|Home | About | Journals | Submit | Contact Us | Français|
Dickkopf 1 (DKK1), an inhibitor of Wnt signaling, not only functions as a head inducer during development, but also regulates joint remodeling and bone formation, which suggests roles for DKK1 in the pathogenesis of rheumatoid arthritis and multiple myeloma. We recently demonstrated that levels of DKK1 in palmoplantar dermal fibroblasts are physiologically higher than those observed in non-palmoplantar dermal fibroblasts. Thus, the DKK1-rich mesenchyme in palmoplantar dermis affects the overlying epithelium and induces a palmoplantar phenotype in the epidermis. More specifically, DKK1 suppresses melanocyte function and growth via the regulation of microphthalmia-associated transcription factor (MITF) and β-catenin. Furthermore, DKK1 induces the expression of keratin 9 and α-Kelch-like ECT2 interacting protein (αKLEIP) but down-regulates the expression of β-catenin, glycogen synthase kinase 3β, protein kinase C and proteinase-activated receptor-2 (PAR-2) in keratinocytes. Treatment of reconstructed skin with DKK1 reproduces the hypopigmentation and thickening of skin via Wnt/β-catenin signaling. These studies elucidate why human palmoplantar skin is thicker and paler than non-palmoplantar skin via the secretion of DKK1 by fibroblasts that affect the overlying epidermis. Thus, DKK1 may be useful for reducing skin pigmentation and for thickening photo-aged skin and palmoplantar wounds caused by diabetes mellitus and rheumatic skin diseases.
Our research groups have performed a series of collaborative studies elucidating how human palmoplantar skin is regenerated using non-palmoplantar skin. Our group reported that human palmoplantar skin defects can be treated with non-palmoplantar epidermis (Yamaguchi et al, 1999), which is easily obtained from donors without forming scars (Yamaguchi et al, 2005a). Thin and fragile non-palmoplantar epidermis can differentiate into mechanically strong palmoplantar epidermis due to the effects of the underlying palmoplantar dermis (Hanafusa et al, 2007; Yamaguchi et al, 2001; Yamaguchi et al, 2005b; Yamaguchi et al, 2004b; Yamaguchi and Yoshikawa, 2001), which is the first clinically proven site-specific regeneration observed for a human tissue. It is paradoxical that artificial wound closures using skin flaps and conventional skin grafts inhibit ideal regeneration, even in salamanders, since those operations interfere with formation of the wound epidermis (Muneoka et al, 2008). We hypothesized that wound coverage with only the epidermal component would not interfere with remodeling of the underlying microenvironment but nonetheless would enhance appropriate differentiation. More specifically, we then established that adult human palmoplantar dermis at least in part determines the fate of the overlying epidermis.
These studies indicated that factors regulated by the palmoplantar dermis can induce the palmoplantar phenotype in grafted non-palmoplantar epidermis. Chang et al. recently reported that HOX gene expression patterns observed in the fetus are retained in adult human fibroblasts using microarray analyses of various sites of cadaver skin, which suggested that topographically different fibroblasts play an essential role in regulating the positional identity of the epidermis, even in adult tissues (Chang et al, 2002). That group subsequently proved that noncoding RNAs obtained from fibroblasts regulate the expression patterns of HOX transcription factors with site specificity (Rinn et al, 2007) and that HOXA13, which is highly expressed in mouse palmoplantar fibroblasts, induces palmoplantar epidermis via the enhancement of Wnt5a (Rinn et al, 2008). Yasuda et al. reported that expression levels of fibronectin are lower in palmoplantar fibroblasts, which may explain why palmoplantar epidermis differs from non-palmoplantar epidermis (Yasuda et al, 2006).
Our group identified dickkopf 1 (DKK1), an inhibitor of the Wnt signaling pathway and a head inducer during development (Glinka et al, 1998), as an abundantly expressed factor in palmoplantar fibroblasts at mRNA (Yamaguchi et al, 2004a) and at protein levels (Yamaguchi et al, 2007b). DKK1 plays key roles in the development of osteolytic lesions in multiple myeloma (Tian et al, 2003) and in the bone-destructive pattern observed in rheumatoid arthritis through tumor necrosis factor (TNF)-α (Diarra et al, 2007). Here we review briefly the effects of DKK1 on skin pigmentation and thickness by inducing the palmoplantar phenotype.
Melanocytes regulate skin pigmentation (Yamaguchi et al, 2007a), and they are affected by external factors such as ultraviolet radiation (Yamaguchi et al, 2006) and by internal factors secreted from fibroblasts and keratinocytes (Yamaguchi and Hearing, 2006). We investigated the effects of DKK1 secreted by palmoplantar fibroblasts on melanocyte growth and differentiation (Yamaguchi et al., 2004a). DKK1 (at 100 ng/ml) suppressed the growth of normal human melanocytes and their function, as measured by expression levels of melanogenic proteins, including tyrosinase, dopachrome tautomerase, GP100/Pmel17, MART1 and MITF (microphthalmia-associated transcription factor). These effects were mediated by the Wnt/β-catenin/MITF signaling pathway, since incubation of melanocytes with DKK1 down-regulated the expression of β-catenin and glycogen synthase kinase (GSK) 3β and up-regulated the expression of Ser9-phosphorylated GSK3β and protein kinase Cα.
We performed microarray analyses and compared gene expression patterns in normal human melanocytes treated with or without DKK1 (Yamaguchi et al., 2007b). The results (http://www.ncbi.nlm.nih.gov/geo/ GSE5515) indicated involvement of apoptosis pathways, including TNF-α and Gadd45, and of melanocyte receptors for DKK1 other than LRP5/6 and Kremen1/2, which explains the suppression of growth and of melanogenesis.
Melanin synthesized in melanosomes within melanocytes is transferred to keratinocytes through the protease-activated receptor (PAR)-2 (Seiberg et al, 2000). We investigated the effects of DKK1 on melanosome transfer and on the expression of PAR-2 in normal human keratinocytes (Yamaguchi et al, 2008). DKK1-transfected keratinocytes and DKK1-treated reconstructed skin equivalents showed less melanin uptake than mock-transfected keratinocytes and non-treated skin equivalents, respectively. DKK1 treatment decreased expression levels of PAR-2, and the expression of PAR-2 was lower in reconstructed skin equivalents treated with DKK1 and in palmoplantar epidermis in vivo.
Thus, DKK1 not only affects melanocytes and suppresses their proliferation and differentiation, but also decreases melanin transfer from melanocytes to keratinocytes via the suppression of PAR-2, explaining why palmoplantar epidermis is hypopigmented.
Keratin 9 is a marker for palmoplantar epidermis, since it is expressed exclusively in suprabasal layers of palmoplantar epidermis and is only sparsely detected in acrosyringia, epidermal keratinocytes around sweat gland ducts, and in non-palmoplantar epidermis (Knapp et al, 1986). We had already proven that palmoplantar fibroblasts can induce keratin 9 expression in non-palmoplantar keratinocytes (Yamaguchi et al., 1999).
We investigated whether DKK1 alone can recapitulate the palmoplantar phenotype in non-palmoplantar keratinocytes in terms of skin thickness (Yamaguchi et al., 2008). Treatment of keratinocytes with DKK1 resulted in stimulation of proliferation and density, as measured by the MTT assay (DKK1 vs. control; 0.412±0.069 vs. 0.358±0.085 A405; p=0.012; n=6) and by the average size/area of individual keratinocytes (DKK1 vs. control; 641±148 vs. 1097±366 μm2; p=0.006; n=8), respectively, as recently reported for human mesenchymal stem cells (Gregory et al, 2005). Treatment of reconstructed human skin with DKK1 induced a thickened epidermis.
We performed microarray analyses using normal human keratinocytes either treated or not treated with DKK1 (http://www.ncbi.nlm.nih.gov/geo/ GSE9211), and we identified three genes responsible for inducing the thickened phenotype: keratin 9, αKLEIP and β-catenin. Keratin 9 seems to be essential to support the skin against severe mechanical stimuli (Knapp et al., 1986). αKLEIP is involved in cytokinesis and cell contraction (Hara et al, 2004), suggesting that keratinocytes become smaller due to severe internal tension. Decreased expression of (membranous type) β-catenin may result in the shrinkage (or the decreased space occupation) of cells (Fedi et al, 1999). Treatment of keratinocytes with DKK1 increased the expression of αKLEIP and keratin 9 and decreased the expression of β-catenin through regulation of the Wnt/β-catenin signaling pathway. These expression patterns were identical in in vivo specimens.
DKK1 is responsible for thickened and hypopigmented palmoplantar epidermis, and it may be useful for inducing thickening of specific skin areas such as photo-aged skin and palmoplantar wounds caused by diabetic foot ulcers or rheumatic ulcers and/or for reducing skin pigmentation. These cumulative studies elucidate why human palmoplantar skin is thicker and paler than non-palmoplantar skin, due to the secretion of DKK1 by fibroblasts that affect the overlying epidermis (Figure 1). We are further investigating the interaction of DKK1 with Wnt/β-catenin signaling pathways and several key molecules including PAR-2, αKLEIP and keratin 9. HOXA13, Wnt5a and fibronectin may interact with DKK1 to regenerate and to maintain the homeostasis of adult palmoplantar skin.
This research was supported by a grant-in-aid from the Ministry of Education, Culture, Sports, Science, and Technology (Japan; no. 18689028) and by the Intramural Research Program of the National Cancer Institute at NIH. We thank Dr. Ichiro Katayama, Chairman of the Department of Dermatology, Osaka University Graduate School of Medicine (Osaka, Japan), for encouragement.