¹Human Medical Genetics Program, University of Colorado Health Sciences Center, Aurora, Colorado, USA

Correspondence: Dr. Richard A. Spritz, Human Medical Genetics Program, University of Colorado Health Sciences Center, Mail-stop 8300, PO Box 6511, Aurora, Colorado 80045, USA. E-mail: richard.spritz@uchsc.edu

Abstract

Mice transgenic for the Kit Val620Ala mutation, which in humans has been associated with progressive piebaldism, exhibit dominant white spotting but show no evidence of progressive depigmentation. These results are consistent with the previous hypothesis that progressive piebaldism might result from digenic inheritance, of the KIT^V620A mutation that causes piebaldism and a second, unknown locus that causes progressive depigmentation.

Lady Macbeth's anguished lament speaks to some of the most puzzling conundrums in investigative dermatology. The white-spotting disorder piebaldism, because of its visually striking patches of white skin (leukoderma) and hair (poliosis), was known to the ancient Romans and is thought to be the first genetic disorder to have been recognized to exhibit autosomal dominant inheritance (Morgan, 1786). We have learned a lot over the centuries. We now know that the leukodermal patches result from an almost total lack of melanocytes from the involved regions. We now know that human piebaldism, and the homologous murine mutant "dominant white spotting" (W), result from loss-of-function and dominant-negative missense mutations in the KIT gene, and that there is a general correlation between genotype and severity of the resultant clinical phenotype (Spritz, 1994). We now know that KIT encodes the cell surface transmembrane tyrosine kinase receptor for KIT ligand (mast-cell growth factor; stem-cell factor; steel factor), and that KIT-dependent signaling plays key roles in melanogenesis, gametogenesis, and early stages of hematopoiesis. Nevertheless, the most obvious clinical questions remain unanswered. Why is developmental lack of melanocytes in piebaldism so localized, resulting in white spots rather than diffuse hypopigmentation? Why do the white spots preferentially involve the forehead, ventral chest and trunk, backs of the elbows, and fronts of the knees? Why are the congenital leukodermal patches of piebaldism so stable over a patient's life, neither filling in nor growing larger, whereas the acquired white spots of vitiligo may wax and wane? And why are the white spots of piebaldism utterly resistant to medical treatment, whereas those of vitiligo may fill in after ultraviolet therapy and immunotherapy?

Why is developmental lack of melanocytes in piebaldism so localized?

Although the leukodermal patches of piebaldism typically lack melanocytes, they frequently contain hyperpigmented islands, abnormally pigmented macules, and freckles, all of which contain apparently normal melanocytes. Furthermore, the leukodermal patches often exhibit marginal hyperpigmentation, compared with the surrounding skin, which appears essentially normal. In human piebaldism, and in mice with "dominant white spotting" (W), the leukodermal patches are remarkably stable, usually being evident at birth and changing little over the course of patients' lives, although limited filling in has been reported in some milder cases of piebaldism (Davis and Verdol, 1976). Together, these phenomena suggest that deficient KIT-dependent signaling results in a primary defect of migration of neural crest melanoblasts to the skin during embryologic development and, furthermore, interferes with local migration of skin melanoblasts and melanocytes during postnatal life.

Recently, Richards and co-workers (2001) reported a mother and daughter with the remarkable phenotype of progressive piebaldism, the result of a novel amino acid substitution, Val620Ala (V620A), located in a highly conserved region of the KIT intracellular tyrosine kinase domain. In both mother and daughter, congenital skin and hair depigmentation progressed from infancy through at least mid-childhood, ultimately resulting in a relatively severe piebald phenotype. The biological basis for the highly atypical progressive clinical course in this family was not studied but would seem to imply reduced survival of skin melanocytes and self-renewing melanoblasts, in addition to defective melanoblast migration during development. Interestingly, several other family members had a tendency to localized hair graying but did not carry the KIT mutation. Accordingly, the authors speculated that progressive piebaldism in this family might result from digenic co-inheritance of the KIT^V620A mutation and an unknown dominant-acting unlinked mutation responsible for progressive hair graying.

In this issue, Tosaki et al. describe detailed investigation of the biology of the Kit^V620A substitution in transgenic mice, aiming to create an animal model for progressive piebaldism. Another Kit tyrosine kinase domain substitution, D790N, had been shown to act as a dominant-negative in transgenic mice (Ray et al., 1991); this suggested that Kit^V620A might behave similarly, providing an animal model in which to study the biology of progressive melanocyte loss. The authors accordingly engineered the V620A substitution in mouse Kit complementary DNA, prepared four independent transgenic mouse lines, and carefully assessed the resultant phenotypes. Surprisingly, whereas all four transgenic lines exhibited dominant white spotting and reduction of mast-cell numbers consistent with the W-mutant phenotype, none of the lines showed any evidence of postnatal progression of the pigmentary phenotype. Thus, the Kit^V620A transgenic mice exhibit typical dominant white spotting but fail to model the most interesting aspect of the homologous human mutant phenotype, progressive depigmentation.

Tosaki and colleagues (2006) offer several hypotheses as to why the Kit^V620A transgenic mice do not manifest progressive depigmentation. They point out that KIT-dependent signaling is required both for developmental migration of melanoblasts to the skin and also for melanoblast/melanocyte maintenance and survival (Nishikawa et al., 1991; Okura et al., 1995; Kunisada et al., 2001; Botchkareva et al., 2001), and they imply that KIT^V620A might affect the latter more in humans than in mice. However, this would require that the KIT-V620A polypeptide exert a stable dysfunctional effect, in addition to or instead of acting as a straightforward dominant-negative. Perhaps more likely is the original hypothesis of Richards et al. (2001), that the family with progressive piebaldism segregated two unlinked mutations: KIT^V620A, causing typical dominant piebaldism, and a mutation in a second, unknown locus, causing progressive hair graying. Very recently, hair graying has indeed been linked to incomplete maintenance of self-renewing melanoblasts, in which Pax3 and Mitf seem to play key roles (Steingrimsson et al., 2005). Although Richards and co-workers (2001) found no mutations of MITF in their original family, digenic inheritance involving KIT and some other locus remains an attractive hypothesis to account for progressive piebaldism in this family. Nevertheless, a full understanding of this phenomenon may await identification of genes specifically involved in hair graying in humans, and so, for now, atypical progressive piebaldism associated with KIT^V620A must be added to the list of conundrums, rather than to the list of puzzles solved.