1. ^*Basic Research Laboratory, Kanebo Ltd., Odawara, Japan
2. ^†Department of Dermatology, Teikyo University of Medicine, Tokyo, Japan

Correspondence: Hidehiko Baba, Basic Research Laboratory, Kanebo Ltd, 5-3-28, Kotobukicho, Odawara, 250-0002 Kanagawa, Japan. Email: h.baba@oda.cos.kanebo.co.jp

Received 19 April 2004; Revised 6 June 2004; Accepted 24 June 2004.

To the Editor:

Stem cell factor (SCF) is a multifunctional growth factor that is critical for the survival and maturation of melanocytes, and which also affects dermal mast cell growth, accumulation and degranulation. It is one of the membrane-anchored growth factors that are processed to soluble forms by proteolytic cleavage, and which function both as soluble and as membrane molecules (^{Ashman, 1999}). Recently,^{Hachiya et al (2001)} reported that ultraviolet B (UVB) irradiation increases membrane-bound SCF content in human epidermis and in cultured keratinocytes, and they suggested the participation of this type of SCF in UVB-induced melanogenesis, in which this molecule would bind to the c-kit receptor on melanocytes, resulting in persistent c-kit activation. That study however, did not fully examine the relationship of soluble type SCF, as they could not detect it. On the other hand, one report showed that chronic UVB exposure increases the number of dermal mast cells, accompanied by increased epidermal SCF expression in albino hairless mice (^{Kligman and Murphy, 1996}), whereas another study showed that subcutaneous injection of recombinant SCF or transgene expression to produce soluble SCF in the epidermis causes cutaneous mast cell hyperplasia and cutaneous hyperpigmentation (^{Grichnik et al, 1995;Kunisada et al, 1998}). These observations led us to speculate that epidermal SCF also acts as a soluble factor in UV-exposed skin. Thus, in this study, we examined the possibility that stimulation of SCF release occurs in UVB-exposed cultured keratinocytes and in human epidermis.

Normal human epidermal keratinocytes were purchased from Kurabo (Osaka, Japan) and were cultured in MCDB153HAA medium (Kyokuto, Tokyo, Japan) supplemented with ethanolamine (0.1 mM), phosphoethanolamine (0.1 mM), hydrocortisone (0.5 nM), epidermal growth factor (0.1 ng per mL, Sigma, St Louis, Missouri), insulin (5 g per mL, Sigma), and bovine pituitary extract (BPE; 4 L per mL, Kurabo). At sub-confluence, the cells were transferred to starvation medium (without BPE) 24 h prior to UVB stimulation, which was carried out twice at a 24-h interval using FL20SE lamps (Toshiba, Tokyo, Japan). The culture media were harvested 24 h after the second UVB irradiation and were analyzed using an ELISA. As shown in Figure 1a, the amount of SCF in the culture medium was increased by UVB stimulation in a dose-dependent manner up to 20 mJ per cm², and then decreased at a dose of 40 mJ per cm² with an accompanying reduction of cell viability. Further, western blot analysis revealed that UVB irradiation caused an increase of the 31-kDa SCF protein band, which peaked at a dose of 20 mJ per cm² (Figure 1b). In addition, the molecular weight of that band corresponded to the cleaved form of full-length membrane-bound SCF. These results show that UVB irradiation at non-cytotoxic doses increases the release of SCF by cultured human keratinocytes. Our in vitro findings were further supported by an experiment using a living skin equivalent (MEL-300, Kurabo), in which UVB stimulation also resulted in increased SCF content in the culture medium (data not shown). In a previous study, performed under conditions that induced a significant reduction of cell viability by UVB irradiation, a UVB-induced upregulation of SCF release could not be detected (^{Hachiya et al, 2001}). Therefore, it seems that responses induced by UVB irradiation at non-cytotoxic levels were not considered in that study.

Figure 1.

UVB stimulation increases SCF release from cultured human keratinocytes. (a) Effects of UVB irradiation on SCF release and cell viability of keratinocytes. The bars and polygonal line show the SCF content in culture medium and cell viability, respectively. SCF content was determined using an ELISA (Quantikine, R&D Systems, Minneapolis, Minnesota). Cell viability was measured using an alamarBlue Assay (Alamar Biosciences, Sacramento, California) 24 h after the second UVB irradiation. Each value is presented as the meanSD of three determinations. ^**p<0.01, ^*p<0.05 as compared with the control by Dunnett's test. (b) Detection of soluble SCF by western blotting. Samples were concentrated 10-fold using a Centricon YM-10 (Millipore, Bedford, Massachusetts) and were separated by gel electrophoresis. Rabbit anti-SCF (Immuno-Biological Laboratories, Gunma, Japan) was used as a primary antibody.

Full figure and legend (35K)

Following approval from the Ethical Committee of Kanebo Basic Research Laboratory, we also administered UVB irradiation, which was performed with 2 MED of UVB (0.09–0.12 J per cm²; FL20SE lamps), to the upper-back area (2.0 cm 2.0 cm) of 10 healthy male Japanese volunteers who gave their written informed consent, after which analyses of SCF and c-kit expression were performed 3 and 10 days following the irradiation on skin biopsy specimens. Five millimeter-punch biopsies were taken from each exposed region following the experimental irradiation. As a control, non-exposed skin areas were also biopsied from all subjects. SCF was detected by immunohistochemistry throughout the epidermis in intact skin, as previously reported (Figure 2a:^{Longley et al, 1993;Weiss et al, 1995}). Three days after the irradiation, cytoplasmic staining of SCF was found to be reduced in epidermal keratinocytes, whereas membrane-associated SCF staining was clearly seen in the epidermal upper spinous and granular layers (Figure 2b). Thereafter, those levels returned to nearly normal by day 10 (Figure 2c). Further, a decreased expression of c-kit on melanocytes was observed on day 3 after UVB (Figure 2d–f).

Figure 2.

Changes of immunoreactivity toward SCF/c-kit in human epidermis following UVB exposure. Immunostaining for SCF (a–c) and c-kit (d–f) at pre-exposure (a, d), and 3 (b, e) and 10 days (c, f) after exposure. Acetone fixed-cryostat 5-m sections were incubated with polyclonal rabbit anti-SCF (Santa Cruz Biotechnology, California) or a polyclonal rabbit anti-c-kit (DAKO, Carpinteria, California). Immunolocalization of the primary antibody was performed using the avidin–biotin peroxidase complex method (ABC-Elite kit, Vector Laboratories, Burlingame, California), with nickel-diaminobenzidine as the substrate.

Full figure and legend (200K)

Changes in epidermal SCF-immunostaining patterns to reduced cytoplasmic staining and membrane-associated staining have been reported in lesional urticaria pigmentosa (UP), a skin disease characterized by dermal accumulations of mast cells and increased epidermal melanin. Such changes are considered to result from the altered metabolism of SCF, which induces an elevation of soluble SCF in the skin (^{Longley et al, 1993;Weiss et al, 1995}). On the other hand, it has been reported that the c-kit receptor is internalized and degraded after binding its soluble ligands (^{Shimizu et al, 1996}), and our experiment showed decreased c-kit expression on day 3 (Figure 2e). Therefore, altered SCF immunoreactivity observed on day 3 might be the consequence of enhanced SCF liberation in the epidermis.

From reports that serum SCF levels are elevated in patients with various skin diseases (e.g. atopic dermatitis, scleroderma) and that those levels recover with clinical improvement (^{Kanbe et al, 2001;Yamamoto et al, 2001}), there is every probability that SCF shedding occurs in cutaneous tissues. Furthermore, it has been revealed that several proteases are involved in the release of SCF (^{Pandiella et al, 1992;Gallea-Robache et al, 1997;Longley et al, 1997;Heissig et al, 2002}), and the profiles of some of these enzymes in the skin are altered by UV exposure (^{Fisher et al, 1997}). Thus, it seems that UV irradiation of the skin establishes a condition in which SCF shedding occurs.

In conclusion, SCF release from cultured epidermal cells was increased by UVB radiation, which also altered the SCF/c-kit expression pattern in human epidermis. These results suggest that SCF/c-kit signaling in UV-exposed skin occurs through soluble SCF, as well as through the membrane-bound form.