Margarete Niebuhr Department of Dermatology and Allergology Hannover Medical School Ricklinger Str. 5 D-30449 Hannover Germany
Background: Atopic dermatitis (AD) is often complicated by an enhanced susceptibility to bacterial skin infections, especially with Staphylococcus aureus. Toll-like receptors (TLR), especially TLR-2 recognizes cell wall components of S. aureus, e.g. lipoteichoic acid (LTA) and peptidoglycan (PGN). A heterozygous TLR-2 R753Q polymorphism occurs in a frequency of 11.5% in adult AD patients and has been shown to be associated with a severe phenotype of AD.
Methods: The aim of this study was to investigate the impact of TLR-2 agonists (LTA, PGN and Pam3Cys) on cytokine production in human monocytes from AD patients with the TLR-2 R753Q polymorphism compared with that of AD patients with ‘wild type’ TLR-2 and control individuals to elucidate the functional role of the TLR-2 R753Q polymorphism.
Results: Monocytes from AD patients with the TLR-2 R753Q mutation produced significantly more IL-6 and IL-12 compared with that of AD patients with nonmutated TLR-2 upon stimulation with TLR-2 agonists.
Conclusion: We show for the first time functional differences in TLR-2 responsiveness of monocytes from AD patients with the TLR-2 R753Q mutation compared with wild type AD patients in a ligand-dependent manner.
Clinical implication: Our data support the emerging concept that AD patients have a dysbalance in innate and aquired immunity. TLR-2 may be essential in the pathogenesis and maintenance of AD and may be involved in the enhanced susceptibility to skin infections with S. aureus and in a higher inflammatory response in patients with AD carrying the TLR-2 polymorphism.
Atopic dermatitis (AD) is a chronic inflammatory skin disease of the atopic syndrome (1, 2). One hallmark of AD is the striking susceptibility to colonization with Staphylococcus aureus: 80–100% of patients with AD are colonized with S. aureus (3, 4). In contrast, S. aureus can be isolated from the skin of only 5–30% of healthy individuals, mainly from intertriginous areas (4). Moreover, a relationship between disease severity, extent and S. aureus colonization of lesional and nonlesional skin has been noted (3) which can be due to IgE-mediated hypersensitivity (5, 6) or production of exotoxins with superantigenic properties (7, 8). It was shown that disease severity can be improved with antistaphylococcal treatment (9).
Both the innate and the adaptive immune system participate in recognition and defence of microbial pathogens. More recently, the role of the innate immune system has gained large attention in the pathogenesis of AD (10, 11). Activation of the innate immune system is mediated by pattern recognition receptors (PRRs), which are present on immune cells and recognize pathogen-associated molecular patterns (PAMPs). Further key elements of innate immunity are human defensins. Especially, human β defensin 2 (HBD-2) and 3 (HBD-3) are thought to be involved in cutaneous immune defence (12). However, their expression is decreased in AD (13) which implicates an altered host defence mechanism within this compartment of the immune system.
Toll-like receptors act as PRRs comprising a family of (currently) 13 receptors in humans with distinct recognition profiles. In this context, TLR-2 has emerged as a principle receptor in combating Gram-positive bacteria, especially S. aureus (14). TLR-2 deficient mice revealed a higher intradermal S. aureus growing rate upon subcutaneous injection compared with wild type mice.
Recently, the prevalence of several TLR single nucleotide polymorphisms (SNP) among adult AD patients from our department was analysed. In humans suffering from septic shock, the TLR-2 R753Q SNP was found to be associated with S. aureus infection (15). A high frequency (12%) of adult SD patients whose skin is known to be colonized with S. aureus who carried the TLR-2 R753Q SNP was found. Moreover, the TLR-2 R753Q mutation was associated with a severe phenotype compared with that of AD patients without this mutation (16). However, it has not yet been elucidated whether this mutation or stimulation with TLR-2 ligands has any functional impact on the pathogenesis of AD.
Toll-like receptor-2 forms heterodimers with TLR-1 and TLR-6 to interact with a rather broad spectrum of ligands. Studies using knockout mice identified TLR-1 as the co-receptor required for the recognition of triacylated lipoproteins and lipopeptides such as Pam3Cys (17), while diacylated components such as lipoteichoic acid (LTA), which is a component of the cell wall of S. aureus, interact with TLR-2/TLR-6 heterodimers (17, 18). Peptidoglycan (PGN) is a major constituent of the cell wall of Gram-positive bacteria which induces signal transduction via TLR-2, NOD 1 (Card4) and NOD 2 (Card15) receptors, respectively. NOD molecules including NOD1 and NOD2, are a family of intracellular pattern recognition proteins involved in bacterial detection (19, 20).
Interestingly, a recent study by Hasannejad et al. (21) showed a general impairment of TLR-2-mediated proinflammatory cytokine production by monocytes from AD. In this study, the authors investigated the effect of the synthetic TLR-1/TLR-2 ligand Pam3Cys on the production of TNFα and IL-1ß.
In this study, we investigated the effects of TLR-2 stimulation with the staphylococcal cell wall components PGN and LTA and with the synthetic TLR-2 agonist Pam3Cys on cytokine secretion in patients with AD carrying the TLR-2 R753Q SNP compared with wild type AD patients and with healthy controls.
Our data confirm a functional role of TLR-2 on monocyte function. Here, we measured the cytokines IL-6 and IL-12. We did not find a general impairment of TLR-2-induced cytokine production in AD. However, we found significant functional differences in monocytes carrying the SNP compared with ‘wild type’ monocytes from patients with AD. Thus, our data indicate that a dysregulation in TLR-2-mediated effects might be an explanation for the severe phenotype in TLR-2 R753Q SNP AD patients. Therefore, TLR-2 might be essential in the linkage between innate and adaptive immunity in the pathogenesis of AD and responsible for the enhanced susceptibility to skin infections with S. aureus.
Materials and methods
Preparation of monocytes
Peripheral blood mononuclear cells (PBMC) were isolated by Lymphoprep density-gradient centrifugation from informed consented healthy donors, wild type AD patients and AD patients with the TLR-2 R753Q SNP. Atopic dermatitis was determined by the diagnostic criteria of Hanifin and Rajka (22). CD14+ cells were purified by negative selection according to the manufacturers’ instructions (Miltenyi Biotec, Bergisch Gladbach, Germany). Monocytes were cultured in RPMI medium 1640 supplemented with 10% heat-inactivated fetal calf serum (Gibco, Eggestein, Germany), 2 mM glutamine (Seromed, Berlin, Germany) and 1% penicillin/streptomycin (Seromed). Before initiation of further experiments, cells were analysed by two-colour flow cytometry (FACScalibur; Becton Dickinson, Heidelberg, Germany) for contaminating CD3+ T cells, CD20+ B cells and CD56+ natural killer cells. CD14+ monocytes had a purity of at least 86%.
Stimulation of monocytes
Cells were either unstimulated or stimulated for various periods of time with PGN from S. aureus (1, 5 or 10 g/ml; Invivogen, Toulouse, France) which acts via TLR-2 and NOD, LTA from S. aureus isolated as previously described (23) (0.1, 1 and 10 μg/ml; University of Konstanz, Konstanz, Germany) which acts via TLR-2/TLR-6 and Pam3Cys (0.1, 1 and 10 μg/ml; EMC microcollections, Tübingen, Germany) which acts via TLR-2/TLR-1. For IL-12 experiments, cells were primed for 2 h with IFN-γ (200 U/ml; R&D Systems, Wiesbaden, Germany) as previously described (24). Lipopolysaccharide (LPS) was not detected in any reagent, as determined by the Limulus amebocyte assay (Haemochrom Diagnostika, Essen, Germany).
Intracellular staining for IL-6 and IL-12
For intracellular staining, BD Cytofix Cytoperm Plus kit with BD GolgiPlug (Biosciences, San Diego, CA, USA) was used following the supplier’s instructions. Cells were stained with anti-human IL-12p40/p70 and IL-6 or isotype control (Becton Dickinson) and analysed by flow cytometry (FACScalibur; Becton Dickinson).
Monocytes were stimulated as indicated and the supernatants were harvested and analysed for IL-6 (Duo Set; R&D Systems, Minneapolis, MN, USA) and IL-12p70 (Ready-Set-go, eBioscience, San Diego, CA, USA) using a commercially available enzyme-linked immunosorbent assay (ELISA) system following the manufacturer’s instructions.
mRNA isolation, RT and LightCycler PCR
mRNA was isolated from 1 × 105 monocytes either unstimulated or stimulated for 2, 4, 6, 8 and 24 h with PGN (10 μg/ml), 8 and 24 h with LTA (10 μg/ml) and 8 and 24 h with Pam3Cys (10 μg/ml) using the High Pure mRNA isolation kit (Roche Molecular Biochemicals, Mannheim, Germany) according to the supplier’s instructions.
The cDNA was synthesized with integrated removal of genomic DNA in the samples by using a QuantiTect reverse transcription kit (Qiagen, Hilden, Germany). Quantitative real-time RT-PCR (qRT-PCR) was performed on a LightCycler PCR (Roche, Molecular Biochemicals) using Quantitect SYBR Green PCR kit (Qiagen) and Quantitect Primer for IL-6 and IL-12p40 and GAPDH (house keeping gene).
Specific targets were amplified using the following programme: (PCR initial activation step) 15 min, 95°C and ramp 20°C per min; (denaturation) 15 s, 94°C, ramp 2°C per min; (annealing) 20 s, 55°C, ramp 2°C per min; (extension) 20 s, 72°C, ramp 2°C per min. For quantitative analysis, standard curves for IL-6 and IL-12p40 were created. These standard curves describing the PCR efficiencies of the target and the reference gene (GAPDH) allowed an efficiency-corrected quantification using the Relative Quantification software (Roche, Molecular Biochemicals).
Statistical analyses were performed using the Student’s t-test and paired t-test. The software used to perform the statistical analysis was sigma stat for Windows. Data are shown as the mean ± SEM (standard error of the mean). In the figures, *P-value < 0.05, **P < 0.01 and ***P < 0.001.
Prevalence of the TLR-2 R753Q SNP mutation in patients with AD
We screened all adult patients with AD who visited our outpatient clinic in 2006 for the TLR-2 R753Q SNP mutation. Until now, we tested a total of 176 adult AD patients and could reproduce and extend the published data (16) now finding a prevalence of 11% in adult AD patients in this larger cohort.
TLR-2-mediated cytokine secretion is not generally impaired in monocytes from patients with AD compared with monocytes from healthy donors
To study the functional role of TLR-2 on monocyte function, we measured the cytokines IL-6 and IL-12 upon stimulation with three different TLR-2 ligands like Pam3Cys, PGN and LTA. Monocytes were stimulated for various periods of time with PGN (1, 5 or 10 μg/ml), LTA (0.1, 1 and 10 μg/ml) or Pam3Cys (0.1, 1 and 10 μg/ml).
Interleukin-6 was directly induced by all TLR-2 ligands both in monocytes from healthy individuals (n = 6) and in AD patients (n = 13). As TLR-2 ligands did not induce IL-12 in ‘resting’ monocytes, cells were primed for 2 h with 200 U/ml IFN-γ prior to stimulation with TLR-2 before the determination of IL-12. Monocytes both from healthy individuals and from AD patients produced IL-12 under this condition. We did not find a general impairment of TLR-2-induced IL-6 or IL-12 production in monocytes from the 13 patients with AD (data not shown).
No individual of the healthy control group but 6/13 AD patients carried the heterogeneous TLR-2 R753Q SNP. In further evaluations, we differentiated the tested individuals into three groups: healthy individuals (‘controls’), AD patients without the TLR-2 R753Q SNP (‘wild type AD’) and AD patients with the TLR-2 R753Q SNP (‘SNP AD’).
In AD, the TLR-2 R753Q SNP is associated with a higher interleukin-12p40/p70 expression and secretion by monocytes stimulated with the TLR-2/TLR-1 ligand Pam3Cys
IL-12p70 production upon stimulation with TLR-2 ligands was investigated at both the protein and mRNA level, respectively. Monocytes from TLR-2 R753Q SNP AD patients, ‘wild type’ AD patients and healthy controls were stimulated with LTA, Pam3Cys and PGN to elucidate whether signalling of ligands requiring the TLR-2/TLR-1 heterodimer (Pam3Cys), the TLR-2/TLR-6 heterodimer (LTA) and the TLR-2/NOD pathway (PGN) is affected. IFN-γ was used as priming signal, as previously described (24). Then, cell-culture supernatants were analysed for IL-12p70 by ELISA, the production of intracellular IL-12p40/p70 was determined by flow cytometry and mRNA was harvested for IL-12p40 PCR analysis.
Monocytes from TLR-2 R753Q SNP AD patients prestimulated with IFN-γ secreted significantly more IL-12p70 compared with that of AD patients with nonmutated TLR-2 and healthy controls upon stimulation with Pam3Cys. Stimulation with LTA yielded a similar trend, whereas stimulation with PGN resulted in similar IL-12p70 secretion when ‘wild type’ monocytes and monocytes carrying the TLR-2 R753Q SNP were compared (Table 1).
Table 1. IL-12 and IL-6 secretion upon TLR-2 stimulation with PGN, LTA and Pam3Cys in monocytes from wild type AD patients and TLR-2 R753Q SNP AD patients
Wild type AD ± SEM in pg/ml
SNP AD ± SEM in pg/ml
1 × 105 monocytes were stimulated for 48 h with Pam3Cys (1 μg/ml), LTA (1 μg/ml) or PGN (5 μg/ml). For IL-12 induction, cells were prestimulated for 2 h with IFN-γ (200 U/ml). Cell culture supernatants were analysed for concentrations of IL-6 and IL-12 by ELISA. The mean value ± SEM of n = 6 TLR-2 R753Q SNP AD patients, n = 7 wild type AD patients and n = 6 healthy controls is shown in pg/ml as well as P-values.
6.7 ± 2.5
8.8 ± 2.8
IFN + Pam3Cys
18.0 ± 9.3
54.7 ± 14.4
IFN + LTA
11.0 ± 2.9
26.8 ± 13.6
IFN + PGN
80.4 ± 5.79
37.5 ± 6.5
9.4 ± 4.9
5.1 ± 3.7
460.9 ± 264.9
781.7 ± 191.4
176.7 ± 42.6
622.0 ± 103.3
413.2 ± 180.3
305.6 ± 83.2
Intracellular IL-12p40/p70 expression showed a similar trend upon Pam3Cys stimulation with up-regulation of a higher percentage of IL-12 positive monocytes in TLR-2 R753Q SNP AD patients compared with nonpolymorphic AD patients, as detected by flow cytometry (data not shown). No significant difference of IL-12p40 mRNA expression between monocytes from TLR-2 R753Q SNP AD patients and ‘wild type’ AD patients was detected upon TLR-2 stimulation as demonstrated by efficiency-controlled qRT-PCR (data not shown).
In AD, the TLR-2 R753Q SNP is associated with a higher interleukin-6 production by monocytes stimulated with the TLR-2/TLR-6 ligand LTA and the TLR-2/NOD ligand PGN
The effect of TLR-2 stimulation on IL-6 production in monocytes was studied at both the protein (Fig. 2) and mRNA level (Fig. 3A, B).
Monocytes from TLR-2 R753Q SNP AD patients secreted significantly more IL-6 compared with ‘wild type’ AD patients upon stimulation with LTA (Fig. 2).
IL-6 mRNA was upregulated by all TLR-2 agonists tested in this study: 8 h stimulation with LTA (Fig. 3B), PGN (Fig. 3A, B) and Pam3Cys (Fig. 3B) yielded in an upregulation of IL-6 mRNA expression in monocytes from TLR-2 R753Q SNP AD patients compared with wild type AD patients. This effect was significant upon 8 h PGN stimulation (Fig. 3A).
No effects of TLR-2 stimulation were observed using intracellular IL-6 staining as detected by flow cytometry (data not shown).
Atopic dermatitis is a chronic inflammatory skin disease, which is frequently complicated by an enhanced susceptibility to bacterial skin infections, especially with S. aureus. This may be due to an impaired innate immune defense against these bacteria (10, 11). Recently, the TLR-2 R753Q SNP was found at high frequency (12%) among adult AD patients subsequently seen in our outpatient clinic and was associated with a severe phenotype compared with that of AD patients without this mutation. In contrast, the TLR-2 R753Q SNP occurs at a frequency of only 3% in healthy controls (16). We screened a total of 78 adult AD patients from our department and 39 healthy controls in that previous study (16). Here, we continued screening all adult patients with AD who visited our outpatient clinic in 2006 and 2007. Until now, we tested a total of 176 adult AD patients and could reproduce and extend the published data now finding a prevalence of 11% in adult AD patients in this larger cohort.
In another study, Weidinger et al. (25) showed no increased frequency of the TLR-2 SNP in AD patients in a genetic analysis of German parent-offspring trios. Weidinger et al. (25) analysed not only adults but also children and the study population was smaller (n = 42) which may explain the different results in both studies. The contradictory results between both studies motivated us to continue the screening process in our adult outpatient population suffering from AD. In a larger cohort of 176 subjects, we could now reproduce and extend our previous findings (16) with a high TLR-2 R753Q SNP prevalence of about 11% in AD patients.
In functional experiments, we investigated TLR-2-mediated effects on monocytes, which are known to emigrate into the skin in eczema and differentiate into macrophages and inflammatory dendritic epidermal cells in AD. We used highly purified monocytes to exclude interfering effects from other cells in the PBMC fraction. For example Mrabet-Dahbi et al. (26) recently described differences in T cell responses to TLR-2 ligands associated with the TLR-2 R753Q SNP in patients with AD. First, we confirmed a functional role of TLR-2 on monocyte function when we measured the cytokines IL-6 and IL-12 upon stimulation with three different TLR-2 ligands. As there is no selective TLR-2 ligand known so far, we chose to stimulate monocytes in parallel with a synthetic TLR-2/TLR-1 ligand (Pam3Cys), a natural TLR-2/TLR-6 ligand (LTA) and another natural TLR-2/NOD ligand (PGN).
We did not find a general impairment of TLR-2-induced effects on monocytes isolated from patients with AD compared with healthy controls. Monocytes from wild type AD patients produce less IL-12 and IL-6 upon stimulation with TLR-2 ligands compared with monocytes from healthy controls (Figs 1–3). This is in line with Hasannejad et al. (21), who found a selective impairment in IL-1β and TNF-α secretion in monocytes from patients with AD. These findings can contribute to the known susceptibility of AD patients to S. aureus. Hassanejad et al. (21) had not characterized TLR-2 in regard of a possible R753Q mutation which might have influenced the results. Moreover, they did not compare the effects of the TLR-1/TLR-2 ligand to effects of PGN and LTA both acting via TLR2 and other co-receptors (NOD and TLR-6, respectively). TLR-2 R753Q mutant AD patients have a more severe phenotype compared with ‘nonmutant’ AD patients (16). The enhanced production of the proinflammatory cytokines IL-12 and IL-6 may explain the enhanced skin inflammation in patients with the TLR-2 R753Q SNP.
To compare TLR-2-mediated effects on monocytes with and without the TLR-2 R753Q SNP, we investigated the levels of IL-12 production. Interleukin-12 is an important proinflammatory and immunoregulatory cytokine mainly produced by monocytes/macrophages, which polarizes and stimulates Th1 cells (27, 28). In a mouse model, TLR-2-induced IL-12 has been shown to play a significant role in host defences (29). In AD, IL-12 is considered to be a ‘key player’ during the chronification of eczema (30).
Here, we found a significantly enhanced IL-12 secretion upon Pam3Cys stimulation in monocytes from patients with AD carrying the TLR-2 SNP. Moreover, we observed a similar trend upon LTA stimulation but not upon PGN stimulation which indicates a critical role of the co-stimulatory molecules.
Schröder et al. (31) compared TLR-2-mediated effects (secretion of TNF-α and IFN-γ) in a human whole blood model. In that study, the authors also described significant differences in individuals heterozygous for the TLR-2 polymorphism Arg753Gln in response to TLR-2/TLR-1 ligands but not in response to TLR-2/TLR-6 ligands which is in line with our observation.
Here, we also investigated IL-6 which has previously been shown to be upregulated in human monocytes upon stimulation with PGN and LTA (32). We could confirm the induction of IL-6 in monocytes isolated from healthy individuals and from patients with AD. Interestingly, IL-6 was induced upon LTA stimulation (TLR-2/TLR-6) to a significantly higher degree in TLR-2 R753Q SNP AD patients compared with ‘wild type’ AD patients on both the protein and the mRNA level. Moreover, a higher induction of IL-6 mRNA in TLR-2 R753Q SNP AD patients was observed after PGN (TLR-2/NOD) stimulation, whereas Pam3Cys had no different effects on IL-6 production by monocytes carrying the TLR-2 SNP.
Hassanejad et al. (21) described differences in IL-1β and TNF-α expression in distinct monocyte subpopulations [CD14dim (proinflammatory) and CD14bright (classical) monocytes] upon Pam3Cys stimulation in patients with AD: a reduction of TLR-2-mediated proinflammatory cytokine production in patients with AD was mostly observed in proinflammatory CD14dim monocytes expressing high FcεRI. Therefore, it could be possible that the ability of classical monocytes to produce IL-6 and IL-12 in response to TLR-2 ligands may be alternatively enhanced to compensate for the impaired ability of proinflammatory monocytes to produce TNF-α.
Interleukin-6 has been shown to play an important role as proinflammatory cytokine in AD which is upregulated in the skin of humans and in animal models of AD (33). Thus, the differences of IL-6 production in response to the staphylococcal cell wall components LTA and PGN may directly contribute to severity of AD in patients carrying the TLR-2 SNP.
In conclusion, we show for the first time differences in cytokine production by monocytes carrying the TLR-2 R753Q SNP compared with ‘wild type’ monocytes isolated from patients with AD which are dependent on the nature of the TLR-2 ligand. Our data support the emerging concept that AD patients have a dysbalance in innate as well as aquired immunity. Further investigations are necessary to elucidate the function of TLR-2 in patients with AD to explain the high susceptibility to staphylococcal infections and provide new therapeutic options for altering the dysbalance in innate and acquired immunity.
We would like to thank Gabriele Begemann and Kathrin Baumert for their excellent technical assistance. The authors have no conflicting financial interests on this work which was supported by a grant from the Deutsche Dermatologische Gesellschaft (DDG).