Quantum Cascade Laser-Based Mid Infrared Spectroscopy (QCL-MIDIR) for Real-Time Monitoring of Capsid Concentration and Empty/Non-Empty Percentage of Adeno-Associated Viral Particles

Abstract Adeno-associated virus (AAV) is a leading viral vector platform for in vivo gene delivery. Yet, the optimal production of these biologics remains challenging due to incorrectly filled (capsids containing partial genome information) AAV particles, residual host cell proteins, and residual DNA from the host cell (the baculovirus (BV) used to infect the Spodoptera frugiperda (Sf9) cells present in the cell culture lysate). Empty and partially filled AAV capsids adversely affect therapeutic efficacy and must be monitored closely. Current analytical methods for product quantification and characterization are labor-intensive, costly, and hampered by extended turnaround times and low throughput. The optimal AAV production requires more efficient, cost-effective, and fit-for-purpose analytical techniques capable of extensive characterization of AAV particles via on-line analysis. Here, we demonstrate on-line applications of quantum cascade laser-based mid-infrared spectroscopy (QCL-MIDIR) to characterize AAV9 particles and estimate the concentration and the percentage of empty or non-empty capsids. We have developed partial least squares regression (PLSR) models to predict the concentration and the percentage of empty and non-empty capsids. Our model can detect AAV9 particles with a limit of detection (LODmin) of 7.39 × 1012 capsids per mL (cp/mL) and down to 15.02% non-empty capsids. Together, our analytical platform provides a reliable, comprehensive approach to assessing and characterizing key quality attributes, including concentration and the percentage of empty or non-empty capsids, thereby meeting the demand for rapid, high-throughput analysis during development.