(T1430-09-53) Comprehensive AAV Characterizations Using HRMS: Peptide Mapping, Capsid Proteins, Intact ssDNA, and Intact Virus

Purpose: Analytical regulatory requirements for AAV-based gene therapies are currently not well established. However, monitoring protein and ssDNA heterogeneity for AAV may help define regulated analytics and accelerate AAV clinical and process development. Considerations in developing analytics for AAV include low concentrations, limited quantities, and large analyte masses that challenge commercially available analytical techniques. Here, we present methods for LC-TOF-MS analysis of both trypsin digested AAV5 and intact capsid proteins of several AAV serotypes. We also demonstrate the utility of charge-detection MS (CDMS) to analyze ssDNA isolated from AAV9 by anion-exchange chromatography (AEC). Additionally, CDMS, AEC, and size-exclusion chromatography measurements of the level of AAV8 without its targeted complement ssDNA are compared.

Methods: Tryptic digestion of 1.25 µg of AAV5 capsid protein occurred after surfactant removal using denaturing SEC. UHPLC-TOF-MS with a C18 RP column was used for the peptide analysis. LC-MS with intrinsic fluorescence detection (LC-FLR-MS) was used for the relative quantification and characterization of intact capsid proteins for several serotypes using a C4 RP column and DFA mobile phase modifier. Isolated ssDNA and Intact AAV8 samples were buffer exchanged in 20 mM ammonium acetate prior to nanospray CDMS analysis. ssDNA was isolated from AAV by digestion with proteinase K and purified using AEC. Intact AAV samples were analyzed directly by AEC with intrinsic fluorescence detection and by SEC using UV absorbance at 260 nm and 280 nm.

Results: Using only 1.25 µg of AAV sample, a combination of denaturing UHPLC SEC fractionation and an optimized 1-hour MS-friendly tryptic digest generated reproducible results and 98% sequence coverage for each of the highly homologous AAV capsid proteins (VP1, VP2, and VP3). Less than 0.4% standard deviation (n=3) was observed for all site-specific PTMs and degradants including N-acetylation, deamidation, oxidation, methylation, and phosphorylation, observed with the level above 0.5% using a benchtop LC-ToF-MS system. Using DFA as the mobile phase modifier, a 20-min RPLC-MS method, using 100 ng of capsid protein, was developed to achieve high chromatographic resolution of the capsid proteins with effective MS sensitivity. PTMs such acetylation and phosphorylation were identified. Additionally, fluorescence detection allowed for the assessment of changes in the relative abundances of VP1, VP2, and VP3 at a mass load of 50 ng. CDMS analysis was applied to the characterization of intact AAV8 virus and ssDNA isolated from AAV9. For intact AAV virus, the extent that the capsid contains its requisite ssDNA, no ssDNA, and a ssDNA fragment (full, empty, and partial) is a critical quality attribute as it impacts product efficacy and safety. Charge heterogeneity observed by AEC for an AAV8 sample determined by CDMS to be essentially empty was predominantly due to VP protein charge heterogeneity and not presence of AAV8 with a partial complement of ssDNA. CDMS also demonstrated a linear response for the level of empty AAV8 in addition to a high correlation with determinations by FLR-AEC and dual wavelength (260 nm and 280 nm) UV-SEC methods. For the analysis of ssDNA isolated from AAV9, CDMS data demonstrated that upon release from the capsid the ssDNA partially degraded or paired with its respective complement ssDNA.

Conclusion: UHPLC-Tof-MS and CDMS provide in-depth AAV protein characterization and intact AAV integrity information with a few micrograms of sample.

References: . J. Mol Biol. 2016 January 29; 428(2 Pt A): 292–300. doi:10.1016/j.jmb.2015.06.019.
2. Optimized reversed phase LC/MS methods for intact protein analysis and peptide mapping of adeno-associated virus (AAV) proteins. https://doi.org/10.1089/hum.2021.046


Size Exclusion Chromatography showing the empty and filled AAV capsids (A280/A260). The linear correlation plots showing good agreement with regard to both slope and correlation between the measured and the predicted percent empty capsid.


Anion Exchange Chromatography showing the empty and filled AAV capsids. The linear correlation plots showing good agreement with regard to both slope and correlation between the measured and the predicted percent empty capsid.


Charge Detection Mass Spectrometry showing the empty and filled AAV capsids. The linear correlation plots showing good agreement with regard to both slope and correlation between the measured and the predicted percent empty capsid.