vitro studies have begun to unravel the biochemical function of the protein product of the CF 
gene called the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) which appears to 
be a highly regulated chloride channel (Reviewed in Collins, 1992]. Patients with CF have 
mutations that render the channel nonresponsive to elevations in intracellular cAMP [Kerem et 
al., 1989]. 
Cloning the CF gene also suggested new strategies for treatment based on gene therapy. In an 
important first step towards this goal, we reported in 1990 the functional correction of 
defective chloride transport in a CF cell in vitro by retroviral-mediated transfer of a normal 
CF gene [Drumm et at., 1990]. Translating the advance made in vitro to a practical and 
effective therapy for CF lung disease was not immediately obvious. Existing experimental 
paradigms of gene therapy were largely based on transplantation of autologous cells that had 
been genetically corrected ex vivo with recombinant retroviruses. Safe and effective 
application of this model to the treatment of CF lung disease seems unlikely. We also considered 
administration of recombinant retroviruses to the human airway as a way of directly 
transferring the therapeutic gene to the appropriate target cell in vivo. We found this to be an 
extremely inefficient process unless the airway has been injured and the epithelial cells are 
regenerating [Engelhardt et al., 1992a, see Appendix D]. This, in fact, was the predicted result 
in light of the requirement of cell division for retroviral integration. 
The development of effective gene therapies for CF will likely require the creation of new 
approaches that are capable of targeting genes to airway epithelial cells by direct 
administration in the airway. Furthermore, gene transfer must be accomplished in a non- 
replicating, stable target cell and the recombinant gene should be expressed for a prolonged 
period of time without pathology. In a series of studies performed in Cotton rats, Crystal and 
colleagues have provided convincing evidence that recombinant adenoviruses may have the 
properties described above necessary for safe and effective gene transfer to the human airway 
[Rosenfeld et al., 1991 and 1992]. 
Adenoviruses are nonenveloped DNA viruses responsible for many types of self-limited 
respiratory illnesses in humans. Many serotypes of adenoviruses exist; some of which have 
been exposed to a large proportion of the population in the United States. 
Adenoviruses have been the subject of extensive investigation in the laboratory [Horwitz, 
1990] . This has proven to be a powerful system for studying basic mechanisms of gene 
regulation. The adenovirus exists as a 35 kb double stranded DNA molecule with multiple early 
and late transciptional units. The type of recombinant virus proposed in this protocol has been 
used as a vector system in many in vitro and in vivo settings. The El region is deleted so as to 
render the virus defective in replication. Under some conditions, such as high Multiplicity of 
Infection (MOI), El deleted viruses can replicate in recipient cells. The recombinant minigene 
is placed under the control of a heterologous promoter in place of El sequences. 
Perricaudet and colleagues have developed El deleted Ad5 for use in vaccines [Levrero et al., 
1991] . They have performed experiments in which hepatitis B surface antigen expressing 
virus was injected into nonhuman primates. Interestingly, no immune response was observed. 
Of relevance to this protocol was the study by Smith et al. in 1970 in which volunteers were 
exposed to type 4 adenovirus by inhalation as part of a vaccination trial [Smith et al., 1970]. 
No substantial morbidity was noted. 
Recombinant DNA Research, Volume 16 
[811] 
