In cystic fibrosis (CF) there is an excessive inflammatory response to lung infections with Pseudomonas aeruginosa, which causes significant morbidity and mortality. Mice deficient in the cystic fibrosis conductance transmembrane regulator homolog (Cftr) have exaggerated production of proinflammatory cytokines in epithelial lining fluid and increased mortality in response to chronic bronchopulmonary infection with mucoid P. aeruginosa, compared with infected wild-type littermates. Whether delivery of CFTR to CF airways by an adenoviral vector (Ad2/CFTR-16) decreases cytokine production and mortality in response to chronic bronchopulmonary infection with mucoid P. aeruginosa was tested. CF mice [stock Cftrtm1Unc-TgN(FABPCFTR)#Jaw] were anesthetized with isoflurane and inoculated intranasally with either Ad2/CFTR-16, diluent (sucrose), or empty vector (Ad2/EV). Two weeks later, mice were anesthetized with 2.5% Avertin and inoculated transtracheally with P. aeruginosa-laden agarose beads (PA M57–15). The cumulative 10-day survival of mice pretreated with Ad2/CFTR-16 was significantly higher compared with mice pretreated with sucrose but not significantly higher than mice pretreated with Ad2/EV. After adjusting for differences in experiment, we found weight loss at 3 days for mice treated with Ad2/CFTR-16 to be significantly less than for the sucrose- or Ad2/EV-treated groups. However, cytokine responses were similar in all groups 3 days after infection. In conclusion, the observed survival advantage of adenoviral delivery of CFTR to the CF lung may be due either to CFTR expression or possibly to proinflammatory effects of the adenoviral vector, or both.
- animal model
- gene therapy
- lung infection
- host response
- cystic fibrosis
- cystic fibrosis transmembrane conductance regulator
cystic fibrosis is one of the most common recessive, lethal genetic diseases in Caucasians. In cystic fibrosis, defective function of CFTR in airway epithelial cells and submucosal glands results in chronic disease of the respiratory tract, manifested by airway obstruction and recurrent infections of the lung and paranasal sinuses, that begins early in life (6, 11). The cystic fibrosis lung is particularly susceptible to Pseudomonas aeruginosa, and this organism plays a critical role in the development and progression of pulmonary disease in cystic fibrosis (11). The pathogenesis of P. aeruginosa lung infection in cystic fibrosis may occur via a series of steps, in which absent or dysfunctional CFTR ultimately leads to reduced volume of epithelial lining fluid, impairing mucociliary clearance of bacteria. P. aeruginosa may thrive in this environment; once they reach a critical mass, a biofilm forms. This phenotypic change further protects the bacteria from antibiotic killing of host defenses, and bacteria persist in the lung (5). Results of human (14), animal (10), and tissue culture (1, 8, 12, 19) studies indicate that with the same stimulus, the response to infection with P. aeruginosa is exaggerated in cystic fibrosis compared with wild type. Because it is the infection, inflammation, and bronchiectasis that ultimately take the patients' lives in cystic fibrosis, any definitive therapy of cystic fibrosis must ameliorate this aspect of the disease if it is to be effective.
CFTR gene therapy has been proposed for the definitive treatment of cystic fibrosis. Gene transfer with adenoviral vectors has been considered for this purpose, since adenovirus infects the airways, it can be delivered by aerosol, and this vector can produce high-level, long-term (at least 70 days) expression in nondividing cells (16). Although delivery of CFTR to cystic fibrosis airways by adenoviral vectors is possible in mice, the impact of this treatment on secondary effects of the disease manifestations has not yet been determined. The purpose of these studies was to determine whether CFTR delivery improves the outcome of a cystic fibrosis murine model of P. aeruginosa lung infection.
MATERIALS AND METHODS
Mice. C57BL/6J male mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Before being used in an experiment, they were acclimated for at least 3 days. Breeding pairs of stock Cftrtm1Unc-TgN(FABPCFTR)#Jaw mice (gut-corrected Cftr knockout mice; stock #2364) and the congenic B6.129P2-Cftrtm1Unc mice (stock #2196) were originally purchased from The Jackson Laboratory, and cystic fibrosis and wild-type mice, respectively, were bred at Case Western Reserve University Animal Resource Center in a specific pathogen-free facility (see Ref. 21 for details of animal husbandry and for pathogens tested). Mice were fed autoclaved Purina Mouse Chow 5010 or irradiated Harlan Teklad 7960. Autoclaved city water in bottles with sipper tubes was available at all times. All mice were maintained in static isolator units (Lab Products, Seaford, DE) on combination size corncob bedding (The Andersons, Maumee, OH) and placed in fresh isolator cages immediately after instillation of adenovirus or P. aeruginosa-laden agarose beads. Gut-corrected Cftr knockout male mice used in these studies ranged in age from 4.6 to 12.6 wk old, with an average age of 9 wk at the onset of the experiment. Heterozygote wild-type male mice used in age-specific infection studies were 6, 9, or 12 wk of age at the time of inoculation with P. aeruginosa-laden agarose beads.
Agarose bead model of chronic bronchopulmonary infection with P. aeruginosa. The agarose bead model of chronic Pseudomonas endobronchial infection was initially developed in rats (3) and later modified for use in mice (17). Briefly, agarose was mixed with tryptic soy broth containing mucoid P. aeruginosa strain M57–15, a clinical isolate from a cystic fibrosis patient, grown to late log phase. The agarose-broth mixture was added to mineral oil that was equilibrated at 50–55°C, rapidly stirred for 6 min at room temperature, and then cooled over 10 min. The agarose beads were washed once with 0.5% deoxycholic acid and sodium salt (SDC) in PBS, pH 7.4, once with 0.25% SDC in PBS, and four times with PBS. Bead diameter was measured by inverted light microscopy using the imaging software Image ProPlus (Media Cybernetics, Baltimore, MD). Quantitative bacteriology was performed on an aliquot of homogenized bead slurry to determine the number of colony forming units (CFU) per milliliter.
Mice were anesthetized with intraperitoneal injections of 2.5% Avertin (2,2,2-tribromoethanol and tert-amyl alcohol in 0.9% NaCl; 0.015 ml/g body wt) and placed in dorsal recumbency, and the ventral cervical area was surgically prepared after plucking the hair by swabbing the area sequentially with Betadine scrub, Betadine solution, and an alcohol swab. A transverse skin incision was made, and the trachea was visualized by blunt dissection. Transtracheal insertion of a 22-gauge 1-inch intravenous catheter was used to instill 50 μl of diluted bead slurry directed preferentially into the right mainstem bronchus.
Bronchoalveolar lavage. Mice were killed by carbon dioxide narcosis followed immediately by exsanguination via direct cardiac puncture. After death, bronchoalveolar lavage (BAL) was performed in situ using a 22-gauge bead-tipped feeding needle ligated to the trachea to prevent backflow. One or three 1-ml aliquots of sterile PBS were used to lavage the lungs for survival or cytokine studies, respectively, and pooled.
Analysis of cytokines in bronchoalveolar lavage fluid. Immediately after collection and before centrifugation, BAL fluid from mice killed 3 days after P. aeruginosa infection was treated with protease inhibitors to a final concentration of 0.1 mM PMSF and 5 mM EDTA. Supernatants were sterile filtered and stored at -70°C until cytokine analysis. Murine TNF-α, IL-1β, IL-6, macrophage inflammatory protein-2 (MIP-2), and keratinocyte chemoattractant (KC/N51; KC) were measured using commercially available ELISAs according the manufacturer's recommended protocols (R&D Systems, Minneapolis, MN). The BAL supernatants were assayed in duplicate, compared with known standards, and the values were corrected for the respiratory epithelial lining fluid (ELF) volume recovered by measuring urea dilution (15). Values that fell below the limits of detection were assigned a value equal to the limits of detection for the assay. A 10-μl aliquot of unprocessed BAL fluid was dotted onto a tryptic soy agar (TSA) plate and examined for the presence or absence of bacterial growth 24 h after being incubated at 37°C. The remaining fluid was centrifuged at 100 g for 10 min at 4°C, the pellet resuspended in 1.0 ml of PBS, and a total leukocyte count was performed using a hemacytometer. A differential cell count was performed on cytocentrifuge preparations (Cytospin 3; Shandon, Pittsburgh, PA) stained with hematoxylin and eosin (H&E) using standard morphometric methods. One study assessed the response of gut-corrected Cftr knockout mice to chronic lung infections with mucoid P. aeruginosa compared with C57BL/6J wild-type mice.
Lung histopathology. After death and BAL, the right lung was inflation fixed with 2% paraformaldehyde in PBS for at least 48 h, cut once midsagitally, and embedded in paraffin. Tissue sections were cut throughout the entire tissue block such that 5-μm sections were taken at regular intervals that encompassed the full craniocaudal range of sections. Sections were stained with H&E using standard techniques and examined for the presence or absence of inflammation using morphometric techniques, as described elsewhere (20).
Quantitative lung bacteriology. Mice were killed 3 days after infection with P. aeruginosa (4–5 × 104 CFU/mouse). The right and left lungs were removed aseptically and placed in 50 ml of cold, sterile PBS. The lungs were then homogenized, and quantitative bacteriology was performed on 10-fold dilutions of the lung homogenate by plating on TSA plates in duplicate.
Effect of age on host response to P. aeruginosa lung infection. Wild-type mice (littermates to the congenic B6.129P2-Cftrtm1Unc mice) were inoculated transtracheally with P. aeruginosa-laden agarose beads on day 0. One experiment was performed to assess lung inflammation 3 days after chronic bronchopulmonary infection with mucoid P. aeruginosa (BAL cytokines and absolute and relative cell counts) and another experiment was performed to assess cumulative survival 10 days after infection.
Adenoviral vectors. Ad2/CFTR-16 is an Ad2-based vector with most of the E1 region (nucleotides 357–3328) deleted and replaced with the CFTR expression cassette (16). Ad2/CFTR-16 contains a cytomegalovirus enhancer-promoter-driven human CFTR as the transgene, followed by a bovine growth hormone poly(A) signal and retains wild-type E2 and E4 regions. The E3 region of Ad2/CFTR-16 has a 1,549-bp deletion in the E3B region corresponding to Ad2 nucleotides 29292–30840.
Instillation of adenoviral vectors. Mice were anesthetized with isoflurane using a vaporizer (VetEquip, Pleasanton, CA) delivered into a clear container into which the animals were placed. The anesthetized animals were removed from the container and immediately inoculated intranasally with 50 μl of one of the following preparations: 109 plaque forming units/mouse of the vector Ad2/CFTR-16 (16) or the empty control vector Ad2/EV (similar to Ad2/CFTR-16, but without CFTR) or 5% sucrose in PBS (the viral vector diluent). All mice survived this procedure.
CFTR gene expression. Gut-corrected Cftr knockout mice were inoculated intranasally with Ad2/EV or Ad2/CFTR-16 and killed 2 wk later. BAL was performed using a single 1-ml aliquot of PBS. After BAL, the left lung was snap-frozen in liquid nitrogen and stored at -80°C until analysis. Quantitative analysis of CFTR was determined as previously described and expressed as the number of copies per 500 ng of lung RNA (16). The right lung was inflation fixed and used to quantitatively assess lung inflammation, as above. One experiment was performed to examine gene transfer and lung inflammation 2 wk after inoculation with adenoviral vector alone.
Evaluating host response in mice pretreated with adenoviral vectors and infected with mucoid P. aeruginosa. Gut-corrected Cftr knockout mice were inoculated intranasally with 50 μl of 5% sucrose, Ad2/EV, or Ad2/CFTR-16. Two weeks later, mice were inoculated transtracheally with P. aeruginosa-laden agarose beads. Mice were then killed 3 or 10 days after P. aeruginosa infection by carbon dioxide inhalation followed by exsanguination via cardiac puncture. Sera were collected for urea analysis. BAL was performed by cannulating the trachea in situ with a 22-gauge 1-½-inch bead-tipped feeding needle, instilling 1-ml aliquots of sterile PBS, and collecting the fluid by gentle aspiration. Lungs were wither processed for histological examination or quantitative bacteriology.
Nine experiments were performed to evaluate the response of gut-corrected Cftr knockout mice to lung infection with P. aeruginosa after pretreatment with Ad2/CFTR-16: two to evaluate quantitative lung bacteriology and three to evaluate the host inflammatory response 3 days after infection with P. aeruginosa, and four to evaluate cumulative survival 10 days after P. aeruginosa infection. For experiments assessing the host inflammatory response 3 days after P. aeruginosa lung infection, beads were diluted 10-fold in sterile PBS (0.12–1.89 × 105 CFU/mouse; average bead diameter 110 μm), and BAL was collected using three aliquots of PBS, and for survival experiments, beads were diluted 20-fold (0.97–2.69 × 104 CFU/ mouse; average bead diameter 116 μm), and BAL was collected using a single aliquot of PBS. Mice that did not survive the surgical procedure or died due to complications from surgery were not included in data analyses. There were no significant differences in surgical deaths based on pretreatment with Ad2/EV, Ad2/CFTR-16, or sucrose (diluent) control (18.5%, 8.9%, and 18.2%, respectively). These high mortality rates are attributed to the delicate nature of cystic fibrosis mice compared with wild-type mice, which experienced <5% mortality under the same surgical conditions (21). Data from two mice pretreated with Ad2/CFTR-16 and killed 3 days after inoculation with P. aeruginosa-laden agarose beads were not included; one died just before it was to be killed, and the trachea from another broke during the BAL collection procedure, so no fluid was collected.
Animal welfare. After P. aeruginosa infection, most mice displayed clinical signs of infection with scruffy coat, moderate dehydration, and decreased activity. Mice that displayed more severe clinical signs (severely dehydrated, hunched posture, appeared painful and unsteady during ambulation) were killed. Mice did not appear severely ill before 3 days after infection. Case Western Reserve University's Animal Care and Use Committee approved all animal protocols. The Animal Resource Center at Case Western Reserve University is an American Association for Accreditation of Laboratory Animal Care-approved institution.
Statistical analysis. Data are expressed as means ± SD. Statistical analyses were performed using the statistical software packages SigmaStat V2.03 (Jandel Scientific, San Rafael, CA) and SAS version 8.2 (Cary, NC). One-way ANOVA was used to analyze age, body weight, and change in body weight. For other variables, the nonparametric Kruskal-Wallis and Wilcoxon's rank sum test were used due to non-normality. When pairwise comparisons among three groups were made, multiple testing was controlled for using the Bonferroni method with significance criterion P < 0.05/3 = 0.0167. When necessary, group comparisons were made after controlling for the experiment by using either Wilcoxon's tests stratified by experiment or ANOVA, with the experiment as a blocking factor. Similarly, analysis of covariance (ANCOVA) was used to make comparisons among groups after adjusting for both experiment and age. Survival curves were calculated using the Kaplan-Meier method. Survival rates at a fixed time point were compared using Pearson's χ2 or Fisher's exact test, and logistic regression.
Response of gut-corrected cystic fibrosis mice to chronic bronchopulmonary infection with mucoid P. aeruginosa. Yu and colleagues (23) reported an increased vulnerability of underweight, calorie-restricted mice to the impact of P. aeruginosa infection. Therefore, whether gut-corrected Cftr knockout mice differed from wild-type mice with respect to body weight and after challenge with P. aeruginosa-laden agarose beads was tested. There were no differences (P > 0.05) in body weight between homozygous wild-type littermates to B6.129P2-Cftrtm1Unc and stock Cftrtm1Unc-TgN(FABPCFTR)#Jaw (gut-corrected Cftr knockout) mice at 7 (4.5 ± 0.7 and 4.8 ± 0.8 g, respectively), 14 (8.1 ± 1.4 and 8.0 ± 1.3 g, respectively), and 21 (11.1 ± 0.7 and 11.3 ± 1.9 g, respectively) days of life (n = 33–37 per group). However, the gut-corrected Cftr knockout mice had significantly greater inflammation compared with C57BL/6J wild-type controls with respect to weight loss (Fig. 1), mortality (80.0% vs. 14.3%, respectively; P = 0.015; Fig. 2), and greater levels of inflammatory mediators present in the ELF compared with C57BL/6J controls (P ≤ 0.006; Wilcoxon's rank sum test) such as the acute-phase cytokine IL-6 (18.7 ± 11.0 and 5.5 ± 2.3 ng/ml ELF, respectively) and the murine neutrophil chemokine KC (11.7 ± 4.3 and 5.1 ± 2.5 ng/ml ELF, respectively; Fig. 3) in response to chronic bronchopulmonary infection with mucoid P. aeruginosa.
Effect of age on host response to P. aeruginosa lung infection. The age range of mice was anticipated to be wide for two reasons. First, empirical evidence suggested that the breeding efficiency of the gut-corrected Cftr knockout mice was poor. The average litter size of gut-corrected Cftr knockout mice was 4.5 pups per litter (range between 1 and 12; n = 6,682 pups from 1,500 litters born 10/22/1999 to 7/1/2003). Females were noted not to have the presence of a vaginal plug within 18–24 h after a male was dropped in the cage, and attempts to synchronize the estrus cycle were not effective. In addition, mice were to be inoculated with adenovirus and then infected with P. aeruginosa 2 wk later. Therefore, whether age affected the response to P. aeruginosa-laden agarose beads was tested. There were no significant differences in host response with regard to cytokine production, absolute or relative cell counts in BAL fluid, or cumulative survival rates among wild-type mice aged 6, 9, or 12 wk (Table 1). The primary cell type found in the infected lung 3 days after infection was the neutrophil; lymphocytosis was not observed.
Response of gut-corrected Cftr knockout mice to P. aeruginosa after Ad2/CFTR-16 delivery: survival experiments. Figure 4 illustrates the cumulative survival of mice over time following mucoid P. aeruginosa infection on day 0. Mice pretreated with Ad2/CFTR-16 had the highest survival rate (48.3%), which in unadjusted comparisons was significantly higher than survival in mice pretreated with sucrose (18.5%, P = 0.0188) but did not differ significantly from those pretreated with Ad2/EV (28%, P = 0.13). When differences in survival among experiments were adjusted for using logistic regression, the Ad2/CFTR-16 group had significantly higher odds of cumulative survival compared with the sucrose group [odds ratio (OR) = 6.82, P = 0.0127], but not compared with the Ad2/EV group (OR = 2.91, P = 0.14). After controlling for experiment and treatment, we found the age of the mice to be not significantly related to survival.
Data on initial age, weight, change in body weight at 3 and 10 days, and lymphocyte and macrophage variables are summarized by group in Table 2. After controlling for differences among experiments, we found between-group differences for change in body weight 3 days after infection as well as for relative lymphocyte counts. With the use of ANOVA to analyze change in body weight at 3 days, both experiment (P < 0.0001) and treatment group (P = 0.007) were significant. After we adjusted for differences in experiment using ANOVA, the weight loss at 3 days for mice treated with Ad2/CFTR-16 was significantly less than for the sucrose- or Ad2/EV-treated groups (P values = 0.005 and 0.009, respectively). After we controlled for experimental effects, mice inoculated with Ad2/CFTR-16 had significantly greater relative numbers of lymphocytes compared with mice pretreated with sucrose (P = 0.0041).
Evaluation of lung inflammation 2 wk after intranasal administration with adenoviral vectors. Because of the unusual pattern of lymphocytosis in mice pretreated with Ad2 after inoculation with P. aeruginosa-laden agarose beads, Ad2-treated mice 2 wk after Ad2 administration that were not inoculated with P. aeruginosa were examined. Mice inoculated with Ad2/CFTR-16 had significantly greater (P < 0.05) relative and absolute numbers of lymphocytes and areal lung inflammation compared with mice inoculated with Ad2/EV (Table 3). Relative numbers of neutrophils were also significantly increased in Ad2/CFTR-16-treated mice compared with those treated with Ad2/EV.
Quantitative analysis of CFTR 2 wk after intranasal inoculation with adenoviral vector. The number of DNA copies in the left lung of cystic fibrosis mice that received Ad2/EV did not differ from that of mice that received Ad2/CFTR-16 (6.67 ± 0.41 × 106 and 6.38 ± 1.07 × 106, respectively). CFTR mRNA expression averaged 15 ± 13 copies/500 ng of lung RNA in mice inoculated with Ad2/EV (n = 3), whereas expression averaged 3.3 ± 2.7 × 104 copies/500 ng of lung RNA in mice inoculated with Ad2/CFTR-16 (n = 6; P = 0.024; Wilcoxon's rank sum test).
Evaluation of inflammation 3 days after P. aeruginosa infection after pretreatment with adenoviral vectors. Results combining data from three individual experiments are shown in Table 4. Because the mice pretreated with Ad2/CFTR-16 were significantly older than mice treated with sucrose (P = 0.012), we carried out further analyses of variables in Table 4, adjusting for both experiment and initial age using ANCOVA models. To better meet assumptions for the ANCOVA, variables were transformed by taking log[%PMN/(100-%PMN)], and using natural logarithms of the other cytokine and lymphocyte and polymorphonuclear neutrophil (PMN) variables, where the constant 1 was added to absolute and relative lymphocyte counts before taking logs to avoid taking logarithms of 0. After we adjusted for age and experiment, absolute and percent lymphocytes were significantly higher in the Ad2/CFTR-16- and Ad2/EV-treated groups compared with the sucrose-treated group (Fig. 5). Also, after we adjusted for age, Ad2/CFTR-16-treated and sucrose-treated groups differed in absolute neutrophil number (P = 0.008).
Areal lung inflammation. Untreated sentinel gut-corrected Cftr knockout mice (Fig. 6A) had little to no lung inflammation (3.4 ± 3.8; n = 3). Two weeks after Ad2/EV (Fig. 6B) or Ad2/CFTR-16 (Fig. 6C) administration, gut-corrected Cftr knockout mice inoculated with Ad2/EV had significantly (P = 0.004) less areal inflammation in the right lung (46.8 ± 3.3%; n = 3) compared with gut-corrected Cftr knockout mice inoculated with Ad2/CFTR-16 (68.1 ± 7.6%; n = 5), with an increase in mononuclear cellular infiltrate in the peribronchial and perivascular spaces. There were no differences in areal lung inflammation in the right lung between gut-corrected Cftr knockout mice pretreated with sucrose (n = 18), Ad2/EV (n = 19), or Ad2/CFTR-16 (n = 21) followed by infection with P. aeruginosa-laden agarose beads 3 days before death (95.2 ± 2.8%, 95.0 ± 3.7%, and 95.3 ± 2.2%, respectively); representative lung sections with P. aeruginosa-laden agarose beads are shown in Fig. 6, D–F, respectively.
Quantitative lung bacteriology. There were no significant differences in bacterial lung load among mice pretreated with sucrose (n = 16), Ad2/EV (n = 17), or Ad2/CFTR-16 (n = 19), followed by infection with P. aeruginosa-laden agarose beads (5.78 ± 7.18, 7.45 ± 7.67, and 7.30 ± 6.29 × 107 CFU/lung, respectively).
Cystic fibrosis is caused by a defect in CFTR, which frequently leads to chronic lung infections with mucoid P. aeruginosa. The cystic fibrosis host inflammatory response to P. aeruginosa is exaggerated and ultimately leads to the death of the patient. It is not known whether expression of exogenous CFTR in the lung will decrease the host inflammatory response or improve survival in response to chronic bronchopulmonary infection with P. aeruginosa. The purpose of this series of studies was to determine whether delivery of CFTR to cystic fibrosis airway epithelia could improve survival and decrease the host inflammatory response to lung infections with P. aeruginosa using a cystic fibrosis mouse model.
Stock Cftrtm1Unc-TgN(FABPCFTR)#Jaw mice bear the S489X mutation in Cftr and the human CFTR transgene driven by the fatty acid binding promoter (FABP), which drives CFTR expression primarily in the intestinal tract. These so-called gut-corrected Cftr knockout mice have been shown to be cystic fibrosis-like with respect to nasal electrophysiology, nasal nitric oxide synthase 2 expression (18), and mRNA expression (24). Until now, no studies of lung infection with P. aeruginosa-laden agarose beads in these gut-corrected Cftr knockout mice had been reported. The studies described here clearly illustrate that these cystic fibrosis mice respond with excessive lung inflammation and increased mortality to chronic bronchopulmonary infection with mucoid P. aeruginosa compared with wild-type mice, much like previously published results (9, 10). Therefore, use of these gut-corrected Cftr knockout mice is an appropriate alternative to the Cftr knockout mice, which are difficult to maintain since they require special animal husbandry for survival (7). In addition, since the gut-corrected Cftr knockout mice grow normally and they can be fed a normal mouse chow (24), it is unlikely the host response is due to nutritional deficits (23).
The use of adenoviral vectors to deliver genes to correct genetic diseases has been studied extensively. For the adenoviral vector used in this study, CFTR mRNA persists in mouse lung for 70 days. Data from transgenic mice indicate that human CFTR can correct the murine cystic fibrosis ion transport defect in airway epithelium (16). These properties encouraged us to undertake experiments with this vector to determine whether expression of normal human CFTR in the airways of cystic fibrosis mice affects their response to lung challenge with mucoid P. aeruginosa.
It was anticipated that cystic fibrosis mice would have a range of ages, due in part to a perception of poor breeding efficiency, and that the mice aged 2 wk from the time of adenoviral vector administration to the time of mucoid P. aeruginosa lung infection. Here, the data showed that the gut-corrected Cftr knockout mice had poor breeding efficiency with low and variable litter sizes. Because space in the vivarium to house large numbers of breeding pairs to compensate was not available, the resulting range of ages was large. Therefore, determining the effect of age was important. Previously, it was shown that initial body weight was not a significant factor in determining clinical outcome after chronic bronchopulmonary infection with mucoid P. aeruginosa (21). However, the effect of age on the host response was not addressed. In wild-type C57BL/6J mice, age (over the range of 6–12 wk) was not found to affect survival or the inflammatory responses 3 days after chronic bronchopulmonary infection with mucoid P. aeruginosa. However, when data from the experiments using the gut-corrected Cftr knockout mice were examined, age was a significant covariate, and therefore data in these analyses were adjusted for age.
The most significant finding in this series of studies was that pretreatment of gut-corrected Cftr knockout mice with Ad2/CFTR-16, but not Ad2/EV, improved survival after challenge with P. aeruginosa-laden agarose beads, compared with gut-corrected Cftr knockout mice pretreated with vehicle alone. Initial examination of the data suggested that the outcome varied by experiment. Differences between experiments were probably due to variations in the inoculum of P. aeruginosa that was applied. Although the P. aeruginosa-laden agarose beads were prepared in as similar a fashion as possible from experiment to experiment, the final inoculum varied over nearly a 10-fold range. However, the effect of Ad2/CFTR-16 on the survival of cystic fibrosis mice following P. aeruginosa-laden agarose bead challenge persisted after statistical adjustment for the effect of age and experiment.
At least two explanations are possible for the differences in survival. One is that continuing expression of human CFTR in the lungs of the cystic fibrosis mice, documented by quantitative RT-PCR, affords protection against death due to inoculation with P. aeruginosa-laden agarose beads. The fact that Ad2/CFTR-16 confers significant protection compared with sucrose vehicle, but Ad2/EV does not, supports this explanation. However, expression of CFTR is modest, and there is not a statistically significant difference between the survival of mice pretreated with Ad2/CFTR-16 and Ad2/EV, which would be expected if the major contributor to the outcome was the expression of the transgene. Moreover, the other hallmark of the response of cystic fibrosis mice to P. aeruginosa-laden agarose beads, the increased inflammatory response, was not diminished in the mice pretreated with Ad2/CFTR-16 compared with those treated with sucrose.
Another possibility is that the inflammatory response to adenovirus itself may both confound the results and contribute to a survival advantage observed for the mice treated with Ad2/CFTR-16. Both Ad2/CFTR-16 and Ad2/EV provoked significant pulmonary inflammation, with increased neutrophils and lymphocytes in BAL fluid and increased area of the lung subsumed by inflammation 2 wk after vector administration, even before the P. aeruginosa-laden agarose beads were applied. This inflammatory response was greater in mice pretreated with Ad2/CFTR-16 than in those pretreated with Ad2/EV. There are possible explanations for this phenomenon. If this preexisting inflammatory response resulted in more prompt killing or containment of the bacterial challenge, it could explain the survival advantage conferred by treatment with adenovirus, with Ad2/CFTR-16 conferring greater advantage than Ad2/EV. However, P. aeruginosa colony counts in the lungs were indistinguishable among the three treatment groups on day 3 after P. aeruginosa-laden agarose bead administration, so there is no evidence for increased bacterial killing in the adenovirus-treated animals. In addition, there was no direct correlation between levels of CFTR expression and bacterial clearance in one study using free P. aeruginosa delivered transtracheally to mice expressing different amounts of human CFTR or murine Cftr in the lung (4). It is also possible that expression of a foreign gene, in this case human CFTR, provides an additional inflammatory stimulus compared with empty vectors. Although these gut-corrected Cftr knockout mice express human CFTR in the gut and should therefore recognize it as “self,” its presentation in the context of adenoviral vector may alter this assumption. Transgenic mice bearing adenoviral vector genes, who should see adenovirus as self, were able to produce a robust inflammatory response against a challenge with adenovirus, that is “break tolerance” (2). Therefore, it is possible that there was no longer tolerance to CFTR once adenovirus was administered to the gut-corrected Cftr knockout mice in these studies.
The lung lymphocytosis was a striking feature of the inflammatory response to adenoviral vectors 14 days after administration and was augmented by challenge with P. aeruginosa-laden agarose beads at day 17 (3 days after bead challenge). Lymphocytosis persisted in the surviving animals 10 days after P. aeruginosa-laden agarose bead challenge (24 days after adenoviral vectors were administered), although these data were censored by the death of most of the animals in the experiment. All groups of animals (pretreated with sucrose, Ad2/EV, or Ad2/CFTR-16) produced an exuberant neutrophilic response to P. aeruginosa-laden agarose beads, and the cytokine responses (TNF-α, IL-1β, IL-6, MIP-2, and KC) were indistinguishable among the groups 3 days after challenge (the last day the data are not censored by death of the animals). In the P. aeruginosa agarose bead model in rats, induction of the T helper type 1 (Th1) lymphocytic response before challenge with P. aeruginosa-laden beads resulted in a better outcome than induction of a Th2 response (13). It is possible that adenoviral challenge produced a Th1 response, as reported by others (22), which may have conferred immunological protection on the animals, thereby improving cumulative survival in mice receiving Ad2/EV or Ad2/CFTR-16. In addition, changes in CD8+ vs. CD4+ response remains to be investigated.
In conclusion, these experiments indicate that the gut-corrected Cftr knockout mice respond to chronic bronchopulmonary infection in an exaggerated fashion to mucoid P. aeruginosa compared with wild-type mice. In addition, administration of Ad2/CFTR-16 to the gut-corrected Cftr knockout mice conferred a survival advantage to subsequent challenge with P. aeruginosa-laden agarose beads compared with administration of the sucrose vehicle, but not Ad2/EV, and this survival improvement occurred without reduction in either inflammatory response or lung burden of bacteria. Therefore, correction of the CFTR defect in airway epithelial cells using an adenoviral vector may not necessarily improve the clinical outcome of cystic fibrosis patients. The beneficial effect of adenoviral delivery of CFTR observed in cystic fibrosis mice may be due either to CFTR expression or due to other effects of the adenoviral vector itself, probably via proinflammatory effects, or to some combination of these factors.
Genzyme Corporation provided the adenoviral vectors.
A. Scaria is an employee of and has stock ownership of Genzyme. S. Wadsworth is an employee of Genzyme and has stock ownership of and patents pending for Genzyme.
We express our appreciation to the following: the Case Western Reserve University (CWRU) Pediatric Inflammatory Mediator Core, especially Christopher Statt, for performing cytokine assays; personnel in CWRU Cystic Fibrosis Animal Core, most notably Lisa Hogue, Christiaan van Heeckeren, Merle Fleischer, James Poleman, Veronica Peck, and Alma Wilson for providing expert technical assistance; Claudia Garner for sectioning and staining lung specimens; Neal Heilman for point counting lung histology sections; Dr. Ronald Walenga for helpful discussions; and Donna Hempel for quantifying CFTR mRNA expression.
This work was supported by National Institutes of Health Grants HL-60293 and DK-27651 and a Research Development Program Grant from the Cystic Fibrosis Foundation.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society