Molecular characterization and diagnostic potential of serine proteinase inhibitors from Taenia solium

Serine protease inhibitors (serpins) play essential physiological roles in a wide range of biological processes. Serpins are researched limited in Taenia solium, although some are considered to participate in host immune responses. Tsserpins were identified as typical serpins due to the primary structure of characteristic features: the serpin motif, serpin signature and reaction centre loop (RCL). RCLs of four serpin genes (TsB6, Ts4848, Ts12383 and Ts570) contained the conserved sequences of inhibitory serpins, which may involve in immune regulation. TsEP45 differed greatly from the patterns of representative serpins, suggesting that TsEP45 may be non-inhibitory. The bioinformatic analyses were supposed that Tsserpins might be a potential antigen for diagnosis. The five recombinant Tsserpin proteins were expressed and identified reacting with Cysticercus cellulosae-positive serum samples. The indirect enzyme-linked immunosorbent assay (iELISAs) based on Tsserpins were developed and validated, one of the five Tsserpins, TsEP45, showed excellent diagnostic results with 93·33% sensitivity and 94·12% specificity, respectively. This performance was in perfect accordance with the results of the bioinformatic analysis. This study provided a comprehensive demonstration of sequences and structural-based analysis of Tsserpins. The iELISAs based on five Tsserpins were developed and compared. TsEP45 was the potential species-specific antigen for developing iELISA to detect porcine cysticercosis.


Introduction
Proteases play crucial roles in the life cycle of various helminthic parasites. Parasite proteases are likely to be participated in host-parasite interactions and parasite metabolism, for instance, host tissue invasion and evasion of host immune response (Na et al. 2006). Serine protease inhibitors (serpins) are members of a superfamily proteins identified in a broad range of living beings, including viruses, bacteria, animals and plants (Irving et al. 2000;Gettins, 2002), with a characteristic well-conserved tertiary structure (Roberts et al. 1995). Serpins participate in numerous fundamental biological functions such as blood coagulation (Debrock and Declerck, 1997;Sugino et al. 2003), signalling cascades (O'Donnell and Blackman, 2005), fibrinolysis (Mak et al. 1996), inflammation (Thorgersen et al. 2007), activation of the complement system (Congote, 2007) and in mechanisms associated with host immune modulation (Chopin et al. 1988). Numerous serpins have been identified in parasitic helminths that were involved in immune regulation and parasite survival via interference with the host immune-stimulatory signals (Molehin et al. 2012).
Taenia solium is an important zoonotic parasite that causes a great threat to public health and economic loss in developing countries (Fleury et al. 2011). Despite quite a few researches regarding the biology and immunology of T. solium, the mechanisms associated with host immune evasion or the roles of serpins in these processes were still unknown. In other organisms, serpins contributed to inhibiting blood coagulation, damaging host proteases, and escaping from host immune attacks. In parasitic helminths, serpins appeared to play the similar roles (Molehin et al. 2012). Additionally, serpins were of considerable value as a diagnostic tool. Schistosoma mansoni serpins, Sm-SERPIN and Sm-RP26, showed significantly high reactivity to sera from S. mansoni-infected patients and were diagnostic of that particular species (Tanigawa et al. 2015). The Schistosoma haematobium serpin was also considered as a marked species-specific antigen (Blanton et al. 1994). Serpins also had been shown to be highly immunogenic, displaying strong reactivity to sera from infected rats (Molehin et al. 2014), suggesting that the immunodiagnostic potential of the serpins stemmed from the important physiological roles in host-parasite interactions. In view of the immunodiagnostic potential of serpins, the possible diagnostic value of Tsserpins should be explored.
Through the analysis of previously published T. solium genome and transcriptome data (Tsai et al. 2013) and structure-to-function bioinformatics analysis of the Tsserpins' polypeptides, five Tsserpins were produced as recombinant proteins in Escherichia coli system and identified as diagnostic candidates. The indirect enzyme-linked immunosorbent assays (iELISAs) were established based on the five recombinant proteins to explore a sensitive and specific assay for the diagnosis of porcine cysticercosis.

Ethics statement
This study did not involve the use of endangered or protected species, and was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences. All experiments were conducted maintaining current China laws. The protocol was approved by the Committee on the Ethics of Animal Experiments of Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences.

Parasite materials and serum samples
Healthy pigs of 40-60 days aged without cysticercosis were purchased from a local area in Gaolan, Gansu Province, China, and all pig blood samples were collected as negative control by saphenous bleeding before infection. Thirty pigs were orally infected with 3000 eggs of T. solium that was collected from Dali, Yunnan Province, China. Blood samples were collected by saphenous bleeding 60 days post-infection (dpi) (pigs were confirmed by necropsy examination). Trichinella spiralis-positive sera were from six pigs orally infected with 2000 infective muscle larvae (ML) at 35 dpi (pigs were confirmed by necropsy examination). The sera against Toxoplasma gondii were obtained from six pigs that were inoculated peritoneally with 1000 oocysts. T. gondii infection was confirmed by polymerase chain reaction (PCR) as described previously (He et al. 2016) at 15 dpi.
Cysticercus tenuicollis positive sera (six samples) were collected from naturally infected pigs from local slaughterhouse in Yongdeng, Gansu Province, China. Pigs were provided with nonmedicated feed and water ad libitum in the experiment, housed in ventilated pigsty and monitored every day. At the end of experiments, all pigs were humanely sacrificed, and sera samples were collected and stored at −20°C until use. Meanwhile, the parasites were isolated and stored at −80°C until use in other experiments.

Identification of Tsserpins
Source sequences encoding Tsserpins were analysed utilizing BLAST (Basic Local Alignment and Search Tool) with the BLASTP and BLASTN algorithm against the Taenia Genome Database (TGD) available at http://taenia.big.ac.cn/taenia/index. html and T.solium gene annotation (http: //192.168.51.199/index. html). The serpins sequences were confirmed by inspecting the expected length range of 350-450 amino acids (Molehin et al. 2012) and the presence of two highly conserved motifs (NAVYFKG and DVNEEG) (Han et al. 2000).

In silico analysis
The Tsserpins coding sequences were compared with known entries in GenBank using the BLASTP program. The physicochemical property of the mature Tsserpins including theoretical molecular weight, isoelectric points and amino acid (aa) composition were calculated using the ProtParam software (http://web.expasy.org/protparam/).

Structure modelling
The secondary structure of T.solium serpins were predicted using PredictProtein (http://www.predictprotein.org/). The threedimensional (3D) of Tsserpins and other serpins were predicted by Swiss-Model program. QMENA was used to estimate model reliability and predict quality. The predicted structures were aligned and viewed using DeepView-SWISS pdbViewer v4.1.

Phylogenetic analysis
Amino acid sequences of Tsserpins and the representative parasite serpins from GenBank (accession numbers in Fig. 2) were submitted to the ClustalW 1.83. The phylogenetic analyses were performed by MEGA6.0 for tree building and TreeView for tree drawing.

Total RNA extraction and cDNA synthesis
Total RNA was extracted from 30 mg T.solium cysticerci using Trizol reagent (Invitrogen, USA), according to the manufacturer protocol. All reagents for RNA extraction were prepared using diethypyrocarbonate (DEPC) treated water. The concentration and integrity of extracted RNA samples were determined with a Nandrop® spectrophotometer (Thermo Scientific). cDNA was synthesized from total RNA by One-step Reverse Transcription Kit (clontech) following the manufacturer's instruction.

Cloning and expression of Tsserpins
The full-length coding sequences were amplified from T.solium cysticerci cDNA by PCR using specific primers to each of the following Tsserpins: TsB6, TsEP45, Ts570, Ts4848 and Ts12383. Primers were designed using Primer Premier 5 (Table 1S). Individual reactions for PCR amplification contained a premix cocktail (dNTPs, PCR buffer, DNA polymerases) 25 µL, 1·0 µL cDNA, 1·0 µL of each primer (10 µM), 22 µL ddH 2 O (deionized water). The amplification reactions were carried out using a thermal cycling profile of 95°C for 5 min; 30 cycles at 94°C for 40 s, 65°C for 40 s, 72°C for 90 s and a final extension for 10 min at 72°C. The PCR products were analysed on a 1% (w/v) agarose gel and purified by Agrose Gel DNA Extraction Kit (Clontech).

Protein purification and immunoblotting
Recombinant proteins were purified by a Ni 2+ Sepharose™ 6 Fast Flow purification kit as the manufacturer's instructions (GE, USA). Concentrations of purified protein samples were detected by Pierce™ BCA (bicinchonininc acid) protein assay kit (Thermo Scientific, USA) and identified by Western blotting. Briefly, the recombinant proteins were isolated by SDS-PAGE and transferred to nitrocellulose membrane, and then the membrane was incubated with the positive serum from porcine cysticercosis (1:200 dilution) and subsequently incubated with horseradish peroxidase (HRP)-conjugated rabbit anti-pig IgG (1:10 000 dilution, Sigma), the reactions were detected with DAB (diaminobenzidine).

Statistical analysis
Receiver operating characteristics (ROC) (Greiner et al. 2000) analysis was performed to assess the diagnostic value of antigens. One-way ANOVA tests were performed to compare all pairs of groups. Statistical significance was set at P < 0·05. Data analysis was showed using GraphPad Prism 5.0 (GraphPad Software Inc., San Diego, USA).

Cloning and general characteristics of Tsserpins
The cloned full-length cDNA and general characteristics of Tsserpins are listed in Table 1. According to the MEROPS (http://merops.sanger.ac.uk/inhibitors/index.shtml) classification of protease inhibitors (Rawlings,et al. 2014), Tsserpins are members of inhibitor family I4 (Clan ID). Multiple sequence alignment of Tsserpin amino acid sequences showed low overall identity (21-69%) with representative serpins from GenBank (Fig. 1). Analysis of Tsserpin peptides indicated the presence of three conserved motifs, the serpin signature, serpin motif and the reactive centre loop (RCL) (  (Gettins, 2002)，where the amino acid residues are defined as …P 3 -P 2 -P 1 (cleavage site)-P 1 ′ -P 2 ′ -P 3 ′ …and X is any amino acid. The RCL of TsB6, Ts4848 and Ts12383 were highly similar to typical pattern of RCL, while Ts570 was with residue variation only at position P 16 and P 14 . Unexpectedly, TsEP45 differed greatly from the patterns of other serpins, which showed several variations: P 14 [I] and P 12−9 [LLSV]. No signal peptide or transmembrane domain was found in the peptide sequences. Other post-translational modification sites, such as cAMP-and cGMP-dependent protein kinase phosphorylation, casein kinase II phosphorylation, N-myristoylation, protein kinase C phosphorylation and tyrosine kinase phosphorylation sites were predicted to present in the five Tsserpins (Table 3S).
According to the analysis of hydrophilicity, flexible regions, antigenic index and surface probability, potential B cell epitopes also might be present in all Tsserpins (Table 4S). Phylogenetic analysis showed that the Tsserpins were distributed on different evolutionary branches (Fig. 2). Tsserpins all had the high-sequence homology to Echinococcus granulosus serpins: TsB6 shared 67% sequence identity with SerpinB9 (Accession number EUB60253.1), Ts4848 and Ts12383 shared 79% to E.granulosus Serpin (Accession number CDS22753.1), Ts570 shared 75% with E.granulosus Serpin (Accession number EBU63679.1). TsEP45 was closely related to E.granulosus Serpin (Accession number EUB63680·1) with 81% sequence identity. Tsserpins were distantly related to those of other helminths, especially the evolutionary status of TsEP45 and Ts570 were very ancient, far away from the serpins of several common porcine parasites, including Ascaris suum, T. gondii and T. spiralis.

Structure-based alignment analysis of Tsserpins
The potential tertiary structure of a protein reveals several specific biological functions. 3D protein structures of Tsserpins were predicted by homology modeling. Firstly, template structures were developed by SWISS-MODEL. The search results (Fig. 3) indicated that TsB6, Ts4848 and Ts12383 shared similar structure to SerpinB3 (2zv6, Homo sapiens, human), the structure of TsEP45 was similar to the serine protease inhibitor (3sto, Schistosoma haematobium), and Ts570 had a similar structure to anithrombin-III (2hij, H. sapiens, human). Ramachandran plotting indicated that the structure models were optimal (data not shown).

Expression and identification of Tsserpins
The PCR products were obtained (Fig. 1S) and inserted into pET-30a (+) vector. Then the recombinant plasmids pET-30a (+)-Tsserpins were confirmed by enzyme digestion (Fig. 2S) and DNA sequencing. The recombinant pET-30a (+)-Tsserpins were expressed and analysed by Western blotting. SDS-PAGE analysis showed the recombinant proteins were mainly found in the pellet (Fig. 3S). The peak expression was observed at 37°C when induced with IPTG (isopropyl-β-D-thiogalactoside) at 1·0 mM L −1 , following 6 h induction. Tsserpins were purified by Ni 2+ -affinity chromatography, and a single band for each of recombinant proteins in SDS-PAGE indicated that the recombinant serpins were free of obvious contaminants (Fig. 4). Immunoblotting results showed Tsserpins were reacted with the positive T.solium cysticercus sera (Fig. 5). And no reaction was observed between Tsserpins and the negative control sera from healthy animals.  Guangxue Liu et al.

Evaluation of the serodiagnostic potential of Tsserpins
To evaluate the possibility of Tsserpin as antigens for diagnosis of cysticercosis, we compared the reactivity of the Tsserpin antigens and various serum samples from pigs experimentally infected Cysticercus cellulosae (n = 30) and negative samples with iELISA.
The results showed that TsEP45, TsB6 and Ts570 were specifically reactive with sera from T. solium cysticercosis (Fig. 6). There was no cross-reactivity with positive samples of C. tenuicollis, T. gondii and T. spiralis. The remained two antigens, Ts4848 and Ts12383, detected without disease-specific reactivity ( Fig. 6E and 6F), although detective results of positive sera samples from infected pigs with C. cellulosae were higher than those of negative ones. To validate of diagnostic performance of iELISAs based on TsEP45, TsB6 and Ts570, ROC analysis was employed for further evaluation. The cut-off values of TsEP45, TsB6 and Ts570 were calculated by sera from 51 uninfected pigs (negative control) and 30 C. cellulosae positive sera. The ROC curve and the corresponding statistics were summarized in Fig. 7. It can be concluded that TsEP45 showed the better diagnostic value than TsB6 and Ts570 and showed acceptable results to be a diagnostic antigen in iELISA.

Discussion
Tsserpins were identified as typical serpins due to the presence of characteristic features such as the serpin motif, serpin signature and RCL (Rawlings et al. 2014). Additionally, secondary and tertiary structure prediction analyses indicate that Tsserpins may contain 8-9 α-helices and three β-sheets, features were consistent with the known serpins (Huntington, 2006). The results of the predicted tertiary structure of Tsserpins (Fig. 3) showed that the RCLs of native inhibitory serpins may be exposed and accessible to target proteases, which further strengthens the argument that Tsserpins belong to the serpin superfamily (Rawlings et al. 2014). These observations allowed us to explore Tsserpin biological function based on comparative modelling of known serpins. The conserved 20/21 residue peptide 'P 17 [E]-P 16 [E/K/R]-P 15 [G]-P 14 [T/S]-P 13 [X]-P 12−9 [AGSX]-P 8−1 [X]-P 1 ′ −4 ′ ' within the RCL determined whether the serpin was inhibitory or noninhibitory (Irving et al. 2000). The primary structures of inhibitory and non-inhibitory serpins have different amino acids of the hinge region (P 15 to P 8 ), whereas they are well conserved in inhibitory serpins. In general, P 15 is glycine, P 14 threonine or serine. P 12 to P 9 , are all alanine residues in most inhibitory serpins. Analysis of Tsserpins showed that TsB6, Ts12383, Ts570 and Ts4848 have typical serpin structures, suggesting that they are functional inhibitory serpins. Further analysis of the hinge regions of TsB6, Ts12383, Ts570 and Ts4848 suggested that they are occupied predominantly by hydrophobic amino acid residues, which are deemed to be the construction of the skeleton conformation needed for the inhibitory activity (McCarthy and Worrall, 1997). However, the corresponding regions of non-inhibitory serpins do not conform to this consensus sequence, for example, TsEP45 shows some variations: P 14 [I] and P 12-9 [LLSV]. We can speculate, therefore, that TsEP45 may be a non-inhibitory. Non-inhibitory serpins have evolved by ceasing to act as protease inhibitors and by acquiring the ability to interact with other molecules, performing new functions.
To our knowledge, the divergent functions or the specificity of target proteases mainly depends on the variety of RCLs. The target proteases mostly cleave the bonds between P 1 and P 1 ′ in RCLs (Silverman et al. 2001), which suggested that serpins with different amino acid residues of P 1 and P 1 ′ acted on diverse target proteases. According to the reference (Merckelbach and Ruppel, 2007) of the amino acid sequence of serpin Emu , R 344 (arginine) can be identified as the P 1 residue of Ts12383 and Ts4848, as well as cysteine is found at P 1 ′ . Arginine at P 1 could explain  Parasitology Open 5 readily the inhibition of trypsin and pancreatic elastase (PE) since these proteinases cleave their substrates with high specificity after basic amino acids (Merckelbach and Ruppel, 2007). Likewise, the intracellular serpin B6 (human) inhibits trypsin via a P 1 -R (Riewald and Schleef, 1996). Therefore, we can envisage that Ts12383 and Ts4848 have potential to inhibit the activity of trypsin and PE. Based on the amino acid sequence of α-1-proteinase inhibitor (α1-antitrypsin), M (methionine) can be identified as the P 1 residue of TsB6 and Ts570. Human neutrophil elastase (NE) was inhibited by α1-antitrypsin through P 1 -M. So, we can speculate that TsB6 and Ts570 have similar function to α1-antitrypsin inhibiting the activity of NE. In summary, P 1 -P 1 ′ residues were very important in the RCL, and P 1 played a crucial role in defining specificity (Potempa et al. 1994). Some serpins were specialized for single proteases, while others inhibited more than one substrate; further in vitro studies could provide evidence of the range of protease inhibition. The predictions   Each data point represents an average of three ELISA replicates. Statistical significance was set at P < 0·05 as depicted using asterisks: * = P < 0·05, ** = P < 0·01, *** = P < 0·001.

Parasitology Open 7
above were contributed to a better understanding of T.solium in evading the host immune response. If the Tsserpins were secreted during the oncosphere infection phase, it might be able to block the proteolytic attack of host digestive enzymes, which could be the targets of the intestinal immune system and vaccine candidates. Phylogenetic analysis showed that Tsserpins are distributed on different evolutionary branches, suggesting the serpins had functional polymorphism and early evolutionary divergence. The Tsserpins in the same branch might share similar function, such as Ts12383, Ts4848 and serpin Emu (Echinococcus multilocularis). In addition, the results of phylogenetic analysis showed that the evolutionary status of TsEP45 was very ancient, far away from the serpins of several common porcine parasites, including A. suum, T. gondii and T. spiralis, suggesting that TsEP45 was a special serpin in common parasites and has potential to act as a species-specific antigen for developing iELISA to detect porcine cysticercosis.
Even though Tsserpins do not contain signal peptide sequences by bioinformatics analysis, they would be released extracellularly through a special pathway independent of the endoplasmic reticulum. Examples from other organisms include human plasminogen-activator inhibitor 2 (Ritchie and Booth, 1988) and Hc-serpin from Haemonchus contortus (Yi et al. 2010). In addition, exosomes were playing crucial roles in cell communication and delivery of components (Valadi et al. 2007), and it was a possibility that Tsserpins were secreted by exosomes and inducing host immune responses. Based on this initial characterization, Tsserpins had great potential to act as candidate antigens to detect the cysticercus-specific antibodies. To determine the diagnostic potential of Tsserpins in serological assays, the fusion-proteins were coated as antigen in iELISAs. It was found that TsEP45 could effectively differentiate the following parasites species: T. solium, C. tenuicollis, T. gondii and T. spiralis. By detecting sera samples from cysticercus infected pigs, ROC analysis of TsEP45 iELISA exhibited the high sensitivity (93·33%) and specificity (94·12%), respectively. Particularly, the fusion-protein acting as a diagnostic antigen in iELISAs mainly focused on solving the cross-reactivity between C. cellulosae and C. tenuicollis (P < 0·001). The detection results were in accord with sequences alignment results, TsEP45 shared very low similarity to serpins from other species (data not shown). In summary, TsEP45 was a candidate diagnostic antigen for porcine cysticercosis. Further study should be carried out to clarify the relationship between larva loads and dilution of antibody against TsEP45, and how many cysticerci could be detected with this iELISA. The mentioned above are valuable information for establishing a diagnostic method. Meanwhile, iELISA based on TsEP45 will be used to detect the serum samples collected from naturally infected with C. cellulosae and compared with commercial serological methods to demonstrate the feasibility in the future.