Plant-derived chimeric antibodies inhibit the invasion of human fibroblasts by Toxoplasma gondii

Parasites and in vivo passaging in mice

All parasite and cell culture procedures were carried out in class II type A2 lamina flow hoods. The Hs68 fibroblast cell line derived from human foreskin fibroblast (HFF) was obtained from ATCC (CRL-1446). The cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% (v/v) fetal calf serum (FCS), two mM L-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin. When the cells were 100% confluent, they were maintained in infection medium (as above, but with 1% (v/v) FCS).

The Hs68 cells were inoculated with 4 × 10⁶ T. gondii tachyzoite cells and incubated in a humidified cell culture chamber at 37 °C with a 5% CO₂ atmosphere. After 48 h, most of the T. gondii tachyzoites were egressed from the host cells. A cell scrapper was used to disrupt cells containing rosettes (tachyzoites) and the tachyzoites were passed through a 27-gauge needle and centrifuged at 1,000×g for 15 min. The pellet containing tachyzoites was resuspended in five ml infection medium and filtered through a three-μm Nucleopore polycarbonate filter membrane (GE Healthcare Life Sciences, Marlborough, MA, USA) to remove cell debris. The filtered parasites were then quantified with a hemocytometer and trypan blue staining (Sambrook & Russell, 2001). Native soluble antigen (CAT. No: EV8131) from lysed T. gondii was acquired commercially (Meridian Life Science, Memphis, TN, USA).

Mouse immunization

Three BALB/c mice, 6–8 weeks old, were immunized by the subcutaneous injection of 2.02–4.04 × 10⁶ T. gondii RH tachyzoites emulsified in Freund’s Complete Adjuvant (CAT. No: F5881) (Sigma-Aldrich, St. Louis, MO, USA) at a ratio of 1:1, with an interval of 2 weeks (day 0 and day 14), followed by a second booster injection on day 21 with the same amount of tachyzoites but this time without adjuvant. The immunized mice were bled 4 days after the second booster and the serum was tested for the presence of T. gondii-specific IgG antibodies by immunoblot analysis (Towbin, Staehelin & Gordon, 1979). All methodologies were carried out in accordance with University of Malaya, Animal Care and Use Ethical Committee (ACUC) guidelines and all experimental protocols were approved by the ACUC of the University of Malaya (Clearance number: PAR/18/06/2007/FMY(R). Following the detection of T. gondii-specific IgG antibodies, mouse spleens were harvested, snap-frozen in liquid nitrogen and stored at −80 °C.

Construction of the phage–scFv display library

Total RNA was prepared from homogenized mouse spleen using the RNeasy Mini Kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer’s instructions. The mRNA was purified using the MicroPoly(A) Purist™ small scale mRNA purification kit (Ambion, Foster City, CA, USA) and 7.8 μg was used as the template for cDNA synthesis with the Superscript™ III Reverse Transcriptase (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. Typically, 0.2 μg of mRNA was used per reaction for first-strand cDNA synthesis primed with an oligo-(dT)₂₀ primer (Invitrogen, Carlsbad, CA, USA). The resulting cDNA was used as a template for the amplification of the murine IgG variable heavy (V_H) and light (V_L) chain gene regions as previously described (Nathan et al., 2002) incorporating flanking SfiI and NotI restriction sites in the V_L 5′ and V_H 3′ domains, respectively. All primer sequences are listed in Table S1. The primer combinations were designed to amplify most of the known mouse antibody sequences thus generating a polyclonal pool of V region fragments.

A total of 23 Vκ region PCRs (50-μl volume) were carried out to amplify the κ light chain (LC) variable region as follows: 0.4 μM MSCVκ (Vκ 5′ sense/SfiI) primer mix, 0.4 μM MSCJκ (Vκ 3′ antisense) primer mix, two μl first-strand cDNA, 0.3 mM dNTP mix (Promega, Madison, WI, USA), 1× PCR buffer and 2.0 U DyNAzyme™ II DNA polymerase (Finnzymes, Vantaa, Finland). The Vλ gene region amplifications were prepared using one set of forward and reverse primers only and were set up as above except that the primer mixes were substituted with the MSCVL-1 (Vλ 5′ sense/SfiI) and MSCJKL-B (Vλ 3′ antisense/short linker) primers. Four Vλ PCR reactions were prepared.

The V_H gene region was amplified using a combination of 19 forward primers and three reverse primers, generating a total of 30 V_H reactions prepared as described above. The V_H region PCR reactions included the MSCVH (V_H 5′ sense) and MSCG (V_H 3′ antisense/NotI) primer mixes for the amplification of V_H regions IgG1, IgG2a, IgG2b, IgG3, and IgM. All amplifications began at 94 °C for 3 min followed by 30 cycles of 94 °C for 30 s, 56 °C for 30 s, 72 °C for 90 s, and a final extension step at 72 °C for 2 min. The PCR products were purified using the QIAquick Gel Extraction Kit (Qiagen GmbH, Hilden, Germany).

The V_H and V_L coding sequences were fused using the flanking primers MSCF (5′-GCG GGG CCC AGC CGG CCG AGC TCG-3′) and RSCB (5′-GCC TGC GGC CGC ACT AGT GAC AGA-3′) by splice-overlap extension PCR (SOE-PCR) to produce the scFv fragments. The scFv assembly reaction comprised one μM MSCF primer, one μM RSCB primer, ∼50 ng each of the V_H and V_L coding sequences, one mM dNTP mix, 2× amplification buffer, one mM MgSO₄, 2× PCR_x enhancer solution, and one U Platinum Pfx DNA Polymerase (Invitrogen, Carlsbad, CA, USA). The reactions comprised seven cycles at 92 °C for 1 min, 63 °C for 30 s, 58 °C for 50 s, and 72 °C for 1 min, followed by 23 cycles of pull-through reactions at 92 °C for 1 min, 63 °C for 30 s, and 72 °C for 1 min, followed by a final extension step at 72 °C for 10 min.

The resulting polyclonal pool of scFv fragments was sequentially digested with SfiI and NotI and transferred to the pCANTAB5E phagemid vector (GE Healthcare Life Sciences, Marlborough, MA, USA). The ligation products were introduced into electrocompetent Escherichia coli TG1 cells by electroporation (2.5 kV, 200 Ω, 25 μF, time constant >3.6 ms). Five sets of scFv electroporation transformations were carried out to generate a combinatorial scFv library in which the antibodies were expressed as fusions with the M13 minor phage coat protein pIII after rescue with a helper phage.

Biopanning of the phage-scFv display library

The phage-scFv display library was incubated with T. gondii tachyzoites (2.88 × 10⁴) in a pre-blocked microcentrifuge tube at 8 °C for 2 h while rotating. The tachyzoite cells were then pelleted by centrifuging at 1,000×g, 4 °C, for 5 min and the supernatant was discarded. The cells were vortexed five times with Todd–Hewitt broth (THB) containing 0.05% (v/v) Tween-20 (THB-T) and centrifuged as above. Bound phage were eluted by resuspending cells in 200 μl glycine-HCl (pH 2.2) at 4 °C for 10 min and neutralized by adding 26.7 μl Tris-Cl (pH 8.0). The phage eluate was then introduced to log-phase E. coli TG1 cells and plated on super optimal broth with ampicillin and glucose (SOBAG) overnight at 30 °C, at 100× and 10× dilutions for phage output titering. The phage output titer was calculated from the mean of triplicate sets of each dilution to determine the number of phage captured on the cells. The remaining phage-infected E. coli TG1 suspension was plated on large SOBAG plates and also incubated at 30 °C overnight to obtain single colonies of T. gondii-binding scFv clones. Each scFv clone was verified by PCR for the presence of the full-length scFv sequence and fingerprinted with MvaI (Thermo Fisher Scientific, Waltham, MA, USA). Full-length scFv genes with unique fingerprint patterns were selected and the clones were sequenced using the dideoxy method with the pCANTAB5 sequencing primer pairs S1 and S6 (GE Healthcare Life Sciences, Marlborough, MA, USA).

Monoclonal scFv binding titer assay

Phage rescue of the monoclonal full-length scFv clones was achieved using VCSM13 interference-resistant helper phage (Stratagene, San Diego, CA, USA). The recombinant phage suspensions for each clone (2.03 × 10¹¹ p.f.u.) were blocked with 1% (w/v) bovine serum albumin (BSA) before incubation with T. gondii tachyzoites (1.0 × 10⁵ cells/reaction). Each scFv clone was also tested for binding to HFF cells (1.0 × 10⁵ cells/reaction) as a control. All reactions were incubated at 8 °C for 2 h while rotating. The centrifugation, washing, phage-elution steps, and re-infection steps were carried out as described above. The means of the samples and negative controls were compared using the one-tailed Wilcoxon–Mann–Whitney statistical test (H₀ < H_A, α = 0.05) in SigmaPlot v11.0 (Systat Software Inc., San Jose, CA, USA). The exact p value of the Wilcoxon Mann–Whitney statistics was computed using Analyse-it® v2.22 statistical analysis software for Microsoft Excel.

Sequence analysis and structural modeling of V regions

The full-length V_H and V_L chain sequences were numbered according to the International ImMunoGeneTics (IMGT) unique numbering scheme (Lefranc et al., 2003) and analyzed using V-Quest (imgt.cines.fr/textes/vquest) to determine the CDRs and germline origins. Target-unrelated peptide (TUP) motifs and sequence redundancy were identified using scanner and reporter of target-unrelated peptides (SAROTUP) and MimoDB.

Structural modeling was accomplished using the web-based antibody modeling software Rosetta Antibody: Structure Prediction Server provided by the Department of Chemical and Biomolecular Engineering, Johns Hopkins University (http://antibody.graylab.jhu.edu/) (Sircar, Kim & Gray, 2009). The protocol integrates ab initio CDR H3 loop modeling and simultaneous optimization of CDR backbone dihedral angles and the relative orientation of V_L and V_H. Only first rank predictions were used for each scFv model. The analysis and superimposition of scFv molecular models was generated with PyMOL for Windows.

Production of soluble scFv antibody fragments

Candidate scFv-phages particles were used to infect E. coli HB2151 cells, a non-suppressor strain for the IPTG-inducible expression of antibody fragments. Soluble scFv antibodies were extracted from the bacterial periplasm as previously described (Barbas et al., 2001). Periplasmic extracts were analyzed by 12% (w/v) SDS–PAGE (Laemmli, 1970) and transferred to nitrocellulose membranes which were probed overnight at 4 °C with an anti-E-tag primary antibody (Abcam, Cambridge, UK) diluted 1:1,000 in Tris-buffered saline (TBS) with 2% (w/v) skimmed milk. Blots were washed three times with TBS-T (TBS with 0.05% (v/v) Tween-20), and incubated with a horseradish peroxidase-conjugated anti-rabbit IgG secondary antibody (CAT. No: A0545) (Sigma-Aldrich, St. Louis, MO, USA) diluted 1:1,000 in 2% (w/v) skimmed milk for 1 h with gentle agitation. After three further washes in TBS-T, the signal was detected with SuperSignal West Pico ECL substrate (Thermo Fisher Scientific, Waltham, MA, USA).

Construction of affinity-matured scFv library

Antibody hotspot regions in the CDR were identified using V-Quest. The hotspot motif RGYW in scFv TG130 V_L CDR1 was chosen for the introduction of random mutations to produce an affinity-matured scFv antibody library. Completely overlapping sense and antisense DNA oligomers were designed to generate a library randomizing four nucleotides (two consecutive amino acids) in the hotspot while preventing the introduction of stop codons. The following degenerate oligomers were used, with the randomized sequences underlined and bold: pCANTAB5 E—RGYW Fwd (Sense), 5′-G GCC AGT CAG GAT GTG VNS NCT GCT GTA GCC-3′; pCANTAB5 E—RGYW Rev (Antisense), 5′-GGC TAC AGC AGN SNB CAC ATC CTG ACT GGC C 3′, (V = A or C or G, S = C or G, B = C or G or T). The affinity maturation reaction was carried out using the QuickChange Lightning Site-Directed Mutagenesis Kit (Stratagene, San Diego, CA, USA).

Before site-directed mutagenesis, oligonucleotide adaptors were generated from the degenerate oligomers to improve the reaction performance. Equimolar amounts of the degenerate oligonucleotides (pCANTAB5 E—RGYW Fwd and pCANTAB5 E—RGYW Rev) were resuspended in annealing buffer (10 mM Tris-HCl pH 7.5, 60 mM NaCl) and heated to 95 °C for 10 min before cooling to room temperature overnight to form an oligonucleotide duplex. Following the generation of the degenerate adaptors, the mutagenesis PCR reaction was carried out using 10.6 ng of the phagemid pCANTAB5E-TG130 as a template and 30.0 pmol of the degenerate DNA adaptors (15.0 pmol or 153.4 ng each of the degenerate DNA oligomers pCANTAB5 E—RGYW Fwd and pCANTAB5 E—RGYW Rev). The scFv template and DNA adaptors were mixed with the QuikChange Lightning enzyme (Agilent Technologies, Santa Clara, CA, USA) and its component reagents according to the manufacturer’s protocol in a 50-μl reaction volume and heated to 95 °C for 2 min, followed by 18 cycles at 95 °C for 20 s, 60 °C for 10 s, and 68 °C for 2.5 min, and a final extension step at 68 °C for 5 min. The mutagenesis PCR product was digested with Dpn1 to remove all parental phagemid dsDNA, precipitated with ethanol and introduced into XL-10 Gold® ultracompetent cells (Stratagene, San Diego, CA, USA) according to the manufacturer’s protocol. Transformed clones were verified by colony PCR and sequenced to check for successful sequence diversification. Sequences were aligned using the BioEdit v7.0.5.3 (Hall, 1999). Isolated plasmid DNA was introduced into chemically-competent E. coli TG1 cells (Tu et al., 2005) and rescued using helper phage VCSM13 (Stratagene, San Diego, CA, USA) (Barbas et al., 2001).

Biopanning of the affinity-matured scFv library

The affinity-matured scFv library prepared by phage rescue was precipitated with polyethylene glycol and blocked with 2% (w/v) BSA (1:1 ratio) before biopanning. The recombinant phage (∼2.7 × 10¹² cfu) from the mutated library was incubated in triplicate with T. gondii tachyzoites (∼4.0 × 10⁵ cells) in pre-blocked microcentrifuge tubes and the mixture was rotated at four rpm at 8 °C for 1 h, with VCSM13 helper phage negative controls. The cells were pelleted by centrifugation at 1,000×g, 4 °C for 5 min and the supernatant was discarded. The cells were resuspended in cold THB-T, vortexed and incubated for 5 min at 4,000 rpm, 8 °C. This washing step was repeated for 10 consecutive rounds, and bound phage were eluted by resuspending the cells in glycine-HCl (pH 2.2) as described above. After re-infection, the eluted phages were titered to determine the quantity of captured phages, and glycerol stocks of the eluate were prepared and stored at –80 °C.

Eluted recombinant phages were used to re-infect log-phase E. coli TG1 cells and plated on SOBAG medium. After an overnight incubation at 30 °C, 96 single colonies were picked and each clone was sequenced to characterize the captured affinity-matured scFv antibodies. The five unique affinity-matured phage-scFv clones were rescued using VCSM13 interference-resistant helper phage (Stratagene, San Diego, CA, USA) and assessed by biopanning against T. gondii tachyzoites. The recombinant phage suspensions from each clone (mean titer 2.03 × 10¹¹ cfu/ml) were blocked in 2% (w/v) BSA before incubation with T. gondii tachyzoites (1.0 × 10⁵ cells per reaction) as described above.

Conversion of scFv antibodies into chimeric human IgGs and transient expression

The V_H sequence from each scFv was inserted into the plant expression vectors pTRAkc-HC upstream of the Cγ region using the restriction sites BspHI and NotI, whereas the V_L sequence was inserted into pTRAkc-LC upstream of the Cκ region using the restriction sites NcoI and NotI. The pTRA plant expression vectors are derivatives of pPAM (GenBank: AY027531). The pTRA constructs were introduced into Agrobacterium tumefaciens GV3101 by electroporation (Eppendorf, Hamburg, Germany) using the standard settings of 2.5 kV, 5 ms, 200 Ω, and 25 μF, and the transformants were stored as a 50% (v/v) glycerol suspension at −80 °C. Bacterial suspensions were prepared for agroinfiltration (Piotrzkowski, Schillberg & Rasche, 2012) by diluting acetosyringone-induced bacteria to OD 1.0 with infiltration medium. Combinations of pTRAkc-HC and pTRAkc-LC transformants were mixed at a 1:1 ratio prior to infiltration. The suspensions of the five HC + LC combinations (corresponding to the five scFv variants) were co-agroinfiltrated into N. benthamiana leaves by standard syringe infiltration (D’Aoust et al., 2009) and plants were incubated in phytotrons for 5 days before harvesting.

Infiltrated leaves were weighed before grinding to a fine powder in liquid nitrogen with a mortar and pestle. Total soluble protein was extracted in two ml PBS containing 10 mM sodium disulfide per gram of leaf material. Cell debris was removed by filtering through Miracloth (Merck Millipore, Billerica, MA, USA) followed by centrifugation at 40,000×g, 20 min, 4 °C. The supernatant was adjusted to 500 mM NaCl and allowed to cool on ice for 30 min before centrifuging as above and passing through a 0.45-μm syringe filter.

The chimeric IgGs were purified using Protein A Sepharose (GE Healthcare Life Sciences, Marlborough, MA, USA) (Hehle et al., 2011) and desalted with PD-10 columns (GE Healthcare Life Sciences, Marlborough, MA, USA) using the manufacturer’s gravity protocol. Purified and desalted fractions were analyzed by reducing and non-reducing SDS–PAGE with Coomassie staining to determine antibody integrity. The extracted IgG was detected using alkaline phosphatase (AP)-conjugated human Fc-specific (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA) and human κ LC-specific (Sigma-Aldrich, St. Louis, MO, USA) antibodies, and the signals were developed using NBT substrate. Purified IgG antibody concentrations were determined using a Qubit® 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA).

ELISA and immunoblot analysis of anti-T. gondii chimeric IgG

Toxoplasma gondii antigen (CAT. No: EV8131) (Meridian Life Science, Memphis, TN, USA) was coated onto high-binding enzyme-linked immunosorbent assay (ELISA) plates (Dynex Technologies, Chantilly, VA, USA) at a concentration of 0.5 μg per well in PBS overnight at 4 °C. The plates were washed three times with PBS and blocked with 3% (w/v) BSA in PBS for 2 h at room temperature. HFF cell lysates and mouse ascites (CAT. No: M8273) (Sigma-Aldrich, St. Louis, MO, USA) as controls were coated at a concentration of 0.5 μg per well and processed as above. The recombinant IgGs purified from transformed plants were each diluted to a working concentration of two μg/ml in 3% (w/v) BSA and tested with the antigens for 2 h at room temperature. The ELISA plates were washed four times with PBS-T before detecting binding antibodies with HRP-conjugated anti-human Fc diluted 1:10,000 for 1 h at room temperature. The plates were washed three times with PBS-T, followed by two washes in PBS. The assays were developed with 1-Step™ Ultra TMB ELISA substrate (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. Reactions were stopped after 30 min with 2M H₂SO₄ and readings taken at OD_{450 nm} using a microplate reader (Tecan, Maennedorf, Switzerland).

For immunoblot analysis, T. gondii total antigen lysate (CAT. No: EV8131) (Meridian Life Science, Memphis, TN, USA) was either prepared by boiling in reducing sample buffer (1% (w/v) SDS, 10% (v/v) glycerol, 10 mM Tris-Cl, pH 6.8, one mM EDTA, 0.05 mg/ml bromophenol blue, 5% (v/v) β-mercaptoethanol) for 5 min, or mixed with NativePAGE™ sample buffer (Thermo Fisher Scientific, Waltham, MA, USA), before fractionation on a 12% (w/v) acrylamide gel. The gel was blotted onto a nitrocellulose membrane and blocked with 5% (w/v) BSA in TBST. The blotted membrane was cut into strips and probed with the chimeric IgGs at a concentration of one μg/ml, overnight at 4 °C. Bound IgGs on individual strips were detected using an AP-conjugated anti-human Fc antibody (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA). Treated strips were developed with Western Blue® stabilized substrate for AP (Promega, Madison, WI, USA).

Immunofluorescence microscopy

Acid-washed coverslips were treated with 0.01% (w/v) poly-l-lysine solution (CAT. No: P4707) (Sigma-Aldrich, St. Louis, MO, USA) for 10 min at room temperature, drained, dried, and covered with freshly-harvested parasites (approximately 10⁵ cells per coverslip). The cells were allowed to attach for 1 h at 37 °C followed by fixation in 2.5% formaldehyde for 30 min on ice. The samples were blocked with 3% (w/v) BSA in PBS. To visualize the binding of chimeric antibodies to the surface of intact parasites, fixed coverslips were incubated with each chimeric Ig antibody (five μg/ml) for 16 h at 4 °C, washed five times in PBS-T and stained with AlexaFluor 488-goat-anti-human-IgG (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA) for 1 h. Coverslips were mounted with ProLong Gold (Thermo Fisher Scientific, Waltham, MA, USA) and cured for 24 h. All staining steps were conducted with antibodies diluted in 3% BSA in PBS. Coverslips incubated with the buffer control and secondary antibodies were included as negative controls. Results were visualized using a 100 × lens on an Olympus BX51 microscope.

Invasion and attachment assays

Parasite invasion assays were carried out as described (Tyler & Boothroyd, 2011) with minor modifications. Briefly, washed RH parasites were incubated in DMEM with 2% (v/v) FCS supplemented with 35 μg/ml of each antibody or 1.5% volume of a buffer control (PBS) and incubated at 37 °C for 30 min while rotating. Invasion was encouraged by incubating on coverslips confluent with a monolayer of HFF cells for 2 h at 37 °C in a CO₂ incubator (1 × 10⁷ tachyzoites in 140 μl per coverslip for six-well plates). Infected monolayers were then gently washed twice in PBS and fixed in 2.5% formaldehyde. Extracellular parasites were stained with rabbit anti-T. gondii polyclonal antiserum followed by AlexaFluor 594-donkey-anti-rabbit secondary antibodies. Infected monolayers were first permeabilized with 0.2% Triton X-100 in PBS and then blocked with 3% BSA. Parasites were then stained with goat anti-T. gondii polyclonal antiserum followed by AlexaFluor 488-donkey-anti-goat secondary antibodies. All primary and AlexaFluor-conjugated secondary antibodies were obtained commercially (Abcam, Cambridge, UK). Coverslips were mounted with Prolong Gold containing DAPI (Thermo Fisher Scientific, Waltham, MA, USA) overnight. The numbers of green (intracellular and extracellular) and red (extracellular) tachyzoites were scored on 30 randomly-selected fields on each of two separately-mounted coverslips with different antibody treatments, and visualized using a 100 × lens on an Olympus BX51 microscope. All digital images were acquired and analyzed using the Nikon NIS Elements software. Results are representative of data from at least two independent experiments.

Attachment assays proceeded as described for the invasion assay, except that the HFF monolayers on coverslips were pre-fixed with 2.5% paraformaldehyde before parasite incubation. Freshly-released tachyzoites were incubated with the monoclonal antibodies (25 μg/ml IgG in DMEM per 1.0 × 10⁶ tachyzoites) separately for 30 min at 37 °C while rotating. Pre-treated tachyzoites were then allowed to attach to fixed monolayers on a six-well plate for 15 min at 37 °C. The attached parasites were stained with rabbit anti-T. gondii polyclonal antiserum followed by AlexaFluor 594-donkey-anti-rabbit secondary antibodies. After washing, fixing and mounting, the number of attached parasites was determined in 15 random fields at 40× magnification. All assays were carried out in triplicate.

Plaque assay

Tachyzoites were pre-treated at 37 °C for 30 min before infection with the individual monoclonal antibodies, each at a concentration of 2.5 μg/ml. Confluent monolayers of HFF cells were grown in six-well plates and seeded with 1.0 × 10⁴ tachyzoites per well for 2 h at 37 °C. The monolayers were washed and replenished with fresh DMEM, before further incubation for 5 days at 37 °C. The monolayers were then fixed in absolute ethanol for 5 min and stained for 1 min with a crystal violet solution as described (Martínez-Torrecuadrada et al., 2005). Images were captured using a Nikon Eclipse TS100 inverted microscope, with 20× magnification. A total of 20 random fields per coverslip were captured for each treatment and the plaque area was measured using Nikon NIS Elements software. All assays were carried out in triplicate.

Biopanning a phage–scFv display library to isolate T. gondii-specific antibodies

Mice were immunized with T. gondii tachyzoites over a period of 3 weeks and were found to be seropositive for infection by immunoblot (Fig. S1). The mice were sacrificed to harvest the spleens, and scFv antibody fragments were constructed by amplifying V_H and V_L gene fragments from the isolated splenic mRNA and then fusing the V_H and V_L regions by SOE–PCR with an intervening seven-amino-acid flexible linker sequence (GGSSRSS) (Fig. S2). At least 10 separate rounds of PCR were carried out during the construction of the scFv fragments to increase library size and diversity.

The scFv fragments were cloned into the phagemid vector pCANTAB5E and expressed as a phage display scFv library with a resulting complexity of 1.62 × 10⁴ independent transformants. Antibodies with appropriate binding activities were selected by a single round of solution-phase biopanning with T. gondii tachyzoites. Multiple rounds of biopanning were undesirable because this resulted in the enrichment of mostly truncated antibody fragments with stop codons in the variable regions (Figs. S3–S7).

The polyclonal biopanned scFv library had a sixfold higher binding titer to the parasite cells than the un-panned scFv library, indicating enrichment for T. gondii-binding scFvs. Biopanning resulted in the selection of four different antibody fragments with T. gondii binding activity, namely scFvs TG64 (GenBank ID: JN104603), TG69 (GenBank ID: JN104604), TG116 (GenBank ID: JN104605), and TG130 (GenBank ID: JN104602). The corresponding amino acid sequences were analyzed using IMGT V-Quest (imgt.cines.fr/textes/vquest) (Table S2). Antibody germline alignments are provided in Fig. S8. Based on this analysis, the divergence of the selected variable domain sequences from germline counterparts is summarized in Table 1.

Sequence divergence from the germline repertoire reflects somatic hypermutation events in B cells during affinity maturation (MacLennan, 1994; Li et al., 2004) and indirectly indicates high affinity and/or efficacy. V-Quest analysis showed that scFv TG130 contained the most amino acid point mutations in the CDRs, that is, six residues in total (V_H and V_L chains), most of which were localized in CDR3 (Table 1), which usually contributes the most significant contacts in antibody–antigen interactions. In contrast, scFvs TG64, TG69, and TG116 contained fewer deviations from germline sequences (TG64 = 0, TG69 = 2, TG116 = 2) potentially indicating a lower degree of affinity maturation. The sequences were then analyzed using SAROTUP (http://immunet.cn/sarotup/) to exclude false-positive antibody fragments (Huang et al., 2010). This indicated that scFvs TG64, TG69, and TG116 (but not TG130) contained TUP motifs (Table S3).

Phage rescue was used to produce recombinant phage particles fused to the selected scFv clones for a monoclonal biopanning assay. Each of the scFv–phage clones was titered before and after biopanning (input and output titers, respectively) by eluting the scFvs from tachyzoites to determine their recovery (Fig. 1). Beginning with an average phage–scFv input titer of 10¹¹ cfu/ml for each clone, we found that scFv TG116 had the highest T. gondii binding titer (5.62 × 10⁵ cfu/ml) and scFv TG130 the lowest (2.05 × 10⁵ cfu/ml). However, statistical analysis revealed no significant difference in the ability of three of the four isolated scFvs to bind T. gondii and HFF cells (p = 0.05, n = 12) despite initial promising binding titers: P_TG64 = 0.44, P_TG69 = 0.11, and P_TG116 = 0.50), indicating that these antibodies were not specific for the parasite target.

The remaining antibody (scFv TG130) was more promising and showed significant T. gondii binding specificity (p = 0.05, n = 12) with a fivefold higher binding titer for the parasite compared to HFF (P_TG130 = 0.0303; Fig. 1). Antibody specificity should take precedence over titers as an indication of bona fide antigen-specific binding. These results are consistent with our bioinformatic analysis showing that scFvs TG64, TG69, and TG116 contained TUPs, were less mature and more likely to be nonspecific. In contrast, scFv TG130 combined unique antibody CDR sequences with high target specificity, justifying its further characterization and functional validation.

Structural analysis of scFv TG130

We analyzed the structure of scFv TG130 using the Rosetta Antibody: Structure Prediction Server (Sircar, Kim & Gray, 2009) which allowed the generation of a three-dimensional molecular model for TG130 and its germline counterpart. The structural divergence of scFv TG130 from its germline counterpart was then investigated by molecular superimposition. Structural alignment of superimposed ribbon diagrams representing the germline and scFv TG130 V-regions showed significant structural changes at the antigen-combining site (Fig. 2). Although the CDR-L3 region of scFv TG130 had twice as many mutations (four substitutions) as the CDR-H3 region (two substitutions), molecular superimposition revealed that the greatest structural deviation was located within the heavy chain (HC) CDR-H3 loop rather than CDR-L3. The structural alignment showed that the greatest deviation point for the V_L chain was the L3 Thr⁹⁴ germline residue, which had been mutated to Tyr⁹⁴ in scFv TG130, with a root mean square deviation (RMSD) of 2.523 Å. In the V_H chain CDR H3 region, the greatest deviation point was the Ala⁹⁶ germline residue, which was mutated to Asp⁹⁶ in the scFv TG130 fragment, with a RMSD of 4.729 Å.

The molecular structure of TG130 indicated that six of the 15 somatic mutations relative to the germline counterpart (His → Tyr-L88 (CDR-L3), Tyr → Asn-L89 (CDR-L3), Thr → Tyr-L91 (CDR-L3), Pro → Tyr-L93 (CDR-L3), Ala → Asp-H100 (CDR-H3), and Trp → Gly-H101 (CDR-H3)), were fixed within the CDR-H3 or CDR-L3 loops, and thereby potentially involved in antigen binding. Nine other mutated residues were located in Vernier positions that do not make direct contact with the antigen. There was a higher percentage of amino acid residue changes within the CDRs compared to the framework regions, with 14.3% of mutations within the CDRs and 5.1% within the framework regions.

Additional in vitro affinity maturation of scFv TG130

In order to further improve the affinity of scFv TG130 for T. gondii, the RGYW hotspots motif (R = A or G, Y = C or T, W = T or A) was modified by in vitro site-directed mutagenesis. The resulting affinity-matured antibody library had a complexity of 4.0 × 10⁴ independent clones. The affinity-matured library was subjected to a further round of biopanning and 96 antibody clones eluted from the parasite cells yielded four distinct clones selected for further characterization, with Arg-Ala (RA15), Arg-Thr (RT51), Arg-Ser (RS55), and Leu-Thr (LT3) substitutions at positions 27 and 28 of V_L CDR1 (Table 2). Multiple codon usage was also observed for each of these clones, suggesting strong selection for the specific replacement residues. RA15, RT51, RS55, LT3, and the parental scFv TG130 were selected for the generation of chimeric IgG antibodies.

Table 2:

Sequences of the RGYW-mutant scFv obtained after parasite selection.

ScFv ID	Nucleotide sequences	Amino acid mutations
Parental Ab—TG130	AGT Act	Ser Thr
Alanine mutant—RA15	AGG Gct	Arg Ala
	CGC Gct
	CGG Gct
RT51	AGG Act	Arg Thr
	CGC Act
RS55	AGG Tct	Arg Ser
	CGG Tct
LT3	CTG Act	Leu Thr
	CTC Act

DOI: 10.7717/peerj.5780/table-2

Note:

Affinity maturation of scFv TG130 at RGYW hot spot motifs resulted in site-directed mutagenesis at amino acid positions 27 and 28 of V_L CDR1. The unique sequences of the mutant scFv antibodies isolated from biopanning selection against T. gondii are listed, along with the parental antibody. The mutated amino acid residues are color-coded according to its side-chain properties—Uncharged Polar (Red), Hydrophobic (Blue), Charged Polar (Green).

Generation and transient expression of T. gondii-specific chimeric IgGs

Chimeric IgG antibodies were constructed by transferring the V_H and V_L regions from the parental scFv (TG130) and the four affinity-matured derivatives (RA15, RT51, RS55, and LT3) into the binary plant expression vectors pTRAkc-HC and pTRAkc-LC, containing the human Cγ and Cκ domains, respectively. A. tumefaciens cultures separately transformed with the HC and LC constructs were co-infiltrated into N. benthamiana leaves to produce full-length IgG molecules (Fig. 3). Subsequent purification by Protein A affinity chromatography produced 32.8–72.1 μg of purified antibody per gram infiltrated fresh leaf tissue, specifically 48.6 μg/g TG130, 32.8 μg/g RA15, 45.9 μg/g RT51, 64.9 μg/g RS55, and 72.1 μg/g LT3.

The resulting antibody preparations had an average of 93% purity as estimated from SDS–PAGE and gel densitometry using Image J. Under reducing conditions, both the antibody’s HCs and LCs are visible in the eluate lane at the expected sizes of 25 and 60 kDa, respectively. With non-reducing conditions, the antibody appeared as a higher molecular weight band with the size of approximately 170 kDa, consistent with the tetrameric form of an intact IgG. Along with this tetrameric IgG, several more multimeric antibody bands were observed on the immunoblots, revealing minor degradation of the antibodies. However, the intact tetrameric IgG form showed a stronger band intensity and is validated by analytical HPLC to be the major species.

An indirect ELISA was used to quantify the binding of each recombinant antibody to T. gondii lysates, with detergent-solubilized HFF lysate and control mouse ascites (standard commercial preparation) incorporated as non-specific controls (Fig. 4). All five recombinant IgG showed significant binding to T. gondii compared to the control antigens. However, contrary to our expectations, the parental antibody IgG1 showed higher binding activity than the variants selected from the affinity-matured antibody library.

The antibodies were characterized in more detail by immunoblot analysis using both reduced and non-reduced fractions of T. gondii lysate (Fig. 5). We found that IgG TG130 (parental), RT51 and RA15 showed detectable binding to the parasite antigens in both the reduced and non-reduced fractions, whereas IgG RS55 and IgG LT3 consistently showed no detectable binding to any fractions. The lysates were fractionated by reducing SDS–PAGE and the three positive antibodies (parental TG130 plus RT51 and RA15) highlighted multiple bands, the most prominent of represented antigens of ∼60 and ∼30 kDa. Under non-reducing conditions, the same antibodies detected antigens of ∼55 and ∼60 kDa. IgG RA15 generated particularly weak binding signals.

Immunofluorescence assays (IFAs) were carried out to visualize the binding of the chimeric antibodies to parasite surface proteins in the native context (Fig. 6). Despite the variable results obtained in the ELISA and immunoblot experiments, we observed specific immunostaining for all IgGs to the surface of T. gondii tachyzoites (Fig. 6). This ambiguity across different immunoassays may reflect differences in the abundance and the folding of the parasite antigens. We used lysed, detergent-solubilized parasite antigens for the ELISA and immunoblots, but native antigens present on intact cells for the IFAs. The negative controls yielded negligible binding signals in all experiments as anticipated.

Inhibition of parasite invasion by recombinant IgGs

Finally, we investigated the ability of the chimeric antibodies to inhibit the invasion of human fibroblasts using the red/green invasion assay (Figs. 7A and 7B). We found that three of the chimeric IgGs caused a significant decline in parasite invasion efficacy relative to untreated controls, namely 75% for parental TG130, 69% for RT51, and 72% for RS55, whereas IgG RA15 showed no significant inhibition of invasion (Fig. 7C).

In order to determine whether the inhibition of parasite invasion reflected an impairment of parasite binding to host cells, we measured the ability of IgG-treated parasites to bind host cells fixed in paraformaldehyde, which renders them more rigid and resistant to invasion. With the exception of RA15, all treated parasites showed a partial (35–67%) reduction in host cell attachment compared to the untreated controls, but the attachment scores differed significantly only for IgG RS55 (48.5% inhibition) as shown in Fig. 7D. IgG RA15 showed little improvement in the ability to block parasite attachment compared to the untreated control (8.3% inhibition).

A plaque assay was used for further validation of the effect of chimeric IgGs on invasion, because this assay recapitulates all the steps of the lytic cycle. Remarkably, parental IgG TG130 and RT51 achieved an 87% reduction in mean plaque area compared to untreated controls (Fig. 7G). The parasites treated with these antibodies formed substantially smaller plaques compared to controls even at a low IgG concentration of 2.5 μg/ml (Figs. 7E and 7F). Collectively, these results demonstrate that plant-derived IgGs, particularly TG130 (parental), and RT51 are highly efficacious ligands to inhibit the invasion of host cells by T. gondii in vitro.