Streptococcal impetigo induces Th1‐preponderant activation of T lymphocytes with subsequent anergy to superantigenic exotoxins in patients with atopic dermatitis

SIR, Impetigo is a contagious superficial pyogenic infection of the skin caused by Staphylococcus aureus (S. aureus) or Streptococcus pyogenes (S. pyogenes). Streptococcal impetigo, which appears as a non‐bullous, crusty eruption, readily occurs in patients with atopic dermatitis (AD).1 We investigated the changes in peripheral blood eosinophil counts associated with streptococcal impetigo of nine patients with AD (aged 12–25 years; two males and seven females; four patients in addition had poststreptococcal glomerulonephritis and eight had simultaneous secondary infection with S. aureus).2 In all but one of the patients, the total peripheral blood leucocyte count was elevated on presentation with active impetigo. In contrast, the eosinophil count was decreased when the patients presented with active impetigo, began to increase 7–10 days after admission (impetigo subsiding) and rose further in another 7–11‐day period with clearance of the impetigo. Contrariwise, there was no tendency for a change in eosinophil count in six AD patients whose impetigo was purely staphylococcal (five males aged 14–44 years and one 23‐year‐old female). In five patients with streptococcal impetigo, the serum IgE levels fell between admission and 10–46 days later, by 16–75%. However, in patients with staphylococcal impetigo, the changes in IgE were variable (−22% to a 43% elevation). Hence, after impetigo caused by streptococci, the eosinophil count falls quickly, and the IgE level declines slowly. This observation raised the possibility that S. pyogenesmodulates the Th2‐predominant immunological profile in AD patients.3 We cultured peripheral blood mononuclear cells (PBMC) from three normal subjects for 72 h in the presence of S. pyogenes isolated from one patient and treated with mitomycin C (MMC),4 and quantified interferon‐γ (IFN‐γ) and interleukin (IL)‐4 levels in the culture supernatant. MMC‐treated S. pyogenesstimulated PBMC to produce IFN‐γ; the IFN‐γ concentration of the S. pyogenesaddition group and that of the control were 558 ± 78 and 11 ± 2 (mean ± SD, units/mL; P< 0.001), respectively. No detectable IL‐4 was present in either supernatant. Similar results were obtained using superantigenic streptococcal pyrogenic exotoxin (SPE) A.5 Cutaneous infection with S. pyogenes seems to alter systemic immunity in AD patients by releasing superantigens. PBMC taken from three patients (cases 1–3, concomitant infection with both species) on admission (active impetigo) were cultured with staphylococcal enterotoxin B (SEB), toxic shock syndrome (TSS) toxin‐1 (TSST‐1), SPEA, or SPEC, as described previously.6 In a representative, healthy subject (Fig. 1, left column), SEB and TSST‐1 induced T‐cell proliferation at a concentration as low as 0.1 ng/mL, while the mitogenic activities of SPEA and SPEC became apparent at 1 ng/mL. In patients 1 and 2, while PBMC responded normally to SEB and TSST‐1, they responded poorly to SPEA and SPEC (proliferative responses to these toxins were not substantial at 1 ng/mL). Although less apparent, patient 3 exhibited a similar pattern. In addition, SPEA and SEB stimulated PBMC from a normal subject to produce IFN‐γ at comparable levels, whereas PBMC from patients 1 and 2 on stimulation with SPEA produced smaller amounts of IFN‐γ than with SEB, at the time of active impetigo (Fig. 1, right column). When the SPE‐stimulated proliferation and IFN‐γ production in case 1 were tested again after the impetigo had healed (11 days after admission), they increased markedly (data not shown). Vβ2 and Vβ8 are the T‐cell receptor elements possessed by T‐cell populations that are reactive with both SPEA and SPEC.5 We monitored the percentage of these T cells in patient 1 by flow cytometry. When taken at the time of active impetigo, PBMC contained 5.3% Vβ2+ cells (normal, 4.7 ± 1.8: mean ± SD) and 3.1% Vβ8+ cells (normal, 3.4 ± 1.2). Neither the percentage of Vβ2+ cells (5.7%) nor Vβ8+ cells (2.3%) was substantially increased after clinical improvement. Thus, the degree of T‐cell proliferation in response to streptococcal toxins did not correlate with the number of T cells bearing relevant Vβ, suggesting that T‐cell clonal anergy, but not clonal deletion, is the mechanism of hyporesponsiveness, as observed in TSS.7 Poor responses of peripheral blood mononuclear cells (PBMC) to streptococcal pyrogenic exotoxin (SPE) A and SPEC and reduced production of IFN‐γ by PBMC in response to SPEA at the active stage of impetigo. (Left column) PBMC from cases 1–3 at the time of active impetigo and PBMC from a normal subject were cultured for 72 h in the presence of staphylococcal enterotoxin‐B (SEB), toxin shock syndrome toxin‐1 (TSST‐1), SPEA, or SPEC at the indicated concentration. (Right column) PBMC were cultured for 72 h in the presence or absence of SPEA or SEB (triplicate experiments). IFN‐γ in the culture supernatants was quantified by an enzyme‐linked immunosorbent assay. Vertical bars represent SD. *P< 0.005, **P< 0.05, ***statistically not significant, compared with the corresponding SEB. These findings suggest that superantigens released from S. pyogenes induce a transient burst of Thl‐preponderant cytokine production, sufficient to depress eosinophil recruitment and, subsequently, temporary T‐cell anergy to these exotoxins. The Th1 skewing effect of the superantigen was also reported with TSST‐1.8 Circulating superantigen was detected in patients with streptococcal TSS.9 Such diffusion of exotoxins from lesions to the blood may also occur in cutaneous infection with S. pyogenes. Given that Th2 cells play an important part in AD,3 it was to be expected that streptococcal impetigo might improve the clinical condition of AD. However, we did not see any such remarkable improvement of AD skin lesions during the course of impetigo in our patients. Presumably, the unsustained production of IFN‐γ is not sufficient to affect the disease process of AD or, alternatively, Th1 cells in addition to Th2 cells may be involved in producing the eruption of AD.10