Abstract
Following recent advances in post-thrombectomy stroke care, the role of neuroinflammation and neuroprotective strategies in mitigating secondary injury has gained prominence. Yet, while neuroprotection and anti-inflammatory agents have re-emerged in clinical trials, their success has been limited. The neuroinflammatory response in cerebral ischemia is robust and multifactorial, complicating therapeutic approaches targeting single pathways. In this study, we aimed to characterize early inflammatory gene dysregulation following ischemic stroke using translational in-silico and in-vivo approaches. Using an in vivo ischemic stroke model, transcriptomic analysis revealed significant dysregulation of inflammatory genes. Graph theory analysis then showed a rich-club organization among stroke-related genes, with highly connected core nodes. The expression levels of the core genes identified within this network significantly explained radiological outcomes, including T2-signal hyperintensity (R(2) = 0.57, P < 0.001), mean diffusivity (R(2) = 0.52, P < 0.001), and mean kurtosis (R(2) = 0.65, P < 0.001), correlating more strongly than non-core genes. Similar findings were observed with functional and cognitive outcomes, showing R(2) values of 0.58, 0.7, 0.54, and 0.7 for neurological severity scores, corner tasks, passive avoidance, and novel object recognition tasks, respectively (P < 0.001). Using in-silico analysis, we identified a set of upstream regulators directly interacting with core network nodes, leading to simulations that highlighted C3-targeted therapy as a potential treatment. This hypothesis was then confirmed in vivo using a targeted C3 inhibitor (CR2-fH), which reversed gene dysregulation in the neuroinflammatory network and improved radiological and functional outcomes. Our findings underscore the significance of neuroinflammation in stroke pathology, supporting network-based therapeutic targeting and demonstrating the benefits of targeted complement inhibition in enhancing outcomes through modulation of the neuroinflammatory network core. This study's approach, combining graph theory analysis along with in-silico modeling, offers a promising translational pipeline applicable to stroke and other complex diseases.