Abstract 708: Using Global Proteomics and Network Science to Explore Therapeutic Targets for Abdominal Aortic Aneurysm (AAA)

Objective: To understand mechanisms critical to aneurysm development and identify potential therapeutic targets, we performed global proteomics and network analysis in mouse models of AAA. Methods and Results: AAAs were produced by luminal perfusion of the infrarenal aorta of C57BL/6 (wild-type) mice with elastase or subcutaneous infusion of angiotensin II (AngII) in Apoe-/- and Ldlr-/- mice (on congenic C57BL/6 background). Aortas were harvested at two intervals corresponding to developing or end-point aneurysm phenotypes that are specific to each model (n=3 for each procedure, interval, and genotype). Aortas were dissected into 8 segments spanning the arch to infrarenal portion, resulting in combined 288 aortic segments for label-free proteomics. Proteins were classified as significantly correlated to aneurysm according to their fold-change over control. Additional correlations were derived from a high-dimensional data analysis tool (“Xina”), developed in our laboratory, which enables clustering of proteins according to model, genotype, aortic region, and interval. Using our criteria, we identified lists of 159 proteins in the Apoe-/- and 158 proteins in the Ldlr-/- mice (AngII infused), and 173 proteins in wild-type mice (elastase perfused). Network analysis of these protein lists reveal commonalities between the models to include protein pathways related to protein translation, immune function, platelet activation, and extracellular matrix organization. Pathways enriched following AngII infusion included phagosome function and platelet aggregation, while pathways enriched in the elastase model included apoptosis, smooth muscle cell contraction, and Fc gamma R-mediated phagocytosis. Conclusion: Identification of pathways and proteins shared among these models may present master regulators of aneurysmal events and identify therapeutic targets. Conversely, identifying aneurysmal events unique to each model will provide valuable information regarding the model(s) chosen for use in a study and help contextualize model-specific findings.