Pulsed-Field Gel Electrophoresis of Staphylococcus aureus Isolates from Atopic Patients Revealing Presence of Similar Strains in Isolates from Children and Their Parents

Abstract

Skin colonization with Staphylococcus aureus is often associated with atopic dermatitis, and staphylococcal enterotoxins have been implicated in the etiology of atopic disease. In this study, the colonization of patients with atopic dermatitis and their parents was investigated in order to evaluate the possibility of intrafamiliar transmission. S. aureus strains were isolated from 30 of 45 patients (66%). In 19 of 29 families (65%), at least one parent carried S. aureus, and the overall colonization rate of the parents was 48%. All strains were typed by pulsed-field gel electrophoresis (PFGE), and the presence of enterotoxin genes in the strains was assayed by multiplex PCR. A high percentage (84%) of the isolates present on the children and on at least one of their parents displayed identical PFGE and enterotoxin patterns as well as identical antibiotic resistance profiles, indicating intrafamiliar transmission. Forty-five percent of the strains did not carry any enterotoxin gene. The most frequently found enterotoxin genes were seg and sei, which were present in 36% of the strains, and seb, which was found in 24% of the strains. The other toxin genes occurred only in low frequencies. Most strains were resistant to penicillin (82%), and 15% showed resistance to more than one antibiotic. Intermediately-vancomycin-resistant S. aureus or methicillin-resistant S. aureus strains were not detected. In conclusion, this study indicates that the colonization rate of parents of atopic children is rather high and may increase the risk of recolonization of the child.

Staphylococcus aureus is a gram-positive pathogen that has been implicated in the pathogenesis of atopic dermatitis (AD). AD is one of the most frequent chronic inflammatory skin diseases and is found in 10% of all German children starting school (35). It is characterized by a dysregulation of the immune system that involves an increased Th2 response with an enhanced level of Th2 cytokines in the acute phase, a decreased level of other cytokines (e.g., tumor necrosis factor alpha), and decreased production of antistaphylococcal peptides of the innate immune system, i.e., β-defensin 3 and cathelicidins (12, 30). This peptide deficiency, the impaired skin integrity, and the increased expression of fibrinogen (9) favor colonization with S. aureus in AD, which is found in a higher percentage on the skin of affected children than on the skin of a healthy control group (23). Many strains of S. aureus are able to secrete superantigens, i.e., enterotoxins and toxic shock syndrome toxin. The production of superantigens may lead to an aggravation of AD (3), and a reduction of colonization has been shown to be effective in reducing the severity of the disease (14). Furthermore, exposure to superantigens may lead to the sensitization of the patient and the production of specific immunoglobulin E antibodies against enterotoxin A and enterotoxin B. These antibody titers were found to be elevated in patients with severe skin lesions (17, 25). Therefore, a reduction of S. aureus colonization in AD patients is desired. This study was conducted in order to assess the colonization of atopic patients and transmission of the strains between children and their parents. Moreover, whereas the roles of enterotoxins A, B, and C and toxic shock syndrome toxin in AD have been studied (7, 17), in the last years, a number of new enterotoxin genes have been detected in S. aureus (for a review, see references 21 and 40), and a further aim of this study was to determine the frequency of these genes in strains isolated from patients with AD. Besides, there is an increasing incidence of community-acquired methicillin-resistant S. aureus (34), and the first reports of infection of AD patients with methicillin-resistant strains in the far east (1, 16) and the United States (38) have been published, which led us to determine the antibiotic susceptibilities of our isolates.

MATERIALS AND METHODS

Strains and media.

A total of 67 isolates from patients and their parents who visited the outpatient clinic at the Department of Dermatology, University of Bonn, were characterized in this study. The samples were taken from lesional skin and the anterior nares (patients) or the anterior nares only (relatives) using Metaswab tubes (Mast, Brescia, Italy) and were cultured within 24 h on Columbia blood agar (Becton Dickinson, Heidelberg, Germany). The identification of S. aureus was carried out by testing the production of clumping factor and free coagulase. S. aureus strains were stored as frozen stocks in 50% glycerol at −70°C and were cultured in brain heart infusion medium (Oxoid, Wesel, Germany) at 37°C with aeration unless indicated otherwise.

PFGE.

Chromosomal DNA for the SmaI restriction digest was purified from the strains as described previously (12). Pulsed-field gel electrophoresis (PFGE) was performed using the Chef DRIII system (Bio-Rad, Munich, Germany) by employing pulsed-field certified agarose (1%) (Bio-Rad) at 6 V/cm, a field angle of 120°, and switch times of 5 to 15 s for 7 h and 15 to 60 s for a further 19 h. A chromosomal DNA digest of S. aureus NCTC8325 served as the mass standard.

Antimicrobial susceptibility tests.

Antimicrobial susceptibility tests against penicillin, methicillin, oxacillin, cefazolin, cefoxitin, imipenem, gentamicin, netilmicin, erythromycin, clindamycin, doxycycline, ciprofloxacin, levofloxacin, fosfomycin, teicoplanin, vancomycin, mupirocin, fusidic acid, and linezolid (Oxoid, Hampshire, England) were performed according to protocols recommended by the Clinical and Laboratory Standards Institute or DIN 58940-4 by employing the disk diffusion method or Etest (AB Biodisk, Solna, Sweden).

PCR of enterotoxin genes.

Attempts to perform a PCR employing whole cells were unsuccessful. Therefore, chromosomal DNA was purified using Genomic-Tips 20/G (Qiagen, Hilden, Germany) after cell lysis in the presence of lysostaphin (Sigma-Aldrich, Taufkirchen, Germany) or by employing the Instagene Matrix kit (Bio-Rad), both in accordance with the manufacturers' instructions. DNA was analyzed using a Nanodrop 1000 spectrophotometer (Peqlab Biotechnologie, Erlangen, Germany).

The presence of enterotoxin genes was screened by two multiplex PCR assays covering the sea, seb, sec, seh, and sej (group 1) genes or the sed, see, seg, and sei (group 2) genes. The primers seb-sec forward, seb reverse, sei forward, and sei reverse were described previously by Løvseth et al. (26), whereas all other primers were designed as described previously by Monday and Bohach (28), using the 16S rRNA gene as an internal control for both groups. The PCR conditions were 95°C for 10 min for the initial denaturation step, followed by 20 cycles (denaturation step at 95°C for 1 min, annealing step at 68°C for 45 s, and primer elongation step at 72°C for 1 min), which were followed by 30 cycles with an annealing step at 62°C and a final elongation step at 72°C for 10 min. When only sei or seg had been detected in a strain, the presence of the second gene was probed using additional alternative primers for seg and sei that had been designed according to nucleotide exchanges described previously (6): seghinmut (5′-ATGTCTCCACCTGTTGAAGG-3′), seihinmut (5′-CAACTTGAATTTTCAACCGGTACC-3′), and seirueckmut (5′-CAGGCAGACCATGTCCTG-3′).

These primers were employed in a conventional PCR (Go Taq polymerase; Promega GmbH, Mannheim, Germany) (annealing step at 60°C and 30 cycles). All positive multiplex PCR signals were verified by conventional PCR by employing the specific primers at an annealing temperature of 60°C for 30 cycles.

The following S. aureus reference strains were employed: S. aureus Mu50 (sea, sei, and seg) (20), S. aureus ATCC 14458/NCTC10654 (seb) (2), S. aureus LT 759/04 (Reference Centre for Staphylococci, Bonn, Germany) (sea and sec), S. aureus ATCC 23235/NCTC10656 (sed and sej) (8), and S. aureus ATCC 27664 (see) (5).

RESULTS

Colonization of atopic children and their families with S. aureus.

One hundred fifty-six samples originating from 45 patients and their relatives were evaluated in this study. Representative skin lesions and the anterior nares of the patients were swabbed. Forty-three S. aureus isolates were obtained from 30 patients (66%). Fifteen children were not colonized. The 30 colonized children (17 patients 1 to 3 years of age, 6 patients 4 to 6 years of age, 3 patients 7 to 9 years of age, and four patients 9 to 17 years of age; mean age, 4.8 years) belonged to 29 different families. Relatives who accompanied these patients to the clinic were also probed for colonization with S. aureus using nose swabs. For 23 families, both father and mother were examined. Here, 13 mothers (57%) and 9 fathers (39%) were positive for S. aureus, and in 19 of the 23 families, at least one parent was colonized (Table 1). In total, the colonization rate of the parents was 48%. In six cases, it was not possible to obtain a sample from the father.

TABLE 1.

Comparison of isolates obtained from different family members^a

Source of isolate	Localization	PFGE	Strain	Enterotoxin gene(s)	Antibiotic resistance
Patient 2	Nose	Unrelated	D2
Patient 2, mother	Nose		D3		Pen
Patient 3	Hand	Unrelated	D4		Pen, Fus
Patient 3, mother	Nose		D5		Pen
Patient 8	Nose	Similar	D12	A, B, C	Pen
Patient 8, mother	Nose
Patient 9	Nose	Similar	D13	A	Pen, Net
	Foot
Patient 9, mother	Nose
Patient 10	Nose	Similar	D14	G, I	Pen
	Eyelid
Patient 10, mother	Nose
Patient 11	Nose	Unrelated	D15	G, I	Pen
Patient 11, mother	Nose		D16	G, I	Pen
Patient 12	Nose	Similar	D17	G, I	Pen
Patient 12, father	Nose
Patient 13	Nose	Similar	D18	G, I	Pen
Patient 13, mother	Nose
Patient 14	Hand	Similar	D19		Pen
Patient 14, father	Nose
Patient 15	Nose	Similar	D20		Pen, Ery
Patient 15, mother	Nose
Patient 16	Nose	Similar	D21	B, G, I	Pen, Mup, Ery, Fus
	Hand
Patient 16, mother	Nose
Patient 16, father	Nose
Patient 17	Cheek	Similar	D22	B, G, I	Pen
Patient 17, father	Nose
Patient 18	Nose	Similar	D23	B	Pen
	Forehead
Patient 18, father	Nose
Patient 19	Cheek	Similar	D24	G, I	Pen
Patient 19, mother	Nose
Patient 20	Ear	Similar	D25	B	Pen
Patient 20, father	Nose
Patient 23	Nose	Similar	D28	A	Pen
	Arm
Patient 23, mother	Nose
Patient 24a	Hand	Similar	D29		Pen
Patient 24b	Face
Patient 24b	Nose	Similar	D30		Pen
Patient 24, mother	Nose
Patient 24, father	Nose	Related	D31
Patient 26	Nose	Similar	D34	B, G, I	Pen
	Hand
Patient 26, father	Nose
Patient 27	Nose	Related	D35		Pen
	Neck	Similar	D36
Patient 27, mother	Nose
Patient 27, father	Nose	Related	D37	B, G, I	Pen, Net

^aSimilar strains displayed identical PFGE patterns, whereas in related strains, at least one band shift had occurred. Abbreviations: Mup, mupirocin; Fus, fusidic acid; Pen, penicillin; Net, netilmicin; Ery, erythromycin.

All 67 isolates were characterized by PFGE. The analyses showed that in 16 (84%) of the 19 families where an isolate had been obtained from at least one parent, isolates with an indistinguishable PFGE pattern were present on the patient and his (her) mother or father. In 10 cases, the patient strain colonized the mother, and in six cases, a similar strain was present on the father. In three cases, strains with related patterns were isolated from the father. In one family (patient 16), all members were colonized by the same strain (Table 1). For 15 patients, isolates had been collected from infected skin and nose, and 11 patients (73%) were colonized with identical strains in both locations (Table 2). As a consequence, the S. aureus isolates obtained in this study could be grouped into 38 different strains.

TABLE 2.

Comparison of isolates obtained from different locations

Source of isolate	Localization	PFGE	Strain	Enterotoxin gene(s)	Antibiotic resistancea
Patient 4	Nose	Similar	D6	D, J
	Arm
Patient 5	Nose	Related	D7		Pen
	Knee		D8	B, G, I	Pen
Patient 6	Nose	Similar	D9		Pen
	Knee
Patient 7	Nose	Unrelated	D10	G, I
	Leg		D11	G, I
Patient 9	Nose	Similar	D13	A	Pen, Net
	Foot
Patient 10	Nose	Similar	D14	G, I	Pen
	Eyelid
Patient 16	Nose	Similar	D21	B, G, I	Pen, Mup, Ery, Fus
	Hand
Patient 18	Nose	Similar	D23	B	Pen
	Forehead
Patient 21	Nose	Similar	D26		Pen
	Arm
Patient 22	Nose	Similar	D27		Pen
	Foot
Patient 23	Nose	Similar	D28	A	Pen
	Arm
Patient 24b	Face	Related	D29		Pen
	Nose		D30		Pen
Patient 26	Nose	Similar	D34	B, G, I	Pen
	Hand
Patient 27	Nose	Unrelated	D35		Pen
	Neck		D36
Patient 28	Nose	Similar	D38	D, G, I, J	Pen, Net, Gen
	Perioral

^aAbbreviations: Mup, mupirocin; Fus, fusidic acid; Pen, penicillin; Net, netilmicin; Ery, erythromycin.

Distribution of enterotoxin genes.

All isolates were tested for the presence of enterotoxin genes by employing two multiplex PCRs. Although sei and seg are encoded on the same pathogenicity island, the enterotoxin gene cluster egc, frequently only sei or seg was detected using the primer pairs described previously (26, 28). This effect may be caused by the polymorphisms that were found in egc and that impede the annealing of some primers (6). Indeed, an analysis of the polymorphisms described and the primers used showed that some strains harbor a one-base exchange in the 5′ terminus of primer “seg forward” as well as in the sei primers. The design of new primers led to a complete detection rate of sei and seg genes.

Seventeen strains (45%) did not carry any of the enterotoxin genes tested (Tables 1 and 2). For the enterotoxigenic strains, the seg/sei combination was found most frequently and was present in 14 strains (36%). Furthermore, nine strains (24%) were characterized by the enterotoxin B gene seb. The other enterotoxins were present in lower frequencies, ranging from 2.6% to 5.2% of the isolates; see and seh were not found (Fig. 1). If the enterotoxin gene cluster (egc) that carries seg and sei is counted as one enterotoxin locus, then 13 strains carried one enterotoxin-encoding locus, 6 strains were characterized by two loci, and 2 strains harbored three different enterotoxin loci.

FIG. 1.

Percent frequency of enterotoxin genes in 40 genetically distinct strains of S. aureus. The first bar shows the frequency of strains that carried none of the genes tested.

Antibiotic resistance.

The antibiotic resistances of all 67 isolates were tested. The isolates that belonged to single strains did not differ in their susceptibilities. Eighteen percent of the strains did not show any resistance, and the majority of the strains (31 strains [82%]) were resistant to penicillin (Tables 1 and 2). Only a few strains additionally showed resistance to other antibiotics, i.e., erythromycin, gentamicin, netilmicin, mupirocin, and fusidic acid. The percentage of multiply resistant strains was low (four strains were resistant to two substances, and two single strains were resistant to three and four agents, respectively). The strain that was resistant to four agents, penicillin, fusidic acid, mupirocin (MIC > 1,024 mg/liter), and erythromycin, colonized both parents and the child and carried the enterotoxin genes seb and egc. Methicillin-resistant S. aureus or strains showing intermediate vancomycin resistance were not detected.

DISCUSSION

Colonization of atopic patients with S. aureus has been well documented, and some older studies indicated colonization rates well over 90% (10, 24). In Bonn, only 66% of the patients were found to be colonized with S. aureus; however, this result corresponds to other studies performed in the last years that detected between 57% and 64.2% of colonized patients (16, 33, 39). It has been shown that colonization with S. aureus depends on the Th2 response and that even anti-inflammatory treatment will reduce skin colonization (13), indicating that colonization is dependent on the severity of the disease and that pretreatment with steroids or antibiotics will inhibit the growth of S. aureus. In contrast to other studies (see, e.g., reference 27), the antibiotic and steroid treatments had not been suspended before the samples were taken, which may explain the rather low isolation rate.

The colonization rate for the parents (48%) was higher than that of the average population. For example, a study that elucidated the colonization rate of young menstruating women found only 26% colonization (31); in contrast, here, 57% of the mothers of atopic children were colonized. There may be multiple reasons for the high colonization rates of the parents: (i) the treatment of the infected lesions is normally carried out by the parents and offers excellent opportunities for transmission of the bacteria, (ii) there is close physical contact between parents and especially small children, and (iii) the parents of atopic children are often characterized by susceptibility to atopic disease as well and therefore may be prone to colonization by S. aureus. In this context, it is interesting that one father of our group still suffered from atopic eczema.

In order to investigate whether transmission within the families had taken place, the bacteria were typed. The PFGE, enterotoxin, and antibiotic resistance patterns showed that in the large majority of the families (84%), identical strains had been isolated from the child and his or her parent(s), indicating transmission within the family. In one case, the strain isolated from the skin of the child was identical to the strain that was present in the nose of the mother, whereas a different strain colonized the nose of the child itself. Even if the strains were not identical, related strains were often present within one family, indicating that strain transmission had taken place some time ago. For 73% of the patients, identical strains were obtained from nose and skin, and in 13% of the cases, related strains were present. This high frequency of identical and related strains is best explained by the fact that nasal carriage of S. aureus has been shown to be associated with hand carriage (32, 37), and the children may transfer the nose strains to their lesions in this way.

The production of enterotoxins by strains that colonize atopic patients has been well studied, but most analyses examined only the excretion of enterotoxins A, B, and C and toxic shock syndrome toxin 1 (41, 42). In most of those studies, seb was the most commonly found enterotoxin gene (42), and likewise, seb was one of the most frequently encountered enterotoxin genes seen in our strain collection; it was found in 9 of the 38 strains in this study.

However, the most frequently isolated enterotoxin genes were seg and sei, indicating the presence of the egc enterotoxin gene cluster. In 14 of 38 strains, both genes were detectable, giving an isolation rate of 37% for all strains or of 66% with regard to 21 enterotoxin-positive strains. Recent studies indicated that egc seems to be rather frequently found in animal isolates (25%) (36) and even more so in human isolates (55%) (4). Mempel et al. (27) previously tested for the presence of egc in isolates from patients with AD and also found the gene cluster in 48% of the isolates. Therefore, the frequency of egc in our isolates is not surprising.

In addition to enterotoxin I and enterotoxin G, egc often encodes the enterotoxin-like superantigens M, N, and O (sem-seo) (19) and sometimes enterotoxin U (22). It has been demonstrated that the egc enterotoxins are able to stimulate T-cell proliferation, and some cases of egc-mediated staphylococcal toxemias have been observed (18). However, although the enterotoxins of the egc gene cluster seem to be the most common staphylococcal enterotoxins, the serum levels of neutralizing antibodies against these toxins are lower than those against the classical enterotoxins (sea to sed) (15), and the toxins seem to be produced in lower quantities (29). The impact of these enterotoxins on AD remains to be evaluated.

It has been demonstrated that skin colonization with S. aureus is associated with an exacerbation of eczema in atopic disease, and therefore, current therapeutic approaches include anti-inflammatory and antimicrobial treatment in order to reduce or eliminate colonization (13, 14). Failure of treatment and recolonization of the skin may result from nasal carriage, antibiotic resistance of the strains, and contamination during treatment (11). The results demonstrate that the noses and/or skin of many children and their parents were colonized and indicate that intrafamiliar spread had occurred. In conclusion, it should be determined whether the children might benefit from a surveillance of the carrier status of their own noses and those of their family members in order to avoid intrafamiliar “ping pong” infections.