Activating TIE2 mutations with distinct features cause a spectrum of venous malformations

Venous malformations (VMs) are localized defects of angiogenesis consisting of abnormally enlarged venous channels with disorganized surrounding smooth muscle cells (SMCs). VMs can cause pain and diminish quality of life. While mostly sporadic, 1-2% of venous malformations occur as an autosomal dominant inherited trait, cutaneomucosal venous malformation (VMCM). Earlier studies demonstrated that TIE2 germline mutations cause VMCM (Vikkula et al., 1996), while somatic mutations in the same gene cause common sporadic unifocal venous malformation (Limaye et al., 2009). In the frame of this thesis, we have expanded the number of mutations identified in VMCM and in unifocal VMs. In addition, we have shown that TIE2 mutations are also responsible for multifocal forms of venous malformation (BRBN and sporadically occurring multifocal venous malformations). Interestingly, in contrast to unifocal VMs, most multifocal VMs are caused by TIE2 double cis mutations. Thus, TIE2 mutations have been identified in patients with a variety of venous malformations, illustrating the extensive phenotypic heterogeneity of TIE2-mediated venous anomalies. However, distinct mutations are enriched in the different clinical subtypes, and different venous entities also differ in terms of mutational mechanisms (germline, mosaic, or somatic events that occursimultaneously or serially). In cultured endothelial cells (HUVECs), we show that different TIE2 mutations have both common and mutation-specific effects. All mutations identified in these studies (including those that truncate the TIE2 C-terminal tail) induce ligand-independent phosphorylation of the receptor. We also show that the activation of PI3K/AKT is common in cells expressing mutated TIE2. Another common feature is the disruption of the normal EC cobblestone morphology. On the other hand, we also show genotype-phenotype correlation of mutations with distinct cellular effects, with clinically different venous phenotypes. In collaborative works, we also show that we are able to recapitulate the pathognomonic features of human VMs in a mouse xenograft model, using TIE2-mutant human ECs. This allowed us to better characterize the different mutants and to test molecular therapies. This led to a therapeutic pilot study in which rapamycin shows improvement of the symptoms of affected patients.