Amino acid analysis and molecular weight (MW) and electronic point (pI) determination of protein were performed using BioXM 2.6. Multiple sequence alignment was performed using MUSCLE¹ and visualized online using ESPript 3.0² (Edgar, 2004; Robert and Gouet, 2014). Protein sequences were primarily selected from pathogens, yeast, and humans. The accession numbers for these sequences were 4QLZ and 4QMB for S. japonicum, EAX99718.1 for Trichomonas vaginalis G3, CCD82372.1 for S. mansoni, CCF72873 for Babesia microti strain RI, NP_218145.1 for Mycobacterium tuberculosis H37Rv, CAB37743.1 for Helicobacter pylori, 1IPW for Escherichia Coli, 1E9G for S. cerevisiae, and AAH01022.3 for Homo sapiens.

Inorganic pyrophosphate metabolism was analyzed by enzymatic reaction analysis, a literature search, and sequence alignment. Enzymatic reactions involving PPi were searched using the search term “diphosphate” in the Enzyme Structures Database.³ Further confirmation was done “on a one by one basis” to ascertain the participation of PPi in these reactions. Furthermore, based on PPi location at product or substrate positions in the reactions, these catalysts were divided into PPi-generating enzyme and PPi-utilizing enzyme (including PPi hydrolyase). Enzymes involved in PPi metabolism were searched⁴ and investigated in humans and S. japonicum. Furthermore, PPi transporter and member-integrated PPi metabolism-related enzymes were investigated using a literature survey based on the PPi homeostasis model (Villa-Bellosta et al., 2013). Thereafter, a comparison of the PPi metabolic pathway between blood flukes and human hosts was performed using a sequence alignment search of genes and proteins related to PPi metabolism. To search for progressive ankylosis protein homology (ANKH) in Schistosoma, the query sequences (Supplementary File 1) were downloaded from UniProtKB. The search term “ANKH family” was used in the “Family and Domains > Protein family” field Databases and included S. haematobium, S. japonicum, and S. mansoni genomes.

A forward primer (5′-AATGGGTCGCGGATCCATGTCGGT TGAACGTGGG-3′) and a reverse primer (5′-GGTGGTGGTGCTCGAGTTAAATATTCGTATTACAAAAAT GCC-3′) were used to amplify the SjPPase gene with cDNA as the template. The 25-μl reaction mixture comprised of 2× Taq PCR MasterMix (12.5 μl), 10 μM forward primer (1 μl), 10 μM reverse primer (1 μl), cDNA template (1 μl), and water (9.5 μl). The cycling conditions were set as follows: initial denaturation at 95°C for 5 min, 30 cycles of PCR including denaturation at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 1 min. The full-length gene of soluble PPase from S. japonicum was inserted by the In-fusion cloning method into plasmid pET-28a (+) between two restriction sites: BamH I and Xho I. The recombinant plasmid (SjPPase-pET-28a) was transformed into E. coli BL21 (DE3), and then the bacteria-harboring SjPPase-pET-28a was conserved in LB medium containing 15% glycerol and 50 mg kanamycin per liter at -80°C for protein expression. The product of small-scale expression was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting to verify the expression level of the target protein. The prepared bacteria was revived in 50 ml LB medium containing kanamycin at 37°C, and shaken at 165 rpm for 7 h. The suspension was then decanted into 1 L of LB culture, incubated at 37°C, and shaken at 165 rpm. When OD₆₀₀ approached 1.4, 0.7 mM isopropyl-β-d-thiogalactoside (IPTG) was added to the culture; shaking speed and temperature were altered and set at 195 rpm and 24°C, respectively. After 9 h, the suspension was harvested by centrifugation at 5,000 rpm and a temperature of 4°C for 8 min. The pellet was re-suspended in nickel-nitrilotriacetic acid (Ni-NTA) column buffer A [300 mM NaCl, 50 mM phosphate buffer (pH 7.4), 10 mM imidazole, and 10% glycerol] containing 5 mM phenylmethylsulfonyl fluoride (PMSF) and lysed by sonication. The lysate was clarified by centrifugation at 13,000 rpm and temperature of 4°C for 30 min, and the supernatant was filtered using a 0.22-μm filter membrane. The flowed fraction was loaded into an affinity chromatography column of Ni-NTA resin at a 3-ml/min flow rate using an ÄKTA purifier 10 (GE Healthcare, Chicago, IL, United States). The resin was washed with Ni-NTA column buffer A until the value of UV approached 0 mU, washed again with a mixture of Ni-NTA column buffer B [300 mM NaCl, 50 mM phosphate buffer (pH 7.4), 500 mM imidazole, and 10% glycerol] and Ni-NTA column buffer A at a volumetric proportion of 1:9 until the value of UV approached 20 mU, then eluted with 20 ml buffer B. The eluted protein was dialyzed overnight against diethyl aminoethyl (DEAE) column buffer A [20 mM Tris-HCl (pH 8.0), 50 mM NaCl, 5 mM EDTA, and 10% glycerol]. The protein solution was loaded into DEAE resin at a 3-ml/min flow rate using ÄKTA purifier 10. The resin was washed with DEAE column buffer A until the value of UV approached 0 mU, then eluted with 20 ml DEAE column buffer B [20 mM Tris-HCl (pH 8.0), 500 mM NaCl, 5 mM EDTA, and 10% glycerol]. Each fraction was performed by SDS-PAGE for purification analysis and molecular mass analysis. The eluted protein was dialyzed overnight against a dialysis buffer [50 mM Tris-HCl (pH 7.4), 100 mM NaCl, and 10% glycerol]. The purification products were analyzed using SDS-PAGE, and the purified protein was concentrated to approximately 18 mg/ml using a microcon-10 kDa centrifugal filter (MilliporeSigma, Burlington, MA, United States). The S. cerevisiae PPase gene (ScPPase, 1E9G) was synthesized and further constructed in pET-28a (+). The ScPPase recombinant protein was prepared using the method for rSjPPase.

We prepared eight samples containing 100 μg rSjPPase in 100 μl PBS. First, these were equally separated into two groups in which 5 μl of 1 mM Mg²⁺ was added to one group while the other group was without Mg²⁺. Next, four samples from each group were treated with 5 μl of 1 mM glutathione (GSH), 5,5’-dithiobis(2-nitrobenzoic acid) (DNTB), imidodiphosphate (IDP), and 10 μl PBS for 5 min at room temperature. Thereafter, all samples were centrifuged at 3,000 rpm (4°C) for 2 min, and native-PAGE was performed for 3 h at 240 V in an ice bath.

Inorganic pyrophosphatase activity assay was based on a previously described method for HSP90 ATPase activity (Avila et al., 2006). The effects of some molecules, such as magnesium ions (Mg²⁺), PPi, sodium fluoride (NaF), and methylene diphosphonic acid (MDP), on the catalytic activities of SjPPase, were investigated. The detailed conditions are described in Supplementary Table 1. Each group comprised two reactions, test and control reactions, each of which was made up of a 100-μl reaction system. The SjPPase or ScPPase enzyme and magnesium ions, as well as inorganic pyrophosphate and NaF or MDP, were mixed to a 50-μl volume—VC signifies “various concentrations.” VC_A represents a series of magnesium ions concentrations, including 500, 250, 100, 50, and 25 μM. Various PPi concentrations represented by VC_B included 2,000; 1,000; 500; 250; 100; 50; 10; and 5 μM. Various NaF concentrations represented by VC_C included 1 and 0.5 mM. Various concentrations of MDP represented by VC_D included 2 and 1 mM. The reaction was incubated for 45 min and monitored every 2.5 min. A fluorescence microplate reader was used for fluorescence detection at an excitation of 560 nm and emission detection at 590 nm. All reactions were processed at 25°C. Protein concentration was tested using a bicinchoninic acid assay (BCA assay) using bovine serum albumin as the standard. The curves were analyzed by nonlinear fitting using the GraphPad Prism software, version 5.0 (GraphPad, San Diego, CA, United States). Data for ScPPase and Mg²⁺ ions concentration-response were generated by nonlinear regression to Michaelis–Menten kinetics. Data for SjPPase PPi concentration-response were analyzed by nonlinear regression to substrate inhibition.

Crystallization screening was performed in 96-well plates with sitting drops consisting of 1 μl protein solution mixed with an equal amount of precipitant. The index^TM PEG/Ion Screen^TM and the PEG/Ion 2 Screen^TM kit were used for screening (Hampton Research, Aliso Viejo, CA, United States). Colorless rectangular rod-like crystals were obtained after 2 weeks from the condition 0.2 M sodium malonate, 20% polyethylene glycol (PEG) 3,350. Crystallization conditions were optimized by fine-tuning the pH value, additive type, and salt and precipitation agent concentration. The final conditions (Supplementary Table 2) suitable for X-ray diffraction were reproduced manually in large volumes by the hanging-drop method in 24-well plates, equilibrating 1 μl protein and 1 μl precipitant against 1 ml precipitant. The protein concentration for crystallization was 16–18 mg/ml.

The crystals were collected at a wavelength of 0.9707 Å using an ADSC Quantum 315r CCD detector on a BL17U1 beamline at Shanghai Synchrotron Radiation Facility (SSRF), China. Data were processed and reduced using the HKL2000 package (Otwinowski and Minor, 1997), and relevant statistics are shown in Table 1.

The structures of SjPPase-Pi and apo-PPase were solved by the molecular-replacement (MR) method using yeast PPase (PDB code: 1E9G) as an initial search model on Phenix (Adams et al., 2010). Thereafter, refinements were conducted using REFMAC, and the models were checked and rebuilt using COOT (Emsley and Cowtan, 2004). Statistical data for the final rounds of refinement are presented in Table 1. The electron density maps were calculated using Phenix, and the figures were generated using PyMOL software (PyMOL molecular graphics system, version 1.3.1; Schrödinger, Inc, New York, NY, United States). The atomic coordinates and structure factors (code 4QLZ and 4QMB) were deposited in the Protein Data Bank.⁵

A structural comparison of the SjPPase and ScPPase ligand and cofactor binding sites was performed. A series of ScPPase PDBs (1e6a, 1e9g, 1m38, 1wgi, 2ihp, 2ik0, 2ik1, 2ik2, 2ik4, 2ik6, 2ik7, 2ik9, and 117e) and SjPPase 4qlz were used for searching the binding sites of ligands and apo-PPase. Combined with the multi-sequencing alignment results, some identical or varying residues between SjPPase and ScPPase at binding sites were found.

Docking simulations were carried out using the C-DOCKER module (Discovery Studio, version 2.1; Accelrys, San Diego, CA, United States). The X-ray crystal structure of SjPPase was used for docking calculation. All water molecules in the crystal structure were removed, while Co²⁺ (M1 and M2) ions in the active site were kept as part of the protein. The CHARMm-force field was applied for docking. The region within 12 Å of Pi was chosen as the active binding site. Random conformations of PPi and MDP were minimized using CHARMm-based molecular dynamics (1,000 steps), which were then docked into the defined binding site. The other parameters were set as default. The CDOCKING ENERGY scoring function was used to rank binding poses.

The 864 bp CDS encoded a 287-aa putative protein with a theoretical molecular weight and isoelectric point (pI) estimated at 32.7 kDa and 6.05, respectively. The protein contained only one pyrophosphatase region. Multiple sequence alignment shows that SjPPase contained the characteristics of a family I PPase, as DXDXXD signature motif and other 12 conserved residues (Figure 1). In addition, the SjPPase protein sequence exhibited 52% identity with ScPPase.

According to enzymatic reaction analysis, intracellular PPi (iPPi) was mainly generated from reactions catalyzed by nucleotidyltransferases, together with alkyl and aryl transferases, while PPi is a by-product. Generated PPi was hydrolyzed by PPase (EC 3.6.1.1) or alkaline phosphatase (AP, EC 3.1.3.1) or utilized by pentosyltransferases and phosphotransferases with an alcohol group as an acceptor. Based on a literature survey (Villa-Bellosta et al., 2011; Foster et al., 2012), PPi was pumped out from cells by ANKH. Extracellular PPi (ePPi) was hydrolyzed by tissue-nonspecific alkaline phosphatase (TNAP) or ectonucleotide pyrophosphatase phosphodiesterase-3 (NPP3). In addition, a part of ePPi was produced from pyrophosphorolysis of nucleoside triphosphate (NTP) catalyzed by ectonucleotide pyrophosphatase phosphodiesterase-1 (NPP1). All of these formed the general picture of iPPi and ePPi metabolism (Figure 2). We then further searched for protein sequences related to PPi metabolism in Schistosoma and humans (Supplementary Table 3). We observed that PPi transporter ANKH exists in the human genome (accession number: NP_473368.1) but was absent in the Schistosoma genome. Based on the preceding observations, the comparison of iPPi and ePPi metabolism between humans and Schistosoma is shown in Figure 2.

The Western blotting results suggested that target proteins were expressed in the supernatant of lysates (data not shown). Protein was purified using a Ni-NTA column followed by the DEAE column. The results (Figure 3A) were displayed by SDS-PAGE using coomassie blue staining.

The interaction between PPase and micromolecules was based on the following: the divalent metal-ions as the co-iron of PPase, two substrate affinity sites on the surface of PPase, and cystine residues as the regulatory sites for enzyme activity. A 12% polyacrylamide gel was run for the designed mixtures to improve our understanding of the function of SjPPase micromolecules (Figure 3B). The major components of lanes 1, 2, 3, and 8 were at the same level of electrophoretic distance, while lanes 6 and 7 were at a higher level and lane 4 was at a lower level. The sample for lane 5 was precipitated after pre-treatment. The comparison between the nontreated sample (lane 1) and the sample treated with DNTB (lane 5) revealed that DNTB could destroy the structure of SjPPase. Having observed the differences between lanes 5 and 4 (samples treated with DNTB and Mg²⁺, respectively), this study shows that Mg²⁺ could stabilize the SjPPase structure. IDP can interact with apo-SjPPase (lanes 1 and 7) and the Mg²⁺-SjPPase complex (lanes 6 and 8). There were weaknesses in the interaction between GSH and SjPPase (lanes 1 and 3, or lanes 2 and 8).

The PPase activity was analyzed using a Pi Per^TM Phosphate assay Kit (Molecular Probes, Eugene, OR, United States). We investigated the influence of PPi on SjPPase and ScPPase activity (Figures 4A,B). In contrast to ScPPase (higher PPi concentration and higher activity), SjPPase presents a maximum activity at a concentration of 250 μM PPi. Inhibition of enzyme activity was observed when the concentration of PPi was over 250 μM. In addition, the effect of various concentrations of Mg²⁺ on the SjPPase activity was determined (Figure 4C). The enzyme activity of SjPPase increases with the concentration of Mg²⁺ from 25 to 500 μM. Subsequently, it was observed that the inhibitors NaF and MDP inhibited the enzymatic activity of SjPPase (Figure 4D).

The structures of apo-SjPPase (4QMB) and SjPPase-Pi (4QLZ) were similar. Overall, the root mean square deviation (RMSD) between the two structures was 0.454 Å (Supplementary Table 4). The two structures consist of two molecules in the asymmetric unit and also form a homodimer (Figure 5A). The two subunits in one asymmetric unit contain residues 1–284 and 1–283 out of a total of 287 residues, respectively. Superimposition of the two subunits leads to RMSDs of 0.29 Å (4QLZ) and 0.12 Å (4QMB). The monomers (Figure 5B) were arranged in a compact globular shape consisting of several β-sheets and α-helixes, one 3₁₀-helix, and some loops. The conservative residues of active sites were located at the core of the structure.

The apo-SjPPase and SjPPase-Pi had some differences. The main distinction is that there were ligands (four cobalt ions and one phosphate) in one SjPPase-Pi unit but not in apo-SjPPase. By alignment of secondary structure, it becomes apparent that changes occur in residues 132–157 involving four conserved residues (Gly139, Asp145, Asp150, and Lys152) (Figure 1). Due to the interaction between these ligands and SjPPase, the β_132–145-turn_146–147-β_148–157 of 4QMB becomes β_132–141-loop₁₄₂-β_143–145-turn_146–147-β_148–149-loop_150–151-β_152–157 of 4QLZ (Figure 1).

The active site of SjPPase is shown in Figure 5C. The overall shape and size of the active site were very similar to those of ScPPases (Tuominen et al., 1998). Ligands including magnesium acted as metal cofactors, and phosphate hydrolyzed products were incorporated in this structure. These ligand interactions were mainly derived from the hydrogen donors/receptors and coordination bond atoms. From the perspective of the six-state mechanism of PPase (Oksanen et al., 2007), the structure of the SjPPase-Pi complex is in the F intermediate catalytic cycle, which is the stage at which one Pi is released after PPi hydrolysis. However, the F intermediates differ in the two structures (4QLZ and 2IHP). Therefore, the Pi in the ScPPase-Pi complex structure is at the P1 site (Figure 6A), and the Pi in the SjPPase-Pi complex structure is at the P2 site (Figure 6B).

Molecular studies were conducted using the CDOCKER program in the Discovery Studio 2.1 software package to provide information on the competitive inhibition of MDP on SjPPase. The X-ray crystal structure of SjPPase was used for docking calculation. The representative positions of MDP and PPi considered in this work are shown in Figure 7. Generally, MDP can mimic PPi precisely to interact with SjPPase at the same binding site. In addition, MDP exhibited a similar binding mode compared to PPi (Figures 7A,B), and the same was also observed in yeast PPase (Oksanen et al., 2007). A few important hydrogen bonds are involved in the interactions. The docking models illustrated that PPi might form four hydrogen bond interactions with three key residues in the active site of SjPPase (Arg76, Tyr190, and Lys191) (Figure 7C). In the case of MDP, except for the four hydrogen bonds, an extra hydrogen bond was formed between the carbonyl oxygen atom and Arg76, as shown in Figure 7D. Moreover, MDP has a lower CDOCKER energy than PPi (MDP: –186.23 kcal/mol; PPi: –171.50 kcal/mol). Furthermore, MDP could serve as a perfect PPi mimic; it is not surprising that MDP exhibits a significantly competitive inhibition effect on SjPPase.

Furthermore, docking simulations were performed using the CDOCKER program within Discovery Studio 2.1 software package to deepen our understanding of substrate inhibition of SjPPase by PPi. The simulation was carried out using the X-ray crystal structure of SjPPase. Results obtained show that PPi bound just at the entrance of the product release channel (Figure 8A), which was formed by Lys54, Lys71, Asn72, Lys74, Arg76, Tyr190, Lys191, and Lys196 (Figure 8B). PPi might form three hydrogen bond interactions with three key residues in the product release channel (Arg76, Tyr190, and Lys191) (Figure 8B). Product (Pi) was supposed to be released through the flexible positively charged channel (Oksanen et al., 2007), which could provide a low-energy pathway out by passing Pi along the chain of lysines (Oksanen et al., 2007). As shown in the simulation models, PPi blocked the exit of Pi and cut off its release pathway (Figure 8C), which might explain the mechanism of action of the substrate inhibition of SjPPase.