DOI: 10.1111/exd.12735
Original Article
www.wileyonlinelibrary.com/journal/EXD
Inhibition of hedgehog signalling attenuates UVB-induced skin photoageing Wanyeon Kim1,2,*, EunGi Kim3,*, Hee Jung Yang1, TaeWoo Kwon3, SeoYoung Han3, Sungmin Lee3, HyeSook Youn1,2, Youngmi Jung1,3, ChulHee Kang4 and BuHyun Youn1,2,3 1 Department of Biological Sciences, Pusan National University, Busan, South Korea; 2Nuclear Science Research Institute, Pusan National University, Busan, South Korea; 3Department of Integrated Biological Science, Pusan National University, Busan, South Korea; 4Department of Chemistry, Washington State University, Pullman, Washington, USA Correspondence: BuHyun Youn, Department of Biological Sciences, Pusan National University, Busan, 609-735, South Korea, Tel.: 82-51-510-2264, Fax: 82-51-581-2962, e-mail: [email protected] *These authors contributed equally to this work.
Abstract: The hedgehog (Hh) signalling pathway regulates normal development and cell proliferation in metazoan organisms, but its aberrant activation can promote tumorigenesis and progression of a variety of aggressive human cancers including skin cancer. Despite its importance, little is known about its role in photoageing, a type of UV-induced skin lesions. In this study, we investigated the involvement of Hh signalling in the photoageing process as well as the use of an Hh-regulating alkaloid compound as a novel therapeutic drug to regulate photoageing in keratinocytes. We found that UVB induced Hh signalling by the expression of Hh ligands and Hh-mediated transcription factors, respectively. Moreover, UVB-induced Hh activation relied on mitogen-activated protein kinase (p38, ERK and JNK) activity and
inflammatory responses (upregulation of COX-2, IL-1b, IL-6 and TNF-a), resulting in premature senescence and photoageing in vitro and in vivo. Notably, a selected Hh inhibitor, evodiamine, mediated photoageing blockade in a mouse skin model. Taken together, our findings demonstrated that Hh signalling is associated with UVB-induced photoageing, while pharmacological inhibition of Hh signalling significantly reduced experimental photoageing, indicating its potential for use as a therapeutic target for this disease.
Introduction
collagens. Moreover, activation of transcription factor NF-кB induces the expression of pro-inflammatory cytokines involved in immunoregulation and cell survival such as IL-1, IL-6, vascular endothelial growth factor, TNF-a and MMPs (6). Taken together, these findings suggest that these inflammatory responses lead to UVB-induced photoageing in skin. Hedgehog (Hh) ligands consist of Sonic Hh (SHH), Desert Hh (DHH) and Indian Hh (IHH), which are secreted proteins (7). Hh ligands are modulated by post-translational modifications, autoproteolytic cleavage and lipidation. Secreted Hh ligands bind to Hip1, Patched (Ptch)-1 and Ptch-2 of the cell surface. Smoothened (Smo) is inhibited by Ptch under normal conditions. When Hh ligands bind to Ptch, Smo is activated, leading to transcriptional activation of glioma (GLI). GLI1 lacks a transcriptional repressor domain, while cleaved GLI2 and GLI3 have repressor functions. GLI2 is mainly accountable for the activator function in response to Hh signalling, while GLI3 is known to be responsible for the activity of repressor. In the absence of Hh signal, GLI2 is phosphorylated by protein kinase A, casein kinase I and glycogen synthase kinase 3b, promoting interaction of the E3 ligase Slimb and b-transducin repeat-containing protein. The ubiquitinated GLI2 is partially processed by the proteasome, after which cleaved GLI2 translocates into the nucleus and represses target gene expression. Activated Smo facilitates the transcriptional activation of GLI2 through the inhibition of phosphorylation. The activated forms of GLI2 stimulate the expression of target genes such as GLI1, CCND1 and BCL2 in the nucleus (8,9). Deregulation of the
The skin functions as a defensive barrier against environmental factors, primarily ultraviolet (UV) irradiation. Chronic exposure to sunlight over the maximal cumulative solar UV dose leads to wrinkling, dryness, pigmentation and premature ageing of the skin, which is known as photoageing (1,2). UV radiation consists of UVA (320–400 nm), UVB (280–320 nm) and UVC (200– 280 nm). As UVC is filtered out by the ozone layer, UVA and UVB are considered to be the primary factors responsible for induction of photoageing (3). UVB is far more mutagenic and energetic than UVA and leads to more deleterious outcomes (4). After penetrating the epidermis of the skin, UVB can cause a variety of adverse effects, such as epidermal thickness, dermal elastosis, degradation of collagen by increased matrix metalloproteinases (MMPs), inflammatory infiltration and vessel ectasia in skin. Consequently, it stimulates erythema, inflammation, hyperplasia, formation of wrinkles and photoageing, a type of UVB-induced skin lesions, which may lead to cancer development. UVB-induced photoageing is initiated by production of reactive oxygen species (ROS), which activates several receptors for epidermal growth factor, interleukin (IL)-1, keratinocyte growth factor and tumor necrosis factor (TNF)-a. Activation of these receptors induces downstream signalling pathways of mitogen-activated protein kinases (MAPKs) such as p38, JNK and ERK, which leads to an increase of the nuclear transcription complex activator protein 1 (AP-1), c-Jun and c-Fos (4,5). Subsequently, MMPs as target genes of AP-1 are upregulated and induce degradation of
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
Key words: evodiamine – hedgehog – inflammation – photoageing – UVB
Accepted for publication 24 April 2015
611
Kim et al.
Hh signalling pathway initiates tumor formation and mediates survival of the brain, muscle and skin cancers (10). Several studies have provided evidence that Hh signalling is responsible for aberrant proliferation of skin cells and induction of genomic instability, leading to the development of basal cell carcinoma (BCC), a type of skin cancer (11,12). Aberrant Hh signalling leads to an increase in GLI expression and activation, which consequently induces tumor progression in the skin. Although many studies have been conducted to investigate the association between Hh signalling and tumorigenesis of skin cells, little is known about the role of Hh signalling in photoageing.
(a)
(b)
Evodiamine is a quinolone alkaloid isolated from the fruit of Evodia rutaecarpa that has the ability to protect against oxidative damage and modulate inflammatory response through the inhibition of several proteins, including inducible nitric oxide synthase, cyclo-oxygenase 2 (COX-2) and hypoxia-inducible factor 1a (13,14). Moreover, evodiamine has been shown to exert anticancer activity against many cancer types, including induction of apoptosis and inhibition of cell proliferation, invasion and metastasis (15,16). Nevertheless, the function of evodiamine associated with antiphotoageing activity and regulation of Hh signalling in skin is not known. This study was conducted to clarify the mechanisms and key molecules that confer UVB-induced skin photoageing. We demonstrated that UVB-induced activation of Hh signalling and subsequent activation of the MAPK pathway and inflammatory response play important roles in induction of premature senescence in HaCaT cells and wrinkle formation in a mouse model. Our findings provide a possible explanation of how Hh signalling is involved in UVB-induced skin damage. Furthermore, we suggest that Hh signalling has great potential for use as a pharmaceutical target to prevent photoageing, and evodiamine is a promising candidate for the inhibition of Hh signalling.
Materials and methods
(c)
The materials and methods used in this study, including cell culture; drug treatment; transient transfection; UVB irradiation; Western blot analysis; cell viability assay; luciferase reporter assay; real-time quantitative RT-PCR (qRT-PCR); assays for the production of prostaglandin E2 (PGE2), IL-1b and IL-6; MMP activity assay; senescence-associated b-galactosidase staining; animal protocol; evaluation of wrinkle formation; haematoxylin and eosin (H&E) staining; immunohistochemistry (IHC); and statistical analysis, are described in detail in the Supporting Information.
Results Hedgehog signalling is upregulated by UVB in HaCaT cells Hh signalling is important for tumor formation and growth in skin (8). Based on this information, we analysed the expression of SHH (a most representative Hh ligand) in normal skin and squamous cell
(d)
(e)
(f)
(g)
612
Figure 1. Hh signalling is upregulated by UVB in HaCaT cells. (a) To estimate the expression levels of SHH in skin cancer, data derived from the Oncomine database (www.oncomine.org) using biopsies of normal skin (n = 6) and SCCs (n = 5) were analysed (17). (b) To measure Hh signalling in response to UVB irradiation, the protein levels of Smo and cleaved form of SHH and IHH were measured. At 3 h after UVB irradiation, the cells were harvested and the cell lysates were subjected to Western blot analysis. (c) To select candidates to regulate Hh signalling, several natural products (evodiamine (Evo), silibinin, rhamnetin, cirsiliol, nordihydroguaiaretic acid (NDGA), caffeic acid phenethyl ester (CAPE) and curcumin) were screened. After pretreatment with these compounds and UVB irradiation, mRNA levels of GLI1 and GLI2 were analysed by qRT-PCR. *P > 0.05 compared with non-irradiated cells as well as P < 0.05 compared with UVBirradiated cells, **P < 0.05 compared with non-irradiated cells as well as P < 0.05 compared with UVB-irradiated cells. (d) To determine the non-cytotoxic dose of evodiamine, the viability of cells was assessed by a triazolyl blue tetrazolium bromide assay. Cells were treated for 2 h with evodiamine at various concentrations up to 25 lM. (e) To estimate Hh signalling in response to UVB irradiation and treatment with evodiamine or GDC-0449 (a Smo inhibitor), the mRNA levels of GLI1 and GLI2 were assessed by qRT-PCR. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells. (f) To measure Hh signalling in response to pretreatment with evodiamine or GDC-0449 and UVB irradiation, protein levels of GLI1 and GLI2 (full-length as an activator form) from the cytoplasm to the nucleus were measured by Western blot analysis after cytoplasmic or nuclear fractionation. Tubulin and lamin B were used as markers for the cytoplasmic extract (CE) and nuclear extract (NE), respectively. (g) To determine the activation of Hh signalling in response to pretreatment with evodiamine or GDC-0449 and UVB irradiation, GLI-specific promoter activity was measured by a luciferase assay. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
Regulation of hedgehog-mediated photoageing
carcinomas (SCCs) using the Oncomine database (www.oncomine.org). We found that SHH expression in SCCs was higher than that found in normal skin (Fig. 1a) (17). In addition, several literatures demonstrated that Hh signalling is associated with UVBinduced skin cancers including SCCs and BCCs (11,18,19). Because UVB-induced photoageing may lead to the development of skin cancers (20), we hypothesized that activation of Hh signalling might be responsible for not only skin cancers, but also photoageing induced by UVB irradiation. To monitor alterations of the Hh pathway in response to UVB exposure, we measured the expression of Hh ligands and Smo in HaCaT cells after UVB irradiation. As shown in Fig. 1b, the protein levels of Smo, SHH and IHH increased in a dose-dependent manner. We next assessed the mRNA levels of SHH and IHH in skin tissues of aged Sprague Dawley (SD) rats to determine whether Hh signalling is associated with the general ageing process. The mRNA levels of both Hh ligands in skin tissues of SD rats gradually increased from the tissues of 6-m-old to 18-m-old rats and then markedly decreased in 24-m-old rats (Fig. S1), suggesting that Hh signalling is not directly correlated with normal ageing. In addition, we found that mRNA levels of both GLI1 and GLI2 were induced by UVB irradiation (Fig. 1c). Because several natural products have been reported to regulate the signalling pathway without adverse effects in various cancer cell lines, a large number of natural products were prepared for screening (21,22). The primary screening based on the biological/pharmacological activity of natural products was conducted by measuring transcriptional activity of GLI1/2 under UVB irradiation. Subsequently, we analysed the physical and chemical properties of compounds according to solubility and molecular weight along with the Lipinski-like criteria (23). After selection of fewer than 10 candidates, the compounds were retested at various concentrations to determine the potency and cytotoxicity of each compound. With an optimal concentration of each candidate, we assessed alteration of GLI1/GLI2 mRNA expression in response to treatment of each compound and UVB irradiation (Fig. 1c). Ultimately, one quinolone alkaloid, evodiamine, was selected for further investigation. To determine the concentration of evodiamine that could be used without cytotoxicity, HaCaT cells were treated with different concentrations of evodiamine (Fig. 1d). Evodiamine did not affect cell viability at concentrations up to 5 lM. To determine whether evodiamine can regulate the Hh signalling pathway, we measured the expression of GLI1 and GLI2 as the last effectors of Hh signalling. We found that evodiamine is able to suppress UVB-induced expression of GLI1 and GLI2 (Fig. 1e,f). GDC-0449, a Smo inhibitor, was treated to confirm the inhibitory effects of evodiamine on UVB-induced Hh signalling. In addition, a GLI-specific luciferase reporter assay was conducted to measure the transcriptional activity of GLIs in response to UVB irradiation (Fig. 1g). We observed that UVB increased the transcriptional activity of GLIs and this effect was downregulated by pretreatment of evodiamine or GDC-0449. Taken together, these findings suggest that UVB exposure can promote Hh signalling, which is inhibited by evodiamine treatment.
(a)
(b)
(c)
(d)
(e)
UVB-induced inflammation is downregulated by the inhibition of hedgehog signalling in HaCaT cells
Figure 2. UVB-induced inflammation is downregulated by the inhibition of Hh signalling in HaCaT cells. (a) To estimate the activation of MAPKs (p38, ERK and JNK) in response to pretreatment with evodiamine or GDC-0449 and UVB irradiation, phosphorylation of MAPKs was detected. At 3 h after UVB irradiation, the cells were harvested and the cell lysates were subjected to Western blot analysis. (b) To determine inflammation-related genes in response to pretreatment with evodiamine, SB203580 (a p38 inhibitor), PD98059 (an ERK inhibitor) or SP600125 (a JNK inhibitor) and UVB irradiation, the mRNA levels of COX-2, IL-1b, IL-6 and TNF-a were assessed by qRT-PCR. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells. (c) To measure Hh signalling in response to pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, protein levels of COX-2 and TNF-a were verified by Western blot analysis. (d) To measure COX-2 activity in response to pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, PGE2 in the media was detected by enzyme immunoassay. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells. (e) To measure cytokines secreted in response to pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, IL-1b and IL-6 in the media were analysed by enzyme immunoassay. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells.
UVB-induced photoageing is known to be associated with inflammation through the MAPK signalling pathway (4). In addition, several studies have shown that Hh signalling is accelerated via activation of MAPKs and overexpression of COX-2 (24,25). The
transcription of pro-inflammatory genes, including those encoding IL-1b, IL-6, IL-12 and TNF-a, can be also increased by GLIs under inflammatory responses such as bacterial infection,
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
613
Kim et al.
cytokine-mediated paracrine signalling or nitric oxide signalling (26–28). In this study, we determined whether the activation of MAPKs and expression of COX-2 and other pro-inflammatory genes associated with Hh signalling were modulated by UVB irradiation and evodiamine. As shown in Fig. 2a, phosphorylation of p38, ERK and JNK increased following UVB irradiation, but this response was decreased by the inhibition of Hh signalling with pretreatment of evodiamine or GDC-0449. In particular, we observed that treatment of GDC-0449 resulted in significant decrease in ERK phosphorylation and partial decrease in p38 and JNK phosphorylation. To further elucidate the involvement of inflammation in UVB-induced photoageing, we measured the expression of COX-2, IL-1b, IL-6 and TNF-a in HaCaT cells. We found that the mRNA levels of these inflammation markers were increased by UVB irradiation (Fig. 2b). Increased expression of these genes was reduced by pretreatment with evodiamine, GDC0449 or MAPK inhibitors. Based on the results shown in Fig. 2a and b, an ERK inhibitor, PD98059, was used to confirm the antiphotoageing effects of evodiamine and GDC-0449 on regulation of UVB-induced Hh signalling. Next, the expression levels of COX-2 and TNF-a were elevated following UVB irradiation in HaCaT cells, but protein expression was reduced to the basal levels by evodiamine, GDC-0449 or PD98059 (Fig. 2c). The inhibitory effects of evodiamine on the UVB-induced expression of COX-2 and TNF-a were reversed by overexpression of GLI1 or GLI2 (Fig. S2a,b). PGE2 is one of the most abundant metabolites of arachidonic acid produced by COX-2. To confirm that COX-2 activity was elevated in response to UVB irradiation, we analysed the production of PGE2 in HaCaT cells. PGE2 generation was increased by UVB irradiation and reduced by evodiamine, GDC0449 or PD98059 (Fig. 2d). As IL-1b and IL-6 act as pro-inflammatory cytokines after being released from the cells, we measured IL-1b and IL-6 secreted into the cell culture media. As shown in Fig. 2e, the secretion of IL-1b and IL-6 was significantly increased by UVB irradiation. In contrast, pretreatment with evodiamine, GDC-0449 or PD98059 suppressed the release of both cytokines. The inhibitory effects of evodiamine on the UVB-induced secretion of IL-1b and IL-6 were rescued by overexpression of GLI1 or GLI2 (Fig. S2c). These results demonstrated that UVB induces MAPK activation and inflammation, which is negatively regulated by the inhibition of Hh signalling by treatment with evodiamine.
UVB-induced cellular senescence is reversed by the inhibition of hedgehog signalling in HaCaT cells UVB-induced photoageing is known to be mediated by increased MMP expression and activation along with collagen degradation (15,29). To determine the effects of UVB on MMP expression, we measured changes in the levels of photoageing-related MMPs. The expression of three MMPs, MMP1, MMP2 and MMP9, increased following UVB irradiation, but this response was downregulated by the inhibition of Hh signalling with pretreatment of evodiamine, GDC-0449 or PD98059 (Fig. 3a). In contrast, the expression of tissue inhibitor of metalloproteinase 1 (TIMP1), a cellular inhibitor of MMPs, was reduced in response to UVB irradiation and rescued by the inhibition of Hh signalling. To determine the activity of MMPs in response to UVB irradiation, we performed zymography and activity assays of MMPs (Fig. 3b,c). Active MMPs increased following UVB irradiation and were reduced by evodiamine, GDC-0449 or PD98059. Next, we analysed cellular
614
(a)
(c)
(b)
(d)
Figure 3. UVB-induced cellular senescence is rescued by the inhibition of Hh signalling in HaCaT cells. (a) To estimate the effects of evodiamine, GDC-0449 or PD98059 on UVB-induced photoageing, protein levels of photoageing markers including MMP1, MMP2, MMP9 and TIMP1 were observed by Western blot analysis. (b) To measure the effects of evodiamine, GDC-0449 or PD98059 on UVB-induced photoageing, the activities of MMP1, MMP2 and MMP9 were assessed by b-casein or gelatin zymography. Conditioned media of HT1080 cells were used as positive controls for active MMP2 and MMP9. (c) To measure the effects of evodiamine, GDC-0449 or PD98059 on UVB-induced photoageing, the activities of MMP1, MMP2 and MMP9 were analysed by enzyme immunoassay. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVBirradiated cells. (d) The inhibitory effects of evodiamine, GDC-0449 or PD98059 on UVB-induced senescence were measured by senescence-associated b-galactosidase staining. After pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, cells were fixed and stained by senescence-associated b-galactosidase staining solution (top). The graphs showed the percentage of the cells with staining-positive cells (bottom). *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells. Scale bar = 50 lm.
senescence by senescence-associated b-galactosidase staining. Although b-galactosidase is activated under acidic conditions, senescence-associated b-galactosidase activity is detected at suboptimal pH levels (pH 6) (30). Thus, cellular senescence can be monitored using X-gal staining by activating senescence-associated b-galactosidase at pH 6. In our experiment, UVB irradiation significantly increased cellular senescence, but this effect was reversed
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
Regulation of hedgehog-mediated photoageing
(a)
(b)
Figure 4. UVB-induced photoageing is attenuated by the inhibition of Hh signalling in a mouse model. (a) A UVB-induced photoageing mouse model was established in response to UVB irradiation for 12 week (three times per wk). After topical application of evodiamine or intraperitoneal injection of GDC-0449 or PD98059 prior to UVB irradiation, wrinkle formation at the dorsal skin was detected (left). The degree of wrinkle formation was evaluated according to the scale described in Table S2 (right). *P < 0.05 compared with non-irradiated mice, **P < 0.05 compared with UVB-irradiated mice. (b) Mouse dorsal skin was fixed and embedded in paraffin. The sections were stained with H&E staining (top). Thickness levels of epidermis (E) and dermis (D) were estimated (bottom, left) and presented as a graph (bottom, right). *P < 0.05 compared with non-irradiated mice, **P < 0.05 compared with UVB-irradiated mice. Scale bar = 50 lm. (c) To measure Hh signalling and inflammation-related factors in a mouse model in response to pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, GLI1, GLI2, COX-2 and IL-1b were detected by IHC staining. Scale bar = 50 lm. (d) A schematic diagram illustrating Hh signalling in response to UVB irradiation and the antiphotoageing effects of evodiamine in HaCaT cells. UVB irradiation increases the expression of Hh ligands and Hh-mediated transcription factors, GLI1 and GLI2. This stimulates MAPK activation and inflammatory responses accompanied with alterations of MMP and TIMP1 expression, leading to premature senescence and photoageing. The UVB-induced photoageing phenotype in keratinocytes can be attenuated by treatment of evodiamine as a novel compound for the inhibition of Hh signalling.
UVB-induced photoageing is attenuated by the inhibition of hedgehog signalling in a mouse model
(c)
(d)
by pretreatment with evodiamine, GDC-0449 or PD98059 (Fig. 3d). Our findings indicate that UVB-induced cellular senescence is suppressed by the inhibition of Hh signalling.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
To confirm the correlation between Hh signalling and photoageing in vivo, a mouse model of photoageing was established. Following UVB irradiation for 12 week, wrinkles formed in the dorsal skin (Fig. 4a); however, wrinkle formation was suppressed when evodiamine, GDC-0449 or PD98059 was applied. As shown in Fig. 4b, UVB irradiation also increased the thickness of both the epidermis and dermis, but this effect was attenuated by treatment with evodiamine, GDC-0449 or PD98059. In addition, we conducted IHC assays to measure the protein expression of GLI1, GLI2, COX-2 and IL-1b. We confirmed in vivo that the levels of GLI1 and GLI2 were elevated in photoaged skins, but recovered by pretreatment with evodiamine, GDC-0449 or PD98059 (Fig. 4c). The levels of two pro-inflammatory factors, COX-2 and IL-1b, were also increased in mice irradiated with UVB. Expression of these factors was then reduced by administration of evodiamine, GDC-0449 or PD98059. Collectively, our data suggest that UVB-induced Hh signalling is closely associated with photoageing and evodiamine is a promising inhibitor for signalling to overcome UVB-induced photoageing.
Discussion A large number of patients have been suffering from skin photoageing caused by overexposure to UVB (3). Although the human immune system has protective molecules involved in absorbing harmful UV light and scavenging ROS, including pigments and proteins, these molecules cannot heal the accumulated damage caused by prolonged and chronic UVB exposure (1). Accordingly, it is necessary to develop reagents and antioxidants capable of blocking UVB-induced photoageing, especially those with molecular mechanisms of action (31). Therefore, to identify the molecular pathway associated with UVB-induced photoageing, we focused on Hh signalling based on the high correlation of photoageing with skin cancer and crucial roles of Hh signalling in skin tumorigenesis. In this study, we found for the first time that Hh signalling is mainly responsible for UVB-induced photoageing. Specifically, Hh ligands were upregulated by UVB, leading to increased expression of GLI1 and GLI2 (Fig. 1), which affected MAPK (p38, ERK and JNK) activation and inflammatory responses (upregulation of COX-2, IL-1b, IL-6 and TNF-a) lead-
615
Kim et al.
ing to premature senescence and photoageing (Fig. 2). We also identified evodiamine as a novel compound for negative regulation of UVB-induced photoageing through the inhibition of Hh signalling in vitro and in vivo. The results of this study demonstrated the involvement of Hh signalling in the UVB-induced photoageing process, which is different from that of normal ageing (Fig. 1). In contrast to Hh activation in response to UVB, the Hh ligands of rat skin tissues were found to be upregulated at young and adult age (6- to 18-m-old), but significantly downregulated at old age (24-m-old) (Fig. S1). Although further experimental evidence is needed to confirm this phenomenon, one possible explanation is that activation of Hh signalling is essential to controlling growth and metabolism during the growth period of adults (32,33). Nevertheless, the results of the present study imply that the functions of Hh signalling might be indirectly involved in the normal senescence process of time-dependent ageing. We also observed that administration of GDC-0449 significantly downregulated ERK phosphorylation and partially decreased p38 and JNK phosphorylation under UVB irradiation (Fig. 2a). It was shown that inflammatory genes, including those encoding COX-2, IL-1b, IL-6 and TNF-a, were downregulated by treatment of a p38 inhibitor or a JNK inhibitor as similar as an ERK inhibitor (Fig. 2b). These results indicate that ERK acts as a major mediator of UVBinduced Hh signalling and p38 and JNK act as partial mediators leading to induction of inflammatory gene expression through crosstalk pathways for UVB-induced photoageing. Hh activation under UVB irradiation was accompanied by the activation of several stress-responsive proteins, including p38, JNK, COX-2 and several cytokines (Fig. 2). Thus, we assumed that Hh signalling might specifically participate in stress-induced premature ageing including UVB-induced photoageing and is a major regulatory pathway for the activation of MAPKs and inflammatory response to cause such a phenotype. It is known that pro-inflammatory cytokines, including IL-1b, IL-6 and TNF-a, play a significant role in photoageing through MMP activation and alterations of tissue components such as subcutaneous fat (34,35). Our data are supported by several studies that showed that overexpression of SHH could trigger an increase in TNF-a and IL-6 mRNA and that GLI2 is a transcription factor for the expression of IL-1b and IL-6 (25– 27). Moreover, it has been reported that mice with activation of Hh signalling by knockdown of Ptch show a marked reduction of their BCC when they are also COX-2 deficient (25). When it comes to ERK and p38, there are several clues for direct association between Hh signalling and activation of MAPK pathway. ERK1/2 is phosphorylated and activated by SHH stimulation under various cellular responses, such as myoblast differentiation, morphogenesis and cancer progression (5,36). One report demonstrated that SHH-induced signalling mediates rapid increases in p38 phosphorylation and selective inhibition of p38 significantly downregulates SHH-dependent expression of GLI1 (37). Although there is no direct evidence of Hh signalling and JNK activation, a few studies have shown that treatment of Hh antagonist reduces JNK phosphorylation and that SHH is generated in a JNK-dependent manner under inflammation of acute liver injury (38,39). In addition, GLI1 is identified as a substrate of JNK and ERK in a recent study of computational prediction and experimental verification (40). Collectively, these accumulated data support our
616
findings that Hh signalling might stimulate inflammation through induction of pro-inflammatory cytokines and COX-2 as well as via crosstalk with the MAPK pathway, leading to UVB-induced photoageing. These results also indicate that controlling Hh signalling could be essential to UVB-induced skin lesions. After UVB absorption in the epidermis, a great deal of ROS is generated and several cytokines in keratinocytes are subsequently released, resulting in stimulation of MMP expression and suppression of TIMP expression (41). Increased protein levels of MMPs, which can be suppressed by TIMP, lead to impaired skin structure and formation of wrinkles through degradation of collagen and elastin in skin (6). As MMP1 is a collagenase and MMP2 and MMP9 are gelatinases, they can contribute to UVB-induced photoageing (42). In the present study, we showed that the expression of MMP1, MMP2 and MMP9 increased and expression of TIMP1 decreased in response to UVB exposure, but that these effects were reversed by evodiamine treatment (Fig. 3). However, evodiamine is not known to have antioxidant activity to scavenge excessive ROS that can affect MMP expression. Thus, these findings indicate that evodiamine is capable of modulation of MMP and TIMP expression through direct involvement of their transcription, such as inhibition of Hh signalling, not via regulation of ROS levels under UVB exposure. In addition to alteration of MMP expression, we found that administration of evodiamine induced a significant decrease in premature senescence, wrinkle formation and the thickness of the epidermis and dermis (Fig. 3,4). Taken together, these findings suggest that evodiamine might be a promising agent for antiwrinkle formation and antiphotoageing properties. Considering the importance of the effects of Hh signalling on multifunctional activities, we attempted to identify novel reagents for Hh inhibition. As there is direct evidence of activation of Hh signalling in skin cancer, many groups have developed selective Hh inhibitors, especially those targeting Smo. These approaches have gradually shown to be curative in the treatment of aggressive skin cancers that are locally advanced or metastatic. As cyclopamine, a phytochemical, has been identified as a Smo inhibitor, various molecules have been developed to render greater potency and specificity (43). As a result of these efforts, vismodegib was approved by the FDA in 2012. However, it was suggested that this drug should not be administered during pregnancy due to its teratogenic and fetotoxic properties (44). In addition to vismodegib, other molecules known to lead to Hh inhibition such as LED225 and IPI-926 have been investigated in clinical trials. Unfortunately, a study has recently reported that LED225 can induce taste disturbance as undesired effects (45). Therefore, we focused on novel phytochemicals as alternatives of previously developed Hh inhibitors because natural products derived from plants have been reported to have various biological activities, but a relatively safe level of cytotoxicity. We identified evodiamine as a potent Hh inhibitor from screening of plant-derived chemicals based on the molecular properties. When keratinocytes were treated with evodiamine alone and then irradiated with UVB, mRNA expressions of both GLI1 and GLI2 were downregulated to normal levels (Fig. 1c, e). In contrast, it is well known that GLI1 and GLI2 play crucial roles in the development of skin cancers such as BCC (46,47). Indeed, the maintenance of normal GLI1 and GLI2 levels is important as both are transcription factors of Hh signalling involved in cell growth transition from resting state in various tissues including
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
Regulation of hedgehog-mediated photoageing
skin and hair follicles (48). Recently, it has been reported that GLI1-positive cells in the perivascular niche of many organs are involved in regulation of adult tissue repair (49,50). Considering specified wound-healing mechanism in the skin (51), UVB-damaged skin cells would be hypoproliferative and, consequently, the recovery rate against UVB-induced damage would be negatively affected if a compound such as silibinin downregulates either GLI1 or GLI2 much less than normal levels. Thus, evodiamine is expected to be able to modulate Hh signalling to normal levels and have the potential for use as not only a novel lead compound for skin photoageing, but also for skin cancer. Furthermore, our results provided a clue that evodiamine influences on key factors regulated by Hh ligands although we could not suggest a specific binding partner of evodiamine. We observed that the effects of evodiamine were similar to those of GDC-0449, a Smo inhibitor, showing decrease of GLI1/2 transcriptional activity, MAPK phosphorylation, pro-inflammatory genes induction and MMP activation. Treatment of evodiamine, consequently, led to the inhibition of Hh signalling and photoageing phenotype. From these results, we assumed that Smo activity, at least in part, might be one of major targets affected by evodiamine. Accordingly, further structural and molecular studies of evodiamine and proteins in Hh signalling should be performed. Moreover, this will enable optimization of the capacity of target inhibition through structure–activity relationship analysis to advance the potency and specificity of evodiamine while minimizing possible adverse effects. To date, the exact molecular mechanism underlying Hh signalling in skin photoageing has been elusive. In the present study, we demonstrated that Hh signalling could be crucial for conferring UVB-induced photoageing (Fig. 4d). Here, we provide the first evidence of a novel mechanism of photoageing induced by Hh
References 1 Gilchrest B A. J Invest Dermatol 2013: 133: E2– E6. 2 Ichihashi M, Ando H. Exp Dermatol 2014: 23 (Suppl 1): 43–46. 3 Pandel R, Poljsak B, Godic A et al. ISRN Dermatol 2013: 2013: 930164. 4 Yaar M, Gilchrest B A. Br J Dermatol 2007: 157: 874–887. 5 Elia D, Madhala D, Ardon E et al. Biochim Biophys Acta 2007: 1773: 1438–1446. 6 Kim S R, Jung Y R, An H J et al. PLoS ONE 2013: 8: e73877. 7 Briscoe J, Therond P P. Nat Rev Mol Cell Biol 2013: 14: 416–429. 8 Harrison W, Cochrane B, Neill G et al. Oncogene 2014: 33: 2432–2440. 9 Katoh Y, Katoh M. Curr Mol Med 2009: 9: 873–886. 10 Pasca di Maglian M, Hebrok M. Nat Rev Cancer 2003: 3: 903–911. 11 Pantazi E, Gemenetzidis E, Trigiante G et al. Cell Death Dis 2014: 5: e1028. 12 Berman D M, Karhadkar S S, Maitra A et al. Nature 2003: 425: 846–851. 13 Yu H, Jin H, Gong W et al. Molecules 2013: 18: 1826–1843. 14 Liu Y N, Pan S L, Liao C H et al. Shock 2009: 32: 263–269. 15 Jiang J, Hu C. Molecules 2009: 14: 1852–1859. 16 Liao C H, Pan S L, Guh J H et al. Carcinogenesis 2005: 26: 968–975. 17 Nindl I, Dang C, Forschner T et al. Mol Cancer 2006: 5: 30.
signalling with inflammatory response and MAPK activation. We also demonstrated that evodiamine might be a potential inhibitor of Hh signalling, resulting in attenuation of UVB-induced photoageing. The results of this study indicated that regulation of Hh signalling with a pharmacologic reagent, evodiamine, could be a promising strategy for inhibiting inflammation and photoageing and decreasing skin cancer risk.
Acknowledgements This research was financially supported by the Ministry of Education (MOE) and National Research Foundation of Korea (NRF) through the Human Resource Training Project for Regional Innovation (2012H1B8A2026225) and supported by a 2-Year Research Grant of Pusan National University. The authors thank the Aging Tissue Bank for providing aged skin tissues.
Author contributions WK, YJ, CK and BY designed the study. WK, EK, HJY, TK, SH, SL and HY performed the research. WK, EG, TK, HY, YJ, CK and BY analysed the data. WK, EK and CK and BY wrote the manuscript. BY supervised the research.
Conflict of interest The authors have declared no conflicting interests.
Supporting Information Additional supporting data may be found in the supplementary information of this article. Data S1.Materials and methods. Figure S1. The mRNA levels of Shh and Ihh in skin tissues of aged SD rats were measured to determine whether Hh signalling is associated with the general ageing process. Figure S2. The inhibitory effects of Hh signalling by evodiamine were rescued by overexpression of GLI1 or GLI2. Table S1. Primers for determining the expression levels of Hh signalling- and inflammation-related genes. Table S2. Grading of UVB-induced skin wrinkles in a mouse model. Appendix S1. References.
18 Athar M, Li C, Tang X et al. Cancer Res 2004: 64: 7545–7552. 19 Brellier F, Marionnet C, Chevallier-Lagente O et al. Cancer Res 2004: 64: 2699–2704. 20 Epstein J H. J Am Acad Dermatol 1983: 9: 487– 502. 21 Mann J. Nat Rev Cancer 2002: 2: 143–148. 22 Kim W, Seong K M, Youn B. Exp Mol Med 2011: 43: 323–333. 23 Kim W, Youn H, Kwon T et al. Pharmacol Res 2013: 70: 90–101. 24 Kessaris N, Jamen F, Rubin L L et al. Development 2004: 131: 1289–1298. 25 Tang J Y, Aszterbaum M, Athar M et al. Cancer Prev Res (Phila) 2010: 3: 25–34. 26 Elsawa S F, Almada L L, Ziesmer S C et al. J Biol Chem 2011: 286: 21524–21534. 27 Ghorpade D S, Sinha A Y, Holla S et al. J Biol Chem 2013: 288: 33037–33048. 28 Saqui-Salces M, Coves-Datson E, Veniaminova N A et al. PLoS ONE 2012: 7: e48039. 29 Adachi H, Murakami Y, Tanaka H et al. Exp Dermatol 2014: 23 (Suppl 1): 32–36. 30 Aliouat-Denis C M, Dendouga N, Van den Wyngaert I et al. Mol Cancer Res 2005: 3: 627–634. 31 Mizutani T, Masaki H. Exp Dermatol 2014: 23 (Suppl 1): 23–26. 32 Teperino R, Aberger F, Esterbauer H et al. Semin Cell Dev Biol 2014: 33: 81–92. 33 Varjosalo M, Taipale J. Genes Dev 2008: 22: 2454–2472. 34 Borg M, Brincat S, Camilleri G et al. Climacteric 2013: 16: 514–521.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
35 Li W H, Pappas A, Zhang L et al. J Dermatol Sci 2013: 71: 58–66. 36 Lu J T, Zhao W D, He W et al. Acta Pharmacol Sin 2012: 33: 691–700. 37 Atkinson P J, Dellovade T, Albers D et al. J Neurochem 2009: 108: 1539–1549. 38 Weston C R, Wong A, Hall J P et al. Genes Dev 2003: 17: 1271–1280. 39 Pratap A, Panakanti R, Yang N et al. Pharm Res 2010: 27: 2492–2504. 40 Whisenant T C, Ho D T, Benz R W et al. PLoS Comput Biol 2010: 6: e1000908. 41 Zaid M A, Afaq F, Syed D N et al. Photochem Photobiol 2007: 83: 882–888. 42 Kim M S, Oh G H, Kim M J et al. Photochem Photobiol 2013: 89: 911–918. 43 Lin T L, Matsui W. Onco Targets Ther 2012: 5: 47–58. 44 Von Hoff D D, LoRusso P M, Rudin C M et al. N Engl J Med 2009: 361: 1164–1172. 45 Kumari A, Ermilov A N, Allen B L et al. J Neurophysiol 2015: 113: 1034–1040. 46 Dahmane N, Lee J, Robins P et al. Nature 1997: 389: 876–881. 47 Tojo M, Kiyosawa H, Iwatsuki K et al. Br J Dermatol 2003: 148: 892–897. 48 Green J, Leigh I M, Poulsom R et al. Br J Dermatol 1998: 139: 911–915. 49 Kramann R, Schneider R K, DiRocco D P et al. Cell Stem Cell 2015: 16: 51–66. 50 Zhao H, Feng J, Seidel K et al. Cell Stem Cell 2014: 14: 160–173. 51 Heng M C. Int J Dermatol 2011: 50: 1058–1066.
617
Original Article
www.wileyonlinelibrary.com/journal/EXD
Inhibition of hedgehog signalling attenuates UVB-induced skin photoageing Wanyeon Kim1,2,*, EunGi Kim3,*, Hee Jung Yang1, TaeWoo Kwon3, SeoYoung Han3, Sungmin Lee3, HyeSook Youn1,2, Youngmi Jung1,3, ChulHee Kang4 and BuHyun Youn1,2,3 1 Department of Biological Sciences, Pusan National University, Busan, South Korea; 2Nuclear Science Research Institute, Pusan National University, Busan, South Korea; 3Department of Integrated Biological Science, Pusan National University, Busan, South Korea; 4Department of Chemistry, Washington State University, Pullman, Washington, USA Correspondence: BuHyun Youn, Department of Biological Sciences, Pusan National University, Busan, 609-735, South Korea, Tel.: 82-51-510-2264, Fax: 82-51-581-2962, e-mail: [email protected] *These authors contributed equally to this work.
Abstract: The hedgehog (Hh) signalling pathway regulates normal development and cell proliferation in metazoan organisms, but its aberrant activation can promote tumorigenesis and progression of a variety of aggressive human cancers including skin cancer. Despite its importance, little is known about its role in photoageing, a type of UV-induced skin lesions. In this study, we investigated the involvement of Hh signalling in the photoageing process as well as the use of an Hh-regulating alkaloid compound as a novel therapeutic drug to regulate photoageing in keratinocytes. We found that UVB induced Hh signalling by the expression of Hh ligands and Hh-mediated transcription factors, respectively. Moreover, UVB-induced Hh activation relied on mitogen-activated protein kinase (p38, ERK and JNK) activity and
inflammatory responses (upregulation of COX-2, IL-1b, IL-6 and TNF-a), resulting in premature senescence and photoageing in vitro and in vivo. Notably, a selected Hh inhibitor, evodiamine, mediated photoageing blockade in a mouse skin model. Taken together, our findings demonstrated that Hh signalling is associated with UVB-induced photoageing, while pharmacological inhibition of Hh signalling significantly reduced experimental photoageing, indicating its potential for use as a therapeutic target for this disease.
Introduction
collagens. Moreover, activation of transcription factor NF-кB induces the expression of pro-inflammatory cytokines involved in immunoregulation and cell survival such as IL-1, IL-6, vascular endothelial growth factor, TNF-a and MMPs (6). Taken together, these findings suggest that these inflammatory responses lead to UVB-induced photoageing in skin. Hedgehog (Hh) ligands consist of Sonic Hh (SHH), Desert Hh (DHH) and Indian Hh (IHH), which are secreted proteins (7). Hh ligands are modulated by post-translational modifications, autoproteolytic cleavage and lipidation. Secreted Hh ligands bind to Hip1, Patched (Ptch)-1 and Ptch-2 of the cell surface. Smoothened (Smo) is inhibited by Ptch under normal conditions. When Hh ligands bind to Ptch, Smo is activated, leading to transcriptional activation of glioma (GLI). GLI1 lacks a transcriptional repressor domain, while cleaved GLI2 and GLI3 have repressor functions. GLI2 is mainly accountable for the activator function in response to Hh signalling, while GLI3 is known to be responsible for the activity of repressor. In the absence of Hh signal, GLI2 is phosphorylated by protein kinase A, casein kinase I and glycogen synthase kinase 3b, promoting interaction of the E3 ligase Slimb and b-transducin repeat-containing protein. The ubiquitinated GLI2 is partially processed by the proteasome, after which cleaved GLI2 translocates into the nucleus and represses target gene expression. Activated Smo facilitates the transcriptional activation of GLI2 through the inhibition of phosphorylation. The activated forms of GLI2 stimulate the expression of target genes such as GLI1, CCND1 and BCL2 in the nucleus (8,9). Deregulation of the
The skin functions as a defensive barrier against environmental factors, primarily ultraviolet (UV) irradiation. Chronic exposure to sunlight over the maximal cumulative solar UV dose leads to wrinkling, dryness, pigmentation and premature ageing of the skin, which is known as photoageing (1,2). UV radiation consists of UVA (320–400 nm), UVB (280–320 nm) and UVC (200– 280 nm). As UVC is filtered out by the ozone layer, UVA and UVB are considered to be the primary factors responsible for induction of photoageing (3). UVB is far more mutagenic and energetic than UVA and leads to more deleterious outcomes (4). After penetrating the epidermis of the skin, UVB can cause a variety of adverse effects, such as epidermal thickness, dermal elastosis, degradation of collagen by increased matrix metalloproteinases (MMPs), inflammatory infiltration and vessel ectasia in skin. Consequently, it stimulates erythema, inflammation, hyperplasia, formation of wrinkles and photoageing, a type of UVB-induced skin lesions, which may lead to cancer development. UVB-induced photoageing is initiated by production of reactive oxygen species (ROS), which activates several receptors for epidermal growth factor, interleukin (IL)-1, keratinocyte growth factor and tumor necrosis factor (TNF)-a. Activation of these receptors induces downstream signalling pathways of mitogen-activated protein kinases (MAPKs) such as p38, JNK and ERK, which leads to an increase of the nuclear transcription complex activator protein 1 (AP-1), c-Jun and c-Fos (4,5). Subsequently, MMPs as target genes of AP-1 are upregulated and induce degradation of
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
Key words: evodiamine – hedgehog – inflammation – photoageing – UVB
Accepted for publication 24 April 2015
611
Kim et al.
Hh signalling pathway initiates tumor formation and mediates survival of the brain, muscle and skin cancers (10). Several studies have provided evidence that Hh signalling is responsible for aberrant proliferation of skin cells and induction of genomic instability, leading to the development of basal cell carcinoma (BCC), a type of skin cancer (11,12). Aberrant Hh signalling leads to an increase in GLI expression and activation, which consequently induces tumor progression in the skin. Although many studies have been conducted to investigate the association between Hh signalling and tumorigenesis of skin cells, little is known about the role of Hh signalling in photoageing.
(a)
(b)
Evodiamine is a quinolone alkaloid isolated from the fruit of Evodia rutaecarpa that has the ability to protect against oxidative damage and modulate inflammatory response through the inhibition of several proteins, including inducible nitric oxide synthase, cyclo-oxygenase 2 (COX-2) and hypoxia-inducible factor 1a (13,14). Moreover, evodiamine has been shown to exert anticancer activity against many cancer types, including induction of apoptosis and inhibition of cell proliferation, invasion and metastasis (15,16). Nevertheless, the function of evodiamine associated with antiphotoageing activity and regulation of Hh signalling in skin is not known. This study was conducted to clarify the mechanisms and key molecules that confer UVB-induced skin photoageing. We demonstrated that UVB-induced activation of Hh signalling and subsequent activation of the MAPK pathway and inflammatory response play important roles in induction of premature senescence in HaCaT cells and wrinkle formation in a mouse model. Our findings provide a possible explanation of how Hh signalling is involved in UVB-induced skin damage. Furthermore, we suggest that Hh signalling has great potential for use as a pharmaceutical target to prevent photoageing, and evodiamine is a promising candidate for the inhibition of Hh signalling.
Materials and methods
(c)
The materials and methods used in this study, including cell culture; drug treatment; transient transfection; UVB irradiation; Western blot analysis; cell viability assay; luciferase reporter assay; real-time quantitative RT-PCR (qRT-PCR); assays for the production of prostaglandin E2 (PGE2), IL-1b and IL-6; MMP activity assay; senescence-associated b-galactosidase staining; animal protocol; evaluation of wrinkle formation; haematoxylin and eosin (H&E) staining; immunohistochemistry (IHC); and statistical analysis, are described in detail in the Supporting Information.
Results Hedgehog signalling is upregulated by UVB in HaCaT cells Hh signalling is important for tumor formation and growth in skin (8). Based on this information, we analysed the expression of SHH (a most representative Hh ligand) in normal skin and squamous cell
(d)
(e)
(f)
(g)
612
Figure 1. Hh signalling is upregulated by UVB in HaCaT cells. (a) To estimate the expression levels of SHH in skin cancer, data derived from the Oncomine database (www.oncomine.org) using biopsies of normal skin (n = 6) and SCCs (n = 5) were analysed (17). (b) To measure Hh signalling in response to UVB irradiation, the protein levels of Smo and cleaved form of SHH and IHH were measured. At 3 h after UVB irradiation, the cells were harvested and the cell lysates were subjected to Western blot analysis. (c) To select candidates to regulate Hh signalling, several natural products (evodiamine (Evo), silibinin, rhamnetin, cirsiliol, nordihydroguaiaretic acid (NDGA), caffeic acid phenethyl ester (CAPE) and curcumin) were screened. After pretreatment with these compounds and UVB irradiation, mRNA levels of GLI1 and GLI2 were analysed by qRT-PCR. *P > 0.05 compared with non-irradiated cells as well as P < 0.05 compared with UVBirradiated cells, **P < 0.05 compared with non-irradiated cells as well as P < 0.05 compared with UVB-irradiated cells. (d) To determine the non-cytotoxic dose of evodiamine, the viability of cells was assessed by a triazolyl blue tetrazolium bromide assay. Cells were treated for 2 h with evodiamine at various concentrations up to 25 lM. (e) To estimate Hh signalling in response to UVB irradiation and treatment with evodiamine or GDC-0449 (a Smo inhibitor), the mRNA levels of GLI1 and GLI2 were assessed by qRT-PCR. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells. (f) To measure Hh signalling in response to pretreatment with evodiamine or GDC-0449 and UVB irradiation, protein levels of GLI1 and GLI2 (full-length as an activator form) from the cytoplasm to the nucleus were measured by Western blot analysis after cytoplasmic or nuclear fractionation. Tubulin and lamin B were used as markers for the cytoplasmic extract (CE) and nuclear extract (NE), respectively. (g) To determine the activation of Hh signalling in response to pretreatment with evodiamine or GDC-0449 and UVB irradiation, GLI-specific promoter activity was measured by a luciferase assay. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
Regulation of hedgehog-mediated photoageing
carcinomas (SCCs) using the Oncomine database (www.oncomine.org). We found that SHH expression in SCCs was higher than that found in normal skin (Fig. 1a) (17). In addition, several literatures demonstrated that Hh signalling is associated with UVBinduced skin cancers including SCCs and BCCs (11,18,19). Because UVB-induced photoageing may lead to the development of skin cancers (20), we hypothesized that activation of Hh signalling might be responsible for not only skin cancers, but also photoageing induced by UVB irradiation. To monitor alterations of the Hh pathway in response to UVB exposure, we measured the expression of Hh ligands and Smo in HaCaT cells after UVB irradiation. As shown in Fig. 1b, the protein levels of Smo, SHH and IHH increased in a dose-dependent manner. We next assessed the mRNA levels of SHH and IHH in skin tissues of aged Sprague Dawley (SD) rats to determine whether Hh signalling is associated with the general ageing process. The mRNA levels of both Hh ligands in skin tissues of SD rats gradually increased from the tissues of 6-m-old to 18-m-old rats and then markedly decreased in 24-m-old rats (Fig. S1), suggesting that Hh signalling is not directly correlated with normal ageing. In addition, we found that mRNA levels of both GLI1 and GLI2 were induced by UVB irradiation (Fig. 1c). Because several natural products have been reported to regulate the signalling pathway without adverse effects in various cancer cell lines, a large number of natural products were prepared for screening (21,22). The primary screening based on the biological/pharmacological activity of natural products was conducted by measuring transcriptional activity of GLI1/2 under UVB irradiation. Subsequently, we analysed the physical and chemical properties of compounds according to solubility and molecular weight along with the Lipinski-like criteria (23). After selection of fewer than 10 candidates, the compounds were retested at various concentrations to determine the potency and cytotoxicity of each compound. With an optimal concentration of each candidate, we assessed alteration of GLI1/GLI2 mRNA expression in response to treatment of each compound and UVB irradiation (Fig. 1c). Ultimately, one quinolone alkaloid, evodiamine, was selected for further investigation. To determine the concentration of evodiamine that could be used without cytotoxicity, HaCaT cells were treated with different concentrations of evodiamine (Fig. 1d). Evodiamine did not affect cell viability at concentrations up to 5 lM. To determine whether evodiamine can regulate the Hh signalling pathway, we measured the expression of GLI1 and GLI2 as the last effectors of Hh signalling. We found that evodiamine is able to suppress UVB-induced expression of GLI1 and GLI2 (Fig. 1e,f). GDC-0449, a Smo inhibitor, was treated to confirm the inhibitory effects of evodiamine on UVB-induced Hh signalling. In addition, a GLI-specific luciferase reporter assay was conducted to measure the transcriptional activity of GLIs in response to UVB irradiation (Fig. 1g). We observed that UVB increased the transcriptional activity of GLIs and this effect was downregulated by pretreatment of evodiamine or GDC-0449. Taken together, these findings suggest that UVB exposure can promote Hh signalling, which is inhibited by evodiamine treatment.
(a)
(b)
(c)
(d)
(e)
UVB-induced inflammation is downregulated by the inhibition of hedgehog signalling in HaCaT cells
Figure 2. UVB-induced inflammation is downregulated by the inhibition of Hh signalling in HaCaT cells. (a) To estimate the activation of MAPKs (p38, ERK and JNK) in response to pretreatment with evodiamine or GDC-0449 and UVB irradiation, phosphorylation of MAPKs was detected. At 3 h after UVB irradiation, the cells were harvested and the cell lysates were subjected to Western blot analysis. (b) To determine inflammation-related genes in response to pretreatment with evodiamine, SB203580 (a p38 inhibitor), PD98059 (an ERK inhibitor) or SP600125 (a JNK inhibitor) and UVB irradiation, the mRNA levels of COX-2, IL-1b, IL-6 and TNF-a were assessed by qRT-PCR. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells. (c) To measure Hh signalling in response to pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, protein levels of COX-2 and TNF-a were verified by Western blot analysis. (d) To measure COX-2 activity in response to pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, PGE2 in the media was detected by enzyme immunoassay. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells. (e) To measure cytokines secreted in response to pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, IL-1b and IL-6 in the media were analysed by enzyme immunoassay. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells.
UVB-induced photoageing is known to be associated with inflammation through the MAPK signalling pathway (4). In addition, several studies have shown that Hh signalling is accelerated via activation of MAPKs and overexpression of COX-2 (24,25). The
transcription of pro-inflammatory genes, including those encoding IL-1b, IL-6, IL-12 and TNF-a, can be also increased by GLIs under inflammatory responses such as bacterial infection,
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
613
Kim et al.
cytokine-mediated paracrine signalling or nitric oxide signalling (26–28). In this study, we determined whether the activation of MAPKs and expression of COX-2 and other pro-inflammatory genes associated with Hh signalling were modulated by UVB irradiation and evodiamine. As shown in Fig. 2a, phosphorylation of p38, ERK and JNK increased following UVB irradiation, but this response was decreased by the inhibition of Hh signalling with pretreatment of evodiamine or GDC-0449. In particular, we observed that treatment of GDC-0449 resulted in significant decrease in ERK phosphorylation and partial decrease in p38 and JNK phosphorylation. To further elucidate the involvement of inflammation in UVB-induced photoageing, we measured the expression of COX-2, IL-1b, IL-6 and TNF-a in HaCaT cells. We found that the mRNA levels of these inflammation markers were increased by UVB irradiation (Fig. 2b). Increased expression of these genes was reduced by pretreatment with evodiamine, GDC0449 or MAPK inhibitors. Based on the results shown in Fig. 2a and b, an ERK inhibitor, PD98059, was used to confirm the antiphotoageing effects of evodiamine and GDC-0449 on regulation of UVB-induced Hh signalling. Next, the expression levels of COX-2 and TNF-a were elevated following UVB irradiation in HaCaT cells, but protein expression was reduced to the basal levels by evodiamine, GDC-0449 or PD98059 (Fig. 2c). The inhibitory effects of evodiamine on the UVB-induced expression of COX-2 and TNF-a were reversed by overexpression of GLI1 or GLI2 (Fig. S2a,b). PGE2 is one of the most abundant metabolites of arachidonic acid produced by COX-2. To confirm that COX-2 activity was elevated in response to UVB irradiation, we analysed the production of PGE2 in HaCaT cells. PGE2 generation was increased by UVB irradiation and reduced by evodiamine, GDC0449 or PD98059 (Fig. 2d). As IL-1b and IL-6 act as pro-inflammatory cytokines after being released from the cells, we measured IL-1b and IL-6 secreted into the cell culture media. As shown in Fig. 2e, the secretion of IL-1b and IL-6 was significantly increased by UVB irradiation. In contrast, pretreatment with evodiamine, GDC-0449 or PD98059 suppressed the release of both cytokines. The inhibitory effects of evodiamine on the UVB-induced secretion of IL-1b and IL-6 were rescued by overexpression of GLI1 or GLI2 (Fig. S2c). These results demonstrated that UVB induces MAPK activation and inflammation, which is negatively regulated by the inhibition of Hh signalling by treatment with evodiamine.
UVB-induced cellular senescence is reversed by the inhibition of hedgehog signalling in HaCaT cells UVB-induced photoageing is known to be mediated by increased MMP expression and activation along with collagen degradation (15,29). To determine the effects of UVB on MMP expression, we measured changes in the levels of photoageing-related MMPs. The expression of three MMPs, MMP1, MMP2 and MMP9, increased following UVB irradiation, but this response was downregulated by the inhibition of Hh signalling with pretreatment of evodiamine, GDC-0449 or PD98059 (Fig. 3a). In contrast, the expression of tissue inhibitor of metalloproteinase 1 (TIMP1), a cellular inhibitor of MMPs, was reduced in response to UVB irradiation and rescued by the inhibition of Hh signalling. To determine the activity of MMPs in response to UVB irradiation, we performed zymography and activity assays of MMPs (Fig. 3b,c). Active MMPs increased following UVB irradiation and were reduced by evodiamine, GDC-0449 or PD98059. Next, we analysed cellular
614
(a)
(c)
(b)
(d)
Figure 3. UVB-induced cellular senescence is rescued by the inhibition of Hh signalling in HaCaT cells. (a) To estimate the effects of evodiamine, GDC-0449 or PD98059 on UVB-induced photoageing, protein levels of photoageing markers including MMP1, MMP2, MMP9 and TIMP1 were observed by Western blot analysis. (b) To measure the effects of evodiamine, GDC-0449 or PD98059 on UVB-induced photoageing, the activities of MMP1, MMP2 and MMP9 were assessed by b-casein or gelatin zymography. Conditioned media of HT1080 cells were used as positive controls for active MMP2 and MMP9. (c) To measure the effects of evodiamine, GDC-0449 or PD98059 on UVB-induced photoageing, the activities of MMP1, MMP2 and MMP9 were analysed by enzyme immunoassay. *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVBirradiated cells. (d) The inhibitory effects of evodiamine, GDC-0449 or PD98059 on UVB-induced senescence were measured by senescence-associated b-galactosidase staining. After pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, cells were fixed and stained by senescence-associated b-galactosidase staining solution (top). The graphs showed the percentage of the cells with staining-positive cells (bottom). *P < 0.05 compared with non-irradiated cells, **P < 0.05 compared with UVB-irradiated cells. Scale bar = 50 lm.
senescence by senescence-associated b-galactosidase staining. Although b-galactosidase is activated under acidic conditions, senescence-associated b-galactosidase activity is detected at suboptimal pH levels (pH 6) (30). Thus, cellular senescence can be monitored using X-gal staining by activating senescence-associated b-galactosidase at pH 6. In our experiment, UVB irradiation significantly increased cellular senescence, but this effect was reversed
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
Regulation of hedgehog-mediated photoageing
(a)
(b)
Figure 4. UVB-induced photoageing is attenuated by the inhibition of Hh signalling in a mouse model. (a) A UVB-induced photoageing mouse model was established in response to UVB irradiation for 12 week (three times per wk). After topical application of evodiamine or intraperitoneal injection of GDC-0449 or PD98059 prior to UVB irradiation, wrinkle formation at the dorsal skin was detected (left). The degree of wrinkle formation was evaluated according to the scale described in Table S2 (right). *P < 0.05 compared with non-irradiated mice, **P < 0.05 compared with UVB-irradiated mice. (b) Mouse dorsal skin was fixed and embedded in paraffin. The sections were stained with H&E staining (top). Thickness levels of epidermis (E) and dermis (D) were estimated (bottom, left) and presented as a graph (bottom, right). *P < 0.05 compared with non-irradiated mice, **P < 0.05 compared with UVB-irradiated mice. Scale bar = 50 lm. (c) To measure Hh signalling and inflammation-related factors in a mouse model in response to pretreatment with evodiamine, GDC-0449 or PD98059 and UVB irradiation, GLI1, GLI2, COX-2 and IL-1b were detected by IHC staining. Scale bar = 50 lm. (d) A schematic diagram illustrating Hh signalling in response to UVB irradiation and the antiphotoageing effects of evodiamine in HaCaT cells. UVB irradiation increases the expression of Hh ligands and Hh-mediated transcription factors, GLI1 and GLI2. This stimulates MAPK activation and inflammatory responses accompanied with alterations of MMP and TIMP1 expression, leading to premature senescence and photoageing. The UVB-induced photoageing phenotype in keratinocytes can be attenuated by treatment of evodiamine as a novel compound for the inhibition of Hh signalling.
UVB-induced photoageing is attenuated by the inhibition of hedgehog signalling in a mouse model
(c)
(d)
by pretreatment with evodiamine, GDC-0449 or PD98059 (Fig. 3d). Our findings indicate that UVB-induced cellular senescence is suppressed by the inhibition of Hh signalling.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
To confirm the correlation between Hh signalling and photoageing in vivo, a mouse model of photoageing was established. Following UVB irradiation for 12 week, wrinkles formed in the dorsal skin (Fig. 4a); however, wrinkle formation was suppressed when evodiamine, GDC-0449 or PD98059 was applied. As shown in Fig. 4b, UVB irradiation also increased the thickness of both the epidermis and dermis, but this effect was attenuated by treatment with evodiamine, GDC-0449 or PD98059. In addition, we conducted IHC assays to measure the protein expression of GLI1, GLI2, COX-2 and IL-1b. We confirmed in vivo that the levels of GLI1 and GLI2 were elevated in photoaged skins, but recovered by pretreatment with evodiamine, GDC-0449 or PD98059 (Fig. 4c). The levels of two pro-inflammatory factors, COX-2 and IL-1b, were also increased in mice irradiated with UVB. Expression of these factors was then reduced by administration of evodiamine, GDC-0449 or PD98059. Collectively, our data suggest that UVB-induced Hh signalling is closely associated with photoageing and evodiamine is a promising inhibitor for signalling to overcome UVB-induced photoageing.
Discussion A large number of patients have been suffering from skin photoageing caused by overexposure to UVB (3). Although the human immune system has protective molecules involved in absorbing harmful UV light and scavenging ROS, including pigments and proteins, these molecules cannot heal the accumulated damage caused by prolonged and chronic UVB exposure (1). Accordingly, it is necessary to develop reagents and antioxidants capable of blocking UVB-induced photoageing, especially those with molecular mechanisms of action (31). Therefore, to identify the molecular pathway associated with UVB-induced photoageing, we focused on Hh signalling based on the high correlation of photoageing with skin cancer and crucial roles of Hh signalling in skin tumorigenesis. In this study, we found for the first time that Hh signalling is mainly responsible for UVB-induced photoageing. Specifically, Hh ligands were upregulated by UVB, leading to increased expression of GLI1 and GLI2 (Fig. 1), which affected MAPK (p38, ERK and JNK) activation and inflammatory responses (upregulation of COX-2, IL-1b, IL-6 and TNF-a) lead-
615
Kim et al.
ing to premature senescence and photoageing (Fig. 2). We also identified evodiamine as a novel compound for negative regulation of UVB-induced photoageing through the inhibition of Hh signalling in vitro and in vivo. The results of this study demonstrated the involvement of Hh signalling in the UVB-induced photoageing process, which is different from that of normal ageing (Fig. 1). In contrast to Hh activation in response to UVB, the Hh ligands of rat skin tissues were found to be upregulated at young and adult age (6- to 18-m-old), but significantly downregulated at old age (24-m-old) (Fig. S1). Although further experimental evidence is needed to confirm this phenomenon, one possible explanation is that activation of Hh signalling is essential to controlling growth and metabolism during the growth period of adults (32,33). Nevertheless, the results of the present study imply that the functions of Hh signalling might be indirectly involved in the normal senescence process of time-dependent ageing. We also observed that administration of GDC-0449 significantly downregulated ERK phosphorylation and partially decreased p38 and JNK phosphorylation under UVB irradiation (Fig. 2a). It was shown that inflammatory genes, including those encoding COX-2, IL-1b, IL-6 and TNF-a, were downregulated by treatment of a p38 inhibitor or a JNK inhibitor as similar as an ERK inhibitor (Fig. 2b). These results indicate that ERK acts as a major mediator of UVBinduced Hh signalling and p38 and JNK act as partial mediators leading to induction of inflammatory gene expression through crosstalk pathways for UVB-induced photoageing. Hh activation under UVB irradiation was accompanied by the activation of several stress-responsive proteins, including p38, JNK, COX-2 and several cytokines (Fig. 2). Thus, we assumed that Hh signalling might specifically participate in stress-induced premature ageing including UVB-induced photoageing and is a major regulatory pathway for the activation of MAPKs and inflammatory response to cause such a phenotype. It is known that pro-inflammatory cytokines, including IL-1b, IL-6 and TNF-a, play a significant role in photoageing through MMP activation and alterations of tissue components such as subcutaneous fat (34,35). Our data are supported by several studies that showed that overexpression of SHH could trigger an increase in TNF-a and IL-6 mRNA and that GLI2 is a transcription factor for the expression of IL-1b and IL-6 (25– 27). Moreover, it has been reported that mice with activation of Hh signalling by knockdown of Ptch show a marked reduction of their BCC when they are also COX-2 deficient (25). When it comes to ERK and p38, there are several clues for direct association between Hh signalling and activation of MAPK pathway. ERK1/2 is phosphorylated and activated by SHH stimulation under various cellular responses, such as myoblast differentiation, morphogenesis and cancer progression (5,36). One report demonstrated that SHH-induced signalling mediates rapid increases in p38 phosphorylation and selective inhibition of p38 significantly downregulates SHH-dependent expression of GLI1 (37). Although there is no direct evidence of Hh signalling and JNK activation, a few studies have shown that treatment of Hh antagonist reduces JNK phosphorylation and that SHH is generated in a JNK-dependent manner under inflammation of acute liver injury (38,39). In addition, GLI1 is identified as a substrate of JNK and ERK in a recent study of computational prediction and experimental verification (40). Collectively, these accumulated data support our
616
findings that Hh signalling might stimulate inflammation through induction of pro-inflammatory cytokines and COX-2 as well as via crosstalk with the MAPK pathway, leading to UVB-induced photoageing. These results also indicate that controlling Hh signalling could be essential to UVB-induced skin lesions. After UVB absorption in the epidermis, a great deal of ROS is generated and several cytokines in keratinocytes are subsequently released, resulting in stimulation of MMP expression and suppression of TIMP expression (41). Increased protein levels of MMPs, which can be suppressed by TIMP, lead to impaired skin structure and formation of wrinkles through degradation of collagen and elastin in skin (6). As MMP1 is a collagenase and MMP2 and MMP9 are gelatinases, they can contribute to UVB-induced photoageing (42). In the present study, we showed that the expression of MMP1, MMP2 and MMP9 increased and expression of TIMP1 decreased in response to UVB exposure, but that these effects were reversed by evodiamine treatment (Fig. 3). However, evodiamine is not known to have antioxidant activity to scavenge excessive ROS that can affect MMP expression. Thus, these findings indicate that evodiamine is capable of modulation of MMP and TIMP expression through direct involvement of their transcription, such as inhibition of Hh signalling, not via regulation of ROS levels under UVB exposure. In addition to alteration of MMP expression, we found that administration of evodiamine induced a significant decrease in premature senescence, wrinkle formation and the thickness of the epidermis and dermis (Fig. 3,4). Taken together, these findings suggest that evodiamine might be a promising agent for antiwrinkle formation and antiphotoageing properties. Considering the importance of the effects of Hh signalling on multifunctional activities, we attempted to identify novel reagents for Hh inhibition. As there is direct evidence of activation of Hh signalling in skin cancer, many groups have developed selective Hh inhibitors, especially those targeting Smo. These approaches have gradually shown to be curative in the treatment of aggressive skin cancers that are locally advanced or metastatic. As cyclopamine, a phytochemical, has been identified as a Smo inhibitor, various molecules have been developed to render greater potency and specificity (43). As a result of these efforts, vismodegib was approved by the FDA in 2012. However, it was suggested that this drug should not be administered during pregnancy due to its teratogenic and fetotoxic properties (44). In addition to vismodegib, other molecules known to lead to Hh inhibition such as LED225 and IPI-926 have been investigated in clinical trials. Unfortunately, a study has recently reported that LED225 can induce taste disturbance as undesired effects (45). Therefore, we focused on novel phytochemicals as alternatives of previously developed Hh inhibitors because natural products derived from plants have been reported to have various biological activities, but a relatively safe level of cytotoxicity. We identified evodiamine as a potent Hh inhibitor from screening of plant-derived chemicals based on the molecular properties. When keratinocytes were treated with evodiamine alone and then irradiated with UVB, mRNA expressions of both GLI1 and GLI2 were downregulated to normal levels (Fig. 1c, e). In contrast, it is well known that GLI1 and GLI2 play crucial roles in the development of skin cancers such as BCC (46,47). Indeed, the maintenance of normal GLI1 and GLI2 levels is important as both are transcription factors of Hh signalling involved in cell growth transition from resting state in various tissues including
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
Regulation of hedgehog-mediated photoageing
skin and hair follicles (48). Recently, it has been reported that GLI1-positive cells in the perivascular niche of many organs are involved in regulation of adult tissue repair (49,50). Considering specified wound-healing mechanism in the skin (51), UVB-damaged skin cells would be hypoproliferative and, consequently, the recovery rate against UVB-induced damage would be negatively affected if a compound such as silibinin downregulates either GLI1 or GLI2 much less than normal levels. Thus, evodiamine is expected to be able to modulate Hh signalling to normal levels and have the potential for use as not only a novel lead compound for skin photoageing, but also for skin cancer. Furthermore, our results provided a clue that evodiamine influences on key factors regulated by Hh ligands although we could not suggest a specific binding partner of evodiamine. We observed that the effects of evodiamine were similar to those of GDC-0449, a Smo inhibitor, showing decrease of GLI1/2 transcriptional activity, MAPK phosphorylation, pro-inflammatory genes induction and MMP activation. Treatment of evodiamine, consequently, led to the inhibition of Hh signalling and photoageing phenotype. From these results, we assumed that Smo activity, at least in part, might be one of major targets affected by evodiamine. Accordingly, further structural and molecular studies of evodiamine and proteins in Hh signalling should be performed. Moreover, this will enable optimization of the capacity of target inhibition through structure–activity relationship analysis to advance the potency and specificity of evodiamine while minimizing possible adverse effects. To date, the exact molecular mechanism underlying Hh signalling in skin photoageing has been elusive. In the present study, we demonstrated that Hh signalling could be crucial for conferring UVB-induced photoageing (Fig. 4d). Here, we provide the first evidence of a novel mechanism of photoageing induced by Hh
References 1 Gilchrest B A. J Invest Dermatol 2013: 133: E2– E6. 2 Ichihashi M, Ando H. Exp Dermatol 2014: 23 (Suppl 1): 43–46. 3 Pandel R, Poljsak B, Godic A et al. ISRN Dermatol 2013: 2013: 930164. 4 Yaar M, Gilchrest B A. Br J Dermatol 2007: 157: 874–887. 5 Elia D, Madhala D, Ardon E et al. Biochim Biophys Acta 2007: 1773: 1438–1446. 6 Kim S R, Jung Y R, An H J et al. PLoS ONE 2013: 8: e73877. 7 Briscoe J, Therond P P. Nat Rev Mol Cell Biol 2013: 14: 416–429. 8 Harrison W, Cochrane B, Neill G et al. Oncogene 2014: 33: 2432–2440. 9 Katoh Y, Katoh M. Curr Mol Med 2009: 9: 873–886. 10 Pasca di Maglian M, Hebrok M. Nat Rev Cancer 2003: 3: 903–911. 11 Pantazi E, Gemenetzidis E, Trigiante G et al. Cell Death Dis 2014: 5: e1028. 12 Berman D M, Karhadkar S S, Maitra A et al. Nature 2003: 425: 846–851. 13 Yu H, Jin H, Gong W et al. Molecules 2013: 18: 1826–1843. 14 Liu Y N, Pan S L, Liao C H et al. Shock 2009: 32: 263–269. 15 Jiang J, Hu C. Molecules 2009: 14: 1852–1859. 16 Liao C H, Pan S L, Guh J H et al. Carcinogenesis 2005: 26: 968–975. 17 Nindl I, Dang C, Forschner T et al. Mol Cancer 2006: 5: 30.
signalling with inflammatory response and MAPK activation. We also demonstrated that evodiamine might be a potential inhibitor of Hh signalling, resulting in attenuation of UVB-induced photoageing. The results of this study indicated that regulation of Hh signalling with a pharmacologic reagent, evodiamine, could be a promising strategy for inhibiting inflammation and photoageing and decreasing skin cancer risk.
Acknowledgements This research was financially supported by the Ministry of Education (MOE) and National Research Foundation of Korea (NRF) through the Human Resource Training Project for Regional Innovation (2012H1B8A2026225) and supported by a 2-Year Research Grant of Pusan National University. The authors thank the Aging Tissue Bank for providing aged skin tissues.
Author contributions WK, YJ, CK and BY designed the study. WK, EK, HJY, TK, SH, SL and HY performed the research. WK, EG, TK, HY, YJ, CK and BY analysed the data. WK, EK and CK and BY wrote the manuscript. BY supervised the research.
Conflict of interest The authors have declared no conflicting interests.
Supporting Information Additional supporting data may be found in the supplementary information of this article. Data S1.Materials and methods. Figure S1. The mRNA levels of Shh and Ihh in skin tissues of aged SD rats were measured to determine whether Hh signalling is associated with the general ageing process. Figure S2. The inhibitory effects of Hh signalling by evodiamine were rescued by overexpression of GLI1 or GLI2. Table S1. Primers for determining the expression levels of Hh signalling- and inflammation-related genes. Table S2. Grading of UVB-induced skin wrinkles in a mouse model. Appendix S1. References.
18 Athar M, Li C, Tang X et al. Cancer Res 2004: 64: 7545–7552. 19 Brellier F, Marionnet C, Chevallier-Lagente O et al. Cancer Res 2004: 64: 2699–2704. 20 Epstein J H. J Am Acad Dermatol 1983: 9: 487– 502. 21 Mann J. Nat Rev Cancer 2002: 2: 143–148. 22 Kim W, Seong K M, Youn B. Exp Mol Med 2011: 43: 323–333. 23 Kim W, Youn H, Kwon T et al. Pharmacol Res 2013: 70: 90–101. 24 Kessaris N, Jamen F, Rubin L L et al. Development 2004: 131: 1289–1298. 25 Tang J Y, Aszterbaum M, Athar M et al. Cancer Prev Res (Phila) 2010: 3: 25–34. 26 Elsawa S F, Almada L L, Ziesmer S C et al. J Biol Chem 2011: 286: 21524–21534. 27 Ghorpade D S, Sinha A Y, Holla S et al. J Biol Chem 2013: 288: 33037–33048. 28 Saqui-Salces M, Coves-Datson E, Veniaminova N A et al. PLoS ONE 2012: 7: e48039. 29 Adachi H, Murakami Y, Tanaka H et al. Exp Dermatol 2014: 23 (Suppl 1): 32–36. 30 Aliouat-Denis C M, Dendouga N, Van den Wyngaert I et al. Mol Cancer Res 2005: 3: 627–634. 31 Mizutani T, Masaki H. Exp Dermatol 2014: 23 (Suppl 1): 23–26. 32 Teperino R, Aberger F, Esterbauer H et al. Semin Cell Dev Biol 2014: 33: 81–92. 33 Varjosalo M, Taipale J. Genes Dev 2008: 22: 2454–2472. 34 Borg M, Brincat S, Camilleri G et al. Climacteric 2013: 16: 514–521.
ª 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Experimental Dermatology, 2015, 24, 611–617
35 Li W H, Pappas A, Zhang L et al. J Dermatol Sci 2013: 71: 58–66. 36 Lu J T, Zhao W D, He W et al. Acta Pharmacol Sin 2012: 33: 691–700. 37 Atkinson P J, Dellovade T, Albers D et al. J Neurochem 2009: 108: 1539–1549. 38 Weston C R, Wong A, Hall J P et al. Genes Dev 2003: 17: 1271–1280. 39 Pratap A, Panakanti R, Yang N et al. Pharm Res 2010: 27: 2492–2504. 40 Whisenant T C, Ho D T, Benz R W et al. PLoS Comput Biol 2010: 6: e1000908. 41 Zaid M A, Afaq F, Syed D N et al. Photochem Photobiol 2007: 83: 882–888. 42 Kim M S, Oh G H, Kim M J et al. Photochem Photobiol 2013: 89: 911–918. 43 Lin T L, Matsui W. Onco Targets Ther 2012: 5: 47–58. 44 Von Hoff D D, LoRusso P M, Rudin C M et al. N Engl J Med 2009: 361: 1164–1172. 45 Kumari A, Ermilov A N, Allen B L et al. J Neurophysiol 2015: 113: 1034–1040. 46 Dahmane N, Lee J, Robins P et al. Nature 1997: 389: 876–881. 47 Tojo M, Kiyosawa H, Iwatsuki K et al. Br J Dermatol 2003: 148: 892–897. 48 Green J, Leigh I M, Poulsom R et al. Br J Dermatol 1998: 139: 911–915. 49 Kramann R, Schneider R K, DiRocco D P et al. Cell Stem Cell 2015: 16: 51–66. 50 Zhao H, Feng J, Seidel K et al. Cell Stem Cell 2014: 14: 160–173. 51 Heng M C. Int J Dermatol 2011: 50: 1058–1066.
617
