SUMMARY Brain arteriovenous malformations (bAVMs) are composed of abnormal connections between arteries and veins that lack an intervening capillary network. As a result, high-pressure blood from feeding arteries shunts directly into veins. These vascular lesions become distended and highly remodeled, resulting in a tangle of enlarged blood vessels that are prone to rupture. Indeed, bAVMs are a leading cause of hemorrhagic stroke in children and young adults. All current treatment modalities for bAVMs, including surgery, embolization or radiation carry a significant risk of disability or death, and these options are not available for ~20% of bAVM patients due to excessive risk. Because of these complications, alternative medical strategies with lower morbidities such as targeted pharmacological therapies are desperately needed. However, we first need a clear understanding of the biology underlying bAVM development and maintenance. The majority of bAVMs occur sporadically without a family history of the disease. Using whole exome sequencing, we recently identified somatic, activating mutations in the KRAS gene, which encodes a GTPase that is involved in signal transduction. The identified mutations were confined to the endothelium and result in KRAS being locked in a GTP-bound ‘ON’ state. Notably, we have established mouse and zebrafish models of endothelial-specific expression of mutant KRAS, which have revealed the sufficiency for these genetic lesions to drive disease. We have gone on to demonstrate, through transcriptional profiling of cultured cells and in vivo studies in zebrafish expressing mutant KRAS, that many KRAS-induced molecular and cellular changes require MEK/ERK activity. Much remains to be learned regarding the etiology of sporadic bAVMs and our cell culture, mouse and zebrafish models will enable us to define the molecular, cellular and morphological changes that are involved in the initiation and maintenance of bAVMs. We will utilize our expertise in animal models of bAVMs, imaging, cell biology, signaling and single-cell RNA sequencing, to gain unprecedented insight into the bAVM disease process. This information will be leveraged for the design of pharmacological interventions to improve patient outcomes. Our proposal will: 1) define the threshold of KRAS mutant endothelial cells that can remodel vessels, 2) identify the vascular bed(s) that are susceptible to active KRAS expression, 3) determine how KRAS mutations impact hemodynamic signaling and bAVM progression, 4) uncover the cell-autonomous and non-cell autonomous mechanisms of mutant KRAS, and 4) determine the requirement for KRAS and MEK activation for bAVM maintenance in our pre-clinical models. Together, these studies will expand our understanding of bAVM pathogenesis and will assess whether inhibition of the KRAS/MEK pathway may be a viable therapeutic target to pursue in human patients with bAVM.