Epidermal expression of receptor for advanced glycation end products (RAGE) is related to inflammation and apoptosis in human skin.

Abstract

Receptor for advanced glycation end products (RAGE), which is composed of one extracellular ‘V’-type and two ‘C’-type immunoglobulin domains, is a pattern-recognition receptor that participates in various cellular responses linked to inflammatory reactions, tissue damage and regeneration. The membrane-bound full-length form of RAGE (mRAGE) is a signal-transducing receptor that leads to activation of the transcription factor nuclear factor-kappa B (NF-κB) in a pro-survival response 1, 2 and a pro-apoptotic response of caspase 8 activation 3. In addition, there are other RAGE proteins which lack the cytoplasmic region and are soluble forms, that is a splice variant named endogenous secretory RAGE (esRAGE) 4 and soluble RAGE (sRAGE), which is enzymatically cleaved from mRAGE. Those two soluble forms of RAGE are considered to suppress the RAGE-mediated intracellular signalling as decoy receptors. We previously reported the distribution of mRAGE and esRAGE in normal human adult tissues using RAGE domain-specific antibodies 5. Although RAGE is expressed in healthy human skin, its characterization and physiological functions in that tissue are still unclear. In this study, to understand the physiological rather than the pathophysiological role of RAGE in the skin, we examined the expression of total RAGE (mRAGE, esRAGE and sRAGE), mRAGE only, cytokines and cell proliferation/apoptosis markers using immunohistochemistry and real-time PCR in normal human skin specimens. We asked whether protein and/or gene expression levels of RAGE are associated with cell proliferation/differentiation, inflammation and/or apoptosis in human epidermis. All skin samples were obtained at the Kitayama Clinic from surgical resections of 67 patients who had undergone a skin excision surgery with signed informed consent. Immunohistochemistry and real-time PCR were performed. For additional details on materials and methods, see Data S1. We first investigated the expression of total RAGE (mRAGE+esRAGE+sRAGE), which has a common extracellular V-domain detected by anti-RAGE(V), and mRAGE only, which possesses a cytoplasmic tail detected by anti-RAGE(C), in human skin. We observed positive staining signals of total RAGE in epidermal keratinocytes of the spinous layer, dermal fibroblasts, microvascular endothelial cells, sebocytes and basal cells in sebaceous glands and clear cells in sweat glands (eccrine glands) (Fig. 1). Staining with the anti-RAGE(C) antibody showed positive signals in epidermal keratinocytes, dermal fibroblasts and endothelial cells, but only very faint staining in sweat and sebaceous glands (Fig. 1). Those staining patterns indicate the distinct expression of mRAGE in keratinocytes, fibroblasts and endothelial cells. No positive staining was seen using a non-immune control IgG (data not shown). Next, the expression of total RAGE and mRAGE was quantitatively evaluated in the epidermis [ranging from 0 to 300 according to the modified H-score method]. The data showed a marked individual variation but no statistically significant differences in skin from different areas of the body (Fig. S1). We also found no statistically significant correlations of age, sex and sun-exposed (cheek and neck)/sun-protected (back, abdomen and buttock) regions with total RAGE or mRAGE (Table 1a). However, the expression of total RAGE was positively correlated with epidermal thickness (r = 0.295, P < 0.05); this was also defined using a stepwise linear regression model (β = 0.303, P < 0.04, R2 = 0.073) (Table 1a). We next addressed gene expression patterns in the epidermis. The expression level of AGER, which encodes the membrane-bound full-length isoform of RAGE, was positively, significantly and strongly correlated with TNF-α, IL-1α, S100B (a RAGE ligand) (r = 0.591, P < 0.001; r = 0.730, P < 0.001; r = 0.356, P < 0.001, respectively), Fas and Bax (proapoptotic factors) (r = 0.803, P < 0.001; r = 0.724, P < 0.001, respectively) and weakly with Involucrin (an epidermal differentiation marker) (r = 0.258, P < 0.04) and proliferating cell nuclear antigen (PCNA) (r = 0.248, P < 0.05), but not with age, sex or epidermal thickness (Table 1b). A stepwise linear regression model shows that AGER expression was positively correlated with Bax (β = 0.310, P < 0.01), Fas (β = 0.267, P < 0.001) and S100B (β = 0.538, P < 0.001) (Table 1b). The correlation diagrams are shown in Figure S2. Unexpectedly, mRAGE and total RAGE protein expression levels were not significantly associated with those gene expression levels (data not shown). These results suggest that AGER expression may cooperate and coordinate with apoptosis and inflammation rather than with the keratinization process in normal epidermis. RAGE has been recently shown to have many distinct biological functions such as pro-survival/pro-apoptotic responses 1-3, inflammatory reactions 6, host defenses 7 and efferocytosis. Such RAGE-derived responses could depend on certain conditions and situations between cellular components and ligand species/concentrations. The exposure of keratinocytes to lower or higher concentrations of S100A8/A9 differentially activated the pro-survival or pro-apoptotic cascade via RAGE, respectively 8. However, the situation between RAGE (mRAGE, esRAGE or sRAGE) and its ligands remains unclear in healthy human skin in vivo. In this study, we defined the expression of RAGE, especially mRAGE, in non-pathological and lesion-free normal human epidermis, suggesting its physiological contributions to homoeostasis and the constant turnover of the epidermis. Quantitative analysis showed that mRAGE expression was significantly correlated with epidermal thickness (Table 1). This finding is consistent with experimental data that the deletion of RAGE in keratinocytes diminishes the inflammatory hypertrophy of the epidermis in mice 9. We found that AGER expression is positively and significantly correlated with the expression of genes encoding S100B, Fas, Bax, IL-1α and TNF-α (Table 1). Our findings demonstrate that mRAGE expression is dominant in keratinocytes of healthy human epidermis and suggest that mRAGE could monitor and respond to acute reactions and cellular responses, including inflammation and apoptosis, rather than to long-term alterations during ageing to maintain homoeostasis of the skin. MI, YY, HY, TH and KH conceived and designed the experiments. MI and KH performed the experiments. MI, YK and KH analysed the data and provided reagents/materials/analysis tools. MI, YY and TF wrote the manuscript. The authors declare no conflict of interest. Figure S2. Total RNAs were extracted from the epidermis and were then subjected to real-time quantitative RT-PCR analyses of AGER, IL-1α, TNF-α, S100B, Fas and Bax. These transcript expression levels were normalized to RPLP0 ribosomal protein expression. Correlations between AGER and the other genes were calculated. Analyses were carried out using Pearson's product-moment correlation coefficient and a forward-backward stepwise multivariate linear regression model analysis. Data S1. Materials and Methods. Data S2. Supplementary Reference. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.