Genetic study of a family segregating Waardenburg-Shah syndrome

Abstract

Waardenburg-Shah syndrome (WS4, MIM_277580) is a congenital developmental disorder characterized by pigmentary abnormalities of the skin, eyes and hair, sensorineural deafness and intestinal aganglionosis (HSCR; Hirschsprung disease). Mutations in the coding regions of EDN3, EDNRB, or SOX10 account for 65-85% of the WS4 patients. These mutations are not fully penetrant, contributing to the phenotypic variability of WS4.

We screened these genes in a three-generation family (14 individuals; three members affected with HSCR only and one affected with “partial’ WS4 –iris heterochromia and HSCR-). A novel heterozygous missense mutation was identified in EDNRB. EDNRB encodes the EDNRB receptor, which is essential for the differentiation of the neural crest cells into melanocytes, enteric ganglia and Schwann cells. The mutation was present in four affected and three unaffected family members. In the EDNRB isoforms 1 and 2, the mutation results in the replacement of the translation initiation codon methionine (Met) with a valine (Val) and such replacement (M1V) would theoretically abolish the use of the translation initiation codon. However, in EDNRB isoform 3, the replacement is at Met91 (M91V) and is predicted benign. Since different EDNRB transcripts are expressed concomitantly in the still developing newborn's gut, we theorized that the intra-familial variability of the phenotype could be related to the expression ratio between benign and damaging isoforms.

We examined the consequences of M1V or M91V in their respective isoforms. Constructs containing either wild-type cDNA of isoform 1 and 3 or their mutated counterpart were transiently transfected into Human Embryonic Kidney 293 cell (HEK293). Confocal and immunoblot experiments showed that EDNRB M1V generated a shortened protein (starting from Met46); the wild-type-EDNRB isoform 3 or its mutant (p.M91V) were only found in the cytosol. Although EDNRB M1V was able to generate a shorter protein, the later failed to translocate onto the cell membrane, theoretically, affecting signal-transduction. Isoform 3 did not seem to play a role as cellular receptor.

We also identified a c.-248G/A rare change at the 5’-untranslated region (5’UTR) of EDN3 (EDNRB ligand) which was predicted to affect translation efficiency. The presence of this variant in affected individuals but not in healthy carriers of the EDNRB mutation, suggests that both variants are necessary for the disease manifestation. Variations within the disease phenotype may be due to each individual’s genetic background.

To identify other susceptibility loci, we carried out whole-genome linkage scan in this family using a high density SNP assay. Merlin software was used for parametric and non-parametric linkage. A susceptibility locus on chromosome 4q13.3-q24 was identified by both nonparametric and parametric linkage analyses, with LOD scores of 1.204 and 1.7109 respectively. Haplotype analysis refined the region to a 27.76 cM interval, in which genes involved in neuron development reside.

To conclude, the novel EDNRB M1V mutation in this family may lead to HSCR and/or WS4 when in conjunction with other genetic lesions, such as the EDN3 5’UTR rare variation and/or a not yet identified susceptibility locus on chromosome 4q13.3-q24.