Vigilant Keratinocytes Trigger Pathogen-Associated Molecular Pattern Signaling in Response to Streptococcal M1 Protein

Abstract

The human skin exerts many functions in order to maintain its barrier integrity and protect the host from invading microorganisms. One such pathogen is Streptococcus pyogenes, which can cause a variety of superficial skin wounds that may eventually progress into invasive deep soft tissue infections. Here we show that keratinocytes recognize soluble M1 protein, a streptococcal virulence factor, as a pathogen-associated molecular pattern to release alarming inflammatory responses. We found that this interaction initiates an inflammatory intracellular signaling cascade involving the activation of the mitogen-activated protein kinases extracellular signal-regulated kinase (ERK), p38, and Jun N-terminal protein kinase and the subsequent induction and mobilization of the transcription factors NF-κB and AP-1. We also determined the imprint of the inflammatory mediators released, such as interleukin-8 (IL-8), growth-related oncogene alpha, migration inhibitory factor, extracellular matrix metalloproteinase inducer, IL-1α, IL-1 receptor a, and ST2, in response to streptococcal M1 protein. The expression of IL-8 is dependent on Toll-like receptor 2 activity and subsequent activation of the mitogen-activated protein kinases ERK and p38. Notably, this signaling seems to be distinct for IL-8 release, and it is not shared with the other inflammatory mediators. We conclude that keratinocytes participate in a proinflammatory manner in streptococcal pattern recognition and that expression of the chemoattractant IL-8 by keratinocytes constitutes an important protective mechanism against streptococcal M1 protein.

INTRODUCTION

In order to cause infection, invasive microorganisms have to surmount protective barriers, such as skin, the respiratory mucosa, or the gastrointestinal tract. Among these potential ports of entry, the skin is the largest integumentary organ. Notably, the skin is constantly colonized not only with commensals but also sometimes with pathogenic bacteria (1), and thus, special precaution is needed to prevent their invasion. This is indeed a difficult task, and because of their large size, phagocytosing cells cannot continuously patrol the entire skin surface and scan for intruders. Therefore, the skin has to rely on sophisticated alert systems that send inflammatory signals, which in turn evoke the recruitment of neutrophils, macrophages, and other immune cells. Keratinocytes are the most prevalent cell type of the skin, and they are equipped with such alert systems (2, 3). Among these, Toll-like receptors (TLRs) play an important role in the early host response to an invading pathogen. Once activated, TLRs can trigger a number of immune reactions, such as the secretion of chemotactic cytokines (4) and mobilization of antimicrobial peptides (5, 6). The TLR arsenal in keratinocytes is extensive, and it has been found that these cells express functional TLR1, TLR2, TLR3, TLR4, TLR5, and TLR9 (7, 8).

The Gram-positive bacterium Streptococcus pyogenes (group A streptococcus) is an important human pathogen causing a variety of superficial infections of the skin (impetigo and erysipelas) and throat (pharyngitis) (9). Under rare and unfortunate circumstances, these complications can lead to severe life-threatening systemic infections, such as necrotizing fasciitis and streptococcal toxic shock syndrome, often associated with high rates of morbidity and mortality (10–12). To cause these conditions, S. pyogenes has evolved a panel of secreted and surface-bound virulence factors that enable the bacterium to effectively infect the human host. Among these factors, M and M-like proteins are probably the best-characterized streptococcal virulence determinants. Due to their high abundance, these proteins are used to classify different serotypes, and over 200 have currently been identified (13). M and M-like proteins are composed of two polypeptide chains that form an alpha-helical coiled-coil structure which can interact with a number of plasma proteins, such as factor H, fibrinogen, and serum albumin. It is worth mentioning that M1 protein is normally anchored to the bacterial cell wall, but it can also be released from the bacterial surface, as seen, for instance, in patients with necrotizing fasciitis (14). The release can occur endogenously or by the action of host-derived proteinases (14, 15). Once released, M1 protein can cause systemic inflammatory reactions by targeting neutrophils, monocytes, and T cells (14, 16–18). Interestingly, the interaction of M1 protein with these cells involves different receptors. While the interaction with neutrophils is mediated by cross-linking β₂ integrins (14), M1 protein binds to TLR2 on monocytes (16) and to T-cell receptors on T cells (18). These interactions can cause a number of inflammatory reactions, such as the mobilization of heparin binding protein, a potent inducer of vascular leakage and sepsis biomarker (19), and various proinflammatory and Th1-type cytokines (14, 16–18). Though it has been shown that the release of these proteins contributes to the pathology of an infection, it has not been reported whether the release of M1 protein in proximity to the site of infection can act as a pathogen-associated molecular pattern (PAMP) that helps to alert the innate immune system.

In this study, we investigated if keratinocytes react immunologically to M1 protein released from the bacterial surface. We further wished to gain insight into the molecular mechanisms by which these cells respond to streptococcal infection. Our data show that when keratinocytes encounter M1 protein, it induces an inflammatory intracellular pathway resulting in the release of proinflammatory mediators. Together our findings suggest that the recognition and proinflammatory response of keratinocytes to this streptococcal surface protein is an important step in alarming the host defense about the intruder and preventing further development into a severe infection.

MATERIALS AND METHODS

Cell culture and reagents.

Cells of the HaCaT cell line, a spontaneously immortalized human keratinocyte cell line, were cultured in serum-free keratinocyte-SFM without calcium chloride (Invitrogen, Carlsbad, CA) and were supplemented with 25 μg/ml bovine pituitary extract, 0.2 ng/ml epidermal growth factor, and 1× Gibco 100× antibiotic-antimycotic (Life Technologies, Carlsbad, CA) (20). M1 protein was purified according as previously described (21). Briefly, an isogenic AP1 mutant strain expressing an M1 protein lacking the cell wall-anchoring region was used. Consequently, the streptococcal protein accumulated in the growth medium, from which it was purified on the basis of its high affinity to human fibrinogen. M1 protein is the only fibrinogen-binding protein of the AP1 strain (15, 22), and after purification, no other contaminants are detected (16). All other streptococcal proteins (protein M4, protein M5, protein M22, and protein SIC) were purified as described earlier (23–25). For Western blot analysis, antibodies to phospho-extracellular signal-regulated kinase 1/2 (phospho-ERK1/2), phospho-ribosomal r6 kinase 1 (phospho-RSK1), phospho-Jun N-terminal protein kinase (phospho-JNK), phospho-p38, and IκBα were purchased from Cell Signaling Technologies (Beverly, MA). TLR-blocking antibodies, a polyclonal antibody (PAb) control, PAb to human TLR2, and PAb to human TLR4 were purchased from InvivoGen (San Diego, CA). The ERK, p38, and JNK inhibitors FR180204, SB202190, and SP600125, respectively, were purchased from Sigma-Aldrich (St. Louis, MO, USA) (26). A preincubation with blocking antibodies and inhibitors was used before M1 protein was added to the cell-antibody-inhibitor supernatants.

RNA isolation and microarray analysis.

Total cellular RNA was isolated from cultured HaCaT cells using a GenElute mammalian total RNA miniprep kit (Sigma-Aldrich, St. Louis, MO) according to the manufacturer's instructions. Cells were seeded into 12-well plates and treated with 5 μg/ml M1 protein for 6 h or left untreated before isolation. Gene expression analyses were performed at the SCIBLU Genomics Centre, Lund University, Lund, Sweden. Briefly, purity and concentration were determined with an ND-1000 spectrophotometer (NanoDrop Technologies Inc., Wilmington, DE), and integrity was determined using a model 2100 bioanalyzer (Agilent Technologies Inc., Santa Clara, CA). Microarray experiments were performed in triplicate with three control samples and three M1 protein-stimulated samples. Gene expression was analyzed using the Human Gene (v2.0) ST array (Affymetrix, Santa Clara, CA) according to the manufacturer's instructions. Affymetrix chip and experimental quality analyses were performed using Expression Console software (v1.1.2; Affymetrix) by applying the robust multichip average normalization method. Genes were considered differentially expressed when the fold change in expression was either ≥1.6 (upregulated) or ≤0.6 (down-regulated) and the P value was less than 0.0001.

Western blot analysis.

Cells were lysed at 4°C for 30 min in Lysis Buffer 6 (R&D Systems, Minneapolis, MN). Total protein concentrations in lysates were measured with an ND-1000 spectrophotometer (NanoDrop Technologies Inc.). Equal amounts of total protein were loaded onto 10% SDS-polyacrylamide gels in sample buffer. Proteins were transferred to polyvinylidene difluoride (PVDF) membranes (Merck Millipore, Billerica, MA) and blocked with Tris-buffered saline containing 0.1% Tween 20 and 5% nonfat dry milk. Membranes were incubated with the appropriate primary antibody overnight at 4°C, followed by incubation with horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature. Blots were developed by using SuperSignal West Pico chemiluminescent substrate (Thermo Fisher Scientific, Waltham, MA) and visualized with a ChemiDoc MP imaging system (Bio-Rad Laboratories, Hercules, CA).

EMSA.

Nuclear extracts were prepared from 8 × 10⁶ cells and lysed using a Pierce nuclear and cytoplasmic extraction reagent kit (Thermo Fisher Scientific). Electrophoretic mobility shift assays (EMSAs) were performed using a Pierce LightShift chemiluminescent EMSA kit according to the manufacturer's instructions. Briefly, AP-1 or NF-κB consensus oligonucleotide probes (for AP-1, 5′-CGC TTG ATG AGT CAG CCG GAA-3′; for NF-κB, 5′-AGT TGA GGG GAC TTT CCC AGC C-3′) were end labeled with biotin (InvivoGen, San Diego, CA). Reaction mixtures consisted of 10 μg nuclear extract in binding buffer, and the probe-protein complexes were separated on a 6% polyacrylamide gel before they were electrotransferred onto a 0.45-μm-pore-size Biodyne B nylon membrane (Thermo Fisher Scientific). Membranes were cross-linked using the ChemiDoc MP imaging system (Bio-Rad Laboratories) with UV light for 10 min. Probes were detected using a chemiluminescent nucleic acid detection module (Thermo Fisher Scientific) and visualized with the ChemiDoc MP imaging system (Bio-Rad Laboratories).

Human XL cytokine array.

Keratinocyte culture supernatants were analyzed with a Proteome Profiler human XL cytokine array (R&D Systems) according to the manufacturer's protocol. Cells were seeded to 70 to 80% confluence in 12-well plates and then either treated with 5 μg/ml M1 protein or left untreated for 24 h. Supernatants were collected and centrifuged at 500 × g for 5 min. Membranes were incubated with 400 μl of the supernatant overnight at 4°C. Chemiluminescence was detected by using the ChemiDoc MP imaging system, and the mean pixel density was analyzed by the use of Image Lab software (v4.1; Bio-Rad Laboratories).

ELISA.

Cytokine levels in the keratinocyte culture supernatants were measured by a sandwich enzyme-linked immunosorbent assay (ELISA; DuoSet ELISA development kit; R&D Systems) according to the manufacturer's protocol. Briefly, the plates were incubated overnight at 4°C with 100 μl capture antibody diluted in coating buffer. The contents of the wells were blocked with coating buffer containing 0.5% bovine serum albumin. Standards and samples diluted in assay buffer were added to the wells, and then detection antibody diluted in assay buffer was added and the mixture was incubated. The contents of the wells were incubated with streptavidin conjugated to horseradish peroxidase before the visualization of antibody binding by adding 100 μl 3,3′,5,5′-tetramethylbenzidine (TMB) solution, and the reaction was stopped by the addition of 1.8 M H₂SO₄. The absorbance at 450 nm was measured with a microplate reader (Viktor3 model 1420; PerkinElmer, Waltham, MA). All measurements were performed in duplicate. Data are presented as the level of cytokine expression as a percentage of that induced by M1.

Immunoelectron microscopy.

HaCaT cells were incubated with M1 protein for 30 min at 37°C. For transmission electron microscopy, samples were embedded in Epon resin, sectioned, and subjected to antigen retrieval with meta-periodate as recently described in detail (27). Sections were labeled with rabbit anti-M1 protein and rat anti-TLR2 followed by gold-conjugated goat anti-rabbit immunoglobulin (5 nm) and goat anti-rat immunoglobulin (10 nm) antibodies. Samples were observed in a Philips/FEI CM 100 transmission electron microscope at the Core Facility for Integrated Microscopy, Panum Institute, University of Copenhagen.

Statistical analysis.

All data are expressed as the mean ± standard deviation (SD) from at least three independent experiments. Differences were assessed by the Student t test. A probability (P) value of <0.05 was considered significantly significant. Analyses were performed using GraphPad Prism statistical software (version 6.0a for Mac OS X; GraphPad Software Inc., La Jolla, CA).

Microarray data accession number.

The data received in this analysis have been deposited in NCBI's Gene Expression Omnibus (GEO) database and are accessible through GEO series accession number GSE61993.

RESULTS

M1 protein triggers an inflammatory response in human keratinocytes.

In order to study the role of M1 protein in evoking an inflammatory response in human keratinocytes, RNA extraction and microarray analysis were performed following 6 h of incubation with M1 protein. The data received from this analysis have been deposited in NCBI's Gene Expression Omnibus (28) and are accessible through GEO series accession number GSE61993. The analysis revealed 498 differently expressed genes, of which 304 genes were upregulated and 194 genes were downregulated. Having a closer look at the upregulated genes with a function in inflammation, we found 18 genes coding for proteins that are involved in intracellular signaling. Out of those, TLR4, TRAF4, TICAM1, and MYD88 are parts of the TLR and intereleukin-1 (IL-1) receptor (IL-1R) pathways (Table 1). In addition, we identified a panel of genes that act downstream of these signaling networks and function as transcription factors in the AP-1 complex (FOSB, FOS, JUN, JUND, JUNB) or NF-κB complex (REL, NFKB1, NFKB2). Other genes involved in inflammation and intracellular signaling code for transcription factors (CEBPB, CEBPD, CITED4), the bradykinin receptor (BDKRB1), as well as the plasminogen activator and its receptor (Table 1). Together, these findings suggest that M1 protein is able to trigger inflammatory signaling cascades in human keratinocytes.

TABLE 1.

Increased expression of genes for inflammatory signal transducer receptors and intracellular signaling in keratinocytes in response to streptococcal M1 protein^a

Gene	Gene product	FCb	Function
FOSB	FBJ murine osteosarcoma viral oncogene homolog B	7.3	AP-1 transcription factor complex
FOS	FBJ murine osteosarcoma viral oncogene homolog	5.8	AP-1 transcription factor complex
JUN	Jun protooncogene	4.2	AP-1 transcription factor complex
TLR4	TLR4	2.7	Recognition and activation of innate immunity
JUND	Jun D protooncogene	2.5	AP-1 transcription factor complex
REL	c-REL, v-rel reticuloendotheliosis viral oncogene homolog (avian)	2.5	NF-κB transcription factor complex
TRAF4	TNF receptor-associated factor 4	2.4	Mediator of signal transduction of IL-1R/Toll-like and TNF receptors
JUNB	Jun B protooncogene	2.3	AP-1 transcription factor complex
CEBPB	CCAAT/enhancer binding protein (C/EBP) beta	2.3	Transcription factor with inflammatory responses
PLAUR	Plasminogen activator, urokinase receptor	2.3	The receptor of urokinase, plasminogen activator
TICAM1	TLR adaptor molecule 1	2.0	Mediator of signal transduction of TLRs
NFKB1	Nuclear factor of kappa light polypeptide gene enhancer in B cells 1	1.9	NF-κB transcription factor complex (p105, p50)
MYD88	Myeloid differentiation primary response gene 88	1.9	Essential signal transducer of IL-1R and TLR
BDKRB1	Bradykinin receptor B1	1.8	The receptor of bradykinin
CEBPD	CCAAT/enhancer binding protein (C/EBP) delta	1.7	Transcription factor with inflammatory responses
PLAU	Plasminogen activator, urokinase	1.7	A serine protease converting plasminogen to plasmin and degrading extracellular matrix
NFKB2	Nuclear factor of kappa light polypeptide gene enhancer in B cells 2	1.7	NF-κB transcription factor complex (p100, p52)
CITED4	Cbp/p300-interacting transactivator with Glu/Asp-rich carboxy-terminal domain 4	1.7	Transcriptional coactivator

^aDetermined by microarray analysis.

^bFC, fold change between M1 protein-induced and control mRNA expression.

Once activated, keratinocytes are known to release a number of soluble factors involved in cell proliferation, host defense, or inflammation (29–31). Table 2 provides a summary of the 20 upregulated genes coding for effector molecules that are produced upon stimulation with M1 protein. The levels of expression of genes coding for IL-8 and chemokine C-C motif ligand 20 (CCL20) were increased as much as 32.9- and 20.2-fold, respectively, but a considerable rise in the levels of expression of CXCL1 and CXCL3 was also noted. The four genes code for chemokines that are known for their ability to attract immune cells, including neutrophils, lymphocytes, and monocytes (32–34). Other genes that were found to be upregulated were TNFAIP, IL1RL1, CSF2, IL1A, IL1B, and IL1F9, which are important in either inducing or modulating an inflammatory environment. Finally, we identified a third group of genes, namely, HAS2, CTGF, HBEGF, IL-24, IL-20, RNASE7, VEGFA, F2RL1, and TGFA, that have a more direct effect on wound healing and the location of an infection rather than supporting immune responses (Table 2). The remaining genes that were upregulated mainly code for transcription factors with unknown or undefined targets, while most of the downregulated genes identified play a role in cell cycle regulation, cell replication, metabolic processes, and the transcription of regulatory genes. Altogether, the microarray analysis revealed that stimulation of keratinocytes with M1 protein leads to multiple intracellular signaling events and induction of inflammatory reactions.

TABLE 2.

Increased expression of genes for inflammatory mediators in keratinocytes in response to streptococcal M1 protein^a

Gene	Gene product	FCb	Function
IL-8	IL-8	32.9	Inducer of chemotaxis in target cells, primarily neutrophils
CCL20	CCL20, MIP3A	20.2	Strongly chemotactic for lymphocytes but also an attractor of neutrophils
HAS2	Hyaluronan synthase 2	17.7	Producer of hyaluronan molecules
TNFAIP	TNF-α-induced protein 3	10.6	Anti-inflammatory protein
CXCL1	Chemokine (C-X-C motif) ligand 1 (GROα)	10.0	Neutrophil chemoattractant activity
CTGF	Connective tissue growth factor	9.8	Matricellular protein important for wound healing
HBEGF	Heparin-binding epidermal growth factor-like growth factor	9.5	Membrane-anchored chemotactic mitogen
IL1RL1	IL-1R-like 1 (ST2)	7.5	The receptor of IL-33
IL-24	IL-24	5.5	Cytokine of the IL-10 family, important for wound healing
IL-20	IL-20	5.4	Regulator of proliferation and differentiation of keratinocytes during inflammation
CSF2	Granulocyte-macrophage colony-stimulating factor 2	4.9	A cytokine that controls the production, differentiation, and function of granulocytes and macrophages
RNASE7	RNase, RNase A family 7	4.8	Skin-derived antimicrobial protein 2 (SAP-2) with RNase activity
IL1B	IL-1β	4.7	A proinflammatory cytokine
VEGFA	Vascular endothelial growth factor A	3.7	Mediator of increased vascular permeability
IL1A	IL-1α	3.4	A proinflammatory cytokine
IL-11	IL-11	3.1	A proinflammatory cytokine
CXCL3	Chemokine (C-X-C motif) ligand 3 (GROγ)	2.8	Chemotactic for monocytes and neutrophils
F2RL1	Coagulation factor II (thrombin) receptor-like 1, protease-activated receptor 2 (PAR2)	2.7	Modulator of inflammatory responses and a sensor for proteolytic enzymes
TGFA	Transforming growth factor α	2.6	A ligand for the epidermal growth factor receptor
IL1F9	IL-1 family, member 9	2.5	A ligand to IL-1R-like 2

^aDetermined by microarray analysis.

^bFC, fold change between M1 protein-induced and control mRNA expression.

M1 protein triggers MAPK activation and subsequent AP-1 and NF-κB transcription factor activity.

TLR signaling has been described to activate mitogen-activated protein (MAP) kinase (MAPK) and NF-κB transcription factor activity in order to induce inflammatory responses (35). As our microarray analysis also revealed increased expression of AP-1 and NF-κB subunits, downstream targets of MAPKs, we wished to study if the stimulation of keratinocytes by M1 protein leads to MAPK activation. Thus, whole-cell lysates from M1 protein-activated keratinocytes were subjected to Western blot analysis using antibodies against the phosphorylated state of extracellular signal-regulated kinases (ERKs), downstream ribosomal r6 kinase (RSK), p38 MAP kinase (p38), and Jun N-terminal kinase (JNK) (Fig. 1a). GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was used as a loading control, and nonstimulated keratinocytes were used to determine the background level of phosphorylation. The results show a fast activation of ERK, especially ERK2, already at 30 min after M1 protein exposure. RSK activation was also noted after 30 min; however, in contrast to the findings for ERK1/2, the signal continued to rise for up to 1 h. This was followed by p38 activation after 1 h and the activation of JNK, where phosphorylation was detected after 2 h. Together, these findings suggest that the activation of MAPKs is highly regulated and has a specific order to orchestrate their executive functions.

FIG 1.

MAPK and NF-κB activation following M1 protein exposure. Lysates from HaCaT cells stimulated with 5 μg/ml streptococcal M1 protein collected at the indicated time points were analyzed by Western blotting. (a) Phosphorylation of ERK1/2 at Thr202/Tyr204 and Thr185/Tyr187, RSK at Ser380, p38 at Thr180/Tyr182, and JNK at Thr183/Tyr185. p, phosphorylated. (b) Degradation of IκBα. Representative blots from three independent experiments are shown. GAPDH was used as a loading control.

Activation of the transcription factor NF-κB and translocation of NF-κB from the cytoplasm into the nucleus are important steps in the TLR signaling pathway. They are initiated by the phosphorylation, ubiquitination, and subsequent degradation of the NF-κB inhibitor, IκBα. Indeed, we observed that at approximately 2 h after stimulation with M1 protein, IκBα starts to disappear, pointing to activation of NF-κB (Fig. 1b).

AP-1 is also a transcription factor and, like NF-κB, is induced upon TLR signaling. To follow the mobilization of the two transcription factors to their target destination, electrophoretic mobility shift assays (EMSAs) were carried out. The EMSA results showed the binding of both AP-1 and NF-κB consensus sequence-binding probes, demonstrating the presence and DNA binding activity of both AP-1 and NF-κB inside the nucleus in keratinocytes at 2 to 4 h of stimulation with M1 protein (Fig. 2). In summary, the data show that stimulation with M1 protein triggers a chain of signaling events in keratinocytes that eventually lead to the translocation and activation of transcription factors inside the nucleus.

FIG 2.

M1 protein induces nuclear localization and activation of AP-1 and NF-κB transcription factors in keratinocytes. The results of EMSAs of nuclear extracts from keratinocytes cultured in the absence or presence of M1 protein (5 μg/ml) obtained at the indicated time points are shown. Ten-microgram nuclear extracts were analyzed by EMSA using labeled oligonucleotide probes corresponding to the consensus AP-1 DNA binding sequence (a) and to the consensus NF-κB DNA binding sequence (b). The results shown are representative of those from three independent experiments yielding comparable results.

Determination of the cytokine imprint from keratinocytes exposed to M1 protein.

In a next series of experiments, we characterized the cytokine response to M1 protein-activated keratinocytes. We used a human XL cytokine array kit which allows the simultaneous measurement of 102 different cytokines, chemokines, and growth factors. Keratinocytes were incubated with M1 protein or left untreated. After 24 h of incubation, supernatants were collected and subjected to semiquantitative cytokine analysis. Figures 3a and b show that 17 secreted proteins were identified, and 10 of those were upregulated upon stimulation. For further characterization, we focused only on those proteins with a major function in inflammation. This group of proteins included IL-8, growth-related oncogene alpha (GROα), macrophage migration inhibitory factor (MIF), extracellular matrix metalloproteinase inducer (EMMPRIN), IL-1α, IL-1Ra, and ST2. The secretion of these seven proteins was further confirmed by ELISA. We also included TLR2 and TLR4 blocking antibodies to investigate the role of TLR signaling in the induction of these cytokines (Fig. 4a). The expression of IL-8 was significantly decreased when the inhibitors were used (for the TLR4 inhibitor, P = 0.0080; for the TLR2 inhibitor, P < 0.0001), especially when TLR2 was blocked. The experiments also show that treatment with the TLR-blocking antibodies seems to slightly upregulate the mobilization of some cytokines, such as IL-1α, GROα, and MIF. Although some of the observed increases were significant, these do not necessarily need to have a biological impact.

FIG 3.

Cytokine imprint from keratinocytes exposed to M1 protein. Keratinocytes were incubated in the presence or absence of M1 protein (5 μg/ml) for 24 h. (a) Cell supernatants were analyzed using a human cytokine XL array. The numbers correspond to the numbers in the bar graph in panel b. (b) Mean pixel densities were quantified by Image Lab software (v4.1; Bio-Rad Laboratories), and the identities of the respective cytokines are indicated. Dkk-1, Dickkopf WNT signaling pathway inhibitor 1; FGF-19, fibroblast growth factor 19; PDGF-AA, platelet-derived growth factor, AA chain; PLAUR, plasminogen activator, urokinase receptor; VEGF, vascular endothelial growth factor.

FIG 4.

M1 protein-induced release of IL-8 is dependent on TLR2. (a) Keratinocytes were preincubated with TLR blocking antibodies (5 μg/ml) for 30 min and then incubated in the presence or absence of M1 protein (5 μg/ml) for 24 h. Cell supernatants were analyzed by ELISA. CTRL, control; AB, antibody. The data represent the average percent ± SD from at least three independent experiments analyzed by the Student t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (b) HaCaT cells were incubated with M1 protein for 30 min at 37°C. Cells were then fixed and prepared for transmission electron microscopy. Samples were first labeled with rabbit anti-M1 protein and rat anti-TLR2, followed by gold-conjugated goat anti-rabbit immunoglobulin (5 nm) and goat anti-rat immunoglobulin (15 nm). The colocalization of M1 protein and TLR2 was observed. Bar, 500 nm.

Based on these findings, we next investigated whether M1 protein is able to interact with TLR2 on the surface of keratinocytes. Figure 4b depicts immune electron micrographs of HaCaT cells that were stimulated with M1 protein. Indeed, immunostaining revealed that M1 protein and TLR2 colocalize, corroborating the TLR2 dependence of IL-8 release.

In order to study if the MAPK signaling pathway is involved in IL-8 release, we also employed a panel of MAPK inhibitors. Figure 5a shows that an ERK inhibitor (FR180204) and a p38 inhibitor (SB202190) significantly block the release of the chemokines (P < 0.0001). Use of a combination of both inhibitors did not further downregulate IL-8 secretion. When a JNK inhibitor, SP600125, was tested, no effect on IL-8 release was seen (data not shown). Interestingly, no significant effect on the secretion pattern upon ERK and p38 inhibition was observed for the other six proteins analyzed (GROα, MIF, EMMPRIN, IL-1α, IL-1Ra, and ST2), except that GROα levels were lowered in the presence of the p38 inhibitor (for p38 inhibition, P = 0.0271; for ERK and p38 inhibition, P = 0.0372). Finally, we tested whether the release of IL-8 is restricted to stimulation with M1 protein or if other M proteins are also capable of inducing its release from keratinocytes. Therefore, we stimulated cells with M1, M4, M5, and M22 proteins, while protein SIC, a non-M or M-like soluble streptococcal protein, served as a control. Figure 5b depicts that, in addition to M1 protein, M5 protein was also able to stimulate the cells, while M4 and M22 proteins as well as protein SIC had no effect. Taken together, these data show that the secretion of IL-8 is dependent on TLR signaling and MAPK pathways, summarized in Fig. 6, while other inflammatory mediators may utilize other pathways.

FIG 5.

Effect of signal transduction inhibitors and different M-protein serotypes on IL-8 release from keratinocytes. (a) HaCaT cells were preincubated with MAPK inhibitors (10 μM the ERK inhibitor FR180204 or 1 μM the p38 inhibitor SB202190, or both) for 1 h or left untreated before the addition of M1 protein. Cell supernatants were collected after 24 h and analyzed by ELISA. The data represent the average percent ± SD from at least three independent experiments. (b) HaCaT cells were incubated with M1, M4, M5, and M22 proteins and protein SIC (5 μg/ml) for 24 h, and the release of IL-8 was determined by ELISA. The data represent the average protein levels ± SDs from at least three independent experiments, which were analyzed by the Student t test. *, P < 0.05; ***, P < 0.001.

FIG 6.

Schematic overview of M1 protein-induced IL-8 expression in human keratinocytes. The binding of M1 protein to TLR2 on keratinocytes induces activation of intracellular myeloid differentiation primary response gene 88 (MyD88). MyD88 is then responsible for transmitting the inflammatory signal by recruiting interleukin-1 receptor-associated kinase 1 (IRAK1) and IRAK4 together with TNF receptor-associated factor 6 (TRAF6). This complex activates transforming growth factor β-activated kinase (TAK) and TAK-binding protein (TAB), both of which can activate the MAPK signaling cascade and the NF-κB pathway. At the end of this chain of events, the localization and promoter binding of AP-1 and NF-κB trigger the transcription of IL-8 inside the nucleus and its subsequent secretion. IKK, IκB kinase; MAPKKK, mitogen-activated protein kinase kinase kinase; MAPKK, mitogen-activated protein kinase kinase.

DISCUSSION

Streptococcal skin infections are often initiated by adherence of the bacteria to epidermal cells. This can eventually provoke local inflammatory reactions, which can then lead to suppurative lesions. The clinical spectrum of the resulting cutaneous and soft tissue infections is broad and ranges from localized impetigo to deeply invasive necrotizing fasciitis. The extent to which these conditions develop into more systemic complications depends on the depth within the tissue that the bacteria reach (11, 36). Though much effort has been undertaken to unravel the bacterial strategies employed to spread and become more invasive, the molecular mechanisms involved are still far from being completely understood.

Streptococcal M and M-like proteins are important virulence determinants that interact not only with cell types such as neutrophils, monocytes, and endothelial cells but also with keratinocytes (16, 37–39), which were the focus of this study. It was already described in 1995 that M and M-like proteins are responsible for bacterial adherence to keratinocytes by binding to membrane cofactor protein (CD46) (39). Later, in 2003, it was reported that this interaction is also important for bacterial entry into epithelial cells (40). However, whether this leads to an induction of host responses was not addressed in those studies.

In the present study, we examined the effect of soluble M1 protein on human keratinocytes to show that the protein is able to trigger multiple immunologic responses in these cells (summarized in Fig. 6). It should be noted that the M1 serotype is considered one of the most invasive streptococcal serotypes (41), and soluble M1 protein is a potent inducer of inflammation. Once the protein has been released from the bacterial surface, it can cause systemic reactions, such as severe pulmonary damage, before the bacteria enter the circulation (14). These previous findings suggest that soluble M1 protein contributes significantly to pathological complications distal to the primary site of infection. To our knowledge, the present study is the first to show that M1 protein can also act as a PAMP, thereby evoking innate immune responses. Interestingly, we observed that M5 protein can cause a response similar to that caused by M1 in keratinocytes when the effect of other M proteins was tested. Since the M5 serotype is also regarded as an invasive serotype (42), this points to a stronger response to strains of this serotype by the host defense.

The immune responses following M1 exposure involve the activation of the TLR signaling pathway through MAPKs and the transcription factors NF-κB and AP-1. The effects of this activation include the keratinocytic production and release of proinflammatory mediators, such as IL-8, GROα, MIF, IL-1α, IL-1Ra, and ST2. The M1 protein has been described to interact with TLR2 on human peripheral monocytes in order to trigger cytokine release (16), but the cell type-specific cytokine imprint in response to M1 protein seems to be very different.

In particular, IL-8 and GROα signaling has been described to play an important function in recruiting neutrophils to the site of infection or inflammation (43, 44). Interestingly, our data point to a special role for IL-8, because not only is this chemokine the most upregulated gene upon stimulation with M1 protein, but also it utilizes a distinct signal pathway via the TLRs and the MAPKs ERK and p38 that was not shared with the other inflammatory mediators tested. M proteins have been described to bind the membrane cofactor protein CD46, although the receptor has not been described to be involved in cytokine production. It is therefore not likely that CD46 is responsible for the other inflammatory mediators that are released upon stimulation with M1 protein. One could speculate that the production and release of these mediators are a consequence of autocrine signaling from prestored vesicles containing IL-1α or tumor necrosis factor alpha (TNF-α) to their receptors, IL-1R and TNF receptor, respectively (30); however, additional experimental support is needed to confirm this hypothesis.

Previous work by Hoffmann and colleagues has shown that the three MAPKs (ERK, p38, and JNK) are important for generating maximum IL-8 expression (45). In our study, inhibition of ERK and p38 significantly reduced IL-8 expression in M1 protein-stimulated keratinocytes, and similar findings have been made when gastric epithelial cells were treated with Helicobacter pylori (46). However, JNK, a c-Jun-activating MAPK (47), had no effect on the generation of IL-8 in our experiments, and others have also reported similar findings. For example, in 2001 Bennett and colleagues described that JNK is not involved in IL-8 expression in lipopolysaccharide-stimulated human monocytes (48).

Our results further show that the two MAPKs ERK and p38 are activated in a time-space manner, but since a combination of both inhibitors did not further decrease the level of IL-8 expression, they might share the same downstream targets for IL-8 induction. We noted that p38 inhibition had the strongest effect on IL-8 expression. This could be due to the p38-related phosphorylation of both AP-1 and NF-κB, which leads to transcriptional activation (49, 50), whereas ERK phosphorylation of AP-1 has only a stabilizing effect (51) and no direct activity on NF-κB has been described. ERK can initiate NF-κB signaling by activation of RSK- or mitogen and stress-activated protein kinase-related phosphorylation of the NF-κB inhibitor, IκBα (52, 53). In addition, p38 has been found to have a stabilizing effect on mRNA for IL-8 (54), which could provide another explanation for our results. Our findings further indicate that streptococcal M1 protein induces highly regulated and complex signaling networks through different MAPKs, triggering proinflammatory actions, such as IL-8 release, in keratinocytes.

In this study, we demonstrate that keratinocytes recognize M1 protein as a PAMP to alert the innate immune system by releasing alarming inflammatory mediators. To this end, we examined the interaction between keratinocytes and streptococcal surface protein M1 and determined the immunologic response of these cells when encountering the bacterial protein. The data presented thus far suggest a keratinocytic release of inflammatory mediators most likely designated for the recruitment of neutrophils. Although it remains to be clarified how critical IL-8 signaling is for preventing severe streptococcal skin infections, together with how critical the other inflammatory mediators released are, our data suggest that neutrophil attraction through IL-8 signaling could be an important protection that keratinocytes exert in response to streptococcal M1 protein.