The non-pathogenic human dependoviruses, adeno-associated viruses (AAV), infect either proliferating or post-mitotic cells. Since all the AAV and helper virus genes can be provided in trans, the entire AAV genome may be replaced with the gene of interest. The resulting recombinant AAV (rAAV) may then be produced transiently in cell culture and the rAAV particles concentrated and purified based on physical properties of the virion. The established rAAV production methods rely on transient transfection of adherent cells which grow at approximately 100,000 cells per cm(2). If each transfected cell produces 10,000 particles of rAAV, then the yield per cm(2) is 10e9 particles per cm(2) or approximately 2x10e11 particles per 15 cm diameter plate. Although the per cell yield of rAAV is relatively high, the ability to scale-up production rapidly becomes untenable with exponential increase of cells required to produce vector. Thus, production of 10e15 particles requires 5,000 to 10,000 tissue culture plates. In order to take advantage of cells growing in suspension, we developed an rAAV production system that utilizes recombinant Autographa californica nuclear polyhedrosis virus, (AcNPV or baculovirus) to deliver the AAV rep and cap genes as well as the rAAV vector DNA into the insect cell line, Spodoptera frugiperda (Sf9) cells. These cells grow at densities of 10e6 to >10e7 cells per ml. We demonstrated that rAAV DNA is rescued and replicated from either plasmid or baculovirus in the presence of AAV Rep 78 and that the level of rAAV DNA replication exceeds the level observed in HEK 293 cells. Similarly, Sf9 cell extracts express the AAV structural proteins, VP-1, VP-2, and VP-3 in greater amounts than HEK 293 cells. By manipulating the initiation codon and contexts, we were able to obtain VP expression in amounts that appear in the mature virion. Two AAV encoded replication proteins, Rep 78 and Rep 52, are necessary for AAV DNA replication. By utilization of a strong and weak insect virus promoters, we were able to achieve relatively high levels of Rep 52 and low levels of Rep 78. Infecting Sf9 cells with three recombinant baculoviruses: Bac-Rep, Bac-VP, and Bac-ITR, produced greater than 10e4 DNAse resistant particles of rAAV2 per Sf9 cells. Infecting the Sf9 cells at 2x10e6 cells per ml resulted in >10e10 particles per ml. We analyzed the particles and have determined that the Sf9 produced rAAV was indistinguishable from HEK 293 produced rAAV based on the following assays: 1. Electron microscopy 2. Protein composition 3. Buoyant density 4. Neutralization with an anti-AAV2 capsid mAb 5. Transduction of 293 cells 6. Inhibition of transduction with heparin 7. In vivo transduction of murine retina and skeletal muscle We have extended the process to include AAV serotypes 1, 4, and 5. However, using the processes developed for rAAV2 production, there are no a priori reasons that would prevent the production of any AAV serotype in Sf9 cells.
Showing the most recent 10 out of 29 publications
