Abstract
Precise and accurate particle size distributions (PSD) of adeno-associated virus (AAV) preparations are an important measure of sample quality. Particularly challenging is the quantitation of AAV particles carrying genetic payload variants, which are considered impurities that must be minimized. While a plethora of methods exist to evaluate the amount of empty and full AAV particles, quantitation of partially-filled capsids having genomes close to the size of full capsids, has proven more challenging. The recently developed high-speed sedimentation velocity protocol (hs-SV-AUC) provides precise and accurate PSDs quickly. Here we assess the limiting resolution of hs-SV-AUC using both simulated and experimental data for a 1:1 mixture of two purified AAV preparations differing in DNA payloads by 696 nucleotides (NT). As few as 12 absorbance scans are needed to achieve baseline resolution of the two AAV components, while also providing AAV composition information via dual-wavelength analysis. For comparison, PSDs were obtained for the same 1:1 AAV mix using low-speed SV-AUC (ls-SV-AUC), mass photometry (MP) and charge detection mass spectrometry (CD-MS). In terms of their ability to separate the two full AAV peaks, the order of resolution of these methods is as follows: hs-SV-AUC > CD-MS > MP ≈ ls-SV-AUC. The advantages and disadvantages of each method are discussed.
