Three approaches have been taken with respect to upstream production of AAV vectors. First, in DNA transfection-based procedures, various modifications have been made to the complementing rep-cap gene cassette in an attempt to enhance specific productivity and to decrease production of rcAAV. One group demonstrated that expression of rep and cap proteins may be limiting (100), but 2 other studies (101,102) suggested that cap proteins were limiting due to down-regulation of cap by increased production of rep. A packaging plasmid that has the Rep78/68 expression down-regulated by changing the initiation codon AUG to ACG was reported to give higher cap expression and higher yields of vector particles (102).
The only adenovirus genes required for full helper function are E1, E2A, E4, and VA, and transfection of the latter 3 genes into cells that contain the E1 genes, such as human 293 cells, can provide full permissivity for AAV (68). The infectious adenovirus can be replaced as the helper with a plasmid containing only the adenovirus E2A, E4, and VA genes (103,104) that, together with the E1A genes supplied by 293 cells, provide a complete helper function in the absence of adenovirus production. Another group (105) used a plasmid containing nearly all the adenovirus genome except the E1 region, but this yielded infectious adenovirus, probably by recombination with the E1 region in the cell. All these systems require transfection with 3 plasmids, for vector, rep-cap, and adenovirus helper function, respectively. In contrast, Grimm et al. (106) combined all 3 adenovirus genes and the rep-cap genes into a single plasmid. In general, all these approaches increased vector productivity compared with earlier systems such as pAAV/Ad (107), and productivities of at least 104 particles per cell have been reported. Nevertheless, these approaches still require DNA transfection and may be unwieldy for production scale-up.
An alternate approach to AAV vector production is to generate stable cell lines that contain the rep and cap complementing genes, the vector genome, or both. To avoid DNA transfec-tion, the cells must still be infected by a helper virus, adenovirus, but this can be removed readily as a result of advances in downstream purification processes (see Section VI.E). Rescue of vector from a producer cell line having the vector stably integrated was demonstrated by transfecting the cells with a rep-cap helper plasmid and infecting with adenovirus (108). Stable cell lines containing a rep gene capable of generating functional rep protein were constructed by Yang et al. (72) who replaced the p5 promoter with a heterologous promoter. Clark et al. (109) generated cell lines containing the rep and cap gene cassettes but deleted for AAV ITRs. Furthermore, the vector plasmid could be stably incorporated into the packaging cells to yield AAV vector producer cell lines (109,110). Producer cell lines provide a scalable AAV vector production system that does not require manufacturing of DNA and may reduce generation of rcAAV. However, a new producer cell line must be generated for each individual AAV vector and this may be laborious.
A modification of the packaging cell line method is to use a cell line containing a rep-cap gene cassette that is then infected with an Ad/AAV hybrid virus. The Ad/AAV hybrid is an E1 gene-deleted adenovirus containing the AAV ITR vector cassette (111,112). After infection of cells containing the rep-cap genes, the AAV-ITR cassette is excised from the Ad/ AAV, amplified, and then packaged into AAV particles. This allows the same packaging cell line to be used for production of different AAV vectors simply by changing the Ad/AAV hybrid virus, but it requires coinfection with adenovirus to provide the E1 gene function. It is worth noting that the Ad/ AAV hybrid viruses can also be used as delivery vehicles for AAV vectors (113-115), but this might suffer some disadvantages such as induction of innate immune responses characteristic of the adenovirus capsid interaction with cells. Another packaging cell system was described (116) in which the packaging cell contains both a rep-cap gene cassette and the AAV ITR vector cassette, and both cassettes are attached to an SV40 replication origin. Also in the cells is a SV40 T antigen gene that is under control of the iei-regulated system such that addition of doxycycline induces T antigen, which in turn results in amplification of the rep-cap and the vector cassettes. Subsequent infection of the cells with adenovirus renders the cells permissive for vector production.
Herpes simplex virus also can be used in production of AAV vectors by generating 2 types of HSV/AAV hybrid viruses. One approach (117,118) uses an HSV/AAV hybrid virus in which the AAV rep-cap genes, under control of their native promoters, were inserted into the HSV genome. This HSV/AAV rep-cap virus could generate AAV vector when infected into cell lines along with a transfected AAV vector plasmid or into cell lines carrying an AAV vector provirus. Alternatively, an HSV/AAV hybrid virus was constructed by inserting an AAV ITR vector cassette between HSV genome replication origins and then packaging this construct into an HSV particle using an HSV amplicon system (119,120). These HSV/AAV vector hybrids were infectious for neural cells, but the biological properties and fate of these HSV/AAV hybrid genomes after infection of cells is complex and not yet well understood. Finally, insect cells can also be used to produce AAV vectors. In this system, baculovirus vectors containing rep-cap gene cassettes or the AAV-ITR vector cassette are used to infect insect cells and AAV vectors can be generated (121).
Was this article helpful?