(468d) Simultaneous Optimization of Viral Vector Titer and Product Quality with Mathematical Modeling of Recombinant Adeno-Associated Virus Production

Recombinant adeno-associated virus (rAAV) is often employed for in vivo gene therapy and vaccine development due to its ability to enable long-term expression of transgenes, low toxicity, and high efficacy. While over 200 rAAV-based clinical trials are ongoing, scale-up and optimization of the manufacturing process remains challenging, limiting clinical potential. Specifically, the need to simultaneously optimize both titer and product quality is difficult, as only 5-30% of produced capsids carry the transgene, causing low effective titer and leading to costly and time-intensive downstream purification. This challenge is partially due to a lack of mechanistic understanding of the complex interactions between cell growth kinetics, viral vector production, and metabolism in viral vector-producing HEK cells. Mathematical modeling has the potential to meet this need, but foundational work has not yet been translated to experimental improvements. Further, the development of integrated models of cellular metabolism and mechanistic models of viral vector production remains an unmet need. Towards this goal, we have adopted an approach that uses a combination of mechanistic modeling and experiments investigating rAAV2 production via transient transfection in HEK cells to identify and optimize experimentally tunable conditions to simultaneously improve titer and product quality. Specifically, we developed a set of ordinary differential equations (ODEs) to describe the dynamics associated with a multi-component transiently transfected rAAV production system and the effects of modulating component amounts and ratios. The model includes explicit descriptions of cell viability, plasmid uptake, synthesis of viral and packaging proteins, viral genome synthesis, helper plasmid functionality, and capsid synthesis, packing, and secretion. Key model outputs include overall rAAV titer, the ratio of full to total capsids, and viable cell density, all over time. We will present results showing calibration of model parameters to in-house training data, identification of key mechanistic bottlenecks affecting performance, and model-guided design of experimental interventions to mitigate these bottlenecks. Ultimately, optimization of rAAV titer and product quality will have a major impact on production processes, leading to a decrease in production costs and an increase in the number of patients that can access rAAV-based treatments. Similar modeling strategies could also be applied to optimize other viral vector production systems, such as production of rAAV via stable cell lines, production of different rAAV serotypes, and production of lentiviral vectors.