Gene Therapy for CF using Cationic Liposome Mediated Gene Transfer: Phase I Trial 
increase in short-circuit current induced by forskolin was 11.04, ±5.17. These results indicate 
that lipid mediated gene transfer successfully introduced a cyclic AMP dependent chloride 
secretion that was barely detectable in control tissues. An increase in cAMP dependent current 
is the anticipated result of human CFTR expression in a tissue with otherwise minimal 
endogenous forskolin stimulable chloride secretion. 
These results are among the first to demonstrate transfer of an active human CFTR to 
an intact mammalian epithelium in vivo . Previous reports have demonstrated correction of the 
Cl (0 transport defect in the trachea of transgenic (cf/cf) mice using DOTMA/DOPE/DNA 
complexes. However, 1) the presence of an endogenous cAMP stimulated Cl' channel (distinct 
from CFTR within the airways of these mice and 2) the young age and severe gastrointestinal 
manifestations (including malnutrition) of the CF animals limit the use of the CF mouse in gene 
transfer based experiments targeting the pulmonary epithelium. The rat trachea may represent 
an alternative model for studies of CFTR gene transfer in this context, as suggested by efficient 
reporter gene transfer and relatively low activity of the endogenous rat CFTR protein. The 
discrepancy between in vitro and in vivo transfection efficiency (Figures 2 and 5) may relate to 
differences between the cell types studied, the conditions used, or to the poorly understood 
mechanisms within cells which allow the expression of plasmid DNA based genes under the 
regulation of eukaryotic promoters. We have observed that although only 5% of human 
respiratory epithelial cells express reporter genes when driven by the SV40 early promoter in 
vitro (Fig. 2), virtually 100% of these cells receive the lipid/DNA conjugate and rapidly 
transport the exogenous DNA to the cell nucleus (Fig. 3). We also have recently demonstrated 
that promoters of increasing strength not only lead to enhanced overall signal, but to an 
increased proportion of cells expressing Lac Z (10% using the SV40 promoter vs >90% using 
a Vaccinia/Lac operon plasmid-based promoter) (data not shown). These results indicate that 
delivery of plasmid DNA to respiratory epithelial cells in culture is much more efficient than 
the fraction of cells which are able to express the DNA at detectable levels, and implicates 
features inherent in the cells (rather than the particular lipid formulations) as contributing to the 
observed efficiency of lipid based gene expression. Such cell-specific factors might contribute 
to differences in the efficiency of gene transfer observed in vitro compared with cellular 
expression in vivo . 
These results represent the first in vivo demonstration of cationic lipid gene transfer to 
the respiratory epithelium of animals other than mice ‘using a lipid formulation other than 
lipofectin. The finding of successful gene transfer using cationic liposomes in a second animal 
species together with high activity in CF and normal human respiratory epithelium in vitro 
increases the probability that the results can be extrapolated to human subjects. Considering this 
information, efforts aimed to develop human based trials of lipid DNA conjugates to correct the 
CF defect in airways of patients with CF now seem appropriate. Human based trials will be 
required in order to evaluate both the efficacy of CFTR gene transfer and the possible safety or 
toxicity of this approach in vivo . 
m. METHODOLOGY 
m.A. Introduction; 
The pre-clinical data and background information provided above indicate that lipid 
mediated CFTR transfer to respiratory epithelium may constitute an important alternative to 
viral-based DNA delivery. The overall design of our Human Protocol will be a non-randomized, 
Recombinant DNA Research, Volume 18 
[827] 
