1986 a and b; Berger et al, 1989; Wilmott et al., 1990; McElvaney et al., 1991; Meyer et al., 
1919; Trapnell et al., 1991; and Nakamura et al, 1992). 
V.A.3 Infusion of recombinant adenoviruses. Toxicity due to the adenovirus can be caused by 
several mechanisms. 
V.A.3.a Ectopic or overexpression of CFTR. 
We have performed a series of studies in animals to address these issues. The most compelling 
experiments involved the generation of transgenic mice that had been programmed to express 
high levels of human CFTR in the airway. One set of animals expressed CFTR from a promoter of 
the SPC gene that is very active in the distal airway including the alveolar cells and the distal 
bronchiolar cells. Another set of animals expressed CFTR from a promoter of the CC10 gene 
that is very active in most of the cells of the proximal airway. All of the transgenic lines 
expressed high levels of human CFTR protein (which is active in mouse cells) in the 
appropriate cells and none of them demonstrated pathology or clinical abnormalities. This 
suggests that expression of recombinant CFTR in the wrong cell or unregulated, overexpression 
in the correct cell will not be deleterious. 
V.A.3.b Replication of the virus and expression of viral proteins. 
The virus has been engineered in a way that the region responsible for activating its 
transcriptional units, called El, has been deleted. In principal, therefore, the recombinant 
genome should not replicate or express viral proteins in the recipient cell. In practice, certain 
cell types under special conditions can transactivate the viral genes using endogenous cellular 
factors. 
We have established a model of the CF airway that is based on the growth of human airway 
xenografts in nu/nu mice to better address these important issues in a relevant biological 
setting. This model was used to evaluate the possibility of viral protein expression and 
replication of the El deleted CFTR virus in the context of a human CF epithelium. These studies 
failed to detect viral protein expression or ongoing replication of the virus in human xenografts 
that were stably expressing the recombinant gene. Any replication and spread of the virus in 
this model would rapidly disseminate throughout the xenograft because it could not be contained 
by the immune system, which is absent in these nu/nu mice. We believe therefore that this is 
an extremely stringent test for viral protein expression and replication. The MOI we have 
selected to use in vivo in this study is estimated to be far below that required to cause cytopathic 
effects in vitro based on transactivation of the El deleted genome by cellular factors. The 
xenograft experiments have now been performed with cells derived from both cystic fibrosis 
and non cystic fibrosis patients with similar results, arguing against increased virulence of the 
virus in the relevant biological setting of a deficiency of CFTR. Our recent studies using non- 
human primates confirm the relative safety of this approach to airway directed gene transfer. 
Animals exposed to high concentration of recombinant adenoviruses had no significant gross or 
histopathology. 
It is important to consider the consequences of the unexpected dissemination of the recombinant 
virus or a replication competent derivative. Previous experience with adenoviruses in humans 
suggests that they are relatively benign. Adenoviruses are endemic in the population causing a 
variety of self limited infections. Infection with Ad5 causes mild respiratory infections 
characterized by Coryza, pharyngitis, fever, malaise, and tonsillitis. The relative safety of 
adenoviruses was evaluated in a series of experiments, performed in the late 1960's and early 
1970's, in which volunteers were exposed via the respiratory tract to wild types forms of 
adenoviruses including the serotype used in this study (Ad5). For example, in a study by Couch 
[834] 
Recombinant DNA Research, Volume 16 
