INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

The staphylococcal superantigen toxic shock syndrome toxin 1 (TSST-1) is implicated as the major cause of toxic shock syndrome (TSS) (3, 45). In contrast to conventional antigens, superantigens bind to a nonpolymorphic region of major histocompatibility complex class II (MHC-II) molecules on antigen-presenting cells (APCs) outside the peptide groove and do not require prior processing for presentation to the T-cell receptor (TCR) (19). The MHC-superantigen complex binds to a relatively nonpolymorphic region of the T-cell receptor (TCR) bearing specific V determinants (for example, TSST-1 activates only those T cells bearing V2-TCR in human peripheral blood mononuclear cells [PBMC]) (18). This trimolecular interaction between the superantigen, MHC-II, and the TCR leads to massive proliferation of T cells and uncontrolled release of proinflammatory cytokines including interleukin 1 (IL-1), IL-2, gamma interferon (IFN-), tumor necrosis factor (TNF), and others (17, 23, 38). It is believed that the massive release of these proinflammatory cytokines, particularly TNF- and IFN-, is the key factor leading to the life-threatening complications of TSS (29, 30, 40).

Cytokines are the primary modulators of the immune response, and have either proinflammatory or anti-inflammatory functions (43). They are derived either from APCs or from T-helper cells, which can be categorized into two major subsets, Th1 and Th2, based on their cytokine production profiles (34). Th1 effector cells produce predominantly proinflammatory cytokines such as IFN-, IL-2, TNF-, and TNF-, which are associated with cell-mediated immunity. Th2 effector cells produce largely anti-inflammatory cytokines such as IL-4, IL-5, IL-6, IL-10, and IL-13, which are associated with humoral immunity (1, 33). Both T-cell subsets are capable of cross-regulating and suppressing each other through a complicated network of cytokine-mediated signaling (1, 9, 37). For example, IFN- produced by Th1 cells inhibits the development of Th2 cells (11), whereas IL-4 and IL-10 produced by Th2 cells inhibit Th1 development (33, 47). IL-6 may possess both proinflammatory and anti-inflammatory effects depending on the particular model system being studied (14). There is evidence for the presence and activity of each of these Th subsets during the immune response to bacterial superantigens (9, 12, 30, 31, 50). In the present study, we sequentially monitored the Th1 and Th2 cytokine profiles following TSST-1 stimulation of human PBMC in vitro. To further clarify the possible mechanism of TSST-1-induced cytokine production, a panel of loss-of-function, single-amino-acid-substitution TSST-1 mutant toxins were studied in parallel. These mutant toxins include G31R, which was previously found to be defective in MHC-II binding (24), as well as S14N (W. W. S. Kum, R. W. Y. Hung, S. B. Cameron, and A. W. Chow, submitted for publication) and H135A (6, 7), which were previously found to be defective in TCR interaction.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Purification of WT rTSST-1 and mutant toxins. Wild-type recombinant TSST-1 (WT rTSST-1) and mutant toxins S14N, G31R, and H135A were obtained by random and site-directed mutagenesis as described previously (24). WT and mutant toxin genes were transformed into Staphylococcus aureus RN4220, and expressed toxins were purified from lipopolysaccharide (LPS)-free culture supernatants using a combination of preparative isoelectric focusing and chromatofocusing as reported previously (23). Toxin purity was assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 14% acrylamide gel and silver staining, and LPS activity was monitored by the Limulus amoebocyte lysate test (sensitivity limit, 10 pg per ml) (23).

Preparation of human PBMC and culture conditions. Fresh human PBMC from random healthy adult donors were obtained by centrifugation of leukopheresis packs over Histopaque 1.077 (Pharmacia Fine Chemicals, Dorval, Quebec, Canada) as described previously (46). Cells at the interface were washed three times in Hanks' balanced salt solution (Stem Cell Technologies Inc., Vancouver, British Columbia, Canada). Approximately 2 × 10⁶ human PBMC were cultured in growth medium containing RPMI 1640 medium (Stem Cell) supplemented with 10% fetal bovine serum (heat inactivated at 56°C for 30 min; HyClone Laboratories, Inc., Logan, Utah), 2 mM L-glutamine (GIBCO BRL, Burlington, Ontario, Canada) and 20 µg of polymyxin B sulfate (to neutralize any possible LPS contamination) (Sigma Chemical Co., St. Louis, Mo.) per ml in 24-well tissue culture plates (Falcon Labware; Becton Dickinson Canada Inc., Mississauga, Ontario, Canada) and incubated at 37°C in 5% CO₂ with WT rTSST-1 or mutant toxins S14N, G31R, and H135A at a concentration of 1 nM. WT rTSST-1 and the mutant toxins S14N and G31R stimulated equivalent maximal levels of T-cell proliferation at this concentration as determined by [³H]thymidine incorporation (24). In order to study the kinetic profiles of cytokine secretion, cells were exposed to the toxins over a period of 3 days. Culture supernatants were collected at 0.5 h, at 1-h intervals from 1 to 6 h, at 2-h intervals from 8 to 24 h, at 48 h, and at 72 h. Stimulated supernatants were microcentrifuged at 800 × g for 5 min, transferred to fresh vials, and frozen at 70°C until analysis.

Cytokine assays. Culture supernatants from cells stimulated with WT rTSST-1, S14N, G31R, or H135A at various time intervals were assayed for different cytokines by enzyme-linked immunosorbent assay (ELISA), using commercial sandwich ELISA kits containing recombinant human cytokine standards, murine monoclonal capture antibodies, and biotinylated goat anti-human cytokines detecting antibodies as described previously (22, 46). The sources of various ELISA kits for the different cytokine assays (and their sensitivity limits) are as follows: (i) TNF- (62 pg/ml), TNF- (16 pg/ml), IFN- (62 pg/ml), IL-1 (125 pg/ml), IL-2 (125 pg/ml), IL-6 (20 pg/ml), and IL-12 (31 pg/ml) from R&D Systems, Minneapolis, Minn.; (ii) IL-4 (62 pg/ml), IL-10 (62 pg/ml), and transforming growth factor  (TGF-) (62 pg/ml) from PharMingen, San Diego, Calif.

Data analysis. The GraphPad PRISM version 3.0 software (GraphPad Software, Inc., San Diego, Calif.) was used for data analysis. Cytokine assays by ELISA were determined in duplicate, and data were obtained from three different donors. Time-dependent changes in cytokine levels between each toxin treatment group were assessed by one-way analysis of variance with repeated measures. Bonferroni's test was used for multiple comparisons between the different toxin treatment groups. Differences were considered significant if the 2-tailed probability of the null hypothesis was less than five percent (P < 0.05).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Toxin purity and characterization. All toxins were LPS-free and migrated as single protein bands on SDS-PAGE with approximate molecular masses of 22 kDa, and each reacted with rabbit polyclonal anti-TSST-1 antibody by Western immunoblotting (data not shown). Previously reported biologic properties of these toxins including MHC class II binding constants, maximal T-cell proliferation (22, 24), V2-specific T-cell proliferation, and TCR down-regulation in human PBMC (Kum et al., submitted) are summarized in Table 1.

Secretion profiles of the early-phase (within 3 to 4 h) cytokines. Three proinflammatory cytokines, IFN-, IL-2, and TNF-, were detected in culture supernatants within 3 to 4 h following stimulation with WT rTSST-1 (Fig. 1A to C).

View larger version (38K): [in this window] [in a new window]   FIG. 1.   Detection of proinflammatory and anti-inflammatory cytokines by ELISA in human PBMC stimulated by WT rTSST-1, and loss-of-function single-amino-acid-substitution mutant toxins G31R (MHC-II binding defective) and S14N and H135A (both TCR interaction defective). Results (means ± standard errors of the means) were from three different donors. Panels: A, IFN-; B, IL-2; C, TNF-; D, IL-1; E, IL-6; F, TNF-; G, IL-10; H, IL-12.

(i) IFN-. IFN- was induced by WT rTSST-1 very early in human PBMC, being first detectable in culture supernatants within 3 h, increasing rapidly until 8 h, and more gradually through 48 h (Fig. 1A). The MHC-II-binding-defective mutant toxin G31R demonstrated a kinetic pattern similar to that of WT rTSST-1 but with a slightly delayed onset (~5 h) and blunted response throughout the observation period. In contrast, the mutant toxins with defective TCR interaction either did not induce any IFN- secretion throughout the entire 72-h period (H135A) or demonstrated a markedly delayed response at 48 h with no detectable IFN- secretion in the early phase (S14N) (P < 0.001 compared to WT rTSST-1) (Fig. 1A).

(ii) IL-2. Human PBMC stimulated with WT rTSST-1 also exhibited a biphasic IL-2 response, which was first detectable at 3 h, with rapid increase in the first 8 h, and a more gradual increase through 48 h (Fig. 1B). G31R induced a similar secretion profile but with a delayed onset at ~6 h. Surprisingly, even though mutant toxin S14N had retained the ability to induce V2-specific T-cell proliferation in a previous study (Kum et al., submitted), it did not induce any detectable IL-2 production throughout the 72-h observation period. H135A again did not induce any cytokines, as shown by a comparison with RPMI controls.

(iii) TNF-. TNF- was first detectable at ~3 to 4 h following stimulation with WT rTSST-1, and its secretion profile also appeared to follow a biphasic course, with a rapid rate of production during the early phase until 8 h and a more gradual but sustained rate of production thereafter through 48 h (Fig. 1C). G31R induced similar levels of TNF- but with a slightly delayed onset (~8 h). The production of TNF- induced by S14N was markedly delayed (detectable only at 48 h), and quantifiable levels were markedly diminished compared to those observed with WT rTSST-1 (P < 0.001). H135A again did not induce any cytokines at any time.

Secretion profiles of the late-phase (16 to 48 h) cytokines. Most of the other cytokines induced by WT rTSST-1 in human PBMC were detected only at ~16 h or later (Fig. 1D to H). These included both proinflammatory (IL-1, TNF-, IL-12) as well as anti-inflammatory (IL-6, IL-10) cytokines. G31R induced a similar cytokine profile, except that production of IL-1 and IL-12 was markedly reduced. H135A again did not induce any cytokines at any time. With the exception of IL-10, S14N induced lower levels of all cytokines than did G31R and WT rTSST-1 during the 72-h observation period.

(i) IL-1. Secretion of the Th1 cytokine IL-1 by WT rTSST-1 in human PBMC was not observed until ~16 h, attaining peak levels at 48 h (Fig. 1D). IL-1 secretion was induced by both mutant toxins S14N and G31R, with similar kinetic profiles, although the concentrations induced by each of these were markedly reduced compared to that induced by WT rTSST-1 (P < 0.001).

(ii) IL-6. WT rTSST-1 and G31R induced similar secretion profiles of IL-6, which was first detected at ~16 h, reaching peak levels at 48 h (Fig. 1E). Significantly reduced production of IL-6 was induced by S14N compared to that induced by WT rTSST-1, being detectable only at ~20 h and with markedly diminished levels at 48 h (P < 0.001) (Fig. 1E).

(iii) TNF-. Both WT rTSST-1 and mutant toxin G31R induced the secretion of TNF-, with similar kinetic profiles, being detectable only after 24 h followed by a gradual increase in concentration at 48 to 72 h (Fig. 1F). The mutant toxin S14N induced TNF- with a kinetic profile similar to those of WT rTSST-1 and G31R, although slightly reduced in comparison.

(iv) IL-10. Secretion levels of the Th2 anti-inflammatory cytokine IL-10 in human PBMC were similar for WT rTSST-1 and G31R, being detectable only at ~48 h, followed by a gradual decline in concentration at 72 h (Fig. 1G). In contrast, the induction of IL-10 by S14N continued to increase at 72 h.

(v) IL-12. Like those of TNF- and IL-10, secretions of the proinflammatory cytokine IL-12 in human PBMC stimulated with WT rTSST-1, G31R, and S14N were not observed until ~48 h (Fig. 1H). The IL-12 responses to both G31R and S14N were markedly blunted compared to what was observed with WT rTSST-1 (P < 0.01), with G31R producing a higher level than S14N.

Absence of detectable anti-inflammatory cytokines IL-4 and TGF-. Th2 anti-inflammatory cytokines IL-4 and TGF- were not detected at any time in any of the culture supernatants stimulated with either WT rTSST-1, G31R, S14N, or H135A, even though the assays had acceptable sensitivity limits for detection (62 pg/ml for each cytokine).