Clinical and Experimental Immunology REVIEW doi:10.1111/j.1365-2249.2005.02980.

The role of microorganisms in atopic dermatitis

B. S. Baker Summary
Atopic dermatitis (AD) is a common, fluctuating skin disease that is often
associated with atopic conditions such as asthma and IgE-mediated food
allergy and whose skin lesions are characterized by a Th-2 cell-mediated
response to environmental antigens. The increasing prevalence and severity of
atopic diseases including AD over the last three decades has been attributed to
decreased exposure to microorganisms during early life, which may result in
an altered Th-1/Th-2-balance and/or reduced T cell regulation of the immune
response. Patients with AD exhibit defects in innate and acquired immune
responses resulting in a heightened susceptibility to bacterial, fungal and viral
infections, most notably colonization by S. aureus. Toxins produced by S.
aureus exacerbate disease activity by both the induction of toxin-specific IgE
and the activation of various cell types including Th-2 cells, eosinophils and
keratinocytes. Allergens expressed by the yeast Malazessia furfur, a compo-
nent of normal skin flora, have also been implicated in disease pathogenesis in
a subset of AD patients.
Microorganisms play an influential role in AD pathogenesis, interacting
Accepted for publication 14 October 2005 with disease susceptibility genes to cause initiation and/or exacerbation of
E.mail: [email protected] disease activity.

CARD4/Nod1, CARD15/Nod2, TLR2 and TLR4 [9–11] have

Introduction
been implicated in AD. Environmental factors that interact
Atopic dermatitis (AD) is a common, chronic fluctuating with susceptibility genes in AD, inducing the production of
skin disease with prevalence in children of between 7 and IgE antibodies and activation of Th-2 cells, include foods,
17%. The disease presents commonly within the first 2 years house dust mite (HDM) allergens, secondary microbial
of life, and in approximately two-thirds of cases will persist infections and stress. This review focuses on the role of
into adulthood. A high proportion of infants with AD have a microorganisms in AD, and discusses the mechanisms
positive family history of AD, asthma or allergic rhinitis in involved in the prevention, initiation and exacerbation of
one or both parents, and will go on to develop further atopic disease activity by microbial components.
complications later in childhood. AD is a multifactorial dis-
ease in which both hereditary and environmental factors
The hygiene hypothesis
play a role. Four chromosomal regions of linkage to the dis-
ease, which differ from atopy-associated loci, have been Over the last 30 years there has been a continuously increas-
identified in genome scans of children with AD [1,2]. Inter- ing prevalence of AD and other atopic allergic disorders,
estingly, these linkage sites closely correspond to four loci especially in western industrial countries where it can be as
reported to contain susceptibility genes for psoriasis suggest- high as 20–37% of the population [12]. Environmental fac-
ing that they may contain genes that control skin inflamma- tors such as increased air pollution, indoor exposure to
tion and immunity in both diseases [2]. In addition, several HDM antigens in less ventilated modern homes, and dietary
candidate genes coding for proteins with immunological changes have been implicated, but there is little consistent
function, such as the high affinity IgE receptor present on evidence that these factors account for the observations. A
mast cells, Th-2 cytokines, RANTES and IRF2 [3–8], and more relevant factor appears to be a reduction in the expo-
receptors involved in the response to microorganisms, sure to infections during early childhood, which may result

© 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9 1
B. S. Baker

from modern lifestyle factors such as the use of antibiotics, a

MECHANISM 1
reduction in family size (allergic sensitization is higher in
the first-born, but is less frequent in children from large
families), and an increase in hygiene and living standards Reduced
[13]. The relationship between decreased exposure to Pathogens/PAMPS
microbial antigens associated with a western lifestyle and the
increasing severity and prevalence of atopic diseases has
become known as the ‘hygiene hypothesis’ [14]. Innate
immune cells, such as macrophages and dendritic cells, PRR Reduced
express pattern recognition receptors (PRRs) that recognize
IL-12
pathogen-associated molecular patterns (PAMPs) associated
with microorganisms; activation of these receptors induces a Th1
Th-1 type response. It was therefore suggested that a lack of
microbial antigen-induced immune deviation from the Th- T
Th2
2 cytokine profile that predominates at birth to a Th-1 type IL-4 Naive
profile could explain the development of enhanced Th-2 cell
responses to allergens (Fig. 1) [15]. However, this explana-
tion did not did not take into account the fact that chronic
parasitic worm (helminth) infections which induce strong ALLERGY
Th-2 responses and high IgE levels are not associated with
allergy, or that the prevalence of Th-1-associated autoim-
mune diseases have also increased at the same time as aller-
gic diseases [16].
Subsequently, a second mechanism was proposed for the MECHANISM 2
switch to an atopic phenotype; a defect in the stimulation of
dendritic cells by nonpathogenic microorganisms in gut- Reduced
associated lymphoid tissue leading to reduced production of Non-Pathogens/PAMPS
IL-10-producing regulatory T cells (Fig. 1) [16,17]. Differ-
ences in the bacterial colonization of the gut have been
reported in children with AD; Enterococci and Bifidobacteria
were reduced, and Clostridia and Staphylococcus aureus
PRR
numbers increased [18,19]. These changes have been attrib- Reduced
uted, at least in part, to a frequent use of antibiotics, which Tr activation
has consistently been shown to markedly increase the risk of + +
developing AD, even into the antenatal period [20]. This has CD4 CD25
led to the use of diet supplementation with Lactobacillus Tr1
(marketed as probiotics) in both the prevention and treat- Th3
ment of AD [20]. A randomised, controlled trial in which
oral Lactobacillus CG supplementation was given prenatally
to mothers with an allergic family history, and postnatally
for 6 months to their offspring, resulted in an almost 50%
Th1 Th2
reduction in AD frequency [21]. Furthermore, a significant
reduction in SCORAD (disease severity scoring of atopic
dermatitis) after one month of Lactobacillus GG supplemen-
tation was observed in a randomised controlled trial of 31
infants with AD and cow’s milk intolerance [22]. It was
Autoimmune Allergy
subsequently demonstrated that L. reuteri and L. casei
prime monocyte-derived dendritic cells, via DC-SIGN Diseases
(DC-specific intercellular adhesion molecule 3-grabbing Fig. 1. Two possible mechanisms for the increased prevalence of allergy
nonintegrin), to drive the development of IL-10-producing as a result of decreased exposure to microbial antigens.
regulatory T cells [23]. These findings suggested a possible
mechanism for the beneficial effects of Lactobacillus admin- Although a recent review of 64 studies published between
istration in AD, and provided support for the proposal that 1966 and August 2004 failed to find any convincing evidence
reduced T cell regulatory activity may be responsible for the that exposure to a specific infection such as tuberculosis
increased prevalence of allergy. reduces the risk of AD, there was, in contrast, some evidence

2 © 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9
 The role of microorganisms in atopic dermatitis

that common viral and bacterial childhood infections were increased susceptibility to infection [28]. Indeed, IgE has
positively associated with an increased risk of AD expression been shown to inhibit neutrophil adhesion, phagocytosis
[20]. However, a protective effect was associated with endot- and respiratory burst, which may affect clearance of micro-
oxin exposure in a farming environment, day care atten- organisms from the skin [32]. A major factor in increased S.
dance or having a dog during infancy, situations in which aureus colonization of AD skin is a defective cutaneous
chronic (nonpathological) microbial stimulation may take innate immune response, involving decreased production of
place. Thus it is possible that the effects of infections in early antimicrobial peptides and the expression of functional
childhood on the development of AD may depend partly variants of the TLR and Nod/CARD receptors for microbial
upon the nature and degree of exposure, rather than components (Fig. 2).
infection per se. Clearly further research into the early prim-
ing of a child’s immune system by microorganisms is neces-
sary in order to establish the factors required to induce the
Defective innate immunity in AD: antimicrobial
development of AD.
peptides
The innate immune system of the epidermis is the first line of
Microbial colonization of AD skin
defence against invasion by microorganisms, which gain
Patients with AD are highly susceptible to certain cutaneous entry after the skin is damaged. Anti-microbial peptides
bacterial, fungal and viral infections [24]. AD patients have form part of this defence system, three of which are triggered
an increased incidence of warts caused by the human papil- by injury or inflammation of the skin: the β defensins HBD-
lomavirus, and of cutaneous fungal infections such as that 2 and HBD-3, and a cathelicidin, hCAP18/LL-37 [33]. HBD-
caused by Trichophyton rubrum. They are particularly sus- 2 shows microbicidal activity against predominately Gram-
ceptible to severe infections caused by herpes simplex type 1 negative organisms such as E. coli and P. aeruginosa, and
virus (eczema herpeticum or Kaposi’s varicelliform), vac- yeasts, but is relatively ineffective against Gram-positive bac-
cinia virus (eczema vaccinatum) coxsackieA virus and teria such as S. aureus [34]. In contrast, HBD-3 and hCAP18/
molluscum contagiosum virus. These viral infections can LL-37 are more potent, broad-spectrum antibiotics that kill
represent serious complications in AD, and if not treated both Gram-positive and Gram-negative organisms and the
promptly have the potential to be life threatening. However, yeast C. albicans [35,36]. In addition, HBD-2 can enhance
bacterial colonization with Staphylococcus aureus is the most the innate immune response of the epidermis, and provide a
common skin infection in AD (> 90% of patients compared link with the acquired immune response by inducing up-reg-
to 5% of normal individuals) and occurs on both lesional ulation of costimulatory molecules and the maturation of
and, to a lesser extent, nonlesional AD skin [25,26]. Further- immature dendritic cells in a TLR4-dependent manner [37].
more, there appears to be a causative relationship between Beta-defensins can also act as chemoattractants for imma-
the numbers of bacteria present on the skin and the severity ture dendritic cells and memory T lymphocytes via the
of disease in AD patients, whilst treatment-induced removal chemokine receptor CCR6 [38].
of the bacteria is associated with improvement in skin lesions Immunostaining, measurement of specific mRNA by real-
in most cases [27,28]. Various factors are involved in the time reverse-transcriptase-PCR or GeneChip microarray
altered skin colonization by S. aureus in AD including an analysis for HBD-2, HBD-3 and hCAP18/LL-37 in acute and
altered epidermal barrier, increased bacterial adhesion, chronic lesions from patients with AD showed a significant
defective bacterial clearance, and decreased innate immune decrease in expression as compared to that of psoriasis
responses. patients [39,40]. IL-4 and/or IL-13, has been shown to sup-
S. aureus are tightly attached to the uppermost corneo- press the TNF-α or IFN-γ-induced up-regulation of HBD-2
cytes, and can penetrate the epidermis via the intercellular and HBD-3 mRNA in keratinocytes and normal skin
spaces probably as a result of lipid deficiencies in AD skin. explants suggesting that the reduced levels of antimicrobial
In AD, the average pH of the skin is slightly more alkaline, peptides may be explained by the predominance of Th-2
and sphingosine levels are decreased in both lesional and type cytokines in AD skin lesions [40]. Clinical isolates of S.
nonlesional stratum corneum [29,30]. In addition, the dry- aureus from AD patients can be killed by a combination of
ness and cracking of AD skin, as a result of transepidermal HBD-2 and hCAP18/LL-37 at the concentrations found in
water loss caused by altered lipid content, may facilitate psoriatic lesions, but the levels present in AD skin are too low
bacterial colonization. Furthermore, Th-2 cytokines such as to be effective.
IL-4 in atopic skin increase expression of fibronectin and The innate skin defence system of patients with AD is
fibrinogen, receptors that mediate the adhesion of S. aureus further compromised by a deficiency of dermcidin-derived
to stratum corneum [31]. In a proportion of AD patients antimicrobial peptides in sweat, which correlates with infec-
who respond poorly to anti-inflammatory treatment, per- tious complications [41]. Dermcidin, a peptide with no
sistent S. aureus colonization is associated with higher total homology to other antimicrobial peptides, is specifically and
IgE levels suggesting that IgE may contribute to an constitutively expressed in sweat glands in the dermis of skin,

© 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9 3
B. S. Baker

Altered
barrier MICROORGANISMS
function

Low levels of
antimicrobial
molecules Altered PRR expression

HBD 2/3
LL-37
IL-8, iNos

ind
its
sweat

ib

uc
h
in

es
Fig. 2. Defective innate immunity in AD duct
skin.Altered intracellular and extracellular PRR IL-4 IFN-γ
expression, and decreased antimicrobial mole- IL-13 TNF-α
low dermcidin
cule production (partly due to inhibition by Th- low IgA
2 cytokines) results in an impaired innate
immune response to microorganisms, which gain
entry to the epidermis as a result of an altered
barrier function. HBD-2/HBD-3, human β-
defensin-2/3; LL-37, a member of the cathelicidin sweat
family; iNos, induced nitric oxide synthetase. gland

secreted into sweat and transported to the epidermal surface ines, antimicrobial peptides and inducible enzymes in the
[42]. In common with HBD-3 and hCAP18/LL-37, dermci- skin, via activation of transcription factors, activator protein
din has a broad spectrum of activity against a variety of patho- (AP)-1 and nuclear factor (NF)-κB [47]. Two single
genic microorganisms. In healthy individuals, a significant nucleotide polymorphisms (SNPs) have been described for
reduction in viable bacterial cells on the skin surface occurs each of the receptors, which result in changes in amino acid
after sweating, but this is not the case in patients with AD [41]. sequences. One of the TLR2 polymorphisms (Arg753Gln),
Furthermore, the rate of sweat production, and the secretion which is located within the intracellular part of the receptor
of IgA in sweat are reduced in AD patients contributing to the and has been particularly associated with S. aureus infec-
impaired innate immune response [43,44]. tions, was found to be present in a higher frequency in AD
The expression of the innate immune response genes, IL- patients compared to controls [11]. The subgroup of AD
8 (CXCL8) and induced nitric oxide synthetase (iNos) was patients carrying this polymorphism had increased disease
also decreased in AD compared to psoriatic skin [40]. IL-8 is severity characterized by markedly elevated IgE antibodies to
a chemokine that attracts polymorphonuclear leucocytes S. aureus superantigens and HDM allergens. In addition, a
into the skin where they phagocytose and kill bacteria, whilst further subgroup of AD patients expressed a higher fre-
iNos can kill viruses, bacteria and fungi through production quency of the two TLR4 polymorphisms, Asp299Gly and
of nitric oxide. In common with the antimicrobial peptides, Thr399Ile, than controls [11]. These cosegregating polymor-
production of IL-8 and iNos is also inhibited by Th-2 phisms, located in the extracellular domain of the receptor,
cytokines. have previously been reported in patients with septic shock,
particularly that induced by Gram-negative bacteria, and are
linked to LPS hyporesponsiveness [48].
Defective innate immunity in AD: TLR and
Intracellular PRRs are represented by the Nod (nucle-
Nod/CARD proteins
otide-binding oligomerization domain) family, which
Toll-like receptor 2 (TLR2) and TLR4 are members of a includes Nod2/caspase recruitment domain containing pro-
family of PRRs that recognize various conserved microbial tein (CARD) 15 and the closely related Nod1/CARD4
components or PAMPs. TLR2 recognizes components of protein. Both proteins detect peptidoglycan, the major
Gram-positive bacteria and yeasts, such as lipotechoic acid, component of the bacterial cell wall, although the specifici-
peptidoglycan (although recent evidence suggests that this ties of the two receptors are distinct and nonoverlapping
may be contaminating lipoteichoic acid [45]), lipoproteins [49]. Polymorphisms in the Nod2/CARD15 gene that result
or zymosan, whilst the PAMPs detected by TLR4 include the in changes in peptidoglycan recognition have been reported
Gram-negative bacterial component, LPS [46]. Recognition to be associated with susceptibility to Crohn’s disease, a Th-
of PAMPs by TLRs initiates a signalling cascade that results 1-mediated inflammatory disease of the bowel [50,51].
in the production of proinflammatory cytokines, chemok- Three of these Crohn’s-associated Nod2/CARD15 polymor-

4 © 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9
 The role of microorganisms in atopic dermatitis

phisms have been investigated in children with asthma and response, which is largely mediated by the release of staphy-
allergy by PCR-based restriction enzyme assays [10]. Chil- lococcal enterotoxins (SE), such as SEA, SEB and toxic shock
dren with the polymorphic C2722 allele had a more than 3- syndrome toxin (TSST)-1, also referred to as ‘superantigens’
fold risk to develop allergic rhinitis and an almost 2-fold risk [53] (Table 1). SEB applied to intact normal skin or the non-
for AD. More recently, an association of Nod1/CARD4 poly- lesional skin of patients with AD can induce erythema and
morphisms with AD have been reported in a study that dermatitis, and, in some AD patients, a flare of their disease
examined the effects of 11 SNPs, covering the complete gene, in the elbow flexure of the same arm to which the toxin was
on atopy phenotypes [9]. One Nod1/CARD4 haplotype and applied [54]. Furthermore, 14 of 68 patients recovering from
three polymorphisms (rs2907748, rs2907749, rs2075822) toxic shock syndrome caused by TSST-1, but no patients
were significantly associated with AD in a population-based recovering from Gram-negative sepsis, developed chronic
cohort, case-control population, and/or family–based asso- eczematous dermatitis [55]. These findings suggest that
ciation analysis. These polymorphisms were also associated superantigens can initiate, exacerbate and maintain inflam-
with asthma and total serum IgE levels, but not with allergic mation associated with AD.
rhinoconjunctivitis or specific sensitization. The major effects of staphylococcal superantigens in AD
Peptidoglycan from S. aureus has been shown to induce are likely to be mediated via the polyclonal activation of
the production of various keratinocyte-derived mediators superantigen-specific TCR Vβ families of T cells. T cells
including GM-CSF, a cytokine that is overproduced in AD expressing Vβ chains specific for the superantigen accumulate
skin lesions [52]. It remains to be established whether this selectively in superantigen-treated skin, but not in skin
occurs via PRR stimulation; however, keratinocytes are treated with sodium lauryl sulphate, supporting this
known to express several members of the TLR family [47], hypothesis [56]. In the peripheral blood of AD patients whose
and intracellular Nod/CARD receptors may also be present skin is colonized by superantigen-secreting S. aureus, a rele-
in these cells. vant skewing of superantigen-reactive Vβ families was
Thus the reduced production of antimicrobial peptides observed in CD4+ and CD8+ T cells coexpressing cutaneous
and other innate immune factors, together with an impaired lymphocyte antigen (CLA), a skin homing receptor [57].
recognition of microbial antigens as a result of functional Superantigens up-regulate CLA expression by T cells via stim-
polymorphisms in the genes coding for PRRs, are major fac- ulation of IL-12 production, thus promoting their homing to
tors contributing to the susceptibility to infection and result- the skin [58].
ing exacerbation of disease activity in patients with AD. T cell activation by superantigens may be further aug-
mented by an inhibitory effect (at least by SEB) on the
immunosuppressive activity of circulating CLA+CD4+ CD25+
Cellular activation by staphylococcal superantigens
regulatory T cells, which, surprisingly, are increased in
The main consequence of increased colonization of AD skin patients with AD [59]. Furthermore, SEB-reactive (Vβ3+,
by S. aureus is exacerbation of the inflammatory immune Vβ12+ or Vβ17+) CD4+ T cells producing Th-2 cytokines in

Table 1. Role of staphylococcal superantigens in AD.

Superantigen-induced effects Functional consequences
Application of SEB to uninvolved skin Induces local erythema and dermatitis
Sometimes causes disease flare
Superantigen-specific IgE production Release of histamine by mast cells and basophils
Facilitates toxin presentation by Langerhans cells to T cells
Correlates with disease severity
SEC1-specific IgG2 production Decreased in AD; may affect phagocytosis of S. aureus by
polymorphonuclear leucocytes
Selective Vβ family expansion in lesional skin and blood Activation of Th-2 cells in a Vβ-specific manner
CLA expression on activated Th-2 cells, circulating and in lesional skin Induces homing of T cells to skin
Apoptosis of superantigen-reactive T cells SEB-reactive Th-2 cells in AD are apoptosis-resistant, thus helping to
maintain information
Inhibition of CLA+CD4+CD25+ regulatory T cell activity Increases inflammation
Dermal eosinophils Inhibits apoptosis, increases expression of surface activation antigens
and enhances oxidative burst
Langerhans cells and macrophages Stimulates production of IL-1, TNF-α and IL-12
HLA-DR+ keratinocytes Transient Ca2+ mobilization
TNF-α production
Presentation of superantigen to Th-2 cells

© 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9 5
B. S. Baker

AD patients are more resistant to SEB-induced apoptosis Conversely, a subgroup of patients with AD showed a defi-
than corresponding SEB-reactive Th-1 cells from healthy ciency in the production of IgG2, but not of IgG1 or IgG4
individuals [60]. antibodies against toxin SEC1, which was associated with a
In addition to T cells, superantigens can also mediate severe disease phenotype [71]. This appeared to be specific
effects on other cell types such as eosinophils, Langerhans for SEC1 as levels of IgG2 antibodies against SEB, or another
cells, macrophages and keratinocytes. During flares of AD, bacterial antigen, pneumococcal capsular polysaccharide,
eosinophils are recruited to the skin by chemoattractants were normal in these patients. IgG2 antibodies are poor com-
such as RANTES (regulated on activation, normal T plement activators, but can effectively mediate polymorpho-
expressed and secreted) and eotaxin, where they are acti- nuclear leucocyte phagocytosis of S. aureus via the FcγRIIa
vated and undergo degranulation and cytolytic degenera- receptor on the cell surface [72]. Although the functional
tion, the products of which promote inflammation and role of anti-SEC1 IgG2 antibodies in AD pathogenesis
tissue damage [61,62]. It has been demonstrated that SEB in remains unclear, it is possible that such a selective decrease in
AD skin lesions is localized predominately to eosinophils in some patients with AD may contribute to the persistence of
the dermis, as well as to a lesser extent on Langerhans cells SEC1-producing S. aureus on lesional skin and the resulting
and IgE-bearing cells [63]. Superantigens modulate the exacerbation of disease activity.
effector function of eosinophils, and probably the course of
AD, by inhibiting eosinophil apoptosis, increasing expres-
Malassezia-induced IgE- and Th-2 cell-mediated
sion of activation antigens on the eosinophil surface, and
responses
enhancing oxidative burst of eosinophils in vitro [64]. They
also bind to HLA-DR on Langerhans cells and macrophages Malassezia (formerly known as Pityrosporum orbiculare/
and stimulate them to produce IL-1, TNF-α and/or IL-12. ovale) is part of the normal human skin flora and is most
These cytokines either up-regulate the expression of adhe- abundant at sites of high sebum production such as the scalp,
sion molecules on endothelial cells, or increase CLA expres- chest and back where it colonizes the stratum corneum.
sion on T cells, respectively, thus facilitating the recruitment Most healthy individuals have developed IgG antibodies to
of CLA+ memory T cells to the skin. In addition, kerati- Malassezia, but in 30–80% of AD patients, IgE and/or T cell
nocytes that have been induced to express MHC Class II reactivity to the organism is present [73]. Patients with AD
molecules by stimulation with IFN-γ can interact with affecting mainly the head and neck region appear to be more
superantigens, resulting in transient intracellular Ca2+ likely to produce Malassezia-specific IgE antibodies, coincid-
mobilization and the release of proinflammatory TNF-α ing with the higher levels of yeast colonization in these areas,
[65,66]. HLA-DR+ keratinocytes can also present superanti- than patients with AD located elsewhere on the body [74].
gens to T cells; because KCs do not synthesize IL-12, this Malassezia-specific IgE antibodies are rarely produced by
results in the activation of Th-2 rather than Th-1 cells [67]. atopic patients whose skin is unaffected [75,76]. This,
Two other S. aureus products, staphylococcal protein A and together with the high prevalence of type I hypersensitivity
α-toxin also induce the production of TNF-α by kerati- to Malassezia as compared to other fungi in AD suggests that
nocytes, with additional cytotoxic effects exerted by the lat- Malassezia-specific antibodies are pathogenically significant
ter [68]. [77,78]. Histamine release tests have confirmed the biologi-
cal activity of circulating Malassezia-specific IgE antibodies
in approximately 70% of AD patients, supporting a role for
Superantigen-specific IgE and IgG2 antibodies
these antibodies in the disease process [78].
Superantigen-secreting S. aureus have been isolated from The defective skin barrier in AD may allow the whole
over 50% of AD patients, and many of these patients produce Malassezia yeast cells and their allergens, nine of which have
IgE antibodies specific for the toxins found on their skin been isolated and cloned so far, to enter the skin and be taken
[69,70]. In contrast, although SEA, SEB or TSST – secreting up by Langerhans cells. In vitro studies have shown that the
S. aureus have also been isolated from the lesional skin of process of internalization of Malassezia causes maturation of
psoriasis patients, their sera did not contain IgE antibodies dendritic cells and the production of proinflammatory and
to the toxins [69]. Basophils and mast cells from AD patients immunoregulatory cytokines, but not IL-12, thereby favour-
with antitoxin IgE antibodies release histamine on exposure ing the induction of a Th-2 type response [79]. Indeed,
to toxins, but only those toxins against which they have higher blood and skin Th-2 type responses to Malassezia
raised specific IgE antibody [69]. Furthermore, there is a cor- have been demonstrated in vitro in AD patients compared to
relation between the presence of IgE antibodies specific for that of normal controls [80]. Furthermore, it has been dem-
staphylococcal superantigens and both the severity of AD, onstrated that a positive atopy patch test to Malassezia in vivo
and total serum IgE levels [71]. Thus toxins produced by S. correlates with a Th-2-like peripheral blood mononuclear
aureus also exacerbate AD by activating mast cells, basophils cell response in AD patients in vitro [81].
and other Fcε-receptor bearing cells carrying the relevant Malassezia also exerts other proinflammatory effects such
antitoxin IgE antibody. as activation of the alternative complement pathway, and the

6 © 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9
 The role of microorganisms in atopic dermatitis

stimulation of keratinocytes to produce a variety of inflam- 8 Nishio Y, Noguchi E, Ito S, Ichikawa E, Umebayashi Y, Otsuka F,
matory cytokines such as IL-6, IL-8 and TNF-α, which may Arinami T. Mutation and association analysis of the interferon reg-
contribute to its role in AD pathogenesis [82,83]. However, ulatory factor 2 gene (IRF2) with atopic dermatitis. J Hum Genet
2001; 46:664–7.
antifungal treatment studies have so far proved inconclusive
9 Weidinger S, Klopp N, Rummler L et al. Association of NOD1
as to the clinical relevance of Malassezia-induced allergy in
polymorphisms with atopic eczema and related phenotypes. J
AD. This may be explained by a lack of selection of relevant
Allergy Clin Immunol 2005; 116:177–84.
patient subgroups and the use of inappropriate measures of 10 Kabesch M, Peters W, Carr D, Leupold W, Weiland SK, von Mutius
clinical outcome [84]. E. Association between polymorphisms in caspase recruitment
domain containing protein 15 and allergy in two German popula-
tions. J Allergy Clin Immunol 2003; 111:813–7.
Conclusion
11 Ahmad-Nejad P, Mrabet-Dahbi S, Breuer K et al. The Toll-like
AD is a chronic inflammatory skin disease whose initiation receptor 2 R753Q polymorphism defines a subgroup of patients
and clinical activity is modified by exposure to, and interac- with atopic dermatitis having severe phenotype. J Allergy Clin
tion with, microorganisms. In genetically predisposed indi- Immunol 2004; 113:565–7.
12 Ring J, Kramer U, Schafer T, Behrendt H. Why are allergies increas-
viduals a lack of chronic exposure to microbial antigens in
ing? Curr Opin Immunol 2001; 13:701–8.
early life increases the risk of disease. Once manifested,
13 Williams HC, Strachan DP, Hay RJ. Childhood eczema: disease of
defects in skin immune surveillance mechanisms such as the advantaged? Br Med 1994; 308:1132–5.
decreased production of antimicrobial peptides and the 14 Strachan DP. Hay fever, hygiene and household size. Br Med 1989;
expression of functionally altered PRRs result in exacerba- 299:1259–60.
tion of disease activity by skin colonizing microorganisms, 15 Matricardi PM, Bonini S. High microbial turnover rate preventing
particularly S. aureus and Malassezia furfur. Manipulation of atopy: a solution to inconsistencies impinging on the Hygiene
the innate immune response to microorganisms could pro- hypothesis? Clin Exp Allergy 2000; 30:1506–10.
vide a novel approach to the treatment of AD, as suggested by 16 Yazdanbakhsh M, Kremsner PG, van Ree R. Allergy, parasites, and
the beneficial effects of administration of probiotic bacteria the hygiene hypothesis. Science 2002; 296:490–4.
in preliminary studies. TLR ligands are also being tested for 17 Romagnani S. The increased prevalence of allergy and the hygiene
hypothesis: missing immune deviation, reduced immune suppres-
their ability to shift an allergen-specific Th-2 immune
sion, or both? Immunol 2004; 112:352–63.
response to that of protective Th-1 immunity, although the
18 Bjorksten B, Sepp E, Julge K, Voor T, Mikelsaar M. Allergic devel-
long-term effects of such a strategy are, at present, unknown opment and the intestinal micro-flora during the first year of life. J
[85]. Allergy Clin Immunol 2001; 108:516–20.
19 Watanabe S, Narisawa Y, Arase S, Okamatsu H, Ikenaga T, Tajiri Y,
Kumemura M. Differences in faecal micro-flora between patients
References
with atopic dermatitis and healthy control subjects. J Allergy Clin
1 Lee YA, Wahn U, Kehrt R et al. A major susceptibility locus for Immunol 2003; 111:587–91.
atopic dermatitis maps to chromosome 3q21. Nat Genet 2000; 20 Flohr C, Pascoe D, Williams HC. Atopic dermatitis and the
26:470–3. ‘hygiene hypothesis’: too clean to be true? Br J Dermatol 2005;
2 Cookson WO, Ubhi B, Lawrence R et al. Genetic linkage of child- 152:202–16.
hood atopic dermatitis to psoriasis susceptibility loci. Nat Genet 21 Kalliomäki M, Salminen S, Arvilommi H, Kero P, Koskinen P, Iso-
2001; 27:372–3. lauri E. Probiotics in primary prevention of atopic disease: a ran-
3 Cox HE, Moffatt MF, Faux JA, Walley AJ, Coleman R, Trembath domised placebo-controlled trial. Lancet 2001; 357:1076–9.
RC, Cookson WO, Harper JI. Association of atopic dermatitis to 22 Majamaa H, Isolauri E. Probiotics: a novel approach in the man-
the beta subunit of the high affinity immunoglobulin E receptor. agement of food allergy. J Allergy Clin Immunol 1997; 99:179–85.
Br J Dermatol 1998; 138:182–7. 23 Smits HH, Engering A, van der Kleij D et al. Selective probiotic
4 Kawashima T, Noguchi E, Arinami T, Yamakawa-Kobayashi K, bacteria induce IL-10-producing regulatory T cells in vitro by
Nakagawa H, Otsuka F, Hamaguchi H. Linkage and association of modulating dendritic cell function through dendritic cell-specific
an interleukin 4 gene polymorphism with atopic dermatitis in Jap- intercellular adhesion molecule 3-grabbing nonintegrin. J Allergy
anese families. J Med 1998; 35:502–4. Clin Immunol 2005; 115:1260–7.
5 He JQ, Chan-Yeung M, Becker AB, Dimwit-Ward H, Ferguson AC, 24 Lubbe J. Secondary infections in patients with atopic dermatitis.
Manfreda J, Watson WT, Sandford AJ. Genetic variants of the IL13 Am J Clin Dermatol 2003; 4:641–54.
and IL4 genes and atopic diseases in at-risk children. Genes Immun 25 Leyden JJ, Marples RR, Kligman AM. Staphylococcus aureus in the
2003; 4:385–9. lesions of atopic dermatitis. Br J Dermatol 1974; 90:525–30.
6 Liu X, Nickel R, Beyer K et al. An IL13 coding region variant is 26 Ring J, Abeck D, Neuber K. Atopic eczema: role of microorganisms
associated with a high total serum IgE level and atopic dermatitis in on the skin surface. Allergy 1992; 47:265–9.
the German multicenter atopy study (MAS-90). J Allergy Clin 27 Nilsson EJ, Henning CG, Magnusson J. Topical corticosteroids and
Immunol 2000; 106:167–70. Staphylococcus aureus in atopic dermatitis. J Am Acad Dermatol
7 Nickel RG, Casolaro V, Wahn U et al. Atopic dermatitis is associ- 1992; 27:29–34.
ated with a functional mutation in the promoter of the C-C 28 Guzik TJ, Bzowska M, Kasprowicz A, Czerniawska-Mysik G,
chemokine RANTES. J Immunol 2000; 164:1612–6. Wójcik K, Szmyd D, Adamek-Guzik T, Pryjma J. Persistent skin

© 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9 7
B. S. Baker

colonization with Staphylococcus aureus in atopic dermatitis: rela- ing does not occur via peptidoglycan recognition. EMBO Reports
tionship to clinical and immunological parameters. Clin Exp 2004; 5:1000–6.
Allergy 2005; 35:448–55. 46 Takeda K, Akira S. Toll-like receptors in innate immunity. Int.
29 Rippke F, Schreiner V, Doering T, Maibach HI. Stratum corneum Immunol 2005; 17:1–14.
pH in atopic dermatitis: impact on skin barrier function and col- 47 Pivarcsi A, Nagy I, Kemeny L. Innate immunity in the skin: how
onization with Staphylococcus aureus. Am J Clin Dermatol 2004; keratinocytes fight against pathogens. Curr Immunol Rev 2005;
5:217–23. 1:29–42.
30 Arikawa J, Ishibashi M, Kawashima M, Takagi Y, Ichikawa Y, 48 Lorenz E, Mira JP, Frees KL, Schwartz DA. Relevance of mutations
Imokawa G. Decreased levels of sphingosine, a natural antimicro- in the TLR4 receptor in patients with gram-negative septic shock.
bial agent, may be associated with vulnerability of the stratum cor- Arch Int Med 2002; 162:1028–32.
neum from patients with atopic dermatitis to colonization by 49 Girardin SE, Hugot J-P, Sansonetti PJ. Lessons from Nod2 studies:
Staphylococcus aureus. J Invest Dermatol 2002; 119: 433–9. towards a link between Crohn’s disease and bacterial sensing.
31 Cho SH, Strickland I, Tomkinson A, Fehringer AP, Gelfand EW, Trends Immunol 2003; 24:652–8.
Leung DY. Preferential binding of Staphylococcus aureus to skin 50 Hugot JP, Chamaillard M, Zouali H et al. Association of NOD2
sites of Th-2 mediated inflammation in a murine model. J Invest leucine-rich repeat variants with susceptibility to Crohn’s disease.
Dermatol 2001; 116:658–63. Nature 2001; 411:599–603.
32 al-Mohanna F, Parhar R, Kawaasi A, Ernst P, Sheth K, Harfi H, al- 51 Ogura Y, Bonen DK, Inohara N et al. A frameshift mutation in
Sedairy S. Inhibition of neutrophil functions by human immuno- NOD2 associated with susceptibility to Crohn’s disease. Nature
globulin E. J Allergy Clin Immunol 1993; 92:575–66. 2001; 411:603–6.
33 Gallo RL, Murakami M, Ohtake T, Zaiou M. Biology and clinical 52 Matsubara M, Harada D, Manabe H, Hasegawa K. Staphylococcus
relevance of naturally occurring antimicrobial peptides. J Allergy aureus peptidoglycan stimulates granulocyte macrophage colony-
Clin Immunol 2002; 110:823–31. stimulating factor production from human epidermal kerati-
34 Schroder J-M, Harder J. Human beta-defensin-2. Int J Biochem nocytes via mitogen-activated protein kinases. FEBS Lett 2004;
Cell Biol 1999; 31:645–51. 566:195–200.
35 Harder J, Bartels J, Christophers E, Schroder J-M. Isolation and 53 Proft T, Fraser JD. Bacterial superantigens. Clin Exp Immunol
characterisation of human β-defensin-3, a novel human inducible 2003; 133:299–306.
peptide antibiotic. J Biol Chem 2001; 276:5707–13. 54 Strange P, Skov L, Lisby S, Nielsen PL, Baadsgaard O. Staphylococ-
36 Zanetti M. Cathelicidins, multifunctional peptides of the innate cal enterotoxin B applied on intact normal and intact atopic skin
immunity. J Leuk Biol 2004; 75:39–48. induces dermatitis. Arch Dermatol 1996; 132:27–33.
37 Biragyn A, Ruffini PA, Leifer CA et al. Toll-like receptor 4-depen- 55 Michie CA, Davie T. Atopic dermatitis and staphylococcal super-
dent activation of dendritic cells by β-defensin 2. Science 2002; antigens. Lancet 1996; 347:324.
298:10255–9. 56 Skov L, Olsen JV, Giorno R, Schlievert PM, Baadsgaard O, Leung
38 Yang D, Chertov O, Bykovskaia SN et al. Beta-defensins: linking DY. Application of Staphylococcal enterotoxin B on normal and
innate and adaptive immunity through dendritic and T cell CCR6. atopic skin induces up-regulation of T cells by a superantigen-
Science 1999; 286:525–8. mediated mechanism. J Allergy Clin Immunol 2000; 105:820–6.
39 Ong PY, Ohtake T, Brandt C, Strickland I, Boguniewicz M, Ganz 57 Davison S, Allen M, Vaughan R, Barker J. Staphylococcal toxin-
T, Gallo RL, Leung DY. Endogenous antimicrobial peptides and induced T cell proliferation in atopic eczema correlates with
skin infections in atopic dermatitis. N Engl J Med 2002; increased use of superantigen-reactive Vbeta-chains in cutaneous
347:1151–60. lymphocyte-associated antigen (CLA)-positive lymphocytes. Clin
40 Nomura I, Goleva E, Howell MD et al. Cytokine milieu of Exp Immunol 2000; 121:181–6.
atopic dermatitis, as compared to psoriasis, skin prevents 58 Leung DY, Gately M, Trumble A, Ferguson-Darnell B, Schlievert
induction of innate immune response genes. J Immunol 2003; PM, Picker LJ. Bacterial superantigens induce T cell expression of
171:3262–9. the skin-selective homing receptor, the cutaneous lymphocyte-
41 Rieg S, Steffen H, Seeber S, Humeny A, Kalbacher H, Dietz K, associated antigen, via stimulation of interleukin 12 production. J
Garbe C, Schittek B. Deficiency of dermcidin-derived antimicro- Exp Med 1995; 181:747–53.
bial peptides in sweat of patients with atopic dermatitis correlates 59 Ou LS, Goleva E, Hall C, Leung DY. T regulatory cells in atopic
with an impaired innate defense of human skin in vivo. J Immunol dermatitis and subversion of their activity by superantigens. J
2005; 174:8003–10. Allergy Clin Immunol 2004; 113:756–63.
42 Schittek B, Hipfel R, Sauer B et al. Dermcidin: a novel human anti- 60 Lin Y-T, Wang CT, Hsu CT, Wang LF, Shau WY, Yang YH, Chang
biotic peptide secreted by sweat glands. Nat Immunol 2001; BL. Differential susceptibility to Staphylococcal superantigen
2:1133–7. (SsAg)-induced apoptosis of CD4+ T cells from atopic dermatitis
43 Stern UM, Hornstein OP, Salzer B. Do training-dependent differ- patients and healthy subjects: The inhibitory effect of IL-4 on SsAg-
ences in perspiration exist between healthy and atopic subjects? J induced apoptosis. J Immunol 2003; 171:1102–8.
Dermatol 2000; 27:491–9. 61 Leiferman KM, Ackerman SJ, Sampson HA, Haugen HS, Venencie
44 Imayama S, Shimozono Y, Hoashi M, Yasumoto S, Ohta S, Yon- PY, Gleich GJ. Dermal deposition of eosinophil-granule major
eyama K, Hori Y. Reduced secretion of IgA to skin surface of basic protein in atopic dermatitis. N Engl J Med 1985; 313:282–5.
patients with atopic dermatitis. J Allergy Clin Immunol 1994; 62 Maeda K, Yamamoto K, Tanaka Y, Anan S, Yoshida H. The rela-
94:195–200. tionship between eosinophils, OKT6-positive cells and house dust
45 Travassos LH, Girardin SE, Philpott DJ, Blanot D, Nahori MA, mite (HDM) antigens in naturally occurring lesions of atopic der-
Werts C, Boneca IG. Toll-like receptor 2-dependent bacterial sens- matitis. J Dermatol Sci 1992; 3:151–6.

8 © 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9
 The role of microorganisms in atopic dermatitis

63 Morishita Y, Tada J, Sato A, Toi Y, Kanzaki H, Akiyama H, Arata J. 74 Devos SA, van der Valk PG. The relevance of skin prick tests for
Possible influences of Staphylococcus aureus on atopic dermatitis – Pityrosporum ovale in patients with head and neck dermatitis.
the colonizing features and the effects of staphylococcal enterotox- Allergy 2000; 55:1056–8.
ins. Clin Exp Allergy 1999; 29:1110–7. 75 Scalabrin DM, Bavbek S, Perzanowski MS, Wilson BB, Platts-Mills
64 Wedi B, Wieczorek D, Stunkel T, Breuer K, Kapp A. Staphylococcal TA, Wheatley LM. Use of specific IgE in assessing the relevance of
exotoxins exert proinflammatory effects through inhibition of fungal and dust mite allergens to atopic dermatitis. a comparison
eosinophil apoptosis, increased surface antigen expression with asthmatic and nonasthmatic control subjects. J Allergy Clin
(CD11b, CD45, CD54, and CD69), and enhanced cytokine- Immunol 1999; 104:1273–9.
activated oxidative burst, thereby triggering allergic inflammatory 76 Broberg A, Faergemann J, Johansson S, Johansson SG, Stranne-
reactions. J Allergy Clin Immunol 2002; 109:477–84. gard IL, Svejgaard E. Pityrosporum ovale and atopic dermatitis
65 Wakita H, Tokura Y, Furukawa F, Takigawa M. Staphylococcal in children and young adults. Acta Derm Venereol 1992;
enterotoxin B upregulates expression of ICAM-1 molecules on 72:187–92.
IFN-gamma-treated keratinocytes and keratinocyte cell lines. J 77 Kim TY, Jang IG, Park YM, Kim IIO, Kim CW. Head and neck der-
Invest Dermatol 1995; 105:536–42. matitis: the role of Malassezia furfur, topical steroid use and envi-
66 Ezepchuk YV, Leung DY, Middleton MH, Bina P, Rieser R, Norris ronmental factors in its causation. Clin Exp Dermatol 1999;
DA. Staphylococcal toxins and protein A induce cytotoxicity and 24:226–31.
release of tumor necrosis factor from human keratinocytes. J Invest 78 Nissen D, Petersen LJ, Esch R, Svejgaard E, Skov PS, Poulsen LK,
Dermatol 1996; 107:603–9. Nolte II. IgE-sensitization to cellular and culture filtrates of fungal
67 Goodman RE, Nestle F, Naidu YM, Green JM, Thompson CB, extracts in patients with atopic dermatitis. Ann Allergy Asthma
Nickoloff BJ, Turka LA. Keratinocyte-derived T cell costimulation Immunol 1998; 81:247–55.
induces preferential production of IL-2 and IL-4 but not IFN- 79 Buentke E, Scheynius A. Dendritic cells and fungi. APMIS 2003;
gamma. J Immunol 1994; 152:5189–98. 111:789–96.
68 Travers JB, Norris DA, Leung DYM. The keratinocyte as a target for 80 Tengvall Linder M, Johansson C, Zargari A, Bengtsson A, van der
staphylococcal bacterial toxins. J Invest Dermatol Symp Proc 2001; Ploeg I, Jones I, Harfast B, Scheynius A. Detection of Pityrosporum
6:225–30. orbiculare reactive T cells from skin and blood in atopic dermatitis
69 Leung DY, Harbeck R, Bina P, Reiser RF, Yang E, Norris DA, Han- and characterisation of their cytokine profiles. Clin Exp Allergy
ifin JM, Sampson HA. Presence of IgE antibodies to staphylococcal 1996; 26:1286–97.
exotoxins on the skin of patients with atopic dermatitis. J Clin 81 Johansson C, Eshagi H, Linder MT, Jakobson E, Scheynius A. Pos-
Invest 1993; 92:1374–80. itive atopy patch test reaction to Malassezia furfur in atopic derma-
70 Nissen D, Pedersen LJ, Skov PS, Vejlsgaard GL, Poulsen LK, Jarlov titis correlates with a T helper 2-like peripheral blood mononuclear
JO, Karlsmark T, Nolte H. IgE-binding components of staphylo- cell response. J Invest Dermatol 2002; 118:1044–51.
coccal enterotoxins in patients with atopic dermatitis. Ann Allergy 82 Belew PW, Rosenberg EW, Jennings BR. Activation of the alterna-
Asthma Immunol 1997; 79:403–8. tive pathway of complement by Malassezia ovalis (Pityrosporum
71 Mrabet-Dahbi S, Breuer K, Klotz M, Herz U, Heeg K, Werfel T, ovale). Mycopathologica 1980; 70:187–90.
Renz H. Deficiency in immunoglobulin G2 antibodies against sta- 83 Watnabe S, Kano R, Sato II, Nakamura Y, Ilasegawa A. The effects
phylococcal enterotoxin C1 defines a subgroup of patients with of Malassezia yeasts on cytokine production by human kerati-
atopic dermatitis. Clin Exp Allergy 2005; 35:274–81. nocytes. J Invest Dermatol 2001; 116:769–73.
72 Bredius RG, de Vries CE, Troelstra A, van Alphen L, Weening RS, 84 Aspres N, Anderson C. Malassezia yeasts in the pathogenesis of
van de Winkel JG, Out TA. Phagocytosis of Staphylococcus aureus atopic dermatitis. Austr J Dermatol 2004; 45:199–207.
and Haemophilus influenzae type B opsonized with polyclonal 85 Leung T-f, Tang NLS, Wong GWK, Fok T-f. CD14 and Toll-like
human IgG1 and IgG2 antibodies. Functional hFc gamma RIIa receptors: Potential contribution of genetic factors and mecha-
polymorphism to IgG2. J Immunol 1993; 151:1463–72. nisms to inflammation and allergy. Curr Drug Targets – Inflamm
73 Scheynius A, Johansson C, Buentke E, Zargari A, Linder MT. Allergy 2005; 4:169–75.
Atopic eczema/dermatitis syndrome and Malassezia. Int Arch
Allergy Immunol 2002; 127:161–9.

© 2006 British Society for Immunology, Clinical and Experimental Immunology, 144: 1–9 9