Spotlight on Spotted Mice: A Review of White Spotting Mouse Mutants and Associated Human Pigmentation Disorders

Authors

  • Laura L. Baxter,

    1. Mouse Embryology Section, Genetic Disease Research Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD 20892, USA
    Search for more papers by this author
  • Ling Hou,

    1. Mouse Embryology Section, Genetic Disease Research Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD 20892, USA
    Search for more papers by this author
  • Stacie K. Loftus,

    1. Mouse Embryology Section, Genetic Disease Research Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD 20892, USA
    Search for more papers by this author
  • William J. Pavan

    1. Mouse Embryology Section, Genetic Disease Research Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD 20892, USA
    Search for more papers by this author

* Address reprint requests to William J. Pavan, National Institutes of Health, National Human Genome Research Institute, Genetic Disease Research Branch, 49 Convent Drive, Building 49, Room 4A82, Bethesda, MD 20892-4472, USA. E-mail: bpavan@nhgri.nih.gov

Abstract

Mutation of genes that regulate neural crest-derived melanoblast development and survival can result in reduction and/or loss of mature melanocytes. The reduction in melanocyte number in the skin and hair follicles manifests itself as areas of hypopigmentation, commonly described as white spotting in mice. To date ten genes have been identified which are associated with white-spotting phenotypes in mouse. Seven of these genes are associated with neural crest and melanocyte disorders in humans. This review summarizes the phenotypes associated with mutation of these genes in both mouse and man. We describe our current understanding of how these genes function in development, and explore their complex roles regulating the various stages of melanocyte development.

Abbreviations –
E

embryonic day

HMG

high mobility group

RPE

retinal pigmented epithelium

WS

Waardenburg syndrome

WS IV

Waardenburg–Shah syndrome

Introduction

Mouse coat color mutants have been selectively bred for aesthetic reasons for thousands of years. Mouse coat coloration variants include differences in hue (e.g. black, brown, dilute, agouti), differences in color patterning (e.g. brindling, light ventrum), and white spotting (areas lacking pigmentation). For the last century, these mice have proven to be valuable assets to geneticists and embryologists and have been instrumental for many important discoveries. In the early 1900s, coat color variants were vital for testing theories of Mendelian traits, linkage, independent assortment and quantitative traits (1, 2). In the mid- to late 1900s, they were used to establish the first linkage maps of the mouse genome and have provided valuable tools for developmental geneticists studying sites of gene actions. Elucidation of the genetic basis of these mutants during the past two decades has provided insights into human diseases with similar molecular alterations (3, 4).

These fascinating coat coloration patterns are controlled by regulation of growth and function of the pigment-producing cell, the melanocyte, and by interactions of melanocytes with dermal keratinocytes and hair cells. Melanocyte precursors, called melanoblasts, are generated from the neural crest during embryonic development, and migrate along a dorsal-lateral pathway beneath the ectoderm. After extensive migration and proliferation, melanoblasts colonize skin and hair follicles where they differentiate into melanocytes and begin production of melanin pigment.

Color variations in mice are due to alterations in melanin production, distribution, or deposition into hair and/or skin. In contrast, white spotting mutants arise from improper melanoblast development or survival. These spotting mutants exhibit characteristic spotting patterns that include belly spots, head spots, belts spanning the caudal trunk region, piebald spotting, and peppering (Fig. 1). The appearance of these white spots, reflecting absence of mature melanocytes, is caused by defects at various stages of melanocyte development, including proliferation, survival, migration, invasion of the integument, hair follicle entry and melanocyte stem cell renewal. White-spotting defects can also be associated with disorders in other cell populations, related either by a common neural crest origin or by pleiotropic gene functions in different tissue derivatives.

Figure 1.

Examples of spotting patterns in mutant mice. (A) A belted mouse demonstrating the white band proximal to the hindlimb. (B) A KitWLacZ/+ mouse demonstrating a white head blaze, belly spot and small dorsal spot on the back. (C) A piebald spotting pattern demonstrating extensive white spotting between the limbs. (D) A Pax3Spl/+ mouse exhibiting the characteristic white belly spot.

Many white spotting traits have been identified in mouse and man, and 10 of the genes responsible have been cloned (Table 1). Important work has also been done with zebrafish orthologs of these genes (5), however due to space constraints these will not be addressed in this review. In this review, we summarize phenotypes and gene function associated with the cloned white spotting mutants in mouse and man. For a recent review summarizing all murine pigmentation phenotypes, see (6). We also discuss developmental anomalies that lead to white spotting, interactions among white spotting genes, and uncloned white spotting mutants.

Table 1.  Cloned white spotting genes and phenotypes in mouse and man
GeneHuman locationHuman disorderMouse mutantMouse location (cM location)
Pax32q35WS1, WS3Splotch1 @ 44
Mitf3p13WS2AMicrophthalmia6 @ 40
Sox1022q13.1WS4, neuropathiesDominant Megacolon15 @ 46.6
Ednrb13q22WS4Piebald lethal14 @ 51
Edn320q13WS4Lethal spotting2 @ 104
Kit4q12PiebaldismDominant white spotting5 @ 42
Kitl12q22 Steel10 @ 57
Snai2/Slugh8q11WS2/PiebaldismWhite spotting16 @ 9.4
Adamts2012q12 Belted15 @ 44.4
Mcoln 31p22.3 Varitint-waddler3 @74.8

Developmental origins of white spotting

The two major theories of how white spotting patterns occur in mice both state that a very small number of melanocyte precursors go on to populate the entire coat of adult animals. By analyzing mutant spotted mice that were selectively bred to have small areas of pigmentation, Schaible proposed that the dermal melanocyte population of mice arises from only 14 melanocyte stem cells. He postulated that these stem cells migrate laterally before undergoing proliferation, thus establishing distinct ‘pigment centers’ of melanoblast proliferation (7). This theory predicts that white spotting results from loss of one or more of these stem cells, or from failure of these cells to expand to fill intervening gaps.

The second theory of how white spotting occurs was based on the analysis of chimeric mice generated from aggregation of two embryos with different coat color mutations (8). Mintz interpreted the resulting dorsal to ventral striped patterns to reflect lateral melanoblast migration from the neural crest, with no midline crossing and little rostral-caudal dispersion of melanoblast progeny. By analyzing many chimeric animals, Mintz hypothesized that 17 progenitor cells on each side of the neural tube give rise to the melanocyte lineage, and the presence of a spotting mutation causes selective death of a subset of these precursors, leading to absence of melanoblasts in specific regions which cannot be compensated for by proliferating melanocytes generated by neighboring progenitors. In support of this hypothesis, generation of mosaic embryos by injection of a retroviral construct expressing tyrosinase into albino embryos resulted in the appearance of pigmented stripes with a similar width to those seen in aggregation chimeras (9). However, other independent analyses of chimeric mice concluded that these stripes were frequently broken by neighboring stripes, suggesting that the cells generated by melanoblast progenitors mixed at their boundaries (10–13).

A recent study readdressed this question using both mosaics and chimeras. The mosaics contained LacZ driven by the Dct promoter (Dct–LacZ) and underwent rare rearrangement that enabled labeling of individual melanoblast clones. The chimeric embryos used Dct–LacZ as one partner, thus allowing the visualization of regional melanoblast localization. This work supported the presence of a larger number of melanoblast progenitors than proposed by Schaible or Mintz, and revealed extensive axial mixing between clones rather than sharply defined boundaries between descendants of melanoblast stem cells (14).

While it is technically not feasible to assess these theories using in situ hybridization analysis of early stage embryos, insights into the mechanisms of spotting have been obtained from analysis of embryonic melanoblast gene expression patterns. In situ hybridization analysis of the melanocyte genes Dct, Kit, Mitf, Matp, and Pmel17 show a non-uniform distribution of melanoblasts located laterally to the neural tube (15–20). Melanoblasts are first detected at embryonic day (E) 10.5 in the head and cervical region, and by E11.5 appear in the trunk and tail. Fewer melanoblasts are observed in the trunk region, compared with the head, cervical, and tail regions (Fig. 2A).

Figure 2.

The non-uniform and non-random distribution of melanoblasts in mouse embryos correlates with pigmentation patterns present in piebald spotting. (A) During embryonic development, melanoblasts show variable density along the rostral-caudal axis. Large numbers are seen around the developing optic cup, in the cervical region, and caudal to the hindlimbs, as shown by dark gray shading; fewer melanoblasts are seen in other regions of the head, and along the trunk, as shown by pale gray shading. (B) In the adult mouse, areas of hypopigmentation correlate to embryonic regions of low melanoblast density. White spotting is often seen as a head blaze and in the trunk, shown with gray shading. These areas, as well as white feet, white tail, and ventral spotting likely reflect an initially reduced melanoblast number and a longer distance of melanoblast migration to reach these anatomical regions.

Based on analysis of Dct expression in wild type and piebald-lethal embryos, it was proposed that many white spotting mutations that exhibit ‘piebald type’ spotting patterns reduce total melanocyte number, rather than selectively affecting stem cells responsible for populating a given area (21). Since the distribution of wild type melanoblasts is non-uniform along the rostral-caudal axis, areas remaining pigmented coincide with regions containing larger numbers of melanoblasts (head and tail), while areas of white spotting coincide with areas of lower melanoblast density (trunk) (Fig. 2B). Additional support of this hypothesis is seen in non-uniform patterns of melanoblast distribution as visualized with antibodies to c-kit and Dct, which showed larger numbers of melanoblasts in the head and caudal regions and smaller numbers in the trunk region at E10.5–12.5. In addition, injection of c-kit antibody during early embryonic development caused the appearance of spotting patterns that paralleled the regions of low melanoblast density (22).

While this theory can explain spotting patterns arising from mutation of the genes MITF, EDNRB, KIT, KITL, EDN3, PAX3, SOX10, all of which have a generalized reduction in pigmented areas, it cannot explain the etiology of regional spotting patterns in other mouse mutants, such as those displaying white belts (belted) or peppering of the coat (patchwork). In these mutants, other mechanisms must play a role. One example is seen in detailed analysis of patch (Ph) and rump white (Rw) mice, which have genomic alterations that result in misexpression of Kit in embryonic development. Ph mice display white spotting that is restricted to the belt region, and Rw mice display spotting extending from the lumbar region to the hindlegs and tail. Analysis of Ph and Rw mutant embryos with a Dct–LacZ transgenic reporter showed a nearly complete absence of melanoblasts in the trunk from E11.5 to 15.5. By E16.5, however, Jordan and Jackson identified a novel wave of melanoblast proliferation late in development that accounted for the population of a portion of these depleted trunk regions (23).

Additional embryological studies are needed to determine the precise number of melanoblasts that populate the entire mouse, and the effects of the various white spotting mutations on the development and distribution of melanoblasts. Insight into the functional pathways affected in these white spotting mutants has come from the identification of 10 known genes for spotted mice (Table 2). Functional analysis of these genes has shown them to be involved in diverse functions including transcriptional gene regulation, cell signaling to regulate of cell survival and proliferation, and cell migration.

Table 2.  Cell types affected in white spotting mutants
Cell type/structureGenesPhenotype
MelanocytesPax3, Mitf, Sox10, Ednrb, Edn3, Kit, Kitl, Slugh, Adamts20, Mcoln3Hypopigmentation
Cochlear melanocytesSox10, Kit, Kitl, Slugh, MitfDeafness
Cochlear hair cells, vestibular cellsMcoln3Deafness, circling
Retinal pigmented epitheliumMitfMicrophthalmia
Neural tubePax3Exencephaly, spina bifida
Enteric GangliaSox10, Ednrb, Edn3Megacolon
OsteoclastsMitfOsteopetrosis
Mast cellsMitfMast cell deficiency
Skeletal/craniofacial structuresPax3Wide nasal bridge
HeartPax3Developmental defects
Hematopoietic stem cellsKit, KitlAnemia
Primordial germ cellsKit, KitlSterility

Mitf– Microphthalmia

The first microphthalmia mutant (mi) was identified in the 1940s (24), but 50 yr passed before the gene responsible for mutations at this locus was cloned via transgenic insertion (25, 26). Mutations in the basic–helix–loop–helix–leucine–zipper protein transcription factor Mitf (Microphthalmia-associated Transcription Factor) result in abnormal melanocyte differentiation and reduction in melanocyte numbers (25, 26). Mitf is expressed in developing neural crest-derived melanoblasts and in the neuroepithelium-derived retinal pigmented epithelium (RPE) of the eye beginning at E10, and Mitf mutations cause loss of neural crest-derived melanoblasts and abnormal eye development (15, 16). A subset of Mitf mutations can also affect the development of osteoclasts and mast cells, causing defective bone resorption and mast cell deficiency (27). At least 23 different murine mutant alleles of Mitf have been identified, both recessive and semi-dominant, displaying a wide range of hypopigmentation, from moderate spotting to complete absence of mature melanocytes (28).

Mutation of human MITF is associated with Waardenburg Syndrome (WS) type IIa, a disorder characterized by regions of skin hypopigmentation (leukoderma), ocular pigmentation defects, and deafness caused by the loss of melanocytes of the inner ear (29). Tietz syndrome is associated with heterozygosity for a dominant negative form of MITF, where the phenotypes of hypopigmentation and deafness are more pronounced (30, 31).

Mitf is a key transcription factor in initiating transcription of many important melanocyte-specific genes and has been referred to as the melanocyte master regulatory gene (32, 33). The melanogenic enzymes tyrosinase (Tyr), Tyrosinase related protein 1 (Tyrp1), and Dopachrome tautomerase (Dct) contain E-box binding sites in their promoters that can be bound directly by Mitf (34–36). Interestingly, recent work suggests that Mitf is not the sole regulator of Tyr and Tyrp1. Transfection of Mitf into B16 melanoma cells does not increase Tyr and Tyrp1 expression, although expression of a dominant negative Mitf isoform reduces melanin synthesis and Tyr and Tyrp1 expression (37). Recent work has also shown that Mitf directly regulates the expression of the melanogenic proteins Pmel17 and Melan-a, but controls expression of the transporter Matp indirectly (18, 20, 38, 39). Not only does MITF play a role in expression of pigmentation genes, it also functions in cell survival by directly regulating Bcl2 in melanocytes and osteoclasts (40). Further studies are needed to understand the role of MITF in melanoblast survival, differentiation, and disease.

Slug

SLUG (murine Snai2 or Slugh) encodes a zinc finger transcription factor of the snail family (41) that is expressed in migratory neural crest cells in the mouse (42). The initial description of a targeted deletion of Snai2 was reported to not affect neural crest generation, migration, or development in mice (42), however another report of Snai2 deletion mice described strain-dependent phenotypes of white spotting in addition to pigment dilution (43). This group also described hematopoetic and gonadal defects in Snai2 knock-out mice, and proposed that Snai2 acts in the same cellular pathway as Kit and Kitl (43). Consistent with SLUG function in human development, three unrelated individuals with Piebaldism have recently been identified with deletion mutations in SLUG (44). Additionally, two individuals with WS2 have been identified that harbor deletions in SLUG, in which the phenotypes of deafness and abnormal iris pigmentation were seen (45). These data along with in vitro activation of the SLUG promoter by MITF suggest that SLUG functions downstream of MITF to regulate melanocyte development (45).

Sox10– Dominant Megacolon

The Dominant megacolon (Dom) mouse mutant, which exhibits white spotting and megacolon in heterozygotes and embryonic lethality in homozygotes, arose as a spontaneous mutation (46). Dom was shown to be the result of a point mutation that introduces a premature termination codon in the gene Sox10 (47, 48). Sox10 (Sry-like HMB box 10) is a member of the high mobility group (HMG) family of transcription factors, showing HMG domain homology to the testis determining factor SRY. A LacZ gene replacement of Sox10 recapitulates the phenotypes of hypopigmentation and megacolon seen in the original Dom allele (49), however, dominant negative effects have been proposed for truncated Sox10 alleles (50, 51).

Mutations in SOX10 are associated with the neurocristopathy Waardenburg–Shah syndrome (also called WS IV), in which pigmentary defects are seen in combination with features of Hirschsprung disease (52–55). This disorder is characterized by leukoderma, a white forelock and eyelashes, abnormal iris pigmentation, deafness, and enteric aganglionosis (56–58). Recently WS IV patients with additional neurological phenotypes have been shown to harbor SOX10 mutations (59, 60). Additional genotype-phenotype correlations will be required to understand the variation of this disease.

Sox10 is expressed in neural crest cells and in neural crest-derived structures during embryonic development (47, 48, 61–64), and is required for proper development and survival of neural crest-derived melanocyte and glial lineages [for review see (65)]. Sox10 acts to promote the survival and differentiation of melanocytes and glia, and a complete understanding of its interacting factors and downstream targets remains to be fully elucidated. In melanoblasts, Sox10 has been shown to strongly activate Mitf expression (50, 53) and to regulate Dct expression (51).

Pax3– Splotch

The Splotch mouse mutant (66) was shown to be due to a loss of function mutation in Pax3, a paired-box homeodomain transcription factor (67–69). Pax3 is one of a group of 9 Pax genes, which are highly conserved across species and contain a paired DNA binding domain. Mice that harbor heterozygous Pax3 mutations display ventral spotting, with occasional dorsal spotting on the back or tail. Homozygous Pax3 mutations are embryonic lethal, and display limb abnormalities, exencephaly, heart defects, and abnormal tail morphology.

In man, heterozygous loss of function PAX3 mutations are associated with WS types I and III. Several rare cases of homozygous PAX3 mutations causing WS III (or Klien–Waardenburg syndrome) have also been identified (70–72). Phenotypes of WS I include abnormal placement of the inner canthus of the eye, leading to a wide nasal bridge; hypopigmentation most often manifested as white blaze of hair at the forehead or leukoderma; heterochromia iridis; and deafness due to abnormal development of cochlear melanocytes. WS III includes skeletal abnormalities and cardiopulmonary defects together with the phenotypes seen in WS I. A missense mutation in PAX3 also is responsible for a dominant craniofacial-deafness-hand syndrome, in which craniofacial anomalies, hearing loss, and hand deformities are seen (73, 74).

Pax3 plays a key role in early neural crest development, and is crucial for the proper formation of nervous, muscular, and cardiovascular systems. Ectopic Pax3 expression triggers the formation of cellular aggregates and directs a mesenchymal to epithelial transition through induction of hepatocyte growth factor/scatter factor competence (75). Neural tube development requires Pax3 function, and one function for Pax3 is to regulate neural tube closure via inhibition of p53-mediated apoptosis (76). While the precise function of Pax3 in melanocyte development is not completely understood, one role for Pax3 is in transcriptional activation of Mitf in concert with Sox10 (50, 53, 77).

Mucolipin3 – Varitint-Waddler

The varitint-waddler (Va, VaJ) mouse displays phenotypes of white spotting, deafness, circling, reduced fertility, and reduced viability (78, 79). Recent work identified mutations in the predicted six transmembrane domain protein Mucolipin3 (Mcoln3) as responsible for the varitint-waddler phenotypes (80). Mcoln3 shows sequence similarity to the non-selective transient-receptor-potential ion channel family. To date, Mcoln3 expression has been demonstrated only in the sensory hair cells of the inner ear, which show developmental defects in Va mice. Mcoln3 shows punctate cytoplasmic staining by immunofluorescence in these cells, consistent with vesicular localization and suggesting Mcoln3 could affect vesicular trafficking in melanocytes (80). Future studies are needed to explore the role of this interesting gene in development and determine if it plays a role in human disease.

Kit– Dominant White Spotting and KitL – Steel

Dominant white spotting (W) and Steel (Sl) mutant mice arise spontaneously and occur frequently (81), with 76 and 44 alleles currently identified, respectively (28). Most alleles of W and Sl are semidominant, with heterozygotes showing head and belly spots. Homozygotes or compounds of two different mutations are often lethal, and those that survive are black-eyed white, sterile, and anemic. W is caused by a mutation in the Kit (also known as c-Kit) proto-oncogene that encodes a tyrosine kinase family receptor (82). Sl encodes Kit ligand (Kitl) and is also known as steel factor (SL), stem cell factor (SCF) or mast cell growth factor (MGF) (83–87). Kitl produces two forms, a transmembrane form and a soluble form (88). The soluble form is needed for melanocyte precursor dispersal on the lateral pathway and/or for their initial survival in the migration staging area, while the membrane-bound form is required for melanocyte precursor survival in the dermis (89). Activation of Kit by Kitl leads to receptor dimerization and autophosphorylation of specific tyrosine residues in the kinase domain. This activates downstream signal transduction molecules, such as MAPK, PI3 kinase, JAK/STAT, and Src family members and causes Kit-mediated cellular actions [for review see (90)]. Kit signaling is required for normal development of three migratory cell populations: melanocytes, blood cells, and primordial germ cells (19, 91, 92). In addition, it has been reported that Kit and Kitl mutants have defects in the intestinal pacemaker system, T-cell precursors, hippocampal learning and hearing (93–95).

Mutations in human KIT are associated with piebaldism, a rare autosomal dominant disorder. Human piebaldism is characterized by patches of white skin and white hair (96–100). Kit mutations are also associated with human gastrointestinal stromal tumors, urticaria pigmentosa, and aggressive mastocytosis where KIT proteins were constitutively activated (101, 102). To date, no aberrant development of KITL has been found in human patients.

Kit is mainly expressed in developing melanoblasts, hematopoietic stem cells, and primordial germ cells, while Kitl is expressed in tissues associated with the Kit-expressing cells and in the neural tube (17, 89, 103–105). Analyses of W and Sl mutant mice showed that Kit signaling functions are necessary for the survival and/or migration of melanoblasts (89, 106, 107). Likewise, the injection of Kit antibody into early mouse embryos blocks proper melanocyte development (22, 108), and the ectopic expression of Kitl promotes development of melanocyte precursors (109). Kit signaling also affects specific gene expression during melanocyte differentiation (110). In melanocytes and melanoma cells, KIT signaling leads to an increase in MITF phosphorylation, which is associated with an enhanced recruitment of the transcriptional coactivator P300/CBP and a concomitant stimulation of MITF transcriptional activity. This increase is transient and followed by rapid ubiquitination and proteasome-mediated degradation of MITF (111–114).

Ednrb– Piebald-Lethal and Edn3 – Lethal Spotting

Like mice with defects in Kitl/Kit signaling, mutations in Piebald-lethal (sl ) and Lethal spotting (ls) also disrupt normal melanocyte development (for reviews, see (4, 115). The recessive mutants sl and ls arose spontaneously (116). Mutation of the G-protein coupled endothelin receptor-B (Ednrb) is responsible for sl (117), and ls encodes Endothelin 3 (Edn3), a 21-residue peptide ligand for Ednrb (118). Activation of Ednrb by Edn3 leads to the activation of downstream signal transduction pathways, including PKC, CamKII, and MAPK (119). Ednrb signaling exerts pleiotropic effects on mouse development, and its function is required for the normal development of two migratory neural crest-derived cell populations, melanocytes and enteric ganglia (117, 118).

Mutations at the human EDNRB and EDN3 loci are associated with WS IV and Hirschsprung disease. As described above, WS IV is characterized by pigmentary defects and enteric aganglionosis (56–58). ABCD syndrome has been also identified as a homozygous nonsense mutation in the EDNRB gene (120). The role of EDNRB signaling in human disease has been recently reviewed (115).

Ednrb is expressed in developing neural crest cells, melanoblasts, and enteric ganglia in mouse embryos (117, 121–124), while Edn3 is expressed in tissues associated with Ednrb-expressing cells (121, 125). Studies of mice with mutant Ednrb alleles showed that Ednrb signaling functions are necessary for the development of melanoblasts and enteric neural precursors (21, 117, 118, 122). Likewise, transgenic expression of EDN3 rescues aganglionosis and piebaldism in lethal spotted mice (126). Ednrb is also expressed in other neural crest cells and non-neural crest-derived lineages, such as the neural tube (123). Experiments in Ednrb mutant mice have shown that Ednrb is not needed for melanoblast formation, but is needed for migration of melanoblasts and enteric neuroblast precursors prior to cell differentiation, between E10.5 and E12.5 (127), and is also required for melanoblast development in the epidermis beyond E12.5 (22).

Although Ednrb is expressed in unspecified neural crest cells and melanocyte precursors, it is not clear that it acts solely in a cell-autonomous manner (117, 124, 128). By cross-explantation of embryonic tissue and neural crest cells, it has been found that the melanoblasts of the hypomorphic Ednrbs (piebald) allele show increased survival on in vitro cultured wild type skin compared with mutant skin (129). To address the question of whether melanocyte development depends entirely on the cell-autonomous action of Ednrb, we performed a series of tissue recombination experiments in vitro using neural crest cell cultures from EdnrblacZ embryos. We have found that Ednrb plays a significant role during melanocyte differentiation by sequential cell-autonomous and non-autonomous actions (L. Hou et al., unpublished results).

Adamts20– Belted

The belted (bt) mouse mutant was first described in 1945 and is characterized by a dorsal white stripe located just proximal to hind limbs that is often accompanied by a belly spot (130). To date there have been 12 bt alleles identified, all exhibiting a very similar belting phenotype (28). Recently, mutations in the gene Adamts20, encoding a disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motif protein, have been identified in three bt alleles (131). Currently it is unclear if these three mutant alleles result in hypomorphic alleles or functional nulls.

Unlike many of the other genes mutated in the classic spotting mutants, Adamts20 is not expressed in migrating melanoblasts. Instead Adamts20 expression precedes expression of the melanoblast marker Dct and is found in the mesenchyme through which melanoblasts migrate (131). This expression pattern along with homology to a secreted metalloprotease family thought to interact with the extracellular matrix (132) is consistent with Adamts20 having a role in melanoblast migration. Since the Adamts20 pattern of expression is more expansive than the proximal hindlimb region that is unpigmented, it is unclear how the bt phenotype occurs. Functional redundancy with other metalloproteases outside of the belted region could compensate for mutated Adamts20, or the proximal hindlimb region could be susceptible to spotting due to a reduction in melanocyte number.

Initial experiments seeking to address what cell type was affected in the bt mouse using skin graft cultures from the predicted unpigmenting regions came to differing conclusions (133, 134). Mayer and Maltby used skin isolated from E12 to 12.5 day embryos and demonstrated that the melanocytes were present within the grafted skin, but unable to populate the hair follicle itself. They concluded that melanoblasts were able to occupy the entire skin graft from the white spotted area, but were unable to form differentiated melanocytes within the hair follicles in this region. They proposed that the Adamts20bt mutation acts at the hair follicle by either blocking melanoblast entry into follicles or by preventing melanoblast differentiation within the follicle. The expression of Adamts20 in developing dermis and hair follicles is consistent with Adamts20bt acting in a non-melanocyte intrinsic manner. However, similar experiments by Schaible using later aged embryos for skin explants were unable to confirm these findings (134). Further studies are needed to determine the timing and mode of action by which Adamts20 is needed for melanocyte development and how its mutation leads to the regional spotted phenotype.

Uncloned white spotting loci

Evidence exists that several additional genes remain to be discovered that are necessary for melanocyte development and survival. For example, family studies have identified two human WS2 loci for which the genes responsible have not been cloned (135–137). In mouse, at least 15 uncloned loci that cause white spotting have been identified and characterized (Table 3). Some, such as belly spot and tail and fleck, appear to affect lineages in addition to the melanocyte lineage, and are crucial for embryonic development based on embryonic lethality in homozygotes. Others, such as patchwork, appear to affect later stages of melanocyte development and survival, resulting in melanocyte loss upon entry into the hair follicle (138). New mutants with white spotting are being generated by large scale ENU mutagenesis screens, providing evidence that additional genes required for normal melanocyte development remain to be discovered and characterized (139). Identification of the genes responsible for these phenotypes will reveal more details of the genetics of neural crest-melanocyte development.

Table 3.  Uncloned white spotting mouse mutants. Data from (28)
Mouse mutantLocationHeterozygous phenotypesHomozygous phenotypes
  1. aDuring final preparation of this manuscript, Kumba was shown to be caused by a mutation in the transcription factor Zic2 (141).

Belted2  Dorsal belt
Belly spot and tail16 @31.5short kinked tail, ventral spotting, malocclusion, white feet, spine and eye abnormalitiesEmbryonic lethal
Belly spot with white toes Belly streak1Belly spot, white hind toes 
cryptorchidism with white spotting, deletion region5 @ 84 Cryptorchidism; dorsal/ventral spotting or belt
Extra toes spotting7 @58Polydactyly, head and belly spot 
Fleck Belly spot, white tail and hindpawsEmbryonic lethal
Flecking Head spot2 @102 Head, belly spot Head spot
Kumbaa14Belly spot, curly tail 
Patchwork10 Variegated pigmentation
Recessive spotting5 @ 42Thin belly spotHead, belly spot
Roan14 @ 61Microspotting 
Variable spotting9 @ 19Small belly spotBelly, feet, tail, head spots
White nose15 @ 46.35 Belly, nose spots
White toes7Belly spots, white hindpaws 

Summary

The cloned white spotting mutants demonstrate how readily mouse models can recapitulate human disorders, and will no doubt aid in identification of more genes responsible for human pigmentation anomalies. Once genes important for melanocyte development are identified, numerous tools are available for dissecting specific gene function. Early embryonic neural tubes can be dissected and cultured to allow neural crest cells to migrate and proliferate in vitro. This system allows visualization of developing melanocytes along with the potential for introduction of exogenous DNA through retroviral infections. These infections can be precisely targeted when transgenic neural tubes containing retroviral receptors expressed under the control of melanocyte lineage-specific promoters are used (140).

In addition, the genes responsible for white spotting mutants have been useful as diagnostic melanoma markers. Since they act early in development, targeting them for melanoma intervention may prove challenging; their downstream targets, however, could reveal modes of intervention for melanoma suppression or treatment. Therefore, elucidation of factors necessary for proper development and survival is crucial to understand both the normal and malignant biology of melanocytes. Our understanding of the processes involved in melanocyte development and disease will undoubtedly come from using these spotting mutants in a combination of experiments involving functional genomics, developmental genetics and embryology.

Ancillary