Marfan syndrome (MFS), a heritable connective tissue disorder, is caused by mutations in the gene coding for fibrillin-1 (FBN1), an extracellular matrix protein. One of the three major categories of FBN1 mutations involves exon-skipping. To rapidly detect such mutations, we developed a long RT-PCR method. Either three segments covering the entire FBN1 coding sequence or a single 8.9 kb FBN1 coding segment were amplified from reverse-transcribed total fibroblast RNA. Restriction fragment patterns of these RT-PCR products were compared and abnormal fragments were directly sequenced. Six exon-skipping mutations were identified in a panel of 60 MFS probands. All skipped exons encode calcium binding epidermal growth factor (EGF)-like domains and maintain the reading frame. In five probands, exon-skipping was due to point mutations in splice site sequences, and one had a 6 bp deletion in a donor splice site. Pulse-chase analysis of labelled fibrillin protein revealed normal levels of synthesis but significantly reduced matrix deposition. This dominant-negative effect of the mutant monomers is considered in the light of current models of fibrillin assembly. Probands with this type of FBN1 mutation include the most severe forms of MFS, such as neonatally lethal presentations.

Marfan syndrome (MFS) is an autosomal dominantly inherited connective tissue disorder characterized by cardiovascular, ocular and skeletal manifestations. Previously, mutations in the fibrillin-1 gene on chromosome 15 (FBN1) have been reported to cause MFS. We have now screened 44 probands with MFS or related phenotypes for alterations in the entire fibrillin coding sequence (9.3 kb) by single strand conformation analysis. We report four unique mutations in the fibrillin gene of unrelated MFS patients. One is a 17 bp deletion and three are missense mutations, two of which involve 8-cysteine motifs. Another missense mutation was found in two unrelated individuals with annuloaortic ectasia but was also present in unaffected relatives and controls from various ethnic backgrounds. By using allele-specific oligonucleotide hybridization, we screened 65 unrelated MFS patients, 29 patients with related phenotypes and 84 control individuals for these mutations as well as for a previously reported mutation and two polymorphisms. Our results suggest that most MFS families carry unique mutations and that the fibrillin genotype is not the sole determinant of the connective tissue phenotype.

No abstract

A human × Chinese hamster (CH) somatic cell hybrid subclone deficient in HPRT and containing only human chromosome 18 was irradiated with 7000 rad and fused to a thymidine kinase deficient CH cell line. Radiation-rescued hybrid cell lines, selected in HAT medium, were analyzed for human DNA with human interspersed-repeat sequence primers. Size and number of human chromosome fragments retained in a subset of hybrids were determined by FISH. A panel of 98 radiation hybrids (RH) was selected and analyzed for 90 chromosome 18-specific STSs by PCR, and for the D18Z1 centromeric marker by Southern blotting. STSs were developed from previously mapped RFLP loci and from published sequences. In addition, 32 novel STSs were generated from an 18-specific lambda library and from 18-specific YACs previously localized to chromosome bands by FISH. Marker retention frequency varied from 8–65 % with an average of 24%. In selected RH the STS typing data were correlated with the chromosome 18 regions retained using ‘reverse FISH’ of IRS-PCR products from the RH to normal metaphase chromosomes. The order and intermarker distances of loci were determined using two-point and multipoint maximum likelihood methods. The resulting RH map covers most of chromosome 18 with four groups of tightly linked markers and three regions of loosely linked markers, one around the centromere and two on the long arm. More than a third of the markers are polymorphic and allow integration with the linkage map. This RH map provides a framework for establishing a clone contig of the entire chromosome 18.