E2 - Other Nonviral Delivery (including lipid nanoparticles and exosomes)
221: New Biological Insights into Extracellular Vesicle Associated AAV Help Improve Vector Yield and Provides Rational for Intraluminal AAV Packing
Type: Oral Abstract Session
Presentation Details
Session Title: Exosomes, Virus like Particles, and LNPs
Introduction. Extracellular vesicle-associated adeno-associated virus vectors (EV-AAV) represent a unique population of vectors generated during production in 293 cells, in contrast to conventional non-enveloped AAV capsids. Several studies have provided experimental evidence that EV-AAV provides desirable gene delivery aspects such as greater resistance to antibody neutralization and transduction of organs and tissues in vivo in mice. Despite these promising data, there is a great deal of characterization of the EV-AAV system that needs to be performed. Free conventional AAV vectors co-isolating with EV-AAV has hampered the definition of accurate yields and functional activity of EV-AAV. Results. Here we use an optimized density gradient to answer some outstanding questions regarding EV-AAV characteristics, as well as using it to assess optimization of the production system. Using a variety of molecular biology and biochemistry methods, we determined that gradient fractions 1-6 contained EV-AAV and fractions 7-10 conventional AAV. First, we quantitated the percentage of AAV genomes on the gradient were in the EV-AAV fractions of AAV serotypes 1, 2, 5, 6, and 9. We found the average value ranged from 2.4% (AAV5) to 11.95% (AAV6). AAV6 and AAV9 had the highest total yields of genomes in the EV-AAV fractions. Next, we assessed whether intraluminal EV-AAV9 could mediate functional transduction of cells (as opposed to AAV9 bound to the surface of EVs, or co-isolating free AAV9 capsids). To do this we performed an immunoprecipitation using an anti-AAV9 antibody conjugated to magnetic beads. Interestingly, we observed that approximately 20% of AAV genomes in the EV-AAV9 pooled fractions were resistant to AAV9-pulldown, suggesting a substantial portion of EV-AAV9 are intraluminal (with no bound AAV on the EV surface). To understand if these membrane-encapsulated vectors could transduce cells and resist antibody neutralization we performed neutralization assays using intravenous immunoglobulin (IVIg) and HeLa cells. We compared transduction in the presence or absence of IVIg with conventional AAV9, purified EV-AAV9, and EV-AAV9 with all surface-bound or free AAV removed by anti-AAV9 pulldown (ultra-pure EV-AAV9, UP-EV-AAV9). Interestingly, we found that in the absence of IVIg, UP-EV-AAV9 gave slightly higher (1.3-fold) transduction efficiency to EV-AAV9 and 2.5-fold higher transduction efficiency than conventional AAV9. UP-EV-AAV9 was similarly resistant to antibody neutralization to EV-AAV9 and much more resistant (6-fold at a 1:500 dilution of IVIg) to conventional AAV9. Finally, we tested whether heterologous expression of membrane associated accessory protein (MAAP) from AAV8 (MAAP8) with an AAV9 Cap (AAV9 MAAP null) production system would alter yields of EV-AAV9. Interestingly, we found that MAAP8/AAV9 Cap increased yields of EV-AAV9 by 3-fold compared to the cis-expression of AAV9 Cap/AAV9 MAAP. Conclusions. Gradient purification allowed us to accurately quantitate EV-AAV and conventional AAV yields in producer cell media. Our anti-AAV9 pulldown assay suggests that AAV9 capsids inside EVs are functional at transduction and therefore do not require AAV9 on the EV surface to mediate transgene expression. The antibody neutralization data also suggests that surface bound AAV9 capsids are not required for antibody evasion by EV-AAV as UP-EV-AAV9 was similarly resistant to IVIg as EV-AAV9. Thus, strategies aimed at increasing AAV packing inside EVs are warranted. The experiments with heterologous expression of MAAP8 suggest that yields of EV-AAV9 can be improved using this approach. In sum, these data provide important information for the further development of the EV-AAV vector system.
Ming Cheng1, Demitri de la Cruz1, Carrie Ng1, Zachary Elmore2, Aravind Asokan2, Casey A. Maguire1
1The Massachusetts General Hospital, Charlestown, MA,2Duke University, Durham, NC"
Find This Session