Waardenburg Syndrome, Type 1

A number sign (#) is used with this entry because Waardenburg syndrome type 1 (WS1) is caused by heterozygous mutation in the PAX3 gene (606597) on chromosome 2q36.

Waardenburg syndrome type 3 (WS3; 148820) is also caused by mutation in the PAX3 gene.

Description

Waardenburg syndrome type 1 is an autosomal dominant auditory-pigmentary syndrome characterized by pigmentary abnormalities of the hair, skin, and eyes; congenital sensorineural hearing loss; and 'dystopia canthorum,' the lateral displacement of the ocular inner canthi (reviews by Read and Newton, 1997, Tamayo et al., 2008, and Pingault et al., 2010).

Clinical Variability of Waardenburg Syndrome Types 1-4

Waardenburg syndrome is an auditory-pigmentary syndrome characterized by pigmentary abnormalities of the hair, including a white forelock and premature graying; pigmentary changes of the iris, such as heterochromia irides and brilliant blue eyes; and congenital sensorineural hearing loss. Waardenburg syndrome has been classified into 4 main phenotypes. WS type 1 is distinguished by the presence of dystopia canthorum. WS type 2 (WS2; see 193510) is distinguished from type 1 by the absence of dystopia canthorum. WS type 3 (WS3; 148820) has dystopia canthorum and upper limb abnormalities. WS type 4 (WS4; see 277580), also known as Waardenburg-Shah syndrome, has the additional feature of Hirschsprung disease (reviews by Read and Newton, 1997 and Tamayo et al., 2008).

Genetic Heterogeneity of All Types of Waardenburg Syndrome

Waardenburg syndrome is genetically heterogeneous. WS1 and WS3 are both caused by mutation in the PAX3 gene. See WS2A (193510) for a discussion of genetic heterogeneity of WS type 2, and WS4A (277580) for a discussion of genetic heterogeneity of WS type 4.

Clinical Features

Waardenburg (1951) first delineated the syndrome that bears his name, describing it as a disorder combining anomalies of the eyelids, eyebrows, and nasal root with congenital deafness.

Feingold et al. (1967) noted that the white forelock may be present at birth and later disappear in patients with Waardenburg syndrome. Arias (1980) suggested that visceral and cranial malformations (such as Hirschsprung megacolon) are associated with Waardenburg syndrome type 1.

Yoshino et al. (1986) evaluated the incidence of dystopia canthorum in a 3-generation family with Waardenburg syndrome type 1 and concluded that it is the most frequently expressed sign of the condition.

Da-Silva (1991) reported 2 large multigenerational Brazilian kindreds with WS type 1. The total number of affected individuals was 73. The major manifestations were telecanthus, which was the only constant anomaly, prominent nasal root (78%), round or square tip of nose, hypoplastic alae, smooth philtrum, bushy eyebrows with synophrys (76%), sensorineural deafness (67%), heterochromia or hypoisochromia iridis, hypopigmented ocular fundus, white forelock (29%), premature graying (44%), and hypopigmented skin lesions (55%).

Winship and Beighton (1992) reviewed phenotypic variation of Waardenburg syndrome on the basis of an analysis of 68 affected children.

In a craniofacial anthropometric study of 51 WS type 1 individuals, da-Silva et al. (1993) concluded that the most discriminating parameters were, from clinical measurements, increased intercanthal distance and decreased philtrum length, and, from roentgenographic measurements, decreased nasal bone length and increased lower facial height.

Read and Newton (1997) provided a review of the clinical features and molecular basis of Waardenburg syndrome and other auditory pigmentary syndromes.

From a systematic literature search, Song et al. (2016) determined that the prevalence of hearing loss in patients with Waardenburg syndrome differed according to the genotype: the prevalence in those with WS1 due to PAX3 mutations was 52.3%.

Other Features

Giacoia and Klein (1969) documented the occurrence of bilateral cleft lip in Waardenburg syndrome. Arias (1971) observed a black forelock in place of white forelock.

Goodman et al. (1988) observed absence of vagina and of right-sided uterine adnexa in an 18-year-old woman with WS1. They postulated that these are related to Waardenburg syndrome because of altered invasion of neurons in early embryogenesis.

Carezani-Gavin et al. (1992) reported a patient with Waardenburg syndrome type 1 who also had a meningomyelocele at the L3-S1 level and Arnold-Chiari malformation. Characteristic features of WS1 were present, including hearing loss, dystopia canthorum, broad nasal root, and narrow nasal tip. There was a family history of the disorder, but none of the other affected individuals had neural tube defects. In the family reported by Carezani-Gavin et al. (1992), Hoth et al. (1993) identified a heterozygous mutation in the PAX3 gene (R56L; 606597.0014).

Chatkupt et al. (1993) stated that spina bifida had been noted in at least 4 patients with Waardenburg syndrome. They reported the cases of brothers with both Waardenburg syndrome and lumbosacral myelomeningocele. The mother had features of Waardenburg syndrome. Spina bifida occurs with the 'Splotch' mutation, which molecular studies indicate is the homologous disorder in the mouse (see 606597).

Inheritance

Waardenburg syndrome type 1 is an autosomal dominant disorder (Pardono et al., 2003).

Jones et al. (1975) found evidence of paternal age effect in new mutations for autosomal dominant Waardenburg syndrome.

Kapur and Karam (1991) described a family in which 3 children with this disorder were born to normal, unrelated parents. Germline mosaicism was postulated.

Diagnosis

Laestadius et al. (1969) provided normal standards for the measurement of inner canthal and outer canthal distance. Standards were also presented by Christian et al. (1969).

In place of the measurement of inner canthal distance, the Waardenburg Consortium (Farrer et al., 1992) recommended the W index: a composite measure including the inner canthal, inner pupillary, and outer canthal distances. Normal and dystopic subjects had W values (mean +/- SD) of 1.76 +/- 0.16 and 2.61 +/- 0.19, respectively (Newton, 1989); the Waardenburg Consortium recommended a threshold W value of 2.07.

Pardono et al. (2003) studied 59 patients with Waardenburg syndrome from 37 families (30 with type 1, 21 with type 2, and 8 isolated individuals without telecanthus). All patients were examined for the presence of 8 cardinal diagnostic signs: telecanthus, synophrys, iris pigmentation disturbances, partial hair albinism, hearing impairment, hypopigmented skin spots, nasal root hyperplasia, and lower lacrimal dystopia. The authors noted that some patients with type 1 may not have dystopia canthorum, but that it is present in 95 to 99% of patients with WS type 1. Using their own data as well as those collected from the literature, the authors estimated the frequencies of these 8 cardinal signs of Waardenburg syndrome based on a sample of 461 affected individuals with type 1 and 121 with type 2.

Cytogenetics

Ishikiriyama et al. (1989, 1989) reported a 20-month-old boy with dystopia canthorum, sensorineural deafness, heterochromia iridis, partially albinotic ocular fundi, and partial leukoderma. Cytogenetic studies showed a paracentric inversion (2)(q35q37.3); his parents had normal chromosomes. Ishikiriyama et al. (1989) suggest that the gene for Waardenburg syndrome type 1 may be located on chromosome 2q35 or 2q37.3. Kirkpatrick et al. (1992) described WS type 1 in a child with del(2)(q35q36.2). Because of this report and that of Lin et al. (1992) of deletion of 2q37 without features of WS1, Ishikiriyama (1993) concluded that the WS1 gene is located at 2q35.

Mapping

On the basis of an analysis of mouse and hamster mutants as models for Waardenburg syndrome(s), Asher and Friedman (1990) predicted that the gene(s) would be found to be on chromosome 2q near fibronectin-1 (135600), on chromosome 3p near the protooncogene RAF1 (164760) or 3q near rhodopsin, (RHO; 180380), or on chromosome 4p near the protooncogene KIT (164920).

Foy et al. (1990) demonstrated linkage of the Waardenburg syndrome to placental alkaline phosphatase (ALPP; 171800), which had previously been assigned to 2q37; the peak lod score was 4.76 at a recombination fraction of 0.023. These findings suggest that the distal breakpoint responsible for the paracentric inversion is at the site of the Waardenburg syndrome, namely, 2q37.3. This region of chromosome 2 is homologous to mouse chromosome 1, which contains the 'Splotch' locus (Sp) (PAX3). This patchy pigment mutation is accompanied by a malformation of the inner ear and severe CNS malformation in the homozygote. Whether the heterozygote is deaf is unclear.

In a family with 11 affected individuals spanning 4 generations, Asher et al. (1991) confirmed the assignment of WS1 to 2q. No recombination was found with ALPP at 2q37 or with FN1 (135600) at 2q34-q36. Grundfast et al. (1991) found no obligatory crossovers between WS1 and ALPP. For a collection of families, they obtained a maximum lod of 12.5 at theta = 0.31 for the WS1/ALPP linkage.

Molecular Genetics

Tassabehji et al. (1992) identified variations in the PAX3 gene in 6 of 17 unrelated patients with Waardenburg syndrome type 1 using primers to amplify exons followed by testing for heteroduplex formation on polyacrylamide gels. No variants were seen in any exon in 50 normal controls. In 3 families that were tested, the variant was found to be familial in 2 and apparently de novo in the third. The variant bands showed perfect linkage to WS in the families studied. One family was found to have a heterozygous 18-bp deletion in the central region of exon 2, resulting in loss of amino acids 29 to 34 (606597.0001).

Simultaneously and independently, Baldwin et al. (1992) identified a heterozygous mutation in the PAX3 gene (P50L; 606597.0002) in affected members of a large Brazilian family with Waardenburg syndrome type 1 reported by da-Silva (1991). There were 49 affected persons in 6 generations, and more than 78% of the affected individuals had hearing loss.

Baldwin et al. (1995) described 10 additional mutations in the PAX3 gene in families with WS type 1. Eight of these mutations were in a region of PAX3 where only 1 mutation had previously been described. Taken together with previously reported mutations, these mutations covered essentially the entire PAX3 gene. All but 1 of the mutations were 'private;' only 1 mutation had been reported in 2 apparently unrelated families. Baldwin et al. (1995) also cataloged 16 previously reported mutations and 5 chromosomal abnormalities affecting the 2q35 region that were associated with WS.

Modifier Genes

Work in the hamster model for Waardenburg syndrome suggested to Asher and Friedman (1990) that modifier genes may account for the intrafamilial variation in phenotype in Waardenburg syndrome.

Reynolds et al. (1996) sought to determine whether the W-index is influenced primarily by allelic variation in the PAX3 disease gene or other major loci, by polygenic background effects, or by all of these potential sources of genetic variation. They studied both WS1-affected individuals and their WS1 unaffected relatives. After adjustment of the W-index for WS1 disease status, segregation analyses by the regression approach indicated major-locus control of this variation, although residual parent-offspring and sib-sib correlations were consistent with additional (possibly polygenic) effects. Separate analyses of WS1-affected and WS1-unaffected individuals suggested that epistatic interactions between disease alleles at the PAX3 WS1 locus and a second major locus influenced variation in dystopia canthorum. Reynolds et al. (1996) suggested that their approach should be applicable for assessing the 'genetic architecture' of variation associated with other genetic diseases.

While mutations in PAX3 seem to be responsible for most, if not all, WS1 cases, it is not clear what accounts for the reduced penetrance of deafness. Stochastic events during development may be the factors that determine whether a person with a PAX3 mutation will be congenitally deaf or not. Alternatively, genetic background, nonrandom environmental factors, or both may be significant. Morell et al. (1997) compared the likelihood for deafness in affected subjects from 24 families with PAX3 mutations and in 7 of the families originally described by Waardenburg. They found evidence that stochastic variation alone does not explain the differences in penetrance of deafness among WS families. Their analyses suggested that genetic background in combination with certain PAX3 alleles may be important factors in the etiology of deafness in WS1.

Genotype/Phenotype Correlations

In a series of patients with Waardenburg syndrome, Tassabehji et al. (1994) found a number of previously unidentified PAX3 mutations. These included a chromosomal deletion, a splice site mutation, and an amino acid substitution that closely corresponded to the molecular changes seen in the 'Splotch-retarded' and 'Splotch-delayed' mouse mutants, respectively. These mutations confirmed that Waardenburg syndrome is produced by gene dosage effects and showed that the phenotypic differences between 'Splotch' mice and humans with Waardenburg syndrome are caused by differences in genetic background rather than different primary effects of the mutations.

Chalepakis et al. (1994) studied the functional consequence of the mutations described in 606597.0001 and 606597.0006 on DNA binding and compared the results with those in the 'Splotch' mouse. Combining the phenotypic features of heterozygous mutants and considering that molecular defects ranging from single point mutations to large deletions cause similar phenotypes, they excluded the possibility that the mutated allele in heterozygotes interferes with the function of the wildtype allele. Contrariwise, they considered both WS and 'Splotch' mutants to represent loss-of-function mutations.

Zlotogora et al. (1995) reported a large kindred in which many individuals had Waardenburg syndrome type 1 caused by a heterozygous mutation in the PAX3 gene (S84F; 606597.0009). However, there was 1 child, born of consanguineous parents, who had a severe phenotype consistent with WS type 3 (148820): this patient was found to be homozygous for the S84F mutation. The child presented with dystopia canthorum, partial albinism, and very severe upper limb defects. Since all Pax3 mutations in mice lead to severe neural tube defects and intrauterine or neonatal death, the survival of the homozygote in this case and the absence of neural tube defects were unexpected.

Ayme and Philip (1995) observed exencephaly in a fetus with possible homozygous Waardenburg syndrome. The fetus was the product of a mating between a gypsy brother and sister, both of whom had Waardenburg syndrome.

Baldwin et al. (1995) stated that their analysis of a total of 30 PAX3 mutations causing WS type 1 or type 3 demonstrated little correlation between genotype and phenotype. Deletions of the entire PAX3 gene resulted in phenotypes indistinguishable from those associated with single-base substitutions in the paired domain or homeodomain of the gene. Moreover, 2 similar mutations in close proximity could result in significantly different phenotypes, WS type 1 in 1 family and WS type 3 in another.

DeStefano et al. (1998) assessed the relationship between phenotype and gene defect in 48 families containing 271 individuals with WS collected by members of the Waardenburg Consortium. They grouped the 42 unique mutations previously identified in the PAX3 gene in these families into 5 mutation categories: amino acid substitution in the paired domain, amino acid substitution in the homeodomain, deletion of the ser-thr-pro-rich region, deletion of the homeodomain and the ser-thr-pro-rich region, and deletion of the entire gene. This classification of mutations was based on the structure of the PAX3 gene and was chosen to group mutations predicted to have similar defects in the gene product. They found that odds for the presence of eye pigment abnormality, white forelock, and skin hypopigmentation were 2, 8, and 5 times greater, respectively, for individuals with deletions of the homeodomain and the pro-ser-thr-rich region compared to individuals with an amino acid substitution in the homeodomain. Odds ratios that differed significantly from 1.0 for these traits may indicate that the gene products resulting from different classes of mutations act differently in the expression of WS. Although a suggestive association was detected for hearing loss with an odds ratio of 2.6 for amino acid substitution in the paired domain compared with amino acid substitution in the homeodomain, this odds ratio did not differ significantly from 1.0.

Pathogenesis

Watanabe et al. (1998) showed that PAX3 transactivates the MITF (156845) promoter. They further showed that PAX3 proteins associated with WS1 in either the paired domain or the homeodomain failed to recognize and transactivate the MITF promoter. These results provided evidence that PAX3 directly regulates MITF, and suggested that the failure of this regulation due to PAX3 mutations causes the auditory-pigmentary symptoms in at least some individuals with WS1.

Bondurand et al. (2000) showed that SOX10 (602229), in synergy with PAX3, strongly activates MITF expression in transfection assays. Transfection experiments revealed that PAX3 and SOX10 interact directly by binding to a proximal region of the MITF promoter containing binding sites for both factors. Mutant SOX10 or PAX3 proteins failed to transactivate this promoter, providing further evidence that the 2 genes act in concert to directly regulate expression of MITF. In situ hybridization experiments carried out in the dominant megacolon (Dom) mouse confirmed that SOX10 dysfunction impaired Mitf expression as well as melanocytic development and survival. The authors hypothesized that interaction between 3 of the genes that are altered in WS could explain the auditory/pigmentary symptoms of this disease.

Population Genetics

Waardenburg syndrome has been described in American blacks (Hansen et al., 1965) and in Maoris (Houghton, 1964) as well as in Europeans.

In the state of South Australia, Waardenburg syndrome is a leading cause of deafness and has a position comparable to porphyria in South Africa, having been introduced by early settlers who had many descendants (Fraser, 1967).

An affected Chinese family was reported by Chew et al. (1968).

In a report from a consortium, Grundfast et al. (1991) concluded that about 56% of WS families are linked to 2q markers. Farrer et al. (1992) estimated that the WS1 gene on chromosome 2 was responsible for approximately 45% of the 44 families in their sample.

Nomenclature

Arias (1971) proposed the delineation of Waardenburg syndrome type 1 and 2 based on the presence or absence of dystopia canthorum. Thus, descriptions of the disorder before this time refer only to 'Waardenburg syndrome' and not to a specific subtype (Pardono et al., 2003; Tamayo et al., 2008).

History

Simpson et al. (1974) and Arias et al. (1975) found a weak suggestion of linkage between Waardenburg syndrome and the ABO locus (110300), known to be located in 9q34. However, Read et al. (1989) excluded linkage to ABO. In a study of 2 large kindreds in northeastern Brazil, da-Silva et al. (1990) could not confirm linkage to ABO. Since a plausible mouse model is 'Steel' (Sl), a dominant mutation on mouse chromosome 10 closely linked to Pep-2, Read et al. (1989) studied polymorphic probes for loci on human chromosome 12 close to PEPB (169900), the human homolog, in 7 families. They excluded a sizable region of 12q as the site of this gene.

Quoting William Harvey's famous observation that 'Nature is nowhere accustomed more openly to display her secret mysteries than in cases where she shows traces of her workings apart from the beaten path,' Duyk et al. (1992) reviewed the forms of deafness, syndromal and nonsyndromal, for which linkage has been established.