ATP Depletion Induces a Loss of Respiratory Epithelium Functional Integrity and Down-regulates CFTR (Cystic Fibrosis Transmembrane Conductance Regulator) Expression*

To mimic the effect of ischemia on the integrity of airway epithelium and expression of cystic fibrosis transmembrane conductance regulator (CFTR), we induced an ATP depletion of the respiratory epithelium from upper airway cells (nasal tissue) and human bronchial epithelial 16HBE14o− cell line. Histological analysis showed that 2 h of ATP depletion led to a loss of the epithelium integrity at the interface between basal cells and columnar cells. The expression of connexin 43 (Cx43, subunit of the gap junctions) and desmoplakins 1 and 2 (DPs 1 and 2, major components of the desmosomes) proteins was inhibited. After 90 min of ATP depletion, a significant decrease of the transepithelial resistance (25%) was observed but was reversible. Similar results were obtained with the 16HBE14o− human bronchial epithelial cell line. ATP depletion led to actin filaments depolymerization. The expression of the mature CFTR (170 kDa) and fodrin proteins at the apical domain of the ciliated cells was down-regulated. The steady-state levels of CFTR, Cx43, DPs 1 and 2 mRNAs, semiquantified by RT-polymerase chain reaction kinetics, remained constant throughout ATP depletion in nasal tissue as in the homogeneous cell population of 16HBE14o− human bronchial epithelial cell line. This suggests that the down-regulation of these proteins might be posttranscriptional. The intercellular diffusion through gap junctions of Lucifer dye was completely inhibited after 90 min of ATP depletion but was reversible. The volume-dependent and the cAMP-dependent chloride secretion were inhibited in a nonreversible way. Taken together, these results suggest that an ATP depletion in human airway epithelium, mimicking ischemia, may induce a marked alteration in the junctional complexes and cytoskeleton structure concomitantly with a loss of apical CFTR expression and chloride secretion function.

In renal grafts, ischemia via cellular ATP depletion induces a series of structural, biochemical, and functional alterations, which lead to a loss of epithelial cell surface membrane polarity (1). Evidence accumulated so far indicates that the dissociation of the actin cytoskeleton and associated surface membrane structures leads to numerous cellular alterations including loss of cell-cell contact, cell extracellular matrix adhesion, and surface membrane polarity of renal proximal tubule cells (1)(2)(3)(4). During renal ischemia, the disruption of junctional complexes of proximal tubule induces a redistribution of the Na/K-ATPase from the basolateral membrane domain to the apical membrane domain (1,2).
The lung is the only solid organ that is transplanted without restoration of systemic arterial supply and where blood flow is reduced to levels insufficient to maintain cellular energy levels (5). Therefore, it is essential to determine the effects of ischemia in airway epithelial structure and function to adequately restore lung function. Lung transplantation is indicated for patients with end-stage respiratory failure (6), such as patients with cystic fibrosis (CF). 1 Cystic fibrosis, the most common and severe autosomal recessive disease among the northern American and European populations, is characterized by a defect in cyclic AMP-dependent chloride channel activity in a number of tissues, in particular the respiratory tract tissue (7). It is caused by mutations in the gene coding for the cystic fibrosis transmembrane conductance regulator (CFTR) (8 -10), which is a cAMP-regulated low-conductance channel (11). Lack or mislocalization of CFTR is regarded as being specific for CF (12,13). In normal airway surface respiratory epithelium, CFTR is restricted to the apical domain of the ciliated cells (13). Whether ischemia in lung transplants may induce a redistribution of CFTR protein from the apical membrane domain remains unknown.
The aim of this study was to investigate the effects of ATP depletion, simulating lung graft ischemia, on epithelium functional integrity and on CFTR expression. Our data demonstrate that ATP depletion induces an alteration of the respiratory epithelium at the interface between basal cells and columnar cells with a disruption of desmosomes and gap junction complexes. The expression of the mature CFTR protein is down-regulated, and the cAMP-mediated chloride secretion is inhibited. Our results suggest that the respiratory epithelial cells may lose polarity and CFTR chloride channel function during the ischemia of the lung transplants.

Human Airway Tissue
Fresh nasal polyps were obtained from non-CF patients undergoing nasal polypectomy due to nasal obstruction. Immediately after the removal, the tissue material was transferred to the laboratory in Hanks' medium, which contained 20 mM HEPES and antibiotics (100 units/ml penicillin and 100 g/ml streptomycin). Pieces of nasal polyp (approximately 1 cm 2 in size) were directly snap-frozen in liquid nitrogen.

Explant Outgrowth Cell Culture
Explants (1-2 mm 2 in size) of human respiratory nasal epithelial tissue from nasal polyps were seeded on 35-mm tissue culture dishes coated with type I collagen. Cells were incubated in RPMI 1640 culture medium supplemented with 2 mM L-glutamine, 1 g/ml insulin, 1 g/ml transferrin, 10 ng/ml epidermal growth factor, 0.5 mg/ml hydrocortisone, 10 ng/ml retinoic acid, and 100 units/ml penicillin/100 g/ml streptomycin, at 37°C. After 3 days in culture, explants were surrounded by a cell outgrowth resulting from both cell migration and cell proliferation (14).

Primary Cell Culture of Respiratory Cells
Dissociated human nasal respiratory epithelial cells were grown on coated type I collagen-carbodiimide (Sigma) Petri dishes in RPMI 1640supplemented culture medium at 37°C.
All cell cultures were performed at 37°C in an atmosphere of 5% CO 2 and 95% air before ATP depletion procedure.

ATP Depletion
Inhibition of glycolysis was accomplished by washing the cells in a glucose-free modified Ringer's buffer supplemented with 2 mM glutamine, followed by a 3-h incubation in this buffer at 37°C to deplete the tissue of endogenous metabolic substrates (3). The composition of the modified Ringer's buffer was 115 mM NaCl, 20 mM Hepes, 5 mM K 2 HPO 4 , 2 mM MgSO 4 , 1 mM CaCl 2 , and 2 mM glutamine. ATP depletion was achieved by adding in the modified Ringer's buffer the respiratory inhibitor antimycin A (10 M) and the glycolytic inhibitor 2deoxyglucose (10 mM) (3,4). Nucleoside contents were analyzed by high performance liquid chromatography (16) from primary culture of nasal respiratory epithelial cells and from 16HBE14o Ϫ human bronchial epithelial cells. The cells were rinsed, harvested, and placed in 0.5 ml of 0.6 N perchloric acid at 4°C. The cells were lysed using an Ultraturrax homogenizer by two pulses of 2 s. The lysates were neutralized with Tris-ethanol-amine carbonate-potassium hydroxide to a pH of 5.0 -6.0. The samples were placed at 4°C for 15 min in order to de-gas. The cell lysates were centrifuged (15 min, 4°C, 4500 rpm), and 20 l of the supernatant were injected on the column (C18 Bondapack, Waters, St. Quentin en Yvelines, France). The height of each nucleoside peak was measured and compared with the height of the standard peaks. Nucleoside contents were calculated and normalized according to the protein concentration of the samples. Samples designed as controls were obtained from primary culture of nasal respiratory epithelial cells and from 16HBE14o Ϫ human bronchial epithelial cells incubated for 5 h in the modified Ringer's buffer.

Ciliary Beating Frequency Analysis
To test the efficiency of the protocol of ATP depletion on the inhibition of cell activity, we analyzed the ciliary beating frequency. The culture dishes were placed on the heated stage of a phase-contrast inverted microscope (Nikon TMS-F) equipped with a CCD video camera (Panasonic WV CD50). The variation in light intensity, resulting from the ciliary beating, was detected by a photodetector placed on a video screen. The signal of the detector was converted into frequency spectrum by fast Fourier transform software. The mean ciliary beating frequency was calculated from this spectrum. Ciliary beating frequency was measured on 50 different ciliated cells/outgrowth (n ϭ 3) at 0, 30, 60, and 90 min of ATP depletion. Reversibility of ATP depletion was checked by incubating the cell culture for 12 h in a fresh RPMI 1640-supplemented medium. Viability of the cells was controlled by trypan blue exclusion staining.

Analysis of the Epithelium Integrity
Light Microscopy-For histological observations, nasal explants (n ϭ 2) were fixed in formalin (15%) for 60 min at room temperature, dehydrated, and embedded in paraffin. Five-m-thick sections were stained with hematoxylin/phloxin/safran. Two explants were fixed in 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS), pH 7.2, for 2 h and rinsed several times (3 ϫ 10 min) in 0.1 M PBS before being dehydrated through graded concentrations of ethanol and embedded in Epon. Semithin sections (1 m) were cut on a Reichert-Jung Ultramicrotome (Ultracut E, Leica, Rueil-Malmaison, France). Sections were stained with toluidine blue and observed under an Axiophot microscope (Zeiss, Le Pecq, France) at a magnification of ϫ40.
Transepithelial Resistance Measurements-The epithelial integrity of nasal outgrowth cell cultures (n ϭ 3) and 16HBE14o Ϫ cell line cultures (n ϭ 3) was quantified by transepithelial resistance measurements using a Millicell-ERS Resistance System (Millipore Co., Bedford, MA).
Nasal outgrowth cell cultures and 16HBE14o Ϫ cell line cultures were grown on type I collagen-coated 12-mm Transwell porous cell culture inserts, 0.4-m pore size transparent collagen membrane (Costar, Cambridge, MA). Transepithelial resistance was measured after a period of 90 min of ATP depletion and compared with that evaluated for cells incubated before ATP depletion (control) or after a 12-h incubation in cell culture control medium following the 90 min of ATP depletion. Data were expressed as means Ϯ standard error (S.E.) of triplicate measurements.

Immunohistochemistry of CFTR Protein and of Junctional Complexes
Nasal explants (n ϭ 7) were fixed in situ in cold methanol for 10 min at Ϫ20°C, dried, embedded in OCT (Tissue-Tek, Miles Inc.), cryofixed in liquid nitrogen, and stored at Ϫ80°C. Cryosections of 5 m thickness were placed onto gelatin-coated glass slides, further air-dried, and stored at Ϫ20°C until used for immunofluorescence microscopy, as described previously (17,18). 16HBE14o Ϫ human bronchial epithelial cells were grown on glass coverslips coated with 200 l of type I collagen associated with carbodiimide (Sigma). Cell cultures were fixed in situ in cold methanol for 10 min at Ϫ20°C and analyzed for immunofluorescence microscopy.The following mouse monoclonal primary antibodies were used: mouse anti-CFTR (MATG 1061; Transgene, Strasbourg, France, raised against a synthetic peptide corresponding to the amino acid sequence 503-515 without the residue 508, thus equivalent to the ⌬F 508 epitope in the NBF1 domain of the human CFTR protein, dilution 1:200; MATG 1104, Transgene, raised against a synthetic peptide corresponding to the amino acid sequence 722-734 of the R domain of the human CFTR protein, dilution 1:200; mAb 24 -1 (Genzyme Corporation, Cambridge, MA) raised against the amino sequence 1377-1480 of the COOH-terminal domain of CFTR, dilution 1:200). To test the specificity of the two CFTR antibodies (MATG 1061 and 1104) and to more accurately assess the location of the CFTR epitopes, we performed peptide competition assays. We used the synthetic peptides corresponding to the epitopes in the NBF1 domain (amino acid positions 503-515 without 508) and in the R domain (amino acid positions 722-734). The blockage was complete at a peptide/IgG ratio of 8 for MATG 1061 and a peptide/IgG ratio of 2 for MATG 1104. Fodrin was detected with mouse anti-fodrin (dilution 1:10, Sigma).

Scanning Laser Confocal Microscopy
Nasal outgrowth cell cultures (n ϭ 2) were grown on a glass coverslip coated with 200 l of type I collagen associated with carbodiimide (Sigma). Cell cultures were fixed in situ in cold methanol for 10 min at Ϫ20°C and used for immunofluorescence microscopy, as described previously. The labeling of the microtubule network was realized with the mouse anti-␤-tubulin (N357, dilution 1:50, Amersham). F-actin filaments were immunodetected as follows; outgrowth were fixed in 3.7% paraformaldehyde for 10 min, permeabilized in 0.5% Triton X-100 for 10 min, incubated in 1% bovine serum albumin-PBS for 5 min, labeled with phalloidin-FITC (dilution 1:50, Sigma) for 1 h, and rinsed twice in PBS. Cell outgrowths were examined with an MRC-600 Bio-Rad confocal system mounted on a Zeiss Axioplan microscope (Ivry/Seine, France). CFTR labeling in 16HBE14o Ϫ cell line was also examined by scanning laser confocal microscopy. Sequential serial sections were collected at 0.3 m pitch.

RNA Extraction
RNA extraction was performed from frozen human nasal polyps incubated in control medium (n ϭ 3) and from frozen human nasal polyps subjected to 2 h of ATP depletion (n ϭ 3). In parallel, RNA extraction was performed from 16HBE14o Ϫ human bronchial epithelial cells incubated in control medium and from 16HBE14o Ϫ human bronchial epithelial cells subjected to 2 h of ATP depletion. Before RNA extraction, glassware was sterilized overnight at 180°C, and all solutions and plasticware were treated for at least 12 h with 0.1% (v/v) aqueous diethylpyrocarbonate solution (Sigma) to inactivate RNases. Frozen nasal tissue was disrupted with an Ultraturrax homogenizer by two pulses of 2 s, lysed in 7-9-fold excess volume of 6 M guanidinum isothiocyanate, 5 mM sodium citrate, pH 7.0, 0.1 M 2-mercaptoethanol, and 0.5% laurylsarcosine. The total RNA was pelleted by ultracentrifugation (15 h, 20°C; 35,000 rpm, SW55Ti rotor; Beckman) through a CsCl cushion (18).
First strand cDNA synthesis was performed with the Superscript preamplification system (Life Technologies, Inc.); an RNA/primer mixture of 3 g of total RNA, 1.6 mM oligonucleotide primers, and diethylpyrocarbonate-treated water was incubated at 70°C for 10 min and placed in ice for at least 1 min. Seven l of a solution of 2.8 ϫ PCR buffer, 7.14 mM MgCl 2 , 1.4 mM dNTP, and 0.03 M dithiothreitol were added to the volume (12 l) of RNA/primers mixture. Two hundred units of Superscript II RT were added to the volume reaction and incubated for 50 min at 42°C. Reverse transcription was terminated at 70°C for 15 min. RNase H (2 units) was added to destroy the RNA strand during 20 min at 37°C. The first strand cDNA was subjected to PCR in a thermal cycler (Biometra, Trio-Thermoblock, Göttingen, Germany). The PCR assay (50 l of reaction volume covered with 50 l of mineral oil) contained 0.5 mM of each oligonucleotide primer, 0.15 mM of dNTP, 1.15 mM MgCl 2 , 1.02 ϫ PCR buffer, 3 l of the RT reaction mix. After RNA/cDNA denaturation at 94°C for 2 min, PCR (cycling parameters: 94°C/2 min 30 s; T°/2 min; 72°C/3 min) was run for 18 -34 cycles. Aliquots (7 l) were withdrawn in intervals of four cycles and subjected to 2% agarose gel electrophoresis. We titrated for the first reaction cycle when the cDNA became visible by ethidium bromide fluorescence during the late exponential phase of PCR and could be identified by a benchtop scanner (model GS-690 Imaging Densitometer, Bio-Rad). The amount of PCR product increased by 1 order of magnitude within four reaction cycles.

Intercellular Lucifer Yellow Diffusion
Lucifer Yellow (LY) CH (Sigma) was microinjected (pressure of injection: 100 hPa) through a microcapillary (diameter of opening of tip: 0.5 m) continuously for 2 min in the cytosol of 3 epithelial ciliated cells/outgrowth. Injections were performed on culture dishes placed in a temperature-controlled chamber (37°C) on a stage of an inverted microscope (Zeiss IM35). The diffusion of the probe in the neighboring cells was followed by using epifluorescence microscopy (Lucifer Yellow: 400 -440 nm excitation, emitted light Ͼ470 nm). Lucifer Yellow diffusion was video-recorded during the 2 min of microinjection with a low level video camera (Lhesa SIT 4036, St. Ouen l'Aumone, France). Intercellular LY diffusion was evaluated after a period of 90 min of ATP depletion and compared with that measured for cells incubated before ATP depletion (control) or after a 12-h incubation in a RPMI 1640 culture medium following the 90-min period of ATP depletion. Images were digitized as a 512 ϫ 512 pixel, 8-bit array using a Sparc-Classic workstation equipped with a video card (Parallax Graphics, Mountain View, CA). Variation of fluorescence intensity was analyzed by a multivariate statistical technique from the temporal image series.

Measurement of Cell Chloride Efflux by SPQ Analysis
3.5 mM of 6-methoxy-N-(3-sulfopropyl) quinolinium (SPQ) (Sigma), was microinjected through a microcapillary for 5 s in the cytosol of 10 ciliated cells/outgrowth (n ϭ 3). Confluent 16HBE14o Ϫ human bronchial epithelial cells were loaded with SPQ in a hypotonic chloride buffer solution (65 mM NaCl, 1.2 mM K 2 HPO 4 , 0.5 mM MgSO 4 , 5 mM Hepes) for 10 min at 37°C. SPQ-loaded cells were then incubated in a nitrate buffer (103 mM NaNO 3 , 2.4 mM K 2 HPO 4 , 1 mM MgSO 4 , 10 mM Hepes) in the presence or absence of 25 M forskolin (Sigma) at 37°C on the stage of the inverted microscope. Intracellular SPQ was excited at 365 nm through a ϫ32 planachromat objective, and emission light at Ͼ395 nm was recorded for 2 s every min for 15 min with a low level video camera (Lhesa). Cell chloride efflux was evaluated after a period of 90 min of ATP depletion and compared with that measured for cells incubated before ATP depletion (control) or after a 12-h incubation in a RPMI 1640 culture medium following the 90-min period of ATP depletion. Images were digitized and analyzed as described for intercellular Lucifer Yellow diffusion.

Statistical Test
Results were expressed as means Ϯ S.E. Student's t test was used to test the differences between the 90-min ATP-depleted cell cultures and the control cell cultures. A value of p Ͻ 0.05 was considered to be significant.

Role of ATP Depletion on ATP Content and on Ciliary
Beating Frequency-To control the ATP depletion efficiency and assess the maintenance of cell viability, nucleosides contents in one primary culture of nasal respiratory cells were measured by high performance liquid chromatography (Fig. 1). ATP depletion for 90 min induced a 93% decrease in ATP content (from 17.7 M to 1.3 M), a 98% decrease in ADP content (from 11 M to 0.2 M), and a 10-fold increase in AMP content (from 0.6 M to 6 M) (Fig. 1, B and C). After incubation of the cells for 12 h in fresh culture medium following the 90-min ATP depletion, a 6.6-fold increase in ATP content (from 1.3 M to 8.6 M), a 13.3-fold increase in ADP content (from 0.2 M to 2.9 M), and a 90% decrease in AMP content (from 6 M to 0.6 M) were observed (Fig. 1D).
Concomitantly, to control the ATP depletion efficiency and assess the maintenance of cell viability of the homogeneous 16HBE14o Ϫ human bronchial epithelial cell line, nucleoside contents were measured by high performance liquid chroma-FIG. 3. Transepithelial resistance measurements. A, before ATP depletion (control), the transepithelial resistance of nasal outgrowth cell culture was found to reach a mean value of 169 Ϯ 7 ⍀. ATP depletion for 90 min induced a significant 25% decrease of the transepithelial resistance (mean value: 127 Ϯ 7 ⍀, p ϭ 0.013). The transepithelial resistance was partly re-established by incubating the cells in fresh culture medium for 12 h (mean value: 143 Ϯ 9 ⍀). B, before ATP depletion (control), the transepithelial resistance of 16HBE14o Ϫ human bronchial epithelial cell line was found to reach a mean value of 238 Ϯ 4 ⍀. ATP depletion for 90 min induced a significant 48% decrease of the transepithelial resistance (mean value: 129 Ϯ 4 ⍀, p ϭ 0.001). The transepithelial resistance was partly re-established by incubating the cells in fresh culture medium for 12 h (mean value: 210 Ϯ 2 ⍀).
tography (data not shown). ATP depletion for 90 min induced a 97% decrease in ATP content (from 63 M to 2.1 M), a 96% decrease in ADP content (from 23 M to 1 M), and a 10-fold increase in AMP content (from 1.8 M to 18 M). After incubation of the cells for 12 h in fresh culture medium following the 90-min ATP depletion, a reversibility of ATP depletion was observed as assessed by a 23-fold increase in ATP content (from 2.1 M to 48.3 M), a 28-fold increase in ADP content (from 1 M to 28 M), and a 50% decrease in AMP content (from 18 M to 9 M).
Compared with control values (11.0 Ϯ 1.6 Hz), we observed a continuous and significant (p Ͻ 0.001) decrease in ciliary beating frequency in one primary culture of nasal respiratory cells after ATP depletion (7.2 Ϯ 1.8 Hz, 5.1 Ϯ 1.1 Hz, 4.4 Ϯ 1.5 Hz after 30, 60, and 90 min of ATP depletion, respectively). The ciliary beating frequency could be partially re-established after incubation of the cells for 12 h in fresh culture medium (10.8 Ϯ 1.3 Hz, p Ͻ 0.01) (data not shown). Therefore, the ciliated cells were still viable after 90 min of ATP depletion. No significant changes could be observed in ciliary beating frequency after a 48-h incubation of the outgrowth in the modified Ringer's medium without any metabolic inhibitors (data not shown).
Respiratory Epithelium Integrity-Compared with normal pseudostratified airway epithelium ( Fig. 2A), the respiratory epithelium integrity was disrupted after 2 h of ATP depletion at the interface between basal cells and columnar cells as shown in paraffin-embedded sections (Fig. 2B, arrows). Sections observed by transmission electron microscopy revealed the maintenance of tight junctions in the respiratory epithelium subjected to 2 h of ATP depletion (data not shown). A prolonged ATP depletion (12 h) induced a complete desquamation of the columnar cells but the basal cell monolayer was not altered.
Transepithelial Resistance Measurements-The change in the permeability of nasal outgrowth cell culture and of 16HBE14o Ϫ human bronchial epithelial cell line after ATP depletion was evaluated by triplicate transepithelial resistance measurements (Fig. 3, A and B).
CFTR and Junctional Complex Distributions-Before ATP depletion, CFTR protein was located at the apical domain of the ciliated cells in the pseudostratified respiratory epithelium of the explant (Fig. 4A). Fodrin (component of a membrane-cytoskeletal complex containing E-cadherin and ankyrin in Madin-Darby canine kidney cells) was detected at the apex of the ciliated cells (data not shown). Cx43 (component of the gap junctions) was detected as a punctate labeling around basal cells (Fig. 4C). DPs 1 and 2 (major components of the desmosomes) were preferentially expressed between basal and columnar cells (Fig. 4E), but a faint labeling could be observed along the basolateral membrane of the columnar cells.
After 2 h of ATP depletion, CFTR protein was detected in the cytoplasm of the ciliated cells and no more CFTR protein was detected at the apical domain of the ciliated cells (Fig. 4B). None of the CFTR cytoplasmic labelings detected in the epithelial cells after 2 h of ATP depletion could be reversed in the membrane by incubating the cells in fresh culture medium for 12 h (data not shown). Fodrin was observed as a cytoplasmic labeling in the ciliated cells (data not shown). Cx43 and DP expression was down-regulated in basal cells (Fig. 4, D and F). ␤-Catenin, E-cadherin, and ZO-1 distributions were not altered (data not shown), which suggested that ATP depletion targeted specific junctional complexes of the basal cell monolayer characterizing the pseudostratified respiratory epithelium. ␤ 1 integrin was still detected all along the basal cell monolayer (data not shown).
Before ATP depletion, CFTR protein was expressed in 16HBE14o Ϫ cell line (Fig. 5A). We observed by confocal microscopy that the CFTR labeling in 16HBE14o Ϫ cell line was distributed at the apical surface of the cell in the plasma membrane and in submembrane vesicles (Fig. 5C) and was negative at a nuclear plane level (data not shown). In contrast to Cx43 protein (data not shown), DPs 1 and 2 (Fig. 5E) and E-cadherin (data not shown) were detected.
After 2 h of ATP depletion, CFTR protein expression was down-regulated and the labeling was localized in the cytoplasm  (magnification, ϫ470). B, CFTR was located in the cytoplasm of the ciliated cells after 2 h of ATP depletion (wide arrows) (magnification, ϫ470). C, Cx43 was detected as a punctate labeling (arrows) around basal cells (magnification, ϫ740). D, Cx43 was not detectable after 2 h of ATP depletion (magnification, ϫ740). E, DPs 1 and 2 were preferentially distributed at the interface between basal and columnar cells (magnification, ϫ470). F, DPs 1 and 2 expression was inhibited after 2 h of ATP depletion (ϫ470). Controls did not exhibit significant labeling compared with the background (data not shown). of the epithelial cells (Fig. 5B), in a nonreversible way (data not shown). We observed by confocal microscopy that CFTR expression was inhibited at the apical surface of the cells (Fig. 5D). CFTR protein could be detected in a very faint labeling in the cytoplasm of the cells at a nuclear plane level (data not shown). DPs 1 and 2 were not expressed (Fig. 5F). In contrast, Ecadherin expression was not altered (data not shown).
Cytoskeleton Organization-Actin filaments and microtubule network were analyzed in the outgrowth of respiratory epithelial cells. Before ATP depletion, stress fiber filaments of polymerized F-actin were detected (Fig. 6A). In contrast, 90 min of ATP depletion led to nonreversible actin filaments depolymerization and no more stress fibers were observed (Fig.  6B). Negative controls performed by omitting phalloidin-FITC incubation step did not exhibit labeling. In contrast, ␤-tubulin labeling revealed that the microtubule network was not altered after 90 min of ATP depletion (data not shown). All negative controls performed by replacing the primary antibodies with nonimmune IgG did not show labeling (data not shown).
CFTR, Cx43, and DP 1 Transcript Expression-CFTR cDNA became visible by ethidium bromide stain at the 26th cycle during late exponential phase of PCR kinetics (Fig. 7, A and B). DP 1 and Cx43 cDNAs appeared at the 18th cycle of the PCR (Fig. 7, A and B). CFTR transcripts in nasal tissue were less abundant than Cx43 and DP 1 mRNAs (Fig. 7A). All kinetics were repeated three times. No significant changes were observed either assayby-assay or sample-by-sample within the control group or within the group of tissues subjected to 2 h of ATP depletion. Compared with the control group of nasal tissues, ATP depletion did not modify the steady-state level of either CFTR or Cx43 and DP 1 mRNA transcripts (Fig. 7A).
ATP depletion did not modify the steady-state level of CFTR and DP 1 mRNA transcripts in the 16HBE14o Ϫ human bronchial epithelial cell line (Fig. 7B). The 16HBE14o Ϫ human bronchial epithelial cell line did not express a level of Cx43 mRNA sufficient to be detected by RT-PCR.
Lucifer Yellow Diffusion in Respiratory Epithelial Cells-Intercellular communication in the outgrowth of respiratory epithelial cells was evaluated by Lucifer Yellow permeabilization analysis (Fig. 8A). Before ATP depletion (control), LY diffused into the neighboring cells (Fig. 8B, a). ATP depletion for 90 min completely inhibited intercellular communication through gap junctions as demonstrated, by the absence of LY diffusion (Fig. 8B, b). Intercellular communication was re-established by incubating the cells in fresh culture medium for 12 h (Fig. 8B, c).
cAMP Chloride Efflux-Before ATP depletion (control), forskolin (25 M) significantly stimulated the chloride efflux in the ciliated cells of the outgrowth (Fig. 9A) compared with the non-forskolin-stimulated cells (p ϭ 0.03). After 90 min of ATP depletion, the non-forskolin-stimulated and forskolin-stimulated cells exhibited a significant decrease in chloride efflux (p ϭ 0.004) (Fig. 9A). Therefore, the volume-dependent chloride efflux and the cAMP-mediated chloride secretion were inhibited. Chloride efflux inhibition, observed after a period of 90 min of ATP depletion, was not reversible by incubating the cells for 12 h in fresh culture medium, and no CFTR protein was detected at the apical domain of the ciliated cells (data not shown).
Similar results were obtained in the 16HBE14o Ϫ human bronchial epithelial cell line (Fig. 9B). Before ATP depletion (control), forskolin (25 M) significantly stimulated the chloride efflux compared with the non-forskolin-stimulated cells (p ϭ 0.027) (Fig. 9B). After 90 min of ATP depletion, the forskolinstimulated cells exhibited a significant decrease in chloride efflux (p ϭ 0.001) and the non-forskolin-stimulated cells showed a decreased chloride efflux (Fig. 9B). Chloride efflux  , ϫ740). C, we observed by confocal microscopy that CFTR protein was distributed at the apical surface of the cells in the plasma membrane and in submembrane vesicles (magnification, ϫ1500) (arrowheads, pitch: 0.3 m, Z resolution: 0.6 m, height: 2.4 m). E, DPs 1 and 2 were distributed at the interface between the epithelial cells (magnification, ϫ740). After 2 h of ATP depletion, CFTR protein expression was down-regulated and the labeling was localized in the cytoplasm of the cells (B, magnification, ϫ740), in a nonreversible way (data not shown). D, we observed by confocal microscopy that CFTR expression was inhibited at the apical surface of the cells (magnification, ϫ1500; pitch: 0.3 m, Z resolution: 0.6 m, height: 0.9 m). F, DPs 1 and 2 expression was inhibited after 2 h of ATP depletion (magnification, ϫ740). Controls did not exhibit significant labeling compared with the background (data not shown).
FIG. 6. Actin cytoskeleton organization. Actin filament network was analyzed in the outgrowth of respiratory epithelial cells. A, before ATP depletion, stress fibers of polymerized F-actin were detected (magnification, ϫ2220; arrowheads). B, 90 min of ATP depletion led to a nonreversible actin filament depolymerization and no more stress fibers were observed (magnification, ϫ2220). Controls performed by omitting phalloidin-FITC incubation step did not exhibit labeling. inhibition, observed after a period of 90 min of ATP depletion, was partly reversed by incubating the cells for 12 h in fresh culture medium, but CFTR protein was detected in the cytoplasm of the epithelial cells (data not shown). DISCUSSION Our report describes for the first time how ATP depletion, simulating ischemia of lung graft, leads to a loss of human respiratory epithelium integrity by an inhibition of the expression of desmosomal and gap junctional complexes situated at the interface between basal cells and columnar cells. Concomitantly, the depolymerization of the actin cytoskeleton network may explain the loss of epithelial cell surface polarity and an inhibition of apical CFTR expression and chloride secretion function.
It has been shown previously in the simple epithelium of renal proximal tubule that ATP depletion is able to disrupt all of the adhesive mechanisms (1,2,20). Our results emphasize the specificity of the pseudostratified organization of the respiratory epithelium and the critical role played by the junctional complexes situated at the interface between basal cells and columnar cells in the maintenance of epithelium integrity. In non-ATP depleted control specimens, desmosomes, situated at the interface between basal and ciliated cells (6,17,18), anchor intermediate filaments such as actin and cytokeratins (21). As described in the rat tracheal epithelium (22), gap junctions and Cx43 were detected in close relationship to well formed desmosomes in the human nasal respiratory epithelium. After ATP depletion, DPs 1 and 2 and Cx43 expressions in basal cells might be posttranscriptionally down-regulated leading to a disrupted integrity of the nasal respiratory epithelium. After an overnight incubation in an ATP-depleted medium, the only remaining basal cells monolayer of the pseudostratified respiratory epithelium demonstrates the weakness of adhesive mechanisms situated at the interface between the columnar and basal cells. The decreased expression of DPs 1 and 2 after ATP depletion was confirmed in 16HBE14o Ϫ cell line. In contrast, the expression of E-cadherin protein was not altered. ZO-1 and cadherin/catenin complexes mediate cell adhesion of the columnar respiratory epithelial cells. These complexes were still detected after ATP depletion and participated in the remaining focal upper adhesion of the ciliated and secretory cells.
Since the epithelium permeability and the transepithelial resistance depend on the expression of junctional complexes (23,24), we analyzed the transepithelial resistance of the respiratory epithelium after ATP depletion. We observed that the down-regulation of the junctional complexes situated at the interface between basal cells and columnar cells after ATP depletion was associated with a decrease of the transepithelial resistance in both nasal outgrowth cells and in 16HBE14o Ϫ cell line.
Since the partial loss of respiratory epithelium integrity following ATP depletion was related to the breakdown of desmosomes/gap junction complexes, we further analyzed the functional effects of ATP depletion at the interface between basal and columnar cells particularly on intercellular communication mediated by junctional complexes. The gap junctions are protein channels situated in the plasma membrane of numerous epithelial cells (25), which provide intercellular communication of second messengers (26) and are responsible for the homeostatic control of cell growth or differentiation (25). A switch in Cx protein family expression is associated with selective changes in junctional permeability (27). Communication defect results from abnormally low levels of translation and phosphorylation of the gap junctional proteins (28). Our data report that ATP depletion inhibited, in a reversible way, gap junction channel permeability and LY diffusion between respiratory epithelial cells. We hypothesize that ATP depletion in the respiratory epithelium might produce non-functional gap junctions with unphosphorylated Cx43 and abnormal low levels of connexin translation.
We observed by confocal microscopy that ATP depletion in- FIG. 9. cAMP-mediated chloride efflux. A, SPQ was microinjected for 5 s in the ciliated cells of the outgrowth. Fluorescence variation within the microinjected cells was evaluated by recording images for 2 s every min for 15 min. Forskolin (25 M) significantly stimulated cAMP-mediated chloride secretion (*, p ϭ 0.04). ATP depletion (90 min) inhibited not only volume-dependent chloride secretion (**, p ϭ 0.004), but also cAMP-mediated chloride secretion (***, p ϭ 0.004). CFTR chloride channel function was not reversible (data not shown). B, similar results were obtained in the 16HBE14o Ϫ human bronchial epithelial cell line. Before ATP depletion (control), forskolin (25 M) significantly stimulated the chloride efflux compared with the non-forskolin-stimulated cells (*, p ϭ 0.027). After 90 min of ATP depletion, the forskolin-stimulated cells did not exhibit significant chloride efflux (***, p ϭ 0.001) and the non-forskolin-stimulated cells presented a decreased chloride efflux. Chloride efflux inhibition, observed after a period of 90 min of ATP depletion, was partly reversible by incubating the cells for 12 h in fresh culture medium (data not shown). duced a depolymerization of stress fibers of F-actin similar to that observed in proximal tubule cells (3,29). Moreover, fodrin, which is described to form complexes with actin filaments (30), was detected as a faint labeling in the cytoplasm of the ciliated cells. These results suggest that after ATP depletion, the respiratory epithelial cell polarity was damaged by a disruption of the cytoskeleton network as shown earlier by Molitoris et al. (1,2) in the kidney proximal tubular cells. Cell polarity is required for cAMP-mediated chloride secretion (31). CFTR is considered an actin-binding protein, and actin filaments participate in the activation of CFTR chloride channel in bronchial epithelial cells (32).
Therefore, we analyzed CFTR expression and CFTR chloride channel activity because we speculated that actin filaments depolymerization might down-regulate the apical trafficking, expression, and function of the CFTR chloride channel as demonstrated in proximal tubule cells for the Na/K ATPase pump (1,2).
ATP depletion induced a redistribution of CFTR, which shifted from an apical to a cytoplasmic localization in the ciliated cells of the respiratory epithelium as in the homogeneous epithelial cell population of 16HBE14o Ϫ cell line. Since CFTR mRNA concentration, semiquantified by RT-PCR kinetics, was retained, we suggest that ATP depletion would have a post-trancriptional effect on CFTR expression.
In primary cell culture of nasal cells, representing a heterogeneous cell population that is perfectly representative of the in vivo airway cell population, as in the homogeneous epithelial cell population of 16HBE14o Ϫ cell line, our results demonstrate that the CFTR chloride channel function is inhibited after ATP depletion as reported earlier by Wersto et al. (33). CFTR chloride channel is regulated by phosphorylation of the R domain (34), which is mediated by protein kinase C (35)(36)(37). CFTR chloride conductance is regulated by hydrolytic ATP and nonhydrolytic ATP allosteric binding (38). ATP depletion might therefore down-regulate the CFTR chloride channel function. We hypothesize that ATP depletion would down-regulate the chloride efflux not only by a nonreversible internalization of CFTR but also by a lack of availability of ATP to run the sodium pump for chloride loading of the cell. This would also contribute to the lack of response to forskolin in the absence of ATP.
In summary, this study provides evidence that ATP depletion leads to a loss of respiratory epithelium integrity at the interface between basal and columnar cells by disrupting desmosomes and gap junction complexes. Abnormal CFTR protein expression and chloride secretion function were observed after ATP depletion in respiratory epithelial cells, which exhibited a loss of polarity associated with depolymerized actin cytoskeleton network. These results suggest that ATP depletion, simulating ischemia, may induce a marked alteration in the junctional complexes and cytoskeletal structure associated with a loss of apical CFTR expression and chloride secretion function in airway epithelium of lung transplants.