Coat colors are determined by melanin (eumelanin and pheomelanin). Melanin is synthesized in melanocytes and accumulates in special organelles, melanosomes, which upon maturation are transferred to keratinocytes. Melanocytes differentiate from undifferentiated precursors, called melanoblasts, which are derived from neural crest cells. Melanoblast/melanocyte proliferation and differentiation are regulated by the tissue environment, especially by keratinocytes, which synthesize endothelins, steel factor, hepatocyte growth factor, leukemia inhibitory factor and granulocyte-macrophage colony-stimulating factor. Melanocyte differentiation is also stimulated by alpha-melanocyte stimulating hormone; in the mouse, however, this hormone is likely carried through the bloodstream and not produced locally in the skin. Melanoblast migration, proliferation and differentiation are also regulated by many coat color genes otherwise known for their ability to regulate melanosome formation and maturation, pigment type switching and melanosome distribution and transfer. Thus, melanocyte proliferation and differentiation are not only regulated by genes encoding typical growth factors and their receptors but also by genes classically known for their role in pigment formation.
Melanocytes are neural crest-derived cells that synthesize melanin pigments (Hearing, 1993, 2000; Ito, 2003; Mayer, 1973; Rawles, 1947). Undifferentiated precursors of melanocytes, melanoblasts, are derived from neural crest cells in embryonic skin (Mayer, 1973; Rawles, 1947). Melanoblasts invade the epidermis (Mayer, 1973) and colonize there. Mouse epidermal melanocytes are known to differentiate from melanoblasts around the time of birth (Hirobe, 1984a). Fully differentiated melanocytes are characterized by pigmentation and well developed dendrites and can be seen mainly in hair bulbs of the skin, where they secrete mature melanosomes into surrounding keratinocytes, giving rise to melanized hairs (Hirobe, 1995; Mann, 1962; Peters et al., 2002; Slominski and Paus, 1993). Hair bulb melanocytes are derived from epidermal melanoblasts and melanocytes (Hirobe, 1992a). In the hairy skin of mice, epidermal melanocytes are found only during the early weeks after birth (Hirobe, 1984a). However, in glabrous skin such as the ear, nose, foot sole and tail of mice, epidermal melanocytes are also found in adult mice (Quevedo and Smith, 1963).
Melanin synthesis occurs in specialized organelles called melanosomes (Seiji et al., 1963). Melanosome maturation is categorized into four stages: unmelanized immature premelanosomes in stages I and II, and melanized melanosomes into stages III and IV (Fitzpatrick et al., 1969). In mice, coat colors are regulated by melanosome transfer from melanocytes to neighboring keratinocytes in hair bulbs. Melanosomes are produced in varying sizes, numbers and densities in melanocytes. Melanosomes in hair bulb melanocytes are passed on to the hair shaft where the final distribution patterns of the pigment are determined. This distribution plays an important role in determining the coat coloring of mice (Silvers, 1979). Eumelanin-containing melanosomes (eumelanosomes) are elliptical, with longitudinal depositions of pigments in intraluminal fibrils (Hearing et al., 1973; Hirobe and Abe, 1999; Sakurai et al., 1975). In contrast, pheomelanin-containing melanosomes (pheomelanosomes) are spherical, with granular depositions of pigments in multivesicular bodies found in yellow phase agouti melanocytes as well as in yellow melanocytes (Sakurai et al., 1975; Takeuchi, 1985). Thus, differences in melanin synthesis correspond to differences in melanosome morphology.
Proliferation and differentiation of mouse melanocytes during development is regulated by numerous genetic and epigenetic factors (Hirobe, 1992a). Among the genetic factors, the coat color genes are the most important (Bennett and Lamoreux, 2003; Hirobe and Abe, 1999; Lamoreux et al., 2001, 2010; Silvers, 1979). In mice, more than 300 genes are involved in melanocyte proliferation and differentiation; about half of these have been cloned and their functions clarified (Mouse Genome Informatics). However, many unknown genes and their functions still remain to be investigated. Moreover, epigenetic factors from the surrounding tissue environment, such as keratinocytes and fibroblasts, the pituitary gland, other organs and the blood supply, as well as environmental factors such as ultraviolet (UV) radiation and ionizing radiation are also important for the regulation of melanocyte proliferation and differentiation. In this article, studies of genetic and epigenetic control of proliferation and differentiation of melanocytes are reviewed and discussed in detail.
Alpha-melanocyte-stimulating hormone (MSH) and adenosine 3′:5′-cyclic monophosphate (cAMP)
Among the various strains of mice, C57BL/10JHir mice (B10 mice; Figure 1A) possess the greatest number of epidermal melanoblasts and melanocytes (Hirobe, 1992a). Thus, B10 mice are useful for in vivo and in vitro studies of the mechanisms regulating melanocyte proliferation and differentiation. We have developed serum-free, chemically defined culture media to optimize and maintain melanoblasts and melanocytes from epidermal cell suspensions taken from newborn B10 mice (Table 1). Mouse epidermal melanoblasts preferentially proliferate in melanoblast-defined medium (MDM). After 14 days, almost all keratinocytes that exist predominantly in the early stage die, and a pure culture of melanoblasts can be obtained. By using MDM, many candidate melanogens have been tested for their differentiation-stimulating activity. Media supplemented with α-MSH (0.01–100 nM) from primary culture induce almost complete melanocyte differentiation as well as melanosome maturation and dendritogenesis (Hirobe, 1992c). In addition, α-MSH injected subcutaneously into newborn B10 mice greatly stimulates melanocyte differentiation, melanosome formation, and maturation and dendritogenesis (Hirobe, 1992a). These results from in vitro and in vivo studies strongly suggest that differentiation of mouse melanocytes from melanoblasts at the newborn stage is regulated by α-MSH. However, when pure melanoblasts cultured in MDM for 14 days were cultured in MDM supplemented with α-MSH for further 14 days, they failed to differentiate, suggesting that induction of melanocyte differentiation by α-MSH requires the presence of keratinocyte-derived factors.
Table 1. Four kinds of culture media used in this review
Melanoblasts or melanocytes at 14 days
Cell yield/35 mm dish × 104 at 14 days
Ins, insulin (10 μg/ml); BSA, bovine serum albumin (0.5 mg/ml); EA, ethanolamine (1 μM); PEA, phosphoethanolamine (1 μM); SE, sodium selenite (10 nM); α-MSH, melanocyte-stimulating hormone (100 nM); DBcAMP, dibutyryl adenosine 3′:5′-cyclic monophosphate (0.5 mM); bFGF, basic fibroblast growth factor (2.5 ng/ml). All these components were added to Ham’s F-10 medium.
Ins, BSA, EA, PEA, SE
MDM + α-MSH
MDM + DBcAMP
MDMD + bFGF
Pure melanoblasts (90%) and melanocytes (10%)
Dibutyryl cAMP (DBcAMP) and 3-isobutyl-1-methylxanthine (IBMX) also induce melanocyte differentiation, melanosome formation and dendritogenesis in MDM (Hirobe, 1992c). Cyclic AMP exerts its effect through protein kinase A (PKA) (Abdel-Malek et al., 1995). PKA phosphorylates and activates the cAMP-responsive element-binding protein (CREB) (Bertolotto et al., 1998; Tachibana, 2000). CREB then binds to the cAMP-responsive element (CRE) in the M promoter of microphthalmia-associated transcription factor (Mitf) (Busca and Ballotti, 2000). The resulting transient increase in Mitf-M expression may lead to up-regulation of TYR, TRP1 and TRP2 (Busca et al., 2000), enabling differentiation.
Proopiomelanocortin (POMC) produces adrenocorticotropic hormone (ACTH) and α-, β- and γ-MSH (Smith and Funder, 1988). Alpha-MSH is identical to N-acetyl-ACTH1–13. Full-length ACTH (ACTH1–39) as well as ACTH fragments induce almost complete differentiation of mouse epidermal melanocytes in MDM. ACTH4–12 is the minimal message sequence required to induce complete differentiation of mouse epidermal melanocytes (Hirobe and Abe, 2000). In human skin, POMC-derived peptides such as ACTH1–14 are present in the epidermis and in cultured keratinocytes (Wakamatsu et al., 1997). They increase adenylate cyclase and the melanin content and dendricity of cultured human melanocytes. However, no expression of the POMC gene has been observed in the epidermis and dermis of B10 mouse skin from 13–19 days’ gestation or in cultured keratinocytes and melanoblasts/melanocytes derived from newborn B10 mice (Hirobe et al., 2004a). Thus, in mice, α-MSH and ACTH/ACTH fragments may be mostly derived from the pituitary gland via the bloodstream.
Steroid hormones and fatty acids
Sex hormones such as estrogen, progesterone and androgen stimulate melanogenesis in a sex-dependent manner (Snell and Bischitz, 1960). When MDM is supplemented with these hormones, differentiation of B10 melanocytesis induced (Hirobe, unpublished). However, only 30–40% of the cells are differentiated melanocytes. Other steroid hormones, such as hydrocortisone and dexamethasone, show a similar tendency to induce differentiation of mouse melanocytes in culture (Hirobe, 1996). At present, it is unclear why the induction of melanocyte differentiation by sex and steroid hormones is incomplete. Dexamethasone also induces differentiation of melanocytes from embryonic stem cells in culture (Yamane et al., 1999).
Fatty acids are components of cell membranes and play important roles in signaling pathways in cells. Linoleic acid (an unsaturated fatty acid) is reported to accelerate degradation of tyrosinase, whereas palmitic acid (a saturated fatty acid) inhibits degradation of tyrosinase in mouse melanoma cells (Ando et al., 2004). In contrast, palmitic acid increased tyrosinase activity in the epidermis derived from C57BL/6J tail skin in organ culture (Shono and Toda, 1981). We have shown that saturated fatty acids such as palmitic acid, palmitoleic acid and stearic acid as well as unsaturated fatty acids such as oleic acid induce differentiation of B10 melanocytes cultured in MDM, similarly to differentiation of sex hormones (Hirobe et al., unpublished data). Further studies on the mechanisms of action of fatty acids in the melanocyte differentiation are required.
Iron and its compound
Among the variety of minerals present in organisms, iron is the most essential factor for normal development of skin and its appendages such as hair and nails (Lansdown, 2001). However, excess iron elicits hyperpigmentation in human skin through oxidation of tyrosine (Lansdown, 2001). A complex of ferrous chloride and ferric chloride (Sugi and Yamashita, 1991), known as ferrous ferric chloride (FFC®; Akatsuka Corporation, Mie, Japan), is one of the dimers of Fe (II) and Fe (III). FFC-supplemented MDM induces differentiation of melanocytes, but only 30–40% of the cells in the melanoblast-melanocyte population are differentiated melanocytes (Hirobe, 2007). However, when epidermal cell suspensions were treated with FFC plus herbal medicines (Chinese wolfberry and Siberian ginseng), almost all melanocytes could be induced to differentiate (Hirobe, 2009a). These results suggest that FFC can stimulate melanocyte differentiation in combination with herbal medicines. In human melanocytes in culture, FFC also stimulated differentiation, melanogenesis and dendritogenesis (Hirobe, 2009b). Further, FFC-containing skin lotions painted on the dorsal skin of newborn B10 mice accelerated pigmentation and hair growth (Hirobe, 2009c). Moreover, FFC-containing skin lotions painted on the top of the covers of culture dishes stimulated differentiation, melanogenesis and dendritogenesis of mouse and human melanocytes without being added to MDM (Hirobe, in press). These results suggest that FFC can stimulate melanocyte differentiation, melanogenesis and dendritogenesis from a distance.
Factors controlling melanocyte proliferation
Keratinocytes are involved in regulating the proliferation and differentiation of melanocytes
Ultrastructures of pure melanoblasts cultured in MDM (Figure 2A), melanocytes cultured in melanocyte-differentiation medium (MDMM; Table 1, Figure 2B) and melanocyte-proliferation medium (MDMD; Table 1, Figure 2C) and melanoblasts cultured in melanoblast-proliferation medium (MDMDF; Table 1, Figure 2D) are similar to those of melanoblasts and melanocytes in the epidermis of neonatal B10 mouse skin (Hirobe and Takeuchi, 1977). That is, the number and distribution of stage I–IV melanosomes, the Golgi apparatus and mitochondria are similar between cultured epidermal melanoblasts/melanocytes and melanoblasts/melanocytes in the epidermis. In the initial stage of primary culture of B10 mouse epidermal cell suspensions in MDMD or MDMDF, keratinocytes proliferate well and melanoblasts and melanocytes start to proliferate around the keratinocyte colony, suggesting that keratinocytes produce and release melanocyte mitogens and melanogen in cooperation with DBcAMP and/or basic fibroblast growth factor (bFGF) (Hirobe, 1992b,c, 1994).
To identify these keratinocyte-derived factors, candidate substances were tested for their mitogenic and melanogenic effects on melanocytes by adding them to the culture media. Of various candidates, endothelin (ET) 1 (Edn1), ET2 (Edn2), ET3 (Edn3), leukemia inhibitory factor (LIF), steel factor [SLF or kit ligand (kitl)], hepatocyte growth factor (HGF) and granulocyte-macrophage colony-stimulating factor (GMCSF, Csf2) are known to be derived from keratinocytes (Hirobe, 2005). These factors are involved in regulating proliferation of neonatal mouse epidermal melanoblasts and melanocytes as well as differentiation of melanocytes. They increase mitotic indices of melanoblasts and melanocytes as well as the percentage of melanoblasts and melanocytes in the S phase of the cell cycle. Using transgenic mice, kitl (Kunisada et al., 1998), HGF (Kunisada et al., 2000) and Edn3 (Aoki et al., 2009) were also found to be keratinocyte-derived mitogens and melanogens.
Keratinocyte-derived factors that inhibit proliferation of melanoblasts and melanocytes are not well known. Interleukin (IL) 1α was revealed to inhibit proliferation of undifferentiated melanoblasts taken from B10 mice and cultured in MDMDF irrespective of the presence or absence of keratinocytes, but it inhibited proliferation of differentiated melanocytes cultured in MDMD in the presence only of keratinocytes. Moreover, IL-1α induces differentiation of melanocytes and also stimulates tyrosinase activity, melanin synthesis and dendritogenesis of melanocytes irrespective of the presence or absence of keratinocytes (Hirobe and Ootaka, 2007). These results suggest that IL-1α is involved in inhibiting the proliferation of neonatal mouse epidermal melanoblasts alone, and, in addition, it is involved in inhibiting the proliferation and differentiation of melanocytes in cooperation with keratinocyte-derived factors. IL-1α could also be involved in stimulating differentiation, melanogenesis and dendritogenesis.
Keratinocyte-derived factors controlling proliferation and differentiation of mouse melanocytes are thought to possess common characteristics. That is, these molecules are related to hematopoiesis or blood vessel formation, and they are supplied to target tissues via the bloodstream. ETs are produced by endothelial cells (Yanagisawa et al., 1988). HGF is related to angiogenesis by its stimulation of the proliferation and movement of endothelial cells (Bussolino et al., 1992). In addition, Kitl (Nocka et al., 1990), LIF (Tomida et al., 1984) and Csf2 (Morstyn and Burgess, 1988) are functional in hematopoietic cells. It is possible that during development, melanoblasts and melanocytes are influenced by factors released from blood vessels during their migration (Manova and Bachvarova, 1991; Matsui et al., 1990; Steel et al., 1992). After colonizing the epidermis, melanocytes require the supply of these factors from the surrounding tissue environment, especially from keratinocytes, for their further proliferation and differentiation.
UV and ionizing radiation
UV irradiation increases the proliferation, differentiation, dendritogenesis and melanogenesis of mouse and human melanocytes (Furuya et al., 2002, 2009; Hirobe, 2005; Imokawa, 2004; Naganuma et al., 2001; Quevedo and Fleischmann, 1980; Szabo, 1967). Pigmented spots are induced in the skin of hairless mice long after cessation of repeated UVB irradiations (Furuya et al., 2002), which were found to stimulate proliferation and differentiation of epidermal melanocytes from the skin of these pigmented spots (Furuya et al., 2002; Hirobe, 2005). The stimulation was elicited by the surrounding keratinocytes rather than the melanocytes themselves. Keratinocytes derived from irradiated skin seem to produce more mitogens and melanogens for melanocytes than produced by those from non-irradiated skin. Antibodies against numerous growth factors and cytokines have been tested for their mitogenic and melanogenic effect, and Csf2 was shown to be involved in regulating proliferation and differentiation of melanocytes in pigmented spots (Hirobe, 2005). The quantity of Csf2 secreted from keratinocytes derived from the pigmented spots was greater than that from keratinocytes derived from control mice. Moreover, immunohistochemistry revealed that expression of Csf2 was dominant in keratinocytes derived from the pigmented spots, but not from the normal skin of non-irradiated mice. Thus, Csf2 is a keratinocyte-derived factor involved in regulating proliferation and differentiation of epidermal melanocytes from pigmented spots developed after UVB irradiation. In human melanocytes, Edn1 and Kitl play major roles in regulating proliferation and differentiation of melanocytes from UV-irradiated skin (Hachiya et al., 2001; Hirobe et al., 2010a; Imokawa, 2004).
Ionizing radiation also affects proliferation and differentiation of mouse melanocytes. It has been reported to exert several actions on mouse melanocytes, including selective killing (Chase, 1949; Potten, 1968; Reams and Schaeffer, 1968), somatic mutation (Fahrig, 1975; Russell and Major, 1957), inhibition of proliferation and/or differentiation (Hirobe, 1995), stimulation of differentiation of melanocyte stem cells (Inomata et al., 2009) and stimulation of melanogenesis (Quevedo and Isherwood, 1958). Ionizing radiation of embryonic skin produces white spots (patches of white hair) in the mid-ventrum and tail tips of B10 mice (Hirobe et al., in press). The frequency of occurrence of ventral white spots in B10 mice differs with the type and amount of linear energy transfer (LET) of irradiation (γ-rays and heavy ions) in a dose-dependent manner. Moreover, in the skin of 18-day-old B10 mouse embryos, γ-rays and heavy ions greatly inhibited proliferation and differentiation of epidermal melanoblasts and melanocytes as well as hair bulb melanocytes (Hirobe et al., in press), suggesting that ionizing radiation affects proliferation and differentiation of mouse melanocytes during development.
FFC stimulates proliferation of mouse and human epidermal melanoblasts and melanocytes (Hirobe, 2007, 2009b,c); it also stimulates the proliferation from a distance (Hirobe, in press). It is possible that FFC can stimulate the production and release of mitogens for melanoblasts and melanocytes, since FFC can stimulate proliferation and differentiation of mouse and human keratinocytes and fibroblasts (Hirobe, 2007, 2009b,c).
Basic FGF stimulates proliferation of undifferentiated melanoblasts from newborn B10 mice in the presence of DBcAMP and keratinocyte-derived factors such as Edn1, Edn2, Edn3, Kitl, LIF, HGF and Csf2 (Hirobe, 2005). However, anti-bFGF antibody failed to inhibit proliferation of melanoblasts in primary cultures of epidermal cell suspensions in MDMDF or MDMD, suggesting that bFGF is not a keratinocyte-derived factor. In mice, bFGF appears to be derived from fibroblasts in the dermis (Figure 3). In humans, bFGF is known to be a keratinocyte-derived mitogen for melanocytes (Halaban et al., 1988; Yamaguchi and Hearing, 2009). Basic FGF binds to its specific receptors, FGFR-1/FGFR-2 (Johnson and Williams, 1993), activates MK (Imokawa et al., 1997) and elicits upregulation of proteins required for melanoblast proliferation (Figure 3).
The signaling pathways of these keratinocyte-derived factors form a complex network with other factors (Figure 3). Proliferation of mouse melanoblasts requires three signaling pathways: PKA by cAMP elevators, PKC by Edn1/Edn2/Edn3 and MK by Kitl/Csf2/LIF/HGF/bFGF. Proliferation of mouse melanocytes requires two signaling pathways: PKA and PKC or MK. Differentiation and/or melanogenesis/dendritogenesis of mouse melanocytes require two signaling pathways: PKA and PKC or MK.
Regulation by genes controlling melanocyte numbers
In the epidermis, B10 mice possess numerous melanoblasts and melanocytes, whereas C3H/HeJms (C3H) mice (A/A) possess extremely few. Results of genetic crosses between B10 and C3H mice indicate that the melanocyte and melanoblast-melanocyte populations in the epidermis of newborn mouse skin are controlled by semi-dominant genes (Hirobe, 1995). Epidermal melanocytes of newborn B10 mice are capable of proliferating during healing of skin wounds (Hirobe, 1988a), but proliferative responses to skin wounding in newborn C3H mice showed that the melanocyte population did not increase after wounding, despite a slight increase in melanoblast-melanocyte populations. Pigment-producing melanocytes in mitosis were frequently found in B10 and its congenic agouti mice (B10-A/A) but not in C3H mice (Hirobe, 1988b). Genetic crosses between B10 and C3H mice indicate that the proliferation of mouse epidermal melanocytes during healing of skin wounds is controlled by semi-dominant genes. The increase in the size of the melanoblast-melanocyte population in the epidermis of C3H mice seems to be the result of the division of melanoblasts, since the number of melanocytes did not increase and mitosis of pigment-producing melanocytes was not observed. Therefore, it is conceivable that B10 melanocytes possess proliferative activity even at a late stage of differentiation, whereas the activity of C3H melanocytes is restricted to the early stage of differentiation. This assumption is supported by results from cell culture experiments where C3H melanocytes failed to proliferate in MDMD, but proliferated well in MDMDF (Hirobe, unpublished). Thus, the genes controlling proliferation of epidermal melanocytes during healing of skin wounds may determine how long epidermal melanocytes retain their proliferative capacity during differentiation.
Regulation by the coat color genes
To clarify the mechanisms of regulation of proliferation and differentiation of melanocytes by coat color genes, characteristics of proliferation and differentiation of mouse melanocytes in culture were compared between two different strains of mice that possess the same genetic background except for one allele in the topical coat color locus, i.e. congenic strains were compared. Alleles from eight coat color loci, namely agouti (A), brown (b, Tyrp1b), albino (c, Tyrc), dilute (d, Myo5ad), recessive yellow (e, Mc1re), pink-eyed dilution (p, Oca2p), slaty (slt, Dctslt) and sash (Wsh, KitW-sh), were introduced to B10 background by repeated backcrosses, and congenic lines of B10 mice were prepared (Figure 1): B10-A/A (Figure 1B), -b/b (Figure 1C), -c/c (Figure 1D), -d/d (Figure 1E), -e/e (Figure 1F), -p/p (Figure 1G), -slt/slt (Figure 1H) and -Wsh/Wsh (Figure 1J). The ru2d (Hps5ru2-d) allele is a spontaneous autosomal recessive mutation that occurred in B10 mice in our laboratory (B10-ru2d/ru2d,Figure 1I).
Expression of the agouti pattern formation is altered by genic substitutions at the agouti locus (Sakurai et al., 1975). Animals homozygous for the a allele produce black eumelanin only (Silvers, 1979). The switch between eumelanin and pheomelanin synthesis is regulated by α-MSH and agouti protein, the product of the A allele expressed in the hair bulb (Barsh, 1996). The agouti protein is produced and released from dermal papilla cells in the hair bulb. A recent study showed that loss and gain of function of β-catenin in dermal papilla cells resulted in yellow and black mice, respectively. In addition, β-catenin activity in dermal papilla cells regulates melanocyte activity (eumelanogenesis) via a mechanism that is independent of the agouti protein (Enshell-Seijffers et al., 2010). These results suggest that β-catenin plays an important role in agouti pattern formation as well as in eumelanogenesis.
The number of melanoblasts and melanocytes after birth does not differ between black and agouti mice (Hirobe and Abe, 1999). The proliferation of agouti melanocytes cultured in MDMD is also similar to that of black melanocytes (Table 2). Agouti melanocytes (Figure 4B) exhibit normal morphology (dendritic, polygonal or epithelioid) and a similar degree of pigmentation to black melanocytes (Figure 4A). In addition, there is no difference in TYR, Tyrp1, Dct and Kit activity between black and agouti melanocytes (Table 3). Melanosomes of black and agouti melanocytes are evenly distributed within the melanocytes, and elliptical stage IV melanosomes are observed (Figure 5A,B). However, there are fewer stage IV melanosomes in agouti melanocytes than in black melanocytes (Table 2).
Table 2. Effects of the coat color genes on proliferation and differentiation of mouse epidermal melanocytes in culture
Effects of the coat color genes on proliferation, differentiation, melanosome formations, eumelanin (Eu) and pheomelanin (Phe) synthesis, reactivity to l-tyrosine (Tyr) in melanocytes cultured in MDMD for 14 days.
Table 3. Effects of the coat color genes on expression of melanocyte-specific proteins, tyrosinase (TYR), tyrosinase-related protein-1 (Tyrp1), TRP2 (Dct), Kit and Kitl in melanocytes cultured for 7 days
MDMD + l-Tyr (2 mM)
Results of immunocytochemical staining (TYR, Tyrp1, Dct, Kit and Kitl) of primary melanoblasts/melanocytes cultured for 7 days are shown. Epidermal cell suspensions derived from 0.5-day-old wild-type (B10) and congenic mice that carry many coat color mutant genes were cultured in MDMD.
Chemical analysis of melanin produced in cultured epidermal melanocytes revealed that the content of pyrrole-2,3,5-tricarboxylic acid (PTCA, a degradation product of eumelanin; Ito and Fujita, 1985; Ito and Wakamatsu, 1994) in agouti melanocytes is similar to that in black melanocytes (Hirobe et al., 2004b). Also, the content of 4-aminohydroxyphenylalanine (4-AHP, a degradation product of pheomelanin; Wakamatsu and Ito, 2002; Wakamatsu et al., 2002) in agouti melanocytes cultured in MDMD is similar to that in black melanocytes (Table 2), as are the PTCA/AHP ratios. However, a 1.5-fold increase in AHP, and a 37-fold increase in 5-S-cysteinyldopa (5-S-CD, a precursor of pheomelanin), was observed in culture media derived from agouti melanocytes cultured in MDMD (Hirobe et al., 2004b). Moreover, a 11-fold increase in AHP content in the epidermis of 3.5-day-old agouti mice and a 95-fold increase in the epidermis of 5.5-day-old agouti mice were observed compared with black mice. Analysis of the A allele using reverse transcription-polymerase chain reaction (RT-PCR) revealed that cultured keratinocytes and melanocytes did not express the A allele. Moreover, the agouti protein was expressed in the dermis of 0.5-, 3.5- and 5.5-day-old agouti mice, but not in the dermis of black mice or in the epidermis of agouti or black mice. These results suggest that epidermal melanoblasts of agouti mice can be influenced by the agouti protein produced in the dermis and can continue to synthesize pheomelanin in culture conditions. Pheomelanin production in the epidermis of 3.5- and 5.5-day-old agouti mice may be the result of the influence of the agouti protein produced in the dermis.
B (Tyrp1), the wild-type allele at the brown locus, produces black eumelanin, whereas b (Tyrp1b), the recessive allele, produces brown eumelanin. The coat color of brown mice is lighter than that of black mice, whereas the tyrosinase activity in brown mice is higher than in black mice (Foster, 1965; Hirobe, 1984b; Tamate et al., 1989). The proliferation rate of brown melanocytes cultured in MDMD is similar to that of black melanocytes (Table 2), and brown melanocytes possess normal morphology (dendritic, polygonal or epithelioid), but their pigmentation (Figure 4C) is much lower than that of black melanocytes. TYR, Dct and Kit activity in brown melanocytes does not differ from that in black melanocytes, but TRP1 activity is greatly reduced (Table 3). Although brown melanosomes are evenly distributed within melanocytes, their morphology is very different from that of black melanosomes. Elliptical melanosomes and mature stage IV melanosomes are rarely observed (Table 2). Brown melanosomes are mostly spherical stage III melanosomes with granular or lamellar depositions of pigments. In addition, eumelanin is decreased threefold in brown melanocytes, whereas pheomelanin is increased fourfold (Hirobe et al., 1998; Ozeki et al., 1995; Tamate et al., 1989). The PTCA/AHP ratio in brown melanocytes is one-tenth of that in black melanocytes. The formation of elliptical eumelanosomes requires plenty of eumelanin and higher TRP1 activity.
C (Tyr), the wild-type allele of the albino locus, produces melanin, whereas c (Tyrc), the recessive allele, produces no pigment in the coat and eyes (Silvers, 1979; Tanaka et al., 1990; Yamamoto et al., 1989). The c allele is a point mutation at nucleotide residue 387 (G to C transversion) causing a Cys to Ser substitution at position 85 in one of the cysteine-rich domains of the tyrosinase molecule (Shibahara et al., 1990). This mutation reduces tyrosinase activity completely. However, the effects of the c allele on proliferation of melanoblasts remains unclear. We previously studied the effects of the c mutation on proliferation of melanoblasts cultured in MDMD and MDMDF, and found that the proliferation rate of albino melanoblasts was about one-half that of black melanocytes (Hirobe et al., 1998), suggesting that cell proliferation is active in epidermal melanocytes with full melanogenesis such as black melanocytes but not in epidermal melanoblasts with no melanogenesis. In other words, proliferation and differentiation of epidermal melanocytes in culture are linked. Albino melanoblasts exhibit normal morphology (dendritic, polygonal or epithelioid), but no pigmentation was observed (Figure 4D). Expression of TYR in albino melanoblasts is not observed, whereas expression of Tyrp1, Dct and Kit is similar to that in black melanocytes (Table 3). Melanosomes are evenly distributed within albino melanoblasts, and morphology of stage I and II melanosomes is similar to that of black melanocytes (Figure 5D). Further, the number of stage I and II melanosomes is greatly increased compared with black melanocytes (Table 2), probably due to the inhibition of stage III and IV formation by the c mutation.
The recessive allele of the dilute locus, d (Myo5ad), dilutes hair pigmentation in mice. Myosin Va encoded by the dilute locus is a dimer of two 190 kDa heavy chains. The N-terminal head region consists of actin- and ATP-binding sites and functions as a motor domain for short-range movement along actin filaments of the cytoskeleton (Westbroek et al., 2001; Wu et al., 1997). However, the effects of the dilute allele on proliferation and differentiation of melanocytes are not fully clear. We investigated the role of the dilute allele on proliferation and differentiation of melanoblasts and melanocytes cultured in MDMD/MDMDF and found that the proliferation rate of dilute melanoblasts and melanocytes was similar to that of black melanoblasts and melanocytes (Hirobe et al., 1998). The rate of differentiation of dilute melanocytes cultured in MDMD was also similar to that of black melanocytes (Figure 4E), as was reactivity to dopa and dopa-premelanin reactions. Dilute melanocytes were dendritic, polygonal or epithelioid in morphology, but their melanosomes were distributed around the nucleus (Figure 4E). A few melanosomes were observed in the peripheral region of the cytoplasm as well as in dendrites. Expression of TYR, Tyrp1, Dct and Kit in dilute melanocytes was similar to that in black melanocytes (Table 3). Dilute melanosomes were distributed around the nucleus, and the number and morphology of stage I–IV melanosomes was similar to that of black melanosomes (eumelanosome type, Figure 5E). These results suggest that the dilute allele is involved in regulating the transport of melanosomes from the perinuclear region to the dendrites, rather than in regulating dendrite formation. These findings are consistent with results of molecular analyses of the dilute allele (Mercer et al., 1991; Provance et al., 1996; Wei et al., 1997; Wu et al., 1997).
Recessive yellow allele
The extension locus increases brown/black eumelanin in hair follicular melanocytes when dominant, but it blocks eumelanin synthesis, extending the range of red/yellow pheomelanin when recessive (Silvers, 1979). The recessive yellow allele results from a frameshift in Mc1r that produces a prematurely terminated, nonfunctioning receptor (Robbins et al., 1993). In addition to the frameshift mutation, the e allele possesses a conservative point mutation, Val101Ala (Robbins et al., 1993). Moreover, the e allele (using B10-e/e skin) stimulates pheomelanin synthesis in the epidermis and dermis as well as hair follicles (Hirobe et al., 2007a). In e/e mice, melanoblasts and melanocytes are greatly reduced in number (Tamate et al., 1986).
Since we could not obtain a pure culture of yellow melanocytes producing pheomelanin only from e/e mice, we investigated proliferation and differentiation of cultured recessive yellow melanocytes producing mainly eumelanin. The addition of DBcAMP to culture media can elicit upregulation of the PKA pathway and stimulate eumelanogenesis in melanocytes (Tamate and Takeuchi, 1984). The proliferation rate of e/e melanoblasts or melanocytes cultured in MDMDF or MDMD is decreased (by around one-half) compared with that of black melanoblasts and melanocytes (Figure 4F). Differentiation of melanocytes cultured in MDMD is also delayed in e/e mice (Hirobe et al., 2007b). Although the expression of TYR and Kit in e/e melanocytes is similar to that in black melanocytes, expression of Tyrp1 and Dct is decreased (Hirobe et al., 2007b). The number of stage III melanosomes does not change, while the number of stage IV melanosomes is decreased (Figure 5F). Excess l-Tyr added to MDMD rescued the reduced proliferation rate of e/e melanocytes (Hirobe et al., 2007b). l-Tyr also stimulated TYR activity and expression of Tyrp1, Dct and Kit as well as maturation of stage IV melanosomes and eumelanin synthesis. These results suggest that the e mutation affects proliferation and differentiation of melanocytes and l-Tyr rescues the reduced proliferation and differentiation of e/e melanocytes by stimulating TYR activity and expression of Tyrp1 and Dct as well as melanosome maturation and eumelanin synthesis. Even at the higher cAMP levels elicited by DBcAMP-supplemented MDMD and MDMDF, proliferation of e/e melanoblasts and melanocytes is greatly inhibited, suggesting that the PKA pathway elicited by excess DBcAMP in e/e melanocytes is different from the PKA pathway elicited by wild-type Mc1r in black melanocytes. The altered PKA pathway in e/e melanocytes may affect cross-talk with PKC or MK, and consequently proliferation and differentiation may be inhibited. l-Tyr is thought to rescue the altered PKA pathway as well as the altered cross-talk between PKA and PKC/MK.
Eumelanin and pheomelanin content in the dorsal hairs of female B10-e/e mice is greater than that seen in male mice, suggesting that the expression of the recessive yellow allele is regulated in a sex-dependent manner (Hirobe et al., 2007a). We have suggested that estrogen is a main factor in determining the higher content of eumelanin and pheomelanin in the hair of female e/e mice (Hirobe et al., 2010b).
Proliferation and differentiation of mouse melanocytes cultured in MDMD is greatly inhibited by the p mutation (Figure 4G) and l-Tyr rescues both proliferation and differentiation (Hirobe et al., 2002), although most of melanins and their precursors fail to accumulate in p/p melanosomes (Wakamatsu et al., 2007). Moreover, in p/p melanoblasts, only a few stage I and II melanosomes are observed (Figure 5G), whereas l-Tyr greatly increases the number of stage II, III and IV melanosomes (Table 2). The p/p allele greatly inhibits eumelanin synthesis, but not pheomelanin synthesis. Production of pheomelanin in p/p melanocytes is not influenced by the agouti, non-agouti black and recessive yellow alleles (Hirobe et al., 2011).
Pink-eyed dilution melanoblasts possess smaller but more numerous mitochondria than black melanocytes (Hirobe et al., 2002). Treatment of p/p melanoblasts with l-Tyr decreased the number of mitochondria (Hirobe et al., 2008). Media supplemented with 2,4-dinitrophenol (DNP), an inhibitor of mitochondrial function, stimulated both proliferation and differentiation of p/p melanoblasts, and simultaneous DNP and l-Tyr treatment dramatically induced differentiation of p/p melanocytes (Hirobe et al., 2008). These results suggest that the p allele is involved in regulating the function of mitochondria.
As it appears that mitochondria are well developed in p/p melanoblasts and melanocytes, it is possible that apoptosis occurs. Inhibitors of apoptosis, such as caspase-9 inhibitor (C9I) and Bax-inhibiting peptide (BIP), stimulate proliferation and differentiation of cultured p/p melanoblasts but not of black melanoblasts and melanocytes. The number of apoptotic melanoblasts and keratinocytes in culture derived from p/p mice is greater than in P/P mice (Hirobe et al., unpublished). The number of apoptotic melanoblasts and keratinocytes in p/p mice can be decreased by treatment with C9I and BIP. Moreover, expression of caspase-9 and Bax in p/p melanoblasts and keratinocytes is greater than in P/P melanoblasts and keratinocytes. These results suggest that the increased apoptosis is related to reduced proliferation and differentiation of p/p melanoblasts.
On a non-agouti background, slaty homozygotes possess a slightly diluted coat pigmentation (Green, 1972). In addition to the original slaty mutation, slaty light (Sltlt/DctSlt-lt; more severe effect) and slaty 2J (slt2J/Dctslt−2J; similar phenotype) have been identified (Budd and Jackson, 1995). The slaty mutation is known to change an arginine to a glutamine in the first copper-binding domain of Dct, which converts DC to DHICA in the eumelanin synthesis pathway (Korner and Pawelek, 1980; Jackson et al., 1992; Tsukamoto et al., 1992); the mutation also produces about 10–30% of the activity of wild-type Dct in eye extracts (Jackson et al., 1992). Dct is produced by both wild-type and slaty mutant cDNA, but the protein level of Dct in the slaty mutant is greatly reduced (Kroumpouzos et al., 1994).
The slaty mutation does not affect proliferation of cultured epidermal melanoblasts and melanocytes in MDMD (Figure 4H). However, differentiation and expression of TRP2 in cultured slaty melanocytes is greatly inhibited (Hirobe et al., 2006). The slaty mutation affects both eumelanin and pheomelanin synthesis in a developmental stage-specific and skin site-specific manner.
In slaty melanocytes, numerous spherical melanosomes with granular depositions of pigments, black-type elliptical melanosomes with longitudinal deposits of pigments in intraluminal fibrils and a mix of the two melanosome types are observed (4:1:1) (Figure 5H). Moreover, in slaty melanocytes, there is a large decrease in mature stage IV melanosomes, whereas immature stage III melanosomes are more numerous than in black melanocytes (Hirobe and Abe, 2007). In slaty melanocytes, spherical and mixed type melanosomes gradually decrease after birth, whereas elliptical melanosomes gradually increase. These results suggest that the slaty mutation blocks melanosome maturation at stage III and affects melanosome morphology (elliptical or spherical) in a developmental stage-specific manner.
Inhibition of eumelanin synthesis by the slaty mutation can be partly restored by the addition of excess l-Tyr to MDMD (Hirobe et al., 2006). Eumelanin and pheomelanin may be accumulated with difficulty in slaty melanocytes and are easily released during skin development. l-Tyr is thought to stimulate this release. Perhaps l-Tyr acts directly on melanoblasts and melanocytes and activates factors involved in regulating eumelanin synthesis (Coughlin et al., 1988; Hirobe, 2005; Imokawa, 2004). Another possibility is that l-Tyr acts on the tissue environment, especially keratinocytes, and stimulates synthesis of melanogenic factors controlling eumelanin synthesis (Hirobe, 2005; Imokawa, 2004).
When l-Tyr is added to MDMD, it stimulates melanosome maturation and increases elliptical melanosomes but decreases spherical melanosomes (Hirobe et al., 2006), suggesting that l-Tyr restores the reduced melanosome maturation and changes the altered morphology of melanosomes affected by the slaty mutation. l-Tyr may act directly on melanocytes and activate factors involved in regulating pigmentation. As excess l-Tyr restores maturation of stage IV elliptical melanosomes, slaty melanosomes are thought to possess a normal pathway related to l-Tyr transport. Thus, the possibility exists that the l-Tyr transport system from the cytoplasm to melanosomes is affected by the slaty mutation. If this is true, melanin synthesis would be increased by excess l-Tyr, and maturation of stage IV melanosomes would be stimulated. Furthermore, l-Tyr increases the total number of melanosomes, suggesting that l-Tyr stimulates de novo melanosome formation. It has been reported that α-MSH stimulates differentiation of epidermal melanocytes of black mice in vivo (Hirobe and Takeuchi, 1977). Differentiation stimulated by α-MSH is associated with an increase in the total number of melanosomes. Similar mechanisms in α-MSH and l-Tyr seem to be involved in the stimulation of de novo melanosome formation.
In 2006, a spontaneous autosomal recessive mutant of brown coat color with ruby eyes occurred in the offspring of a B10 mouse dam. The phenotype of this mutant was similar to that of ruby-eye (ru/Hps6ru) or ruby-eye 2 (ru2/Hps5ru2). Human Hermansky–Pudlack syndrome (HPS) is a recessively inherited disease that affects several organs such as the skin (hypopigmentation), eyes (low visual acuity), blood cells (prolonged bleeding) and lungs (interstitial pulmonary fibrosis) (Wei, 2006). Many distinct types of human HPS have been described (Wei, 2006). In mice, many naturally occurring hypopigmentation models of HPS have been characterized (Wei, 2006). Human HPS5 corresponds to mouse Hps5ru2 (ru2), and HPS6 to Hps6ru (ru) (Zhang et al., 2003). RT-PCR analysis revealed that this novel mutation, named ru2d/Hps5ru2-d, elicited a frameshift by 997G deletion in Hps5 (Hirobe et al., unpublished). To clarify the mechanism of the hypopigmentation, the characteristics of proliferation and differentiation of ru2d/ru2d epidermal melanoblasts and melanocytes cultured in the media were investigated. The proliferation of ru2d/ru2d melanoblasts and melanocytes did not differ from that of B10 (Figure 4I). However, the differentiation of ru2d/ru2d melanocytes was greatly inhibited. TYR activity, expression of TYR, Tyrp1, Dct and eumelanin synthesis were markedly decreased in ru2d/ru2d melanocytes (Tables 2 and 3). However, the addition of excess l-Tyr to MDMD rescued the reduced differentiation via increased TYR activity, expression of TYR, Tyrp1, Dct and Kit and eumelanin synthesis (Tables 2 and 3). These results suggest that the ru2d allele inhibits melanocyte differentiation but that the impaired differentiation is rescued by excess L-Tyr.
In ruby-eye 2d melanocytes, elliptical melanosomes are observed, although many immature stage III melanosomes and fewer stage IV melanosomes are observed (Figure 5I). The number of stage IV melanosomes is much smaller than in black melanocytes. The total number of melanosomes in ruby-eye 2d melanocytes is also less than in black melanocytes. l-Tyr markedly increases the number of stage IV melanosomes and the total number of melanosomes in ruby eye 2d melanocytes (Hirobe et al., unpublished). These results suggest that the ruby-eye 2d mutation markedly inhibits melanosome formation and maturation, but its inhibition can be restored by l-Tyr.
Sash forms a dominant spotting pattern (W-locus). This mutation occurred spontaneously in a pair set up to provide a (C3H × 101) F1 hybrid stock (Silvers, 1979). The original mutant had a broad sash of white around its body in the lumbar region and produced offspring like itself when bred with a normal animal. The semi-dominant sash mutation is characterized by a sequence inversion near the Kit gene that leads to ectopic expression of Kit (Duttlinger et al., 1993). Homozygous B10-Wsh/Wsh mice possess almost all-white body hair except for the ear and in heterozygous mice, the center of the body is covered with white hair. Primary cultures of epidermal cell suspensions of sash mice have not detected any melanoblasts or melanocytes. However, co-culture of black melanoblasts/melanocytes with sash keratinocytes stimulated proliferation of black melanoblasts/melanocytes in MDMDF. These results suggest that the sash allele affects early melanoblast development without affecting the production of mitogens for melanoblasts in keratinocytes. Moreover, human epidermal melanocytes can be grown in hair follicles of B10-Wsh/Wsh mice. After plucking out all the reconstituted hairs, the secondary hairs re-grew in the same area and their colors were lighter than the first reconstituted hairs (Ideta et al., 2006). These results also support the assumption that sash keratinocytes possess a normal function in the melanocyte environment.
Action of the coat color genes
The coat color genes that were the focus of this review mostly act directly on melanocytes, whereas the agouti and non-agouti black alleles act on the tissue environment, especially on fibroblasts in dermal papilla. The sash and slaty alleles affect melanoblast migration and differentiation. The albino and pink-eyed dilution alleles influence melanoblast proliferation. The brown, pink-eyed dilution and slaty alleles control formation of stage I and II melanosomes in melanoblasts. The albino, pink-eyed dilution, recessive yellow, slaty and ruby-eye 2d alleles affect expression and activity of tyrosinase in melanocytes. The brown, pink-eyed dilution, slaty and ruby-eye 2d alleles affect melanosome maturation, especially stage IV melanosome maturation. The agouti and non-agouti black and recessive yellow, respectively, affect the types of melanin synthesized (eumelanin or pheomelanin). Finally, the dilute allele is involved in regulating melanosome transfer from melanocyte dendrites to keratinocytes (Figure 6). Precise studies of the mechanisms of action of the coat color genes on melanoblasts and melanocytes using congenic mice that carry the coat color mutant genes enable us to understand the total context of the regulation of melanocyte proliferation and differentiation.
The author expresses his thanks to Prof./Drs Takeuchi (Tohoku University), Ito/Wakamatsu (Fujita Health University), Abe (Yamagata University), Kawa/Mizoguchi/Soma (St Marianna University), Takeuchi/Hotta/Yoshihara (Okayama University), Furuya/Akiu/Naganuma/Fukuda/Ideta/Ifuku/Hara/Horii (Shiseido), Nishikawa/Osawa (RIKEN), Eguchi-Kasai/Sugaya/Murakami (National Institute of Radiological Sciences), Ogawa/Ishizuka (Joetsu University of Education) and Ootaka/Terunuma/Kiuchi (Chiba University) for their collaboration in the original papers. The author also expresses his thanks to Prof. Ito and Dr. Hearing (NIH) for recommending him to write this review.