Development of a new serotyping ELISA for Toxoplasma gondii type II, type III and Africa 1 lineages using in silico peptide discovery methods, well categorized feline and human outbreak serum samples

For the amplification of GRA6 and GRA7 genes, T. gondii strains that were previously isolated from stray cats were used [3]. Among these strains, 19 were type II, two were type III and one was of Africa 1 genotype in which type III and Africa 1 strains are rather rare for the region. Moreover, after complete genome sequencing of Africa 1 it is accepted as an individual lineage (without recombination) (L. Galal personal communication). During PCR, GRA6 and GRA7 genes were amplified as previously described with minor modifications using 5-GTAGCGTGCTTGTTGGCGAC-3 and 5-TACAAGACATAGAGTGCCCC-3 primers and 5-ACCCTATATTGGGGCTTGCT-3 and 5-ACACTGTCCTCGAGCTCCTA-3 primers [20, 21]. For each amplification reaction, 25 µl final volume contained 2 µl template DNA, 12.5 µl master mix (DreamTaq, Thermofisher), 1 µl from each primer (10 pmol/µl), and 8.5 µl distilled water. PCR was conducted using the following conditions: one cycle of 15 min at 95 °C for initial denaturation, 35 cycles of 94 °C for 30 s, 65 °C for 1.5 min, 72 °C for 1 min, and a final extension step at 60 °C for 30 min.

Purified PCR amplicons were sequenced by ABI Prism 3100 genetic analyzer. Sequences generated after translation into amino acid sequences by MEGA7 software were aligned with type II (Beverley; AAF60335.1) and type III (NED; AAF60337.1) reference sequences for GRA6 and type II (Beverley; ABV82436.1) and type III (NED; ABE69205.1) for GRA7. In addition, all detected polymorphisms were compared to the NCBI sequence databases to analyze whether they were novel.

Linear B cell epitopes belonging to GRA6 and GRA7 proteins were predicted by Bcepred taking into consideration hydrophilicity, flexibility/mobility, accessibility, polarity, exposed surface and turns properties (https://​webs.​iiitd.​edu.​in/​raghava/​bcepred/​bcepred_​submission.​html) [24]. Genotype specific epitopes were identified among predicted B cell epitopes which locate in the C-terminus of the GRA6 and GRA7 proteins and show properties of hydrophilicity, accessibility, flexibility and exposed surface or combinations of them [19, 25, 26]. In addition, the antigenicity value of each genotype specific B cell epitope was also analyzed by VaxiJen 2.0 (http://​www.​ddg-pharmfac.​net/​vaxijen/​VaxiJen/​VaxiJen.​html). Table 1 shows epitope sequences, antigenicity value and epitope properties. Peptides harboring genotype specific epitopes were synthesized by Elabscience^® (Elabscience Biotechnology, Inc., Wuhan, China).

3-D structure of each epitope was constructed by I-TASSER Server (http://​zhanglab.​ccmb.​med.​umich.​edu/​I-TASSER) and Protein Data Bank (PDB; http://​www.​rcsb.​org/​pdb/​) was used to retrieve the 3-D structure of heavy and light chains belonging to the variable region of B cell receptor (PDB number: 5DRW). Each epitope was docked to B cell receptor by ClusPro Server (https://​cluspro.​bu.​edu/​home.​php) and visualized on UCSF Chimera 1.14 tool.

Serum samples (n:11) collected from cats naturally infected with type II, type III or Africa 1 genotype and confirmed to be anti-T. gondii IgG positive by a commercial ID.VET ELISA kit (ID Screen^® Toxoplasmosis Indirect Multi-species, France) [3] were used for the detection of discriminative capacity of each epitope synthesized. The ID.VET ELISA kit was also used for the confirmation of negative group cat serum samples (n:9). Human sera (n:30) collected from school students exposed to an outbreak of toxoplasmosis as well as from mother/newborn pairs (n:3) were used for validation of serotyping epitopes. In addition, serum samples obtained from humans (n:38) and cats (n:24) confirmed to have anti-T. gondii IgG antibodies in our previous studies were used to reveal the serotype profile of humans and cats. During peptide-ELISA, 18 human as well as nine cat seronegative sera were used as negative control to determine the cut off value.

Peptide-ELISA was performed as described by Doel [27]. Briefly, each well of the plate (Nunc^®, Denmark) was coated by peptide which was diluted to 2 µg/well in distilled water and incubated overnight at 37 °C until all distilled water evaporated. The well was blocked with a 1×PBS containing 0.5% BSA (v/w) for 1 h at room temperature and then washed three times with PBS-T [1×PBS containing 0.1% Tween 20 (v/v)]. The sera belonging to cats or humans were diluted to 1/33 in 1×PBS containing 0.5% BSA (v/w) and 0.1% Tween 20 (v/v) and incubated for 2 h at room temperature. The wells were washed thrice with PBS-T. Anti-human IgG antibody with HRP (Sigma, Germany) or anti-Feline IgG antibody (Invitrogen-Thermo Fisher) was diluted 1/5.000 in PBS-T and incubated for 40 min at room temperature. The wells were washed thrice with PBS-T and bound antibodies were visualized after adding 3, 3′, 5, 5′ tetramethylbenzidine (TMB) substrate. The reaction was stopped by adding 75 µl of 2 N sulfuric acid and the results were evaluated in a micro titer plate reader (Bio-Tek ELx808™, USA) at 450 nm. The cut off values of ELISAs were calculated using the receiver operating characteristic (ROC) analysis.

Among T. gondii isolates from lineages type II, III and Africa 1 genotype, polymorphisms in GRA6 protein were detected at 20 different amino acid positions (Table 1). An insertion with six amino acids in length (YGGRGE) that is found in type III and Africa 1 isolates was remarkable. Also, polymorphisms in GRA7 protein were detected at 23 different amino acid positions (Table 2). Among these polymorphisms, valine and glutamine alterations found at positions 6 and 37 in type II genotype as well as phenylalanine alteration found at position 60 in type II and III were detected for the first time.

A total of 22 B cell epitopes specific to each of the three lineages were predicted using Bcepred. Twelve of them were derived from GRA6 protein while the remaining 10 of them belong to GRA7 protein. Among GRA6 epitopes, five were specific to type II, five were specific to type III and the remaining two were specific to Africa 1 lineage. Of the GRA7 epitopes, four were specific to type II, four were specific to type III and the remaining two were specific to Africa 1 lineage. All epitopes were predicted to have high antigenicity value ranging from 0.8714 to 3.0947 by VaxiJen 2.0. Similarly, all epitopes were predicted to have properties of hydrophilicity and accessibility whereas most of them also had properties of flexibility and/or exposed surface. Detailed information about selected epitopes is presented in Table 3. Also, as the number of amino acid of an epitope was smaller than 10, it was synthesized as two repeat peptide.

All selected epitopes were docked to the variable region of the light chain of B cell receptor with a high binding energy ranging from − 460.1 to − 742.7 whereas some epitopes (CYGGRGEGGGEDDRRA; CQEVPESGKDGQEVPESGKDG; CGLTRTGGSGGGSGSGPALEQEVPESGKDG and CEDGEDARQGGSGGGSGSGEDGEDARQ) were not correctly dock to the variable region of the heavy chain of B cell receptor although they had high binding energy (Figs. 1 and 2).

Discriminative capacity of each peptide was analyzed by peptide-ELISA using well categorized sera obtained from domestic cats naturally infected with genotypes belonging to type II, III and Africa 1 lineages. Initially, all serum samples were probed by all epitopes to detect the discriminative capacity of each epitope as well as sensitivity and specificity values (Table 4). Among 22 epitopes, GRA6II/c epitope derived from GRA 6 protein reacted with all serum samples from cats infected with type II while no reaction was occurred with sera from cats infected with type III or Africa 1 genotype (Fig. 3-A). As all the remaining epitopes related to type II did not react with all cat sera infected with type II genotype, only GRA6II/c peptide was accepted as a marker for serotype II due to high discrimination capacity. Similarly, GRA7III/d peptide derived from GRA 7 protein reacted strongly with serum from cat infected with type III but at the same time it also reacted weakly with two serum samples collected from cats infected with type II (Fig. 3-B). In order to increase the serotyping capacity of GRA7III/d peptide, type III specific antigen mixture containing GRA7III/a, GRA7III/b, GRA7III/c and GRA7III/d peptides was used in peptide-ELISA. Type III peptide mixture also reacted strongly with sera of cat infected with type III. However, the mixture also reacted weakly with sera from cats infected with type II (Fig. 3-C). Any other type III related peptide, except GRA7III/d, didn’t react with sera of cat infected with type III. Interestingly, GRA7 Africa 1/a epitope reacted strongly with serum samples from cat infected with type III and type II, but did not reacted with serum from a cat infected with Africa 1 genotype. Another Africa 1 specific epitope “GRA6 Africa 1/b” reacted with serum sample collected from a cat infected with Africa 1 genotype. Also, this peptide showed the cross-reactivity among different genotypes by reacting with all serum samples from cats infected with type II and III (Fig. 3-D; Table 4).

These findings obtained from analyses of cat sera demonstrated that utilization of GRA6II/c, GRA7III/d, and GRA6 Africa 1/b peptides enabled serotyping of serum samples infected with T. gondii and paved a way for the construction of a serotyping schema (Fig. 4). According to this schema, if a serum sample containing anti-T. gondii IgG antibody is found positive by GRA6II/c, it can be accepted as serotype II because this peptide only reacted with sera from cat infected with type II. On the other hand, if a serum sample is found negative by GRA6II/c, it should be studied by GRA7III/d and, if it reacts with GRA7III/d, it can be accepted as serotype III because this peptide was single peptide that reacts with serum sample from cat infected with type III. Finally, a serum sample that is found negative by both GRA6II/c and GRA7III/d should be studied by GRA6 Africa 1/b epitope and if it is found positive, this serum can be accepted as Africa 1 genotype or, if negative, it can be considered as either related to a strain from a lineage different from the first three lineages (type II, type III and Africa 1) or related to a different recombinant or atypical strain. This serotyping schema was used during validation and serotyping of the human and cat sera which were categorized in our previous studies [3, 28, 29].

Validation of the serotyping schema which is composed of GRA6II/c, GRA7III/d, and GRA6 Africa 1/b was used to determine the serotype responsible for a toxoplasmosis outbreak and in mother/newborn pairs. When these human serum samples were tested by GRA6II/c, GRA7III/d, and GRA6 Africa 1/b epitopes, all of them reacted with GRA6II/c. We therefore concluded that this toxoplasmosis outbreak was linked to a serotype II strain. All human sera collected during outbreak reacted with GRA6 Africa 1/b however four serum samples (patient 10, 13, 21 and 28) had an AV that is very close to the cut off value (Table 5). Additional validation was performed by serum samples with mother/newborn pairs (n:3) and among these serum pairs, all of them reacted with GRA6II/c and thus, all cases were accepted to be caused by serotype II. Also, all serum samples reacted with GRA6 Africa 1/b epitope as expected (Table 6).

Among human serum samples, serotype II was highly prevalent when compared to type III and Africa 1 serotypes. The prevalences of serotype II, serotype III and Africa 1 were therefore 86.8% (33/38), 7.9% (3/38) and 2.6% (1/38), respectively (Table 7).

Serotype profile was also investigated in serum samples collected from cats. According to obtained results, all serum samples (100%) were successfully serotyped using the validated peptide-ELISA approach. Among serotyped serum samples, 23 of them (95.8%) were serotype II and the remaining one (4.16%) were serotype III (Table 8).

Based on optimization and validation results that we expected, 38 human sera and 24 cat sera should have given reaction with GRA6 Africa 1/b epitope but six of human sera and six of cat sera did not give reaction with GRA6 Africa 1/b epitope despite they were serotyped as serotype II or III.