| To whom crrespondence should be addressed: Ryo Iwamoto, Department of Cell Biology, Research Institute for Microbial Diseases, Osaka University, 3-1, Yamadaoka, Suita, Osaka 565-0871, Japan. Tel: +81–6–6879–8288, Fax: +81–6–6879–8289 E-mail: riwamoto@biken.osaka-u.ac.jp Abbreviations: EGF, epidermal growth factor; HB-EGF, heparin-binding EGF-like growth factor; sHB-EGF, soluble form of HB-EGF; proHB-EGF, membrane-anchored form of HB-EGF; EGFR, EGF receptor; tRA, all-trans retinoic acid. |
Heparin-binding EGF-like growth factor (HB-EGF) is a member of the EGF-family of growth factors, including EGF, TGF-α, amphiregulin (AR), epiregulin (EPR), betacellulin (BTC), and epigen (Harris et al., 2003). HB-EGF binds to and activates EGF receptor (EGFR/ErbB1) (Higashiyama et al., 1991) and ErbB4 (Elenius et al., 1997). Theoretically, HB-EGF can activate not only EGFR and ErbB4 but also ErbB2 and ErbB3 since each ErbB receptor can form heterodimers following ligand binding (Holbro and Hynes, 2004). Like other EGF-family members (Massague and Pandiella, 1993), HB-EGF is first synthesized as a membrane-anchored form (proHB-EGF), and then cleaved at the juxtamembrane domain, resulting in the shedding of soluble HB-EGF (sHB-EGF) (Goishi et al., 1995). sHB-EGF is a potent mitogen and chemoattractant for a number of cell types including vascular smooth muscle cells, fibroblasts and keratinocytes (Higashiyama et al., 1993; Raab and Klagsbrun, 1997). ProHB-EGF forms a complex with other membrane proteins at the cell-cell contact site (Nakamura et al., 1995) and can transduce biological signals to neighboring cells in a non-diffusible manner (Iwamoto and Mekada, 2000). Thus, proHB-EGF is also thought to be a biologically active molecule. HB-EGF has been implicated in a number of physiological and pathological processes (Raab and Klagsbrun, 1997). Studies using mice lacking the HB-EGF gene or expressing a mutant form of HB-EGF revealed that HB-EGF is involved in heart development and function (Iwamoto et al., 2003; Yamazaki et al., 2003), normal lung morphogenesis (Jackson et al., 2003), eyelid development (Mine et al., 2005) and wound healing processes in the skin (Shirakata et al., 2005).
Retinoids, the derivatives of vitamin A, are important regulators of epithelial cell homeostasis. Retinoids are integral to cutaneous cellular proliferation and differentiation as evidenced by the characteristic hyperkeratosis observed in vitamin A deficiency (Miller, 1989). All-trans retinoic acid (tRA) is the major biologically active metabolite of vitamin A. tRA and related synthetic retinoids have been used widely in the therapy of several skin disorders, including acne (Pedace and Stoughton, 1971), photo-aged skin (Kligman et al., 1984), and psoriasis (Goldfarb et al., 1988). However, the effects of retinoids are often conflicting: retinoids can either stimulate or inhibit cell proliferation depending on the cell type affected or on the conditions used in cell cultivation (Amos and Lotan, 1990; Fisher and Voorhees, 1996; Rogers, 1997; Varani et al., 1989, 1991). Although the mechanism by which retinoids regulate the proliferation of epidermal cells is not completely understood, it has been shown that treatment of normal adult human skin and keratinocytes with tRA not only leads to thickening of the skin but also induces significant expression of several genes including HB-EGF (Bernard et al., 2002; Stoll and Elder, 1998; Varani et al., 2001). Induction of HB-EGF has also been reported in the skin of mice treated topically with tRA (Chapellier et al., 2002; Xiao et al., 1999). These studies suggested that HB-EGF might be involved in the signaling cascades underlying retinoid-induced epidermal hyperplasia.
Regarding the physiological function of HB-EGF in skin epidermis, we have shown that HB-EGF promotes keratinocyte migration but not proliferation during the eyelid development (Mine et al., 2005) and the skin wound-healing process (Shirakata et al., 2005). Thus, it has remained unclear whether HB-EGF functions in keratinocyte proliferation in vivo when skin is treated with tRA, and which forms of HB-EGF (the membrane-anchored form or the soluble form) contribute to this process, if HB-EGF is involved. To address these issues, we investigated the process of tRA-induced epidermal hyperplasia using three types of HB-EGF mutant mice: keratinocyte-specific HB-EGF null mice (K5-HBdel/del) (Shirakata et al., 2005), knock-in mice expressing the uncleavable form of HB-EGF (HBuc/uc) (Yamazaki et al., 2003), and knock-in mice expressing wild-type HB-EGF as a wild-type control (Iwamoto et al., 2003). Our results indicate that HB-EGF in its soluble form plays a pivotal role in tRA-induced epidermal hyperplasia. We also suggest that both EGFR and ErbB2 contribute to the proliferation of basal layer keratinocytes, which are stimulated by tRA-induced HB-EGF.
Generation and maintenance of HB-EGF null mice (HBdel/del), wild-type HB-EGF cDNA knock-in mice (HBlox/lox) (used as an experimental wild-type control), keratinocyte-specific HB-EGF null mice (K5-HBdel/del), and knock-in mice expressing the uncleavable mutant form of HB-EGF (HBuc/uc) were carried out as described previously (Iwamoto et al., 2003; Shirakata et al., 2005; Yamazaki et al., 2003).
Adult (8–12 week-old) female mice were used for all experiments. Prior to topical tRA treatment, hair was removed using an electric clipper under pentobarbital-anaesthetization. A single dose of tRA (200 nmol) was diluted with 400 μl of acetone just prior to use, and was applied daily to the entire back skin.
For hematoxylin-eosin staining, skin samples were fixed with 4% paraformaldehyde (PFA), dehydrated and embedded in paraffin. Four-μm sections were stained with hematoxylin-eosin. For lacZ staining, the tissues were fixed with 4% PFA for 2 h, and then stained with 5-bromo-4-chloro-3-indolyl β-D-galactoside (X-gal). The stained tissues were fixed again 4% PFA, embedded in paraffin, sectioned at 4 μm and stained with nuclear fast red. For Ki67 staining, sections were treated with 10 mM sodium citrate buffer (pH 6.0) and heated to unmask the Ki67 antigen, and were incubated for 1 h at room temperature with an antibody against Ki67 (Novocastra) diluted 1:1000 in PBS. Biotinylated anti-rabbit IgG (Vector) diluted 1:500 in PBS was used as secondary antibody. Ki67 immunolocalization was performed using Vectastain Elite ABC and peroxidase substrate DAB kits (Vector). To determine the number of keratinocytes per unit length of epidermis, 5 micrographs from 5 independent sections were taken at high-power (×400) for each mouse skin. The number of keratinocytes was then counted on the 5 micrographs and pooled. The micrograph of the same 5 fields was also used to estimate the thickness of epidermis with Adobe Photoshop 6.0 software. Means±SD were calculated from skin samples of at least 5 animals of each genotype.
RNA was isolated from skin samples using SV Total RNA Isolation System (Promega) according to the manufacturer’s protocol. Reverse transcription was performed using ReverTra Ace (Toyobo) and random primers. PCR amplifications were carried out in a final volume of 25 μl containing an aliquot of the cDNA, 0.5 μM primers and 0.6U KOD Dash (Toyobo). The reaction mixture was kept at 95°C for 2 min for denaturation, followed by 26 amplification cycles for HB-EGF or 22 cycles for β-actin. Each cycle consisted of denaturation at 95°C for 10 sec, primer annealing at 64°C for 30 sec and extension at 72°C for 30 sec. The following primer pairs were used: for HB-EGF, 5'-GGTATCTGCACTCCCCGTGGATGC-3' and 5'-ATGAAGCTGCTGCCGTCGGTGATGCTGA-3'; and for β-actin, 5'-GTGGGCCGCTCTAGGCACCA-3' and 5'-CGGTTGGCCTTAGGGTTCAGGGGG-3'. Quantitative PCR was performed in a real-time format on the 7700 ABI Prism Sequence Detector (Applied Biosystems). Gene specific primers and probes were purchased as Assays-on-Demand Gene Expression Products (Applied Biosystems). The assay IDs of HB-EGF, AR and TGF-α specific primers and probe were Mm00439307_m1, Mm00437583_m1 and Mm00446231_m1, respectively. 18S rRNA primers and probe were used to normalize the input of each cDNA sample. An aliquot of 5 μl of the reverse transcribed cDNA was added to 20 μl of reaction mixture containing 12.5 μl of Platinum Quantitative PCR SuperMix-UDG (Invitrogen), ROX Reference Dye (Invitrogen) and 1X Primers and Probe. The TaqMan cycling conditions consisted of 2 min degradation of the preamplified templates at 50°C and then 40 cycles of denaturation at 95°C for 15 sec, followed by annealing and extension at 58°C for 60 sec.
For the detection of total ErbBs and phosphorylated ErbBs, back skin from 11 week-old wild-type (C57BL6/J) or HBdel/del mice were treated with 200 nmol of tRA for indicated periods. The isolated full back skin was homogenized in lysis buffer (Iwamoto et al., 2003). For detection of total ErbB1-4 proteins and phospho-ErbB1-4 proteins in wild-type skin (Fig. 5), the skin lysate was centrifuged at 20,000×g for 30 min, then each ErbB was immunoprecipitated from the supernatant. The precipitated ErbB and phospho-ErbB were detected by immunoblotting using anti-phosphotyrosine antibody as described previously (Iwamoto et al., 2003). For detection of total ErbB1, 2 proteins and phospho-ErbB1, 2 proteins in wild-type and HBdel/del skin (Fig. 6), the skin lysate was centrifuged at 20,000×g for 30 min, and then the supernatant was subjected to immunoblotting using anti-EGFR or anti-ErbB2 antibody (Santa Cruz Biotechnology), or anti-phospho-EGFR or anti-phospho-ErbB2 antibody (Cell Signaling Technology), respectively.
Topical application of tRA induced a marked increase in epidermal thickness (Fig. 1). In addition, the number of Ki67-positive cells increased markedly (in parallel with skin thickening) in the tRA-treated skin (Fig. 1). These proliferating cells were predominantly localized in the basal layer epidermis, as previously reported (Varani et al., 2001).
 View Details | Fig. 1. Epidermal hyperplasia induced by topical tRA treatment in mouse skin. Normal adult mouse skins were treated for 4 days with vehicle (Veh) or tRA (tRA). Skin sections were stained with hematoxylin and eosin (H & E) or with an antibody to Ki67 (Ki67) to detect proliferating cells. Scale bar: 50 μm. |
In situ hybridization analysis indicated that topical tRA treatment of normal mouse skin induces HB-EGF mRNA expression exclusively in suprabasal keratinocytes (Xiao et al., 1999). To confirm this, we analyzed the expression of HB-EGF in this process using a different technique. For the HB-EGF expression analysis, we used the keratinocyte-specific heterozygotic (HBdel/+/K5-Cre (K5-HBdel/+)) HB-EGF knockout mice (Shirakata et al., 2005). The targeting vector contains wild-type HB-EGF cDNA flanked by loxP sites linked to the LacZ gene as a reporter for the expression of HB-EGF, which is driven by the native HB-EGF promoter when HB-EGF cDNA is deleted by Cre recombinase (Iwamoto et al., 2003). Thus, it is possible in K5-HBdel/+ mice (and also in K5-HBdel/del mice) to identify the expression of HB-EGF in the epidermal keratinocytes by staining for LacZ. The LacZ construct contains the nuclear targeting signal (NTR-LacZ) and staining the nuclei with X-Gal can readily identify cells expressing HB-EGF.
As shown in Fig. 2a, in untreated adult skin, HB-EGF expression was not detectable. However, in tRA-treated skin, HB-EGF expression was induced predominantly in the suprabasal keratinocytes. HB-EGF expression was apparent by 24 h following tRA treatment and the expression gradually increased over the subsequent days (Fig. 2b). The EGFR-ligands including EGF (Weissman and Aaronson, 1983), TGF-α (Coffey et al., 1987), HB-EGF (Hashimoto et al., 1994) and AR (Cook et al., 1991) are known to stimulate keratinocyte proliferation. Among them, EGF expression has been reported to be unaffected by tRA treatment (Xiao et al., 1999). Thus, to investigate the specific role of HB-EGF in tRA-induced epidermal hyperplasia, we examined the induction of TGF-α, AR, and HB-EGF mRNA in tRA-treated skin. tRA was topically applied to the back skin and the total RNA was harvested at several time points. The expression of each EGFR ligand was analyzed by quantitative RT-PCR. HB-EGF mRNA was rapidly induced after tRA treatment, reaching a peak of 5-fold induction at 24 h, whereas AR and TGF-α mRNAs were weakly induced, with a maximum increase of 1.3-fold and 3-fold even at 72 h, respectively (Fig. 2c). These results suggest that HB-EGF is one of the major inducible genes of the EGFR-ligands examined in tRA-treated epidermis.
 View Details | Fig. 2. Expression of HB-EGF mRNA induced by tRA. (a) HB-EGF expression in the suprabasal layer of tRA-treated skin. K5-HBdel/+ mouse skins were treated for 4 days with vehicle (Veh) or tRA (tRA). Skin sections were stained with X-gal (LacZ). Scale bar: 50 μm. (b) Time course of induction of HB-EGF expression detected by LacZ staining. K5-HBdel/+ mouse skins were untreated (day 0) or treated for 24 h (day 1), 48 h (day 2), or 72 h (day 3) with tRA. Skin sections were stained with X-gal. Scale bar: 100 μm. (c) The quantitative RT-PCR analysis of the induction of HB-EGF, amphiregulin and TGF-α mRNA. Normal adult mouse skins were treated with tRA. The y-axis represents relative mRNA levels expressed as fold induction over basal levels in vehicle-treated mice. The x-axis shows the time of sampling after tRA application. Data are mean +/– SD of trip-licate measurements and were normalized against the 18S rRNA signal. p-values were determined by Student’s t-test HB-EGF versus Amphiregulin and TGF-α independently. Asterisks show statistically significant points in both cases (p<0.05). |
To examine directly the contribution of HB-EGF in tRA-induced skin hyperplasia, we compared the rate of hyperplasia between K5-HBdel/del mice and the HBlox/lox mice after tRA treatment. We confirmed that tRA treatment induced HB-EGF mRNA expression in HBlox/lox skin but not in K5-HBdel/del skin (Fig. 3).
 View Details | Fig. 3. Expression of HB-EGF mRNA induced by tRA in HB-EGF mutant mice. Wild-type HBlox/lox mice (lox/lox), keratinocyte-specific HB-EGF null K5-HBdel/del mice (K5-del/del) and uncleavable mutant HBuc/uc mice (uc/uc) were treated for 4 days with vehicle (–) or tRA (+). HB-EGF and β-actin mRNA expression were analyzed by RT-PCR. Disruption of HB-EGF mRNA expression in HB-EGF null skin, and normal HB-EGF mRNA expression in HBuc/uc mice were confirmed. |
In HBlox/lox skin, tRA treatment induced overt hyperplasia, while in K5- HBdel/del skin the hyperplasia was significantly reduced (Fig. 4a; compare ‘lox/lox’ and ‘K5-del/del’). To analyze these observations more quantitatively, the hyperplasia was assessed by measuring the epidermal thickness and the keratinocyte cell number (Fig. 4b). The increase in skin thickness of HBlox/lox skin treated with tRA was approximately 4.1-fold, as compared with the vehicle-treated skin, whereas that of the K5-HBdel/del skin demonstrated an approximately 2.8-fold increase. Thus, the rate of hyperplasia measured by the epidermal thickness in K5-HBdel/del skin was 68% of the hyperplasia measured in tRA-treated control skins. The keratinocyte cell number of HBlox/lox skin treated with tRA was increased approximately 2.9-fold, as compared with the vehicle-treated skin, whereas that of the K5-HBdel/del skin demonstrated an increase of approximately 2.5-fold. Thus, the rate of hyperplasia measured by the cell number in K5-HBdel/del skin was 86% of the hyperplasia measured in tRA-treated control skins. These results indicate that HB-EGF is significantly involved in the skin hyperplasia induced by tRA treatment.
 View Details | Fig. 4. Reduced epidermal hyperplasia in HB-EGF mutant mice. (a) Histological views of three types of HB-EGF mutant mice treated with tRA. Wild-type HBlox/lox mice (lox/lox), keratinocyte-specific HB-EGF null K5-HBdel/del mice (K5-del/del) and uncleavable mutant HBuc/uc mice (uc/uc) were treated for 4 days with vehicle (Veh) or tRA (tRA). Skin sections were stained with hematoxylin and eosin. Scale bar: 50 μm. (b) Quantitative analysis of epidermal hyperplasia assessed by measuring the epidermal thickness (left panel) and the keratinocyte cell number (right panel). The number of keratinocytes in epidermis was counted in a unit area (230 μm×290 μm). Values from vehicle-treated mice (open bar) and tRA-treated mice (solid bar) represent mean +/– SD (n>5). *p-values<0.0001, **p-values<0.005 |
HB-EGF is first synthesized as a membrane-anchored form (proHB-EGF), and then the soluble form (sHB-EGF) is released from the cell surface by ectodomain shedding (Goishi et al., 1995). Moreover, cell culture studies showed that proHB-EGF is biologically active as a juxtacrine growth factor that signals to neighboring cells in a non-diffusible manner (Iwamoto and Mekada, 2000). Thus, to investigate which form of HB-EGF is involved in tRA-induced skin hyperplasia, we analyzed this process by using knock-in mice that express an uncleavable mutant form of proHB-EGF (HBuc) (Yamazaki et al., 2003). We confirmed that tRA treatment induced HB-EGF mRNA expression in HBuc/uc skin, as it did in HBlox/lox skin (Fig. 3), and that any soluble form of HB-EGF proteins was not detected in tRA-treated mutant skins (Yamazaki et al., 2003).
As observed in K5-HBdel/del skin, tRA-induced hyperplasia was significantly reduced in HBuc/uc skin as compared with the HBlox/lox skin (Fig. 4a; compare ‘lox/lox’ and ‘uc/uc’). The thickness of HBlox/lox skin treated with tRA was increased approximately 4.1-fold, in comparison to that of vehicle-treated, whereas that of HBuc/uc skin was increased approximately 2.5-fold (Fig. 4b). Thus, the rate of hyperplasia measured by the epidermal thickness in HBuc/uc skin was 61% of the hyperplasia measured in tRA-treated control skins. The keratinocyte cell number of HBlox/lox skin treated with tRA was increased by approximately 2.9-fold, as compared with vehicle-treated skin, whereas that of the HBuc/uc skin was increased by approximately 2.2-fold (Fig. 4b). Thus, the rate of hyperplasia measured by the cell number in HBuc/uc skin was 76% of the hyperplasia measured in tRA-treated control skins. These results indicate that ectodomain shedding of proHB-EGF is required in skin hyperplasia induced by tRA treatment, and that sHB-EGF but not proHB-EGF functions in this process.
EGFR/ErbB1 is the receptor for HB-EGF (Higashiyama et al., 1991) and it is thought to be important for keratinocyte proliferation, differentiation, and migration (Jost et al., 2000). EGFR can function not only in the homodimer form, but also in the heterodimer one with ErbB2, ErbB3, or ErbB4 (Holbro and Hynes, 2004). To investigate which ErbB is activated in tRA-induced hyperplasia, we examined the tyrosine phosphorylation level of each ErbB receptor in tRA-treated or vehicle-treated wild-type skins (Fig. 5). Tyrosine phosphorylation of EGFR (ErbB1) and ErbB2 was detected at 2 to 72 h after treatment with tRA, while phosphorylated EGFR and ErbB2 were not detected in vehicle-treated skin. Phosphorylated ErbB3 was constitutively detected in both vehicle- and tRA-treated skin. Phosphorylated ErbB4 was not detected, consistent with the previous report that ErbB4 protein is either not present or is present at undetectable levels in the skin (Kiguchi et al., 2000; Plowman et al., 1993; Xian et al., 1997). These results indicate that EGFR and ErbB2 are selectively activated in tRA-induced skin hyperplasia.
 View Details | Fig. 5. Tyrosine phosphorylation of the ErbB receptor family in normal skin induced by tRA. Normal adult mouse skins were untreated (–) or treated for the indicated time with vehicle or tRA. After each incubation period, the skin was lysed and each ErbB receptor was immunoprecipitated from lysates with its corresponding antibody. The immunoprecipitates were analyzed by Western blotting using anti-phosphotyrosine antibody or each of the four corresponding anti-ErbB antibodies. |
To determine whether tRA-induced EGFR and ErbB2 tyrosine phosphorylation is reduced in HB-EGF null mice, we compared the tyrosine phosphorylation level of EGFR and ErbB2 induced by tRA between wild-type and HBdel/del skins. As shown in Fig. 6, phosphorylation of both EGFR and ErbB2 was drastically reduced in HBdel/del skin. These results provide evidence that HB-EGF transduces its signal by means of EGFR and ErbB2 activation, resulting in the promotion of keratinocyte proliferation in tRA-stimulated epidermal hyperplasia.
 View Details | Fig. 6. Reduced tyrosine phosphorylation of the EGFR and ErbB2 in tRA-treated HB-EGF null skin. Wild-type mice (+/+) and HB-EGF null (del/del) mice (two animals were independently used for each treatment) were treated with vehicle or tRA. Twenty-four hours after treatment, the skin was lysed, and total EGFR (ErbB1) and ErbB2, and phospho-EGFR and phospho-ErbB2 in the lysate was analyzed by Western blotting using anti-EGFR and anti-ErbB2 antibody, or anti-phospho-EGFR and anti-phospho-ErbB2 antibody, respectively. |
Epidermal hyperplasia is the most prominent histological change in the skin observed after topical treatment with tRA. Previous studies have suggested that HB-EGF could be involved in the proliferation of keratinocytes following retinoid application using in vitro and ex vivo experimental systems (Chapellier et al., 2002; Stoll and Elder, 1998; Varani et al., 2001; Xiao et al., 1999). In the present study, utilizing three types of mutant mice, HBlox/lox, K5-HBdel/del and HBuc/uc, we present direct evidence indicating the substantial role of secreted HB-EGF in the process of tRA-induced epidermal hyperplasia: 1) topical tRA treatment resulted in prominent expression of HB-EGF in the suprabasal layer of thickened skin, 2) in HB-EGF null skins, tRA-induced hyperplasia was significantly reduced, 3) in the skin of mice expressing an uncleavable HB-EGF, hyperplasia was reduced to a level seen in the skin of HB-EGF null mice. Moreover, the present study demonstrated that tRA treatment resulted in phosphorylation of both EGFR and ErbB2 in wild-type skins, but that phosphorylation was significantly reduced in HB-EGF null skins, suggesting that the expression and ectodomain shedding of proHB-EGF, which generate a soluble form of HB-EGF, followed by HB-EGF-dependent activation of EGFR and ErbB2, are pivotal events contributing to tRA-induced epidermal hyperplasia.
EGFR is implicated as a critical factor in keratinocyte functions including proliferation, differentiation, and migration; however, the relationship between retinoid action and the function of other ErbB receptors in epidermal hyperplasia are not fully understood. It has been reported that antagonists for pan-ErbB tyrosine kinase receptors inhibited tRA-induced hyperplasia in organ-cultured human skin and proliferation of human epidermal keratinocytes in monolayer culture systems (Varani et al., 2001). In this study, we first demonstrate that tRA rapidly and specifically induced autophosphorylation of both EGFR and ErbB2 in normal skin; however, this was significantly reduced in HB-EGF null skins, which indicates that HB-EGF activates both EGFR and ErbB2 in tRA-induced epidermal hyperplasia. HB-EGF does not bind ErbB2 or ErbB3. ErbB2 does not bind directly to any EGF-family ligands (Holbro and Hynes, 2004). ErbB4, an additional binding receptor for HB-EGF and heteropartner of ErbB2 activation, was not detected in mice skin as reported previously (Kiguchi et al., 2000; Plowman et al., 1993; Xian et al., 1997). These findings suggest the possibility that HB-EGF activates the heterodimer form of EGFR and ErbB2 in this process. However, it has been reported that immunostaining for ErbB2 and ErbB3 was accentuated in the upper spinous layers, and that EGFR was strongly labeled in basal cells in human skin epidermis (Piepkorn et al., 2003). This suggests an alternative possibility that HB-EGF first activates EGFR in the basal cells, resulting in induction of other EGF-family ligands, such as neuregulins, that can activate ErbB2/ErbB3 heterodimer in the upper spinous layers.
Confirming the findings of a previous study (Xiao et al., 1999), in the present study we clearly showed that HB-EGF is expressed exclusively in the suprabasal keratinocytes in tRA-induced epidermal hyperplasia. Moreover, EGFR and ErbB2 were activated by tRA treatment, in an HB-EGF-dependent manner. EGFR localizes mainly in the basal keratinocytes (King et al., 1990; Vassar and Fuchs, 1991), and the cells that predominantly proliferate in epidermis localize in the basal layer. These results imply that HB-EGF, expressed in the suprabasal layer, has to translocate to the basal layer of epidermis in order to promote EGFR activation and proliferation of basal layer keratinocytes. Using HBuc/uc mice, we observed that the epidermal hyperplasia was inhibited to a level equivalent to that seen in HB-EGF null mice, indicating that HB-EGF functions in this process in the soluble form, and not the membrane-anchored form. Indeed, we have previously shown that sHB-EGF is produced in tRA-treated epidermis (Yamazaki et al., 2003). In addition, our preliminary experiment demonstrated that the phosphorylation level of EGFR and ErbB2 was dramatically reduced in HBuc/uc mice upon tRA-treatment (data not shown). These results suggest the following model of HB-EGF action in tRA-induced epidermal hyperplasia: 1) topical tRA treatment of skin induces the expression of proHB-EGF in the suprabasal layer of keratinocytes, 2) proHB-EGF is processed by an ectodomain shedding mechanism, resulting in the release of sHB-EGF, 3) sHB-EGF translocates from the suprabasal layer to the basal layer epidermis, and 4) translocated sHB-EGF activates EGFR localized in the basal layer and promotes proliferation of the basal layer keratinocytes, resulting in epidermal hyperplasia.
In skin epidermis, keratinocytes are stratified and the cells contact each other laterally via tight junctions (Tsukita and Furuse, 2002). How can sHB-EGF translocate across these cell-cell junctions? It has been reported, both in cultured keratinocytes and in the skin, that retinoids cause fragility of this cell-cell contact through decreased numbers of tonofilaments in keratinocytes and desmosomal attachments in the epidermis (Hatakeyama et al., 2004; Peck et al., 1977; Wanner et al., 1999; Williams et al., 1981). This suggests that tRA treatment induces epidermal fragility, and as a result, sHB-EGF might be able to translocate through gaps between the keratinocytes to the basal layer.
Phosphorylation levels of EGFR and ErbB2 remained almost at a basal level in tRA-treated HB-EGF null skins, suggesting that HB-EGF is a major factor for tRA-induced EGFR/ErbB2 activation. On the other hand, keratinocyte-specific HB-EGF null mice clearly showed a significant reduction of epidermal hyperplasia following topical tRA treatment; however, the defect was incomplete. These findings suggest that factor(s) other than HB-EGF might be involved in tRA-induced keratinocyte proliferation after activation of EGFR and ErbB2. Although HB-EGF mRNA expression was predominantly induced among the EGFR-ligands tested here (HB-EGF, AR, and TGF-α), other EGF family ligands that were not examined (EGF, EPR, BTC, epigen, and neuregulins) might also be involved in this process. As has been reported previously (Hashimoto et al., 1994), HB-EGF may act as an autocrine growth factor, which induces other EGF-family growth factors in tRA-induced keratinocyte proliferation. Partial reduction of tRA-induced hyperplasia in HB-EGF null skin also suggests that other signaling pathway(s), distinct from the EGF family-ErbB axis pathway, might be involved in this process. Consistent with this view, it has been reported that tRA does not increase HB-EGF expression in RARγ–/– and RARγsb–/– mice, whereas it still exerts a degree of proliferative effect on the epidermis of these mice (Chapellier et al., 2002). Further, it has been shown that expression of keratinocyte growth factor is stimulated by tRA in cultured gingival fibroblasts (Mackenzie and Gao, 2001). Further studies are necessary to reveal the whole molecular mechanism involved in epidermal cell proliferation regulated by retinoids.
We thank I. Ishimatsu, M. Hamaoka, and T. Yoneda for technical assistance, and also thank S. Adachi for support of animal maintenance. This work was supported in part by grants-in-aid from the Ministry of Education, Culture, Sports, Science, and Technology (14580696 to R. Iwamoto, and 16207014 to E. Mekada).
|
Index