Human scFvs That Counteract Bioactivities of Staphylococcus aureus TSST-1

Some Staphylococcus aureus isolates produced toxic shock syndrome toxin-1 (TSST-1) which is a pyrogenic toxin superantigen (PTSAg). The toxin activates a large fraction of peripheral blood T lymphocytes causing the cells to proliferate and release massive amounts of pro-inflammatory cytokines leading to a life-threatening multisystem disorder: toxic shock syndrome (TSS). PTSAg-mediated-T cell stimulation circumvents the conventional antigenic peptide presentation to T cell receptor (TCR) by the antigen-presenting cell (APC). Instead, intact PTSAg binds directly to MHC-II molecule outside peptide binding cleft and simultaneously cross-links TCR-Vβ region. Currently, there is neither specific TSS treatment nor drug that directly inactivates TSST-1. In this study, human single chain antibodies (HuscFvs) that bound to and neutralized bioactivities of the TSST-1 were generated using phage display technology. Three E. coli clones transfected with TSST-1-bound phages fished-out from the human scFv library using recombinant TSST-1 as bait expressed TSST-1-bound-HuscFvs that inhibited the TSST-1-mediated T cell activation and pro-inflammatory cytokine gene expressions and productions.Computerized simulation, verified by mutations of the residues of HuscFv complementarity determining regions (CDRs),predicted to involve in target binding indicated that the HuscFvs formed interface contact with the toxin residues important for immunopathogenesis. The HuscFvs have high potential for future therapeutic application.


Introduction
Superantigens (SAgs) are proteins produced by some bacterial and viral strains that mediate T cell activation by bypassing the conventional peptide-MHC-II presentation to T cell receptor (TCR) [1].Instead, intact (unprocessed) SAgs bind directly to MHC-II molecules on the antigen presenting cells (APC)and simultaneously cross-link TCR-Vβ domains shared by about 5%-20% of circulating CD4 + Toxins 2017, 9, 50 2 of 17 and CD8 + T lymphocytes [1,2].T cell stimulation by SAgs is Lck pathway-independent [3], initiated at the Gα11 (a membrane raft-enriched heterotrimeric G-protein) that stimulates PLCβ and PKC to activate mitogen-activated protein kinases (ERKs) causing nuclear translocation of NF-AT and NF-κB and cytokine gene expressions [3].Massive amounts of cytokines including IL-1β, IL-2, IL-6,TNFα and IFNγ are released from the activated cells into the circulation [1] leading to high fever, rash, skin desquamation (peeling), plasma leakage, obstinate hypotension, and life-threatening multisystem organ failure called toxic shock syndrome (TSS) [2,4].SAgs potentiate host sensitivity to bacterial endotoxin resulting in TNFα-mediated capillary leakage which is the major contributor of the TSS [5,6].
TSS management includes supportive and symptomatic treatment.Antimicrobials and surgical debridement to remove the toxin-producing microorganisms are important.Maintaining blood pressure by fluid therapy is necessary [22,23].Intravenous immunoglobulin (IVIG) confers some benefit to the patients [24].Murine monoclonal antibodies that neutralized endotoxin prevented rabbits from lethal TSS and endotoxin challenge [6].Symptom severity of TSS was mitigated in a rabbit model after giving a mouse monoclonal antibody that neutralized TSST-1 activities [25].Rabbit polyclonal antisera against wild type and TSST-1 mutants (G31R and H135A which affected MHC-II and TCR bindings) protected rabbits even when given late in the course of the TSST-1 challenge [26].In this study, human monoclonal single chain antibodies (HuscFvs) that bound to functionally important residues of TSST-1 were produced by phage display technology.HuscFvs of three phage-transformed Escherichia coli clones inhibited TSST-1 mitogenicity (activation of T cell proliferation) and pyrogenicity (induction of pro-inflammatory cytokine gene expressions and the cytokine secretions).The human scFvs have high potential for testing further as a safe, direct acting anti-TSST-1 remedy.
The rTSST-1 at 10, 100, and 1000 ng/mL were tested for pyrogenicity (ability to induce pro-inflammatory cytokine gene expressions in stimulated cells).Figure 3 shows fold-increase of IL-1β, IL-6, and TNFαgene expressions, respectively, in the human PBMCs after stimulation with rTSST-1 and PHA (positive control) in comparison with non-stimulated cells.Pyrogenicity of the rTSST-1 was not concentration dependent which was conformed to the results reported previously [28].
The rTSST-1 at 10, 100, and 1000 ng/mL were tested for pyrogenicity (ability to induce pro-inflammatory cytokine gene expressions in stimulated cells).Figure 3shows fold-increase of IL-1β, IL-6, and TNFαgene expressions, respectively, in the human PBMCs after stimulation with rTSST-1 and PHA (positive control) in comparison with non-stimulated cells.Pyrogenicity of the rTSST-1 was not concentration dependent which was conformed to the results reported previously [28].To exclude the effect of contaminated LPS in the TSST-1 preparation, an experiment in which 1 µg/mL of LPS (Sigma-Aldrich, St. Louis, MO, USA) was used to stimulate the PBMCs was performed.It was found that the LPS induced only 28% of the cell proliferation compared to 100% stimulation by 1000 ng/mL of TSST-1 (Figure S2).Overall results indicated that the bacterially derived-rTSST-1 acquired the inherent activities of the native counterpart.Thus, the active protein was used further.

Production of HuscFvs
Ten colonies of the HB2151 E. coli transfected with rTSST-1-bound phages derived from the human scFv phage display library by means of phage bio-panning using the rTSST-1 as bait (see Materials and Methods) revealed amplicons of HuscFv coding genes (huscfv; ~1000 bp) (Figure 1C).Only eight clones produced soluble HuscFvs that bound to SDS-PAGE-separated-rTSST-1 (Figure 1D).Even though the HuscFvs bound to the rTSST-1 which was prepared in buffer containing reducing agents (SDS and mercaptoethanol) in the Western blotting, conclusion cannot be made at this stage that the rTSST-1 epitopes are linear sequences [29].After nucleotide sequencing, three clones (nos.35, 53, and 56) showed complete HuscFv sequences, i.e., contiguous sequence coding for IgVH, peptide linker (Gly4Ser1)3, and IgVL.Therefore, huscfvs of these three clones were subcloned into pLATE52 vector and the recombinant vector was used to transform NiCo21 (DE3) E. coli for large scale production of the antibodies.Inclusion bodies were purified from homogenates of the E. coli grown under IPTG induction condition and the HuscFvs were refolded.Figure 1E shows SDS-PAGE-separated purified and refolded HuscFvs.The refolded antibodies retained their binding specificity to the TSST-1 coated on the ELISA well surface with and without BSA as To exclude the effect of contaminated LPS in the TSST-1 preparation, an experiment in which 1 µg/mL of LPS (Sigma-Aldrich, St. Louis, MO, USA) was used to stimulate the PBMCs was performed.It was found that the LPS induced only 28% of the cell proliferation compared to 100% stimulation by 1000 ng/mL of TSST-1 (Figure S2).Overall results indicated that the bacterially derived-rTSST-1 acquired the inherent activities of the native counterpart.Thus, the active protein was used further.

Production of HuscFvs
Ten colonies of the HB2151 E. coli transfected with rTSST-1-bound phages derived from the human scFv phage display library by means of phage bio-panning using the rTSST-1 as bait (see Materials and Methods) revealed amplicons of HuscFv coding genes (huscfv; ~1000 bp) (Figure 1C).Only eight clones produced soluble HuscFvs that bound to SDS-PAGE-separated-rTSST-1 (Figure 1D).Even though the HuscFvs bound to the rTSST-1 which was prepared in buffer containing reducing agents (SDS and mercaptoethanol) in the Western blotting, conclusion cannot be made at this stage that the rTSST-1 epitopes are linear sequences [29].After nucleotide sequencing, three clones (nos.35, 53, and 56) showed complete HuscFv sequences, i.e., contiguous sequence coding for IgVH, peptide linker (Gly 4 Ser 1 ) 3 , and IgVL.Therefore, huscfvs of these three clones were subcloned into pLATE52 vector and the recombinant vector was used to transform NiCo21 (DE3) E. coli for large scale production of the antibodies.Inclusion bodies were purified from homogenates of the E. coli grown under IPTG induction condition and the HuscFvs were refolded.Figure 1E shows SDS-PAGE-separated purified and refolded HuscFvs.The refolded antibodies retained their binding specificity to the TSST-1 coated on the ELISA well surface with and without BSA as determined by indirect ELISA (Figure S3); indicating that the antibodies were refolded properly.The HuscFvs also bound to S. aureus enterotoxin A (SEA) (Figure S3).

Presumptive Residues and Regions of TSST-1 Bound by the HuscFvs
TSST-1 3D structure (PDB 2IJO) and modeled HuscFvs (Table S1) were subjected to intermolecular docking.Interactive modes: salt-bridge, hydrogen, hydrophobic, and van der Waals, were selected from molecular dynamic results.The lowest energy scores for TSST-1 complexed with HuscFv35, HuscFv53, and HuscFv56 were −304± 75.4,−243.3±23.7, and −371.6±15.3 kcal/mol, respectively.Tables 2-4, Figure S4A-C, and Figure S5 show TSST-1 residues and motives that were predicted to form contact interface with HuscFv35, HuscFv53, and HuscFv56.Epitopes of the HuscFv35 and HuscFv56 tend to be conformational, i.e., formed by residues in the separated portions of the TSST-1 molecule that are spatially juxtapose upon the active protein folding, while residues that formed predicted HuscFv53 epitope located mainly between β4 and β5 of the TSST-1 N-terminal domain,suggesting that the epitope might be linear (Figure S5).
Table 2. TSST-1 residues and motives predicted to be bound by residues and domains ofHuscFv35.After residues labeled in red were mutated to alanines, the HuscFv35 lost ability to suppress TSST-1 mitogenicity and pyrogenicity.
Table 3. TSST-1 residues and motives predicted to be bound by residues and domains of HuscFv53.After residues labeled in red were mutated to alanines, the HuscFv53 lost ability to suppress TSST-1 mitogenicity and pyrogenicity.
TSST-1-exposed-human PBMCs added with the HuscFvs had markedly reduced cytokine gene expressions (Figure 5) and the respective cytokine levels (Figure 6), compared to the control HuscFv-treated and non-treated cells (p< 0.05).Both HuscFv35 and HuscFv56 performed better than the HuscFv53.Noantibodies had an effect on the PHA-exposed cells, indicating that the HuscFv inhibitory effect on the TSST-1 pyrogenicity was target specific.The data obtained from PBMCs exposed to rTSST-1 and PHA of Figure 5 do not fit with the data of Figure 3.The reason should be that the experiments were performed on blood samples taken a few months apart, although from the same blood donor.However data of duplicate experiments performed on blood samples taken from the blood donor one or two days apart were not statistically different, as shown by the small error bars of both Figures.
Tables 2, 3 and 4 provide information on the amino acids, their positions, CDRs, and domains of the scFvs that have been predicted to involve in target binding.In order to demonstrate the relevance of the predicted HuscFv residues that formed interface contact with the TSST-1, many of the HuscFv residues which their side chains interacted with TSST-1 were substituted by alanines (marked red in Tables 2, 3 and 4  (1.7%), and HuscFv56 (0.9%).The control HuscFv had some inhibitory activity (placebo effect) on the TSST-1-activated cells (6.5%).The TSST-1-specific-HuscFvs did not affect the PHA-stimulated cells indicating their target specificity.
TSST-1-exposed-human PBMCs added with the HuscFvs had markedly reduced cytokine gene expressions (Figure 5) and the respective cytokine levels (Figure 6), compared to the control HuscFv-treated and non-treated cells (p < 0.05).Both HuscFv35 and HuscFv56 performed better than the HuscFv53.Noantibodies had an effect on the PHA-exposed cells, indicating that the HuscFv inhibitory effect on the TSST-1 pyrogenicity was target specific.The data obtained from PBMCs exposed to rTSST-1 and PHA of Figure 5 do not fit with the data of Figure 3.The reason should be that the experiments were performed on blood samples taken a few months apart, although from the same blood donor.However data of duplicate experiments performed on blood samples taken from the blood donor one or two days apart were not statistically different, as shown by the small error bars of both Figures.
Tables 2-4 provide information on the amino acids, their positions, CDRs, and domains of the scFvs that have been predicted to involve in target binding.In order to demonstrate the relevance of the predicted HuscFv residues that formed interface contact with the TSST-1, many of the HuscFv residues which their side chains interacted with TSST-1 were substituted by alanines (marked red in Tables 2-4).The mutated residues for HuscFv35 were Y27A of VH-FR1; D31A of VH-CDR1; H100A, Q101A, and D108A of VH-CDR3; T165, N166A and Y168A of VL-CDR1; Y185A of VL-FR2; and T192A of VL-FR3; for HuscFv53 were R31A of VH-CDR1; T52A and D57A of VH-CDR2; Y168A of VL-CDR1; and K186 of VL-CDR2; and for HuscFv56 were S31A of VH-CDR1; E55A and E59A of VH-CDR2; Y101A, Y102A, and R104 of VH-CDR3; Y165A of VL-CDR1; Y182 of VL-FR2; N186 of VL-CDR2; and S225 and Y229 of VL-CDR3.The inhibitory activities of the wild type HuscFvs on the TSST-1-mediated cell proliferation and pro-inflammatory cytokine production were abrogated after the residue mutations, i.e., the mHuscFvs could not reduce mitogenicity and pyrogenicity of the TSST-1, as shown in Figures 6 and 7, respectively.
Antibodies of heterologous source have been shown to mitigate symptom severity and rescued animals from the TSS-mediated lethality [30,31].Treatment of human TSS cases is usually performed in the intensive care unit and includes supportive and symptomatic measures, removal of bacteria producing the causative toxin as well as infusion of IVIG thought to contain antibodies to bacterial endotoxin.However, passive immunization and immunotherapy by using homologous (human) antibodies directed to the TSST-1 functionally critical residues has never been performed.Human scFvshave potential applications for immunotherapy of diseases [38][39][40].Thus, the fully human scFvs produced in this study, especially the HuscFv35 and HuscFv56 have high potential for testing further as a safe, direct acting anti-TSST-1 remedy.VL-FR2; N186 of VL-CDR2; and S225 and Y229 of VL-CDR3.The inhibitory activities of the wild type HuscFvs on the TSST-1-mediated cell proliferation and pro-inflammatory cytokine production were abrogated after the residue mutations, i.e., the mHuscFvs could not reduce mitogenicity and pyrogenicity of the TSST-1, as shown in Figure 7 and Figure 6, respectively.Antibodies of heterologous source have been shown to mitigate symptom severity and rescued animals from the TSS-mediated lethality [30,31].Treatment of human TSS cases is usually performed in the intensive care unit and includes supportive and symptomatic measures, removal of bacteria producing the causative toxin as well as infusion of IVIG thought to contain antibodies to bacterial endotoxin.However, passive immunization and immunotherapy by using homologous (human) antibodies directed to the TSST-1 functionally critical residues has never been performed.Human scFvshave potential applications for immunotherapy of diseases [38][39][40].Thus, the fully human scFvs produced in this study, especially the HuscFv35 and HuscFv56 have high potential for testing further as a safe, direct acting anti-TSST-1 remedy.

Recombinant TSST-1 (rTSST-1) Production
TSST-1 gene was retrieved from GenBank no.J02615.Synthetic TSST-1 gene sequence with stop codondeletion and BamH1 and HindIII restriction sites incorporation at the 5′ and 3′ ends, respectively, was inserted into pET21a + DNA (GenScript).The recombinant plasmid was used to transform NiCo21 (DE3) E. coli by means of a highly efficient transformation protocol (New England Biolabs, UK).Appropriately transformed E. coli colony was grown in LB-A broth containing 1 mM IPTG and the 6× His-tagged-rTSST-1 was purified from the bacterial lysate by using Ni-NTA resin (Invitrogen, Waltham, MA, USA).

Mitogenic and Pyrogenic Activities of rTSST-1
Because activated T cells expressed surface CD69 molecules [41]; thus, mitogenicity testing of the rTSST-1 was performed by detecting percentages of CD3 + CD69 + in human PBMCs after exposure to the toxin and controls.Human PBMCs (3× 10 5 cells/well) were cultured in 48-well round-bottom tissue culture plate (Corning) in RPMI-1640 medium (Gibco TM ) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 units/mL penicillin, and 100 µg/mL streptomycin (complete medium) at 37°C in 5% CO2 atmosphere.Various concentrations of rTSST-1 (10-1000 ng/mL) were added appropriately to the cells.Positive control was cells stimulated with 1000 ng/mL phytohemagglutinin (PHA) (Sigma) which is a pan T cell mitogen and negative control was

Recombinant TSST-1 (rTSST-1) Production
TSST-1 gene was retrieved from GenBank no.J02615.Synthetic TSST-1 gene sequence with stop codondeletion and BamH1 and HindIII restriction sites incorporation at the 5 and 3 ends, respectively, was inserted into pET21a + DNA (GenScript).The recombinant plasmid was used to transform NiCo21 (DE3) E. coli by means of a highly efficient transformation protocol (New England Biolabs, UK).Appropriately transformed E. coli colony was grown in LB-A broth containing 1 mM IPTG and the 6× His-tagged-rTSST-1 was purified from the bacterial lysate by using Ni-NTA resin (Invitrogen, Waltham, MA, USA).

Mitogenic and Pyrogenic Activities of rTSST-1
Because activated T cells expressed surface CD69 molecules [41]; thus, mitogenicity testing of the rTSST-1 was performed by detecting percentages of CD3 + CD69 + in human PBMCs after exposure to the toxin and controls.Human PBMCs (3× 10 5 cells/well) were cultured in 48-well round-bottom tissue culture plate (Corning) in RPMI-1640 medium (Gibco TM ) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 units/mL penicillin, and 100 µg/mL streptomycin (complete medium) at 37 • C in 5% CO 2 atmosphere.Various concentrations of rTSST-1 (10-1000 ng/mL) were added appropriately to the cells.Positive control was cells stimulated with 1000 ng/mL phytohemagglutinin (PHA) (Sigma) which is a pan T cell mitogen and negative control was cells in culture medium alone.After 24 h, cells were washed twice with cell washing/blocking reagent (1% heat-inactivated normal serum in PBS), re-suspended in fresh culture medium, added with anti-CD3-PerCP and anti-CD69-PE, and subjected to FACScan flow cytometry with BD Diva software for data acquisition and analysis.Lymphocyte population was identified using forward and side scattered property.T cell population was identified by cells that were CD3 + .The percentage of activated T cells (CD3 + CD69 + ) was determined from an upper-right quadrant (Q2) where as Q4 was CD3 + cells.The Q1 and Q3 were non-activated cell population (CD3 -CD69 -).Results are expressed as percentages of CD3 + CD69 + cells.
For rTSST-1-mediated cell proliferation, human PBMCs were incubated with violet CellTrace TM for 20 min in complete medium before stimulating with rTSST-1 as above.Similar controls were included.After 72 h, cells were washed, stained with anti-CD3-PerCP, and subjecting to FACScan flow cytometry.Results were expressed as percentages of violet/pacific blue stained-CD3 + cells.
Pyrogenicity of rTSST-1 was tested.Human PBMCs (1× 10 5 cells/well) were cultured and stimulated with various concentrations of rTSST-1 as above.Similar controls were included.After 24 h, total RNAs were extracted from cells in individual wells.Expressions of pro-inflammatory cytokine genes including IL-1β, IL-6, and TNFα were determined by quantitative real-time RT-PCR (qRT-PCR) using primers listed in Table S2.One microliter cDNA(50 ng) and 100 nM each PCR primer was put in SYBR Green Master Mix (Applied Biosystems) and subjected to PCR reaction: 95 • C, 10 min then 36 cycles of denaturation at 95 • C for 30 s, annealing at 60 • C for 1 min, extension at 72 • C for 30 s, and hold at 72 • C for 5 min in Strategene Mx3005P QPCR System (Agilent Technologies).Data were analyzed using MxPro QPCR software.β-actin gene was used for normalization.Levels of the pro-inflammatory cytokines in cell culture supernatants were determined by using ELISA kit (Thermo Fisher Scientific, Waltham, MA, USA).

Production of TSST-1-bound HuscFvs
Human scFv phage display library used in this study was constructed previously [42].Briefly, cDNAs were prepared from mRNAs of peripheral blood lymphocytes of multiple human blood donors and used as templates for amplification of immunoglobulin VH and Vκ coding sequences by PCR.The oligonucleotide primers used for the PCR were human degenerate primers designed from all families of human immunoglobulin variable sequences [42].The PCR amplified vh and vl sequences were ligated randomly via a polynucleotide linker (coding for (Gly 4 Ser) 3 ) to generate a repertoire of vh-linker-vl sequences or scfv sequences.The scfvs were ligated with pCANTAB5E phagemid DNAs and the recombinant phagemids were used to transfect TG1 E. coli.After growing the recombinant phagemid-transformed E. coli in the presence of helper phage (M13KO7), complete phage particles which displayed human scFvs as fusion proteins with the phage coat protein (p3) and also carried the respective scfvs in the phage genomes could be obtained from the E. coli culture supernatant.
HuscFv-displayed phage clones that bound to the rTSST-1 were fished-out from the library using the recombinant protein as bait in the biopanning process [42].The phage library was added to an ELISA well pre-coated with 1 µg of purified rTSST-1 and the plate was incubated at 37 • C for 1 h.Unbound phages were removed by washing with buffer and a log phase-grown HB2151 E. coli culture was added to the well containing the antigen-bound phages.The phage transformed bacterial coloniesthat grew on selective agar plates after overnight incubation were screened for the HuscFv genes (huscfvs) by PCR [42].The huscfv-positive clones were grown in 0.2 mMIPTG-conditioned broth to induce HuscFv expressions.Binding of soluble HuscFvs in the bacterial lysates to the SDS-PAGE-separated-rTSST-1 were tested by Western blotting.Nucleotides of the huscfvs coding for rTSST-1-bound-HuscFvs were sequenced, deduced, and canonical complementarity determining regions (CDRs) and immunoglobulin framework regions (FRs) were determined using the IMGT ® Information System [43].

Computerized Simulation for Determining Interactive Residues between TSST-1 and HuscFvs
TSST-1 3D structure was retrieved from RCSB PDB 2IJO.The huscfv 3D structures were modeled by the I-TASSER server [44,45].The I-TASSER-predicted structures were further refined [46,47] and improved to near native states on the automated ClusPro 2.0 antibody-protein docking server.The models from the docking were simulated with NAMD Molecular Dynamics [48].The TSST-1-HuscFv complexes were built and visualized by using PyMol software (PyMol Molecular Graphics System, Version 2 edu, Schrodinger, LLC).

Preparation of Mutated-HuscFvs (mHuscFvs)
Gene sequences coding for HuscFvs which side chains of their residues interacted with TSST-1 (data from molecular dynamics) were substituted by alanines and synthesized (Integrated DNA Technologies, Coralville, IA, USA).The DNA fragments were cloned into pLATE52 and the recombinant vector was used to transform NiCo21 (DE3) E. coli.The HuscFvs were prepared from appropriately transformed E. coli as for the wild type HuscFvs.The mutated HuscFvs (mHuscFvs) were tested for their ability to inhibit TSST-1 activities (mitogenicity and pyrogenicity).

HuscFvs-mediated Inhibition of TSST-1 Activities
For inhibition of TSST-1-mediated T cell activation by the HuscFvs, human PBMCs were added with mixture of rTSST-1 and HuscFvs/mHuscFvs or control HuscFv and kept for 24 h.TSST-1-stimulated cells without antibody treatment, PHA-stimulated cells treated similarly with TSSTS-1-bound-HuscFvs, and normal cells in medium were included in the experiment.After washing, cells were labeled with CD3-PerCP and CD69-PE and analyzed by FACScan Flow cytometry.Total viable lymphocytes were gated by SSC and FSC and CD3 + cells were gated for CD69 + cells.Results were expressed as percentages of CD3 + CD69 + cells.
For inhibiting rTSST-1 pyrogenicity by the HuscFvs/mHuscFv, human PBMCs (5× 10 4 cells/well) in complete medium were added with 1000 ng/mL rTSST-1.The TSST-1-stimulated cells were treated either with TSST-1-bound-HuscFvs/mHuscFv, control/irrelevant HuscFv (did not bind to TSST-1), or medium alone.The antibody:TSST-1 molar ratios were 4:1 (optimal from titration).Triplicate wells were set for each treatment.Cells added with 1000 ng/mL PHA with and without HuscFv-treatments, and cells in medium alone (normal cells) were included.After 24 h, total RNAs were extracted from cells in individual wells and quantified by using NanoDrop instrument.Complementary DNAs were synthesized (RevertAid First Strand cDNA Synthesis kit) and used as templates for quantification of mRNAs of pro-inflammatory cytokines including TNF-α, IL-1B, and IL-6.The quantitative real-time PCR primers for the mRNA quantification are listed in Table S2.β-actin gene was used for normalization.Moreover, cell culture supernatants in all wells were collected and the levels of the pro-inflammatory cytokines were measured using commercial ELISA kits (Thermo Fisher Scientific).Results are the average of the two reproducible experiments.

Statistical Analysis
One way ANOVA followed by post hoc comparison using least significant difference (LSD) and independent t-test were performed for data comparison using SPSS 18.0 statistical software.Significant difference was p < 0.05.

Conclusions
Human single chain antibodies (HuscFvs) to S. aureus TSST-1 that inhibited the TSST-1-mediated T cell activation and pro-inflammatory cytokine gene expressions and productions were generated.The HuscFvs formed interface contact with the TSST-1 residues important for immunopathogenesis of toxic shock syndrome.The HuscFvs have high potential for testing further as a direct acting anti-TSST-1 agent for future clinical use.

Figure 3 .
Figure 3. Pyrogenicity of the rTSST-1 (ability to induce stimulated cells to express pro-inflammatory cytokine genes).Fold increase of IL-1β, IL-6, and TNFα expressions in the human PBMCs after stimulation with rTSST-1 at 10, 100, and 1000 ng/mL and 1000 ng/mL PHA (positive control) in relation to non-stimulated cells (negative control).

Figure 3 .
Figure 3. Pyrogenicity of the rTSST-1 (ability to induce stimulated cells to express pro-inflammatory cytokine genes).Fold increase of IL-1β, IL-6, and TNFα expressions in the human PBMCs after stimulation with rTSST-1 at 10, 100, and 1000 ng/mL and 1000 ng/mL PHA (positive control) in relation to non-stimulated cells (negative control).

Table 1 .
LC-MS/MS Mascot results of peptides generated from recombinant TSST-1 of this study with 30% sequence coverage.

Table 1 .
LC-MS/MS Mascot results of peptides generated from recombinant TSST-1 of this study with 30% sequence coverage.

Table 4 .
TSST-1 residues and motives predicted to be bound by residues and domains ofHuscFv56.After residues labeled in red were mutated to alanines, the HuscFv56 lost ability to suppress TSST-1 mitogenicity and pyrogenicity.