Cystic Fibrosis (CF) is the most common lethal inherited disease, affecting -1/2000
live births. Although the genetic lesion, a mutation in the cystic fibrosis
transmembrane conductance regulator (CFTR) gene, has been elucidated, the exact
mechanism whereby this causes the debilitating disease phenotype is unclear. CF
patients are prone to repeated bacterial infection of the lung; leading to fibrosis of
this tissue, and eventually respiratory failure. Gene therapy has the potential to cure
CF: by introducing a normal copy of the CFTR gene into epithelial cells of the lung,
it may be possible to abolish the lung phenotype. While early gene therapy vectors
utilised a strong viral promoter (e.g. Pcmv) to drive expression of the CFTR cDNA, it
has become apparent that this approach creates problems: expression is short lived
and may not be targeted to the correct cell types at appropriate levels.
CFTR expression is tightly restricted, both temporally and spatially in affected
tissues, notably the lung. Putative regulatory elements have been identified upstream
of the CFTR gene, both proximaliy and distilliy. In addition, a DNase 1
hypersensitive site has been identified in intron 1 of the gene, and has been
associated with an increase in CFTR expression in cells of the gut. Our approach is
to include genomic components of the CFTR locus fused to cDNA in gene therapy
vectors. These 'genomic context vectors' (GCVs) may be capable of recapitulating
the endogenous CFTR expression pattern.
The aim of this project was to investigate the use of the Enhanced Green
Fluorescent Protein (EGFP) as a reporter of CFTR promoter activity. Six vectors
were created coupling portions of the CFTR locus to EGFP in GCVs. Smail plasmids
were made by conventional cloning procedures, while large PAC vectors were made
by a double recombination method employing both homologous and Cre
recombinase/loxP recombination. Some vectors contained an internal ribosome entry
site (IRES) to allow separate translation of proteins from a single mRNA:
These vectors were transfected into permanent cell lines COS"/, MDCK-iowa. T84
and CaC02, in order to assess the effects of the genomic context elements upon
EGFP expression. Several transfection methods were compared to optimise
transfection efficiency, including the liposome DOTAP, the Saint-Mix™ synthetic
amphiphile delivery system, a Polyethyieriimine (PE1) method, and the LID method.
A novel transfection method, the 'SID' method (incorporating Saint-Mix ™ and the
integrin targeting peptide Pb), showed early promise. Ultimately, the LID method
was chosen for further studies, as this method was both, efficient and consistent.
The proximal CFTR 5' region in the pi kbcfproEGFP vector drove expression of
the EGFP transgene at low levels in every celi line analysed. This is in agreement
with previous reports that show basal levels of CFTR expression driven by this
proximal 'housekeeping' region. The additional upstream region in the
PAc65bcfproEGFP vector did not appear to modulate expression in any of the cell
lines analysed. A comparison of the twin vectors pacRC 1 iresEGFP and
pa(PC2 ires EGFP. which differ only in the absence or presence of CFTR intron !
respectively, showed similar levels of expression in the COS7 and MDC'K-iowa cell
lines. Thus, the intron I element does not seem to alter expression in these non-gut
cell lines; this is consistent with reports that show regulation of CFTR expression in
response to the intron element to be specific to cells of the gut epithelium. An
investigation of the intron 1 element in gut cell lines T84 and CaC02 was thwarted
by low transfection efficiency in these cell lines, coupled with inefficient translation
from the IRES.
A comparison of pEGFP-N and pacRC2citivEGFP revealed that large PAC vectors
show an intrinsic reduction in expression in comparison to their small plasmid
counterparts. Further experiments showed that this was not an effect of vector copy
number, and that the effect could not act in trans upon a co-transfected molecule.
These studies also revealed an unexpected interaction: diluting a reporter plasmid
with an anonymous plasmid may actually increase its transfection efficiency.
Permanent cell lines form an important starting point for transfection studies, but
ultimately have proved to be poor models for in vivo gene transfer. An ex vivo
primary air interface sheep tracheal culture was utilised as a more realistic model.
These cultures were characterised by electron microscopy, and demonstrated features
of the native tracheal epithelium. A measure of transepitheliai resistance confirmed
the presence of tight junctions. Cultures were transfected with several of the genomic
context vectors. While PAC vectors had shown a dramatic reduction in expression
relative to their small plasmid counterparts in the in vitro studies, only a small
reduction was seen in the ex vivo cultures, thus PAC vectors, such as GCVs, may
provide a promising approach for gene therapy studies.