of aninterleukin-4-inducing factor from Schistosoma mansoni eggs. Journal ofBiological Chemistry 278, 18384–18392.

Senoo, H., Imai, K., et al., 1996. Three-dimensional structure of extracellular matrixreversibly regulates morphology, proliferation and collagen metabolism ofperisinusoidal stellate cells (vitamin A-storing cells). Cell Biology International20, 501–512.

She, H., Xiong, S., et al., 2005. Adipogenic transcriptional regulation of hepaticstellate cells. Journal of Biological Chemistry 280, 4959–4967.

Smithers, S.R., Terry, R.J., 1965. The infection of laboratory hosts with cercariae ofSchistosoma mansoni and the recovery of the adult worms. Parasitology 55, 701–710.

Tan, J.T., McLennan, S.V., et al., 2008. Connective tissue growth factor inhibitsadipocyte differentiation. American Journal of Physiology: Cell Physiology 295,C740–C751.

Tsukamoto, H., 2005. Fat paradox in liver disease. Keio Journal of Medicine 54, 190–192.

Uemura, M., Swenson, E.S., et al., 2005. Smad2 and Smad3 play different roles in rathepatic stellate cell function and alpha-smooth muscle actin organization.Molecular Biology of the Cell 16, 4214–4224.

WHO, 2002. Prevention and control of schistosomiasis and soil-transmittedhelminthiasis. WHO Technical Report Series 912.

Wickert, L., Steinkruger, S., et al., 2002. Quantitative monitoring of the mRNAexpression pattern of the TGF-beta-isoforms (beta 1, beta 2, beta 3) duringtransdifferentiation of hepatic stellate cells using a newly developed real-timeSYBR Green PCR. Biochemical and Biophysical Research Communications 295,330–335.

Wilson, M.S., Mentink-Kane, M.M., et al., 2007. Immunopathology ofschistosomiasis. Immunology and Cell Biology 85, 148–154.

Wyler, D.J., Tracy, J.W., 1982. Direct and indirect effects of soluble extracts ofSchistosoma mansoni eggs on fibroblast proliferation in vitro. Infection andImmunity 38, 103–108.

Wynn, T.A., Thompson, R.W., et al., 2004. Immunopathogenesis of schistosomiasis.Immunology Reviews 201, 156–167.

Xu, L., Hui, A.Y., et al., 2005. Human hepatic stellate cell lines, LX-1 and LX-2: newtools for analysis of hepatic fibrosis. Gut 54, 142–151.

Zhao, C., Chen, W., et al., 2006. PPARgamma agonists prevent TGFbeta1/Smad3-signaling in human hepatic stellate cells. Biochemical and Biophysical ResearchCommunications 350, 385–391.