SEARCH

SEARCH BY CITATION

Keywords:

  • Kawasaki disease;
  • superantigen;
  • toxin;
  • pathogenesis;
  • serological evidence

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. References

To serologically determine the association of microbial superantigens and the pathogenesis of Kawasaki disease (KD), we conducted a case-control study. Serum IgG and IgM antibodies against staphylococcal enterotoxin A (SEA), SEB, SEC, toxic shock syndrome toxin-1 (TSST-1), and streptococcal pyrogenic exotoxin A (SPEA) were measured by an enzyme-linked immunosorbent assay in 293 serum samples from 65 KD patients on clinical days 1–28 and 120 control samples. The administration of immunoglobulin products, which contain high concentrations of IgG antibodies against all the superantigens, directly elevated antitoxin IgG antibodies in KD patients. In contrast, antitoxin IgM antibodies were not detected in immunoglobulin products. Actually, we found a significant elevation of IgM antibodies against SEA in KD patients in the first (median titre: 0·020, P < 0·01 versus control), second (0·024, P < 0·001), third (0·030, P < 0·001) and fourth (0·038, P < 0·001) weeks, compared to the controls (0·015). Significant differences of IgM antibodies were also true for SEB, TSST-1, and SPEA throughout the first to fourth weeks, and for SEC throughout the second to fourth weeks. The prevalence of KD patients having high IgM titres (> mean + 2SD of control values) to the 5 superantigens was increased with the clinical weeks, and reached 29–43% of KD subjects at the fourth week. This is the first study that describes kinetics of IgM antibodies against superantigens and clarifies the serological significance throughout the clinical course of KD. Our results suggest that multiple superantigens involve in the pathogenesis of KD.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. References

Kawasaki disease (KD) is an acute febrile, systemic vasculitis syndrome of early childhood [1,2]. Despite comprehensive attempts to delineate the causative agents, the aetiology remains to be identified. The epidemiology and clinical features of KD, however, suggest that an infectious agent is the cause or at least an inciting agent [1]. Current evidence suggests that there is an initial infectious trigger consistent with the presence of superantigenic activity. Several investigators have shown selective expansion of T cell receptor (TCR) Vβ2-bearing T cells in peripheral blood during the acute phase of KD [3–5]. Leung et al. [6] showed that a new strain of toxic shock syndrome toxin-1 (TSST-1)-producing Staphylococcus aureus was significantly more frequently isolated from KD patients. However, since other investigators failed to find similar results on these approaches [7–9], the contribution of superantigens (SAgs) to KD has been still debated.

Early serological studies have not shown any evidence of staphylococcal or streptococcal toxin involvement in the pathogenesis of KD [10,11]. In contrast, Nomura et al. recently indicated that TSST-1 [12] and streptococcal pyrogenic exotoxin A (SPEA) [13] contribute to KD in infants younger and older than 6 months of age, respectively. Other investigators showed that streptococcal pyrogenic exotoxin C (SPEC) may be involved [5,14]. To determine a possible association between bacterial SAgs and the pathogenesis of KD, we measured serum antibodies against staphylococcal enterotoxins (SEs), TSST-1, and SPEA in KD patients and control subjects. Analyses based on IgG responses, however, include crucial limitations because immunoglobulin products derived from adult volunteers potentially contain anti-SAg IgG antibodies. These limitations preclude the accurate evaluation on temporal changes of IgG antibodies including early convalescent phase, and on the seroconversion rate. To overcome such limitations, we have focused on the kinetics of IgM antibodies against SAgs. We showed that KD patients had significant elevation of IgM antibodies against one or more of 5 SAgs examined (SEA, SEB, SEC, TSST-1 and SPEA) throughout the first to fourth clinical weeks.

Patients and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. References

This study was conducted at Nishi-Kobe Medical Centre, Department of Paediatrics, and immunoglobulin titres to SAgs were measured at Toray Industries Inc. with approval of the ethical committee at each institute.

Patient population

Between January 1997 and July 2004, infants and children fulfilling the diagnostic criteria for KD [2] were enrolled. We studied 65 KD patients (male/female: 44/21) (Table 1) admitted to our hospital on days 1–9 (day 4·7 ± 2·0). One hundred and twenty disease-free children (male/female: 70/50), who attended our hospital for routine examination before minor elective surgery or for health examination, served as controls (Table 1). We excluded from control subjects, those with:

Table 1.  Demographic features of patients with Kawasaki disease and controls.
 Kawasaki disease (n = 65)Control (n = 120)P value
Age (months)
 Mean ± SD23⋅0 ± 19·524⋅1 ± 21⋅2NS
 Range2–982–98 
Gender (M/F)44/2170/50NS
  • • 
    chronic diseases;
  • • 
    recent medication, surgery or immunoglobulin transfusion;
  • • 
    a history of streptococcal or staphylococcal infections within the previous 6 months.

There were no significant differences in gender or age distribution between KD patients and controls (Table 1).

Treatment procedures

All KD patients were treated with intravenous immunoglobulin (IVIG) of 1 or 2 g/kg and oral aspirin. Immunoglobulin products used were polyethylene glycol-treated (Venoglobulin™-IH, Mitsubishi Pharma, Osaka, Japan) and sulphonated (Venilon™-I, Teijin Pharma Limited, Tokyo, Japan) human immunoglobulin. When clinically refractory to the initial treatment, IVIG was repeatedly administered. Of 65 KD patients, 4 had coronary aneurysms at 1 clinical month, that all resolved 1 years later.

Blood samples and measurement of serum anti-SAgs antibodies

Blood samples were collected with informed consents from one or both of parents, only when blood tests were clinically indicated. After processing to the clinical laboratory for the determination of blood chemistry, the remaining serum was stored at −80 °C until analysis. Serum anti-SAg antibodies were measured by an enzyme-linked immunosorbent assay (ELISA) as previously described [15]. Briefly, 4 toxins from S. aureus (SEA, SEB, SEC and TSST-1) and 1 toxin from S. pyogenes (SPEA) (Toxin Technology, Florida, USA) were used for antigens. Each of the 5 antigens was incubated on a 96-well microplate at 4 °C overnight. After washing, the plates were treated overnight with phosphate-buffered saline (PBS) containing 0·5% bovine serum albumin. Sera diluted 1000-fold for IgG with PBS containing 0·25% bovine albumin and 0·05% Tween 20 were added and incubated for 1 h at 25 °C. As for IgM titre, the conditions were modified; sera were diluted 100-fold and incubation was at 4 °C overnight. After rinsing, peroxidase-conjugated antihuman Fc-specific IgG or IgM was added and incubated for 30 min (or 1 h for IgM antibody) at 25 °C, and the plates were rinsed. Substrate solution containing 3,3’,5,5’-tetramethylbenzidine was then added and incubated for 30 min (or 15 min for IgM antibody) at 25 °C. These ELISA plates were read spectrophotometrically at 450 nm. We also measured anti-SAg IgG and IgM antibodies in immunoglobulin products at an IgG concentration of 5000 mg/dl in 5 lots each. Each sample was examined in duplicate. The intra-assay coefficients of variation for anti-SAg IgG and IgM antibodies were 2·2–11·2% and 3·3–8·0%, respectively, and the interassay coefficients of variation were 4·2–12·3% and 4·5–10·2%, respectively.

Statistics

Data were expressed as mean ± SD unless otherwise noticed. Difference in gender and age distribution between the 2 groups was determined by Fisher's exact test and student-t-test, respectively. Serum total IgM levels were compared by Mann–Whitney U-test. Because values of ELISA titres for both KD and control individuals showed a skewed distribution, median concentrations were first computed for these parameters, and the significance of differences between the 2 groups was assessed by Mann–Whitney U-test. Next, because the distribution after logarithmic transformation (natural logarithm) became normally distributed, logarithmically transformed ELISA titres were used for subsequent statistical analysis. After logarithmic transformation, outlying values of control subjects were excluded according to the one-third rule [16]; if the difference between the extreme value and the next observation is greater than one third of the entire range of all observations, the extreme observation can be excluded. When a logarithmic transformed value of KD patients exceeded the mean + 2SD of that of controls, it was considered high. Significance of differences in the proportion of patients having high anti-SAg antibodies was determined by Fisher's exact test. Correlations between antibody titres and other clinical indices were assessed using Spearman's rank correlation coefficient. Probability values less than 0·05 were considered significant.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. References

Collection of serum samples

We obtained 293 serum samples from 65 KD patients on clinical days 1–28. These were distributed in the first week (days 1–7, n = 128), second week (days 8–14, n = 113), third week (days 15–21, n = 38) and fourth weeks (days 22–28, n = 14). Mean sample numbers from each KD patient were 4·4 ± 1·8 (range 2–8).

IgG antibodies against 5 kinds of SAgs in KD patients, controls and immunoglobulin preparations

The median antitoxin IgG antibodies against SEA, SEB, SEC, TSST-1, and SPEA in control subjects were 0·010 (interquartile range (IQR) 0·004–0·012), 0·008 (IQR 0·005–0·030), 0·048 (IQR 0·027–0·063), 0·019 (IQR 0·005–0·038) and 0·014 (IQR 0·012–0·021), respectively (Fig. 1). In KD patients, we noticed a huge increase in IgG antibodies to 5 SAgs immediately after IVIG therapy. We compared the absolute differences of serum total IgG and anti-IgG antibody levels, designated as ΔIgG and ΔIgG-antitoxin antibodies, before and after (within 48 h) IVIG (Table 2). The ΔIgG-antitoxin antibodies paralleled ΔIgG levels in a dose-dependent manner (Table 2). In addition, two kinds of immunoglobulin products contained a large amount of antitoxin IgG antibodies against SEA, SEB, SEC, TSST-1, and SPEA with mean titres of 0·076, 0·998, 0·937, 0·376 and 0·155, respectively. Hence, ΔIgG-antitoxin antibodies were largely not derived from true immune responses, but from immunoglobulin products.

image

Figure 1. Antitoxin IgG antibodies against superantigens of patients with Kawasaki disease (KD) prior to treatment and controls (C). SEA; staphylococcal enterotoxin A, SEB; staphylococcal enterotoxin B, SEC; staphylococcal enterotoxin C, TSST-1; toxic shock syndrome toxin-1, SPEA; streptococcal pyrogenic exotoxin A. Indicated are median (inner lines), 25/75 percentiles (boxes) and 10/90 percentiles (whiskers).

Download figure to PowerPoint

Table 2.  Absolute differences of serum total IgG and ELISA IgG titre to superantigens before and after intravenous immunoglobulin therapy.
Dose of IVIG treatmentELISA titreΔIgG (mg/dl)
ΔSEAΔSEBΔSECΔTSST-1ΔSPEA
  1. IVIG, intravenous immunoglobulin; Δ, post-IVIG value - pre-IVIG value; SEA, staphylococcal enterotoxin A; SEB, staphylococcal enterotoxin B; SEC, staphylococcal enterotoxin C; TSST-1, staphylococcal toxic shock syndrome toxin-1; SPEA, streptococcal pyrogenic exotoxin A.

1 g/kg (n = 21)
 Mean (A)0⋅0330⋅5180⋅3540⋅0970⋅080 1637⋅9
 SD0⋅0200⋅2160⋅2400⋅0390⋅030  404⋅8
2 g/kg (n = 19)
 Mean (B)0⋅0480⋅7800⋅4730⋅1900⋅108 2692⋅2
 SD0⋅0290⋅5560⋅2420⋅0990⋅071  468⋅3
B/A1⋅45 1⋅51 1⋅341⋅961⋅35 1⋅64

These results lead us to analyse the IgG antibodies in only pretreatment samples (clinical day; 4·4 ± 2·0, range; 1–9 days, n = 53). The median levels of IgG antibodies against SEA (0·013, IQR; 0·009–0·027, P < 0·001 versus controls), SEB (0·043, 0·013–0·069, P < 0·001), SEC (0·055, 0·042–0·068, P < 0·01), and SPEA (0·018, 0·013–0·031, P < 0·05), but not against TSST-1 (0·017, 0·013–0·035) in KD patients were significantly more elevated than those of controls (Fig. 1). To determine the proportion of children having ‘high’ IgG titres, we measured the cut-off points for each antitoxin antibody, and the results were 0·025, 0·184, 0·101, 0·460, and 0·030 for SEA, SEB, SEC, TSST-1, and SPEA, respectively. A significant higher proportion of patients having high IgG antibodies was observed for SEA (KD versus controls; 15/53 versus 5/120, P < 0·001), and SPEA (15/53 versus 7/120, P < 0·001), but not for SEB (5/53 versus 9/120), SEC (9/53 versus 16/120), and TSST-1 (3/53 versus 10/120).

IgM antibodies against 5 SAgs in KD patients, controls and immunoglobulin products

Next we compared serum total IgM and IgM antibodies against the 5 SAgs in paired samples before and after (within 48 h) IVIG treatment. In contrast to total IgG, total IgM levels were not significantly different between pre (171·4 ± 52·9 mg/dl) and post-IVIG treatment (171·3 ± 48·4 mg/dl). Furthermore, the IgM antibodies against the 5 SAgs did not significantly differ between pre- and post-IVIG sera for all SAgs; SEA (median, IQR; 0·021, 0·012–0·030 versus 0·019, 0·010–0·030), SEB (0·035, 0·023–0·053 versus 0·035, 0·025–0·053), SEC (0·055, 0·045–0·084 versus 0·055, 0·041–0·072), TSST-1 (0·031, 0·022–0·056 versus 0·034, 0·023–0·049), and SPEA (0·033, 0·018–0·061 versus 0·031, 0·021–0·059), respectively. ELISA titres for each IgM antibody in both kinds of immunoglobulin preparations were all less than 0·010. Thus, globulin administration did not have any effect on the IgM ELISA titres. We therefore divided the samples according to the clinical course, but not to IVIG-treatment phases.

Kinetic changes in IgM antibodies against SAgs

As shown in Fig. 2a–e, IgM antibodies against SAgs were increased with clinical weeks. The median titres of anti-SEA IgM antibodies in the first (0·020, IQR 0·012–0·032, n = 128, P < 0·01 versus controls), second (0·024, IQR 0·018–0·039, n = 113, P < 0·001), third (0·030, IQR 0·023–0·050, n = 38, P < 0·001), and fourth (0·038, IQR 0·025–0·147, n = 14, P < 0·001) weeks were all significantly higher than those of control subjects (0·015, IQR 0·009–0·023) (Fig. 2a). The significant difference of IgM titres between controls and KD patients through the first to fourth weeks was also true for anti-SEB, anti-TSST-1, and anti-SPEA IgM antibodies (Fig. 2b,d,e). The median titres of anti-SEC IgM antibodies through the second to fourth weeks, but not those of the first weeks significantly differed from the median titres found in controls (Fig. 2c). The significant difference of serum total IgM levels between controls (median; 133 mg/dl, IQR 93·5–167·0 mg/dl) and KD patients through the first to fourth weeks (Fig. 2f). Serum total IgM significantly (P < 0·01) rose from the first week (median; 171 mg/dl, IQR 137·0–211·3 mg/dl) to the second week (248, IQR 194·5–261·5), but thereafter reached a plateau at the third and fourth weeks (Fig. 2f). This was not parallel to the kinetics of the specific antitoxin IgM antibodies (Fig. 2a–e). We showed the kinetics of representative cases showing high antibodies against one or multiple superantigens (Fig. 3). Finally, we examined the prevalence of KD patients having high titres in the 5 SAgs (Fig. 4, Table 3). The cut-off values for respective antitoxin IgM antibodies were 0·059, 0·108, 0·198, 0·094 and 0·098 for SEA, SEB, SEC, TSST-1 and SPEA. The positive numbers among 120 controls for SEA, SEB, SEC, TSST-1, SPEA and anyone of them were 4 (3·3%), 4 (3·3%), 1 (0·8%), 1 (0·8%), 3 (2·5%) and 10 (8·3%), respectively (Fig. 4, Table 3). Of 65 KD patients, 6 and 16 had high IgM antibodies against single and multiple SAgs, respectively (Table 3). The prevalence of KD patients for TSST-1, SPEA and anyone of the 5 SAgs was significantly higher throughout the first to fourth weeks, and was increased over time (Fig. 4). The rates for TSST-1 in the first, second, third and fourth weeks were 10/128 (7·8%) (P < 0·05, versus controls), 9/113 (8·0%) (P < 0·05), 8/38 (21·1%) (P < 0·001), and 6/14 (42·9%) (P < 0·001), respectively. Each rate for SPEA was 10·9% (P < 0·05), 16·8% (P < 0·001), 21·1% (P < 0·001), and 42·9% (P < 0·001), respectively. Significant differences were also found in the second (15·9%, P < 0·01), third (18·4%, P < 0·01) and fourth weeks (35·7%, P < 0·01) for SEA, in the third (18·4%, P < 0·01) and fourth (35·7%, P < 0·01) weeks for SEB and in the fourth week (28·6%, P < 0·001) for SEC (Fig. 4).

image

Figure 2. Kinetic changes in antitoxin IgM antibodies against superantigens (a–e), and those of total serum IgM (f) in patients with Kawasaki disease and controls through the first to fourth weeks. (a) SEA; staphylococcal enterotoxin A (b) SEB; staphylococcal enterotoxin B (c) SEC; staphylococcal enterotoxin C (d) TSST-1; toxic shock syndrome toxin-1 (e) SPEA; streptococcal pyrogenic exotoxin A. Indicated are median (inner lines), 25/75 percentiles (boxes) and 10/90 percentiles (whiskers). *P < 0·01 (versus control), **P < 0·001 (versus control).

Download figure to PowerPoint

image

Figure 3. Representative patients with Kawasaki disease who had high IgM antibodies against one (a, against staphylococcal enterotoxin A) or multiple (b, against 5 all toxins) superantigens. IVIG; intravenous immunoglobulin therapy. Open and closed circles, open and closed squares, and open triangle indicate staphylococcal enterotoxin A, staphylococcal enterotoxin B, staphylococcal enterotoxin C, toxic shock syndrome toxin-1, and streptococcal pyrogenic exotoxin A, respectively.

Download figure to PowerPoint

image

Figure 4. Prevalence of patients having high antitoxin IgM antibodies against superantigens. Open and closed circles, open and closed squares, and open and closed triangles indicate staphylococcal enterotoxin A, staphylococcal enterotoxin B, staphylococcal enterotoxin C, toxic shock syndrome toxin-1, streptococcal pyrogenic exotoxin A, and any of the 5 superantigens, respectively.

Download figure to PowerPoint

Table 3.  Patient number with high (> mean + 2 SD of controls) anti-IgM antibodies against at least one of the 5 microbial superantigens.
Superantigen no.Patient no. with high IgM antibody againstKawasaki disease (n= 65)Control (n= 120)
  1. SEA, staphylococcal enterotoxin A; SEB, staphylococcal enterotoxin B; SEC, staphylococcal enterotoxin C; TSST-1, staphylococcal toxic shock syndrome toxin-1; SPEA, streptococcal pyrogenic exotoxin A.

1SEA 2 3
SEB  3
TSST-1 1 1
SPEA 3 1
2SEA + SPEA 3 1
SEA + SEB 1 
SEA + SEC 1 
TSST-1 + SPEA 2 
3SEA + SEB + SPEA 2 
SEA + TSST-1 + SPEA 2 
SEB + SEC + SPEA  1
4SEA + SEB + SEC + TSST-1 1 
SEA + SEB + TSST-1 + SPEA 1 
SEA + SEC + TSST-1 + SPEA 1 
5SEA + SEB + SEC + TSST-1 + SPEA 2 
Total2210

Relationships between anti-SAg antibodies and clinical indices

We examined the correlations of anti-SAg IgM antibodies from the first to fourth weeks and IgG antibodies in pre-IVIG sera to clinical indices. Ages of control subjects or those of KD patients did not correlate with IgM or IgG ELISA titres for any of the 5 SAgs. There were no significant differences of anti-SAg IgM or IgG antibodies between genders. There were no significant differences in any of 5 kinds of IgM or IgG antibodies between patients developing coronary aneurysms (n = 4) and the counterparts without such lesions (n = 61). White blood cell counts and serum C-reactive protein levels did not correlate with anti-SAg IgG or IgM antibody levels. There was no correlation between IgG and IgM antibodies against any of the 5 SAgs in pre-IVIG samples (n = 53).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. References

The involvement of SAgs in KD has been investigated by a variety of methods such as isolation of SAg-producing pathogens [6,8,9], selective expansion of TCR Vβ-bearing T cells in peripheral blood [3–5,7], and serological immune responses [10–14]. However, exogenous administration of immunoglobulin products can potentially limit the ability to accurately evaluate IgG responses to SAgs in KD patients. Of particular note in the present studies, we surveyed kinetic changes of IgM responses, which were not affected by IVIG treatment. To our knowledge, there has been only one report dealing with IgM antibodies against SAgs by Terai et al. [10], who showed no evidence of significant elevation of anti-TSST-1 IgM antibodies. However, their studies set cut-off levels based on adult volunteers, and examined IgM antibody against only one SAg, i.e. TSST-1 [10]. Thus, this is the first describing kinetic changes of IgM antibodies against numerous SAgs from the acute to early convalescent phase of KD, and is one of the largest serological studies with a sufficient number of age-matched controls.

In the present study, we observed significantly higher titres to all 5 bacterial SAgs in IgM antibody analysis and to SEA, SEB, SEC, and SPEA in IgG antibody analysis. These findings support the hypothesis that KD results from an immunological response that is triggered by any of various different microbial agents. In early studies, any association between these pathogens and KD has not been serologically established [10,11], but these studies might be poorly controlled. Recent studies indicated a single SAg involvement of TSST-1 [12], SPEA [13], or SPEC [14]. These disparities of kind and number of SAgs involved may be due to the following differences; ages of control subjects, immunological classes in analyses, ELISA methods, the timing of blood collection, and geographical populations. Alternatively, because of molecular mimicry between SEB, SEC, and SPEA [17,18], cross-reacted IgG or IgM antibodies [19,20] might be detected in our study. In any case, the specific organism may not be as important as a common mechanism of stimulation to develop the immune system. This may explain why efforts to find a unifying aetiological association, or a consistent TCR Vβ repertoire, have been unsuccessful to date [1].

The mechanisms by which IVIG reserves the immunoregulatory abnormalities in KD remain uncertain. Takei and colleagues [21] showed that IVIG contains anti-SE antibodies that suppress SE-induced T-cell stimulation in vitro. In vivo evidence also indicated that SE-specific antibodies obtained from pooled sera protected mice from a lethal dose of SE [22]. We showed that high concentrations of IgG antibodies against the staphylococcal and streptococcal toxins were present in the IVIG products. These results indicate a high prevalence of antitoxin antibodies in Japanese adults as well as US populations [21,22], and may reflect the virtual absence of KD in adulthood due to widespread immunity. More importantly, we also showed that the administration of immunoglobulin products directly influenced serum IgG-antitoxin antibody levels in KD patients. Our findings suggest that IVIG therapy acts by neutralizing the putative toxin that induces the disease. These actions would presumably eliminate the stimulus that triggers the immune activation and counteract with specific antibodies in SAg-associated pathologies in KD.

The temporal changes of antitoxin antibodies were discordant with those found in serum total IgM levels. While serum total IgM reached a plateau by the second week, antitoxin IgM antibody levels continued to increase during the 4-week observation period. The kinetics of anti-SAg IgM antibodies, however, may be somewhat late, because IgM antibodies are classically the first immunoglobulin class to appear after antigenic stimulation. The exact reason for the delayed responses is not clear. One possible explanation is that the ability to produce immunoglobulins may be perturbed by IVIG. Durandy et al. [23] have indicated that lymphocytes obtained from nonimmunodeficent children treated with intramuscular immunoglobulin have a reduced capacity to proliferate and mature into IgG- and IgM-secreting plasma cells in vitro after stimulation with pokeweed mitogen. In fact, IVIG therapy suppresses polyclonal B cell activation observed in the acute phase of KD [24,25]. An alternative explanation is derived from the cellular immune perturbation caused by KD [23–25]. Microbial SAgs bypass normal antigen presentation by binding to class II major histocompatibility complex molecules on antigen-presenting cells and to TCR Vβ elements on T cells, and trigger T cell activation through this interaction [18,26–29]. In an experimental model, antibody production responded to SE and TSST-1 is T cell-dependent [26,27]. In vitro SAg stimulation of human lymphocyte populations can promote T cell-dependent B cell differentiation and immunoglobulin production [28,29]. Although the dependency on T cell or T cell-driven factors in vivo is not fully clarified in KD patients, cellular immune perturbation may be related to their delayed responses to SAgs. Further studies are necessary to precisely determine the mechanisms of the alteration of immune responses.

In conclusion, our findings represent the first demonstration of significant elevations, although to variable degrees, of IgM antibodies against SEA, SEB, SEC, TSST-1, and SPEA in KD patients. These results suggest the involvement of multiple microbial toxins in the pathogenesis of KD.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. References
  • 1
    Burns JC, Glodé MP. Kawasaki syndrome. Lancet 2004; 364: 53344.
  • 2
    Japanese Kawasaki Disease Research Committee. Diagnostic Guidelines of Kawasaki Disease, 5th edn. Tokyo: Japanese Kawasaki Disease Research Committee, 2002.
  • 3
    Abe J, Kotzin BL, Jujo K, Melish ME, Glode MP, Kohsaka T, Leung DYM. Selective expansion of T cells expressing T-cell receptor variable regions Vβ2 and Vβ8 in Kawasaki disease. Proc Natl Acad Sci USA 1992; 89: 406670.
  • 4
    Curtis N, Zheng R, Lamb JR, Levin M. Evidence for a superantigen mediated process in Kawasaki disease. Arch Dis Child 1995; 72: 30811.
  • 5
    Yoshioka T, Matsutani T, Iwagami S et al. Polyclonal expansion of TCRBV2- and TCRBV6-bearing T cells in patients with Kawasaki disease. Immunology 1999; 96: 46572.
  • 6
    Leung DYM, Meissner HC, Fulton DR, Murray DL, Kotzin BL, Schlievert PM. Toxic shock syndrome toxin-secreting Staphylococcus aureus in Kawasaki Syndrome. Lancet 1993; 342: 13858.
  • 7
    Pietra BA, De Inocencio J, Giannini EH, Hirsch R. TCR Vβ family repertoire and T cell activation markers in Kawasaki disease. J Immunol 1994; 153: 18818.
  • 8
    Leung DYM, Meissner HC, Shulman ST et al. Prevalence of superantigen-secreting bacteria in patients with Kawasaki disease. J Pediatr 2002; 140: 7426.
  • 9
    Horita N, Yokota S, Fuse S, Takamuro M, Tomita H, Sato K, Fuji N, Tsutsumi H. The throat flora and its mitogenic activity in patients with Kawasaki disease. Microbiol Immunol 2004; 48: 899903.
  • 10
    Terai M, Miwa K, Williams T, Kabat W, Fukuyama M, Okajima Y, Igarashi H, Shulman ST. The absence of evidence of staphylococcal toxin involvement in the pathogenesis of Kawasaki disease. J Infect Dis 1995; 172: 55861.
  • 11
    Morita A, Imada Y, Igarashi H, Yutsudo T. Serologic evidence that streptococcal superantigens are not involved in the pathogenesis of Kawasaki disease. Microbiol Immunol 1997; 41: 895900.
  • 12
    Nomura Y, Yoshinaga M, Masuda K, Takei S, Miyata K. Maternal antibody against toxic shock syndrome toxin-1 may protect infants younger than 6 months of age from developing Kawasaki syndrome. J Infect Dis 2002; 185: 167780.
  • 13
    Nomura Y, Masuda K, Yoshinaga M, Takei S, Miyata K. Possible relationship between streptococcal pyrogenic exotoxin A and Kawasaki syndrome in patients older than six months of age. Pediatr Infect Dis J 2003; 22: 7948.
  • 14
    Yoshioka T, Matsutani T, Toyosaki-Maeda T, Suzuki H, Uemura S, Suzuki R, Koike M, Hinuma Y. Relation of streptococcal pyrogenic exotoxin C as a causative superantigen for Kawasaki disease. Pediatr Res 2003; 53: 40310.
  • 15
    Miwa K, Fukuyama M, Sasaki R, Shimizu S, Ida N, Endo M, Igarashi H. Sensitive enzyme-linked immunosorbent assays for the detection of bacterial superantigens and antibodies against them in human plasma. Microbiol Immunol 2000; 44: 51923.
  • 16
    Sasse EA. Determination of reference intervals in the clinical laboratory using the proposed guideline national committee for clinical laboratory standards C28-P. Arch Pathol Laboratory Med 1992; 116: 7103.
  • 17
    Shlievert PM, Bohach GA, Ohlendorf DH et al. Molecular structure of Staphylococcus and Streptococcus superantigens. J Clin Immunol 1995; 15: 4S10S.
  • 18
    Proft T, Fraser JD. Bacterial superantigens. Clin Exp Immunol 2003; 133: 299306.
  • 19
    Kienle E, Buschmann HG. Specificity, cross-reactivity and competition profile of monoclonal antibodies to staphylococcal enterotoxin B and C1 detected by indirect enzyme-linked immunosorbent assays. Med Microbiol Immunol 1989; 178: 12733.
  • 20
    Hynes WL, Weeks CR, Iandolo JJ, Ferretti JJ. Immunologic cross-reactivity of type A streptococcal exotoxin (erythrogenic toxin) and staphylococcal enterotoxins B and C1. Infect Immun 1987; 55: 8378.
  • 21
    Takei S, Arora YK, Walker SM. Intravenous immunoglobulin contains specific antibodies inhibitory to activation of T cells by staphylococcal toxin superantigens. J Clin Invest 1993; 91: 6027.
  • 22
    LeClaire RD, Bavari S. Human antibodies to bacterial superantigens and their ability to inhibit T-cell activation and lethality. Antimicrob Agents Chemother 2001; 45: 4603.
  • 23
    Durandy A, Fischer A, Griscelli C. Dysfunction of pokeweed mitogen-stimulated T and B lymphocytes response induced by gammaglobulin therapy. J Clin Invest 1981; 67: 86777.
  • 24
    Jason J, Gregg L, Han A, Hu A, Inge L, Eick A, Tham I, Campbell R. Immunoregulatory changes in Kawasaki disease. Clin Immunol Immunopthol 1997; 84: 296306.
  • 25
    Leung DYM, Siegel RL, Grady S, Krensky A, Meade R, Reinherz EL, Geha RS. Immunoregulatory abnormalities in mucocutaneous lymph node syndrome. Clin Immunol Immunopathol 1982; 23: 10012.
  • 26
    Williams O, Aroeira LS, Mengel J. T cell-dependent antibody response to staphylococcal enterotoxin B. Scand J Immunol 1995; 42: 30510.
  • 27
    Stohl W, Xu D, Zang S et al. In vivo staphylococcal superantigen-driven polyclonal Ig responses in mice: dependence upon CD4+ cells and human MHC class II. Int Immunol 2001; 13: 1291300.
  • 28
    Fuleihan R, Mourad W, Geha RS, Chatila T. Engagement of MHC-class II molecules by staphylococcal exotoxins delivers a comitogenic signal to human B cells. J Immunol 1991; 146: 16616.
  • 29
    Stohl W, Elliott JE, Linsley PS. Human T cell-dependent B cell differentiation induced by staphylococcal superantigens. J Immunol 1994; 153: 11727.