Portal de Periódicos da CAPES

Staphylococcus aureus Clonal Dynamics and Virulence Factors in Children with Atopic Dermatitis

2005; Elsevier BV; Volume: 125; Issue: 5 Linguagem: Inglês

10.1111/j.0022-202x.2005.23916.x

1523-1747

Hans B. Lomholt, Klaus E. Andersen, Mogens Kilian,

Contact Dermatitis and Allergies

A prospective cohort study was undertaken to determine the clonal dynamics of Staphylococcus aureus colonization and infection during 1 y in children with atopic dermatitis, and to correlate specific clones, accessory gene regulator (agr) groups, and production of virulence factors with eczema activity. Eleven children were examined every 6 wk with swaps taken from active eczema, anterior nose, axillae and perineum, and scoring of eczema activity by severity scoring of atopic dermatitis (SCORAD). Individual S. aureus clonal types were identified and examined for production of superantigens, toxins, and were assigned to agr groups. S. aureus colonization patterns ranged from rare colonization over transient colonization to persistent colonization by a single clone or a dynamic exchange of up to five clones. Production of no single virulence factor including superantigens and toxins was significantly associated with exacerbation of eczema. In four children there was a shift between visits in agr group of colonizing clones. These shifts were associated with an increased SCORAD value of 19 (SE=7, p=0.009). Change of clones belonging to the same agr group was not associated with a higher SCORAD value. In 11 of 12 cases with two different clones co-colonizing a child the clones belonged to the same agr group. In conclusion, this limited group of children with atopic dermatitis showed highly variable colonization patterns of S. aureus, and communication between strains by use of agr encoded octa peptides appeared to be active in vivo. Increased severity of eczema was related to a change in agr group and may have been because of inflammation triggered by the takeover of an antigenically different clone, as agr groups represent ancient phylogenetic lineages. A prospective cohort study was undertaken to determine the clonal dynamics of Staphylococcus aureus colonization and infection during 1 y in children with atopic dermatitis, and to correlate specific clones, accessory gene regulator (agr) groups, and production of virulence factors with eczema activity. Eleven children were examined every 6 wk with swaps taken from active eczema, anterior nose, axillae and perineum, and scoring of eczema activity by severity scoring of atopic dermatitis (SCORAD). Individual S. aureus clonal types were identified and examined for production of superantigens, toxins, and were assigned to agr groups. S. aureus colonization patterns ranged from rare colonization over transient colonization to persistent colonization by a single clone or a dynamic exchange of up to five clones. Production of no single virulence factor including superantigens and toxins was significantly associated with exacerbation of eczema. In four children there was a shift between visits in agr group of colonizing clones. These shifts were associated with an increased SCORAD value of 19 (SE=7, p=0.009). Change of clones belonging to the same agr group was not associated with a higher SCORAD value. In 11 of 12 cases with two different clones co-colonizing a child the clones belonged to the same agr group. In conclusion, this limited group of children with atopic dermatitis showed highly variable colonization patterns of S. aureus, and communication between strains by use of agr encoded octa peptides appeared to be active in vivo. Increased severity of eczema was related to a change in agr group and may have been because of inflammation triggered by the takeover of an antigenically different clone, as agr groups represent ancient phylogenetic lineages. accessory gene regulator pulsed-field gel electrophoresis severity scoring of atopic dermatitis staphylococcal enterotoxin A, B, etc Children with atopic dermatitis are more frequently colonized by Staphylococcus aureus on both lesional and non-lesional skin than healthy children. A higher density of S. aureus in lesional skin is associated with more severe eczema, but to what extent and how the staphylococci aggravate the eczema is not clear. S. aureus produce a multitude of factors with potential skin inflammatory properties including adhesins, cytotoxins, superantigens, capsule, epidermolytic toxins, and a number of extracellular enzymes. In general, it is assumed that staphylococcal infection leads to a worsening of atopic eczema and that antibacterial treatment is beneficial when the children are clinically impetiginized. The role of antibiotics is more controversial when the skin is only colonized and not clinically infected (Williams, 2000Williams R.E.A. The antibacterial-corticosteroid combination: What is its role in atopic dermatitis?.Am J Clin Dermatol. 2000; 1: 211-215Crossref PubMed Scopus (21) Google Scholar; Lübbe, 2003Lübbe J. Secondary infections in patients with atopic dermatitis.Am J Clin Dermatol. 2003; 4: 641-654Crossref PubMed Scopus (87) Google Scholar). The population of S. aureus is comprised of many clonal lineages that produce different combinations of virulence factors (Peacock et al., 2002Peacock S.J. Moore C.E. Justice A. et al.Virulent combinations of adhesin and toxin genes in natural populations of Staphylococcus aureus.Infect Immun. 2002; 70: 4987-4996Crossref PubMed Scopus (457) Google Scholar). Recent findings indicate that clones of S. aureus are able to communicate during co-colonization by means of small octa peptides (pheromones) encoded by the accessory gene regulator (agr) gene and interfere with the colonization of competing clones by bacterial interference (Yarwood and Schlievert, 2003Yarwood J.M. Schlievert P.M. Quorum sensing in Staphylococcus infections.J Clin Invest. 2003; 112: 1620-1625Crossref PubMed Scopus (253) Google Scholar). When the concentration of the agr-encoded octa peptide is high, exotoxin production is upregulated whereas surface molecules including adhesins are downregulated. Four agr groups have been recognized based on sequence variation in the agr gene (Ji et al., 1997Ji G. Beavis R. Novick R.P. Bacterial interference caused by autoinducing peptide variants.Science. 1997; 276: 2027-2030Crossref PubMed Scopus (593) Google Scholar) and these groups appear to reflect separate phylogenetic lineages (Jarraud et al., 2002Jarraud S. Mougel C. Thioulouse J. et al.Relationships between Staphylococcus aureus genetic background, virulence factors, agr groups (alleles), and human disease.Infect Immun. 2002; 70: 631-641Crossref PubMed Scopus (845) Google Scholar). Members of the same group stimulate each other, whereas members of different groups are mutually inhibitory with the exception of groups IV and I, where the interaction is not entirely clear. Only few data are available on the significance of this system in vivo and it is not known whether it plays a role in the colonization of atopic dermatitis. For a better understanding of the interplay between staphylococci and the atopic host, it is important to examine in detail the clonal colonization dynamics over time. Therefore, the present longitudinal study was undertaken in children with atopic dermatitis followed for approximately 1 y. Clones were discerned using pulsed-field gel electrophoresis (PFGE), screened for a number of putative virulence factors, and assigned to one of the four agr groups. Clonal characteristics were compared with eczema activity at the time of colonization. Although the number of children was limited, the results suggest that the agr-related diffusion sensing system in an atopic dermatitis patient plays an important role in determining the colonization and hence eczema activity over time. In total the 11 children were seen on 95 visits and 378 swaps were obtained, 162 of which yielded growth of S. aureus. From many swaps S. aureus colonies with more than one colony morphology could be distinguished and a total of 240 different colonies were isolated. S. aureus were recovered from the perineum on 26 visits, eczema on 58 visits, nose on 55 visits, axillae on 23 visits, and no S. aureus were recovered on 25 visits. The 240 S. aureus isolates were assigned to different clones based on the DNA fingerprint pattern revealed by PFGE. Isolates exhibiting identical band patterns were considered belonging to the same clone as suggested (Tenover et al., 1995Tenover F.C. Arbeit R.D. Goering R.V. Mickelsen P.A. Murray B.E. Persing D.H. Swaminathan B. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: Criteria for bacterial strain typing.J Clin Microbiol. 1995; 33: 2233-2239PubMed Google Scholar). Successive clones from each child were aligned on gels and eventually a total of 28 representative clones were distinguished considering clones from different children as separate clones (Figure 1). By grouping together possibly related clones differing in less than six bands (as defined byTenover et al., 1995Tenover F.C. Arbeit R.D. Goering R.V. Mickelsen P.A. Murray B.E. Persing D.H. Swaminathan B. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: Criteria for bacterial strain typing.J Clin Microbiol. 1995; 33: 2233-2239PubMed Google Scholar) eight groups of unrelated clones could be defined (not shown). Two pairs of siblings shared identical clones (Table I, child 2 clone A and child 3 clone A, and child 10 clone B and child 11 clone A).Table IPatterns of Staphylococcus aureus clonal colonization, virulence factors, eczema activity, and antibiotic treatment in 11 children with atopic dermatitis examined at six to 11 occasions during approximately 1 yChild no.PatternaClone: the first clone identified in each child is arbitrarily designated A, the second B, the third C, etc. Clones from different children with the same designation are not identical. Antibiotic: courses of systemic antibiotics are indicated when administered, D represents dicloxacillin, Z azithromycin, and E erythromycin. Factors: production of the superantigens staphylococcal enterotoxin A, B, and C (SEA, SEB, and SEC), epidermolytic toxins A and B (ETA+B), and β toxin (β).Visit 1Visit 2Visit 3Visit 4Visit 5Visit 6Visit 7Visit 8Visit 9Visit 10Visit 111CloneABABBCDBEBDDBBDBDESCORAD384730324841504441AntibioticDFactorsSECSEC2CloneABAAAAAAAASCORAD834972497363417572AntibioticDZZFactorsSEA, βSEA, βSEA, βSEA, βSEA, βSEA, βSEA, βSEA, βSEA, β3CloneA—AAA————SCORAD351216454019494924AntibioticDDFactorsSEA, βSEA, βSEA, βSEA, β4CloneABAABABACbBold figures and gray shades indicate clones of an agr group different from the former clones colonizing the child. SCORAD, severity scoring of atopic dermatitis; agr, accessory gene regulator.CCCDSCORAD715851787892555556AntibioticDZZ DFactorsSEBSEBSEBSEASEASEASEA5CloneAABABCBAAABSCORAD5544282340305841AntibioticFactorsSEA, βSEA, βSEA, βSEA, β6CloneAAAABAACCDDSCORAD325331404736164250AntibioticZDFactors7CloneAAA—A—AAASCORAD232549134720351012AntibioticFactorETA+BETA+BETA+BETA+BETA+BETA+BETA+B8Clone—AABBBCD———CSCORADNd19115464703142653718AntibioticDDEE DDDFactors9Clone——A——ASCORAD565546543527AntibioticEFactorsSEA, βSEA, β10Clone—ABB—B—BSCORAD2136615329373424AntibioticFactorsSEA, βSEA, βSEA, βSEA, β11Clone———————ASCORAD2421531622Nd824AntibioticDFactorsSEA, βa Clone: the first clone identified in each child is arbitrarily designated A, the second B, the third C, etc. Clones from different children with the same designation are not identical. Antibiotic: courses of systemic antibiotics are indicated when administered, D represents dicloxacillin, Z azithromycin, and E erythromycin. Factors: production of the superantigens staphylococcal enterotoxin A, B, and C (SEA, SEB, and SEC), epidermolytic toxins A and B (ETA+B), and β toxin (β).b Bold figures and gray shades indicate clones of an agr group different from the former clones colonizing the child.SCORAD, severity scoring of atopic dermatitis; agr, accessory gene regulator. Open table in a new tab As given in Table I the 11 children showed variable patterns of S. aureus colonization during the study period. On 16 occasions two clones colonized the same child simultaneously and on two occasions three clones co-existed. In most cases the co-colonizing clones were isolated from different locations, but two simultaneous clones were found three times in the nose, two times in eczema, and once in the perineum. Twenty courses of systemic antibiotics were administered during the study period (Table I) often in combination with local fucidin, chinoform containing topical steroid, or potassium permanganate soak, and once with mupirocin nasally. All clones were susceptible to methicillin and erythromycin. Thirteen clones were relatively resistant to fucidic acid, but the concentration following topical use on skin in vivo is thought to be far above the minimal inhibitory concentration. As shown in Table I S. aureus clones generally persisted or returned in spite of antibiotic and antiseptic treatment, and neither antibiotic treatment nor the use of local steroids seemed to influence clonal shifts. When data were analyzed using a longitudinal autoregressive model, simultaneous S. aureus colonization at all four sites was significantly associated with an increased severity scoring of atopic dermatitis (SCORAD) value of 17 (SE=6, p=0.006) (Table II). In this respect, four children showed simultaneous colonization at all four sites on two, five, three, and four occasions, respectively. Moreover, there was a clear positive trend towards a higher SCORAD value the more of the sites that were colonized, indicating that increased S. aureus colonization was associated with more severe eczema. Colonization at no single site was correlated with significantly more eczema except for the perineum. S. aureus at this location, however, was highly correlated to the simultaneous appearance of the organism at the other sites, so it probably just reflected widespread colonization.Table IIEffect on the eczema activity by Staphylococcus aureus colonization, agr group shift, and change in clones as calculated by means of a longitudinal autoregressive modelVariableEffect on SCORADStandard errorp-valueAgr group shift1970.009Clone shift withinAgr groups-1150.03S. aureus colonization at No site0—— One site-340.51 Two sites750.12 Three sites960.13 Four sites1760.006SCORAD, severity scoring of atopic dermatitis; agr, accessory gene regulator. Open table in a new tab SCORAD, severity scoring of atopic dermatitis; agr, accessory gene regulator. The 28 clones detected were assigned to the four known agr groups with a distribution among groups as given in Table III. In 17 of 18 cases clones co-colonizing the same patient shared agr group. The exception was in patient 8, where clone D assigned to agr group II and clone C assigned to agr group I simultaneously colonized the anterior nose of the patient at one occasion. If a particular combination of two clones was counted only once for each patient, the corresponding numbers were 11 of 12 cases. If the frequency of each agr group in Table III is used as the probability of being colonized with strains from that particular group, then the probability of being colonized with strains belonging to the same agr group can be estimated as (0.462+0.182+0.322+0.042)=0.3480. Then the distribution of co-colonizing strains belonging to the same agr group in 11 of 12 cases does not seem to be a random coincidence (probability=0.000074, binomial distribution).Table IIIVirulence factors detected among 28 Staphylococcus aureus clonesVirulence factorNumber of positive strainsPercentage of positive strainsSuperantigens TSST-100 SEA829 SEB14 SEC14 SED00Cytolytic toxins β toxin725Epidermolytic toxins ETA14 ETB14Adhesins fnbA28100 cna1554Agr groups Agr I1346 Agr II518 Agr III932 Agr IV14Haemolytic activity +1036 ++414 +++1450Proteolytic activity None414 +27 ++829 +++1450TSST-1, toxic shock syndrome toxin 1; SEA, SEB, and SEC, staphylococcal enterotoxin A, B, and C; agr, accessory gene regulator. Open table in a new tab TSST-1, toxic shock syndrome toxin 1; SEA, SEB, and SEC, staphylococcal enterotoxin A, B, and C; agr, accessory gene regulator. Table I shows a shift in overall agr group in four patients associated with a significant increase in SCORAD value of 19 (SE=7, p=0.009) (Table II). In contrast, a shift in clones within the same agr group was significantly associated with a small decrease in SCORAD 11 (SE=5, p=0.03). A shift in clones belonging to different clone groups, but of the same agr group, was not significantly associated with changes in SCORAD. The 28 representative S. aureus clones were tested for a number of putative virulence factors including binding proteins, cytotoxin, epidermolytic toxins, and superantigens. The number of clones positive for each of these factors is given in Table III. No clone was found to produce more than one of the superantigens. The most prevalent superantigen and toxin were staphylococcal enterotoxin A (SEA) and β toxin, and the combined production of these was found in seven clones. For each child the production of superantigens and toxins at the visits is given in Table I in relation to clonal type and eczema activity. When these data were applied to the statistic model none of these factors were associated with a significant increase in SCORAD value, neither when calculated separately for each of the four sampling sites nor when all sites were combined. This study confirmed the abundant colonization of atopic eczema patients with S. aureus. Staphylococcal colonization at all four sites was associated with a significantly higher value of SCORAD 17 (SE=6, p=0.006) and a solid trend was found towards more active eczema the more of the sites that were colonized. Previous studies have shown that the number of S. aureus on the skin is positively correlated to the activity of atopic dermatitis (Williams et al., 1990Williams R.E. Gibson A.G. Aitchison T.C. Lever R. Mackie R.M. Assessment of a contact-plate sampling technique and subsequent quantitative bacterial studies in atopic dermatitis.Br J Dermatol. 1990; 123: 493-501Crossref PubMed Scopus (147) Google Scholar). But the cause relationship in these findings is not known. To study the interplay between S. aureus and eczema over time the clonal dynamics during colonization was examined in detail by using chromosomal DNA fingerprinting. Several different patterns of colonization were revealed with some children continuously harboring a single persistent clone, whereas others experienced a dynamic exchange of up to five clones colonizing different or the same location during the observation period. An intermittent pattern of colonization was observed in several children and one child was colonized on only a single occasion. Although the number of children is limited, these findings agree with the three carriage patterns reported in healthy carriers: non-carriers, intermittent carriers, and persistent carriers, with persistent carriage being more frequent among children (Kluytmans et al., 1997Kluytmans J. van Belkum A. Verbrugh H. Nasal carriage of Staphylococcus aureus: Epidemiology, underlying mechanisms, and associated risks.Clin Microbiol Rev. 1997; 10: 505-520Crossref PubMed Google Scholar; van denbergh and Verbrugh, 1999van denbergh M.F.Q. Verbrugh H.A. Carriage of Staphylococcus aureus: Epidemiology and clinical relevance.J Lab Clin Med. 1999; 133: 525-534Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). The finding that two pairs of siblings shared clones indicates that spread within families takes place as previously reported (Breuer et al., 2002Breuer K. Häussler S. Kapp A. Werfel T. Staphylococcus aureus: Colonising features and influence of an antibacterial treatment in adults with atopic dermatitis.Br J Dermatol. 2002; 147: 55-61Crossref PubMed Scopus (252) Google Scholar). Confounding because of siblings cannot be entirely ruled out. But this is probably a minor problem because exclusion of the four siblings in the analysis revealed a similar increase in SCORAD value 17 (SE=8, p=0.05) associated with agr group shift. Within sibling pairs the shared clones were carried in different patterns over time so that the effect of virulence factors on eczema activity might be expected to occur independently in each child and from common factors in their environment. We initially hypothesized that the appearance of a new clone with a new combination or antigenic makeup of virulence factors leads to a flare of eczema. But no exacerbation of dermatitis was associated with clonal exchanges as long as the clones belonged to the same agr group. In fact, statistical analysis showed a decrease in SCORAD 11 (SE=5, p=0.03). When calculated as a change to another of the eight clone groups defined by PFGE analysis, possibly reflecting a change to a genetically more distant strain than mere clone shifts, there was no significant change in SCORAD. To investigate whether bacterial interference among colonizing strains of S. aureus plays a role in vivo in atopic dermatitis all isolates were assigned to one of the four agr groups. The overall distribution of agr groups was similar to findings reported for isolates from healthy nasal carriers and patients with cystic fibrosis (Goerke et al., 2003Goerke C. Kümmel M. Dietz K. Wolz C. Evaluation of intraspecies interference due to agr polymorphism in Staphylococcus aureus during infection and colonisation.J Infect Dis. 2003; 188: 250-256Crossref PubMed Scopus (34) Google Scholar; Kahl et al., 2003Kahl B.C. Becker K. Friedrich A.W. Clasen J. Sinka B. Von Eiff C. Peters G. Agr-dependent bacterial interference has no impact on long-term colonisation of Staphylococcus aureus during persistent airway infection of cystic fibrosis patients.J Clin Microbiol. 2003; 41: 5199-5201Crossref PubMed Scopus (18) Google Scholar). The finding that in 11 of 12 instances, where more than one clone colonized a child, the co-colonizing clones belonged to the same agr group (probability of mere coincidence=0.000074) suggests that bacterial interference is active in vivo during S. aureus colonization of atopic eczema. Interestingly, among four of the children followed in this study a shift from one agr group to another was observed between two visits. These agr group shifts were associated with a significant flare of eczema 19 (SE=7, p=0.009). Of note, three of the four agr group shifts were preceded by improvement in eczema, but followed by a subsequent flare. One might speculate that a low number of colonizing S. aureus lead to downregulation of virulence factors by negative quorum sensing (Novick, 2003Novick R.P. Autoinduction and signal transduction in the regulation of staphylococcal virulence.Mol Microbiol. 2003; 48: 1429-1449Crossref PubMed Scopus (955) Google Scholar) in the existing strain with ensuing facilitated settlement and spread of the newcomer. As the agr groups appear to reflect ancient phylogenic lineages, the newcomer conceivably produces a new set of antigenically significantly different virulence determinants. Because of a lack of neutralizing antibodies, increased inflammation may result. The subsequent decrease in eczema to "base-line" activity may be because of the eventual induction of such antibodies. All isolates were examined for a number of adhesins, toxins, and superantigens to study the possible effect of single virulence factors in aggravating dermatitis. No clear association was found between production of superantigens and/or β toxin and increased eczema. Two cross-sectional studies have shown more severe eczema in adults and children colonized with superantigen-producing S. aureus as compared with superantigen negative strains (Bunikowski et al., 2000Bunikowski R. Mielke M. Skarabis H. et al.Evidence for a disease-promoting effect of Staphylococcus aureus-derived exotoxins in atopic dermatitis.J Allergy Clin Immunol. 2000; 105: 814-819Abstract Full Text Full Text PDF PubMed Scopus (265) Google Scholar; Zollner et al., 2000Zollner T.M. Wichelhaus T.A. Hartung A. von Mallinckrodt C. Wagner T.O. Brade V. Kaufmann R. Colonisation with superantigen-producing Staphylococcus aureus is associated with increased severity of atopic dermatitis.Clin Exp Allergy. 2000; 30: 994-1000Crossref PubMed Scopus (158) Google Scholar). Two other studies, however, did not confirm this association (Arkwright et al., 2001Arkwright P.D. Cookson B.D. Haeney M.R. Sanyal D. Potter M.R. David T.J. Children with atopic dermatitis who carry toxin-positive Staphylococcus aureus strains have an expansion of blood CD5_B lymphocytes without an increase in disease severity.Clin Exp Immunol. 2001; 125: 184-185Crossref PubMed Scopus (8) Google Scholar; Mempel et al., 2003Mempel M. Lina G. Hojka M. et al.High prevalence of superantigens associated with the ecg locus in Staphylococcus aureus isolates from patients with atopic eczema.Eur J Clin Microbiol Infect Dis. 2003; 22: 306-309PubMed Google Scholar). The interplay between S. aureus virulence factors and eczema is highly complex and although the present study did not support a role for superantigens and/or β toxin, it does not rule out an effect of these in exacerbating atopic eczema under certain circumstances. It is conceivable that the number of bacteria colonizing a surface is crucial. Also, recent findings indicate that the total diversity of S. aureus superantigens may be extreme (Kuroda et al., 2001Kuroda M. Ohta T. Uchiyama I. et al.Whole genome sequencing of meticillin-resistant Staphylococcus aureus.Lancet. 2001; 357: 1225-1240Abstract Full Text Full Text PDF PubMed Scopus (1448) Google Scholar) with many more different superantigens than the classical ones examined so far. In conclusion, the children with atopic dermatitis showed variable patterns of colonization with S. aureus: a few are rarely, some intermittently, and some persistently colonized. Persistent colonization may be with a single clone or a dynamic exchange of several clones. Colonization with S. aureus was associated with more severe eczema, whereas no single virulence factor including the classical superantigens was associated with exacerbation of eczema. Bacterial interference via the agr diffusion sensor appears to play a role determining clonal dynamics in vivo with a change in agr group among colonizing clones significantly associated with a flare in eczema. The sample size of the present study is limited and future studies should be directed at detailed examinations of the interplay between the host immune and inflammatory systems and bacterial virulence factors at the time of shifts in agr groups to confirm and expand the present findings. Eleven children aged 2–11 y, eight boys and three girls, with mild-to-severe atopic dermatitis diagnosed according to the criteria ofHanifin and Rajka, 1980Hanifin J.M. Rajka G. Diagnostic features of atopic dermatitis.Acta Derm Venereol (Stockholm). 1980; 92: 44-47Google Scholar were each followed for a mean of 12 months (range 11–14) at the Department of Dermatology, Odense University Hospital, Denmark, in the period 2001–2002. Two pairs of siblings were included. Consecutive children with atopic dermatitis were included after the parents gave written informed consent. The children were seen approximately every 6th wk and whenever acute exacerbations of the eczema occurred. Treatments were administered according to usual procedures at the department with local corticosteroids, local antiseptics, or antibiotics, and courses of systemic antibiotics when needed. The study protocol was approved by the Ethics Committee of the Funen County, and the study was conducted according to the Declaration of Helsinki Principles. At every visit four swaps were taken with a coal impregnated cotton swap from the anterior nose, axillae, a representative area of active eczema, and the perineum. Samples were transported in Stuart's transport medium (Statens Serum Institut, Copenhagen, Denmark) and were received in the microbiological laboratory within 24 h for incubation on blood agar overnight at 37°C. Representatives of each colonial morphology of putative S. aureus were isolated and subcultured to purity for further analysis. The identity was confirmed by demonstrating coagulase activity. The activity of atopic eczema was determined by using SCORAD (values 0–103) (European Task Force on Atopic Dermatitis, 1993European Task Force on Atopic Dermatitis Severity scoring of atopic dermatitis: The SCORAD index.Dermatology. 1993; 186: 23-31Crossref PubMed Scopus (1600) Google Scholar). Genomic DNA fingerprinting was done by PFGE using standard technology. Each S. aureus isolate was harvested from Todd-Hewitt broth (CM189, Oxoid, Basingstoke, Hampshire, UK) and lysed within the agarose (Bio-Rad Laboratories, Hercules, California) by incubation for 2 h at 37°C in a lysis buffer (TE10:100 buffer; 1% sakrosyl; 1.3 × 104 U per mL lysozyme, Roche, Basel, Switzerland; 100 U per mL mutanolysin, M9901, Sigma-Aldrich, Brønoby, Denmark) and subsequently treated in a solution of proteinase K (1 mg per mL, 705723, Roche; 0.5 M EDTA; 1% sarkosyl; pH 8.0) at 50°C for 14–20 h. After washing in Milli Q water, the DNA-containing agar plugs were incubated in 100 μL 1 × SurE/Cut buffer A (1417959, Roche) containing 10 U of SmaI restriction enzyme (656348, Roche) for 16 h at room temperature. Separation of DNA fragments was performed on a GenePath Strain Typing System using a CHEF-DRIII power module (Bio-Rad Laboratories) according to the manufacturer's recommendations and using the pre-programmed run condition "PSU" (initial ramp time, 5.3 s; final ramp time, 34.9 s; linear progression in ramp time; angle, 120° angel switching; voltage, 6 V per cm; running time 19.5 h). Strains that were compared were tested in the same gel. Banding patterns developed with ethidium bromide were recorded by photography. To examine if failure to eradicate clones of S. aureus from the skin of patients by antibiotic therapy was because of antibiotic resistance, all isolates were examined for susceptibility to methicillin, erythromycin, and fucidic acid using ROSCO susceptibility disks on blood agar (Rosco, Taastrup, Denmark). Allocation of all S. aureus isolates to agr groups I, II, III, and IV was done by detecting PCR products using the four primer pairs and PCR conditions described byPeacock et al., 2002Peacock S.J. Moore C.E. Justice A. et al.Virulent combinations of adhesin and toxin genes in natural populations of Staphylococcus aureus.Infect Immun. 2002; 70: 4987-4996Crossref PubMed Scopus (457) Google Scholar. Gene sequences specific for exfoliative toxins A (eta) and B (etb) and the adhesins fibronection-binding protein A (fnbA) and collagen adhesin (cna) were detected by PCR using primer pairs and condition described byPeacock et al., 2002Peacock S.J. Moore C.E. Justice A. et al.Virulent combinations of adhesin and toxin genes in natural populations of Staphylococcus aureus.Infect Immun. 2002; 70: 4987-4996Crossref PubMed Scopus (457) Google Scholar. All isolates identified as S. aureus were examined for production of SEA, SEB, SEC, and SED and staphylococcal toxic shock syndrome toxin using the latex agglutination kits SET-RPLA and TST-RPLA according to manufacturer's instructions (Oxoid). β-toxin activity was demonstrated by the ability of individual isolates to allow detection of the CAMP reaction (Christie et al., 1944Christie R. Atkins E. Munch-Petersen E. A note on a lytic phenomenon shown by group B streptococci.Aust J Exp Biol Med Sci. 1944; 22: 197-200Crossref Google Scholar) in an isolate of Streptococcus agalactiae on sheep blood agar purchased from Statens Serum Institute (Copenhagen, Denmark). Proteolytic activity was demonstrated on skim milk agar inoculated with each strain in a line. The width of the clear zones developing in the medium around the bacterial growth after overnight incubation was rated (scores 0, 1, 2, and 3) to obtain a semi-quantitative estimate of the proteolytic activity. In order to describe the association between clinical and microbiological parameters on the same child, the data were analyzed by means of a longitudinal model. The standard model to describe such a correlation is the compound symmetry model in which the correlation is independent of the spacing in time between measurements performed on the same child. In order to allow the correlation to depend on the spacing in time, an autoregressive model was preferred, where the correlation between two measurements is given by ρd, where d is the spacing in time (Diggle et al., 2002Diggle P. Heagerty P. Liang K.-Y. Zeger S. Analysis of Longitudinal Data. 2nd edn. Oxford University Press, Oxford2002Google Scholar). Accordingly, because -1<p<1, the longer the time between two measurements, the weaker the correlation. The possible explanatory variables (agr group shift, clone shift, superantigen production, etc.) were added to the model in a stepwise manner in order to give a final "best" model. For the variables included in the final model, their presence was associated with either a positive or a negative change in the SCORAD value. The authors thank Biostatistician Mogens Erlandsen, Department of Biostatistics Aarhus University, Denmark, for statistical analysis, and technician Tove Findahl for performing the laboratory work. The study was supported by the Danish Medical Research Council. The researchers are all independent from funders. This study was approved by the Ethics Committee of the Funen County.