ABS 058: Investigation of Coacervate-Mediated Molecular Delivery on Mouse Behavior After Partial Crush Spinal Cord Injury

Himagowri Prasad ¹ , Laboni Hassan ¹ , Kunyu Li ¹ , Dr. Timothy O'Shea ¹

¹ Boston University Department of Biomedical Engineering

Van Wickle (2025) Volume 1, ABS 058

Introduction: Currently, the Glia Engineering lab is studying the effectiveness of coacervates for protein delivery in central nervous system (CNS) injuries. Coacervates are a class of biomaterials that are formed in aqueous conditions through liquid-liquid phase separation by electrostatic interactions between two polyelectrolytes. In the O’Shea lab, trehalose-based coacervates are formed through the mixing of two oppositely charged branched oligomers formulated with trehalose, a protein-stabilizing disaccharide. Trehalose-based coacervates possess high protein loading capabilities and can be easily dispersed in mouse neural tissue as micron-scale droplets, or could be delivered in bulk as a viscous liquid. In the Glia Engineering Lab, we are investigating the effectiveness of trehalose-based coacervates to deliver protein and small molecules to alter glial proliferation and wound repair function. Our objective for this project was to study the effectiveness of a coacervate-mediated treatment on partial crush spinal cord injury (SCI). One method to measure treatment effect in mice is to assess their functional recovery using behavioral testing. I have assisted members of the lab with analyzing recovery profiles of injured mice over the course of a two week period. In this project, I transformed these results into a novel predictive model using machine learning. Our goal for this project is to predict the behavior analysis score of mice who have sustained a partial SCI after 14 days post injury using their scores from 1 and 3 days post injury. Using data from non-treated mice to build this model, we will be able to assess if the coacervate-mediated treatment advances injury recovery. Finally, we were able to better predict the recovery profile of mice based on the severity of their injury. These results will help advance the biomaterials field by allowing us to understand the effectiveness of the coacervate-mediated treatment based on type of injury and behavioral recovery profile.

Methods: First, we ran open field and gridwalk behavioral analysis tests 1,3,7,10, and 14 days post injury. For the open field, we used the Open Field Test scoring mechanism from 0-5. For the gridwalk, we utilized a novel scoring system. Next, we graphed a mouse’s natural recovery trajectory over time against its open field score, and received an AUC that has trapezoidal intervals. Using each mouse, we calculated their trapezoidal AUC for each DPI interval, then summed the 3-14 trapezoidal AUCs. After calculating and graphing the trapezoidal AUC, we were able to identify three clear “bins” the mice could be categorized as Group 1 (low scoring), Group 2 (intermediate scoring), Group 3 (high scoring). Then for each group, we calculated the mean trace and fitted it to a sigmoidal curve. Lastly, we compared our results with IHC tissue stainings to understand if lesion size is correlated to behavior.

Results: Based on the behavior of mice between 1 and 3 days post injury (DPI), we were able to
separate animals into three distinct groups based on their recovery trajectory over a two
week period. We have seen anecdotally that a low scoring mouse at 3 DPI (score = 0.5) has a larger lesion size than at 14 DPI compared to the lesion at 14 DPI of a higher scoring mouse at 3 DPI (score = 1).

Discussion: We plan to explore this further to determine the relationship between 3 DPI behavior and lesion size in untreated mice. This includes the incorporation of the gridwalk behavioral test scores. We plan to use this data as a predictive model so that upon implementing experimental treatment at 3 DPI to injured mice, we can monitor if animals behavioral recovery is enhanced by the treatment compared to the predicted no treatment recovery profile. These results will strengthen post hoc IHC analysis of spinal cord tissue.