mucosal surfaces, often complemented by superficial mucosal biopsies. Further, the 
effects of other types of toxicants, e.g., O3, have been investigated in nasal surfaces 
(65). Again in these studies, the toxicant is typically delivered by nasal inhalation, 
followed by a series of lavage and biopsy studies focussed on mechanisms of toxicity 
and the direct and immunologic responses to these types of injury. Finally, the nasal 
mucosa have been long used as a model for allergic responses to inhaled foreign 
antigens and other noxious stimuli (66,67). In these studies, a series of potential 
allergens, e.g., ragweed allergens, are instilled into the nasal mucosa, and the 
allergic inflammatory response followed by serial lavage measuring the release of 
mediators, inflammatory cells, and desquamated epithelial cells. These studies 
again have been supplemented by biopsy studies designed to measure the 
inflammatory response in individual cell types, utilizing a combination of 
immunocytochemistry and PCR-based technologies. 
In summary, the nasal mucosa is an excellent model for the respiratory 
epithelium of the lower airways. Large numbers of studies have documented that 
the normal ion transport physiology of the nasal and lower airway regions is very 
similar, and that the nasal epithelium is an accurate predictor of the inflammatory 
responses of the lower airways in response to inhalation/administration of agents 
that are toxic. 
II.D. Xenograft model 
A critical evaluation of the safety and efficacy of adenoviral-mediated 
transfer of the CFTR gene to human airway would be greatly simplified if there was 
an authentic animal model. Critical questions relate to the biology of El deleted 
Ad5 in the context of a human airway and parameters necessary to achieve 
functional correction in this structure. Most currently available animal models have 
limitations. A major problem is that the airway of humans differs substantially in 
function and anatomy from that of other species. For example, many species, 
including rodents, are essentially devoid of the sites of mucous production found in 
humans (i.e., goblet cells of the surface epithelium and submucosal glands). This 
may explain why the recently described murine models of CF, generated by 
mutating CFTR in the germ line, have only mild pathology in the lung (68). 
Another important consideration is the marked variation in infection and 
replication of adenoviruses that is observed in recipient cells from different species. 
Our strategy was to develop a model based on the development of a CF 
human airway xenograft established in a nu/nu mouse. A schematic representation 
of the steps involved in the procedure are shown in Figure 2. 
[448] 
Recombinant DNA Research, Volume 17 
