Materials and Methods

Reagents

Antibodies specific for human TLR2 and TLR4 (polyclonal goat) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA) and from Biocarta (Hamburg, Germany) (clone HTA125 (anti-TLR4) and clone TL2.1 (anti-TLR2)) and were used in a 1:10 and 1:50 dilution, respectively. Purified goat serum (Sigma, Deisenhofen, Germany) and purified mouse IgG (BD-Biosciences, Heidelberg, Germany) served as isotype controls. In order to evaluate the specificity of anti-TLR2 and anti-TLR4 antibodies we used HEK 293 cells transfected with a dominant expression plasmid for either hTLR2 or hTLR4 (kindly provided by C. Kirschning, Institute for Medical Microbiology and Immunology, TU Munich, Germany). As HEK 293 cells do not express any of the investigated receptors we used untransfected cells as negative control.

For analysis of RelA (NF-B subunit p65) translocation, the anti-p65 antibody from Rockland (Gilbertsville, PA) was used (dilution 1:200) followed by a fluorescein isothiocyanate (FITC) conjugated goat antirabbit second-step antibody (dilution 1:100) (Sigma). Griess reagent was purchased from Sigma and used as recommended by the manufacturer. LPS (Escherichia coli) and LTA (S. aureus) were obtained from Sigma and PGN from Fluka (Heidelberg, Germany). LTA and PGN were tested for LPS content using the Limulus assay from BioWhittaker, Walkersville, MD. Aliquots showed LPS contamination of approximately 0.01–0.1 g per mg. Preliminary results using 100 higher concentration of purified LPS (10 g per mL), however, did not show signs of keratinocyte activation in any of the assays used so that we considered LPS as not responsible for the observed biologic phenomena. The PAFR antagonist CV3988 was obtained from Biomol Research Laboratories (Plymouth Meeting, PA).

Cell culture

Primary cultures of human keratinocytes were obtained following standard procedures (^{Mempel et al, 2002}). In brief, surgical specimens of human foreskin were cut into pieces of 0.5 cm² and exposed to dispase 2.4 U per mL (Roche, Mannheim, Germany) for 12 h at 4°C. The epidermis was then mechanically removed from the dermal layer, homogenized by repeated aspirations in Pasteur pipettes, and incubated in ethylenediamine tetraacetic acid (EDTA) trypsin 0.25% (Roche) for 60 min. Cells were washed three times in SFM (Gibco/Life Technologies, Eggenstein, Germany), stained for viability with 0.5% trypan blue (Sigma), and seeded into 75 cm² culture flasks (Becton Dickinson, Heidelberg, Germany) with keratinocyte SFM. Cells were kept at 37°C and 5% CO₂.

Bacterial strains

Parent strains Newman and 8325-4 are standard S. aureus laboratory strains that have been described previously. All strains were kindly provided by Professor T. Foster, Department of Microbiology, Moyne Institute of Preventive Medicine, Trinity College, Dublin, Ireland.

In order to maintain constant numbers of bacteria, staphylococci were treated with mitomycin C (30 g per mL, from Sigma) for 1 h, washed extensively, and adjusted to a stock solution of OD₅₇₀ 0.4, which corresponds to 10⁸ colony-forming units per mL. 25 L of this stock solution were used in 1 mL of SFM to stimulate the cells. This treatment has been shown to inhibit bacterial replication but not the production of virulence factors (^{Tokura et al, 1997}).

Coincubation of keratinocytes and staphylococci

Mitomycin C treated staphylococci were diluted at a concentration of 10⁶ colony-forming units per mL and in antibiotic-free SFM and incubated for 24 h with the staphylococci. Preliminary experiments had shown that a multiplicity of infections (MOI) of 10–50 showed the best results for activation of keratinocytes avoiding cell death.

RNA preparation and RT-PCR

Cells were grown in six-well dishes until approximately 90% confluency and were stimulated for the indicated conditions. Thereafter, the cells were washed and incubated with 1 mL Trizol (Gibco) for 5 min with repeated pipetting. RNA was then isolated following standard techniques including DNAse digest (Qiagen, Hilden, Germany) and dissolved in sterile H₂O. Standardized amounts of RNA (e.g., 1 g) were then subjected to reverse transcription using the Gibco superscript kit (Gibco) and oligo dT primer. The cDNA was then amplified using Taq polymerase from Promega (Mannheim, Germany) following the recommendations of the manufacturer.

For the analysis of TLR expression the primers published by Bauer and Chuang were used with modifications (^{Bauer et al, 2001}, and personal communication;^{Chuang and Ulevitch, 2001}). Previously unpublished primers and primers for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and iNOS were designed using the Primer Express (PE Applied Biosystems, Foster City, CA) software. COX2 primers were taken from^{Honda et al (2000)}. All primers were verified using LPS-stimulated peripheral blood mononuclear cells (10 g per mL for 24 h) HEK 293 and THP-1 cells as positive control and by omitting the reverse transcription step in an otherwise identical set-up as negative control.

Of note, all TLR PCR were run to saturation (40 cycles) in order to detect very minute amounts of specific cDNA whereas reactions for COX2 (30 cycles), iNOS (34 cycles), and GAPDH (28 cycles) were established with various cycle numbers according to the expression pattern of the gene (Table I).

Table I.

Full table

Expression of human COX2, iNOS, and GAPDH was measured using semiquantitative PCR. The following primers were used: iNOS-Fw, ATG CCA GAT GGC AGC ATC AGA; iNOS-Rv, ACC CTG CCA ACG TGG AAT TCA CTC AG; COX2-Fw, TTC AAA TGA GAT TGT GGG AAA ATT GCT; COX2-Rv, AGA TCA TCT CTG CCT GAG TAT CTT; GAPDH-Fw, CAA GTC CCT GAA GGA TGT GGA; GAPDH-Rv, GAG GAG TGG GTG TCG CTG TTT GAA GTC. All primers were obtained from MWG (Ebersberg, Germany).

Real-time PCR for IL-8 transcription

As it has been reported that IL-8 production is tightly regulated after NF-B translocation in human keratinocytes we looked in more detail at the changes in IL-8 mRNA by real-time PCR using the Taqman technique as previously described (^{Rad et al, 2002}).

Taqman primers (MWG) and probes (PerkinElmer, Weiterstadt, Germany) were designed using the primer design software Primer Express. All probes were synthesized by PerkinElmer and labeled with the reporter dye 6-carboxyfluorescein at the 5' end and the quencher dye 6-carboxytetramethylrhodamin at the 3' end. Primers and probes were chosen to span exon junctions or to lie in different exons to prevent amplification of genomic DNA.

Five microliters of RNA were transcribed into cDNA in a total volume of 50 L using 50 U of MultiScribe reverse transcriptase (PerkinElmer) according to the manufacturer's instructions. PCR was performed in a volume of 30 L on the ABI PRISM 7700 sequence detection system (PerkinElmer). For each run, a master mix was prepared on ice containing 15 L of Universal Master Mix (PE Applied Biosystems), primers (1 mol per L), fluorogenic probes (0.32 mol per L), and H₂O. To each well of a 96-well plate, 25 L of master mix and 5 L of cDNA samples were added. All PCR were performed in duplicate. Thermal cycling was initiated with an incubation step at 50°C for 2 min, followed by a first denaturation step at 95°C for 10 min, and continued with 40 cycles of 95°C for 15 s, 58°C for 20 s, and 72°C for 30 s.

The following primers and probes were used: forward, GCCAACAC-AGAAATTATTGTAAAGCTT; reverse, AATTCTCAGCCCTCTTCAA-AAACTT; labeled probe, AGAGCTCTGTCTGGACCCCAAGGAAAAC.

NF-B translocation

We used three different techniques to screen for NF-B translocation. First, nuclear extract gel shifts were performed as previously published (^{Brand et al, 1997;Mackman, 2000}). In brief, nuclear extracts from 510⁶ cells were prepared and analyzed as described. Protein concentrations were determined by the Bradford method (Bio-Rad, Munich, Germany). The prototypic immunoglobulin -chain oligo-nucleotide was used as a probe (5'-CAGAGGGACTTTCCGAGA-3') and labeled by annealing of complementary primers followed by primer extension with the Klenow fragment of DNA polymerase I (Boehringer Mannheim, Germany) in the presence of [-³²P]dCTP (>3000 Ci per mmol; DuPont, Brussels, Belgium) and deoxynucleoside triphosphates (Boehringer Mannheim). Nuclear extracts (5 g protein) were incubated with radiolabeled DNA probes (10 ng; 10⁵ cpm) for 30 min at room temperature in 20 L of binding buffer (20 mmol per L Tris-HCl, pH 7.9; 50 mmol per L KCl; 1mmol per L dithiothreitol; 0.5 mmol per L EDTA; 5% glycerol; 1 mg per mL bovine serum albumin; 0.2% Nonidet P-40; 50 ng of poly(dI-dC) per L). Samples were run in 0.25TBE buffer (10890 mmol per L Tris; 890 mmol per L boric acid; 20 mmol per L EDTA, pH 8.0) on nondenaturing 4% or 6% polyacrylamide gels at 125 V. Nuclear extracts from LPS-stimulated THP-1 cells were used as positive controls. As an additional control, samples were incubated with an excess (10, 100) of nonlabeled B oligonucleotide, which completely abolished binding of the radiolabeled oligonucleotide to the nuclear proteins. To control the nuclear protein content, the nuclear extracts were incubated with a blunt end double-stranded Sp-1 oligonucleotide that was labeled with [-³²P]ATP (>5000 Ci per mmol, DuPont) and T4 polynucleotide kinase (Boehringer Mannheim). Gels were dried and analyzed by autoradiography.

Second, we used the immunostaining technique for p65 translocation as described by^{Song et al (2002)}, and third, in order to quantitate the induction of NF-B translocation, we established the transient transfection of primary keratinocytes with an NF-B–luciferase reporter plasmid from Clontech (Palo Alto, CA). The plasmid was grown using the Topo cloning kit from Invitrogen (Gibco, Eggenfelden, Germany) and purified with the Quiagen Maxi Prep kit. Various amounts of the plasmid were transfected into the keratinocytes using DMRIE-C reagent from Invitrogen following the instructions of the manufacturer. To control for transfection efficacy, a plasmid expressing green fluorescent protein constitutively (kindly provided by Dr Markus Braun-Falco, Department of Dermatology and Allergy, TU Munich) was used in parallel experiments. Usually, this control yielded between 20% and 30% of transfected cells after 24 h. Thereafter, the cells were incubated with the various stimuli for 24 h before the cells were harvested and prepared for both luciferase quantification (using the luciferase reaction buffer from Invitrogen and a luminometer from Berthold Tech, Bad Harzburg, Germany) and total protein determination using the Bradford assay from Bio-Rad.

IL-8 ELISA and Griess reaction

IL-8 protein was quantified using the ELISA from R & D (Wiesbaden, Germany) according to the manufacturer's specification. To this end, cells were grown in 12-well plates and incubated with the respective stimulus for 24 h. Then the cells were washed and the medium was aspirated. Aliquots were used to determine the amounts of IL-8 and the formation of NO using Griess reagent (Sigma) as recommended by the manufacturer. In brief, the accumulation of NO₂^-, a stable end product of NO formation, was measured as an indicator for NO production. 100 L of cell-free culture supernatant were incubated for 15 min with 100 L Griess reagent at room temperature and the absorbance at 540 nm was measured in a microplate reader.

Results

Tlr rna expression in primary human keratinocytes

RNA from unstimulated human keratinocytes was reverse transcribed using oligo dT primers. A panel of hTLR specific primers targeting TLR1–TLR10 specific cDNA was used. This analysis revealed expression of TLR1, TLR2, TLR3, TLR5, and TLR9 but not TLR4, TLR6, TLR7, TLR8, and TLR10 (Figure 1a). The lack of TLR4 transcription under nonstimulatory conditions was surprising but has been previously described by others (^{Kawai et al, 2002}). Repeatedly, TLR3 showed the strongest transcription signal as evaluated in primary keratinocytes of various donors. In order to control for the sensitivity and the specificity of our primers we included HEK293, THP-1 and LPS-stimulated peripheral blood mononuclear cells (Figure 1b) as positive and the corresponding RNA without addition of the reverse transcriptease enzyme (Figure 1c) as negative control.

Figure 1.

Transcription of TLR1–TLR10 was investigated in primary human keratinocytes by RT-PCR. Repeatedly a high expression of TLR1, TLR2, TLR3, and TLR5 was found together with a low transcription signal of TLR9. Interestingly, no transcription signal was found for TLR4. Experiments were repeated in 10 different donors with similar results (M, size marker; 1–10, TLR1-TLR10; 11, GAPDH) (A). As positive control for primer specificity, combined RNA of LPS (10 g per mL) stimulated peripheral blood mononuclear cells, HEK293 and THP-1 cells was used (B). The negative control was performed with an identical reaction mix omitting the reverse transcription step (C). All PCR were run to saturation (40 cycles).

Full figure and legend (180K)

Immunohistochemical staining of cultured human keratinocytes

Primary human keratinocytes were grown on glass slides until confluency. Cells were stained with anti-TLR2 and anti-TLR4. We obtained a clear positive staining for anti-TLR2 (Figure 2a). In contrast, the staining for anti-TLR4 (Figure 2d) showed virtually no difference compared to the isotype control (purified goat serum) (Figure 2g). The specific staining pattern of the antibodies used was verified by using TLR2- and TLR4-transfected HEK 293 cells (Figure 2b,e) as well as untransfected HEK 293 cells (Figure 2c,f). These results paralleled our RT-PCR findings.

Figure 2.

Staining for TLR2 and TLR4 on primary human keratinocytes. Cells were grown on glass slides to confluency. Cells were stained with a polyclonal primary goat anti-TLR antibody (dilution 1:10) and an FITC-labeled rabbit antigoat secondary antibody (dilution 1:50). Repeatedly, the keratinocytes were found to express TLR2 (a) but no specific staining was detected for TLR4 (D). Purified goat serum was used as isotype control (G) and HEK 293 cells transfected with hTLR2 (B) and hTLR4 (E) and untransfected HEK 293 cells (C, F) were used as control for antibody specificity and sensitivity.

Full figure and legend (110K)

NF-B translocation after exposure to S. aureus

Keratinocytes were incubated with 10⁶ staphylococci per mL in SFM. Thereafter, cells were harvested and nuclear extracts were prepared. Gel shift analysis of the separated nuclear extracts showed a clear band for NF-B when the keratinocytes were exposed to S. aureus strain 8325-4 or 50 ng per mL of IL-1 and TNF- (used as positive control) but not when the cells were incubated with medium alone (used as negative control) (Figure 3). Next we screened for translocation of the NF-B subunit p65 (RelA) by immunohistochemistry. As shown in Figure 4 the incubation with S. aureus resulted in a significant translocation of RelA from the cytoplasm into the nucleus of the keratinocytes. Again, IL-1/TNF- was used as positive control whereas medium alone did not induce a significant recruitment of the p65 subunit to the nucleus.

Figure 3.

Nuclear extract gel shift for NF-B. Primary human keratinocytes were grown to confluency in sterile Petri dishes (10⁶ cells) and incubated with TNF- (50 ng per mL)/IL-1 (50 ng per mL) (lane 2), S. aureus strain 8325-4 (lane 3), and medium (lane 4) for 2 h before they were harvested by scraping them off the dish with a sterile cell scraper. Preparation of nuclear extracts and gel shifts were carried out as described in Materials and Methods. LPS-stimulated THP-1 cells served as positive control (lane 1). This experiment showed translocation of NF-B for cytokine and S. aureus stimulated keratinocytes but not for medium-stimulated control cells.

Full figure and legend (115K)

Figure 4.

Staining for RelA in primary human keratinocytes. Cells were incubated with medium (negative control, A), with TNF- (50 ng per mL)/IL-1 (50 ng per mL) (positive control, B), and with S. aureus strain 8325-4 (C) for 2 h before the cells were stained with a primary rabbit anti-p65 antibody followed by an FITC-labeled goat antirabbit antibody. A positive reaction was found after stimulation with cytokines and S. aureus as seen by the changes in staining patterns (unstimulated cells, cytoplasmatic staining pattern; stimulated cells, nuclear staining pattern as denoted by arrows). Note the staining of S. aureus bacteria by their capacity to fix the antibodies through protein A on their surface (C).

Full figure and legend (136K)

In order to quantitate the NF-B activation following incubation with S. aureus, we developed an NF-B-luciferase reporter assay using transient transfections with the luciferase reporter plasmid. This assay showed a significant increase in the activation of NF-B by S. aureus strain 8325-4 and to a lesser extent by strain Newman (Figure 5).

Figure 5.

NF-B translocation was quantified by introducing a luciferase reporter plasmid into the keratinocytes. Transfection efficacy was controlled by parallel transfection of the green-fluorescent-protein-containing control plasmid. Transfection rates were found between 20% and 30% of the cells (as shown in the inserted panel as A and B). Cells were then incubated with S. aureus strains 8324-5 and Newman (mitomycin C inactivated) at an MOI of approximately 50 for 24 h before luciferase units were measured. This experiment showed translocation of NF-B after incubation with S. aureus strain 8325-4 and to a lesser extent with strain Newman. TNF-/IL-1 served as positive control and incubation medium as negative control. Experiments were carried out in triplicate and were repeated at least three times with similar results.

Full figure and legend (34K)

NF-B translocation can be blocked by anti-TLR2 but not anti-TLR4 or a specific inhibitor of the PAFR pathway

We then used the NF-B reporter assay to identify pathogen recognition molecules involved in the activation of human keratinocytes by S. aureus. To this end, keratinocytes were incubated with either anti-TLR2 or anti-TLR4 or the well-described inhibitor of PAFR signaling CV3988. As shown in Figure 6, preincubation with anti-TLR2 but not with anti-TLR4 or the PAFR antagonist CV3988 completely inhibited NF-B activation in the keratinocytes. This inhibition was S. aureus specific as it did not impair the translocation of NF-B following exposure to IL-1/TNF- (data not shown). This experiment showed that NF-B translocation of human keratinocytes is mediated through TLR2 and not TLR4 or PAFR.

Figure 6.

Inhibition of NF-B translocation. Keratinocytes were preincubated for 2 h with the respective anti-TLR2- or anti-TLR4-specific antibody or CV3988 before identical concentrations (MOI 50) of S. aureus strain 8324-5 (mitomycin C inactivated) were added to all culture conditions for a further 24 h. Then, the cells were harvested for determination of luciferase activity. Translocation of NF-B by S. aureus 8325-4 was blocked by preincubation with 10 g per mL of anti-TLR2 (MAB clone TL2.1) but not with anti-TLR4 (MAB clone HTA125) or with the specific PAFR inhibitor CV3988 (10 g per mL), indicating an activation of primary human keratinocytes by S. aureus through the pattern recognition molecule TLR2.

Full figure and legend (10K)

Staphylococcal LTA and PGN both activate human keratinocytes

As various components of staphylococci have been described to act in an immunostimulatory way, we included two commercially available purified cell wall components in our NF-B activation assay. As shown in Figure 7, staphylococcal LTA and PGN both induced NF-B translocation.

Figure 7.

Translocation of NF-B by S. aureus in primary human keratinocytes can be mimicked by the known TLR2 ligands PGN and LTA. Cells were incubated for 24 h with the staphylococcal cell wall components (both in a concentration of 10 g per mL) or strain 8324-5 (MOI 50 and mitomycin C inactivated) before luciferase activity was measured. TNF-/IL-1 and medium served as positive and negative control, respectively.

Full figure and legend (8K)

Exposure of keratinocytes to S. aureus and its cell wall components results in increased transcription of NF-B controlled genes

Our experiments so far established the translocation of NF-B by S. aureus in primary human keratinocytes through a TLR2-dependent mechanism. We next looked for the transcription of NF-B controlled genes by semiquantitative RT-PCR (compared to GAPDH transcription). This experiment revealed for iNOS and COX2 an increase in specific RNA transcription after stimulation. This increase was moderate for COX2 but more pronounced for iNOS. A possible explanation for this difference might be the high level of basal transcription for COX2 already present in unstimulated keratinocytes (Figure 8a).

Figure 8.

Increased transcription of NF-B controlled genes. (A) S. aureus 8325-4 (lane 5) and its cell wall components LTA (lane 3) and PGN (lane 4) induce transcription of the NF-B controlled genes iNOS and COX2. Semiquantitative PCR showed increased transcription of these genes compared to GAPDH. (B) IL-8 transcription was analyzed in more detail by real-time Taqman PCR. Again, we found induction of this gene after stimulation with S. aureus and its components. The figure shows one representative experiment out of three. Medium (A, lane 1) and TNF- (50 ng per mL)/IL-1 (50 ng per mL) (A, lane 2) served as negative and positive control, respectively.

Full figure and legend (55K)

In order to analyze transcription of IL-8, the major mediator of neutrophil attraction, in more detail, we used the quantitative real-time Taqman PCR. This experiment also showed a clear induction of IL-8 by S. aureus and its cell wall components (Figure 8b).

The NF-B-dependent transcription of IL-8 and iNOS is followed by protein translation

In order to search for an enhanced production of the gene products for the iNOS and IL-8 gene we harvested cell culture supernatants after 24 h of stimulation and subjected them to quantification of IL-8 by ELISA and iNOS production by using the Griess reaction (Figure 9a,b). These experiments showed that S. aureus strain 8325-4 as well as the cell wall components LTA and PGN induced IL-8 and iNOS in primary human keratinocytes.

Figure 9.

To investigate staphylococci-induced protein production IL-8 and NO (which reflects iNOS induction) concentrations in culture supernatants of stimulated keratinocyte cultures were evaluated. As shown in panel A, IL-8 was induced by S. aureus and its cell wall components. NO production as measured by the Griess reaction also paralleled the RT-PCR findings of increased proinflammatory gene activation (B). Of note, after challenge with TNF- and IL-1 only a slight increase in NO production was observed. Experiments were carried out in duplicate and were repeated three times.

Full figure and legend (18K)

Discussion

We have investigated the molecular mechanisms leading to the proinflammatory response of human keratinocytes after exposure to S. aureus. These experiments clearly showed that the NF-B-dependent activation of keratinocytes requires TLR2 but not TLR4 or PAFR signaling. Following the contact of the bacteria to this receptor we observed a translocation of the crucial proinflammatory transcription factor NF-B and a transcription as well as a translation of NF-B downstream genes. Interestingly, we found transcription of several TLR genes in the cultured keratinocytes with TLR1, TLR2, TLR3, TLR5, and TLR9 being constitutively transcribed but not TLR4, TLR6, TLR7, TLR8, and TLR10. This study is to our knowledge the first systematic analysis of functional TLR expression in cultured human keratinocytes.

The expression of TLR2 on human keratinocytes has been previously described and the activation of keratinocytes by known TLR2 ligands has been reported (Pivarcsi et al, 2001;^{Uehara et al, 2001;Kawai et al, 2002}). We extended these results by unambiguously identifying TLR2 as the crucial recognition pattern receptor for staphylococci in human keratinocytes. The signaling through TLR2 is followed by the translocation of the proinflammatory transcription factor NF-B and the induction of a proinflammatory response at the mRNA and protein level. This signaling is TLR2 dependent as it can be blocked by an anti-TLR2 specific antibody (but not by anti-TLR4) and cannot be impaired by the specific inhibitor of the PAFR pathway although human keratinocytes have been described to express PAFR (^{Shimada et al, 1998}).

Our finding that primary human keratinocytes do not constitutively express TLR4 mRNA and protein is in accordance with a recent report in which similar results have been published (^{Kawai et al, 2002}) but contrasts with previous publications describing expression of TLR4 as well as signaling of bacterial LPS through this receptor (Pivarcsi et al, 2001;^{Uehara et al, 2001;Song et al, 2002}). One possible explanation might be the inducibility of TLR4 (^{Kawai et al, 2002}), which could account for the different findings.

Given this immediate and reproducible response of cultured keratinocytes to staphylococci and staphylococcal products mediated through TLR2 and given the presence of TLR2 ligands even in coagulase-negative staphylococci (^{Hajjar et al, 2001}), which are present on human skin in high numbers even when it is not inflamed, as well as the estimated 30% of healthy S. aureus carriers, the question remains why the normal human skin does not constantly react to the colonizing bacteria. A possible answer can be found in the expression pattern of TLR2 in normal human skin (^{Kawai et al, 2002}). The expression of this molecule has been found to be strongest in the basal layer(s) of the epidermis whereas only marginal expression was seen in suprabasal layers. Thus, the bacterial products would have to enter through the barrier of the stratum corneum and through the stratum granulosum to gain access to TLR2-expressing keratinocytes. This usually happens only in skin diseases with disturbed barrier functions such as atopic eczema where the staphylococci may contribute to the TH1 inflammatory pattern seen in chronic lesions (which are very often colonized with high numbers of staphylococci) (^{Thepen et al, 1996;Herz et al, 1998}).

Many of the ligands for the TLR molecules have been recently identified and their signaling through the various TLR has been defined in several cell types (for review see^{Janeway and Medzhitov, 2002}). For example, TLR9 has been described to recognize unmethylated CpG DNA (^{Hemmi et al, 2000}), a pathogen-associated signal that is widely suppressed in mammalian DNA. Interestingly, recent work has shown a TH1-like activation of keratinocytes after exposure to such CpG-containing DNA (^{Mirmohammadsadegh et al, 2002}). It is thus conceivable that the expression of TLR9 on the keratinocytes enables the cells to react to pathogen-derived DNA molecules.

TLR5 has been described as a receptor for bacterial flagellin (^{Hayashi et al, 2001}), a microbial component that also plays an important role in the colonization of several bacteria to body surfaces such as the skin.

TLR3 has been identified as a receptor for double-stranded RNA (^{Alexopoulou et al, 2001}). The role of its expression in keratinocytes is not clear at the moment but as for other TLR the recognition of further pattern-associated molecules cannot be excluded.

TLR1 on the other hand has been described to compete with TLR6 for the formation of dimers with TLR2 (^{Hajjar et al, 2001}). For this TLR molecule, an inhibitory effect for a TLR2/TLR6 heterodimer mediated signaling and to a lesser extent for the TLR2 homodimer signaling has been described (^{Hajjar et al, 2001}). As our experiments only found expression of TLR2 but not TLR6 in primary human keratinocytes further experiments are under way to investigate a possible inhibitory signal on TLR2 activation.

As the human skin represents the major barrier to a potentially hostile environment, the development of innate defense mechanisms such as a distinct set of pattern recognition receptors is of vital interest for the host defense. Obviously, the expression of functional receptors for pathogen-associated molecular patterns is required not only on professional epidermal immune cells (e.g., Langerhans cells) but also on epidermal keratinocytes to guarantee normal skin homeostasis.

Future projects are set to elucidate the various interactions of the different TLR in keratinocytes and the mechanisms that orchestrate and regulate the immune response through this family of molecules in human skin.

Notes

^* Pivarcsi A, Reithi B, zell M, et al: Toll-like receptors 2 and 4 are expressed on human keratinocytes and mediate the killing of pathogens. 31st Annual ESDR Meeting, Stockholm, Abstract 224, 2001.

References

References

1.	Alexopoulou L, Holt AC, Medzhitov R & Flavell RA. Recognition of double-stranded RNA and activation of NF-B by toll-like receptor 3. Nature (2001) 413: 732–738. | Article | PubMed | ISI | ChemPort |
2.	Bauer S, Kirschning CJ & Hacker H et al. Human TLR9 confers responsiveness to bacterial DNA via species-specific CpG motif recognition. Proc Natl Acad Sci USA (2001) 98: 9237–9242. | Article | PubMed | ChemPort |
3.	Bibel DJ, Aly R, Shinefield HR & Maibach HI. The Staphylococcus aureus receptor for fibronectin. J Invest Dermatol (1983) 80: 494–496. | Article | PubMed | ISI | ChemPort |
4.	Brand K, Eisele T & Kreusel U et al. Dysregulation of monocytic nuclear factor-B by oxidized low-density lipoprotein. Arterioscler Thromb Vasc Biol (1997) 17: 1901–1909. | PubMed | ISI | ChemPort |
5.	Cario E, Brown D, McKee M, Lynch-Devaney K, Gerken G & Podolsky DK. Commensal-associated molecular patterns induce selective toll-like receptor-trafficking from apical membrane to cytoplasmic compartments in polarized intestinal epithelium. Am J Pathol (2002) 160: 165–173. | PubMed | ISI | ChemPort |
6.	Cho SH, Strickland I, Boguniewicz M & Leung DY. Fibronectin and fibrinogen contribute to the enhanced binding of Staphylococcusaureus to atopic skin. J Allergy Clin Immunol (2001) 108: 269–274. | Article | PubMed | ISI | ChemPort |
7.	Chuang T & Ulevitch RJ. Identification of hTLR10: A novel human toll-like receptor preferentially expressed in immune cells. Biochim Biophys Acta (2001) 1518: 157–161. | Article | PubMed | ISI | ChemPort |
8.	Ezepchuk YV, Leung DY, Middleton MH, Bina P, Reiser R & Norris DA. Staphylococcal toxins and protein: A differentially induce cytotoxicity and release of tumor necrosis factor- from human keratinocytes. J Invest Dermatol (1996) 107: 603–609. | Article | PubMed | ISI | ChemPort |
9.	Hajjar AM, O'Mahony DS, Ozinsky A, Underhill DM, Aderem A, Klebanoff SJ & Wilson CB. Cutting edge: Functional interactions between toll-like receptor (TLR) 2 and TLR1 or TLR6 in response to phenol-soluble modulin. J Immunol (2001) 166: 15–19. | PubMed | ISI | ChemPort |
10.	Hayashi F, Smith KD & Ozinsky A et al. The innate immune response to bacterial flagellin is mediated by toll-like receptor 5. Nature (2001) 410: 1099–1103. | Article | PubMed | ISI | ChemPort |
11.	Hemmi H, Takeuchi O & Kawai T et al. A toll-like receptor recognizes bacterial DNA. Nature (2000) 408: 740–745. | Article | PubMed | ISI | ChemPort |
12.	Herz U, Bunikowski R & Renz H. Role of T cells in atopic dermatitis. New aspects on the dynamics of cytokine production and the contribution of bacterial superantigens. Int Arch Allergy Immunol (1998) 115: 179–190. | Article | PubMed | ISI | ChemPort |
13.	Honda S, Migita K & Hirai Y et al. Induction of COX-2 expression by nitric oxide in rheumatoid synovial cells. Biochem Biophys Res Commun (2000) 268: 928–931. | Article | PubMed | ISI | ChemPort |
14.	Janeway CA, Jr & Medzhitov R. Innate immune recognition. Annu Rev Immunol (2002) 20: 197–216. | Article | PubMed | ISI | ChemPort |
15.	Kawai K, Shimura H, Minagawa M, Ito A, Tomiyama K & Ito M. Expression of functional toll-like receptor 2 on human epidermal keratinocytes. J Dermatol Sci (2002) 30: 185–194. | Article | PubMed | ISI | ChemPort |
16.	Lemjabbar H & Basbaum C. Platelet-activating factor receptor and ADAM10 mediate responses to Staphylococcus aureus in epithelial cells. Nat Med (2002) 8: 41–46. | Article | PubMed | ISI | ChemPort |
17.	Mackman N. Lipopolysaccharide induction of gene expression in human monocytic cells. Immunol Res (2000) 21: 247–251. | PubMed | ISI | ChemPort |
18.	Mempel M, Schmidt T, Weidinger S, Schnopp C, Foster T, Ring J & Abeck D. Role of Staphylococcus aureus surface-associated proteins in the attachment to cultured HaCaT keratinocytes in a new adhesion assay. J Invest Dermatol (1998) 111: 452–456. | Article | PubMed | ISI | ChemPort |
19.	Mempel M, Schnopp C & Hojka M et al. Invasion of human keratinocytes by Staphylococcus aureus and intracellular bacterial persistence represent haemolysin-independent virulence mechanisms that are followed by features of necrotic and apoptotic keratinocyte cell death. Br J Dermatol (2002) 146: 943–951. | Article | PubMed | ISI | ChemPort |
20.	Mirmohammadsadegh A, Maschke J, Basner-Tschakarjan E, Bar A & Hengge UR. Induction of acute phase response genes in keratinocytes following exposure to oligodeoxynucleotides. J Mol Med (2002) 80: 377–383. | Article | PubMed | ISI | ChemPort |
21.	Rad R, Gerhard M & Lang R et al. The Helicobacter pylori blood group antigen-binding adhesin facilitates bacterial colonization and augments a nonspecific immune response. J Immunol (2002) 168: 3033–3041. | PubMed | ISI | ChemPort |
22.	Schwandner R, Dziarski R, Wesche H, Rothe M & Kirschning CJ. Peptidoglycan- and lipoteichoic acid-induced cell activation is mediated by toll-like receptor 2. J Biol Chem (1999) 274: 17406–17409. | Article | PubMed | ISI | ChemPort |
23.	Shimada A, Ota Y, Sugiyama Y, Sato S, Kume K, Shimizu T & Inoue S. In situ expression of platelet-activating factor (PAF)-receptor gene in rat skin and effects of PAF on proliferation and differentiation of cultured human keratinocytes. J Invest Dermatol (1998) 110: 889–893. | Article | PubMed | ISI | ChemPort |
24.	Song PI, Park YM & Abraham T et al. Human keratinocytes express functional CD14 and toll-like receptor 4. J Invest Dermatol (2002) 119: 424–432. | Article | PubMed | ISI | ChemPort |
25.	Thepen T, Langeveld-Wildschut EG, Bihari IC, van Wichen DF, van Reijsen FC, Mudde GC & Bruijnzeel-Koomen CA. Biphasic response against aeroallergen in atopic dermatitis showing a switch from an initial TH2 response to a TH1 response in situ: An immunocytochemical study. J Allergy Clin Immunol (1996) 97: 828–837. | PubMed | ISI | ChemPort |
26.	Tokura Y, Furukawa F, Wakita H, Yagi H, Ushijima T & Takigawa M. T-cell proliferation to superantigen-releasing Staphylococcus aureus by MHC class II-bearing keratinocytes under protection from bacterial cytolysin. J Invest Dermatol (1997) 108: 488–494. | Article | PubMed | ISI | ChemPort |
27.	Uehara A, Sugawara S, Tamai R & Takada H. Contrasting responses of human gingival and colonic epithelial cells to lipopolysaccharides, lipoteichoic acids and peptidoglycans in the presence of soluble CD14. Med Microbiol Immunol (2001) 189: 185–192. | Article | PubMed | ISI | ChemPort |
28.	Walev I, Martin E, Jonas D, Mohamadzadeh M, Muller-Klieser W, Kunz L & Bhakdi S. Staphylococcal -toxin kills human keratinocytes by permeabilizing the plasma membrane for monovalent ions. Infect Immun (1993) 61: 4972–4979. | PubMed | ISI | ChemPort |
29.	Wang Q, Dziarski R, Kirschning CJ, Muzio M & Gupta D. Micrococci and peptidoglycan activate TLR2MyD88IRAKTRAFNIKIKKNF-B signal transduction pathway that induces transcription of interleukin-8. Infect Immun (2001) 69: 2270–2276. | Article | PubMed | ISI | ChemPort |

Acknowledgments

We gratefully acknowledge the skillful technical assistance of S. Bogner, B. Heuser, and G. Roth. Bacterial strains were kindly provided by T. Foster, University of Dublin. We also thank M. Braun-Falco (TU Munich) for providing the GFP-expressing plasmid, C. Kirschning (TU Munich) for providing TLR2 and TLR4 expression plasmids, and S. Bauer (TU Munich) for help with the TLR expression data. This work was funded in part by grant 01GC0104 from the German Federal Ministery of Science and Education (BMBF), grant UW-S15T03 (Project 3b) from the National Genome Research Network (NGFN), and grant KKF 8760160 from the Technical University Munich.