Marfan Syndrome (MFS) is an autosomal dominant disorder of the connective tissue due to mutations of Fibrillin-1 gene (FBN1) in more than 90% of cases and Transforming Growth Factor-Beta-Receptor2 gene (TGFB2R) in a minority of cases. Genotyping is relevant for diagnosis and genotype-phenotype correlations. We describe the FBN1 genotypes and related phenotypes of 81 patients who were referred to our attention for MFS or Marfan-like phenotypes. Patients underwent multidisciplinary pertinent evaluation in the adult or paediatric setting, according to their age. The diagnosis relied on Ghent criteria. To optimise DHPLC analysis of the FBN1 gene, all coding regions of the gene were directly sequenced in 19 cases and 10 controls: heterozygous amplicons were used as true positives. DHPLC sensitivity was 100%. Then, DHPLC was used to screen 62 other cases. We identified 74 FBN1 mutations in 81 patients: 64 were novel and 17 known. Of the 81 mutations, 41 were missense (50.6%), 27, either nonsense or frameshift mutations and predicted a premature termination codon (PTC) (33%), 11 affected splice sites (13.6%), and two predicted in-frame deletions (2.5%). Most mutations (67.9%) occurred in cbEGF-like modules. Genotype was clinically relevant for early diagnosis and conclusion of the diagnostic work-up in patients with incomplete or atypical phenotypes.

TGF-b-receptor 2 (TGFBR2) gene defects have been recently associated with Marfan syndrome (MFS) with prominent cardio-skeletal phenotype in patients with negative fibrillin-1 (FBN1) gene screening. Four mutations have been identified to date in five unrelated families. We screened TGFBR2 gene by direct automated sequencing in two adult patients diagnosed with MFS according to Ghent criteria, and in one girl clinically suspected as affected on the basis of a major cardiovascular criterion and skeletal involvement, all proven not to carry mutations in the exon -intron boundaries of FBN1 gene. We identified two novel and one known TGFBR2 gene mutations in the three unrelated probands. The D446N was identified in a 4-year-old girl with de novo disease characterized by severe cardiovascular disease and skeletal involvement. The M425V and R460H mutations were identified in two familial, autosomal dominant MFSs, both characterized by major cardio-skeletal signs and absence of major ocular signs. The mutation R460H has been recently reported in a family with thoracic aortic aneurysms and dissection. The three mutations are absent in 192 controls and affect evolutionarily conserved residues of the serine/ threonine kinase domain (exon 5). Our data support the recently reported association between TGFBR2 gene and MFS without major ocular signs (MFS2). The number of genotyped cases however is too low to confirm that major ocular signs are characteristically absent in MFS2. Accordingly, all patients proven or suspected to be affected by MFS with negative FBN1 gene screening could benefit from rapid investigation of the TGFBR2 gene.

A relational graph structure based on a bounda ry representation of solid objects is described. In this structure, called face adjacency graph, nodes represent object faces~ whereas edges and vertices are encoded into arcs and hyperarcs. Based on the face adjacency graph, we define a set of primitive face-oriented Euler operators, and a set of macrooperator$ for face manipulation, which allow a compact definition and an efficient updating of solid objects. We briefly describe a hierarchical graph structure based on the face adjacency graph, which provides a representation of an object at different levels of detail. Thus it is consistent with the stepwise refinement process through which the object description is produced.

SummaryThe R2726W mutation in the fibrillin 1 (FBN1, Marfan syndrome) gene segregates with isolated skeletal features of Marfan syndrome and/or high stature. Here we report a family in which two out of four individuals, an 18-year-old son and his mother, a 41-year-old woman, had the R2726W mutation of FBN1. Both family members carrying the mutation were of average height. The son had a Marfan-like phenotype, but his mother did not. The FBN1 R2776W mutation, which is associated with skeletal features of Marfan syndrome, appears incompletely penetrant. Consequently, genetic counselling in the presence of this mutation is difficult.