Venous malformations (VMs), the most common errors of vascular morphogenesis in humans, are composed of dilated, serpiginous channels. The walls of the channels have a variable thickness of smooth muscle; some mural regions lack smooth muscle altogether. A missense mutation resulting in an arginine-to-tryptophan substitution at position 849 in the kinase domain of the receptor tyrosine kinase TIE2 segregates with dominantly inherited VM in two unrelated families. Using proteins expressed in insect cells, we demonstrate that the mutation results in increased activity of TIE2. We conclude that an activating mutation in TIE2 causes inherited VMs in the two families and that the TIE2 signaling pathway is critical for endothelial cell-smooth muscle cell communication in venous morphogenesis.

No abstract

No abstract

Methods SUBJECTSAfter obtaining informed consent, each participant (38 members of the family, fig 1) was interviewed and examined by the same physician (KAP). Physical examinations were performed and family histories were taken. Affected status was assigned on the basis of one or more venous malformations. Other vascular anomalies, such as lateral telangiectatic naevi (port wine stains), were occasionally noted but not included as part of the affected phenotype. Blood was drawn from appropriate subjects and DNA was extracted using established methods.3 GENOTYPES AND LINKAGE ANALYSISAnalyses of simple repeat markers were performed as previously described.3 The venous malformation phenotype was analysed as an autosomal dominant disorder with age dependent penetrance. For estimating the age dependent penetrance, four liability classes were created with age ranges 0-12 (penetrance 0 5), 12-21 (penetrance 0-75) 21-45 (0 90), and >45 (0-95), based on observations from more than 10 years of clinical contact with the entire extended family. Disease frequency was set at 0-0001. Two point linkage analysis was performed on a Sun Sparc station 10 using the MLINK subprogram of the LINKAGE computer package (version 4-9).4 The analysis was repeated using the phenotypic information on only the affected subjects to ensure that information was coming only from those with a definitive disease status. Allele frequency was estimated for the markers using the pedigree maximum likelihood estimation from the family data by the method of Boehnke' and did not differ appreciably from the allele frequencies listed in GDB.