Molecular analysis expands the spectrum of phenotypes associated with GLI3 mutations.

A range of phenotypes including Greig cephalopolysyndactyly and Pallister-Hall syndromes (GCPS, PHS) are caused by pathogenic mutation of the GLI3 gene. To characterize the clinical variability of GLI3 mutations, we present a subset of a cohort of 174 probands referred for GLI3 analysis. Eighty-one probands with typical GCPS or PHS were previously reported, and we report the remaining 93 probands here. This includes 19 probands (12 mutations) who fulfilled clinical criteria for GCPS or PHS, 48 probands (16 mutations) with features of GCPS or PHS but who did not meet the clinical criteria (sub-GCPS and sub-PHS), 21 probands (6 mutations) with features of PHS or GCPS and oral-facial-digital syndrome, and 5 probands (1 mutation) with nonsyndromic polydactyly. These data support previously identified genotype-phenotype correlations and demonstrate a more variable degree of severity than previously recognized. The finding of GLI3 mutations in patients with features of oral-facial-digital syndrome supports the observation that GLI3 interacts with cilia. We conclude that the phenotypic spectrum of GLI3 mutations is broader than that encompassed by the clinical diagnostic criteria, but the genotype-phenotype correlation persists. Individuals with features of either GCPS or PHS should be screened for mutations in GLI3 even if they do not fulfill clinical criteria.

The origin of EFNB1 mutations in craniofrontonasal syndrome: frequent somatic mosaicism and explanation of the paucity of carrier males.

Craniofrontonasal syndrome (CFNS) is an X-linked disorder that exhibits a paradoxical sex reversal in phenotypic severity: females characteristically have frontonasal dysplasia, craniosynostosis, and additional minor malformations, but males are usually mildly affected with hypertelorism only. Despite this, males appear underrepresented in CFNS pedigrees, with carrier males encountered infrequently compared with affected females. To investigate these unusual genetic features of CFNS, we exploited the recent discovery of causative mutations in the EFNB1 gene, which encodes ephrin-B1, to survey the molecular alterations in 59 families (39 newly investigated and 20 published elsewhere). We identified the first complete deletions of EFNB1, catalogued 27 novel intragenic mutations, and used Pyrosequencing and analysis of nearby polymorphic alleles to quantify mosaic cases and to determine the parental origin of verified germline mutations. Somatic mosaicism was demonstrated in 6 of 53 informative families, and, of 17 germline mutations in individuals for whom the parental origin of mutation could be demonstrated, 15 arose from the father. We conclude that the major factor accounting for the relative scarcity of carrier males is the bias toward mutations in the paternal germline (which present as affected female offspring) combined with reduced reproductive fitness in affected females. Postzygotic mutations also contribute to the female preponderance, whereas true nonpenetrance in males who are hemizygous for an EFNB1 mutation appears unusual. These results highlight the importance of considering possible origins of mutation in the counseling of families with CFNS and provide a generally applicable approach to the combined analysis of mosaic and germline mutations.

De novo deletion within the telomeric region flanking the human alpha globin locus as a cause of alpha thalassaemia.

We have identified and characterized a Scottish individual with alpha thalassaemia, resulting from a de novo 48 kilobase (kb) deletion from the telomeric flanking region of the alpha globin cluster which occurred as a result of recombination between two misaligned repetitive elements that normally lie approximately 83 kb and 131 kb from the 16p telomere. The deletion removes two previously described putative regulatory elements (HS-40 and HS-33) but leaves two other elements (HS-10 and HS-8) intact. Analysis of this deletion, together with eight other published deletions of the telomeric region, showed that they all severely downregulated alpha globin expression. Together they defined a 20.4-kb region of the human alpha cluster, which contains all of the positive cis-acting elements required to regulate alpha globin expression. Comparative analysis of this region with the corresponding segment of the mouse alpha globin cluster demonstrated conserved non-coding sequences corresponding to the putative regulatory elements HS-40 and HS-33. Although the role of HS-40 as an enhancer of alpha globin expression is fully established, these observations suggest that the role of HS-33 and other sequences in this region should be more fully investigated in the context of the natural human and mouse alpha globin loci.