Redox Balance Mechanisms in Schistosoma mansoni Rely on Peroxiredoxins and Albumin and Implicate Peroxiredoxins as Novel Drug Targets*

Schistosoma mansoni, a causative agent of schistosomiasis, resides in the hepatic portal circulation of their human host up to 30 years without being eliminated by the host immune attack. Production of an antioxidant “firewall,” which would neutralize the oxidative assault generated by host immune defenses, is one proposed survival mechanism of the parasite. Schistosomes lack catalase, the main H2O2-neutralizing enzyme of many organisms, and their glutathione peroxidases are in the phospholipid class with poor reactivity toward H2O2. Evidence implicates peroxiredoxins (Prx) as providing the main enzymatic activity to reduce H2O2 in the parasite. Quantitative monitoring of Prx mRNAs during parasite life cycle indicated that Prx proteins are differentially expressed, with highest expression occurring in adult stages (oxidative resistant stages). Incubation of schistosomula with Prx1 double-stranded RNA knocked down total Prx enzymatic activity and resulted in lowered survival of cultured parasites compared with controls demonstrating that Prx are essential parasite proteins. These results represent the first report of lethal gene silencing in Schistosoma. Investigation of downstream effects of Prx silencing revealed an abrupt increase of lipid peroxides and the generation of several oxidized proteins. Using mass spectrometry, parasite albumin and actin were identified as the main oxidized proteins. Gene expression analysis showed that schistosome albumin was induced by oxidative stress. This study highlights Prx proteins as essential parasite proteins and potential new targets for anti-schistosome drug development and albumin as a novel, sacrificial oxidant scavenging protein in parasite redox regulation.

Schistosomiasis is a severe, debilitating parasitic disease currently affecting 250 million people in the tropics, resulting in 20 million individuals with severe morbidity and 280,000 deaths annually (1,2). In the coming years the Schistosomiasis Control Initiative and endemic country programs will treat hundreds of millions of people with praziquantel, the single anti-schistosomiasis drug in widespread use (3). There is already clinical and laboratory evidence for the existence of praziquantel resistance parasites (4), and widespread use is expected to generate strong selective pressure for drug resistance. Clearly there is an urgent need for new anti-schistosome drugs.
The presence of the parasite in the host hepatic mesenteries puts them under oxidative stress from immune-generated radicals as well as those potentially generated in the parasite during respiration and the breakdown and consumption of host hemoglobin with the concurrent release of toxic heme and ferrous ions. The addition of one electron to O 2 produces superoxide, which is rapidly reduced to H 2 O 2 by superoxide dismutase. The H 2 O 2 formed is itself able to diffuse and cause cellular damage and must be neutralized to prevent the formation the more damaging hydroxyl radical. In many organisms intracellular H 2 O 2 is eliminated by catalase, glutathione peroxidase (GPx), 2 and/or peroxiredoxins (Prxs). Schistosomes have abundant superoxide dismutase but completely lack catalase and have relatively low levels of GPx (5,6). Furthermore, the known Schistosoma mansoni GPx is in the phospholipid hydroperoxide GPx class (GPx4) with poor reactivity toward H 2 O 2 (7,8). The main function of this GPx class may be to protect biomembranes from oxidative damage. This evidence and our previous work (9 -12) suggest that the schistosome redox system is significantly different from that of the host and that Prxs provide significant, perhaps the vast majority of hydrogen peroxide reducing activity in schistosomes.
Peroxiredoxins are members of a recently identified family of antioxidants involved in the detoxification of hydrogen peroxide and other hydroperoxides (13,14). Prxs are also involved in redox balance and redox signaling processes and affect protein phosphorylation, transcriptional regulation, and apoptosis in animals (14 -18). There are three families of Prx proteins; typical 2-Cys Prx, atypical 2-Cys Prx, and 1-Cys Prx. All Prx have a reactive Cys in the N-terminal portion of the protein.
Both classes of 2-Cys Prx contain a second C-terminal-located Cys. Mammalian cells express six isoforms of Prx; four typical 2-Cys (Prx I-IV), one atypical 2-Cys (Prx V), and one 1-Cys Prx (Prx VI). Both Prx III and Prx IV are found in the mitochondria.
A number of studies confirm that Prx protect against oxidative stress in mammalian cells. Hemolytic anemia and increases in malignancies are seen in cytoplasmic Prx knock out animals (19,20). Depletion of mitochondrial Prxs results in increased intracellular levels of H 2 O 2 , sensitized cells to induction of apoptosis and lower survival rates (21,22).
Three typical 2-Cys Prx are present in S. mansoni (12). The three S. mansoni Prx proteins are ϳ65% identical to each other and to typical 2-Cys Prx from mammals. Although S. mansoni Prx1 displays typical Prx kinetics and reactivity with Trx, Prx2 and Prx3 have unusual, saturable kinetics and higher reactivity with GSH than with Trx (12). Furthermore, Prx1 is unusual in that it is highly resistant to oxidative inac-tivation (12). No atypical 2-Cys or 1-Cys Prx appears to be present in the S. mansoni genome.
In this context we hypothesize that S. mansoni Prxs are essential in parasite redox balance mechanisms. In this study we investigate the expression schistosome Prx proteins during the parasite lifecycle and their subcellular localizations. Because schistosome Prx proteins are biochemically distinct from host Prx (12), if they are essential for parasite survival they may prove to be novel targets for drug development. Because no specific Prx inhibitors are known, to test this hypothesis we have used RNA interference (RNAi) to silence Prx expression in larval parasites (schistosomula). Silencing of Prx leads to decreases in survival of parasites that depended on the level of oxidative stress they experienced. Silencing of Prx expression is not compensated by an increase in GPx expression but, instead, by expression of a parasite albumin. Schistosome albumin along with Prx appears to play a significant role in redox defenses of this important human parasite.

EXPERIMENTAL PROCEDURES
Monoclonal Antibody Production-Recombinant S. mansoni Prxs were produced as His 6 proteins and prepared as described (9,12). Monoclonal antibodies against recombinant S. mansoni Prxs were generated using standard protocols (23). SP2/0 myeloma cells were kindly provided by Dr. Laura Vogel, Illinois State University. The specificities of the antibodies were checked with enzyme-linked immunosorbent assay and Enhanced Chemiluminescence Western blotting (Amersham Biosciences) against recombinant Prx proteins.
Parasite Preparation-Adult S. mansoni (NMRI strain) were obtained by perfusion of 6 -7-week-infected mice (Biomedical Research Institute) by the method of Duvall and DeWitt (24). Eggs were isolated from the liver of infected mice as described (25). Cercariae were obtained from infected Biomphalaria glabrata snails (Biomedical Research Institute) after exposing the snails to light for 1-2 h according to Lewis (25). Schistosomula were prepared from cercariae by mechanically separate the tails by vortexing. The resulting schistosomula were separated from the tails by centrifugation over 40% Percoll at 500 ϫ g for 15 min at 4°C as described (25). Schistosomula were kept at 37°C in 5% CO 2 in RPMI 1640 containing 10 mM Hepes, pH 7, 100 units/ml penicillin, 100 g/ml streptomycin, and 10% fetal calf serum (Atlanta Biologicals) for 4 h, 24 h, 5 days, and 10 days. The 4-h, 24-h, and 5-day schistosomula from in vitro cultures represent skin-stage parasites, and the 10-day cultured schistosomula represent lung-stage parasites. Liver-stage parasites were obtained by perfusion of livers from mice infected 23 days previously with 500 -1000 cercariae. This study was approved by the Institutional Animal Care and Use Committee of Illinois State University (08-2002; Department of Health and Human Services animal welfare assurance number A3762-01).
GPx2 Cloning and Analysis-The S. mansoni EST and genomic data base at The Institute for Genomic Research (TIGR) and the Sanger Institute were queried with the S. mansoni GPx1 sequence (7) to identify novel S. mansoni GPx genes. A TIGR tentative consensus sequence, TC4112, was identified that contained an EST (BF936248), which was generously provided by Dr. Philip LoVerde (State University at Buffalo). The nucleotide sequence of the complete insert of BF936248 was obtained using specific oligonucleotides and fluorescent dideoxy-terminator sequencing on an Applied Biosystems Inc. 310 automated sequencer. This sequence was deposited in GenBank TM with the accession number AY729668. The sequence was examined for the characteristic features of the selenocysteine insertion sequence (SECIS) element using SECISearch 2.0 software program (26). Subcellular Fraction Preparation-Adult parasites were disrupted in 10ϫ isolation buffer (0.33 M sucrose, 1 mM EDTA, 1 mM Tris-HCl, pH 7.4, 1% bovine serum albumin) at 0°C. Homogenates were settled by gravity for 20 min at 0°C to sediment cell debris and unbroken cells and then centrifuged at 900 ϫ g for 30 min at 4°C to collect a nuclear pellet. The resulting supernatant was centrifuged at 10,000 ϫ g for 30 min to collect a mitochondrial pellet. The remaining supernatant was centrifuged at 120,000 ϫ g for 1 h to collect the microsomal fraction. The final supernatant was considered to be the cytosolic fraction. Pelleted fractions were suspended in isolation buffer with 200 M protease inhibitor cocktails. Protein concentrations were determined using the Bradford assay according to the manufacturer's protocol (Bio-Rad). Lactate dehydrogenase and succinate dehydrogenase activities (27,28) were used as controls for cytoplasmic and mitochondrial fractions, and histone and sarco/endoplasmic reticulum calcium ATPase proteins, detected by Western blot, served as controls for the nuclear and microsomal fractions, respectively.
dsRNA Preparation-The entire coding sequence of S. mansoni Prx1 was amplified by PCR using primer (5Ј-AGCTAGCGAGATTTC-CAAAGCATATGGT) and (5Ј-GAAGCTTTCATGTTTCATCTA-CAGATCGTCC) with 2 ng of Prx1-pRSetA as template and was cloned into pCRII-Topo vector (Invitrogen). In this plasmid the Prx1 open reading frame is flanked by SP6 and T7 RNA polymerase promoter sites. The resulting Prx1-pCRII plasmid was linearized and purified, and in vitro RNA transcription was conducted in 20-l reactions according to manufacturer's protocols (Megascript, Ambion). Non-schistosome dsRNA (irrelevant dsRNA) was synthesized from pCRII-TOPO vector (280 base pairs between SP6 and T7 promoter sites) and used as the negative control. Equimolar amounts of single-stranded RNAs were mixed and heated to 68°C for 25 min and cooled to room temperature for annealing. The dsRNAs were checked for purity and integrity by 1.2% agarose gel electrophoresis.
Total RNA Isolation/Reverse Transcriptase-PCR-Total RNA was harvested from parasites after the course of an experiment by TRI reagent following the manufacturer's instructions (Sigma). Complementary DNA was synthesized using 1 g of RNA and oligo (dT) x primer with 200 units of Thermoscript reverse transcriptase in a 25-l reaction according to manufacturer's protocol (Invitrogen).
Quantitative Real-time RT-PCR-Real-time RT-PCR was carried out on Applied Biosystems 7300 real-time PCR system using SYBR green dye (Molecular Probes). Primers for real-time RT-PCR were designed on Primer Express software (Applied Biosystems) and selected for melting temperature (T m ) between 58 and 60°C and amplicon size of 50 -150 base pairs (Table 1). All target sequences and glyceraldehydephosphate-dehydrogenase (GAPDH) were validated for amplification efficiency (User Bulletin #2: ABI PRISM 7700 sequence detection system). Samples of cDNA from different parasite stages were diluted 1:5 with sterile, nuclease-free water and used in real-time PCR or diluted 1:500 for internal control gene levels. The reactions were performed in 25 l according to previously published methods (29). Samples were analyzed in triplicate or quadruplicate, and product purity was checked through dissociation curves at the end of real-time PCR cycles. Statistical analysis of real-time RT-PCR data were carried out with the 2 Ϫ⌬⌬CT method (30).
RNAi Silencing of S. mansoni Prx-Schistosomula were prepared as described above. Methods for RNAi silencing were similar to those previously described (31). Briefly, double-stranded RNAs were added to a final concentration of 25 g/ml (Prx1 or irrelevant non-schistosome RNA control) to 500 g/ml schistosomula in 50 l of water and 250 l of RPMI medium containing 250 units/ml penicillin and 150 g/ml streptomycin in 24-well tissue culture plates. Schistosomula were incubated at 25°C for 25 min and then at 37°C in a water-jacketed incubator with 5% CO 2 atmosphere for the course of the experiment. Fifty microliters of fresh medium was added every 2 days to the wells. The cultured parasites were checked on a daily basis for viability. In the case of parasites under in vitro oxidative stress, H 2 O 2 was added after 24 h of dsRNA treatment to ensure the maximum uptake of dsRNA by schistosomula. Controls without dsRNA treatment were run in parallel. Each treatment was carried out in triplicate, and experiments were done at least twice.
The effect of RNAi on schistosomula cultured at reduced O 2 tension was also investigated. Media were filter-sterilized and purged with different O 2 content gases (0% O 2 , 5% O 2 blood gas, or 20% O 2 ) plus 5% CO 2 before use for 20 min. The concentrations of dissolved O 2 in the media were determined using an O 2 electrode. Parasites were cultured during the course of the experiment under different O 2 tensions at 37°C in a sealed modular incubator chamber (Billups-Rothenberg) flushed with the appropriate gaseous mix. Fifty microliters of freshly purged media were added every 2 days to the wells. The viability of parasites was checked on a daily basis.
Enhanced Chemiluminescent Western Blots-Total proteins from S. mansoni life stages and subcellular fractions were probed with monoclonal antibodies against Prx1, Prx2, and Prx3. The proteins were fractionated by 16% SDS-PAGE, blotted to polyvinylidene difluoride membranes, and blocked with blocking buffer (5% nonfat dry milk, 0.1% Tween 20 in phosphate-buffered saline) overnight at 4°C. The membranes were probed with the primary monoclonal antibodies at a 1:2500 dilution at 25°C for 2 h followed with 2 h of incubation secondary mouse antibody (1:2000) coupled with horseradish peroxidase (Jackson ImmunoResearch Laboratories). Protein bands were visualized using luminol as a substrate according to manufacturer's instruction (Amersham Biosciences). Additional antibodies used were against chicken actin (Developmental Studies Hybridoma Bank, The University of Iowa) detected with a goat anti-mouse IgM secondary antibody (Jackson ImmunoResearch Laboratories), against rabbit skeletal muscle sarco/ endoplasmic reticulum calcium ATPase (Affinity BioReagents), against bovine serum albumin (Sigma), against human ␤-tubulin (Santa Cruz Biotechnology, Inc.), and against human histone H1 (Santa Cruz).
Specific Activity Determinations of S. mansoni Prx and GPx-Reduction of H 2 O 2 or cumene hydroperoxide by schistosomula homogenates was measured in coupled reactions with recombinant Escherichia coli thioredoxin and thioredoxin reductase (12) or glutathione and yeast glutathione reductase (Sigma) by consumption of NADPH at 25°C in Shimadzu UV-1601 spectrophotometer as previously described (12).
Lipid and Protein Oxidation Products-Measurement of malondialdehyde (MDA) and 4-hydroxyalkenals (HAE) was used as an indicator of lipid peroxidation. The assay is based on the reaction of a chromogenic reagent, N-methyl-2-phenylindole, with MDA and HAE at 45°C. The resulting chromophore has maximal absorbance at 586 nm. RNAi and irrelevant dsRNA-treated schistosomula were collected after 0, 3, 6, and 9 days of dsRNA treatment. The samples were sonicated in phosphate buffer, pH 7.4, and 0.5 M butylated hydroxytoluene was added to inhibit any further oxidation of lipids. Sonicates were centrifuged to remove any cellular debris at 3000 ϫ g at 4°C for 10 min. Protein concentrations of the sonicates were determined using the Bradford assay. Two hundred microliters of the sonicate were mixed gently by briefly vortexing with 487.5 l of 10.3 mM N-methyl-2-phenylindole (R1) in acetonitrile, 162.5 l of 10 M FeCl 3 in methanol, and 150 l of methanesulfonic acid. The reactions were incubated at 45°C for 60 min. Absorbance of the samples was measured at 586 nm. Nonspecific changes in absorbance due to sample contribution were corrected by the sample blank (75% acetonitrile, 25% 10 M FeCl 3 in methanol). The MDA concentrations in samples were determined by comparison to the standard curve of MDA in methanesulfonic acid and HCl according to manufacturer's protocols (Oxford Biomedical Research). A range of 0.1-20 M MDA standard was used.
Schistosomula homogenates were derivatized with dinitrophenyl (DNP) hydrazine through the reaction of carbonyls produced by the reaction of proteins with reactive oxygen species with DNP hydrazine (32). Both Prx1 dsRNA-and irrelevant dsRNA-treated schistosomula were collected after 0, 3, 6, and 9 days of treatment. The samples were sonicated in phosphate buffer, pH 7.4, with 2% 2-mercaptoethanol to inhibit further protein oxidation. Sonicates were centrifuged to remove cell debris at 3000 ϫ g at 4°C for 10 min. Protein concentrations of the sonicates were determined using the Bradford assay (Bio-Rad). Five microliters of schistosomula homogenate were used for the DNP hydrazine derivatization reaction according to the manufacturer's protocol (Oxyblot TM protein oxidation detection kit, Chemicon International).
Protein MS Analysis-The schistosomula homogenates from the RNAi treatment (days 3 and 9) were derivatized as described earlier and then neutralized to pH 7.4 with 2 M Tris, pH 9. The neutralized samples were incubated with rabbit anti-DNP antibodies (Sigma) for 4 h at 4°C at 1:150 dilution. The samples were diluted 1:1 with binding buffer (0.1 M Tris, 0.15 M NaCl, pH 8) to obtain the optimum ionic strength and pH for protein A binding. Prepacked AffinityPak TM protein A columns (Pierce) were used for the purification. The column was equilibrated with 5 ml of the binding buffer, the samples were applied to the column, and the column was then washed with 15 ml of binding buffer. The DNP-labeled antibody conjugated proteins were eluted from the column using 4 ml of elution buffer (0.1 M glycine, pH 2.5). The eluted materials were collected in 0.3-ml fractions and immediately neutralized with 30 l of 1 M Tris, pH 9. The eluted fractions were checked by 16% SDS-PAGE for the labeled proteins. The DNP-labeled protein fractions were also checked by Western blot analysis.
Protein bands from Coomassie Blue-stained SDS-PAGE were excised and subjected to in-gel digestion protocol (33) using trypsin (Promega) and analyzed on a Waters Q-ToF Ultima mass spectrometer with a Waters CapLC using a Waters C-18 Symmetry Nanoease column oper-

RESULTS
Monoclonal Antibody Production and Specificity-To understand the role of Prx in parasite redox mechanisms, we generated Prx paralogspecific monoclonal antibodies. Hybridomas were screened to ensure the specificity of the monoclonal antibodies against S. mansoni Prx proteins at least twice by enzyme-linked immunosorbent assay and Western blotting. Hybridomas that produced antibodies with no crossreactivity with different Prx proteins were selected and expanded (results not shown).
Characterization of GPx2, a Second GPx in S. mansoni-Analysis of the S. mansoni EST and genomic databases identified one additional GPx gene. The S. mansoni GPx2 cDNA was found to be 986 bases long encoding a 178-amino acid protein (GenBank TM accession AAU34080). Analysis of the GPx2-predicted peptide identified a 19-amino acid leader sequence with a likely cleavage site between amino acids 18 and 19 (34). The first in-frame UGA codon of the open reading frame is likely to encode a selenocysteine (Sec ϭ U), as was also found in GPx1 (7,8,35) and in 5 of 7 classes of mammalian GPx proteins (36). The second in-frame UGA is the actual stop codon. The validity of the selenocysteine codon is supported by the presence of a SECIS element in the 3Ј-non-coding region of the GPx2 mRNA. The non-Watson-Crick base pairing of the element of the SECIS stem occurs between bases 802-805 and 845-848, and the unpaired adenosine residues occur at bases 818 -820 (Fig. 1C). The SECIS element falls into the form 1 category as predicted by SECISearch 2.0 (26).
Schistosome GPx2-predicted protein has 55% identity over a 149 amino acids with GPx1 from S. mansoni and 48% identity over 160 amino acids with human phospholipid hydroperoxide GPx4 (Fig. 1A). Schistosome GPx2 has less than 36% identity to the other classes of human GPx and clusters with S. mansoni GPx1 and human GPx4 in a phylogenetic analysis (Fig. 1B). Therefore, both schistosome GPx proteins are predicted to be biochemically more similar to the phospholipid hydroperoxide GPx family than other GPx families.
Stage-specific Expression and Subcellular Localization of S. mansoni Redox Proteins-The abundances of mRNAs for schistosome H 2 O 2reducing enzymes during the mammalian stages of the parasite lifecycle were determined using real-time RT-PCR ( Fig. 2A). To standardize results from different samples, GAPDH, tubulin, and acidic ribosomal protein mRNAs and 18 S ribosomal RNA were tested for use as an internal control for real-time PCR. GAPDH was found to be the most consistent and reliable control through parasite different stages (data not shown) and was used in subsequent real-time RT-PCR studies. All target sequences and GAPDH were validated for amplification efficiency, and all were found to have absolute values of the slope of ⌬C T versus log input total RNA less than 0.1 (results not shown). In general, Prx mRNA abundance was Prx1 Ͼ Prx2 Ͼ Prx3, and all three Prx mRNAs were seen to gradually increase during parasite development. GPx mRNAs also showed a trend to increase during parasite development and were, in general, less abundant than Prx mRNAs ( Fig. 2A).
Schistosome Prx proteins were found to be differentially expressed through parasite lifecycle from the infectious cercariae to mature adult stages in the mammalian host and the transmission egg stage (Fig. 2B). Prx mRNA abundance was reflected in the corresponding protein abundance (Prx1 Ͼ Prx2 Ͼ Prx3). Prx1 protein gradually increased in abundance during parasite development. Prx1 was most abundant in eggs and adult females. Very little Prx2 protein was detected in cercariae; there was a large increase in abundance during the first 24 h after transformation and a gradual increase during the remainder of the parasite development. Prx3 protein is absent or undetectable in cercariae and eggs and present at relatively low levels in other stages; adult female worms appear to have the highest levels. Adult female parasites appear to express all Prx proteins at higher levels than adult males. Western blot analysis on adult worm subcellular fractions revealed that Prx1 and Prx2 are cytosolic proteins, whereas Prx3 was localized to the mitochondria (Fig. 2C).
Prx Silencing by RNAi-Our first goal was to determine whether fulllength Prx1 dsRNA could silence the target gene. Using Prx1-specific dsRNA, mRNAs for the entire Prx family were silenced in schistosomula (Fig. 3A). Peroxiredoxin transcripts that were differentially knocked down with Prx1 reduced the most, 250-fold decreased from control parasites; Prx2 and Prx3 decreased 39-and 1.6-fold, respectively. No significant difference in Prx transcript abundances was found between control and non-schistosome, irrelevant dsRNA-treated control para- sites (results not shown). Expression of GPx1 and GPx2 was modestly induced (5.7-and 7.5-fold, respectively) in response to Prx silencing (Fig. 3A).
Treatment with Prx1 dsRNA resulted in large decreases in Prx1 and Prx2 proteins (Fig. 3B). In these studies Prx3 was present at levels below detection by Western blotting (results not shown). Furthermore, treatment of schistosomula with Prx1 dsRNA reduced total Prx enzyme activity, measured as H 2 O 2 reduction in the pres- Parasite survival during Prx RNAi-silencing was increased under reduced oxygen tension (Fig. 3D). When cultured under either 5% O 2 or 20% O 2 , parasites treated with Prx-specific dsRNA had similarly reduced survival compared with controls ( p Ͻ 0.001). However, in 0% O 2 parasites treated with Prx-specific dsRNA had the same survival as controls ( p Ͼ 0.05).
Downstream Effects of Silencing Prx Expression-To investigate downstream effects of Prx silencing by RNAi, we monitored lipid peroxides MDA and HAE and protein carbonyls in dsRNA-treated schistosomula. Lipid peroxides were determined using N-methyl-2phenylindol reaction on days 0, 3, 6, and 9 after RNAi treatment. Significant increases in lipid peroxides were seen at days 6 and 9 in Prx1 dsRNA-treated schistosomula compared with control irrelevant dsRNA-treated schistosomula (Fig. 4A). Lipid peroxides were at or below detectable levels at days 0 and 3 in Prx1 dsRNA-treated schistosomula and at all times in control parasites.
Protein carbonyls were monitored during the course of an RNAi experiment to follow oxidative damage to proteins. Proteins were blotted and analyzed Western blot using an anti-DNP antibody. A single DNP-labeled protein of 85 kDa was seen on day 3 in Prx1 dsRNAtreated schistosomula (Fig. 4B). Several DNP-labeled proteins of 60 -97 kDa were seen on both days 6 and 9 in Prx1 dsRNA-treated schistosomula. No DNP-labeled proteins were detected in control schistosomula (Fig. 4B). After incubation with anti-DNP antibodies, the DNP-labeled proteins were purified by protein A affinity chromatography. The purity of purified proteins was checked by SDS-PAGE (Fig. 4C) and Western blot analysis (not shown). There was a one-to-one correspondence of bands staining on the SDS-PAGE and detected by the Western blot indicating that all proteins purified from the affinity column were DNP-labeled.
The 85-kDa-labeled protein seen on day 3 and the 65-kDa-labeled protein seen on day 6 were identified by nano-liquid chromatography/MS/MS as schistosome actin (GenBank TM accession P53470) in the day-3 sample and schistosome albumin (GenBank TM accession AF418550) in the day-6 sample (Fig. 5A). Sequence coverage of 33.8% for actin and 39.5% for albumin was achieved. We verified the presence of albumin in schistosomula samples by RT-PCR and Western blotting. Albumin mRNA or protein was only detected in parasites after Prx silencing (Fig. 5, B and C). The albumin PCR product was cloned, and the sequence obtained from one colony was a 100% identical match to the S. mansoni albumin accession (results not shown). The S. mansoni albumin gene encodes a protein of 608 amino acids with a predicted M r of 68,225 and with 82-87% identity to rodent serum albumins and 76% identity to human serum albumin (results not shown). Western blot analysis with anti-chicken actin antibodies of homogenate from both control and Prx1 dsRNAtreated schistosomula and affinity-purified labeled DNP identified the expected 43-kDa actin band in control schistosomular homogenates, both the 43-and an 85-kDa species in Prx1 dsRNA-treated schistosomular homogenates, and only an 85-kDa species in the DNP-labeled affinity-purified sample (Fig. 5D).
Expression of Redox Proteins and Albumin in Response to Oxidative Stress-Albumin expression was monitored during parasite development by RT-PCR. No transcripts were detected in any developmental stage of parasites isolated from infected hosts (results not shown). However, when either schistosomula or adult parasites in axenic culture were placed under oxidative stress (20% O 2 plus 100 M H 2 O 2 or 200 M H 2 O 2 ), albumin expression was induced (Fig. 6). Of the H 2 O 2 -neutralizing enzymes in the parasite, only the expression of Prx1 was coordinated with increases in oxidative stress (Fig. 6).

DISCUSSION
Schistosomes survive in their human host blood stream for as long as 30 years. In this environment worms must be able to cope with host immune-and self-generated reactive oxygen species. One defense mechanism the parasite may use is the production of an antioxidant firewall (37). Because schistosomes lack catalase and have low levels of GPx, it has been proposed that Prx plays a major role in the redox defense of the parasite. S. mansoni Prx proteins were found to be biochemically distinct from each other and from the human host Prx proteins (12). The present study extends these observations to better understand schistosome redox mechanisms.
A search of the S. mansoni genome data base (sequenced to 8ϫ coverage) to identify additional H 2 O 2 -reducing proteins identified a single additional GPx gene. Sequence analysis of this gene indicated that schistosome GPx2 has the highest similarity to schistosome GPx1 and the GPx4 class of mammalian phospholipid hydroperoxide GPx. Thus, schistosome GPx2 protein is predicted to have higher specific activity toward phospholipid hydroperoxides as was shown for schistosome GPx1 (8). Schistosome GPx2, like GPx1, is predicted to contain the rare amino acid selenocysteine. Furthermore, GPx2 is predicted to be a secreted protein. Previous studies showed conflicting results for GPx localization in adult schistosome worms. Using a polyclonal serum against the C-terminal 15 amino acids, Roche and co-workers (38) localized GPx specifically to vitelline ducts and cells and eggs. However, using a monoclonal antibody against full-length recombinant GPx (mutant U43C), Mei and LoVerde (6) localized GPx to the tegument and gut epithelium but also at lower levels in muscle and other cells. The most likely explanation for this discrepancy is that the C-terminal peptide antiserum recognized only GPx1, as the C-terminal regions of GPx1 and GPx2 have no similarity, whereas the monoclonal antisera used by Mei and LoVerde (6) recognized both GPx1 and GPx2. The surface/intestinal GPx immunoreactivity may be due to the presence of GPx2, supporting the evidence from the cDNA sequence that GPx2 is a secreted protein. In the absence of catalase and with GPx proteins in the phospholipid hydroperoxide GPx class, Prx does appear to play the major role in schistosome of H 2 O 2 reduction.
Several redox proteins/activities have been shown to increase during development in the human host with adult parasites, which display the highest tolerance to oxidative stress (5,6,37). However, the developmental expression of Prx in schistosomes has not been investigated. Using both real-time RT-PCR and Western blotting it was determined that S. mansoni Prx are differentially expressed through the parasite life cycle with the lowest abundance in larval stages and the highest levels in adults. Shortly after transformation to skin stage parasites, both Prx1 and Prx2 expression rapidly increases. All three Prx transcripts and proteins attained maximum levels in adult worms, especially in females. These observations are in line with those made previously on schistosome redox proteins, suggesting that Prx proteins also play an important role in the host-parasite interaction and neutralization of reactive oxygen species in adult worms.
Cell signaling by intracellular production of H 2 O 2 is an increasingly important mechanism (16 -18). However, because Prx proteins are very abundant in the cytosol of most eukaryotic cells, this creates the problem that H 2 O 2 produced as a signaling molecule would be rapidly reduced by Prx, calling into question the mechanism of signaling. Recently, it was proposed that a transient decrease in Prx activity by high concentrations of H 2 O 2 produced during the signaling process would allow a H 2 O 2 -dependent signal to be propagated (15). Inactivation of Prx occurs thorough oxidation of the peroxidatic cysteine to a cysteine sulfinic acid-containing inactive species. Two novel classes of proteins, sulfiredoxins (39 -41) and sestrins (42), are able to regenerate the cysteine thiol from the cysteine sulfinic acid, regenerating active Prx proteins. From both real-time RT-PCR and Western blotting it appeared that Prx1 was the most abundant Prx paralogue in all mammalian life stages. Schistosome Prx1 is unusual for eukaryotic 2-Cys Prx in that it is highly resistant in oxidative inactivation (12). The high abundance of Prx1 calls into question how, and if H 2 O 2 signaling occurs in schistosomes. Potentially one way to have an abundant resistant Prx and still maintain H 2 O 2 signaling would be to have differential localization of Prx proteins. Prx1 was found to be secreted by eggs (10) and in the tegument of adult worms. 3 If Prx1 were primarily present in parasite secretions or on their surface, redox signaling in schistosome cells through the inactivation of Prx2 could still occur.
As predicted from the cDNA sequences, Prx3 is localized to the mitochondria, whereas Prx1 and Prx2 are cytosolic proteins. Mitochondrial Prx3 was most highly expressed in adult female worms with lower expression in other stages. We have identified a thioredoxin gene encoding a mitochondrial protein and have evidence that the S. mansoni thioredoxin-glutathione reductase mRNA is alternatively spliced to produce a mitochondrial form of the protein. 4 Therefore, the mitochondria of schistosomes have a complete thioredoxin-based redox regulatory system.
Although we used a Prx1 dsRNA in RNAi silencing, all other Prx family members were silenced, as confirmed by real-time RT-PCR and Prx paralogue-specific Western blotting. We attribute this off-target silencing of both Prx2 and Prx3 to the high similarities between their nucleic acid sequences and Prx1 sequence. Overall, Prx1 and Prx2 are 70% identical, but there are regions of 30 bp having Ͼ90% identity. Prx1 and Prx3 are less conserved, at 55% overall identity, but there are regions of 20 -25 bp with Ͼ80% identity. However, the down-regulation of parasite Prx1 transcript (250-fold) was greater than the other Prxs  1.9-fold for Prx2 and Prx3, respectively). This is due in part to higher specificity of annealing of Prx1 mRNA and antisense RNA and due to the intrinsically higher expression of Prx1 mRNA. Non-schistosome, irrelevant dsRNA controls were run in parallel with Prx dsRNA treatments and showed no difference with untreated, control schistosomula.
Silencing of Prx caused a dramatic decline in schistosomula viability compared with control parasites. Silencing of Prx in parasites cultured in 20% O 2 or at the more physiologically relevant 5% O 2 caused significant increases in parasite mortality. Exogenous oxidative stress (200 M H 2 O 2 ) caused 100% killing of the dsRNA-treated schistosomula after 4 days of treatment. Although parasite survival was greatly reduced in 5% O 2 in the host venous system, parasites consume large amounts of RBC and are, therefore, exposed to higher O 2 and reactive heme than might be expected in venous blood and would normally be under significant oxidative stress. These data represent the first report of lethal gene silencing in Schistosoma.
It has been proposed that reactive oxygen species are primarily generated by the autoxidation of respiratory dehydrogenases (43). Schistosomes in the vertebrate host are largely lactic acid fermenters and use little if any oxidative phosphorylation for energy production (44). However, upon silencing of Prx proteins, cultured schistosomes are under increased oxidative stress due to the uncontrolled production of reactive oxygen species within the parasite, as evidenced by increases in lipid and protein oxidation products. Damage to parasite proteins and lipids certainly contribute to the decease in survival of the dsRNA-treated parasites compared with control parasites. This is supported by the evidence that silencing of Prx expression in parasites cultured in very low oxygen tension caused no decrease in survival compared with untreated or irrelevant RNA controls. The source of reactive oxygen species in schistosomes remains to be determined.
Because silencing Prx expression leads to increased oxidative stress in the parasite, we investigated if the worms responded with the induction of other protective redox proteins. The only additional H 2 O 2 -reducing activity known to be present in the worm cytoplasm is GPx. Both GPx1 and GPx2 mRNAs were increased in response to Prx silencing. However, the overall ability to reduce H 2 O 2 by schistosomula homogenate was decreased after Prx silencing despite GPx induction. The Trx-dependent reduction of H 2 O 2 decreased 85% in dsRNA-treated schistosomula and the GSH-dependent reduction of H 2 O 2 , a measure of both Prx and GPx activities (12), decreased 50% in dsRNA-treated schistosomula. Reduction of cumene hydroperoxide, a preferred substrate for S. mansoni GPx and some GSH S-transferases (45) and a secondary substrate for S. mansoni Prx, was increased by ϳ25% in dsRNA-treated parasites. These results suggest that in vivo, Prx are the main source for enzymatic reduction of H 2 O 2 , whereas GPx, GSH S-transferases, and Prx are active in the reduction of lipid hydroperoxides. The delay in the production of lipid hydroperoxides seen in Fig. 5A may be the result of the induction of GPx or the activity of the abundant GSH S-transferases of the parasite.
Surprisingly, a novel schistosome protein, albumin, was found to be the major oxidized protein species resulting from Prx-silencing. Furthermore, albumin expression was induced when parasites were placed under oxidative stress either through Prx silencing or by the addition of H 2 O 2 to the culture media. In contrast, of the known H 2 O 2 reducing proteins in S. mansoni, only Prx1 was coordinately expressed with oxidative stress. An increase in Prx2, GPx1, and GPx2 only occurred under the highest oxidative stress. This strongly suggests that albumin plays a significant role in redox defenses of the parasite. Albumin has previously been suggested to be a sacrificial or scavenger extracellular antioxidant protein involved (46 -48). Albumins may function as antioxidant proteins by binding heavy metals, thus preventing the generation of hydroxyl radicals via the Fenton reaction (49) or through direct reduction of ROS via the free Cys residue in the protein (48,50). By analogy to the mammalian albumins, the S. mansoni albumin protein has one unpaired Cys residue and could potentially directly reduce ROS. Although albumin may play a role in reactive oxygen species scavenging in other organisms, the strong reliance on albumin in redox defenses seen here in schistosomes is unprecedented. It remains to be determined if the oxidized albumin can be regenerated into an active, protective form.
The first oxidized protein detected in Prx-silenced parasites was schistosome actin. Actin is an abundant and ubiquitous cellular protein that functions in the generation and maintenance of cell morphology and polarity, in endocytosis and intracellular trafficking, and in contractility, motility, and cell division. Based on its molecular weight, DNPlabeled actin appeared to be in a dimeric form. This dimeric form of actin was only seen in Prx-silenced parasites. The actin dimers identified in this study differ from those studied previously (51) in that they are not cross-linked by sulfhydryls; they are resistant to reduction by DTT in reducing SDS-PAGE. Actin polymerization dynamics have been implicated in cell death and longevity through its affects on the intracellular redox state (52)(53)(54). Whether schistosome actin is oxidatively modified after Prx silencing due to its involvement in redox responses of the parasite or if its oxidation is due to its abundant nature remains to be determined.
Schistosomes appear to be poorly adapted to cope with oxidative stress. They completely lack catalase and have a restricted repertoire of Prx and GPx proteins. The loss of enzymatic H 2 O 2 -reducing activity from the silencing of Prx is not matched by a compensating increase in GPx or other enzymatic reduction but by the production of a sacrificial protein; this must be an energetically costly protective mechanism for the parasite. Silencing of parasite Prx activity is lethal, indicating that these proteins are essential for survival of the parasite. Because S. man- soni Prxs are biochemically distinct from host Prx and essential for parasite survival, it may be possible to design parasite-specific Prx inhibitors to be used in the future to control schistosomiasis.