The relationships between airway epithelial Cl− secretion-Na+ absorption balance, airway surface liquid (ASL) homeostasis, and lung disease were investigated in selected transgenic mice. 1) To determine if transgenic overexpression of wild-type (WT) human CFTR (hCFTR) accelerated Cl− secretion and regulated Na+ absorption in murine airways, we utilized a Clara cell secretory protein (CCSP)-specific promoter to generate mice expressing airway-specific hCFTR. Ussing chamber studies revealed significantly (∼2.5-fold) elevated basal Cl− secretory currents in CCSP-hCFTR transgenic mouse airways. Endogenous murine airway Na+ absorption was not regulated by hCFTR, and these mice exhibited no lung disease. 2) We tested whether hCFTR, transgenically expressed on a transgenic mouse background overexpressing the β-subunit of the epithelial Na+ channel (β-ENaC), restored ion transport balance and ASL volume homeostasis and ameliorated lung disease. Both transgenes were active in CCSP-hCFTR/β-ENaC transgenic mouse airways, which exhibited an elevated basal Cl− secretion and Na+ hyperabsorption. However, the airway disease characteristic of β-ENaC mice persisted. Confocal studies of ASL volume homeostasis in cultured tracheal cells revealed ASL autoregulation to a height of ∼6 μm in WT and CCSP-hCFTR cultures, whereas ASL was reduced to <4 μm in β-ENaC and CCSP-hCFTR/β-ENaC cultures. We conclude that 1) hCFTR overexpression increases basal Cl− secretion but does not regulate Na+ transport in WT mice and 2) transgenic hCFTR produces increased Cl− secretion, but not regulation of Na+ channels, in β-ENaC mouse airways and does not ameliorate β-ENaC mouse lung disease.
- sodium absorption
- chloride secretion
- CFTR overexpression
in airway epithelia, CFTR functions as a cAMP-activated Cl− channel (2) and also regulates other channels, including the epithelial Na+ channel (ENaC) (35, 28), although, on the basis of Ussing chamber data, recent reports question the capacity of CFTR to regulate ENaC (6, 16). In human cystic fibrosis (CF) airway epithelia, studies of airway surface liquid (ASL) homeostasis utilizing thin-film confocal techniques have suggested that the inability to regulate Na+ absorption, coupled with absent CFTR-mediated Cl− secretion as a function of ASL volume, plays a central role in the pathophysiology of CF airway disease (30, 23). Interestingly, adult mice express little endogenous CFTR in the lower airways (22, 32, 37). The normally low level of CFTR expression, with alternative paths to regulate Na+ absorption and alternative Ca2+-activated Cl− channels to mediate Cl− secretion, may explain why CF mice exhibit relatively normal airway ion transport and do not exhibit spontaneous airway disease.
To determine whether expression of wild-type (WT) human CFTR (hCFTR) could function as a cAMP-regulable Cl− channel and regulate endogenous murine ENaC in mouse airways, we generated transgenic mice expressing hCFTR under the control of the airway-specific Clara cell secretory protein (CCSP) promoter. As a first test for a relationship between hCFTR expression and murine airway ion transport, we measured rates of Cl− secretion and endogenous ENaC-mediated Na+ absorption in tracheas freshly excised from mice overexpressing hCFTR. As a second test of CFTR function in mouse lower airways, we asked whether transgenic expression of hCFTR could correct the pulmonary phenotype of the β-ENaC transgenic mouse. The β-ENaC transgenic model has features of chronic CF lung disease, including accelerated Na+ absorption, ASL volume depletion, goblet cell metaplasia, and mucus plugging of the airways (20). Specifically, we hypothesized that, by crossing CCSP-hCFTR transgenic mice with the β-ENaC mice, we would restore ASL volume by inducing Cl− secretion through CFTR and/or downregulating transgenic overactive ENaC and, consequently, ameliorating ASL depletion and the pulmonary phenotype (22).
Accordingly, CCSP-hCFTR transgenic mice and CCSP-hCFTR × β-ENaC (CCSP-hCFTR/β-ENaC) mice were generated, litters were genotyped, and survival was monitored. Airway bioelectric properties of freshly excised airways and cultured preparations were measured in Ussing chambers, ASL volume/height was measured with confocal microscopy in cultured preparations, and lungs were subjected to histological analysis to characterize the lung phenotype in the transgenic mice.
All mouse studies were approved by the University of North Carolina (UNC) Institutional Animal Care and Use Committee.
Generation of transgenic mice.
Transgenic mice with CCSP-mediated overexpression of hCFTR (referred to as “hCFTR”) were generated using the TG-1 base vector (24), the rat CCSP promoter (13), and hCFTR cDNA. hCFTR cDNA contains the entire coding sequence, as described elsewhere (26, 38), with the non-protein-coding modifications as described by Drumm et al. (8). Mice were generated using standard methods for pronuclear injection conducted in the UNC Animal Models Core Laboratory following Institutional Animal Care and Use Committee-approved protocols. Founder animals (C57Bl/6J × DBA2/J F1) were bred to C57Bl/6J × DBA2/J F1 animals to evaluate germline transmission. Six lines that transmitted the hCFTR sequence were analyzed by RT-PCR for expression of hCFTR in lung and tracheal tissue (not shown). On the basis of these results, three murine lines with expression of hCFTR in the lung and trachea were chosen for colony expansion and two lines, lines 4 and 6, were analyzed further. Line 6 mice were also bred to β-ENaC mice (mixed background, C57Bl/6J × C3H/J). All transgenic mice (hCFTR, β-ENaC, and hCFTR/β-ENaC) were maintained as heterozygotes for each of the transgenes. To minimize the effect of genetic heterogeneity, age- and sex-matched littermate controls were utilized for all experimental purposes, including survival data. Genotyping for the transgene was conducted by the UNC Pathology Mouse Genotyping Core using TaqMan-based PCR for the CFTR and/or the Scnn1b (β-ENaC) transgenes.
Mouse genotyping protocol.
Toe or tail samples were homogenized in tubes with 500 μl of lysis buffer (Applied Biosystems, Foster City, CA) that included proteinase K (10 mg/ml; Invitrogen, Carlsbad, CA). Lysates were incubated at 50°C for 30 min and then at 90°C for 5 min. The pellets were quickly spun down, and the liquid lysates were diluted 1:40 in double-distilled water in 96-well plates; diluted lysates were kept at −20°C for several hours. The crude lysate (10 μl) was used for real-time PCR by addition of 20 μl of reaction mixture. The real-time PCR was performed with a sequence detector (model 7500, Applied Biosystems) using standard cycles (40 cycles at 95°C for 15 min, 95°C for 30 s, and 60°C for 1 min). For hCFTR detection, the forward primer was AGC CTT TAG AGA GAA GGC TG, the reverse primer was TGC TGA TCA CGC TGA TGC GA, and the probe was FAM-CGC CTC TCC CTG CTC AGA ATC TGG-TAMRA. For mouse β-ENaC detection, the forward primer was GAC ACC CAG TAT AAG ATG ACC, the reverse primer was CCT GAG ACA GGA CAT GTA TG, and the probe was FAM-CTG ACT GGC CAT CTG AGG CCT CTG-TAMRA. For endogenous control, the multiplex PCR was applied by the Mus β-actin detection system; the forward primer was CTG CCT GAC GGC CAG GTC, the reverse primer was CAA GAA GGA AGG CTG GAA AAG A, and the probe was TET-CAC TAT TGG CAA CGA GCG GTT CCG-TAMRA. Positive or negative transgenes were determined by the cycle threshold (ΔCT) method.
RT-PCR analysis of CFTR expression.
Total RNA was isolated from freshly isolated mouse tracheas using the RNeasy kit (Qiagen) and reverse-transcribed into cDNA using SuperScript (Invitrogen). PCR was performed using standard procedures and AmpliTaq Gold (Applied Biosystems). Controls included amplifications performed on samples prepared identically with no RT and amplifications performed with no added substrate (water control). For quantification of murine CFTR and hCFTR, a standard was first prepared by amplifying the fragment of the murine CFTR or hCFTR cDNA and cloning it into a plasmid vector (25). The plasmid was purified, quantitated, diluted to 5 × 10−6–5 × 10−9 ng/μl, and used to produce a standard curve. Quantitative PCR was performed using a LightCycler PCR machine and a LightCycler Fast Start DNA Master SYBR Green I kit (Roche Applied Science, Indianapolis, IN). Each reaction was performed in duplicate on RNA isolated from three separate animals (line 4). The average crossover point was determined using the Roche software and converted to copy number per nanograms of total RNA by comparison with standard curves. Primers used to amplify hCFTR were 5′-TGACACACTCAGTTAACCAAGGTCAG-3′ and 5′-CCTCTGAAGAATCCCATAGCAAGCAA-3′. Primers for murine CFTR were 5′-ACGTTCACACCCAACTCAGGCTCC-3′ and 5′-GAAGCAGCCACCTCAACCAGAAAAA-3′.
Analysis of hCFTR protein expression.
Immunoprecipitation of CFTR was performed using procedures similar to those previously described (20, 25). Briefly, tracheas were removed from hCFTR-positive and -negative animals and homogenized in M-PER lysis solution (Pierce, Rockford, IL) containing proteinase inhibitors [protease inhibitor cocktail (Sigma) plus PMSF (catalog no. 7626, Sigma)]. After centrifugation, the supernatant was collected, protein concentration was measured by the bicinchoninic acid method (Pierce), and the samples were stored at −80°C. Monoclonal anti-CFTR antibody 24-1 (R & D Systems, Minneapolis, MN) was added (200 μg of tracheal extract) with 0.09% NP-40 and proteinase inhibitors, and the sample was incubated at 4°C with gentle rocking for 3–5 h. Protein A/G agarose beads (Immunopure, Pierce) were added, and the samples were rocked overnight. The samples were washed thoroughly, and bound proteins were eluted with sample buffer and separated on a 6% Tris-glycine gel (Invitrogen). Proteins were transferred to polyvinylidene difluoride membranes (Hybond-P, Amersham, Piscataway, NJ) in 20% methanol in Tris-glycine transfer buffer (Invitrogen) at 35 V for 2 h. Membranes were blocked in 5% milk, probed with monoclonal antibody 596 (ascites fluid; generously provided by Dr. J. Riordan) (21), and developed with enhanced chemiluminescence Western blotting reagents (ECL, Amersham).
Trachea and lungs were rapidly dissected and frozen directly in optimal cutting temperature (OCT) embedding compound; lungs were inflated with OCT compound before embedding. Thin (6- to 10-μm) sections were cut with a cryotome and mounted on glass slides. Immunostaining with β-ENaC polyclonal serum (9) was performed as previously described (18). For immunostaining with CFTR monoclonal antibodies (19, 21), frozen tissue sections were fixed in cold 100% acetone and processed using the M.O.M. Basic Kit (Vector Laboratories, Burlingame, CA) according to the manufacturer's instructions, with Texas Red-streptavidin (Jackson Immunoresearch Labs, West Grove, PA) used for CFTR detection. In the CFTR costaining experiments, antibodies against β-ENaC or tubulin (Chemicon, Temecula, CA) were added before the last step, and detection was achieved by addition of Texas Red-streptavidin in the presence of FITC-labeled IgG anti-rabbit (β-ENaC) or FITC-labeled IgG anti-rat (tubulin). Tissues were mounted in Vectashield containing 4,6-diamidino-2-phenylindole to label nuclei (Vector Laboratories). Photomicrographs were taken on a confocal microscope (model SP2 AOBS, Leica) using ×20 Plan Apochromat 0.7 numerical aperture and ×40 Plan Apochromat 1.25 numerical aperture lenses and independent laser sources (Leica).
Preparation of native tissue for bioelectric studies.
Adult mice (both sexes) were euthanized by CO2 inhalation, and neonates were killed by an overdose of ketamine-xylazine. The tracheas were immediately excised and mounted in Ussing chambers, as previously described for adult (11) and neonatal mice (10). In most experiments, WT mice and littermate control transgenic mice were studied simultaneously. Ussing chamber measurements of the bioelectric properties of tracheal epithelia were made under short-circuit current (Isc) conditions, as previously described (11). Tissues were bathed bilaterally in Krebs-Ringer bicarbonate solution in most studies; in some preparations, they were bathed bilaterally in Cl−/HCO3−-free Ringer solution (10).
Preparation of cultured tissues for bioelectric studies.
The tracheas were removed from mice euthanized with CO2 and cultured as previously described (12). Once cells were confluent, the T-Col inserts were studied in Ussing chambers fitted with adaptors to accommodate T-Col membranes (Warner Instruments). For electrophysiological characterization, the tissues were bathed bilaterally (10 ml/side) in Krebs-Ringer bicarbonate solution and studied under open-circuit conditions with a voltage clamp (Physiologic Instruments). The transepithelial electrical potential difference was continually recorded, and a constant-current pulse (2–10 μA) was applied across the tissue at 1-min intervals to calculate tissue resistance. From these measurements, the equivalent Isc was calculated. All other details of Ussing chamber techniques have been previously published (11).
Protocols for bioelectric characterization of tracheal preparations.
For bioelectric studies of native and cultured trachea, amiloride (10−4 M) was added to the apical surface to block electrogenic Na+ absorption. Forskolin (10−5 M) and UTP (10−4 M) were added to the apical surface to induce anion secretion via an increase in cellular cAMP and Ca2+, respectively. Inh-172, used to block Cl− secretion through CFTR (33, 36), was a generous gift from Dr. Alan Verkman. All drugs for Ussing chamber studies were purchased from Sigma, with the exception of UTP (Amersham Pharmacia Biotech) and Inh-172.
Measurement of in vitro ASL height.
Thirty microliters of PBS containing 0.2% (vol/vol) Texas Red-dextran (10 kDa; Molecular Probes) were added to the apical surface of freshly washed tracheal cultures to visualize the ASL layer. This volume of PBS results in an initial ASL height of ∼20–30 μm, as previously described (29). Images of the Texas Red-labeled ASL were acquired by laser-scanning confocal microscopy (model 510, Zeiss) using the appropriate filters for Texas Red (540-nm excitation and 630-nm emission). To avoid evaporation of the thin ASL layer, 100 μl of immiscible perfluorocarbon (Fluorinert-77, 3M) were added to the airway surface following the addition of the labeling dye (29). The height of the ASL was measured by averaging the heights obtained from xz scans of five predetermined points on the culture. ASL height was measured immediately following the addition of the Texas Red-dextran (time 0) and at designated times over 24 h in WT, β-ENaC, hCFTR, and β-ENaC/hCFTR-overexpressing mouse tracheal cultures.
Values are means ± SE, with the number of preparations shown in parentheses. To determine if significant differences were present between survival curves, Kaplan-Meier survival analyses (log-ranked) were used. For the bioelectric studies, Student's t-test was used to determine significance when two groups were analyzed. For comparison of more than two groups, a one-way ANOVA was used to determine if there was an overall significant difference between the groups. If there was a statistically significant difference among the groups, the Holm-Sidak test was used for all pair-wise comparisons.
CCSP-hCFTR transgenic mice.
In the two CCSP-hCFTR transgenic lines generated for this study (lines 4 and 6), progeny were produced in the expected Mendelian ratio, and postnatal survival was not affected by the expression of the transgene (Fig. 1A). The hCFTR mice had no observable disease phenotype and exhibited normal lung histology (Fig. 1, B and C).
Analysis of CFTR expression.
The expression of the hCFTR transgene was examined at the mRNA level by quantitative RT-PCR using total RNA isolated from tracheas excised from transgenic and WT animals. To determine the level of transgenic hCFTR expression relative to the endogenous level of murine CFTR, primers specific for murine CFTR and hCFTR were utilized. The level of endogenous murine CFTR was unchanged in the transgenic animals and estimated at less than one copy of CFTR mRNA per cell. As expected, no hCFTR was detected in the WT animals. In contrast, the transgenic animals expressed approximately fivefold higher levels of hCFTR than murine CFTR (∼3 copies/cell). These data demonstrate that although the transgenic animals expressed the hCFTR mRNA in the trachea at levels higher than the endogenous level of murine CFTR, the hCFTR transgene was not highly overexpressed.
To test whether the expressed transgenic mRNA was being translated into CFTR protein, immunoprecipitation was performed on tracheal extracts using monoclonal antibodies specific for hCFTR (21). As shown in Fig. 2, tracheal extracts from transgenic animals contained a protein band of the predicted molecular weight range for hCFTR. The majority of the CFTR protein in the transgenic animals appeared to migrate as a higher-molecular-weight species than CFTR expressed in Calu-3 cells, suggesting that hCFTR in the transgenic murine trachea is a more heavily glycosylated form of CFTR. In contrast, extracts from WT animals exhibited no signal.
Bioelectric characterization of freshly excised hCFTR tracheas.
The bioelectric properties of the two CCSP-hCFTR lines were very similar; thus, bioelectric data are shown only for line 6 (Fig. 3A). The basal Isc of the CCSP-hCFTR transgenic tracheas was ∼2.5-fold greater than that of the WT tracheas (P ≤ 0.001; Fig. 3A). The response to amiloride (an imperfect index of electrogenic Na+ absorption, because amiloride can induce Cl− secretion) did not differ between the WT and CCSP-hCFTR transgenic mice. Accordingly, the post-amiloride residual Isc was significantly greater in the transgenic tracheas. Interestingly, the responses to addition of forskolin and UTP were significantly smaller (P ≤ 0.05) in the tracheas from the CCSP-hCFTR than WT mice.
To test whether the elevated basal Isc in the transgenic mice indeed reflected anion secretion, we studied WT and CCSP-hCFTR transgenic tracheas in bilateral Cl−/HCO3−-free buffer (Fig. 3B). This ion-substitution protocol decreased basal Isc to very low values, which did not significantly differ between genotypes. The response to amiloride in Cl−/HCO3−-free solutions is a better index of Na+ transport rates. The amiloride-induced reductions in Isc were similar for the two genotypes, suggesting that hCFTR did not regulate endogenous ENaC. The responses to forskolin and UTP were attenuated and not significantly different for the two genotypes in the buffer containing no Cl−/HCO3−. Therefore, it appears that the raised basal Isc in the CCSP-hCFTR tracheas reflected raised anion secretion. We did not use bumetanide in an attempt to block Cl− secretion, as we previously demonstrated that, in adult tracheas, bumetanide is ineffective in blocking Cl− secretion (10).
We next investigated whether the elevated anion secretion in the CCSP-hCFTR transgenic tracheas was mediated via CFTR. For these experiments, we applied the CFTR inhibitor Inh-172 (10 μM) to the apical surface (33, 36). Unfortunately, we previously found that this compound was ineffective in blocking CFTR-mediated Cl− secretion in native murine tissue (including gut and gallbladder; unpublished observations). For the CCSP-hCFTR transgenic preparations, this drug inhibited basal Isc by only ∼10% (Δ11.5 ± 2.1 μA/cm2, n = 8). In the WT preparations, the drug decreased basal Isc by 6.2 ± 1.9 μA/cm2, which did not differ significantly from the response exhibited by the CCSP-hCFTR transgenic preparations. We have found this CFTR blocker to be more effective in cultured preparations (see below).
We speculated that the raised basal anion secretion and relative lack of response to forskolin in the CCSP-hCFTR mice might reflect the possibility that endogenous cAMP levels were maintained at relatively high levels as a result of prostaglandin release, as previously noted in canine trachea (1). Therefore, tracheas of each genotype were incubated in 10−5 M indomethacin for 30 min to block prostaglandin synthesis before the bioelectric data were obtained. This maneuver had no significant effect on the basal Isc or the response to amiloride, forskolin, or UTP (data not shown) in either genotype, indicating that endogenous prostaglandins do not appear to be involved in regulating basal Isc in WT or CCSP-hCFTR transgenic mice.
We next investigated whether adenosine, formed locally from released ATP and interacting with apical A2b receptors (15), might be involved in regulating the basal level of cAMP and, hence, the elevated basal Isc of the CCSP-hCFTR transgenic preparations. Thus we treated preparations with apical adenosine deaminase (2 U/ml) and 8-(p-sulfophenyl)theophylline (10−4 M), an adenosine receptor blocker, for 10 min before measuring the bioelectric parameters. Neither the basal Isc nor the response to any drug addition was changed in the WT or hCFTR preparations as a result of this protocol. Thus adenosine did not appear to play a role in the elevated Isc of the transgenic preparations (data not shown).
In a second approach to characterize basal Isc, we applied a compound that decreases Cl− secretion through the Ca2+-activated Cl− conductance (CaCC), i.e., DIDS. DIDS would be expected to significantly decrease the basal Isc if this current were a result of anion secretion mediated by CaCC. This blocker had no significant effect on the residual (post-amiloride) Isc of CCSP-hCFTR or WT mice (Fig. 3C). The forskolin response was attenuated in the WT tracheas (P ≤ 0.05) after application of DIDS, but not in tracheas from CCSP-hCFTR mice (Fig. 3C). As expected, DIDS attenuated the response to UTP, which was small and did not differ significantly between genotypes (compare Fig. 3B with Fig. 3C). We previously noted that forskolin in mouse tracheal epithelia releases Ca2+ and activates CaCC (11), confusing an interpretation of forskolin stimulating increases in Cl− secretion via CFTR. Thus we speculate that the significantly greater forskolin response in DIDS-treated CCSP-hCFTR transgenic than WT mice reflects the elimination of the CaCC-mediated “cross talk” in the response, revealing a CFTR-mediated Cl− current.
To further evaluate the contribution of CaCC to the basal Isc of the tracheas from the CCSP-hCFTR transgenic and WT mice, we depleted the intracellular Ca2+ stores and eliminated the influx of extracellular Ca2+. Accordingly, tracheal preparations were incubated for 30 min in 2 μM thapsigargin and then placed in the Ussing chambers in nominally Ca2+-free Krebs buffer solutions. This maneuver reduced basal Isc in both genotypes but did not significantly decrease the fold difference between WT and hCFTR animals: basal Isc = 20.3 ± 2.0 and 54.9 ± 8.5, respectively (n = 6 in each group, P ≤ 0.01).
Cultured trachea bioelectric properties.
Because we previously found that Inh-172 is more effective in cultured than native tissue, we studied the bioelectric properties of cultured trachea from mice of both genotypes. The bioelectric properties of the cultured tissue were qualitatively similar to those of the freshly excised tracheas (Fig. 3D). The basal Isc was significantly elevated in preparations cultured from CCSP-hCFTR transgenic mice compared with WT mice, whereas there were no differences in amiloride-sensitive current (Fig. 3D). After amiloride treatment, the Isc of the CCSP-hCFTR transgenic preparations was significantly inhibited by Inh-172, whereas the current of the WT preparations was not significantly changed (Fig. 3D). After Inh-172 treatment, the residual currents also did not differ significantly between the preparations of the two genotypes. Interestingly, despite the presence of the CFTR blocker, the forskolin response of the CCSP-hCFTR transgenic preparations was significantly enhanced compared with that of the WT preparations (P ≤ 0.001). In cultured preparations, the cross talk between cAMP and CaCC is decreased (unpublished observations; Ref. 7), which may explain why the response to forskolin is significantly greater in the CCSP-hCFTR cultured cells. The response to UTP in these preparations did not differ between the two genotypes.
Neonatal CF mouse trachea × hCFTR.
While the lower airways of the adult CF mouse exhibit no Cl− secretory defects (11), we previously found that the neonatal CF mouse trachea exhibits a defect in cAMP-mediated Cl− secretion (25). Thus, to determine if hCFTR under the control of the CCSP promoter could correct this defect in the neonatal CF mouse trachea, we crossed the CF mouse (cftrtm1unc) with the hCFTR transgenic mouse. As in the adult mouse, the basal Isc was elevated in the neonatal WT hCFTR transgenic tracheas compared with WT and CF preparations (Fig. 4). The basal Isc was also significantly elevated in CF/hCFTR transgenic mice compared with WT or CF tracheas. There was no significant difference in the response to amiloride between the four groups of mice (Fig. 4). The changes in the residual (post-amiloride) Isc were the same as those in the basal Isc. The response to forskolin was significantly attenuated in the tracheas of the CF vs. WT pups, whereas the pups expressing the hCFTR transgene (on both WT and CF background) exhibited large responses to forskolin (Fig. 4). The magnitude of the bumetanide response paralleled that of the forskolin response. As in the tracheas from adult mice, the UTP response was significantly attenuated in both groups of neonatal tracheas expressing the hCFTR transgene.
Double-transgenic hCFTR × β-ENaC mice.
We crossed the hCFTR mice with the β-ENaC mice to determine if transgenically expressed hCFTR would rehydrate β-ENaC mouse airways directly via increased CFTR Cl− secretory function and/or downregulation of ENaC in the β-ENaC mice. As we previously reported (20), the β-ENaC mice on a mixed-strain background exhibited an increased mortality compared with WT mice (Fig. 5A). Survival of the hCFTR/β-ENaC double-transgenic mice was not significantly different from that of the β-ENaC mice (Fig. 5A).
Double-transgenic (hCFTR × β-ENaC) lung pathology.
As we previously reported (20), the airways of the β-ENaC mice exhibited significant mucus plugging. The airways of the hCFTR/β-ENaC mice also exhibited significant mucus plugging (Fig. 5, B and C), suggesting that the hCFTR did not protect β-ENaC mice from lung pathology.
Immunolocalization of CFTR and β-ENaC.
Localization of CFTR and β-ENaC transgenic protein expression was achieved by immunofluorescence and confocal microscopy analysis using specific antibodies for hCFTR (19, 21) and murine β-ENaC (9). In general, the level of CFTR transgenic protein expression was low and required an amplification step in the immunostaining protocol. Transgenic CFTR protein was scarcely detectable in the large airways (data not shown), and its expression increased distally (Fig. 6, B–D). In the small airways, CFTR was expressed in a continuous pattern, but cells displayed different levels of immunostaining signal. Importantly, transgenic hCFTR expression in the CCSP-hCFTR transgenic and CCSP-β-ENaC/CCSP-hCFTR bitransgenic mice was indistinguishable (Fig. 6, B and C).
Expression of transgenic β-ENaC was identical in the CCSP-β-ENaC/CCSP-hCFTR bitransgenic and CCSP-β-ENaC transgenic mice (Fig. 6). Transgenic β-ENaC was abundantly expressed in a continuous pattern along the superficial epithelium of the large airways (i.e., trachea; data not shown). In the small airways, cells expressing β-ENaC were distributed throughout the epithelium (Fig. 6, A, C, and D) and appeared to be less brightly stained than their counterparts in the large airways. Scattered cells expressing β-ENaC in alveoli were also observed (Fig. 6, A and C).
Expression of transgenic β-ENaC and CFTR was restricted to the nonciliated cells in the large and small airways (data not shown). Although most of the cells appeared to express both proteins (Fig. 6, C and D), some cells exhibited higher levels of transgenic CFTR or transgenic β-ENaC protein (Fig. 6, C and D).
Bioelectric properties of transgenic CCSP-hCFTR/CCSP-β-ENaC mice.
The basal Isc values of tracheas from hCFTR, β-ENaC, and CCSP-hCFTR/CCSP-β-ENaC transgenic mice were statistically different from those of tracheas from WT mice (Fig. 7). While the basal Isc values of the β-ENaC and CCSP-hCFTR transgenic mice did not differ from each other, the basal Isc of the CCSP-hCFTR/β-ENaC double-transgenic mice was significantly greater than that of either single-transgenic group, suggesting that the components that contributed to the raised basal Isc of the two transgenes were additive, i.e., raised Na+ transport in β-ENaC transgenic mice and raised Cl− secretion in hCFTR transgenic mice. The amiloride-sensitive Isc was significantly elevated in the β-ENaC mice and the double-transgenic (hCFTR/β-ENaC) mice. The residual (post-amiloride) Isc was significantly elevated in both transgenic genotypes carrying the hCFTR transgene compared with the WT or the β-ENaC transgenic mice. The response to forskolin was uniformly small and did not differ between groups of mice. The response to UTP was large in the WT and β-ENaC transgenic mice but was significantly attenuated in both groups of mice expressing the hCFTR transgene (Fig. 7).
In vitro ASL height regulation.
We next assessed the role of the hCFTR and β-ENaC transgenes, individually and in combination, in ASL volume homeostasis under physiological “thin-film” conditions in mouse tracheal cultures (29). ASL volume was monitored in cultures that had been exposed to a small volume liquid challenge.
In WT cultures, the added liquid (30 μl, added at time 0) was absorbed over a period of ∼6 h, and the ASL height was maintained at levels corresponding to extended murine cilia (∼6.81 ± 0.6 μm; Fig. 8). This biphasic nature of ASL volume regulation, i.e., rapid removal of liquid from airway surfaces followed by a steady-state volume, suggests that active ion transport systems shifted from an absorptive to a balanced phenotype, involving the regulation of Na+ absorption and Cl− secretion (4). The net effect of this regulation was to maintain ASL at heights compatible with proper mucociliary clearance.
In tracheal cultures from mice overexpressing hCFTR alone (Fig. 8), the initial absorption of the added volume tended to be slightly slower than in paired WT cultures. However, at 24 h following the delivery of the liquid challenge, the steady-state ASL height (6.57 ± 0.5 μm) was not statistically different from that of WT mice (P = 0.69).
In contrast, cultures from β-ENaC transgenic mice exhibited a significantly different pattern of response to the 30-μl liquid challenge (Fig. 8). In β-ENaC transgenic mice, the initial rate of liquid absorption was significantly increased compared with control conditions, consistent with the increased Na+ absorption. More importantly, these cultures failed to slow absorption when ASL height approached that of outstretched cilia (Fig. 8); i.e., they effectively removed all the available liquid (3.78 ± 0.4 μm height) from airway surfaces before an apparent plateau resulted. This pattern of ASL volume hyperabsorption is similar to the hyperabsorption observed in human CF airway cultures (30).
To determine whether the overexpression of hCFTR could offset the excessive ASL volume absorption mediated by unregulated Na+ absorption in β-ENaC cultures (Fig. 8), we measured response to liquid addition in tracheal cultures from mice coexpressing β-ENaC and hCFTR. As shown in Fig. 8, whereas the initial absorption of the added volume was slower in cultures coexpressing β-ENaC and hCFTR (ASL height significantly different at 1 and 2 h), the expression of hCFTR did not modify the steady-state hyperabsorptive phenotype of the β-ENaC mice; i.e., steady-state ASL height was 3.45 ± 0.3 μm. Hence, overexpression of hCFTR failed to significantly alter the dysregulated Na+ absorption due to transgenic β-ENaC expression that produces ASL volume depletion.
A complex interplay between Na+ absorption and Cl− secretion is required to maintain ASL homeostasis. Relationships between these two ion transport processes were tested in two mouse models. 1) Mice overexpressing hCFTR in Clara cells were generated to gain insight into the role of CFTR in the regulation of endogenous murine ENaC. 2) Mice were generated to test whether expression of hCFTR modified disease pathology and survival of β-ENaC transgenic mice.
The first series of studies investigated effects of hCFTR overexpression on a WT background. Ussing chamber studies revealed that the basal Isc was elevated ∼2.5-fold in CCSP-hCFTR transgenic compared with WT tracheas. Several considerations suggest that the elevated basal Isc in CCSP-hCFTR transgenic mice reflected CFTR-mediated Cl− secretion. 1) The elevated basal Isc of CCSP-hCFTR transgenic mouse tracheas was dominated by an anion secretory current, as bilateral removal of Cl− (and HCO3−) reduced the basal Isc of the CCSP-hCFTR transgenic mouse tracheas to levels similar to WT tracheas (Fig. 3B). 2) The CFTR inhibitor Inh-172 exerted a significantly greater inhibition of basal Isc in cultures from CCSP-hCFTR transgenic than WT mice (Fig. 3D). 3) DIDS, which inhibits the CaCC in airway epithelia, did not reduce the basal Isc in CCSP-hCFTR transgenic or WT preparations (Fig. 3C). 4) When CCSP-hCFTR transgenic mice were crossed with CF mice, the tracheas of the hCFTR/CF pups exhibited basal Cl− secretion rates similar to the WT pups (Fig. 4).
hCFTR overexpression also had multiple effects on regulation of anion secretion. Adult CCSP-hCFTR transgenic mouse tracheas responded to forskolin, but the magnitude of the responses was reduced compared with that of WT tracheas (Fig. 3A). Although this effect may appear to be paradoxical because cAMP activates CFTR, we speculate that the expression of a constitutively active CFTR Cl− conductance in the apical membrane of CCSP-hCFTR transgenic mouse tracheas reduced the electrochemical driving force for Cl− secretion. Consequently, despite forskolin activation of both CaCC [via forskolin-dependent Ca2+ mobilization (11)] and hCFTR, the smaller electrochemical driving force for Cl− predicts a smaller Cl− secretory response in CCSP-hCFTR transgenic than WT mice. Similar reasoning pertains to the smaller UTP responses in CCSP-hCFTR transgenic than WT mice (Fig. 3A). In contrast, after DIDS pretreatment, the forskolin response was significantly greater in CCSP-hCFTR than WT tracheas (Fig. 3C). DIDS has been reported to block the forskolin-Ca2+-activated CaCC-mediated Cl− secretory responses (11). Thus the greater forskolin-induced Isc in DIDS-pretreated CCSP-hCFTR than WT tracheas likely reflects the selective contribution of the transgenically expressed hCFTR.
However, we cannot eliminate the possibility that hCFTR transgenically expressed in the murine trachea activated another Cl− channel, e.g., SLC26A9 (3). In human bronchial epithelial cells, SLC26A9 has been found to exhibit a constitutive Cl− conductance that was inhibited by CFTR blockers and required the presence of CFTR to exhibit a function (3). SLC26A9 RNA and expression of the protein have been identified in the murine trachea (34), suggesting that further studies are required to explore this possibility.
In contrast to its function as a cAMP-regulated Cl− channel, transgenically expressed hCFTR did not appear to inhibit endogenous ENaC function in excised tracheas from CCSP-hCFTR transgenic mice (Fig. 3A). Several possibilities may account for the absence of an hCFTR-ENaC regulatory relationship in mice. 1) It may be that most endogenous ENaC is expressed in ciliated cells, whereas the transgenic hCFTR in the CCSP-hCFTR mice is expressed selectively in Clara cells. The expression of endogenous ENaC subunits in WT mice was too low to test this possibility with currently available ENaC antibodies. 2) It is possible that hCFTR cannot regulate the activity of murine ENaC. However, it has been reported that when hCFTR and murine ENaC (α-, β-, and γ-subunits) were coexpressed in oocytes, hCFTR downregulated the amiloride-sensitive Isc (35). 3) It is possible that a third protein is required in tissues that normally exhibit a CFTR-ENaC regulatory relationship, and this putative protein is not expressed in lower airway epithelia of mice.
In a second series of studies, CCSP-hCFTR mice were crossed with β-ENaC mice to investigate whether transgenic overexpression of hCFTR decreased the mortality and lung pathology exhibited by the β-ENaC-overexpressing mouse (21). CCSP-hCFTR/β-ENaC mice exhibited a higher basal tracheal Isc than all other genotypes, suggesting that they simultaneously exhibit Na+ hyperabsorption and raised basal Cl− secretion (Fig. 7). Analyses of the function of each transgene in CCSP-hCFTR/β-ENaC mice supported this conclusion. The amiloride-sensitive (Na+-absorptive) component of Isc was elevated to similar levels in the β-ENaC and CCSP-hCFTR/β-ENaC mice (Fig. 7). In parallel, the post-amiloride Isc, a reflection of constitutive hCFTR-associated Cl− secretion (see above and results), was elevated similarly in the CCSP-hCFTR and the CCSP-hCFTR/β-ENaC double-transgenic mice. Importantly, despite the expression of a new and large basal anion secretory current in the Ussing chamber assays, the hCFTR/β-ENaC mouse did not exhibit an enhanced survival compared with the mouse expressing the β-ENaC transgene alone (Fig. 5), nor was the mucoobstructive lung pathology of the β-ENaC mice altered (Fig. 5).
To investigate why hCFTR failed to regulate transgenically expressed β-ENaC, the simplest explanation, i.e., hCFTR and β-ENaC were expressed in different subpopulations of Clara cells, was tested with immunohistochemical techniques (Fig. 6). β-ENaC appeared to stain more brightly in the upper airways, whereas hCFTR expression increased in distal airways. However, in both regions, immunocytochemical studies suggested that both channels were localized in the same Clara cells. It is also possible that the overexpression of hCFTR, which we estimated to be ∼4- to 5-fold higher than that of endogenous CFTR, was not sufficient to inhibit the high levels of β-ENaC, estimated to be 25- to 100-fold greater than endogenous β-ENaC (20). However, the substantial increase in Cl− secretion associated with hCFTR expression measured electrically argues against this scenario. Finally, it is possible that the Clara cell may not be the cell type that normally expresses CFTR in the murine airway and, thus, is not competent to mediate regulatory interactions between murine β-ENaC and CFTR.
An important question raised by this study is why hCFTR, despite producing a large anion secretory current in Ussing chamber experiments, did not “rescue” the β-ENaC mouse mucoobstructive phenotype in vivo. Insights into the relationships between ion transport rates, ASL volume homeostasis, and lung disease have emerged from combinations of Ussing chamber, Cl−-selective microelectrode studies, and confocal microscopy studies of ASL volume under thin-film conditions. One important concept from these studies is that there are signals in normal thin film of ASL that confer volume-dependent regulation to ENaC (14, 31). A related concept is that regulation of the ENaC activation state controls the rate of Na+ absorption and, indirectly, via the magnitude of the driving force for Cl−, the rate of Cl− secretion (22).
Confocal studies revealed that normal mouse airways absorbed a small bolus of liquid added to the surface relatively rapidly and then achieved a steady-state ASL at approximately the height of outstretched cilia (Fig. 8). The addition of the transgenic hCFTR slowed absorption in cultures from hCFTR WT mice but did not change the overall pattern. This observation emphasizes the relative importance of ENaC regulation, compared with the absolute magnitude of the apical Cl− conductance, for steady-state ASL volume homeostasis. As shown in Fig. 8, airway cultures from transgenic β-ENaC mice in this study, as reported in previous studies (21, 23), absorbed ASL more rapidly and failed to slow absorption until virtually all liquid was depleted (Fig. 8). Importantly, a similar pattern of unregulated volume depletion was also observed in the hCFTR/β-ENaC cultures. We speculate that the failure of hCFTR to induce sufficient Cl− and water secretion to hydrate hCFTR/β-ENaC culture surfaces reflects the absence of volume- and hCFTR-dependent inhibition of ENaC. As in CF, the failure to inhibit ENaC and, consequently, initiate Cl− secretion produced ASL volume depletion, mucus concentration/stasis, and, ultimately, mucoobstructive lung disease (30).
In summary, hCFTR transgenic mice exhibited a constitutive increase in Cl− secretion, but not regulation (inhibition) of ENaC, as measured in Ussing chambers. Furthermore, overexpression of hCFTR was not able to downregulate ENaC hyperfunction or initiate volume secretion in mice overexpressing murine β-ENaC. The failure of transgenically expressed hCFTR to ameliorate β-ENaC lung disease in mice, coupled with recent suggestions that there may not be basal regulatory effects of CFTR on ENaC activity in neonatal pig airways (27), emphasizes the need to understand the mechanisms that produce the time- and airway region-dependent regulation of ENaC critical for ASL homeostasis and lung health.
This work was supported by National Institutes of Health Specialized Center of Research Grant P50 HL-60280 and Grants HL-34322, HL-084934, K01 DK-080847 and P30 DK-065988 and Cystic Fibrosis Foundation Grant R026.
No conflicts of interest, financial or otherwise, are declared by the authors.
B.R.G., L.E.O., and R.C.B. are responsible for conception and design of the research; B.R.G., W.K.O., L.E.O., S.M.K., and B.B. performed the experiments; B.R.G., W.K.O., L.E.O., S.M.K., and B.B. analyzed the data; B.R.G., W.K.O., L.E.O., S.M.K., and B.B. interpreted the results of the experiments; B.R.G. and S.M.K. prepared the figures; B.R.G. drafted the manuscript; B.R.G. edited and revised the manuscript; B.R.G., W.K.O., L.E.O., S.M.K., B.B., and R.C.B. approved the final version of the manuscript.
We thank Marcus Mall for helpful suggestions regarding the manuscript, C. Sun, T. D. Rogers, B. Brighton, J. Hudson, and Weining Yin for excellent technical assistance, and Randy Thresher for many helpful suggestions regarding the study.
- Copyright © 2012 the American Physiological Society