and evaluated for infectious virus daily for three days using plaque assays. As a control, Ad5 
replication competent virus replicated, generating progressively higher titers over this period. 
In contrast, there was no increase in titers of either AvlCFl or AvlNulll. To evaluate the 
potential for viral replication using another, perhaps more sensitive assay, HeLa cells were 
infected with either AvlCFl or AvlCF2 or as a control Ad5 at MOI from 10 to 1000 and 
evaluated by metabolic labelling of DNA using Hirt extraction, restriction endonuclease 
cleavage, gel electrophoresis and autoradiography. Wild-type Ad5 clearly replicated and yielded 
virus-specific labelled DNA bands (Figure 27). In contrast, HeLa cells transduced with either 
AvlCFl or AvlCF2 at MOI of 10-30 pfu/cell showed no labelled virus-specific DNA bands, 
even after long exposure of the autoradiogram. However, at high MOI (100, 500, 1000 pfu/cell) 
some labelling of viral DNA was observed on long exposure on the autoradiogram. Thus, in 
HeLa cells, at high MOI there is a potential for production of infectious virus, albeit at very low 
levels if at all, due to the presence of very low levels of viral DNA replication. Importantly, 
evaluation of the tissue which is the actual target for CF gene therapy in this and other similar 
protocols has shown that the AE1 adenoviral vector AvlCFl did not replicate in bronchial 
epithelial cells even at MOIs of up to 1000 pfu/cell (Figure 28). Similarly, Ad-CFTR did not 
replicate in freshly isolated bronchial epithelial cells (Crystal et.al., RAC proposal, 1992). 
4.3.3 Extent of Infection of Bronchial Tissue: The potential for replication was further 
evaluated in the human tissue which is the actual target of gene therapy in this protocol. To 
accomplish this, freshly isolated human bronchial epithelium was obtained at bronchoscopy from 
a normal individual and infected with MOIs from 10 to 1000 pfu/cell with AvlCFl, or with 
control, Ad5 virus. As expected, Ad5 clearly replicated at all MOIs, while, as a negative 
control, no labelled viral DNA bands were seen in uninfected cells, (Figure 28). In contrast to 
Ad5, no labelling could be seen AvlCFl -infected freshly isolated human bronchial epithelial 
cells at MOI from 10 to 1000 pfu/cell, even after prolonged exposure of the autoradiogram. 
Finally, in a preliminary attempt to address the extent of infection of the respiratory 
epithelium following in vivo administration, we have used an AE1 marker virus (AvlLacZ4) at 
6 x 10 8 pfu, which is introduced via nasal administration to cotton rats lung tissue and have 
evaluated the tissue by X-gal staining, (Figure 29). 72 hours following infection, blue staining 
of nuclei was observed in patchy distribution scattered segmentally in the lung. The punctate 
pattern of infection suggests that the vector was not being transmitted from cell to cell nor from 
bronchus to bronchus, consistent with the lack of in vivo viral replication. The lack of in vivo 
replication and cell-to-cell spread is supported by in vitro infection of several human cell Lnes 
demonstrating a clear dose response relationship between MOI and the percent of cells stained 
with X-gal (Table II and Figure 30). 
Three human pulmonary cell lines H441, A549 and WI-38 cells were infected with AvlLacZ4 
for 12 to 72 hours, at a multiplicity of infection (MOI) of 10 or 50 viral particles per cell. At 
each time point cells were stained for /3-galactosidase. Values represent the percentage of cells 
staining positive. No cytopathic changes were seen over the 72 hour time period. 
4.3.4 Evaluation of the potential for AvlCF2-induced cellular toxicity: Infection by wild- 
type or replication competent adenovirus significantly alters the genetic program of the cells 
evidenced by changes in cell morphology the so-called cytopathic effect (CPE) related to shut 
off of host cell protein synthesis and eventually leading to cell death. To evaluate the effect of 
Recombinant DNA Research, Volume 17 
(387] 
