At this time it is unclear whether individuals with CF will behave more like the cotton rats 
or more like the non-human primates with regard to response to the vector. Human studies are 
required to define the magnitude of the pulmonary inflammation after receiving this construct. 
Our protocol will address this question with measurements of BAL cellular composition, and 
cytokine concentrations. Clinical studies involving administration of wild-type adenovirus 
(serotypes Adi, Ad2, Ad3, Ad4, Ad5, Ad6) in doses of up to 10 6 pfu to the human respiratory 
tract result in only mild self-limited symptoms (Ginberg et.al., 1955, Roder et.al., 1956, Bell 
et.al., 1956, Couch et.al., 1966, Smith et.al., 1970). Importantly, wild-type adenovirus would 
be expected to elicit a more robust pathological response than the AE1 vectors considered at 
present, and in these clinical studies the response was mild. 
Thus, although toxicity studies raise the issue that respiratory administration of AE1, AE3 
recombinant adenoviral vectors such as AvlCF2 may result in some inflammation, this response 
appears to be mild and its significance is unclear. It is expected that such a response will be 
self-limited and should not result in significant pulmonary toxicity, nevertheless, the present 
protocol will seek to verify the safety of AvlCF2 in the human lung and test for local and 
systemic immune response. To this end, we will assess anti-adenovirus antibody production in 
nasal secretions as well as in serum. 
Germane to the potential for an anti-adenovirus vector antibody response, is the possibility, 
or perhaps likelihood, that AvlCF2 vector treatment may be required on a repeated basis. 
Experiments in cotton rats have now demonstrated that the respiratory tissues can be infected 
by an AE1, AE3 recombinant adenoviral vector even in the presence of very high serum anti- 
adenovirus antibody titers. Further, due to the presence of significant amounts of elastase within 
die lungs of CF patients, local degradation of potentially neutralizing antibody may limit the 
utility of this natural defense against the vector in the lungs of CF patients. Thus, it will be 
important to evaluate the ability to repeat the transduction of human respiratory tissues by 
AvlCF2 in vivo to assess the potential utility of adenoviral-mediated gene therapy for the 
pulmonary component of cystic fibrosis. This is one of the aims of this study. 
The E3 region of Ad5 encodes proteins involved in the immune response of the infected cell 
(Ginsberg et.al., 1989; Wold e^.al., 1991). Specifically, the 19 kDa E3 glycoprotein is involved 
in down-regulating MCH class I gene product translocation to the cell surface. This appears to 
have the effect of blocking the natural host defense to the adenovirus infected cells due to a 
reduction in self-recognition by host immune effector cells. Thus, E3 + vectors might, at least 
theoretically, be less easily detected by the host immune system than AE3 vectors. The absence 
of the E3 region has been argued as a safety feature in some CFTR expressing recombinant 
adenoviral vectors (Ad-CFTR, Crystal et.al., RAC proposal, 1992) as has its presence in others 
(Ad2/CFTR-1, Welsh et.al., RAC proposal, 1992). However, since E3 + gene expression is a 
function of El gene expression, and the El region is deleted in all first generation adenoviral 
vectors for gene transfer, the significance of the of presence or absence of the E3 region in 
current vectors is unclear with regard to its effect on immunity or persistence of the vector. 
4.3.2 Replication of AE1 Recombinant Vectors: AE1 recombinant adenovirus vectors are 
replication deficient and can be complemented by wild-type adenovirus. In the absence of wild- 
type adenovirus under certain conditions where cellular Ela-like factors can act in trans 
complementation might also occur. We have evaluated the replication potential of several Avl 
series recombinant adenoviral vectors and compared them to wild-type adenovirus in a variety 
of settings. First, the potential for replication of AvlCFl, Avl Null 1 and Ad5 were evaluated 
in a cultured epithelial cell line, (HeLa) (Figure 26). HeLa cells were infected with 60 pfu/cell 
[386J 
Recombinant DNA Research, Volume 17 
