Assessment of genome packaging in AAVs using Orbitrap-based charge-detection mass spectrometry

Figure 1. AAV bioprocessing and analysis Schematic overview of rAAV production in either HEK293 or SF9 host cells for a designed transgene of around 2.6 kb. After transfection of (or infection with) the corresponding transgene (1) (if replicated properly) they will yield an encapsulated 800-kDa genome (2). Other encapsulated off-target genomes could originate from self-complementary ssDNA dimeric variants (3) as well as truncated genomes (4). These gene products are all encapsulated (5) and will yield a mixture of empty and (partly) filled particles. The initial number of other contaminants (DNA, side product, host cell proteins) can be several orders of magnitude higher than that of rAAV capsids and require extensive purification. rAAV capsids show additionally a wide distribution of empty, partially filled, and filled capsids, requiring additional purification and monitoring. Detailed characterization is important as only a small number of the infectious particles will transduce their genome. This efficiency is sensitive to the presence of empty capsids and capsids filled with off-target genomes. For monitoring this, AUC and EM are the industry standards, whereby the former can also…

Figure 2. AAV bioprocessing can yield a diversity of co-occurring particles that can be sorted and counted by CD-MS (A) Two-dimensional CD-MS histogram of an AAV6a preparation revealing the mass and abundance of co-occurring particles and in (B) the corresponding mass histogram. Besides the distributions originating from the empty and fully filled particles, low-abundant signals are observed in the 2D histograms corresponding to particles containing the self-complementary genome variants (red box) and intermediate truncated variants (purple box). The particles contained within these latter boxes are shown in the mass histogram with an alike color scheme. (C) Two-dimensional CD-MS histogram of an AAV9 preparation and (D) the corresponding mass histogram.

Figure 3. Direct sampling enables efficient AAV particle characterization and quantification by CD-MS (A) Optimized sensitive approach for AAV particle characterization and quantification. Speed and required input amount are significantly reduced as no buffer exchange is needed. Here, the depicted workflow can be executed in ∼30 min per sample. (B) Illustrative 2D histogram of a dataset acquired in 20 min. For fast and easy quantification, all ions that fall in the margins of the drawn polygon are counted as either empty, partially full, or full AAV particles. (C) Corresponding mass histogram and bar plot depicting the most abundant masses for each species as well as their relative abundance.

Figure 4. Quantification of distinct rAAV6b particles (A) Two-dimensional histograms (top row) and corresponding mass histograms of AAV particles mixed at defined ratios (the estimated percentage of empty particles is indicated above). (B) Fractional abundances of empty, partially filled, and fully filled AAV particles from (A) and Figure S3A. Dotted lines indicate the expected estimated values of 75%, 50%, and 25%. For each mixing ratio, CD-MS data gathered at two gas pressure settings are shown (Xe 2 and 3, respectively on the left and right). (C) The relative fractional abundance of the partially and fully filled AAV particles, remains constant with about 71% fully packed and 29% packed with a truncated genome.