Adenovirus expression occurs in two major phases, generally identified as early and 
late (Graham and Prevec, 1991; Horwitz, 1991). In a simplified matter it can be said that 
early proteins are regulatory in nature while late proteins are structural components of the 
virion. Six regions have been described as "early" (El A, E1B, E2A, E2B, E3 and E4) 
although other regions are also expressed soon after infection (LI , IVa2 and IX). El A and 
El B are necessary for viral DNA replication and have been associated with the transforming 
potential of adenoviruses in vitro. E2A codes for a DNA binding protein and E2B for a DNA 
polymerase and the protein found at the 5' ends of the viral genome. E3 is not essential for 
viral replication in vitro and is associated with viral evasion of host immune response. E4 
protein is associated with viral assembly. The studies proposed here use El and E3 
deficient vectors packaged in the 293 human embryonic kidney cell line (described below). 
Two aspects of adenoviral biology have been critical in the production of replication 
incompetent adenoviral vectors. One, the ability to have the regulatory proteins produced 
in-trans and two, the inability of adenovirus cores to package more than 105% of the total 
genome size. The first was exploited by the production of the 293 cells, a transformed 
human embryonic kidney cell line with stably integrated adenoviral sequences from the left- 
hand end (0-11 map units) comprising the El a region of the viral genome (Graham et al., 
1977). This cell line is permissive for the production of viruses defective in El, and has 
been utilized in all previous RAC-approved protocols which use adenoviral vectors (RAC 
9212-034, 9212-035, 9212-036, 9303-041, 9303-042, 9312-067). The 105% limit of 
genome packaging size allowed production of "helper" viral genomes with "stuffer" DNA 
which exceed maximum packaging capacity (Bett et al., 1993). 
Adenoviral genomes are too large to be conveniently manipulated during sub-cloning 
steps to insert the gene of interest. Therefore, the gene of interest is inserted into a deleted 
El region in a plasmid with only a portion of the left-hand end of the adenoviral genome (the 
specific sequences used for this protocol are detailed in Appendix C). The oversized 
genome (helper virus) and the smaller plasmid with the gene of interest are cotransfected 
into 293 cells. The gene of interest plus the defective viral genome must not exceed the 
size limitations for packaging. Virus is produced only when recombination between the two 
plasmids occurs and the gene of interest replaces the stuffer DNA to result in a genome that 
meets the packaging constraints of the adenovirus core. 
Adenoviruses have been extensively characterized and make attractive vectors for 
gene therapy because of their relatively benign symptoms even as wild type infections, their 
ease of manipulation in vitro, the ability to consistently produce high titer, purified virus and 
the broad range of tissue targets. In addition, adenoviral DNA does not get incorporated 
into host cell chromosomes, therefore minimizing concerns about insertional mutagenesis or 
potential germ line effects. 
A primary safety concern is the possible development in vivo of replication competent 
adenoviruses and wide dissemination of such a recombinant in the patient. An additional 
concern is the non-vector contaminants of the preparation which may be of harm to the 
patient. The latter has been addressed by the production of viral vector products in a GLP 
facility specifically built and designed for the purpose of clinical grade vector production 
(described in Appendix C). The former is minimized by production of replication defective 
vectors and testing for the absence of replication competent contaminants at various stages 
of production. Although at high multiplicity of infection even El a deficient viruses may 
replicate, previous experience with Ad5 infections suggest this is a minor risk since Ad5 
infection causes only self-limiting, mild symptoms. The relative safety of Ad5 was 
[718] 
Recombinant DNA Research, Volume 20 
