Thioredoxin Glutathione Reductase as a Novel Drug Target: Evidence from Schistosoma japonicum

Background Schistosomiasis remains a major public health concern affecting billions of people around the world. Currently, praziquantel is the only drug of choice for treatment of human schistosomiasis. The emergence of drug resistance to praziquantel in schistosomes makes the development of novel drugs an urgent task. Thioredoxin glutathione reductase (TGR) enzymes in Schistosoma mansoni and some other platyhelminths have been identified as alternative targets. The present study was designed to confirm the existense and the potential value of TGR as a target for development of novel antischistosomal agents in Schistosoma japonicum, a platyhelminth endemic in Asia. Methods and Findings After cloning the S. japonicum TGR (SjTGR) gene, the recombinant SjTGR selenoprotein was purified and characterized in enzymatic assays as a multifunctional enzyme with thioredoxin reductase (TrxR), glutathione reductase (GR) and glutaredoxin (Grx) activities. Immunological and bioinformatic analyses confirmed that instead of having separate TrxR and GR proteins in mammalian, S. japonicum only encodes TGR, which performs the functions of both enzymes and plays a critical role in maintaining the redox balance in this parasite. These results were in good agreement with previous findings in Schistosoma mansoni and some other platyhelminths. Auranofin, a known inhibitor against TGR, caused fatal toxicity in S. japonicum adult worms in vitro and reduced worm and egg burdens in S. japonicum infected mice. Conclusions Collectively, our study confirms that a multifunctional enzyme SjTGR selenoprotein, instead of separate TrxR and GR enzymes, exists in S. japonicum. Furthermore, TGR may be a potential target for development of novel agents against schistosomes. This assumption is strengthened by our demonstration that the SjTGR is an essential enzyme for maintaining the thiol-disulfide redox homeostasis of S. japonicum.


Introduction
Schistosomiasis is a major tropical disease endemic in Africa, South America and Asia. It is estimated that approximately 600 million people are infected with the causative parasites in 76 countries around the world, and the disease results in about 20 million deaths annually [1][2][3]. In China, this disease caused by Schistosoma Japonicum remains a major public health concern, with more than 30 million people infected [4]. Currently, the control of schistosomiasis mainly depends on drug-based chemotherapy [5]. Since praziquantel was developed in the 1980s, it has replaced other drugs and became the first treatment of choice for schistosomiasis [6,7]. However, praziquantel resistance or tolerance may arise after long-term, repeated large-scale chemotherapy is implemented. Decreased sensitivity of Schistosoma mansoni to praziquantel has been found in some endemic areas [8][9][10][11][12], and several case reports of failure of repeated standard praziquantel treatment to clear Schistosoma haematobium infections have also been reported [13,14]. Although reduced sensitivity of S. japonicum to praziquantel has not been proven to date, the efficacy of the drug was reported to vary in different strains of this species [15]. Therefore, developing new antischistosomal agents to curb the emergence of drug-resistant schisotomes should be given a high priority.
Organisms are continuously attacked by endogenous and exogenous reactive oxygen species (ROS). Affluent antioxidant defense systems maintain the balance of pro-oxidants and antioxidants in a stable redox state and protect the homeostasis of the body. Two major systems exist to detoxify ROS in eukaryotes, the thioredoxin (Trx) system and the glutathione (GSH) system [16]. The Trx system is composed of nicotinamide-adenine dinucleotide phosphate (NADPH), Trx reductase (TrxR) and Trx. The GSH system consists of NADPH, GSH reductase (GR) and GSH. GR drives a tripeptide GSH-dependent system. Meanwhile, TrxR is an NADPH-dependent selenoprotein containing the thiol-disulfide redox active center [17], a GCUG active site motif located at the carboxyl terminus where ''U'' represents selenocysteine. As an important element in the active center of selenoproteins [18,19], selenocysteine is exposed on the highly extended C-terminal surface of TrxR. The Cys and SeCys residues in the reduced state have strong pro-nuclear properties. Many electrophilic compounds can selectively and irreversibly combine with the active center [20,21]. Both the Trx and GSH redox systems play an important role in the anti-oxidative network, regulation of transcriptional factors, cell growth and differentiation, synthesis of DNA, hormone action, cytokine function and metabolism of selenoproteins. At present, TrxR is considered a potential target for the development of novel drugs against tumors and infectious pathogens [20,21].
In contrast to their mammalian hosts, thiol-disulfide redox homeostasis in platyhelminths such as S. mansoni [22], larval Taenia crassiceps (cysticerci) [23] Echinococcus granulosus [24] and Fasciola hepatica [25], fully depend on thioredoxin glutathione redutase (TGR) of linked thioredoxin-glutathione systems. It has been demonstrated that there are no separate TrxR and GR enzymes in S. mansoni and other platyhelminths. The functions are replaced by the unique multi-functional enzyme_TGR. S. mansoni TGR (SmTGR) with the activities of TrxR, GR and glutaredoxin (Grx) has been confirmed by RNA interference as an essential enzyme for S. mansoni survival [26]. Like all eukaryotic TrxR isoforms, TGR is a selenoprotein with a GCUG active motif at the carboxyl terminus [27]. Selenocysteine (U) is recoded by a dedicated UGA codon in the selenoprotein mRNA by translation machinery utilizing a specialized structural element, the selenocysteine insertion sequence (SECIS). Following the silencing of TGR expression or inhibition of its activity in S. mansoni, the parasites die of oxidative damage from the host, suggesting that SmTGR would be a promising and potential drug target [28].
In this study the components involved in maintaining redox balance in S. japonicum were investigated through a bioinformatic approach. Several expressed sequence tags (ESTs) homologous to the SmTGR gene, but no separate TrxR or GR gene, were found implying that the TGR gene may also exist in S. japonicum and plays an important role in redox balance. However, no information on the S. japonicum TGR (SjTGR) enzymatic system is currently available. Here we report the gene cloning, protein expression and purification, as well as enzyme characterization of SjTGR. The effect of auranofin, a known inhibitor of TGR [23,24,26,29], against SjTGR was also assessed. The experimental findings further support TGR as a potential target for development of novel drugs against schistosomiasis.

Ethical statement
The animal work was approved by the Institutional Review Board (IRB00004221) of Jiangsu Institute of Parasitic Diseases, Wuxi, China (Permit Number: JIPDAERP20100724). All animals used in this work were conventionally housed in facilities and were provided food and water ad libitum. Animal studies including infection, intraperitoneal injection, orbital venous bleeding of mice were conducted in accordance with the recommendations in the Guidelines for the Care  Preparation of mRNA and first strand cDNA of S. japonicum adult worms Each mouse was infected with 50 S. japonicum cercariae by abdominal skin penetration. All mice were sacrificed on day 45 post-infection, and S. japonicum adult worms were collected through portal vein perfusion [30]. The mRNA from 0.5 g of S. japonicum adult worms was prepared by using the Illustra QuickPrep TM mRNA purification kit following the manufacturer's procedures. The first strand cDNAs were synthesized using mRNA of S. japonicum adult worms using the SuperScript TM III First Strand Synthesis System in accordance with the manual's instructions.

Cloning of S. japonicum TGR gene
The ESTs of TGR in S. japonicum were BLAST searched with the S. mansoni TGR DNA sequence. The potential open reading frame of SjTGR was deduced by comparing all SjTGR ESTs to the DNA sequence of SmTGR. A pair of gene specific primers were designed according to the DNA sequence of the deduced open reading frame of SjTGR, and the primer sequences were as follows: forward primer: 59-AACATATGCCTCCGATTGATG-GAACA-39 and reverse primer 1: 59-TC AGCAACCGGT-TACCGCTGCAGAGCCCCCA-39.
The first strand cDNA of S. japonicum adult worms was used as the template to amplify the open reading frame DNA fragment encoding SjTGR using a high fidelity DNA polymerase. The specific DNA fragment was cloned into the pGEM-T vector after attaching a deoxyadenine at the 39 terminal by Taq DNA polymerase and confirmed by sequencing at Invitrogen Corporation.

Construction of the recombinant SjTGR expression plasmid
The chimeric gene composed of the SjTGR open reading frame plus a typical bacterial SECIS element was prepared by overlapping PCR using the following primer pair: forward primer: 59-AACATATGCCTCCGATTGATGGAACA-39 and reverse primer 2: 59-GTCGACGGGCGCATAGGTTAACGATTGGT-GCAGACCTGCAACCGATTATTAACCTCAGCAACCGGT-TACCGC-39. The reverse primer 2 included the bacterial SECIS: 59-GGTTAATCGGTTGCAGGTCTGCACCAATCGTTAAC-CTATGCGCCC-39 [31]. The 59 Nde I site and 39 Sal I site were introduced into the chimeric gene by PCR for cloning into the plasmid pET41a containing the corresponding restriction sites. The insertion in the resulting recombinant expression plasmid SjTGR-pET41a was confirmed by DNA sequencing at Invitrogen Corporation.

Expression and purification of recombinant SjTGR
The recombinant SjTGR-pET41a vector and pSUABC vector, which expresses selA, selB and selC to assist selenocysteine incorporation into the selenoprotein, were co-transformed into Escherichia coli BL21 (DE3) competent cells as described by Kuntz and colleagues [26]. The bacteria grown on Luria-Bertani (LB) medium plates containing ampicillin (100 mg/mL) and chloramphenicol (30 mg/mL) antibiotics were selected as positive clones. Single colonies containing the SjTGR-pET41a and pSUABC plasmids were incubated in LB medium containing L-cysteine and seleniate. When the absorbance value of the bacteria culture at 600 nm reached 2.4, IPTG was added at a final concentration of 0.5 mM. The bacteria were then grown at 24uC for 24 h to induce SjTGR expression. The expression products were analyzed on a 12% SDS-PAGE gel stained with Commassie Blue.
To confirm that the recombinant SjTGR was a selenoprotein, the bacteria containing the SjTGR-pET41a and pSUABC plasmids or those containing only the SjTGR-pET41a plasmid were inoculated into LB culture medium or induced in LB medium containing isotope 75 Se-cysteine (SeCys). SeCys incorporation was analyzed on SDS-PAGE gels and autoradiograms.
The recombinant SjTGR was purified by affinity chromatography with an adenosine -29,59-diphosphate agarose (ADPagarose, Sigma) [32][33][34]. Briefly, the bacteria cultures were collected by centrifugation, and the pellets were frozen and thawed repeatedly until they became transparent. Deoxyribonuclease was added to the bacteria lysates to digest the DNA. The lysate solutions were then centrifuged at 30,0006 g for 30 min at 4uC, and the supernatants were passed through 0.45 mm filters. Samples in all steps above were kept on ice.
ADP-agarose column was equilibrated with buffer containing 50 mM potassium phosphate and 3 mM EDTA, pH 7.5. Supernatants containing the expressed products above were diluted ten-fold with the equilibration buffer, applied to the column, which was rinsed with wash buffer (0.4 M potassium phosphate, 3 mM EDTA, pH 7.5) until the ultraviolet absorption of the effluent did not decline further. The recombinant SjTGR proteins were then eluted with buffer containing 50 mM potassium phosphate and 0.5 M KCl, pH 7.5, and the fractions were analyzed by SDS-PAGE. The fractions containing SjTGR were pooled and applied to another ADP-agarose column, and the recombinant SjTGR proteins were eluted specifically with an NADPH linear gradient (0-0.5 mM) in 50 mM phosphate (pH 7.5). The fractions containing purified SjTGR proteins were pooled again and stored at 270uC after being dialyzed and concentrated by ultrafiltration (Amicon Ultra, Millipore, Bedford, MA, USA).

Production of antiserum against recombinant SjTGR and immunoblotting
The purified recombinant SjTGR proteins were mixed with Freund's complete adjuvant (Sigma) to form water in oil emulsions. Each mouse was immunized intradermally in the cervical and back skin with 100 mL of the adjuvanted antigen (containing 50 mg recombinant SjTGR) and boosted two times with the same amount of Freund's incomplete adjuvant antigen at two week intervals. Blood samples were collected via the orbital vein on day 7 after the second boost. Antibody titers of sera against recombinant SjTGR were measured by conventional ELISA.
Purified recombinant SjTGR proteins and the S. japonicum adult worms homogenates were separated on SDS-PAGE gel and transferred to nitrocellulose (NC) membrane. After blocking with 5% milk powder solution for 1 h at room temperature, the NC membrane was incubated with the SjTGR antiserum diluted 1:100 in PBST (PBS with 0.05% Tween-20) for 2 h at room temperature. The membrane was washed 3 times with PBST, for 10 min each time, and then incubated with HRP-conjugated goat anti-mice IgG (1:3000 dilution, Bethyl) at room temperature for 2 h. After washing 3 times as above, the colored signals on the membrane were developed with DAB substrate (6 mg 3,39-Dianinobenzidine dissolved in 10 ml PBS, 20 ml H 2 O 2 ).

SjTGR enzymatic assays
TrxR, GR and Grx activities of recombinant TGR were analyzed based on the methods described by Kuntz et al. [26] with a UV-visible spectrophotometer (UV-2550, SHIMADZU) at 25uC. The TrxR activity was determined by the 5, 59-diothiobis (2-nitrobenzoic acid) (DTNB) reduction assay and the insulin reduction assay. The DTNB reduction assay solution contained 100 mM NADPH, 3 mM DTNB, 10 mM EDTA in 0.1 M potassium phosphate (pH 7.4). The increase of absorbance at 412 nm was monitored during the first 2 min. The extinction coefficient of 13.6 mM 21 cm 21 for TNB was used in the calculation. One enzyme unit is defined as the NADPH-dependent production of 2 mmol of 2-nitro-5-thiobenzoic acid per min at 25uC.
Trx reduction assay was determined by monitoring the decrease of NADPH consumption at 340 nm during the first 2 min. The assay was carried out in the presence of 1 mg/ml insulin, 10 mM SjTrx and 100 mM NADPH in 0.1 M potassium phosphate (pH 7.4) containing 10 mM EDTA. The extinction coefficient of 6.22 mM 21 cm 21 for NADPH was used in the calculation. One unit of TrxR activity is defined as the oxidation of 1 mmol of NADPH per min at 25uC.
The GR activity was determined with GSSG. The reaction mixture contained 100 mM NADPH, 100 mM GSSG in 0.1 M potassium phosphate (pH 7.4) containing 10 mM EDTA. Decrease of NADPH consumption at 340 nm during the first 2 min was monitored. The extinction coefficient of 6.22 mM 21 cm 21 for NADPH was used in the calculation. One unit GR activity is defined as the oxidation of 1 mmol of NADPH per min at 25uC.
The Grx activity was determined by monitoring decrease of NADPH consumption at 340 nm during the first 2 min, and the assay was performed in the presence of 1 mM GSH, 8 mM HED, 0.6 units of yeast GR, 100 mM NADPH in 0.1 M potassium phosphate (pH 7.4) containing 10 mM EDTA. The extinction coeffcient of 6.22 mM 21 cm 21 for NADPH was used in the calculation. One unit of Grx activity is defined as the oxidation of 1 mmol of NADPH per min at 25uC. All assays were performed in triplicate.

Kinetic analysis
Based on the method described by Kuntz et al. [26], SjTGR activities with NADPH, DTNB, SjTrx, GSSG, HED and GSH were determined in the following buffer: 100 mM NADPH, 10 mM EDTA, 0.1 M potassium phosphate (pH 7.4) and 48 nM SjTGR. In the TrxR activity assay, DTNB concentrations varied from 100 to 1000 mM, NADPH concentrations from 5 to 100 mM and Trx concentrations from 9 to 34 mM. In the GR activity, GSSG concentrations were from 10 to 100 mM, and NADPH was from 5 to 100 mM. In the Grx assay, 2 to 10 mM HED, 100 to 1500 mM GSH and 10 to 100 mM NADPH were used. In each reaction, the concentration of one substrate was changed while keeping the others constant. Kinetic parameters were obtained by curve fitting of experimental data using OriginPro 7.0. All assays were performed in triplicate.

Auranofin inhibition of recombinant SjTGR
Inhibition of TrxR, GR and Grx activities of recombinant SjTGR by different concentrations of auranofin was determined in the enzyme assays described above. The 50% inhibitory concentration (IC 50 ) values of auranofin in the various assays were calculated by curve fitting using the SPSS 13.0 statistical software.
The inhibitory constants (K i ) of auranofin for TrxR activity and GR activity of SjTGR were assessed using varying substrate and inhibitor concentrations as follows: DTNB 100, 150, 200, 300, 500, 700, 1000 mM; GSSG 10, 15, 20, 30, 50, 70, 100 mM; auranofin 5, 10, 20, 30 nM. The K i values were calculated using the following equations [35]: Formula (1) is used for calculating K i of competitive inhibition. Where K maxapp is the apparent K max at different inhibitor concentrations, and I ½ is the inhibitor concentration. Formula (2) is used for calculating K i of non-competitive inhibition. Where u maxapp is the apparent u max at different inhibitor concentrations, and I ½ is the inhibitor concentration. All assays were performed in triplicate.

Toxicity of auranofin on S. japonicum adult worms in vitro
Stock solutions of auranofin and praziquantel were prepared by dissolving 1 mg of the drugs in 0.4 ml dimethyl sulfoxide (DMSO) and adding 0.6 ml RPMI 1640 medium. S. japonicum worms obtained from mice (C57BL/6, male, 22,24 g, each infected with 50 cercariae) were washed in RPMI 1640 medium, kept at pH 7.5 with HEPES 20 mM and supplemented with penicillin (100 UI?mL 21 ), streptomycin (100 mg?mL 21 ) and 10% bovine fetal serum (Gibco) [36]. After washing, two pairs of adult worms were transferred to each well of a 24-well culture plate containing 2 mL of the same medium. The worms were cultured for 30 to 60 min at 37uC in a humid atmosphere containing 5% CO 2 , and then different concentrations of auranofin (1, 5, 10, 20 and 30 mg/ml) diluted with RPMI 1640 medium were added. Control worms were treated with equal volumes of RPMI 1640 or DMSO, and worms treated with 30 mg/mL praziquantel were also observed. The worm mobility, tegumental alterations and parasite survival were monitored under an inverted microscope (Leica, Germany) at 1, 6, 24, 48 and 72 h. Parasite death was defined as having no motor activity during 2 min of continuous observation as well as morphological and tegumental alterations [37]. Worms treated with auranofin were homogenized in 0.1 M potassium phosphate (pH 7.4). The homogenates were centrifuged at 20,0006g for 1 h at 4uC, and the supernatants were assayed for TrxR and GR activities as described above. All assays were repeated three times.

Auranofin treatments in vivo
A stock solution of auranofin was prepared by dissolving 2 mg of the drug in 0.4 mL DMSO and adding 0.6 mL of aseptic 0.9% sodium chloride. The stock solution was diluted to 0.4 mg/mL with the aseptic 0.9% sodium chloride before use. All mice (C57BL/6, male, 22,24 g) infected with 30 S. japonicum cercariae by abdomenal skin exposure were administered 6 mg/kg auranofin injected intraperitoneally twice daily for 5 days beginning 35 days post-infection. Control mice were given the same dose of carrier on the same timetable. Mice were sacrificed on day 42 post-infection. Worms were counted by portal vein perfusion. Liver tissues were digested with 5% potassium hydroxyl solution, and the eggs from the liver tissue were counted. The worm reduction rate and egg reduction rate were calculated using the following formulas.
Worm reduction rate (%)~(mean number of adult worms per mouse in control group -mean number of adult worms per mouse in experimental group)=mean number of adult worms in control group.
Egg reduction rate (%)~(mean number of eggs per mouse in control group -mean number of eggs per mouse in experimental group)=mean number of eggs per mouse in control group.

Statistical analysis
The numbers of worms and eggs per gram of liver were expressed as mean values 6 standard errors (SD). Significant reductions in worm and egg burdens after auranofin treatment were determined by two-tailed Student t-test using statistical software SPSS 13.0. Differences between median values were considered to be significant at the level of 5%.

Cloning and characterization of the SjTGR gene
The cDNA fragment encoding the open reading frame of the SjTGR protein was amplified from the mRNA of S. japonicum adult worms using a pair of gene specific primers designed according to the S. japonicum EST sequence, which showed a high degree of homology to the SmTGR gene. The cDNA fragment was composed of 1791 nucleotides encoding 598 amino acids. The predicted molecular weight of SjTGR was 64.9 kDa with an isoelectric point of 6.38. The amino acid sequence of SjTGR was 91% identical to SmTGR and 55% similar to human TrxR. Like SmTGR, the SjTGR protein contained NADPH-and FADbinding domains, a dimer interface domain of pyridine nucleotide disulfide oxidoreductases and the thiol-dislifide redox active center (-CVNVGC-). A glutaradoxin active site (-CPFC-) was found in its N-terminus, and its C-terminus contained a conserved tetrapeptide sequence GCUG (U: selenocysteine), which is the catalytic domain of TGR [27].

Expression and selenium incorporation of recombinant SjTGR
As the synthesis of selenoprotein depends on a highly speciesspecific translation mechanism, use of the bacterial-type SECIS element was necessary in order to express the SjTGR selenoprotein in E. coli host cells. The open reading frame DNA sequence of SjTGR was fused with the SECIS element by PCR to form a chimeric gene following the procedure described by Arnér and colleagues [31]. The chimeric SjTGR gene was cloned directionally into the expression plasmid pET41a at the restriction enzyme sites Nde I and Sal I, resulting in the recombinant plasmid SjTGR-pET41a. Restriction analysis and DNA sequencing confirmed that the chimeric SjTGR gene had been inserted into the pET41a expression plasmid correctly (Fig. 1).
The selenocysteine insertion during selenoprotein translation in E. coli requires an E. coli-type SECIS element just following the UGA codon in the selenoprotein mRNA, as well as the products of the selA, selB, and selC and selD genes [31]. In this study, the recombinant SjTGR-pET41a plasmid was co-transformed into E. coli BL21 with the pSUABC vector expressing the selA, selB and selC genes (encoding selenocysteine synthase, SELB and tRNA Sec , respectively) to aid in the production of SjTGR as a putative selenoprotein. When the cells were induced with 1 mM IPTG for 4 h at 37uC, the SjTGR proteins were expressed in inclusion bodies, which appeared in the precipitate of the bacterial lysate obtained by ultrasonication. Little soluble recombinant SjTGR was found in the supernatant of the bacterial lysate ( Fig. 2A). However, when the growth temperature was reduced to 24uC and the bacteria were induced with IPTG for 24 h beginning at the stationary growth stage (OD 600 nm of culture reaching 2.4) [38], soluble SjTGR proteins were produced successfully (Fig. 2B). The selenocysteine incorporation analysis using 75 Se-labeling cysteine and autoradiograms showed that the products of selA, selB, and selC could increase the yield of SjTGR, suggesting that it is a selenoprotein (Fig. 3).
The recombinant SjTGR protein was purified with an adenosine 29, 59-diphosphate agarose (Sigma A3515) column, used for specific affinity purification of NADPH-binding proteins. The purity of the SjTGR protein sample was more than 95% ( Fig. 4), indicating the NADPH binding capacity of the recombinant SjTGR.

Western blotting with antiserum against SjTGR
High titer (.1:204,800) antiserum against recombinant SjTGR was obtained after immunizing mice with the purified recombinant SjTGR. The antiserum could recognize a single protein band of expected size (MW ,65 kDa) in the supernatant of S. japonicum adult worms homogenates by Western blotting (Fig. 5). The size of the recombinant SjTGR band was similar to that of the natural SjTGR. There were no other protein bands recognized by the antiserum. These results suggested that the recombinant SjTGR proteins were immunoreactive, inducing antibodies specific to the native SjTGR protein, and that no cross-reactive proteins were present in the S. japonicum adult worm.

Activity and kinetics of recombinant SjTGR
The enzyme activities of purified SjTGR were determined using a broad range of substrates, and the results showed that SjTGR possessed the characteristic activities of mammalian TrxR, GR and Grx. The specific activities were 8.1160.34 mmol?min 21 mg 21 with DTNB, 3.3360.08 mmol?min 21 mg 21 with Trx, 2.1960.01 mmol?min 21 mg 21 with GSSG, and 12.1060.39 mmol?min 21 mg 21 in a Grx assay coupled to HED. The TrxR activity (with substrates DTNB and Trx) and the Grx activity of SjTGR were similar to those of SmTGR, although the GR activity of SjTGR was lower than that of SmTGR (Table 1).      Table 2), which were similar to those of SmTGR (Table 3). The catalytic efficiencies of SjTGR on TrxR, GR and Grx were different and found to be in the following order: TrxR (10 5 -10 6 M 21 ?s 21 ).GR (10 5 M 21 ?s 21 ).Grx (10 3 M 21 ?s 21 ).
Although the K cat values of recombinant SjTGR with various substrates were lower than those of SmTGR, the catalytic efficiencies (K cat =K m ) of both enzymes with the same substrate were similar ( Table 3).

Inhibition of recombinant SjTGR by auranofin
Inhibition of recombinant SjTGR by auranofin was observed by adding different amounts of the inhibitor into the enzyme assay system using DTNB, GSSG or HED as substrate. The experimental results showed that auranofin could inhibit the TrxR, GR and Grx activities of recombinant SjTGR, and the inhibitory effects were concentration dependent (Fig. 6A). The IC 50 values were 6.89, 0.47 and 8.12 nM, respectively, for the DTNB, GSSG and HED substrates with 5 nM recombinant SjTGR. The inhibitory effect of auranofin on TrxR activity was competitive, increasing the K m for substrates while the V max remained constant. The inhibitory effect of auranofin on GR activity was non-competitive, altering the V max while the K m for substrates remained constant. The inhibitory constants (K i ) were 0.762 nm for TrxR activity and 0.034 nm for GR activity according to apparent K m and V max values with various concentrations of auranofin (Fig. 6B, 6C).

Toxicity of auranofin in cultured S. japonicum adult worms
The effects of auranofin on cultured worm survival were observed by culturing the worms in medium containing different concentrations of auranofin, and the mobility and death rates of the parasites were monitored. Auranofin at 5 mg/ml (7.40 mM) resulted in a 53.3% decrease of worm viability after 6 h and 66.7% after 12 h. By 24 h, 100% mortality was observed. The median lethal dose (LD 50 ) of auranofin was 2.56 mg/ml (3.78 mM) at 24 h, with 95% confidence interval 2.09,3.02 mg/ml. Auranofin at 5 mg/ml (7.40 mM) reduced 57.1% of the TrxR activity and 71% of the GR activity in the S. japonicum adult worms homogenates compared to that of carrier treated-control worms, and the inhibitory effects were dose-dependent. No TrxR and GR activity was observed in the homogenates when the worms were treated with auranofin at 30 mg/ml (44.22 mM) (Fig. 7A). Additionally, the inhibitory effects of auranofin on TrxR and GR activities increased with time (Fig. 7B, 7C).

In vivo activity of auranofin
ICR mice infected with 30 cercariae were treated with 6 mg/kg auranofin twice daily for 5 days beginning on day 35 postinfection. The results showed that auranofin could reduce worm burden by 41.50% and egg burden by 43.18% at the adult worm stage (Table 4).

Discussion
The antioxidant system plays a critical role in physiological functions of an organism and protects schistosomes from host immune attack [16,25]. Instead of having separate TrxR and GR enzymes, the single TGR protein in platyhelminths carrying these functions can completely deoxidize both Trx and GSSG [22]. TGR proteins of S. mansoni [22], larval T. crassiceps (cysticerci) [23], E. granulosus [24] and F. hepatica [25] have all been determined as potentially important drug targets. S. mansoni adult worms die when the activity of TGR is inhibited or silenced, as TGR produces partial or complete protective effects against oxidative damage in the host [26]. However, little attention has been focused on targeting the redox system of S. japonicum, a helminth endemic in Asia, for development of new schistosomicides.
In this study, the TGR gene of S. japonicum was cloned, and the deduced amino acid sequence of SjTGR revealed that it included NADPH-and FAD-binding domains, a thiol-disulfide redox active center (-CVNVGC-) and a Grx active site (-CPFC-). It was found   to be 91% identical to SmTGR and 55% similar to human TrxR. There were 32 ESTs homologous to SmTGR found by analyzing ESTs of S. japonicum. One longer EST contained a conservative SECIS element in the non-translated region following the  termination codon (TGA) of the TGR open reading frame. The SECIS element mediates the selenocysteine insertion at the penultimate amino acid of TGR functioning in concert with an adjacent cysteine as an essential part of this redox center [39], implying that SjTGR is also a selenoprotein enzyme like the mammalian TrxR isoforms. As selenoprotein synthesis is a highly species-specific process, we fused the SjTGR cDNA with the SECIS sequence of E. coli in constructing the recombinant SjTGR-pET41a plasmid in order to express the SjTGR selenoprotein in E.coli host cells [31]. Co-transformation of this plasmid into E.coli BL21 along with the pSUABC plasmid expressing selA, selB, and selC that provided proteins to aid in the production of selenoproteins successfully yielded a soluble 65 kDa SjTGR protein upon IPTG induction. The selenocysteine incorporation analysis using 75 Se-labeling cysteine and autoradiograms showed that the products of selA, selB, and selC could increase the yield of SjTGR proteins. These results also confirmed that SjTGR is a selenoprotein. Using the same methods previously published for production of recombinant SmTGR and conditions for enzyme kinetics analysis [26], the experimental data in the present study showed that SjTGR is a multifunctional enzyme with TrxR, GR and Grx activities. The GR activity of SjTGR was lower than that in SmTGR (2.1960.01 mmol?min 21 mg 21 vs. 7.2 mmol?min 21 mg 21 ). However, the lower activity was compensated by the lower K m value (  21 ). These findings support the hypothesis that the catalytic efficiency of TGR with Trx was higher than that with GSSG and HED. The Trx/GSSG ratios of 3.3 and 15.7 in T. crassiceps [23] and S. mansoni [26] have also been reported. In all cases, TGR showed a preference toward Trx as substrate.
Compared to the TrxR K m (21.4260.44 mM) and GR K m (83.24624.27 mM) of SjTGR, the Grx K m (10.9762.42 mM) with NADPH was the lowest, which suggests that NADPH, combined with SjTGR, tends to deliver electrons to the Grx activity domain first and then to Trx. In S. mansoni, the electrons flow within the TGR domain from NADPH to the C terminus, then from the C terminus to either Trx or the Grx domain [40]. Electrons are selectively delivered to the acceptor by the C-terminal arm shuttling between SmTrx and the Grx domain [40]. Further study on the direction of electron transfer of SjTGR will be surveyed in our future work.
Compared to the typical human TrxR (2.2610 6 M 21 ?s 21 ) [41] and GR (3.2610 6 M 21 ?s 21 ) [42], it is obvious that the catalytic efficiencies of TrxR from helminths fall into the same range, but their GR catalytic efficiencies were all lower than that of human TrxR (Table 3). One possible reason for this phenomenon is that schistosomes may use the reduced GSH derived from host erythrocytes they ingest to maintain their redox balance, decreasing the requirement of GSSG reduction accordingly. Our results also provide evidence that schistosome TGR is not an efficient GR [40].
High titer (.1:204,800) antiserum was obtained after immunizing mice with the purified recombinant SjTGR. In Western blot analysis, the polyclonal antibodies in this antiserum could recognize a ,65 kDa protein band corresponding to the expected size of SjTGR in the supernatant of S. japonicum adult worms homogenates, indicating that the recombinant SjTGR was immunogenic and induced antibodies reactive to the natural protein. Furthermore, this antiserum did not react with other bands in the worm supernatant. Since the amino acid sequence of TGR is highly homologous to the 56 kDa TrxR and 52 kDa GR of eukaryotic cells, if individual TrxR or GR proteins exist in the S. japonicum adult worm, the corresponding cross-reactive protein bands would have been detected. These results indicate that the S. japonicum adult worm does not express separate TrxR or GR molecules, the functions of which are entirely replaced by TGR. This point was also confirmed by bioinformatic analysis which found no TrxR or GR EST sequences in the S. japonicum EST database (data not published). These data reported here suggest that SjTGR plays a key role in the maintenance of redox balance in S. japonicum.
In the study, the gold compound auranofin, a known inhibitor of TGR, was found to inhibit the activity of the recombinant SjTGR. The IC 50 values of auranofin against TrxR, GR and Grx activities were in the nanomolar range (6.89, 0.47, 8.12 nM) with 5 nM recombinant SjTGR. Of note, the IC 50 value of auranonfin on GR in S. mansoni worms was almost 9 times higher than that in S. japonicum, while the IC 50 values against TrxR and Grx were nearly the same. Interestingly, the inhibitory mechanisms of auranofin against TrxR and GR were distinct. Auranofinmediated inhibition of TrxR activity was competitive, while that of GR activity was non-competitive. This may be due to differences in the structures of auranofin, DTNB and GSSG. Auranofin has a benzene ring structure [43], which is similar to DTNB with two benzene rings connected by a disulfide bond [44], but different from the carbon chain of GSSG [45]. Auranofin and DTNB may combine with the active center of TGR competitively, but auranofin may combine outside of the active center of GR to influence the GR activity of SjTGR in a non-competitive manner. The K i value of auranofin on TrxR and GR was very low (0.762 nM and 0.034 nM), indicating that just a small amount of auranofin could completely inhibit SjTGR activity, similar to the findings in larval T. crassiceps [23], E. granulosus [24] and S. mansoni [43].
Auranofin was confirmed to be toxic to S. japonicum adult worms in vitro, and the dose of 5 mg/ml (7.40 mM) resulted in 100% mortality after 24 h. The calculated LD 50 of auranofin against S. japonicum adult worms in vitro was 2.56 mg/ml (3.78 mM) at 24 h. These results are in agreement with those reported from S. mansoni Table 4. Treatment effect of auranofin on S. japonicum in mice. adult worms [26], E. granulosus larval worms [46] and T. crassiceps metacestode [47]. In vivo, 6 mg/kg auranfin could partially cure S. japonicum-infected mice, with a 41.50% reduction in worm burden and a 43.18% reduction in egg burden. We obtained similar results to those previously observed in S. mansoni treated with auranofin [26]. As SjTGR is highly similar to SmTGR in terms of amino acid sequence and activities, it may be a potentially important drug target as well. However, there are some differences between SmTGR and SjTGR. The Grx activity motif of SjTGR is composed of CPFC compared to CPYC in SmTGR [27]. The differences in amino acid sequence in the TrxR domain of TGR between S. mansoni and S. japonicum are more marked [39]. The potential differences in the molecular structures between SjTGR and SmTGR may result in different inhibitory effects of drugs targeting these proteins. Of course, further protein homology modeling and crystallographic structural studies will be required to better understand TGR as a drug target. Such studies can benefit in the development of new chemotherapeutic drugs targeting multiple schistosome species. Meanwhile, the unique aspects of the SjTGR structure may help to improve potency of these new drugs against S. japonicum.
In conclusion, our study confirms that a multifunctional enzyme SjTGR selenoprotein, instead of separate TrxR and GR enzymes, exists in S. japonicum, which is an extension of findings on TGR in other platyhelminths. SjTGR plays an essential role in maintaining the redox balance in S. japonicum, which further confirms TGR as a potential target for the development of new drugs against schistosomiasis.