rSV40 genome constructs reported to date lacked the 2.6 kb Tag and tag genes, but retained 2 kb containing the SV40 capsid genes, VP1, VP2, and VP3. Since wtSV40 genome is 5.25 kb, deletion of additional SV40 sequences should allow effective packaging of a vector carrying a 4.2 kb version of the human CFTR cDNA. This is shown in Fig. 1. Briefly, pSV5 contains a Tag-deleted SV40 genome with 2 tandem SV40 early promoters (SV40-EP) followed by an engineered polylinker. This modified SV40 genome was cloned as a NotI fragment into pGEM13. An additional 1.8 kb of wtSV40 DNA was excised from pSV5 as a HpaI fragment, to produce pSV5(4.3). Human CFTR cDNA was cloned into the polylinker and rSV40-CFTR was produced from this plasmid according to published procedures (17).

CFPAC cells (CFPAC-1, ATCC) make a mutated form of CFTR that is folded incorrectly (deletion of three nucleotides resulting in a phenylalanine-508 deletion), does not reach the cell membrane, and so is more rapidly degraded than wt CFTR (18). We therefore used these cells to determine the efficiency of rSV40-CFTR transduction and subcellular localization of the delivered protein. CFPAC cells were transduced with rSV40-CFTR or rSV40-HBS, or were mock-transduced. They were then examined by fluorescence microscopy following immunostaining with antibody to CFTR (Fig. 2a). Transduction efficiency, allowing for some cell-to-cell variation in levels of transgene expression, was 85–90% in unselected cells. Higher magnification analysis (Fig. 2b) suggested localization of the immunodetectable CFTR along the cellular membrane.

Therefore, to evaluate more definitively the subcellular localization of the rSV40-CFTR delivered protein, confocal microscopic studies were done. Again, CFPAC cells were transduced with rSV40-CFTR or rSV40-HBS, or mock-transduced, then immunostained for CFTR and viewed using a confocal microscope (Fig. 2c). These studies confirmed that the delivered CFTR localized along the cell membrane. Distribution of the protein was patchy, as has been reported by others (19).

An important measurable activity of CFTR is its function as a cell membrane chloride channel. To test whether rSV40-CFTR transduction of CFPAC cells produced a functional chloride channel, two different assays were used. In the first series of studies, mock- and rSV40-CFTR transduced cells were loaded with the fluorescent Cl-sensitive dye, N- (6-methoxyquinolyl) acetoethyl ester (MQAE, Sigma Chemical Co.) (20), and stimulated by 10 μM forskolin. Permeability to Cl- was assayed as a change in fluorescence as a function of time after adding forskolin. Compared to TC7 cells (CFTR+ normal monkey kidney cells, + control), and mock-transduced CFPAC cells (−control), rSV40-CFTR- restored Cl- channel activity in CFPAC cells to≈50% of wild type levels (Fig. 3).

As an additional test of whether rSV40-CFTR could deliver a functional Cl- channel and to determine if channel activity could be restored in a different cell line, we performed ³⁶Cl isotope tracer efflux assays on IB3–1 cells. CFTR-mutant IB3–1 cells (genotype ΔF508/W1282X) were infected with rSV40-CFTR at an MOI of 100. Mock-transduced T84 cells were the CFTR + control, mock-transduced IB3 cells were the CFTR −control. On day 10 post transduction, the cells were loaded with 5 μCi of ³⁶Cl for 5 hours, and washed to remove excess isotope. The efflux rate was then measured at 30 second intervals for 4.0 minutes. After the final time point, the monolayer was carefully examined for integrity. Flasks with non-intact monolayers were eliminated from the experiment. Samples were counted for Cl³⁶ activity. Transduction with rSV40-CFTR significantly increased Cl- channel activity by this assay, delivering ≈33% of wild type Cl- channel activity (P = 0.01), as compared to control T84 cells (Fig. 4)

Having determined the functional expression of the CFTR channel in vitro, it was necessary to determine if the delivery of a functional chloride channel with rSV40 vectors would also have beneficial effects on a CF lung disease phenotype. Therefore, we used our in vitro characterized vectors to deliver the CFTR gene to Cftr−/− mice that were used to recapitulate a well-characterized Pseudomonas aeruginosa (PA01 strain) agarose bead airway challenge model (21).

Two sets of independent experiments were performed with this model. In the initial Ps-bead challenge experiment matched groups of 5 Cftr−/− mice were treated with 4.0×10⁷ particles of either rSV40-CFTR vector or the irrelevant rSV40-BUGT negative control via intra-tracheal injection and housed in SPF conditions. Due to the highly stable expression of rSV40 vectors it was possible to allow the mice to completely heal form the rSV40 intra-tracheal injection before performing the second intra-tracheal injection with the Pseudomonas bead slurry. Thus, eight weeks later the mice were challenged with 30 ul of a 50 % Pseudomonas aeruginosa (PA01) agarose bead slurry (as determined by 30 minute gravity sedimentation), as previously described (21, 22). This initial bead preparation had a bacterial count of 3.8×10⁶ CFU/ml and resulted in the previously characterized excessive weight loss in Cftr−/− mice. Interestingly, in our studies, mice rescued with rSV40-CFTR had statistically significant less weight loss on the first three days after the bacterial challenge (Fig. 5) when compared to rSV40-BUGT treated mice. On the last and fourth day before sacrifice rSV40-CFTR rescued mice also demonstrated a trend toward less weight loss (p=0.06). The hallmark endpoint of this model has been the excessive weight loss seen in Cftr−/− mice observed in previous studies (21). In our experiments the day 3 weight loss difference between rSV40-BUGT treated mice (~26 %) and rSV40-CFTR rescued mice (~19 %) was ~7 % (Fig. 5).

To determine if these differences in weight lose were consistent with CFTR expression; RT-PCR assays to detect the SV40 driven human CFTR transcript were performed on lung tissue collected 4 days after bacterial challenge. As shown in figure 6, human CFTR expression was detected in mice receiving the rSV40-CFTR, while none was detectable by RT-PCR in mice receiving the BUGT vector. RT-PCR using a no reverse transcriptase control to rule out the possibility of rSV40-CFTR DNA amplification was negative (data not shown). It should be noted that this expression is representative of mice ~ 9 weeks post rSV40-CFTR intra-tracheal delivery and after they have undergone pseudomonas agar bead challenges, thus highlighting the stability of transgene expression achieved by rSV40 gene delivery. This stability is further evidence by a separate experiment in which IHC analysis of lung tissues from mice receiving an AU1 epitope tagged rSV40 vector sacrificed at 2, 4 and 9 weeks post delivery showed no apparent decrease in the expression of the epitope tag at the various time points (data not shown). Further analysis of the mice 9 weeks post delivery revealed that rSV40 transduction as determined by AU1 epitope expression was evident throughout the airways including in mucous producing cells, as determined by overlay of AUI expressing cells and the lysozyme expressing mucous producing cells (figure 7). It has been previously reported that lysozyme is produced by sub-mucosal glands and goblet cells(23, 24).

Further analysis of the lung compartment of these mice revealed that rSV40 mediated correction of the Cftr−/− mice, not only attenuated the weight loss phenotype, but it had also had an effect on the inflammatory cytokine profiles recovered from the broncho-alveolar lavages (BALs). While the cytokine levels in rSV40-CFTR rescued mice all trended lower, IL-1beta was statistically significantly lower (Fig 8). Also importantly for CF related lung disease there was a strong trend (p≤0.1) towards lower KC levels in treated mice. KC is considered the mouse analogue to IL-8 and is mainly responsible for the recruitment of neutrophils into the lung.

In order to determine if this attenuated inflammatory profile in the BALs of rSV40-CFTR recued mice was accompanied by a reduction in the influx of neutrophils and lung pathology, a second set of aged matched mice were IT injected with either rSV40-CFTR or rSV40-BUGT. As in the first experiment described above, 8 weeks after rSV40 delivery, mice were challenged with Ps-bead slurry, albeit of a slightly lower CFU count due to preparation variations (3.0×10⁶CFU/ml). Histopathological analysis of lung sections from this second cohort showed a dramatic reduction in the infiltration and inflammation of the lung tissue (Figure 9, 10). Mice treated with the control vector showed severe inflammatory infiltration of the airways with focal disease in the parenchyma, often accompanied with epithelial disintegration and sloughing. This was accompanied by a prominent presence of neutrophils and type II pneumocyte hyperplasia. These findings were highly attenuated in rSV40-CFTR treated animals; the statistically significant magnitude of these changes can be appreciated in the inflammatory scores obtained from an independent blinded pathologist (table 1). As evidenced in the first set of experiments, KC and other inflammatory cytokine levels in the BAL were lower in CFTR corrected mice and the effect of these attenuated cytokine responses can be appreciated in the significant reduction of neutrophils recovered form the BALs in the second set of experiments (Figure 10).

This vector utilizes a 4.2 kb (full length) human CFTR cDNA paired to 2 tandem SV40 early promoters. rSV40 genome constructs reported to date lacked the 2.6 kb Tag and tag genes, but still carried the 2 kb of wtSV40 capsid protein genes, VP1, VP2, and VP3. Since wtSV40 genome is 5.25 kb, deletion of additional SV40 sequences should allow effective packaging of a vector carrying a 4.2 kb version of the human CFTR cDNA. pSV5 contains a Tag-deleted SV40 genome with 2 tandem SV40 early promoters (SV40-EP) followed by an engineered polylinker. This modified SV40 genome was cloned as a NotI fragment in pGEM13. An additional 1.8 kb of wtSV40 DNA was excised from pSV5 as a HpaI fragment, to produce pSV5(Δ4.3). Human CFTR cDNA was cloned into the polylinker and SV(CFTR) was produced from this plasmid.

This vector utilizes a human bilirubin uridine diphosphate glucuronoside transferase (BUGT) cDNA paired to 2 tandem SV40 early promoters and was used as a control vector for in vivo studies. pSV(HBS) was used as a control vector for in vitro studies.

The resulting Tag-less rSV40 genome, which is carried in pGEM13 (Promega), has been described (12). pSV5(Δ4.3) was made by removing a 1.8 kb HpaI fragment from pSV5. pSVCFTR was made cloning a PCR product from hCFTR into pGEM-TEasy (Promega). A 5′ primer containing an XhoI site (GGCTCGAGGGACCCCAGCGCCCGAGAGA) and a 3′ primer containing a SacII site (CACGTTCTATGTTCCGAAATTGGCGCCGG) were used to make a 4.5 kb PCR product containing the full length CFTR cDNA. This fragment was then cloned into pGEM-TEasy, amplified, and the CFTR cDNA was cut out using SacII and XhoI. This fragment was then cloned into the SacII and XhoI sites in pSV5(Δ4.3), to make pSV5(CFTR). The procedures for making gene delivery vectors from cloned rSV40 genomes have been described in detail elsewhere (17). Briefly, the SV(CFTR) viral genome was excised from this plasmid using NotI, gel purified, recircularized and transfected into COS-7 cells. COS-7 cells, a simian kidney cell line immortalized by stable transfection with a wild type (wt) SV40 genome that is deficient at the origin of replication (16), were maintained in DMEM supplemented with 10% newborn calf serum (NCS, Hyclone). Virus was allowed to grow for 7 days after which lysates were harvested. Lysates were used to amplify SV(CFTR) by infecting COS-7 cells(ATCC, Washington, DC, USA), without helper virus and without additional transfections. Virus was purified by density centrifugation and virus stocks were titered by in situ PCR, as described previously (17). In order to measure the infectivity of replication-deficient SV-40, the viral stocks were titrated by in situ PCR. Infectivity of our rSV40 viral vector stocks was generally approximately 10⁹ infectious units (IU)/ml.

CFPAC cells, a cystic fibrosis pancreatic adenocarcinoma cell line (ATCC, #CRL-1918) (18), were maintained in Iscove’s media supplemented with 10% NCS. Both media were supplemented with 2 mM L-glutamine and 1% penicillin-streptomycin and all cells were grown in 5% CO₂ and 100% humidity. The cells were plated in 4 well slides at a density of 10⁵ cells/well. The cells were transduced with rSV40-CFTR, rSV40-HBS (negative control) or were mock-transduced. Cells were fixed in acetone-methanol (50:50) for 5 minutes. Blocking was performed by 1 hour incubation with 10% goat serum. Cells were incubated for one hour at room temperature with the primary antibody, mouse monoclonal anti-CFTR, C-terminus (Research Diagnostics, Inc., NJ) at a 1:100 dilution, then washed 3 times with PBS, and incubated with 1:100 diluted FITC-conjugated rabbit anti-mouse IgG (Dako, Carpinteria, CA). Mounting medium containing DAPI (Vector Laboratories, Burlingame, CA) was used to visualize the nuclei.

CFPAC cells were transduced with rSV40-CFTR at MOI = 100, or mock transduced. Control cells were TC7 cells, which are normal simian kidney cells (courtesy, Janet Butel, Baylor College of Medicine). To measure Cl- channel activity, a microplate assay technique was used, as described by West and Molloy ²². Specifically, 6d after transduction with rSV40-CFTR, CFPAC cells were plated in 96 well culture plates at a density of 10⁴ cells/well. 48 hours after plating, cells were loaded overnight with 10 mM N-(6-methoxyquinolyl) acetoethyl ester (MQAE). The following day cells were washed with a chloride-containing buffer ( 2.4 mM K2HPO4; 0.6 mM KH2PO4; 10 mM HEPES; 10 mM dextrose; 1 mM MgSO4; 130 mM NaCl). This buffer was then replaced with a chloride-free buffer (2.4 mM K2HPO4; 0.6 mM KH2PO4; 10 mM HEPES; 10 mM dextrose; 1 mM MgSO4; 130 mM NaNO3), containing 10 μM forskolin. Repetitive fluorescence measurements were performed immediately using a Millipore Cytoflour 3050 plate reader (excitation; 360 nm, emission; 460).

IB3–1 cells were seeded at a density of 2×10⁶ cells/T25 flask in LHC-8+ and left overnight in a 37°C humidified CO₂ incubator. The next day monolayers had achieved subjectively between 60–70% confluence. T84 cells were also seeded as a supra-physiologic CFTR positive control and allowed to grow to ~70% at this time. The wells containing IB3–1 cells were assigned into either negative control plates that received PBS treatment or CFTR test plates that received either infected with viral rSV40-CFTR at an MOI of 100 (particles/cell). The virus containing media (LHC-8+) was left on the cells until day 5 and then was changed with 10 ml of fresh LHC-8+ medium. The LHC-8+ medium was replaced again on day 8.

On day 10 cells were loaded with 5mCi of ³⁶Cl for 5 hours at 37°C. Each flask was then transferred into a modified 37°C dry heat incubator and placed upon a metal block heat sink. Each flask was washed 5 times with lactated Ringers solution (LRS) for 30 seconds/wash to remove excess isotope and the efflux rate was then measured at 30 second intervals for 4.0 minutes. All washes and the first efflux collection were obtained by removing the entire fluid volume (1.5 ml) from the flask with a disposable fine tipped transfer pipette, and replacing an equal amount of 37°C lactated Ringers solution (LRS). The replacement solution for second through eighth samples contained a combination of 2.5 mM forskolin, 250 mM 8-Br-cAMP, and 250 mM CPT-cAMP, added to increase intracellular levels of cAMP. At the conclusion of the final time point, the monolayer was carefully examined for integrity. Flasks with non-intact monolayers were eliminated from the experiment. Samples were mixed with CytoSafe scintillation fluid and counted for ³⁶Cl activity. Statistical analyses were conducted by a two tailed t test, P(T≤t). The % change from the maximal post-agonist rate (cpm) and the minimal rate (cpm) prior to that efflux event is defined as the “% change in efflux”. Cells without significant CFTR activity will not stimulate upon exposure to agonist mixture and will display a negative “% change in efflux”. Uninfected negative control IB3 cells (“IB3”), supra-physiologic CFTR positive control T-84 cells, and IB3 cells infected with rSV40-CFTR are shown. t test p values are relative to IB3 neg. control cells.

In the challenge experiments, CFTRtm1Unc-TgN(FABPCFTR)#Jaw mice (gut-corrected CFTR knockout mice) (CFTR−/− )mice were treated intra-trachealy with rSV40-CFTR vector or rSV40-BUGT vector or phosphate-buffered saline (PBS), housed in specific pathogen-free (SPF) conditions for 8 weeks, and then challenged with 30 ul of an approximately 50% Pseudomonas agarose-bead mixture (as determined by 30-min gravity sedimentation) of OD600 2.0 containing ~ 3.0×10⁶ cfu/ml. Mice were randomized with regards to injection order to avoid any unrecognized systematic error. Weights were recorded at time of challenge and at time of sacrifice on day 4. Histopathologic analysis of hematoxylin-eosin stained sections of formalin-fixed, paraffin embedded lung tissue samples taken at the time of sacrifice was performed by a pathologist blinded to the assignment of animals to the two groups.

To ensure the expression of hCFTR by the lungs, following intra-tracheal injections of rSV40-CFTR, lungs were collected from all the CFTR−/− mice. A 100mg of tissue sample was used to extract mRNA using the Qiagen (Valencia, CA) total RNA columns. For Reverse transcriptase first strand synthesis the Invitrogen Superscript III kit was used with the oligo(dt) primers and the optional DNase digestion according to the manufacture’s protocol. This was followed by a PCR using Eppendorf Taq polymerase master mix and gene specific primers for hCFTR (hCFTRfor_5′ aaacttctaatggtgatgacag, hCFTRrev_5′agaaattcttgctcgttgac) or B-actin (B-actin1_5′gctcgtcgtcgacaacggtc, Bactin2_5′caaacatgatctgggtcatcttctc). The no RT controls were subjected to the same PCR conditions and primers but in the absence of reverse transcriptase in the initial cDNA synthesis.

Mice were intra-trachealy injected with an rSV40 vector expressing an AU1 epitope tagged gene, 9 weeks post delivery mice where sacrificed and their lungs were inflated with a 50/50 mixture of PBS and OCT and subsequently frozen in OCT medium. For immunofluorescence, 10 um cryostat sections were processed for indirect immunofluorescence. Blocking was performed by 60 minutes incubation with 10% goat serum in 0.10 M PBS (pH 7.4). Then, cryostat sections were incubated with mouse FITC-labeled anti-AU1 (IgG2a; 1: 100) (Covance, Emeryville, CA) FITC and rabbit anti-human lysozyme (Accurate Chemical and Scientific Corp., Westbury, NY). Incubation with primary antibody was performed for 1 h and followed by incubation for 1 h with secondary antibody TRITC-conjugated goat anti-mouse IgG (Sigma, Saint-Louis, MO), diluted 1: 100. Incubations were at room temperature. Double immunofluorescence was performed as previously described (37). Incubation was followed by extensive washing with PBS. Specimens were finally examined under a Leica DMRBE microscope (Leica Microsystems, Wetzlar, Germany). For negative controls slides were pre-incubated with PBS and 0.01% BSA, non-immune isotype-matches control antibody was substituted for primary antibody, and/or the primary antibody was omitted.

The BAL fluid was retrieved from each animal via cannulation of the exposed trachea and gentle flushing of the lungs with two separate 1 ml aliquots of PBS. Aliquots were pooled for individual animals preceding centrifugation and separation of pelleted cells and supernatant. Cytospin cell preparations made with 200 μls of BAL were stained using Hema 3 (Biochemical Sciences, Swedesboro, NJ) differential stain and relative cell populations were determined using standard morphological criteria. Total cell numbers were calculated by counting the cells on the cytospin slide obtained from the 200 ul BAL aliquot. Assessments of cytokine profiles from the BAL were performed using a commercially available multiplexed kit(Biorad Mouse Multi-Cytokine Detection System; BioRad Laboratories) and the Bioplex Suspension Array System. Simultaneous measurement of several cytokines was performed. All assays were performedaccording to the manufacturer’s protocols. Cytokine concentrations were determinedutilizing Bioplex software with four-parameter data analysis. The sensitivity of the assay is less than 10pg/ml and has a range from 0.2–32,000 pg/ml with an inter and intra-assay CV of less than 10%.