In a series of 50 patients with a broad array of vertebral malformations encompassing the entire length of the spinal column, we only identified a non-synonymous variant in exon 3 of WNT3A, resulting in substitution of alanine for threonine. Our observation of this variant with an overall frequency of 0.35% in the control population studied, and a frequency of 1.1% of African chromosomes included in the control population suggest and the observation that the patient's clinically asymptomatic father harbors the same variant strongly suggest that this variant may represent a polymorphism in the African population. Based on the data obtained from our pilot study, mutations in the WNT3A gene do not appear to be a common etiologic factor in the development of congenital vertebral malformations.
Patient heterogeneity for vertebral malformations and heterogeneity of candidate genes represent major limitations in genetic analyses. It is possible that mutations in WNT3A would have been identified if we had sequenced a cohort of patients characterized by having caudal vertebral malformation phenotypes. Because congenital vertebral malformations represent a rare condition, collection of a phenotypically homogeneous sample is not practical and would require a greater amount of time. DNA sequence analysis was limited to the WNT3A promotor, exons, and intron-exon boundaries. Sequence variants in the introns that could potentially contribute to abnormal transcription would not have been detected. Copy number variations in WNT3A will not have been detected by this methodology. Other genes that regulate the expression of WNT3A have not been examined.
Evidence for genetically heterogeneous etiologies of vertebral malformations includes the identification of mutations in DLL3
in patients with spondylocostal dysostosis [14
] and mutations identified in JAG 1
in patients with Alagille syndrome [16
]. Recently, a few DNA coding sequence alterations with uncertain clinical significance in PAX1
have been identified in a cohort of patients with phenotypically characterized vertebral malformations [17
]. Our inability to detect mutations in WNT3A
, as well as other candidate genes studied, may reflect differences in the regulation of somitogenesis in humans as compared to mice. It is possible that mutations in WNT3A
are lethal in humans. Although our pilot study included 5 stillbirth samples and no mutations in WNT3A
were detected, a larger cohort of stillbirth samples with vertebral malformations would be needed to determine whether mutations in WNT3A
are prevalent among stillbirths with congenital vertebral malformations. Tissue mosaicism for a specific mutation and/or epigenetic factors may have an influence in the development of vertebral malformations in humans. Vertebral malformations may have greater environmental components than genetic components to their origin. A relationship between abnormal distribution of the intersegmental arteries and vertebral malformations was observed in 4 of 11 human embryos and fetuses with congenital vertebral malformations, suggesting the possibility that abnormalities in vascularization during the resegmentation process may contribute towards the development of congenital vertebral malformations [21
Clinical relevance in identifying mutations associated with the development of vertebral malformations and congenital scoliosis is to develop a molecular classification system that could provide prognostic information regarding the severity of progression of congenital scoliosis and the presence of associated birth defects to practitioners and families. Although congenital scoliosis and idiopathic scoliosis represent dichotomous processes, namely a failure to develop normal vertebral structures for the former and a failure to lose normal vertebral structures for the latter, there is evidence that similar genetic factors may play a role in both conditions [22
]. Identification of gene mutations which are associated with the development of congenital scoliosis may also help improve our understanding of the pathogenesis of idiopathic scoliosis.
In summary, we have been unable to detect mutations in the WNT3A gene in a pilot study of 50 patients with vertebral malformations encompassing the entire spine. This study illustrates the problems posed by attempting to design studies aimed at identification of patterning genes contributing towards the development of vertebral malformations. Future studies aimed at investigating the etiology of congenital vertebral malformations should be aimed at the study of additional candidate genes in a more homogeneous patient population.