4. Safety Considerations 
There are a variety of safety issues relevant to the use of a replication 
deficient recombinant adenovirus to treat the respiratory manifestations of 
CF, including: will the vector replicate in human epithelial cells? Does 
the vector express viral genes in addition to the CFTR cDNA? Is respiratory 
infection with the vector associated with transfer of the exogenous gene to 
the germ line? Is respiratory administration of the vector associated with 
damage to the lung? Are anti-vector antibodies elicited by respiratory 
administration of the vector? If so, does reinfection with another adeno- 
virus pose a risk? Is the vector shed following respiratory administration 
of the vector? Are there risks of having some of E3 deleted from the vec- 
tor? If recombination or complementation of the vector occurs in vivo . does 
this pose a risk to the patient and/or environment? Is there a risk of 
malignancy associated with the use of this vector? Does the genome of the 
vector integrate into the genome of the target cells? Does the expression 
of the newly transferred CFTR cDNA need to be regulated? 
The following addresses these questions point by point, and details in 
vitro and in vivo evidence that leads to the conclusion that the potential 
benefits of the use of this vector in the proposed clinical protocol sig- 
nificantly outweigh the potential risks accompanying the use of this vec- 
tor. 
4.1 Will the Vector Replicate in Human Epithelial Cells? 
With the deletion of map units 1.26 to 9.24 (455 to 3327 bp of the left 
hand end) , Ad-CFTR and AvlCFl have been designed to eliminate the regions 
of the adenovirus genome that control replication of adenovirus DNA and 
thus replication of the vector. To evaluate whether an El _ E3~ recombinant 
adenovirus vector replicates in human epithelial cells , human cervical 
carcinoma cells (HeLa) , a human epithelial cell line known to be permissive 
for wild type adenovirus growth (Cinatl et al., 1992), were infected with 
AvlCFl (60 pfu/cell) and the cells evaluated over time by assessing cell 
lysates for the presence of infectious adenovirus (or intact vector DNA) by 
adding the lysate to 293 cells and determining the resulting titer of 
adenovirus . No increase in the amounts of infectious virions was observed 
(Figure 4.1-A). In contrast, the control virus (Ad-dl327, an El + E3~ Ad5- 
based virus) increased over time. 
To determine whether vector DNA replication occurs, HeLa cells were infect- 
ed with AvlCFl (30 to 1000 pfu/cell), the cells incubated with 32 P 04 , = , viral 
DNA extracted (by a modified Hirt procedure, Hirt, 1967), the DNA cut with 
EcoRl, and gel electrophoresis and autoradiography were carried out. Where- 
as Ad-dl327 infection was clearly associated with viral DNA replication at 
all multiplicity of infection (MOI) , AvlCFl replication in HeLa cells was 
MOI dependent (Figure 4.1-B). At low MOI (MOI of 30 or less) no DNA repli- 
cation was observed but at high MOI (MOI of 100 or more) , HeLa did support 
AvlCFl DNA replication. This result is consistent with the observation that 
infection of HeLa at 60 pfu/cell was not associated with an increase in the 
amount of infectious virions (Figure 4.1-A). It is also consistent with 
evidence that at high MOI, HeLa will support replication of Ela~ adenovirus 
Recombinant DNA Research, Volume 16 
[689] 
