Waardenburg syndrome type 2 (WS2) is a dominantly inherited syndrome of hearing loss and pigmentary disturbances. We recently mapped a WS2 gene to chromosome 3p12.3-p14.1 and proposed as a candidate gene MITF, the human homologue of the mouse microphthalmia (mi) gene. This encodes a putative basic-helix-loop-helix-leucine zipper transcription factor expressed in adult skin and in embryonic retina, otic vesicle and hair follicles. Mice carrying mi mutations show reduced pigmentation of the eyes and coat, and with some alleles, microphthalmia, hearing loss, osteopetrosis and mast cell defects. Here we show that affected individuals in two WS2 families have mutations affecting splice sites in the MITF gene.

Waardenburg's syndrome (WS) is an autosomal dominant combination of deafness and pigmentary disturbances, probably caused by defective function of the embryonic neural crest. We have mapped one gene for WS to the distal part of chromosome 2. On the basis of their homologous chromosomal location, their close linkage to an alkaline phosphatase gene, and their related phenotype, we suggested that WS and the mouse mutant Splotch might be homologous. Splotch is caused by mutation in the mouse Pax-3 gene. This gene is one of a family of eight Pax genes known in mice which are involved in regulating embryonic development; each contains a highly conserved transcription control sequence, the paired box. Here we show that some families with WS have mutations in the human homologue of Pax-3. Mutations in a related gene, Pax-6, which, like Pax-3, has both a paired box and a paired-type homeobox sequence, cause the Small-eye mutation in mice and aniridia in man. Thus mutations in the Pax genes are important causes of human developmental defects.

In Williams syndrome (WS), a deletion of approximately 1.5 Mb on one copy of chromosome 7 causes specific physical, cognitive, and behavioral abnormalities. Molecular dissection of the phenotype may be a route to identification of genes important in human cognition and behavior. Among the genes known to be deleted in WS are ELN (which encodes elastin), LIMK1 (which encodes a protein tyrosine kinase expressed in the developing brain), STX1A (which encodes a component of the synaptic apparatus), and FZD3. Study of patients with deletions or mutations confined to ELN showed that hemizygosity for elastin is responsible for the cardiological features of WS. LIMK1 and STX1A are good candidates for cognitive or behavioral aspects of WS. Here we describe genetic and psychometric testing of patients who have small deletions within the WS critical region. Our results suggest that neither LIMK1 hemizygosity (contrary to a previous report) nor STX1A hemizygosity is likely to contribute to any part of the WS phenotype, and they emphasize the importance of such patients for dissecting subtle but highly penetrant phenotypes.

One hundred and thirty-four families or individuals with auditory-pigmentary syndromes such as Waardenburg syndrome (WS) or probable neurocristopathies were screened for mutations in the PAX3 and MITF genes. PAX3 mutations were found in 20/25 families with definite Type 1 WS and 1/2 with Type 3 WS, but in none of 23 with definite Type 2 WS or 36 with other neurocristopathies. The PAX3 mutations included substitutions of conserved amino acids in the paired domain or the homeodomain, splice-site mutations, nonsense mutations and frame-shifting insertions or deletions. No phenotype-genotype correlations were noted within WS1 families. With MITF, mutations likely to affect protein function were found in seven families, five of which had definite Type 2 WS. We conclude that Type 1 and Type 3 WS are allelic and are normally caused by loss of function mutations in PAX3; that Type 2 WS is heterogeneous, with about 20% of cases caused by mutations in MITF, and that individuals with auditory, pigmentary or neural crest syndromes which do not fit stringent definitions of Waardenburg syndrome are unlikely to have mutations in either the PAX3 or MITF genes. The molecular pathology of MITF/microphthalmia mutations appears to be different in humans and mice, with gene dosage having more significant effects in humans than in the mouse.

Supravalvular aortic stenosis (SVAS) is

We describe the complete exon-intron structure of the human elastin (ELN) gene located at chromosome 7q11.23. There are 34 exons occupying approximately 47 kb of genomic DNA. All exons are in-frame, allowing exon skipping without disrupting the reading frame. Microsatellites are located in introns 17 and 18. Deletions of all or large parts of the ELN gene have been previously reported in two patients with supravalvular aortic stenosis (SVAS), and SVAS is also a frequent feature of Williams syndrome, where patients are hemizygous for ELN. We list primer pairs for amplifying each exon, with flanking intron, from genomic DNA to allow detection of point mutations in the ELN gene. We show that some patients with isolated SVAS have point mutations that are predicted to lead to premature chain termination. Knowledge of the genomic structure will allow more extensive mutation screening in genomic DNA of patients with SVAS and other conditions.

Elastin is the protein responsible for the characteristic elastic properties of many tissues including the skin, lungs and large blood vessels. Loss-of-function mutations in the elastin gene are known to cause the heart defect supravalvular aortic stenosis (SVAS). We and others have identified deletions, nonsense mutations and splice site mutations in SVAS patients that abolish the function of one elastin gene. We have now identified an elastin mutation in a patient with a completely different phenotype, the rare autosomal dominant condition cutis laxa. A frameshift mutation in exon 32 of the elastin gene is predicted to replace 37 amino acids at the C-terminus of elastin by a novel sequence of 62 amino acids. mRNA and immunoprecipitation studies show that the mutant allele is expressed. Electron microscopy of skin sections shows abnormal branching and fragmentation in the amorphous elastin component, and immunocytochemistry shows reduced elastin deposition in the elastic fibres and fewer microfibrils in the dermis. These findings suggest that the mutant tropoelastin protein is synthesized, secreted and incorporated into the elastic matrix, where it alters the architecture of elastic fibres. Interference with cross-linking would reduce elastic recoil in affected tissues and explain the cutis laxa phenotype.

No abstract