Cerebral cavernous malformation (CCM), or cavernous angioma, is a common disease that can occur sporadically or familially with autosomal dominant inheritance. CCMs are vascular malformations, predominantly in the brain, consisting of dilated, thin-walled, blood-filled caverns. These lesions can lead to headaches, seizures, focal neurological deficits, and hemorrhagic stroke, but the only available treatment is surgical resection. Familial CCM has been linked to three genes: KRIT1, CCM2, or PDCD10. These genes encode structurally unrelated proteins of poorly understood function, but the three are hypothesized to work as a complex¹.

Generation of an animal model that faithfully recapitulates CCM disease would greatly benefit the study of the natural history and pathophysiology of CCM and the search for therapeutics. In this dissertation, I demonstrate that Pdcd10 and CCM2 signal through distinct pathways, but that loss of heterozygosity is a common genetic mechanism by which both genes lead to CCM disease. I show that Ccm2 acts to suppress the activity of the small GTPase RhoA, whereas Pdcd10 acts through the GCKIII family of kinases. Studies of knockout mice demonstrate that Ccm2 and Pdcd10 serve essential, but different, functions in the endothelium during development. I also examined Pdcd10 function using the fruitfly Drosophila melanogaster , which has a homolog for PDCD10 but not KRIT1 and CCM2. These studies revealed that Pdcd10 regulates lumen formation in the Drosophila tracheal system and genetically interacts with GCKIII, distinguishing it from Ccm2 and indicating that Pdcd10 can function independently of the other CCM genes. Despite these differences in function, loss of heterozygosity at either locus results in the highly penetrant formation of dilated, vascular caverns in mice that phenocopy human CCM histologically and radiologically.

My studies led to the surprising conclusion that the CCM proteins do not share a common signaling mechanism. The distinct pathways, however, lead to a common pathology by the genetic mechanism of loss of heterozygosity. My work has also begun to elucidate the underlying biochemistry behind CCM, which may suggest potential therapies. The establishment of an animal model with highly penetrant disease, similar to human disease, is a powerful first step in developing these therapies.