Human Melanocytes Play An Active Role In The Skin Immune System By Expressing Functional Toll-like Receptors

The human epidermis provides a first defense barrier to a potentially hostile environment. It has been shown that within the epidermis, not only professional immune cells (e.g., Langerhans cells) but also non-immune cells (e.g., keratinocytes and melanocytes) are involved in the immune protection of the host. Although originally identified as a professional producer of melanin, recent studies have revealed that melanocytes exhibit a variety of functions. For example, human melanocytes have the capacity to express HLA-DR, CD40 and adhesion molecules such as ICAM-1 and VCAM-1. In addition, they can produce various soluble mediators of inflammation such as IL-1, IL-6 and IL-8. These studies suggest that melanocytes are not simply pigment-producing cells but also immunocompetent cells. However, it is unclear whether melanocytes can recognize pathogens and play an active role in the skin's local immune defense system.Recognition of pathogens by innate immune cells is mediated by pattern-recognition receptors that recognize conserved pathogen-associated molecular patterns (PAMPs). One major group of pattern-recognition receptors is the Toll-like receptors (TLRs), which transduce signals leading to the activation of NF-κB, which subsequently drive the induction of several pro-inflammatory cytokines and chemokines. Apart from sensing exogenous ligands from microbial components, which is critical for pathogen elimination, TLRs are able to recognize endogenous ligands, such as heat shock proteins (HSPs) and extracellular matrix components, which may be called danger-associated molecular patterns (DAMPs), in analogy to PAMPs.Thus far, more than 10 different TLRs with distinct ligand specificity have been identified. In human skin, the distribution of TLRs is incompletely defined. Several studies demonstrated that keratinocytes display TLRs and respond to corresponding PAMPs by producing pro-inflammatory cytokines. In contrast, the TLR expression pattern of primary human melanocytes has never been systematically examined.Vitiligo is characterized by a loss of melanin-producing melanocytes from the epidermis of patients, which results in skin depigmentation. Both genetic and environmental factors have been implicated to explain the loss of epidermal melanocytes in this disorder. Genetic risk factors are important, but are not sufficient for the disease to occur, and environmental triggers may also contribute to the initiation and/or progression of vitiligo. Human skin is known to be constantly exposed to various pathogens of prokaryotic, eukaryotic, and viral origin. Indeed, a variety of viruses including cytomegalovirus, hepatitis virus, and AIDS virus, have been implicated in the pathogenesis of the vitiligo. In addition, a strong causative link between herpes virus and vitiligo has been noted in the Smyth chicken model. Finally, viral infections have been long associated with a variety or autoimmune disorders. Interestingly, some of them including thyroid disease and typeⅠdiabetes mellitus, occur with increased frequency in patients with vitiligo. These observations suggest that virus might be an important environmental factor that participates in the initiation or acceleration of vitiligo in genetically predisposed individuals.Toll-like receptors (TLRs) play a crucial role in pathogen detection and in the mounting of antimicrobial immune responses. Although initial studies have focused on the TLR expression and function in the immune system, more recent studies have shown that TLR expression is shared by many different cell types including human epidermal melanocytes. We have previously shown that TLRs expressed in melanocytes are functional and respond to their respective ligands by activating the central transcription factor NF-κB and by up-regulating pro-inflammatory mediators. In respect of antiviral responses, TLR3 was shown to be a receptor for double-stranded RNA (dsRNA), which represents either genomic or life cycle intermediate material of many viruses. In the process of viral infection, dsRNA stimulates cellular antiviral activities, and occasionally, cell death, which is another way of protecting the host against virus spreading. In our preliminary study, we found in vitro treatment with poly(I:C), a synthetic analogue dsRNA, resulted in massive death and detachment of cultured human melanocytes. Could human melanocytes utilize TLR3 as an initiator of signaling pathway leading to destruction of themselves in response to viral infections? This hypothesis, as well as its implications in the pathogenesis of vitiligo, needs more evidential supports from both clinical observations and experimental investigations. In the present study, we provided laboratory evidences that melanocytes expressed TLR3. Poly(I:C) treatment induced apoptosis in human melanocytes in a TLR3-dependent manner. This phenomenon required the involvement of IFN-βautocrine signaling but not NF-κB activation. Moreover, dsRNA stimulation led to the activation of downstream signaling pathways, including p38, ERK1/2 and JNK1/2, among which p38 and ERK1/2 controlled both IFN-βsecretion and IFN-βmediated cell death. The fact that human melanocytes dramatically respond to TLR3 ligand indicates that viral infection may play a part in the initiation/amplification of vitiligo by directly triggering melanocyte apoptosis.Primary human melanocyte cultures were obtained from neonatal foreskin. All samples obtained were from surgical procedures with the patients'informed consent following a protocol approved by the Huashan Hospital Fudan University Institutional Review Board. The epidermis was separated from the dermis after an overnight incubation of skin samples in a 0.25% Dispase solution in PBS at 4℃. In order to separate cellular elements, epidermal sheets were incubated at 37℃in a solution of 0.25% trypsin and 0.02% ethylenediaminetetraacetic acid (EDTA) in PBS, for 10 minutes. Cellular suspension was then filtered through 70μm cell strainer and then centrifuged at 1500 rpm for 7 minutes to harvest cells. Then, melanocytes were selectively grown in a defined medium M254 with PMA-free human melanocyte growth supplements. The cells were maintained in a humidified incubator with 5% CO2 at 37℃and were fed every 3-4 d, and further passaged at 1:2 when they became 80% confluent, with experiments being carried out from 3 to 4 passages.To investigate the expression profile of TLRs in human melanocytes, in vitro-expanded melanocytes obtained from healthy subjects were assessed by reverse transcription (RT)-PCR for TLR mRNA using a panel of specific primers for TLR1-TLR10. We found the constitutive mRNA expression of TLRs 1,2,3,4,6,7 and 9 in human melanocytes but not TLRs 5,8, and 10. As a positive control, human PBMC were shown to express all mRNA of TLRs 1-10. The purity of the cultured human melanocytes was verified by RT-PCR analysis for the keratinocyte markers, such as keratin 10 and keratin 14, and the fibroblast marker ASO2, showing that contamination from keratinocytes or fibroblasts was negligible. To measure the TLR protein expression in human melanocytes, flow cytometry and immunofluorescence assays were performed with specific antibodies to human TLR proteins. Ample amounts of TLRs 2,4,7 and 9 were detected in human melanocytes indicated by flow cytometry analysis. TLR3 is generally recognized as an intracellular receptor, but a recent study showed a surface TLR3 on human skin and lung fibroblasts. Thus, we analyzed the specific location of TLR3 expression in human melanocytes. Using the same mAb specific to TLR3, we demonstrated the detection of intracellular TLR3 in human melanocytes but not on the cell surface. The protein expression of TLRs on in vitro-cultured human melanocytes was also demonstrated by immunofluorescence assays, and consistent findings were noted. Compared to the isotype control, a clear positive staining for TLRs 2,3,4,7 and 9 was obtained. These results paralleled our RT-PCR endings. Thus human melanocytes not only expressed TLRs mRNA but also expressed TLRs protein. 3. Regulation of TLR3 in human melanocytesTo determine whether the TLR3 expression was inducible by its activation with poly(I:C), a synthetic surrogate of viral dsRNA, the kinetic of TLR3 mRNA was analyzed by real-time RT-PCR. In a dose-dependent fashion, TLR3 mRNA increased at 6 hours after treatment with poly(I:C) (2.7-fold at 1μg/ml,4.2-fold at 10μg/ml, and 3.6-fold at 100μg/ml) and at 24 hours after stimulation (3.8-fold at 1μg/ml, 13-fold at 10μg/ml and 37-fold at 100μg/ml). These data show that human melanocytes constitutively and inducibly express TLR3, suggesting that melanocytes have the potential to recognize virus-derived pathogenic motifs via TLR3.To characterize the functional relevance of TLRs in human melanocytes, the production of inflammatory cytokines by melanocytes in response to TLR ligands were determined. Human melanocytes were stimulated with various TLR ligands, for example, The TLR2 ligand, peptidoglycan (PGN), the TLR3 ligand, poly (I:C), the TLR4 ligand, LPS, the TLR7 ligand, imiquimod, and CpG 2006, which ligates TLR9. Subsequently, the secretion level of IL-8 and IL-6 were analyzed by ELISA. The maximal concentration of poly (I:C) used in our analysis was restricted to 10μg/ml, as larger dosage led to massive death and detachment of melanocytes. All of the TLR ligands examined potently induced the production of IL-8 in a dose-dependent manner, with maximal secretion being observed at 1μg/ml poly (I:C). Melanocytes also released substantial amounts of IL-6 when stimulated with PGN, poly (I:C), LPS and CpG 2006. On the contrary, no IL-6 production was observed in melanocytes treated with imiquimod, even though the cells expressed TLR7 mRNA and protein.Activated epidermal melanocytes may modulate the recruitment and activation of different immune cells at least in part through the expression of the chemokines. We next analyzed accumulation of mRNAs of inflammatory chemokines such as CCL2/MCP-1, CCL3/MIP-1 and CCL5/RANTES upon different TLR activation using real-time RT-PCR. Here we showed that each TLR ligand induced the release of CCL2, CCL3 or CCL5 with a different kinetics. All the TLR ligands induced significant amounts of CCL3 mRNA in 6-hour stimulated melanocytes, whereas mRNA for CCL2 and CCL5 were significantly induced by all the TLR ligands except imiquimod, the TLR7 ligand. Although some basal expression of CCL2, CCL3 and CCL5 could be observed in nonstimulated cells, there was at least 1-fold increase in their mRNA levels after treatment with TLR ligands.Collectively, these data demonstrated that TLR ligand pretreatment of epidermal melanocytes created a pro-inflammatory milieu by showing elevated cytokines and chemokines.Activation of NF-κB plays a central role in TLR-mediated cellular activation and gene expression in a variety of cell types. Therefore, we tested whether triggering TLRs in human melanocytes can give rise to the activation of NF-κB signaling pathway. TLRs within human melanocytes were stimulated for 6 hour by various ligands and assessed by Western blot analysis to examine phosphorylation of IκBα, which is one of the markers of NF-κB activity. No phosphorylated-IκBαwas observed in unstimulated melanocytes. On the contrary, the clear bands of phosphorylated-IκBαcould be induced by PGN (100μg/ml), poly (I:C) (10μg/ml), LPS (100μg/ml), imiquimod (100μg/ml) or CpG 2006 (5μM) to different degrees, indicating that NF-κB pathway participated in TLR2,3,4,7 and 9 signaling in response to the respective PAMPs.To further substantiate the activation of NF-κB pathway in TLR ligand-treated melanocytes, translocation to the nucleus of the NF-KBp65 was analyzed. Human melanocytes were cultured in slide chambers in the presence of TLR ligands and stained for the NF-KBp65 as described in Materials and methods. The slides were then analyzed for nuclear localization by fluorescence microscopy. In unstimulated melanocytes, the stainings for the NF-κBp65 were mainly visible in the cytoplasm. In contrast, when melanocytes were stimulated with PGN (100μg/ml), poly (I:C) (10μg/ml), LPS (100μg/ml), imiquimod (100μg/ml) or CpG 2006 (5μM), a clear nuclear translocation of the p65 subunit was observed in all the stimulated groups. These data indicated that TLR 2,3,4,7 or 9 ligand engagement led to the activation of NF-κB, which explained the numerous pro-inflammatory genes being activated.We further investigated the responsiveness of melanocytes to dsRNA. Melanocytes exposed to 100 mg/ml poly(I:C) underwent striking changes in cell appearance:cytoplasmic spreading and flattening of the cells and their nuclei were illustrated by phase contrast and phalloidin staining at 8 hours and 24 hours. A more prolonged exposure to poly(I:C) for 48 hours led to a decrease in the melanocyte viability in a dose-dependent way, measured by trypan blue exclusion analysis. The detrimental effect of poly(I:C) on the viability of melanocytes was further quantified by Annexin V/PI staining.22%,47% apoptosis (Annexin V+) were achieved following the treatment with 50,100μg/ml poly(I:C) compared with the 2.5% apoptosis for the control group, indicating that cytotoxicity of poly(I:C) to the melanocytes is largely due to the induction of apoptosis.Several intracellular receptors have been described to bind to dsRNA, such as TLR3, PKR, and MDA-5. To determine whether the apoptosis-inducing effect of poly(I:C) was mediated by TLR3, the receptor expression was efficiently suppressed through infection of a lentivirus designed to stably express a TLR3-specific shRNA. The effect of poly(I:C) on melanocyte apoptosis was then analyzed in control and TLR3 knockdown cells by Annexin V staining. Similar to nontransduced cells, the apoptosis-inducing effect of poly(I:C) was apparent in non-sense shRNA cells, used as control. In contrast, shTLR3 melanocytes were greatly resistant to poly(I:C) toxicity, clearly indicating that the apoptosis-inducing effect of dsRNA depends upon TLR3.Previously, we have demonstrated the NF-κB activation status through IκBα□phosphorylation in melanocytes upon poly(I:C) stimulation. To determine whether the activation of NF-κB pathway was responsible for melanocyte apoptosis induced by poly(I:C), Bay11-7082, a specific pharmacological inhibitor of IκBαphosphorylation was utilized. The poly(I:C)-induced phosphorylation of IκBαwas significantly inhibited by Bay11-7082. However, culture of melanocytes in the presence of Bay11-7082 failed to prevent poly(I:C)-induced apoptosis, even when dose as high as 0.5μM was used. Of note, we found that stimulation of melanocytes with LPS, a strong NF-κB activator, did not alter their apoptosis levels significantly, suggesting that NF-κB activation is not a key mediator of dsRNA-triggered melanocyte death.Because the production of type I IFNs (IFN-a and IFN-(3) induced by TLR3 activation has been reported to trigger apoptosis in several cell types, the role of typeⅠIFNs in dsRNA-induced melanocyte apoptosis was evaluated. Both IFN-αand IFN-Pproduction were strongly induced in melanocytes upon poly(I:C) treatment, and STAT1 phosphorylation was observed, indicative of typeⅠIFN signaling. Neutralization of IFN-βwith specific antibodies strongly inhibited the poly(I:C)-induced apoptosis of melanocytes, while neutralization of IFN-βdid not modify the effect. Phase-contrast observation also confirmed that the rescue of melanocytes from apoptosis was dependent upon the dose of antibodies adopted in the neutralization experiments, demonstrating that IFN-βwas necessary for TLR3-mediated cell death. However, treatment of melanocytes with IFN-αor IFN-β, alone or in a mixture of both for less than 72 hours did not significantly affect melanocyte viability. These results establish that IFN-βsignaling is required for TLR3-triggered cytotoxicity, although it is insufficient to induce cell death by itself.To identify the mechanism of IFN-(3 production upon poly(I:C) stimulation, we first determined the signaling pathways activated by poly(I:C) in melanocytes. Significant phosphorylation of p38 MAPK was detected at 30 minutes after poly(I:C) stimulation, whereas increase of ERK1/2 phosphorylation was observed at 2 hours in melanocytes. Finally, a significant increase of phosphorylation of JNK1/2 was detected at 1 hour after stimulation. To further address the role of activation of these signaling molecules in dsRNA-induced IFN-βproduction and melanocyte apoptosis, we utilized pharmacological inhibitors to p38 MAPK, ERK1/2 and JNK1/2. As shown by Annexin V staining, apoptotic cells in the presence of poly(I:C) strongly decreased when the p38 MAPK or ERK1/2 signaling was blocked (mean of death reduction:73.6% and 50% for p38 inhibition and ERK1/2 inhibition). Accordingly, morphological analysis suggested that the blockade of p38 MAPK or ERK1/2, but not JNK1/2, partially rescued the melanocytes from poly(I:C)-induced death. Furthermore, blockade of p38 MAPK and ERK1/2, inhibited poly(I:C)-stimulated IFN-βsecretion significantly, whereas JNK1/2 blockade had limited effects. Taken together, these results suggest that p38 MAPK and ERK1/2 are involved in both the autocrine secretion of IFN-βand the induction of apoptosis.In summary, this study provides evidence for a TLR expression and response profile of normal human melanocytes, which stress the importance of the melanocytes not only as pigment cells but also as sentinels of skin homeostasis. The present work represents the first detailed mechanistic study on innate immunity signaling pathways activated by dsRNA in melanocytes, and clarifies part of complex molecular responses of melanocyte faced with viral components. On the basis of our results, a possible pathogenetic hypothesis could be that viral dsRNA stimulates TLR3 in human melanocytes and triggers the cellular apoptosis through IFN-β. These findings may add new insight into the possible etiopathogenetic mechanisms leading to melanocyte loss in vitiligo and propose IFN-βas a potential therapeutic target.