Background: Biomanufacturing of adeno-associated virus (AAV) involves multiple processes. The most common approach is triple transfection (TT) of Human Embryonic Kidney 293 (HEK-293) cells with three plasmids: one providing the gene of interest (GOI plasmid, or cisplasmid), one providing the AAV Rep and Cap genes (RepCap plasmid, or trans-plasmid), and a helper plasmid providing critical adenoviral genes (Helper). Another method uses Herpes Simplex Virus (HSV) as a helper virus, in combination with HSV GOI and HSV RepCap plasmids, to produce AAVs, typically in HEK-293 cells. Objective: Using AAV9 as a model capsid, we aimed to determine whether there are differences in selected quality attributes of the final purified AAV9 product across manufacturing platforms. results: For all comparisons, each AAV9 product will be called AAV9-HSV or AAV-TT. It was shown that AAV9-HSV contained molecular weight protein impurities (p=0.0004, ****) and a lower full-to-empty ratio. Mass and %VP ratio differences for each AAV9 capsid were not significant. Methods: A range of biophysical characterization assays was used to assess the reproducibility of the following attributes: hydrodynamic radius, charge heterogeneity, full-to-empty ratio, purity, and molecular weight of the final purified AAV9 from each platform. Results: For all comparisons, each AAV9 product is designated as either AAV9-HSV or AAV9- TT. AAV9-HSV exhibited molecular weight protein impurities relative to AAV9-TT (p = 0.0004, ****) and a lower full-to-empty ratio. Differences in mass and VP ratio for each AAV9 capsid were not statistically significant. conclusion: Specifics to where those differences could have arisen during the manufacturing of each were not investigated and a single lot of each was used. Discussion: The biophysical characterization assays provided information indicating tight overall comparability between platforms, with some variability observed. Conclusions: A single lot of AAV9 from each process was used for all comparisons; a larger sample size is recommended for future validation. Functional assessments, including process steps during manufacturing and biological assays, were not performed and could be explored in future studies to further validate these exploratory findings.
Hoyle et al. (Wed,) studied this question.