Hello, my name is Keean Peterson. I'm a third year medical student at the Wake Forest School of Medicine. This research with my mentor, Dr. Stacy Wolf, is focused on the creation of a novel mouse model of sporadic cerebral arteriovenous malformations. Cerebral AVMs are tortuous high flow lesions characterized by direct arteriovenous connections without an intervening capillary bed. These lesions are highly prone to hemorrhage and are a leading cause of hemorrhagic stroke in children and young adults. Here's an angiogram depicting an intracranial AVM. The pathogenesis of AVMs is largely unknown, however, a host of factors are implicating including genetic predisposition, inflammation, trauma, reaction to injury, and epigenetic influences are all thought to play various roles in development. Some AVMs arise as part of a syndrome such as hereditary hemorrhagic telangiectasia or capillary malformation, arteriovenous malformation syndrome, which exhibit Mendelian or classical inheritance. However, the vast majority of AVMs, upwards of 90%, arise in patients without a family history of disease. Sporadic AVMs were long considered congenital lesions that remain dormant for a long term until presenting later in life. However, newer research suggests that sporadic AVMs, or at least a subset of them, may arise de novo in a neoplastic manner or secondary to local trauma and inflammation, suggesting perhaps that inherited genetic predisposition may occur as a first hit with the AVM developing as a reaction to injury or a somatic second hit mutation. Next generation sequencing is helping to elucidate the genetic drivers of these lesions. Somatic mutations are being identified in a high prevalence of vascular malformations. Multiple studies have shown somatic mutations in the RAS-MAPK signaling axis in the endothelial cells of high flow cerebral arteriovenous malformations. In some studies, the prevalence has been up to 87% when using ultra deep next generation sequencing. Somatic mutations in a parallel signaling pathway, the PI3K mTOR pathway, are strongly associated with low flow lymphatic or venous malformations. So why do these somatic mutations in the RAS-MAPK signaling axis matter? Well, the RAS-MAPK ERK signaling pathway is a driver of cell cycle progression, cell growth, and proliferation. Disregulation of this pathway within endothelial cells could cause uncontrolled angiogenesis and blood vessel sprouting, driving the development of sporadic cerebral AVMs. Sporadic mutations that have been identified commonly affect the RAS-MAPK signaling pathway and are shown here by the red circles. Interestingly, mutations upstream of this signaling axis are known to cause arteriovenous malformations associated with syndromes. For example, mutations in RASA1 or EfrinB4 release RAS from negative feedback and are associated with the development of CM-AVM syndrome, dysregulating the same pathway. Further, mutations in GNAC and the associated proteins upregulate RAS-MAPK signaling and cause neurocutaneous capillary malformations and intracerebral AVMs in Sturge-Weber syndrome. The shared RAS-MAPK signaling abnormalities in both syndromic and sporadic AVMs underscores the importance of this common molecular pathway in AVM pathogenesis. Similar mutations are implicated in malignancy, which have chemotherapeutic agents that may work to target this pathway. Translationally, this means that chemotherapeutics that target this pathway's overactivation in cancer may be effective in slowing or reversing these lesions. In order to test this, an animal model that recapitulates the underlying genetic basis of these sporadic cerebral AVMs is needed. Historically, AVM animal models have either focused on inherited syndromes, like hereditary hemorrhagic telangiectasia, or they relied on surgical anastomosis of intracranial vessels to create an AVM nidus. However, both of these models lack the updated understanding of sporadic cerebral AVM genetics. One recent study tried to rectify this by creating cerebral AVMs through activating H-RAS mutations within the endothelium. However, this technique was limited by ubiquitous RAS upregulation, not limited to the cerebral vasculature, leading to early death of these mice through pulmonary and cardiac AVMs that hemorrhaged. Additionally, limitations of this model included that the H-RAS isoform has yet to be identified in human cerebral AVMs. The goal of this research is to selectively introduce KRAS mutations, which have been found in human sporadic cerebral AVMs, into the cerebral endothelial cells of mice without inducing mutations in the whole body vasculature, which could lead to premature death. In order to accomplish this, a novel AAV serotype, entitled AAV2BR1, that exhibits tropism for neuroendothelial cells, was selected. This serotype targets these microvascular endothelial cells of the brain and is used to deliver the KRAS mutation specifically. In order to accomplish this, the first experiment was to confirm the tropism of this new AAV serotype for the neuroendothelial cells of the brain. Adult wild-type friend leukemia B mice were injected retro-orbitally with either AAV2BR1 packed with luciferin at low, medium, or high doses, or the control. Bioluminescence was then assessed at day 7, 14, and 21 to confirm neuroendothelial delivery. The next step of the experiment was to induce mutant KRAS into the neuroendothelium. Mice were injected with either the optimized dose of the mutant KRAS packed into the vector or a vehicle control. Mice were sacrificed at day 28 via intracardiac perfusion of saline or paraformaldehyde. And cerebral specimens were harvested and have been analyzed so far using western blot and immunofluorescence. Results of experiment 1 showed that a low dose of the AAV2BR1 was sufficient for cerebral penetrance, confirming the tropism of this vector for the neuroendothelium. The next assay was to look at whole-brain PERK1 and 2 activity. PERK is a notable downstream target of the KRAS MAPK pathway and is upregulated in human sporadic cerebral AVMs harboring KRAS pathway mutations. So PERK was measured via western blot, showing no significant difference in whole-brain PERK between the control and KRAS groups, which is not unexpected considering the cerebral endothelial cells for which this AAV was tropic represent a small total proportion of the brain volume. Secondly, we looked at immunofluorescence, which demonstrated an increased double positive population of PERK within CD31 endothelial cells within the KRAS G12V cohort, indicating that PERK activity was selectively increased in endothelial cells in the KRAS G12V cohort. The study's preliminary results show that AAV2BR1 serotype does appear to selectively deliver the gene product to cerebral endothelium consistent with the prior literature. Increased double population of PERK and CD31 positive endothelial cells within brain sections of the KRAS G12V cohort indicate upregulation of RAS MAPK signaling activity in this population. Remember, PERK is a downstream target of the RAS MAPK signaling axis, among other pathways, and activity is shown to be increased in RAS MAPK-driven cerebral AVMs specifically. Our western blot showed no significant difference detected in whole brain PERK1 and 2, which as I mentioned before is not unusual given the fact that cerebral endothelial cells represent a small total proportion of brain volume. Future analysis is on hold due to the COVID-19 pandemic, however, next steps include PCR for detection of KRAS G12V mRNA within endothelial cell isolates and whole brain from the cohorts, as well as liquid latex perfusion to detect vascular malformations and shunting on gross exam of these specimens. Additionally, future cohorts will undergo induction of the mutation at various stages of development to assess the contribution of these mutations on pathogenesis of AVMs depending on timing of induction. And finally, we will assess the effect of BRAF MEK inhibitors of the RAS MAPK pathway on the endothelial RAS MAPK signaling and their utility in AVM prevention and treatment. So, in conclusion, AVMs likely have a genetic first hit as part of their pathogenesis. Up to 87% of sporadic cerebral AVMs and up to 100% of spinal AVMs have activating mutations in the RAS MAPK signaling axis within these endothelial cells. The shared RAS MAPK pathway dysregulation in sporadic and some syndromic AVMs illustrates the likely importance of this signaling pathway in AVM pathogenesis and represents a potential therapeutic target for these lesions. Creation of an animal model that recapitulates this genetic basis is essential to be able to understand the pathogenesis of these lesions and to test the therapeutic targets of this pathway. Our research will continue to refine and report outcomes from this and future experiments in order to assess the contribution of this pathway on AVM pathogenesis and to deliver a sustainable preclinical model of these lesions for testing of translatable treatments. Thank you.

cerebral arteriovenous malformations

AVMs

hemorrhagic strokes

RAS-MAPK signaling pathway

somatic mutations