Interaction of staphylococcal toxic shock syndrome toxin-1 and enterotoxin A on T cell proliferation and TNF secretion in human blood mononuclear cells

Presented in part at the International Congress of Chemotherapy, July 19, 1995, Montreal, Quebec Division of Infectious Diseases, Departments of Medicine, Microbiology and Immunology, University of British Columbia; the Canadian Bacterial Diseases Network; Vancouver Hospital and Health Sciences Centre, Vancouver, British Columbia Correspondence and reprints: Dr Anthony W Chow, Division of Infectious Diseases, Vancouver Hospital and Health Sciences Centre, 2733 Heather Street, Vancouver, British Columbia V5Z 3J5. Telephone 604-875-4148, fax 604-875-4013, e-mail tonychow@interchange.ubc.ca Received for publication August 31, 1998. Accepted February 25, 1999 ML De Boer, WWS Kum, AW Chow. Interaction of staphylococcal toxic shock syndrome toxin-1 and enterotoxin A on T cell proliferation and TNF secretion in human blood mononuclear cells. Can J Infect Dis 1999;10(6):403-409.

S taphylococcal toxic shock syndrome (TSS) is a severe multisystem disease characterized by fever, diffuse skin rash followed by desquamation, hyperemia of the mucous membranes and hypotension or shock.Although initially described primarily in menstruating women in association with tampon use, nonmenstrual cases are increasingly recognized (1).Earlier studies by the authors (2,3) and others (4,5) strongly implicated TSS toxin-1 (TSST-1) as the primary cause of TSS.Apart from TSST-1, the staphylococcal enterotoxins are also strongly linked to TSS (6).By using ELISA and DNA probes, we have examined the production of TSST-1, staphylococcal enterotoxin A (SEA), staphylococcal enterotoxin B (SEB) and staphylococcal enterotoxin C (SEC) in the culture supernatants from 350 Staphylococcus aureus isolates recovered from Vancouver cases of menstrual TSS (MTSS), nonmenstrual TSS (NMTSS), non-TSS S aureus infections and asymptomatic carriers (1,7,8).Interestingly, SEA was commonly coproduced with TSST-1, and this toxin combination was significantly more prevalent in MTSS than in other groups (75% in MTSS compared with 32% in NMTSS and 25% in non-TSS-associated S aureus infections) (7).Epidemiological typing by multilocus enzyme electrophoresis also revealed that MTSS isolates were primarily from a single S aureus clone that produced both TSST-1 and SEA (7,9).Both TSST-1 and SEA belong to a related family of superantigens that induce massive T cell proliferation via specific Vb determinants of the T cell receptor (TCR), and stimulate the release of various proinflammatory cytokines by human, murine and bovine lymphoid cells (10)(11)(12)(13).These biological effects are believed to play a central role in the pathogenesis of superantigenmediated shock (14)(15)(16).Because MTSS is associated with a predominant clone of S aureus, which produces both TSST-1 and SEA, we hypothesized that coproduction of these toxins might enhance the biological activities of either toxin alone.To test this hypothesis, we examined the interaction of purified TSST-1 and SEA, alone or in combination, on the induc-tion of mitogenic response and cytokine secretion of human peripheral blood mononuclear cells (PBMCs) from healthy donors.To our knowledge, this is the first study to examine the effects of superantigen coproduction on the induction of various biological responses believed to be important in the pathogenesis of staphylococcal TSS.

MATERIALS AND METHODS
Purification of TSST-1 and SEA: TSST-1 was purified from culture supernatants of S aureus MN8 by preparative isoelectric focusing and chromatofocusing as previously described (17).SEA was purchased from Toxin Technology (Sarasota, Florida) and further purified by chromatofocusing using a pH 6 to 8 gradient polybuffer exchanger (PBE 94; Pharmacia Fine Chemicals, Uppsala, Sweden).The purity of TSST-1 and SEA preparations was verified by the presence of a single band of approximately 22 kD and approximately 28 kD, respectively, following SDS-PAGE or immunoblotting with polyclonal antisera (17).Purification of human PBMC: Fresh human PBMCs were obtained from healthy donors by the method previously described (11).In brief, cells were obtained by centrifugation of leukopheresis packs over Histopaque 1.077 (Sigma Chemical Co, St Louis, Missouri).Mononuclear cells at the interface were washed with Hank's balanced salt solution before being separated into T cells and non-T cell populations by rosetting with sheep erythrocytes.Monocytes were purified from B lymphocytes by density gradient centrifugation over Percoll (Pharmacia Fine Chemicals, Dorval, Quebec).For isolating purified human T lymphocytes, E-rosetted cells were treated with ammonium chloride to remove sheep erythrocytes, washed three times with Hank's balanced salt solution and suspended in RPMI 1640 (StemCell Technologies Inc, Vancouver, British Columbia) supplemented with 10% heat inactivated (56°C, 30 mins) fetal calf serum (Gibco/BRL Life Technologies Inc, Gaithersburg, Maryland), 2 mM L-glutamine, 25  Mitogenicity assay: Mitogenesis of human PBMCs stimulated by TSST-1 or SEA was assessed by [ 3 H]-thymidine incorporation as previously described (17).A 1:1 ratio of T cells to monocytes (3×10 5 cells/well) in supplemented RPMI 1640 were cultured in 0.2 mL volumes with various concentrations of TSST-1, SEA, or combination for three days at 37°C, 5% carbon dioxode in 96-well round-bottom plates (Falcon Labware, Becton-Dickinson Canada Inc, Mississauga, Ontario).Cells were pulsed with 1 Ci of [ 3 H]-thymidine (6.7 Ci/mmol; ICN Flow Laboratories, Irvine, California) 18 h before completion of incubation, and harvested onto glass-fibre filter paper with an automatic harvester (Skatron, Sterling, Virginia).Samples were counted in a liquid scintillation counter (LS1800, Beckman, Mississauga, Ontario).

Cytokine assays for tumour necrosis factor-alpha, interleukin-1ß and interleukin-6:
A 1:1 ratio of T cells to monocytes (2×10 6 cells/well) in supplemented RPMI 1640 were cultured in 1.0 mL volumes in conjunction with various concentrations of TSST-1, SEA or combination at 37°C and 5% carbon dioxide in 24-well culture plates (Becton-Dickinson).
After approximately 20 h, supernatants were harvested by centrifugation at 800 g for 5 mins, and stored at -70°C until analysis.Tumour necrosis factor-alpha (TNFa ), interleukin (IL)-1b , and IL-6 present in the culture supernatants were assayed by ELISA previously developed in the authors' laboratory (17).
The sensitivity limits for detection were 250 pg/mL for TNFa , 500 pg/mL for IL-1ß, and 625 pg/mL for IL-6.
Statistical analyses: All quantitative data were expressed as means ± standard error of the mean (SEM).Differences in mitogenic responses and cytokine secretion between different groups were compared using the Student's t test for paired samples (two-tailed).P<0.05 was considered significant.

RESULTS
Minimum effective dose of TSST-1 and SEA for inducing mitogenesis in human PBMCs: PBMCs (1:1 ratio of T cells to monocytes) from one healthy donor were stimulated with various concentrations of TSST-1 and SEA (0.001 pg/mL to 100 m g/mL) to generate dose-response curves for mitogenesis (Figure 1).This was performed to select a minimum effective concentration of TSST-1 and SEA for the interaction studies.The concentration range of the dose response to TSST-1 and SEA observed in this donor was similar to that observed by other investigators (17)(18)(19)(20), although the amplitude of the proliferative responses varied from one subject to another (18).Based on these dose-response curves, 1 pg/mL of TSST-1 and 0.1 pg/mL of SEA, respectively, were chosen for further study of the mitogenic responses in human PBMC.Two higher doses of TSST-1 (10 pg/mL and 100 pg/mL) were also tested in some cases.
Can J Infect Dis Vol 10 No 6 November/December 1999 405 TSST-1 and SEA interaction on human peripheral blood mononuclear cells

Interaction of TSST-1 and SEA on mitogenesis of human PBMC:
The mitogenic response of human PBMCs to the combination of TSST-1 at 1.0 pg/mL and SEA at 0.1 pg/mL from nine healthy donors was significantly greater than either toxin alone (P<0.05,paired Student's t test, two-tailed) (Figure 2).In addition, PBMCs from three donors simulated with TSST-1 at 10 pg/mL and at 100 pg/mL, each in combination with 0.1 pg/mL SEA, also exhibited significantly greater T cell mitogenesis than either toxin alone (data not shown).The increased mitogenic effect appeared to be additive rather than synergistic, because the mean proliferative response observed upon coincubation with TSST-1 and SEA did not significantly exceed the sum of the proliferative response with either toxin alone.

Minimum effective dose of TSST-1 and SEA on cytokine secretion from human PBMCs:
The dose responses of human PBMCs to various concentrations of TSST-1 and SEA (ranging from 0.001 to 1 m g/mL) in the induction of TNFa , IL-1ß, and IL-6 were studied in one healthy donor.Although dosedependent effects of TSST-1-and SEA-induced TNFa and IL-6 production were observed with this donor (data not shown), IL-1 ß was detected only at the highest concentration of TSST-1 and SEA tested (1 m g/mL).In light of this, and the inherent variability in cytokine induction from different donors (21,22), the minimum effective dose of TSST-1 and SEA, chosen arbitrarily for further study of cytokine secretion in human PBMCs, were 10 pg/mL and 1.0 or 10 pg/mL, respectively.Interaction of TSST-1 and SEA on TNFa, IL-1ß and IL-6 secretion in human PBMCs: PBMCs (1:1 ratio of T cells to monocytes) of eight healthy donors were stimulated with TSST-1 (10 pg/mL) and SEA (1 pg/mL), either alone or in combination.
Considerable intersubject variability in cytokine responses was observed (Table 1).A significant increase in TNFa was observed in the culture supernatants harvested from five donors stimulated with the combination of TSST-1 and SEA (1.65± 0.27 ng/mL) compared with either TSST-1 alone (0.85± 0.22 ng/mL, P<0.05, paired t test, two-tailed), or SEA alone (0.56± 0.35 ng/mL, P<0.05) (Figure 3A).In contrast with TNFa , the induction of IL-1ß from eight donors stimulated with the combination (   greater than with TSST-1 alone (5.77± 2.8 ng/mL) or SEA alone (3.65± 2.0 ng/mL) (Figure 3B).Increasing the concentration of SEA to 10 pg/mL did not alter the results (data not shown).
Similarly, the induction of IL-6 from eight donors stimulated with the combination (1.19± 0.43 ng/mL) was not significantly greater than with TSST-1 alone (0.99± 0.41 ng/mL), although a significant increase was observed in comparison with SEA alone (0.53± 0.28 ng/mL; P<0.05) (Figure 3C).Again, increasing the concentration of SEA to 10 pg/mL did not change the results (data not shown).

DISCUSSION
The objective of our study was to determine if the combination of minimum effective concentrations of TSST-1 and SEA would induce greater T cell proliferation and cytokine secretion from human PBMCs with compared either toxin alone.This might be suspected because TSST-1 and SEA have similar biological activities, and both bind to MHC class II molecules on human PBMCs (23,24), an event critical for both T cell mitogenesis and cytokine secretion (11,(25)(26)(27).Furthermore, because TSST-1 and SEA share a common MHC class II receptor in human PBMCs (28)(29)(30), a greater response in mitogenesis or cytokine secretion, at minimum effective concentrations, could be expected from the combination of these two toxins rather than with either toxin alone.Our finding that the coincubation of SEA with TSST-1 induced significantly greater T cell proliferation and TNFa secretion supports this notion.The enhanced effects of these toxins in combination are relevant because TSS-associated S aureus frequently co-produce both TSST-1 and SEA (7,31,32) and because these two biological activities are believed to be pivotal to the pathogenesis of TSS (15,16).However, one must be cautious in extrapolating these in vitro results to the clinical setting because the precise amount of TSST-1 and SEA produced in vivo by S aureus during TSS is not known.
In contrast with TNFa induction, we were unable to demonstrate an enhancement in either IL-1ß or IL-6 production by the toxin combination compared with both TSST-1 and SEA alone.There are a number of possible explanations for this apparent disparity.First, it is likely that suboptimal concentrations of TSST-1 and SEA were used for some of the donors studied.Ideally, a dose-response curve for IL-1ß and IL-6 induction should be determined for each PBMC donor, so that the minimum effective dose of TSST-1 and SEA could be more precisely defined for each individual donor.Secondly, the secretion of IL-1ß and IL-6 in PBMC is known to appear later than TNFa after superantigen stimulation (16).Because cytokine secretion in our study was determined at approximately 20 h after stimulation, there may not have been sufficient time for IL-1ß and IL-6 synthesis and secretion to occur in the PBMCs of some donors in contrast with TNFa .In support of this, IL-1ß was below the limits of detection in four of eight culture supernatants, while IL-6 was undetectable in three of eight donors (Table 1).However, because these cytokines appear to be differentially regulated via distinct signal transduction pathways (33)(34)(35), it would not be totally surprising if this apparent disparity in cytokine secretion between TNFa and the other cytokines was still observed when tested under more optimal conditions.Whether the different cytokine activation pathways induced by these toxins are related to the unique MHC Class II-binding sites of these toxins (36)(37)(38) is not known.
Considerable intersubject variability in cytokine secretion patterns was observed in our healthy donors following superantigen stimulation, as has been documented by others (21,22,39).For example, among the eight donors examined for cytokine secretion in the current study, the amount of IL-1ß induced with 10 pg/mL of TSST-1 ranged from less than 500 to 18,000 pg/mL, whereas IL-6 levels ranged from less than 625 to 2935 pg/mL (Table 1).A similar finding was observed following stimulation with 1 pg/mL of SEA.Donor 1 produced high amounts of all three cytokines in response to both TSST-1 and SEA, whereas donor 4 did not produce any detectable cytokine in response to either TSST-1 or SEA.In general, donor PBMCs that did not respond to TSST-1 (eg, donors 4, 5 and 6) also did not respond to SEA in the secretion of identical cytokines.Several factors, including isotypic and allelic polymorphism of MHC Class II in different subjects, could account for the variations observed in the secretion of different cytokines by the same donor or the same cytokine by different donors.First, different superantigens preferentially bind to distinct MHC class II isotypes (40,41) and alleles (36,41), thereby possibly leading to variability in the induction of different cytokines in the same individual.In this regard, the activation of certain T cell subsets by SEA was found to have an influence on the type of cytokine released from human monocytes (42,43).The finding that human TNFa genes are linked to the MHC gene cluster on chromosome 6 (44) also supports the concept that genetic factors may regulate the cytokine response, at least for TNFa .Genetic factors may also provide an explanation for the observation that some individuals appear to be consistently high or low cytokine producers (21).Variation in the Vß determinant for the TCR repertoire among different individuals may also account for some intersubject variability in cytokine responses to different superantigens (10).

CONCLUSIONS
The combination of TSST-1 and SEA induces a significantly greater mitogenic response and TNFa secretion in human PBMCs compared with stimulation with either toxin alone.Whether these in vitro observations also occur in vivo and whether the coproduction of TSST-1 and SEA in TSSassociated S aureus renders this strain more virulent than another that produces only TSST-1 or SEA clearly warrant further investigation.Can J Infect Dis Vol 10 No 6 November/December 1999 Can J Infect Dis Vol 10 No 6 November/December 1999 409 TSST-1 and SEA interaction on human peripheral blood mononuclear cells

DIAGNOSIS
Sinus X-rays were performed (single view in Figure 2), which showed complete opacification of both maxillary sinuses, air fluid levels in both frontal sinuses, opacification of the ethmoid sinuses and a questionable air fluid level in the sphenoid sinus.A computed tomography scan of the sinuses and orbits with infusion showed the soft tissue swelling of the face without extension into the orbits.The nasal cavity was completely occluded, with uptake of radiocontrast into the nasal, ethmoid and maxillary mucosa.The computed tomographic scan also showed complete occlusion of the maxillary sinuses by hypodense material, and partial occlusion of the sphenoid and frontal sinuses as well.Air fluid levels were again seen in the frontal sinuses.A diagnosis of facial erysipelas secondary to pansinusitis was made.
The patient was started on intravenous meropenem (Merrem, AstraZeneca, Mississauga, Ontario) with gradual improvement in the fever and facial swelling over the next four days.No sinus lavage was felt to be indicated, due to the adequate response to antibiotics alone.Blood cultures were negative, and he was changed to oral amoxicillin/clavulanic acid (Clavulin, SmithKline Beecham, Oakville, Ontario) and discharged home on day 5.A follow-up visit with the otolaryngologists was arranged.

DISCUSSION
Facial erysipelas used to be the most common form of erysipelas, now much more common in the lower extremity.It is almost always caused by Streptococcus pyogenes, but may be produced by beta-hemolytic streptococci of other types (ie, groups C or G), as well as other bacteria in more rare circumstances.Preceding skin trauma or portals of entry are common, but facial erysipelas may also be caused by extension of orofacial foci such as odontogenic infections.Sinus infections as a source of facial erysipelas, without orbital involvement, are quite rare.The exact microbial etiology of facial erysipelas when caused by an underlying sinusitis is unknown, but may be polymicrobial in nature.Cultures of erysipelas are infrequently positive, even if performed by 'leading edge aspiration' of the rash.Bacteremia occurs less than 5% of the time.Therapy usually consists of antistreptococcal antibiotics (unless the condition is suspected to be caused by a polymicrobial infection) and local wound care.Blistering and desquamation are common during the healing process.
The two unusual aspects of this patient's presentation are the sinus infection as the likely source of the erysipelas, and the lack of facial pain or pressure in the presence of such an extensive pansinusitis.When encountering a facial erysipelas without evidence of a portal of entry, a search for an underlying infectious source is mandatory.An examination of the teeth and sinuses are necessary.If no source is still evident, sinus x-rays should be completed in all such patients, regardless of symptoms.

Figure 2 )
Figure 2) Mitogenic response of human peripheral blood mononuclear cells from nine healthy donors stimulated with toxic shock syndrome toxin-1 (TSST-1) , staphylococcal enterotoxin A (SEA) or their combination for 72 h.Each value is presented as the mean ± SEM from a minimum of quadruplicate determinations.*Significantly difference from the combination of TSST-1 and SEA (P<0.05;paired Student's t test, two-tailed)

Figure 3 )
Figure 3) Induction of tumour necrosis factor-alpha (TNFa), interleukin (IL)-1 in culture supernatant of human peripherial blood mononuclear cells from healthy donors after stimulation with toxic shock syndrome toxin-1 (TSST-1), staphylococcal enterotoxin A (SEA) or their combination for approximately 20 h.Each value representsthe mean ± SEM of duplicate derminations.*Significantly different from the combination of TSST-1 and SEA (P<0.05,paired Student's t test, two-tailed)

De
Boer et al 41.Herman A, Croteau G, Sekaly R-P, Kappler J, Marrack P. HLA-DR alleles differ in their ability to present staphylococcal enterotoxins to T cells.J Exp Med 1990;172:709-17.42.Fleischer B, Schrezenmeier H, Conradt P. T lymphocyte activation by staphylococcal enterotoxins: role of class II molecules and T cell surface structures.Cell Immunol 1989;120:92-101.43.Gjorloff A, Fisher H, Hedlund G, et al.Induction of interleukin-1 in human monocytes by the superantigen staphylococcal enterotoxin A requires the participation of T cells.Cell Immunol 1991;137:61-71.44.Spies T, Morton CC, Nedospasov SA, Fiers W, Pious D, Strominger JL.Genes for the tumor necrosis factors alpha and beta are linked to the human major histocompatibility complex.Proc Natl Acad Sci USA 1986;83:8699-702.

De Boer et al Interaction de la toxine-1 et de l'entérotoxine A staphylococciques du syndrome de choc toxique sur la prolifération des cellules T et sur la sécrétion du TNFa dans les cellules mononucléaires du sang humain
mM Hepes buffer and 10 m g/mL polymyxin B sulphate.
CONCLUSIONS : Ces données permettent de croire que la coproduction de la TSST-1 et de la SEA par S. aureus pourrait fournir un certain avantage biologique à l'organisme par le biais d'une augmentation de l'effet de ces superantigènes sur l'activation des cellules T et la sécrétion du TNF.

TABLE 1 Induction of tumour necrosis factor-alpha (TNFa) interleukin (IL)-1 and IL-6 in culture supernatants of human peripherial blood mononuclear cells (PBMCs) from healthy donors followng stimulation with TSST-1 and SEA
Human PBMCs from eight donors were stimulated with TSST-1 (10 pg/mL) and SEA (1 pg/mL) for approximately 20 h.TNFa, IL-1 and IL-6 in the culture supernatants were measured by ELISA.Sensitivity limits of detection were 250, 500, and 625 pg/ml, respectively.Results for each donor are expressed as the mean SEM cytokine concentration (pg/mL) in duplicate determinations.NT Not tested; UD Undetectable