Characterization and optimization of the production of adeno-associated viral vectors using a baculovirus expression vector/insect cell system

Adeno-associated viruses exist as a class of virus that have shown promising characteristics as gene therapy vectors. They have further been shown in some instances to also illicit a strong immune response against more threatening viruses like the human immunodeficiency virus. Although recent improvements in the ability to produce these vectors at a reasonable scale in suspension by transient transfection have been reported, their production in insect cells using baculoviruses still promises to be one of the most efficient and scalable platforms to achieve the goal of producing enough material for clinical trials.;It was also found that using high MOIs of BacRep may be problematic in the long run. Instabilities with this baculovirus construct that are atypical of passaging other recombinant baculoviruses resulted in a significant decrease in expression levels over fewer passages. It has been speculated that the instability ensues from the palindromic sequence resulting front the head-to-head orientation of the Rep genes (Kohlbrenner et al., 2005a). The use of low passage BacRep minimizes this effect. An alternate four-baculovirus system developed by Kohlbrenner et al. (2005a) increases the stability by separating the two Rep genes into two baculoviruses. Unlike the triple infection system, it was found, in this work, that capsids are likely the limiting factor when using the quadruple infection system.;The total number of capsids produced always significantly exceeded the number of functional AAV vectors and vector genomes. This implied that there was a potential to increase the efficiency of genome encapsidation. To try and address this limitation using culture parameters, the temperature of the process was modulated. Increasing the temperature during the production phase of the process from 27°C to 30°C increased replication proteins early in the culture, which is thought to be a key reason for the observed increase in SS progeny DNA, ∼ two-fold increase in DNAse resistant particles and ∼ three-fold increase in functional vectors produced. The most significant replication protein increase was Rep78, which inay suggest that the attenuated expression of the Rep78, designed because of its detrimental effects in mammalian cells, may be over-attenuated in the insect cell system.;To fully exploit the insect cell system, the maximum density at which AAV could be produced was investigated. Without adding nutrients during the culture, the synthesis per cell significantly decreased beyond 4x10 6 cells/ml. Medium renewal at the time of infection allowed production at ∼ 1x107 cells/ml although the synthesis per cell was only a fraction of the maximum achieved at lower cell densities. Supplementing the media using a nutrient cocktail increased the specific production but did not achieve the maximum level seen at lower cell densities. Resuspending cells from the early exponential phase, in fresh media, to higher cell densities, maintained specific production levels up to ∼ 1x107 cells/ml. Cell "age" therefore played a role in achieving the highest production levels. The strategies developed in this work yield ∼6.5x1012 functional AAV particles/L of cell culture. (Abstract shortened by UMI.);To optimize the cellular and volumetric production a detailed characterization of the system was undertaken. To gauge the effect of each baculovirus vector used for the production of AAV vectors, the quantity of each, the ratio between them and the time at which they were added, were investigated. The highest titers were achieved when BacRep and BacCap were added at high multiplicities of infection. Manipulating expression levels with the multiplicity of infection is difficult since there is inherent competition between the baculoviruses. This is believed to be the reason why offsetting the ratio of BacRep to BacCap always led to lower production of AAV vectors. Furthermore, the highest AAV production was achieved when all three baculoviruses infected the cell at relatively the same time, otherwise a significant reduction occurred. To minimize the amount of total baculovirus added to the system, BacITR, which only supplies the AAV vector genome, could be added at lower multiplicities of infection without significantly affecting the production of AAV vectors.