点击显示 收起
【摘要】 We examined the effects of intracellular Cl - concentration ([Cl - ] i ) on the epithelial Na channel (ENaC) in a line of Madin-Darby canine kidney (MDCK) cells (FL-MDCK) with a high rate of Na + transport produced by stable retroviral transfection with rENaC subunits (Morris RG and Schafer JA. J Gen Physiol 120: 71-85, 2002). Treatment with cAMP (100 µM 8-cpt-cAMP plus 100 µM IBMX) stimulated ENaC-mediated Na + absorption as well as Cl - secretion via cystic fibrosis transmembrane conductance regulator, which was characterized in -toxin-permeabilized monolayers to have the anion selectivity sequence NO 3 - Br - Cl - I -. With the use of FL-MDCK monolayers in which the basolateral membrane was permeabilized by nystatin, the ENaC conductance of the apical membrane [determined from the amiloride-sensitive short-circuit current (AS- I sc ) driven by an apical-to-basolateral Na + concentration gradient] was progressively inhibited by increasing the [Cl - ] in the basolateral solution (and hence in the cytosol), but it was insensitive to the [Cl - ] in the apical solution. This inhibitory effect of [Cl - ] i occurred regardless of the presence or absence of net Cl - transport. However, from fluorometric measurements using the Cl - -sensitive dye 6-methoxy- N -(3-sulfopropyl)-quinolinium in intact FL-MDCK monolayers on permeable supports, cAMP, which activates both Na + absorption and Cl - secretion, produced a decrease of [Cl - ] i from 76 ± 14 to 36 ± 8 mM ( P = 0.03). Thus it might be expected that activation of Cl - secretion by cAMP would lead to stimulation rather than inhibition of ENaC. In the nystatin-treated monolayers, an increase in [Cl - ] i from 15 to 145 mM decreased AS- I sc from 24.5 ± 1.0 to 10.2 ± 1.6 µA/cm 2. This inhibition of ENaC could be attributed to nearly proportional decreases in the density of ENaC in the apical membrane from 1.91 ± 0.16 to 1.32 ± 0.17 fmol/cm 2 and in the intrinsic channel activity (the average current per ENaC subunit) from 13.3 ± 1.2 to 8.2 ± 1.4 µA/fmol.
【关键词】 epithelial sodium channel surface density chloride channels cystic fibrosis transmembrane conductance regulator
THE AMILORIDE - SENSITIVE EPITHELIAL Na + channel (ENaC; comprising 3 subunits -, -, and -ENaC) in the apical membrane of epithelial cells lining the distal nephron, distal colon, airways, and exocrine gland ducts mediates Na + absorption ( 20 ). The activity of this channel is tightly regulated by several hormones, including aldosterone, AVP, and insulin as well as protein kinases ( 20 ). In addition, it is also regulated by inorganic ions such as Ca 2+, Na +, and Cl - ( 14, 26 ). The regulation of Na + transport via ENaC is essential for maintaining sodium balance, blood pressure, and extracellular fluid volume.
Recently, the inhibition by cystic fibrosis transmembrane conductance regulator (CFTR) of transepithelial Na + absorption via ENaC has been studied intensely because it is hypothesized to explain the pathophysiology of cystic fibrosis in airway epithelia ( 11, 38 ). The inhibitory effects of CFTR on Na + transport are also observed in other organs affected by this disease and in a variety of epithelia that express both transporters ( 21, 27, 33, 45 ). Several potential mechanisms by which CFTR inhibits ENaC have been proposed. For example, CFTR may interfere with protein kinase A-dependent regulation of ENaC ( 55 ), or CFTR may control ENaC through additional regulatory proteins such as PDZ-binding domain proteins ( 5 ) or by direct physical binding ( 28 ). Alternatively, the inhibitory effect of CFTR in some preparations may be due to the rise it produces in intracellular Cl - concentration ([Cl - ] i ) rather than a direct molecular interaction of CFTR with ENaC ( 32 ). Kunzelmann and colleagues ( 6, 30, 32 ) showed that the inhibition of ENaC in Xenopus laevis oocytes by CFTR coexpression required a high extracellular [Cl - ]. Furthermore, this inhibition was not specific to CFTR but could also be produced by coexpression of ClC-0 or ClC-2, and it occurred even when amphotericin was used to permeabilize the cell membrane nonspecifically ( 30 ), leading to the conclusion that ENaC can be inhibited by any agent that increases [Cl - ] i ( 30, 32 ). The importance of [Cl - ] i has also been demonstrated in whole cell patch-clamp studies of M-1 cells ( 34 ), in excised patches of salivary duct in which a variety of anions including NO 3 - on the cytosolic side inhibited ENaC ( 14 ), and in sweat gland ducts in which the basolateral membrane had been permeabilized by -toxin ( 49 ).
As pointed out by Kunzelmann ( 32 ), the effect of activating CFTR or other Cl - channels in a particular epithelium will depend on the direction of the resulting change in [Cl - ] i and on whether that epithelium is primarily involved in NaCl absorption or secretion. In epithelial cells, [Cl - ] i is determined by the steady-state relationship between Cl - transporters in apical and basolateral membranes ( 25 ). In most epithelia that actively secrete Cl -, Cl - is actively accumulated across the basolateral membrane by cotransporters such as NKCC and NCC that maintain [Cl - ] i above its electrochemical equilibrium value. Stimulation of Cl - secretion in such epithelia (e.g., by cAMP, forskolin, or cholera toxin) could be produced by increased apical membrane Cl - channel activity, which would result in a fall in [Cl - ] i, or by an increase in the activity of the basolateral cotransporter in the presence of a finite Cl - conductance in the apical membrane, in which case [Cl - ] i would rise. For example, stimulation of Cl - secretion by cAMP results in a fall in [Cl - ] i in Necturus gallbladder ( 47 ), spiny dogfish rectal gland tubules ( 22 ), canine trachea ( 53 ), and human colonic epithelium ( 7, 39 ), but [Cl - ] i increases with stimulation in human colonic epithelial cells ( 37 ).
We previously developed a line of Madin-Darby canine kidney (MDCK) cells (FL-MDCK) expressing "flagged" rat ENaC subunits, i.e., subunits labeled in the extracellular loop-domain with FLAG epitope as described by Firsov et al. ( 17 ). The FLAG epitope allows the number of ENaC subunits residing in the apical membrane of the transfected MDCK cells to be quantified by the surface binding of 125 I-labeled anti-FLAG antibodies. Our initial studies with intact FL-MDCK monolayers in DMEM medium showed they had a high short-circuit current ( I sc ) and that cAMP treatment produced a rapid transient peak within 5 min followed by a broad peak that decayed over 20 min ( 43 ), as also described in A6 and M-1 cultures ( 8, 29, 34, 44 ). The biphasic response to cAMP in all of these epithelia has been attributed to rapid Cl - secretion by CFTR activation followed by a slower cAMP-dependent activation of ENaC that is blunted by the inhibitory effect of CFTR. However, in a Cl - -free medium, cAMP produced a monotonic and sustained increase in I sc in the FL-MDCK monolayers ( 43 ). For that reason, Morris and Schafer ( 43 ) used a Cl - -free medium to examine the effect of cAMP on the surface density of ENaC subunits and established that it increased in proportion to the increase in amiloride-sensitive (AS)- I sc, demonstrating that cAMP increased Na + transport by an increase in the number rather than a change in the intrinsic activity of the ENaC in the apical membrane ( 43 ).
The present studies were designed to examine whether increased [Cl - ] i inhibits ENaC-mediated Na + absorption in epithelial monolayers of FL-MDCK cells as it does in oocytes and patch-clamped membranes from other epithelia. We also examined whether such a mechanism could account for the later decay in Na + absorption after cAMP stimulation observed in previous studies with this and similar epithelia ( 8, 29, 34, 43, 44 ) and the effect of increases in [Cl - ] i on the density of ENaC subunits in the apical membrane of these FL-MDCK cells.
METHODS
Cell cultures. The cells used in these experiments, referred to as FL-MDCK cells, were a subclone of the MDCK cells ( passages 5 to 35 ) that had been retrovirally transfected to express rat -, -, and -ENaC subunits, each of which contained the FLAG epitope, as described by Morris and Schafer ( 43 ). The cultures were maintained in T-75 flasks at 37°C in DMEM (Life Technologies) supplemented with 10% FBS, 50 mM HEPES (pH 7.4), 1% Pen/Strep-fungizone, and the selection antibiotics: G418 (800 µg/ml), hygromycin (300 µg/ml), and puromycin (5 µg/ml). The cells were split on confluence and seeded on 24-mm Transwell inserts (Costar; catalog no. 3412) at a density of 10 5 cells/cm 2. After being seeded, FL-MDCK cells were grown in DMEM containing FBS and HEPES but without selection antibiotics for 5-7 days. The medium was changed daily. As described by Morris and Schafer ( 43 ), before use in the experiments, the cell monolayers were induced with 1 µM dexamethasone plus 2 mM Na + butyrate in the culture medium overnight.
Electrophysiological studies. The membranes supporting the high-resistance FL-MDCK monolayers were carefully cut from the plastic insets and mounted in Ussing-type chambers in a 37°C incubator. In experiments with intact monolayers, both sides were bathed with 10 ml of Krebs-Ringer bicarbonate (KRB) solution containing (in mM) 113 NaCl, 1.2 Na 2 HPO 4, 25 NaHCO 3, 1.1 CaCl 2, 1.2 MgCl 2, 4.5 KCl, and 10 glucose. All solutions were gassed with 95% O 2 -5% CO 2 at 37°C, and the pH was 7.4.
Transepithelial I sc (µA/cm 2 ) and conductances ( G te; mS/cm 2 ) were measured as described previously ( 43 ). The AS- I sc was defined as the change in I sc produced by the addition of 10 µM amiloride to the apical solution, and the remaining I sc was defined as the non-amiloride-sensitive short-circuit current (NS- I sc ). In those experiments involving cAMP treatment, 100 µM 8-(4-chlorophenylthio)-cAMP (Sigma) plus 100 µM IBMX were added to both the apical and basolateral solutions. In some experiments, the monolayers were kept under open-circuit conditions during the course of the experiment except for intermittent voltage clamping to 0 mV and ±2 mV for a total of 4 s every 30 s to measure I sc and G te.
Permeabilization of the basolateral membrane with -toxin or nystatin. To short-circuit the apical membrane directly and to manipulate the ionic composition of the cytoplasm, we permeabilized the basolateral membrane with the pore-forming agents -toxin or nystatin. FL-MDCK monolayers were incubated with 200 U/ml of -toxin in the basolateral solution at 37°C for 30 min. The agent was then removed before the FL-MDCK monolayers were studied in Ussing chambers. As demonstrated in other epithelia ( 46 ), short-term -toxin treatment of the basolateral membrane did not permeabilize the apical membrane or the junctional complexes because cAMP was able to activate transepithelial transport when added to the basolateral but not to the apical side. In other experiments, nystatin (Ca 2+ 50 mg/ml stock solution in DMSO; Calbiochem, La Jolla, CA) was added to the basolateral solution at a final concentration of 350 µg/ml after the monolayer was mounted in the Ussing chamber (see METHODS in Ref. 36 ). On nystatin addition, the open-circuit voltage and I sc fell to zero in monolayers bathed in symmetrical apical and basolateral solutions, indicating that active transport was abolished by the permeabilization of the basolateral membrane. The transepithelial conductance remained low and stable even 60 min after nystatin addition to the basolateral solution.
After permeabilization of the basolateral membrane with either agent, I sc was allowed to stabilize for 15-20 min before the beginning of each experiment. In the experiments with -toxin, the Ringer-like apical solution contained (in mM) 120 Na + gluconate, 1.2 CaCl 2, 1.2 MgCl 2, 2.4 K 2 HPO 4, 0.6 KH 2 PO 4, 10 glucose, and 10 HEPES, whereas the basolateral solution contained 120 KCl, 1.2 CaCl 2, 1.2 MgCl 2, 2.4 K 2 HPO 4, 0.6 KH 2 PO 4, 10 glucose, and 10 HEPES. Thus there were large transepithelial chemical gradients favoring Na + absorption and Cl - secretion. In the experiments with nystatin permeabilization, the monolayers were bathed in variants of Ringer solution that produced a Na + concentration gradient in the absorptive direction. The apical solution contained (in mM) 120 NaCl, 1.2 CaCl 2, 1.2 MgCl 2, 2.4 K 2 HPO 4, 0.6 KH 2 PO 4, 10 glucose, and 10 HEPES, whereas the basolateral solution contained (in mM) 120 KCl, 1.2 CaCl 2, 1.2 MgCl 2, 2.4 K 2 HPO 4, 0.6 KH 2 PO 4, 10 glucose, and 10 HEPES. The chloride concentration in these two solutions was varied by replacement with gluconate.
Surface labeling of flagged ENaC subunits. ENaC subunits surface density was measured as described by Morris and Schafer ( 43 ). Briefly, anti-FLAG antibody (Sigma; referred to below as M2 Ab) was radiolabeled with 125 I in the UAB Radiolabeling Core Facility in the Comprehensive Cancer Center at the University of Alabama at Birmingham. Labeled M2 Ab was dialyzed for at least 24 h in PBS, and the Ab concentration was determined by a microprotein assay. Specific binding assays were performed on FL-MDCK cells grown on Transwell inserts. After incubation with low- or high-[Cl - ] bathing solutions containing 210 µg/ml nystatin in the basolateral solution for 20 min, the inserts were placed in an ice bath and briefly rinsed with ice-cold PBS, and then incubated for 30 min with ice-cold blocking solution (PBS + 5% FBS). Binding was started on addition of 4 nM M2 Ab in a volume of 500 µl blocking solution per insert. To determine the nonspecific binding, M2 Ab was added to paired inserts together with a 100-fold excess (by weight) of FLAG peptide (Sigma). After 1-h incubation on ice, the inserts were washed four times with 1.5 ml of blocking solution. The M2 antibody remaining bound to the apical surface of the monolayers was then removed with 750 µl of ice-cold acid washing solution (0.5 M NaCl, 0.2 M Na + acetate, pH 2.4). Two acid strips were performed, and the counts were combined for data analysis.
Immunoprecipitation and Western blotting. For immunoprecipitation of CFTR, FL-MDCK monolayers were washed twice with ice-cold PBS, scraped from the Transwell inserts, and lysed in PBS containing 1% Triton X-100 and protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, 1 µg/ml pepstatin, and 1 µg/ml aprotinin) for 10 min at 4°C. The lysate was centrifuged for 10 min at 14,000 g, and the supernatant containing the solubilized proteins was incubated overnight at 4°C with a polyclonal rabbit antibody against the highly conserved nucleotide binding domain (NBD-1) of human CFTR (gift of Dr. J. Collawn, Dept. of Cell Biology, University of Alabama at Birmingham). This IgG antibody was cross-linked to protein A/G-agarose beads with 10 mM dimethylpimelimidate (Pierce Chemical) for 30 min at room temperature, and 500 ng were used per lysate sample. After immunoprecipitation, samples were incubated in Laemmli buffer for 15 min at 37°C and run on 4-12% polyacrylamide gels. The separated proteins were transferred to a nitrocellulose membrane (Nitrobind, Osmonics) and processed for immunoblotting. The membrane blots were blocked with Blotto for 1 h at room temperature and incubated overnight at 4°C with 2 µg/ml of a monoclonal mouse anti-human CFTR NBD-1 antibody (Chemicon, Temecula, CA). They were then incubated with secondary antibody horseradish peroxidase conjugate at 1:4,000 dilution followed by ECL detection (Amersham Pharmacia Biotech) according to the manufacturer's instructions.
For immunodetection of ENaC subunits, FL-MDCK cell lysates (40 µg protein) were separated by electrophoresis on a 4-12% polyacrylamide gel, transferred to nitrocellulose, and probed with antibodies specific to the three rat ENaC subunits (a kind gift from Dr. M. Knepper, National Institutes of Health) at 1:2,000 dilutions.
Intracellular Cl - measurements. Measurements of [Cl - ] i in monolayers of FL-MDCK cells were made using 6-methoxy- N -(3-sulfopropyl)-quinolinium (SPQ; Molecular Probes) as a fluorescent Cl - probe. FL-MDCK monolayers grown on Transwell inserts were incubated overnight at 37°C in a humidified CO 2 incubator with DMEM medium containing 5 mM SPQ. After being loaded, the monolayers were rapidly washed several times and the membranes were cut from the inserts and mounted on custom-made rectangular supports, which were then placed in the cuvette of a Delta-Scan fluorometer (Photon Technologies, Princeton, NJ) at a 45° angle to the excitation beam. The supports had cutouts so that the basolateral as well as the apical surfaces of the monolayers were bathed by the KRB that was perfused continuously through the cuvette at 2 ml/min. Experiments were performed at 37°C, and SPQ fluorescence was measured every second at an emission wavelength of 440 nm in response to the excitation wavelength of 344 nm using Photon Technologies software.
The [Cl - ] i was determined from the fluorescence values using the methods described by Verkman and his collaborators ( 9, 31 ). At the end of every experiment, the fluorescence intensity in the absence of Cl - (F 0 ) was determined by perfusion with a Cl - -free solution containing nigericin (5 µM) and tributylin chloride (10 µM). To determine the fluorescence background, potassium thiocyanate (150 mM) and valinomycin (5 µM) were added to quench intracellular SPQ fluorescence. The fluorescence that was not quenched was subtracted from the measured fluorescence throughout each experiment.
Calibration solutions of varying Cl - concentrations were prepared as mixtures of a solution containing 150 mM KCl plus 10 mM D -glucose and a Cl - -free solution containing 150 mM KNO 3 plus 10 mM D -glucose. These solutions also contained nigericin and tributylin to ensure rapid equilibration of the cytoplasmic Cl - concentration with that of the extracellular solution. The value of (F 0 /F) - 1, where F is the fluorescence at a given Cl - concentration and F 0 is the fluorescence in the Cl - -free solution, was plotted as a function of the Cl - concentration, and the Stern-Volmer constant K sv (M -1 ) was calculated as the slope of the linear regression fit of the data (see Ref. 9 ).
Data analysis. StatView for Macintosh (SAS Institute) was used for standard statistical analyses: ANOVA with Bonferroni/Dunn post hoc testing for multiple comparisons, and paired or nonpaired t -tests as appropriate for single comparisons; n represents the number of experiments. Kaleidagraph for Macintosh (Synergy Software) was used for linear fitting and regression analysis, and the correlation coefficient ( r ) and associated P value of the slope were calculated. In all cases, significance was assumed for P < 0.05.
RESULTS
Expression of ENaC subunits and CFTR in MDCK cells. Our initial experiment was to confirm that the FL-MDCK cells, a subclone of the F F F MDCK cells developed by Morris and Schafer ( 43 ), expressed all three rat ENaC subunits. As shown in Fig. 1, left, the subunit-specific antibodies showed the presence of the -, -, and -ENaC subunit proteins in the FL-MDCK cells, and the levels of all three subunits were increased by overnight induction with Na butyrate and dexamethasone. CFTR has previously been reported to be present in MDCK cells ( 42 ), and our RT-PCR analysis also revealed the expression of CFTR mRNA in FL-MDCK cells (data not shown). To confirm that CFTR protein was present in our MDCK cell line, the whole cell lysates were immunoprecipitated with a polyclonal antibody to CFTR and probed with a monoclonal anti-CFTR (NBD-1 region) antibody. As shown in Fig. 1, right, Western blots of the immunoprecipitates from FL-MDCK cells showed a band of the same size observed for CFTR in HT29 cells, a human "positive control" cell line that exhibits a relatively high expression of CFTR at the protein level.
Fig. 1. Left : Western blots demonstrating the expression of epithelial Na channel (ENaC) subunits in FL-Madin-Darby canine kidney (MDCK) cells. The lanes are grouped according to the subunit-specific antibody with -band at 94 kDa, -band at 98 kDa, and -band at 89 kDa. Cells were incubated overnight before protein extraction either without (-) or with 2 mM Na butyrate + 1 µM dexamethasone (+). Right : Western blots demonstrating the expression of cystic fibrosis transmembrane conductance regulator (CFTR) in FL-MDCK cells. The lanes contained protein immunoprecipitated from HT-29 cells and FL-MDCK cells. The latter were induced overnight as described left. The numbers on the right indicate the molecular mass of proteins in the molecular mass calibration ladder.
Figure 2 is a representative experiment showing the effect of cAMP treatment on transepithelial transport in FL-MDCK cells. In KRB solution, the basal I sc (measured immediately before cAMP treatment) averaged 16.8 ± 1.4 µA/cm 2 ( n = 7). The addition of 100 µM 8-CPT-cAMP plus 100 µM IBMX to the apical and basolateral solutions produced a transient I sc response that peaked and declined within 5 min. This peak was followed by a sustained increase in I sc to 25.8 ± 1.6 µA/cm 2. Addition of 10 µM amiloride to the apical solution rapidly decreased I sc to 9.5 ± 1.4 µA/cm 2, which was further reduced to 7.5 ± 0.9 µA/cm 2 by 200 µM apical glibenclamide. As reported in our previous studies with MDCK and A6 cells ( 43 ), at least two components of I sc are stimulated by cAMP: active Na + absorption via ENaC and active Cl - secretion, a portion of which was tentatively attributed to CFTR.
Fig. 2. Time course of short-circuit current ( I sc ) response to cAMP treatment in Krebs-Ringer bicarbonate solution. A mixture of 100 µM CPT-cAMP + 100 µM IBMX was added as indicated by the bottom bar, followed by 10 µM amiloride and 200 µM glibenclamide addition to the apical solution only ( middle and top bars, respectively). The result is representative of 7 experiments, the average values of which are given in the text.
Characterization of the apical anion channel with -toxin. To examine the characteristics of the apical anion transporter in the absence of the complicating effects of transport processes in the basolateral membrane, we permeabilized the latter membrane with 200 U/ml of -toxin. The permeabilized MDCK monolayers were mounted in Ussing chambers with Na + and gluconate as the major ions in the apical solution and K + and Cl - as the major ions in the basolateral solution. The transepithelial voltage was clamped to 0 and 10 µM amiloride was added to the apical side to block Na + transport. Under these conditions, as shown in Fig. 3, NS- I sc was very low (2.3 ± 0.5 µA/cm 2, n = 5), indicating that the basal ion permeability of the apical membrane was quite low and that the paracellular pathway maintained a high resistance despite the -toxin treatment. The addition of 100 µM cAMP increased NS- I sc to 11.7 ± 1.7 µA/cm 2 ( P < 0.01, n = 5), which was reduced to 3.7 ± 1.0 µA/cm 2 ( P < 0.01, n = 5) by 200 µM glibenclamide added to the apical solution. Because Na + transport was blocked by amiloride, this positive I sc is consistent with Cl - secretion via an anion-selective channel in the apical membrane that is activated by cAMP and inhibited by glibenclamide.
Fig. 3. Time course of non-amiloride-sensitive (NS)- I sc response to cAMP and glibenclamide in FL-MDCK cells in which the basolateral membrane was permeabilized by -toxin. The apical solution contained 135 mM Na + and gluconate as the major ions, and the basolateral solution contained 135 mM K + and Cl - as the major ions. Ten micromolar amiloride was added to the apical solutions to block Na + transport. The result is representative of 5 experiments (summarized in the text and Fig. 4 A ), with average values given in the text.
To determine the anion selectivity of the apical membrane conductance, we performed experiments such as that shown in Fig. 3 with Cl -, I -, Br -, or NO 3 - as the major permeant anion ( Fig. 4 ). As shown in Fig. 4 A, NS- I sc was significantly stimulated by cAMP treatment and inhibited by glibenclamide when Cl - was replaced by any of the other anions. The anion selectivity sequence based on the cAMP-activated NS- I sc was NO 3 - Br - Cl - I - ( Fig. 4 B ). These experiments as well as the presence of CFTR protein indicate that CFTR comprises a significant fraction of the cAMP-activated anion conductance in the apical membrane of FL-MDCK cells.
Fig. 4. Effect of basolateral anion substitution on NS- I sc responses to cAMP monolayers of FL-MDCK cells in which the basolateral membrane was permeabilized by -toxin. A : summary of responses of NS- I sc to cAMP and glibenclamide in basolaterally permeabilized monolayers. In each group of 3 bars, the first is the control condition, the second is after cAMP addition, and the third is after glibenclamide addition. In each group, with I -, Cl -, Br -, or NO 3 - as the major anion in the basolateral solution, the effect of both cAMP and glibenclamide was statistically significant ( P < 0.001). B : mean cAMP-activated NS- I sc (the paired difference between the first and second columns for each anion in A ) obtained with different anions in 4 such experiments.
Regulation of ENaC by intracellular Cl -. It has been shown that Cl - transport or intracellular Cl - is important for the inhibition of ENaC by CFTR in X. laevis oocytes and other epithelia ( 6, 32 ). To examine the relationship between the Na + current via ENaC and Cl - current via CFTR in FL-MDCK cells, we again wished to permeabilize the basolateral membrane so that the apical transporters could be isolated. However, in the -toxin-permeabilized cells, AS- I sc was low and was not increased by cAMP, indicating that ENaC activity was inhibited. Given the fact that -toxin makes plasma membranes permeable to molecules of up to 2-3 kDa ( 57 ), it seems to us most likely that this method of permeabilization produces a very high intracellular Ca 2+ concentration that would inhibit all ENaC activity.
Because of these limitations with the -toxin permeabilization method, we used nystatin permeabilization for those experiments in which AS- I sc was measured. Nystatin is known to create aqueous pores of 4 radius in thin lipid membranes, which exclude Ca 2+ but not smaller monovalent ions such as Na +, K +, and Cl - ( 16, 26 ). In the experiments shown in Fig. 5, FL-MDCK cell monolayers were mounted in Ussing chambers with KRB on both sides, and I sc was measured as usual. Nystatin (350 µg/ml) was then added to the basolateral solution, and both I sc and the transepithelial voltage (measured under open-circuit conditions) fell to zero within 15 min. A short-circuit current was then produced by replacing Na + with K + in the basolateral solution and thus imposing an external gradient for a Na + flux in the absorptive direction.
Fig. 5. Effect of [Cl - ] i on amiloride-sensitive (AS)- I sc in FL-MDCK monolayers with a nystatin-permeabilized basolateral membrane. Na + was present only on the apical side at 120 mM (140 mM K + in the basolateral solution), which provided a favorable concentration gradient for an absorptive Na + flux. A basolateral Cl - gradient: the Cl - concentration in the basolateral solution was kept constant at 10 mM, whereas the apical Cl - concentration was either 10, 50, or 125 mM (gluconate replacement). Mean values are not significantly different. B apical Cl - gradient: the Cl - concentration in the apical solution was kept constant at 10 mM, whereas the basolateral Cl - concentration was either 10, 50, 90, or 125 mM. Each mean for AS- I sc was significantly different, and AS- I sc was negatively correlated with the basolateral Cl - concentration ( r = 0.95, * P < 0.0001). C : no Cl - concentration difference. Apical and basolateral Cl - concentrations were kept equal at 10, 50, or 125 mM. Mean value of AS- I sc at 10 mM Cl - is significantly different from the other 2 higher Cl - concentrations, with no significant difference between mean values of 50 and 125 mM. Mean values of all 3 groups are negatively correlated ( r = 0.83, * P < 0.001) with the basolateral Cl - concentration.
As shown in Fig. 5 A, when the basolateral Cl - concentration was kept at 10 mM (gluconate replacement) while the apical concentration was increased from 10 to 125 mM, thereby increasing the Cl - -absorptive current, AS- I sc was not affected ( n = 29, P = 0.75). However, as shown in Fig. 5 B, when apical Cl - was kept at 10 mM and the basolateral Cl - concentration was increased from 10 to 125 mM, the amiloride-sensitive current was significantly inhibited ( n = 16, P < 0.001). To test whether a Cl - flux was necessary to inhibit ENaC, we prevented a Cl - current by adding symmetrical Cl - concentrations to the apical and basolateral sides of the monolayers. As shown in Fig. 5 C, AS- I sc was inhibited when the Cl - concentration was increased on both sides of the monolayer ( n = 12, P < 0.001). These data results show that ENaC inhibition is proportional to the basolateral Cl - concentration and hence to the cytoplasmic Cl - concentration in these permeabilized cells regardless of the presence or absence of a net Cl - flux.
Effect of cAMP on [Cl - ] i. Given the marked inhibition of AS- I sc by [Cl - ] i in the preceding experiments ( Fig. 5 ), we tested whether a change of [Cl - ] i might explain the late inhibition of ENaC when anion secretion is stimulated by cAMP. We measured [Cl - ] i in FL-MDCK cells grown on Transwell inserts with the fluorescent dye SPQ using the double ionophore technique ( 9, 31 ) as described in METHODS. The fluorescence was measured using standard solutions with varying extracellular Cl - concentrations, and a Stern-Volmer plot for the relationship between (F 0 /F - 1) vs. [Cl - ] i gave a K sv of 9.3 M -1. Based on the SPQ data for five experiments such as the one shown in Fig. 6 A, [Cl - ] i was estimated to be 76 ± 14 mM in the absence of cAMP treatment (control) and rapidly decreased to 35.7 ± 8.7 mM ( P = 0.03) after cAMP treatment ( Fig. 6 B ). The response was sustained, and [Cl - ] i returned to basal levels after cAMP was washed out. The decrease in [Cl - ] i that was produced by cAMP could also be reversed to the basal level by 200 µM glibenclamide ( Fig. 6 C ), indicating again that the Cl - secretion activated by cAMP was mediated through CFTR.
Fig. 6. Effect of cAMP treatment on [Cl - ] i. A : representative experiment showing effect of addition of 100 µM CPT-cAMP and 100 µM IBMX (cAMP) on the relative fluorescent intensity of 6-methoxy- N -(3-sulfopropyl)-quinolinium. B : average calculated [Cl - ] i for 5 experiments like that shown in A (* P = 0.03). C : representative experiment showing effect of addition of CPT-cAMP and IBMX (cAMP), followed by 200 µM glibenclamide on the fluorescent intensity of MEQ. Increases in fluorescence intensity indicate decreases in [Cl - ] i.
Effect of intracellular Cl - on ENaC surface density. To examine whether elevated intracellular Cl - inhibits ENaC by decreasing the number of channel subunits present in the apical membrane or by reducing the intrinsic channel activity, we used the binding of a 125 I-labeled monoclonal antibody (M2) to the FLAG epitope to quantify the density of total ENaC subunits in the apical membrane and compared the surface density of ENaC subunits with AS- I sc. Paired electrophysiological and binding experiments were conducted in monolayers treated with nystatin to permeabilize the basolateral membrane. Under these conditions, electrical coupling between anion channels and ENaC in the apical membrane is avoided, so the AS- I sc becomes proportional to the ENaC-mediated conductance and the total driving force is the chemical gradient of Na + from the apical-to-basolateral side.
As shown in Fig. 7 A, increasing the Cl - concentration in the basolateral solution (and thus presumably [Cl - ] i ) from 15 to 145 mM decreased AS- I sc from 24.5 ± 1.0 to 10.2 ± 1.6 µA/cm 2 ( n = 6, P < 0.001). Because one FLAG epitope was inserted into the extracellular domain of every ENaC subunit, the apical membrane surface density of ENaC subunits could be measured by the specific binding of 125 I-labeled anti-FLAG (M2) antibody. As shown in Fig. 7 B, the specific binding of 4 nM M2 antibody was decreased from 0.65 ± 0.05 to 0.43 ± 0.05 fmol/cm 2 ( n = 7, P < 0.01), when Cl - concentration was increased from 15 to 145 mM. To estimate the surface density of ENaC subunits from saturation binding (B max ), we compared the concentration dependence of M2 Ab binding to that we had observed previously in DMEM ( 43 ) and verified the constant for half-maximal binding k 0.5 of 7.9 nM (data not shown). Using the Michaelis-Menten equation ( 43 ), B max was estimated, respectively, to be 1.91 ± 0.16 or 1.32 ± 0.17 fmol/cm 2 ( P < 0.05) with 15 or 145 mM Cl - in the basolateral solution ( Fig. 8 ). In summary, the 58% reduction in AS- I sc that was produced by this increase in [Cl - ] i was explained by a 31% decrease in the density of ENaC in the apical membrane and a 38% reduction in the intrinsic channel activity (AS- I sc / n ).
Fig. 7. Effect of [Cl - ] i on AS- I sc and specific binding of 125 I-labeled M2 Ab in FL-MDCK monolayers with a nystatin-permeabilized basolateral membrane. A : apical and basolateral Cl - concentrations were kept equal at 15 or 145 mM. Mean value of AS- I sc at 15 mM Cl - is significantly different from 145 mM (* P < 0.001). Average results of 6 experiments are given in the text. B : specific binding of 125 I-labeled M2 antibody in FL-MDCK cells. Apical and basolateral Cl - concentrations were kept equal at 15 or 145 mM. The M2 antibody concentration used in the binding experiments was 4 nM. Mean value of bound M2 antibody at 15 mM Cl - is significantly different from that at 145 mM ( n = 7, * P < 0.05). Averages are given in the text.
Fig. 8. Comparison of Cl - effect on ENaC subunit density in the apical membrane and single-channel properties. Due to the constant driving force for Na + reabsorption that is produced by voltage clamping of the apical membrane in nystatin-permeabilized monolayers, AS- I sc is proportional to the ENaC conductance, which is the product of the number of channels ( n ), the single-channel conductance ( ), and the channel open probability ( P o ). Thus the ratio AS- I sc / n is proportional to · P o.
DISCUSSION
Our previous work showed that the triply transfected FL-MDCK cells, a subclone of the F F F MDCK cells developed by Morris and Schafer ( 43 ), are a useful model epithelium for the study of ENaC regulation. In the current study, we characterized Cl - secretion mediated by anion channels in the apical membrane and examined in greater detail the effect of intracellular Cl - on ENaC surface expression and intrinsic channel activity.
Anion channels associated with NS-I sc. As expected, the FL-MDCK clone used in these studies expressed all three of the transfected ENaC subunits ( Fig. 1, left ), and the presence of the FLAG epitope in the extracellular domain of each subunit was verified by RT-PCR. Epithelial monolayers of these cells uniformly exhibited a high basal I sc, most of which was eliminated by 10 µM apical amiloride ( Fig. 2 ) as previously observed by Morris and Schafer ( 43 ). We hypothesized that anion secretion mediated the NS- I sc because in the presence of apical amiloride, Cl - and HCO 3 - were the only ions present in sufficient concentration to account for the I sc observed, and previous studies showed that the apical membrane of MDCK cells has no measurable K + permeability ( 1 ). The biphasic time course of the I sc response to cAMP ( Fig. 2 ) was consistent with previous studies in MDCK cells ( 43, 54 ) and in A6 cells ( 4, 8, 29, 44, 58 ) that showed a rapid increase in anion secretion followed by a more gradual increase in Na + absorption.
The secretory anion current, represented by NS- I sc, was partially inhibited by 200 µM glibenclamide ( Figs. 2 and 3 ), which is a well-established inhibitor of CFTR at this concentration ( 33, 45 ). The expression of CFTR protein in FL-MDCK cells was confirmed by RT-PCR and immunoprecipitation experiments ( Fig. 1, right ). This finding is in good agreement with Mohamed et al. ( 42 ), who previously reported the presence of CFTR in MDCK type 1 cells (the progenitor for the FL-MDCK cells used in these studies) but not MDCK type 2 cells. In experiments such as that shown in Fig. 3, in which -toxin was used to permeabilize the basolateral membrane and isolate the apical transporters, NS- I sc was increased by cAMP and inhibited by glibenclamide added to the apical solution as would be expected if CFTR was present in the apical membrane. Also, in the -toxin-permeabilized monolayers, the selectivity sequence of the cAMP-activated anion conductance was NO 3 - Br - Cl - I - ( Fig. 4 ), which is a unique characteristic of CFTR ( 2, 3 ).
In preliminary experiments that are not shown here, we also found evidence for a Ca 2+ -activated Cl - channel (CaCC) and the ClC-2 channel in these cells. Thapsigargin, which moderately increases [Ca 2+ ] i in MDCK cells ( 35 ), produced a sustained increase in anion secretion that could be completely blocked by apical DIDS as expected for CaCC ( 19 ). Thus, given the fact that cAMP can increase intracellular [Ca 2+ ] i in the MDCK cells ( 10 ), CaCC may contribute to the fraction of NS- I sc that is not sensitive to glibenclamide ( Figs. 2 and 3 ). A Ca 2+ -activated Cl - conductance has been described in mouse M-1 cortical collecting duct cells ( 41 ) and in primary cultures of rabbit proximal and distal tubule cells ( 52 ). We were unable to detect an RT-PCR product using various primer pairs based on the CaCC sequence from human or mouse; however, it is quite possible that the dog sequence is significantly different. Immunoblot and RT-PCR studies (data not shown) showed that these FL-MDCK cells also expressed the ClC-2 isoform of the Cl - channel that is widely distributed in many similar epithelia ( 27, 45 ) and that can be activated by cAMP acting via PKA as well as by cell swelling and a low extracellular pH ( 12, 27 ). However, we have no information about the localization of ClC-2 to the apical vs. basolateral membrane or about its possible contribution to NS- I sc. Based on these results, we conclude that CFTR is the major but not the only Cl - channel in the apical membrane of FL-MDCK cells and at least one other Cl - channel may contribute to anion secretion and its augmentation by cAMP.
Effect of intracellular Cl - on ENaC. Kunzelmann and collaborators ( 6, 30, 32 ) showed that the inhibition of ENaC in X. laevis oocytes by coexpression of CFTR is not specific to this Cl - channel but can be produced by an increase in intracellular [Cl - ] i due to coexpression of other Cl - channels or even nonspecific membrane permeabilization by amphotericin. The inhibitory effect of [Cl - ] i on ENaC channels has also been demonstrated in patch-clamp studies of other mammalian epithelia (e.g., 14, 34) and in sweat gland ducts in which the basolateral membrane was permeabilized by -toxin ( 49 ).
To determine whether [Cl - ] i had an inhibitory effect on ENaC in FL-MDCK cells, we permeabilized the basolateral membranes with nystatin, which allowed us to study the apical membrane transporters without the complication of the basolateral membrane and to vary the cytosolic Cl - concentration. Nystatin produces aqueous pores of 4- radius that allow rapid equilibration of small monovalent ions but not Ca 2+ ( 16, 26 ). Thus, in the presence of a transepithelial Na + gradient, there was a large AS- I sc in the nystatin-permeabilized cells ( Fig. 5 ) but not in the -toxin-permeabilized cells ( Fig. 3 ), in which the elevation of [Ca 2+ ] i inhibits ENaC. In similar experiments with sweat ducts, Reddy et al. ( 48 ) avoided this inhibition of ENaC during -toxin permeabilization by decreasing the basolateral Ca 2+ concentration. However, in our experiments, we could not reduce the Ca 2+ concentration in the basolateral solution without complete disruption of the epithelium. In our experiments, we found that the addition of nystatin to the basolateral solution in the Ussing chamber rapidly reduced both I sc and the open-circuit voltage to zero in the absence of any transepithelial ion concentration gradients as expected, and the transepithelial conductance remained low, indicating that the basolateral membrane was permeabilized but the paracellular junction remained intact.
In the experiments shown in Fig. 5, we manipulated the direction and magnitude of Cl - current by varying the Cl - concentrations in the external solutions. An Na + concentration gradient from the apical (120 mM Na +, 0 K + ) to basolateral (0 Na +, 120 mM K + ) was used to produce a net driving force for a Na + -absorptive flux, because the apical membrane potential difference is clamped to zero when these permeabilized cells are short-circuited, AS- I sc is a direct measure of Na + absorption via ENaC and is not complicated by the possible effects of changes in the anion conductance on the electrochemical potential driving force for Na + across the apical membrane (see Ref. 24 ). Under these conditions, we found that increases in the basolateral Cl - concentration, and hence in [Cl - ] i, inhibited ENaC, but that changes in the Cl - concentration in the apical solution had no effect. Furthermore, the inhibitory effect of [Cl - ] i on ENaC was solely dependent on the intracellular Cl - concentration rather than the magnitude or direction of the Cl - current.
In the experiments with nystatin-permeabilized monolayers, the Cl - concentration was raised by replacing gluconate in the apical or basolateral solution. Expecting that Cl - is much more permeant than gluconate, one might argue that cell swelling was, in some way, responsible for the inhibitory effect of elevated [Cl - ] i on ENaC [e.g., ClC-2 is activated by cell swelling ( 27 )]; however, when cell swelling was prevented or cell shrinkage was produced by adding mannitol with Cl -, we obtained the same inhibitory effect of [Cl - ] i (data not shown).
Effect of cAMP on [Cl - ] i. Although increased [Cl - ] i inhibited ENaC in the permeabilized FL-MDCK cells in this study as it did in the oocyte and patch-clamp studies, the inhibitory effect of increased anion secretion on ENaC that we observed in intact FL-MDCK monolayers ( 43 ) could not be attributed to an increase in [Cl - ] i if the activation of Cl - channels in the apical membrane decreased [Cl - ] i as reported in other epithelia ( 7, 53 ).
To determine the changes in the intracellular Cl - concentration with cAMP stimulation, we developed a method to measure [Cl - ] i in intact monolayers of our FL-MDCK cells on permeable supports using the Cl - -sensitive fluorescent dye SPQ. Because the two sides of these monolayers were not electrically isolated in the cuvette of the fluorometer, they were effectively short-circuited as they were in the electrophysiological studies. Under control conditions, i.e., in KRB medium and in the absence of cAMP treatment ( Fig. 7 ), [Cl - ] i was 76 ± 14 mM, and cAMP treatment decreased it to 36 ± 9 ( P = 0.03). Although the control [Cl - ] i is somewhat higher than measured in some epithelia that secrete Cl -, e.g., 47 mM in canine tracheal epithelium ( 53 ), it is quite close to the measurement in others, e.g., 61 mM in salivary acinar cells ( 18 ). However, the important point is that [Cl - ] i falls on treatment with cAMP and should activate rather than inhibit ENaC. Thus the late fall in ENaC activity after cAMP treatment, which was observed previously in this ( 43 ) and similar epithelia ( 8, 29, 34, 44 ), cannot be attributed to the effect of stimulated anion secretion on [Cl - ] i.
Mechanism of the inhibitory effect of [Cl - ] i on ENaC. Returning to the effect of elevated [Cl - ] i on ENaC that we observed in the permeabilized monolayers, two mechanisms might account for the inhibition of ENaC by [Cl - ] i : a decrease in the number of ENaC subunits ( n ) in the apical membrane, or a decrease in the intrinsic single-channel activity, i.e., a decrease in the single-channel conductance ( ) or the open probability ( P o ). Because one FLAG epitope was inserted into the extracellular domain of every ENaC subunit, the binding of 125 I-labeled monoclonal antibody (M2) to the FLAG epitope permitted an estimate of the molar density of ENaC subunits ( 17, 43 ). As shown in Fig. 8, the B max estimates for the experiments in Fig. 7 were 1.91 ± 0.16 and 1.32 ± 0.17 fmol/cm 2 in the presence of, respectively, low and high [Cl - ] i ( P < 0.05). Based on a calculated cell density of 2.6·10 6 cells/cm 2 in the confluent monolayers, the subunit density would be 460 subunits per cell at low [Cl - ] i and 320 subunits per cell at high [Cl - ] i.
As shown in Fig. 8, top, the ratio of AS- I sc to Ab bound, i.e., the average current per subunit, depends on the single-channel properties of ENaC and is proportional to the product · P o. The analysis presented in Fig. 8 shows that the average current per subunit is significantly decreased from 13.3 ± 1.2 to 8.2 ± 1.4 µA/fmol by increased [Cl - ] i, indicating that either or P o, or both, is inhibited by [Cl - ] i. Using the value of 4.7-pS conductance observed in rENaC-transfected MDCK cells from the same MDCK cell line ( 26 ), estimates of the range of P o for ENaC can be calculated. In this analysis, it is recognized that the electrochemical potential gradient for Na + across the short-circuited apical membrane is equal to the chemical potential difference ( Na ) imposed by the transepithelial Na + concentration difference. With a nominal cytosolic Na + concentration of 6 mM, Na is 80 mV. If one assumes, for example, that there are four subunits per active ENaC, P o would be 0.22 and 0.14 in the presence of, respectively, low and high [Cl - ] i. 1 It should be noted that these estimates of P o are considerably higher than those in our previous study ( 43 ) or that of Firsov et al. ( 17 ). In other words, it appears that an elevation of [Cl - ] i decreases both the number of ENaC subunits ( n ) and the intrinsic ENaC channel activity ( · P o ). Our conclusion that [Cl - ] i reduces ENaC-mediated Na + transport due to a decrease in n as well as · P o is consistent with that of Marunaka et al. ( 40 ), who reported that an increase in [Cl - ] i decreased the P o of the amiloride-sensitive nonselective cation channel by increasing the closing rate without any significant change in the opening rate in fetal rat alveolar epithelium, but different from that of Dinudom et al. ( 13 ) who showed that trafficking is not involved in the feedback control of Na + channels by intracellular anions in mouse mandibular gland ducts.
In addition to its inhibitory effect on ENaC shown in the present study, intracellular Cl - has been reported to regulate the activity of several other ion transporters, including NKCC1, the Na + /H + exchanger, and a nonselective cation channel ( 50 ). Several laboratories identified Cl - -sensitive proteins that may mediate the effect of changes in [Cl - ] i on ion transporters, such as Cl - -sensitive kinases ( 15, 56 ) and GTP-binding proteins ( 23 ) that may regulate amiloride-sensitive Na + channels. Alternatively, a direct and nonspecific interaction of intracellular Cl - with amiloride-sensitive, nonselective cation channel activity has been proposed in fetal rat alveolar epithelium ( 40 ). Our results do not speak to these possibilities, but they do show that an elevation of [Cl - ] i inhibits ENaC-mediated Na + absorption by approximately proportional decreases in the density of ENaC in the apical membrane and in the intrinsic activity of these channels.
GRANTS
This study was supported by NIH Grant DK-25519-21 and a predoctoral fellowship (R464-CR02) from the Cystic Fibrosis Foundation.
DISCLOSURES
Some of the data have been presented previously in abstract form ( FASEB J 17: A1225, 2003).
ACKNOWLEDGMENTS
We are particularly grateful to Dr. R. G. Morris, who developed the retrovirally transfected cell line used in these studies, and to Dr. D. Bell of the UAB Dept. of Medicine (Nephrology) for the use of the fluorometric equipment in his laboratory. We thank Drs. D. Bell and P. Komolosi for assistance in operating the equipment and designing the experiments with SPQ. We also thank Dr. J. Collawn (UAB Dept. of Cell Biology) for generously providing the anti-CFTR (NBD-1) antibody and Dr. M. Knepper [National Institutes of Health (NIH), Bethesda, MD] for the anti-ENaC antibodies. We also acknowledge the superb technical assistance of Dr. L. Li and M. L. Watkins.
【参考文献】
Aiton JF, Brown CD, Ogden P, and Simmons NL. K + transport in "tight" epithelial monolayers of MDCK cells. J Membr Biol 65: 99-109, 1982.
Anderson MP, Gregory RJ, Thompson S, Souza DW, Paul S, Mulligan RC, Smith AE, and Welsh MJ. Demonstration that CFTR is a chloride channel by alteration of its anion selectivity. Science 253: 202-205, 1991.
Bell CL and Quinton PM. T84 cells: anion selectivity demonstrates expression of Cl - conductance affected in cystic fibrosis. Am J Physiol Cell Physiol 262: C555-C562, 1992.
Bindels RJM, Schafer JA, and Reif MC. Stimulation of sodium transport by aldosterone and arginine vasotocin in A6 cells. Biochim Biophys Acta 972: 320-330, 1988.
Boucherot A, Schreiber R, and Kunzelmann K. Role of CFTR's PDZ1-binding domain, NBF1 and Cl - conductance in inhibition of epithelial Na + channels in Xenopus oocytes. Biochim Biophys Acta 1515: 64-71, 2001.
Briel M, Greger R, and Kunzelmann K. Cl - transport by cystic fibrosis transmembrane conductance regulator (CFTR) contributes to the inhibition of epithelial Na + channels (ENaCs) in Xenopus oocytes coexpressing CFTR and ENaC. J Physiol 508: 825-836, 1998.
Burnham DB and Fondacaro JD. Secretagogue-induced protein phosphorylation and chloride transport in Caco-2 cells. Am J Physiol Gastrointest Liver Physiol 256: G808-G816, 1989.
Chalfant ML, Coupaye-Gerard B, and Kleyman TR. Distinct regulation of Na + reabsorption and Cl - secretion by arginine vasopressin in the amphibian cell line A6. Am J Physiol Cell Physiol 264: C1480-C1488, 1993.
Chao AC, Dix JA, Sellers MC, and Verkman AS. Fluorescence measurement of chloride transport in monolayer cultured cells. Mechanisms of chloride transport in fibroblasts. Biophys J 56: 1071-1081, 1989.
Chase HS Jr and Wong SM. Isoproterenol and cyclic AMP increase intracellular free in MDCK cells. Am J Physiol Renal Fluid Electrolyte Physiol 254: F374-F384, 1988.
Chinet TC, Fullton JM, Yankaskas JR, Boucher RC, and Stutts MJ. Mechanism of sodium hyperabsorption in cultured cystic fibrosis nasal epithelium: a patch-clamp study. Am J Physiol Cell Physiol 266: C1061-C1068, 1994.
Cuppoletti J, Tewari KP, Sherry AM, Kupert EY, and Malinowska DH. ClC-2 Cl - channels in human lung epithelia: activation by arachidonic acid, amidation, and acid-activated omeprazole. Am J Physiol Cell Physiol 281: C46-C54, 2001.
Dinudom A, Harvey KF, Komwatana P, Young JA, Kumar S, and Cook DI. Nedd4 mediates control of an epithelial Na + channel in salivary duct cells by cytosolic Na +. Proc Natl Acad Sci USA 95: 7169-7173, 1998.
Dinudom A, Young JA, and Cook DI. Na + and Cl - conductances are controlled by cytosolic Cl - concentration in the intralobular duct cells of mouse mandibular glands. J Membr Biol 135: 289-295, 1993.
Dowd BF and Forbush B. PASK (proline-alanine-rich STE20-related kinase), a regulatory kinase of the Na-K-Cl cotransporter (NKCC1). J Biol Chem 278: 27347-27353, 2003.
Finkelstein A and Holz R. Aqueous pores created in thin lipid membranes by the polyene antibiotics nystatin and amphotericin B. Membranes 2: 377-408, 1973.
Firsov D, Schild L, Gautschi I, Merillat AM, Schneeberger E, and Rossier BC. Cell surface expression of the epithelial Na channel and a mutant causing Liddle syndrome: a quantitative approach. Proc Natl Acad Sci USA 93: 15370-15375, 1996.
Foskett JK. [Ca 2+ ] i modulation of Cl - content controls cell volume in single salivary acinar cells during fluid secretion. Am J Physiol Cell Physiol 259: C998-C1004, 1990.
Fuller CM, Ji HL, Tousson A, Elble RC, Pauli BU, and Benos DJ. Ca 2+ -activated Cl - channels: a newly emerging anion transport family. Pflügers Arch 443: S107-S110, 2001.
Garty H and Palmer LG. Epithelial sodium channels: function, structure, and regulation. Physiol Rev 77: 359-396, 1997.
Greger R, Mall M, Bleich M, Ecke D, Warth R, Riedemann N, and Kunzelmann K. Regulation of epithelial ion channels by the cystic fibrosis transmembrane conductance regulator. J Mol Med 74: 527-534, 1996.
Greger R and Schlatter E. Mechanism of NaCl secretion in the rectal gland of spiny dogfish ( Squalus acanthias ). I. Experiments in isolated in vitro perfused rectal gland tubules. Pflügers Arch 402: 63-75, 1984.
Higashijima T, Ferguson KM, and Sternweis PC. Regulation of hormone-sensitive GTP-dependent regulatory proteins by chloride. J Biol Chem 262: 3597-3602, 1987.
Horisberger JD. ENaC-CFTR interactions: the role of electrical coupling of ion fluxes explored in an epithelial cell model. Pflügers Arch 445: 522-528, 2003.
Ishibashi K. Intracellular Cl - activity in rabbit proximal convoluted tubule perfused in vitro: regulation by sodium and effects of anion transport inhibitors. Jpn J Physiol 43: 585-597, 1993.
Ishikawa T, Marunaka Y, and Rotin D. Electrophysiological characterization of the rat epithelial Na + channel (rENaC) expressed in MDCK cells. Effects of Na + and Ca 2+. J Gen Physiol 111: 825-846, 1998.
Jentsch TJ, Stein V, Weinreich F, and Zdebik AA. Molecular structure and physiological function of chloride channels. Physiol Rev 82: 503-568, 2002.
Ji HL, Chalfant ML, Jovov B, Lockhart JP, Parker SB, Fuller CM, Stanton BA, and Benos DJ. The cytosolic termini of the - and -ENaC subunits are involved in the functional interactions between cystic fibrosis transmembrane conductance regulator and epithelial sodium channel. J Biol Chem 275: 27947-27956, 2000.
Kleyman TR, Ernst SA, and Coupaye-Gerard B. Arginine vasopressin and forskolin regulate apical cell surface expression of epithelial Na + channels in A6 cells. Am J Physiol Renal Fluid Electrolyte Physiol 266: F506-F511, 1994.
König J, Schreiber R, Voelcker T, Mall M, and Kunzelmann K. The cystic fibrosis transmembrane conductance regulator (CFTR) inhibits ENaC through an increase in the intracellular Cl - concentration. EMBO Rep 2: 1047-1051, 2001.
Krapf R, Berry CA, and Verkman AS. Estimation of intracellular chloride activity in isolated perfused rabbit proximal convoluted tubules using a fluorescent indicator. Biophys J 53: 955-962, 1988.
Kunzelmann K. ENaC is inhibited by an increase in the intracellular Cl - concentration mediated through activation of Cl - channels. Pflügers Arch 445: 504-512, 2003.
Kunzelmann K and Schreiber R. CFTR, a regulator of channels. J Membr Biol 168: 1-8, 1999.
Letz B and Korbmacher C. cAMP stimulates CFTR-like Cl - channels and inhibits amiloride-sensitive Na + channels in mouse CCD cells. Am J Physiol Cell Physiol 272: C657-C666, 1997.
Lien YH, Wang X, Gillies RJ, and Martinez-Zaguilan R. Modulation of intracellular Ca 2+ by glucose in MDCK cells: role of endoplasmic reticulum Ca 2+ -ATPase. Am J Physiol Renal Fluid Electrolyte Physiol 268: F671-F679, 1995.
Loffing-Cueni D, Loffing J, Shaw C, Taplin AM, Govindan M, Stanton CR, and Stanton BA. Trafficking of GFP-tagged F508-CFTR to the plasma membrane in a polarized epithelial cell line. Am J Physiol Cell Physiol 281: C1889-C1897, 2001.
MacVinish LJ, Reancharoen T, and Cuthbert AW. Kinin-induced chloride permeability changes in colony 29 epithelia estimated from 125 I-efflux and MEQ fluorescence. Br J Pharmacol 108: 469-478, 1993.
Mall M, Bleich M, Greger R, Schreiber R, and Kunzelmann K. The amiloride-inhibitable Na + conductance is reduced by the cystic fibrosis transmembrane conductance regulator in normal but not in cystic fibrosis airways. J Clin Invest 102: 15-21, 1998.
Mandel KG, Dharmsathaphorn K, and McRoberts JA. Characterization of a cyclic AMP-activated Cl - transport pathway in the apical membrane of a human colonic epithelial cell line. J Biol Chem 261: 704-712, 1986.
Marunaka Y, Niisato N, O'Brodovich H, and Eaton DC. Regulation of an amiloride-sensitive Na + -permeable channel by a 2 -adrenergic agonist, cytosolic Ca 2+ and Cl - in fetal rat alveolar epithelium. J Physiol 515: 669-683, 1999.
Meyer K and Korbmacher C. Cell swelling activates ATP-dependent voltage-gated chloride channels in M-1 mouse cortical collecting duct cells. J Gen Physiol 108: 177-193, 1996.
Mohamed A, Ferguson D, Seibert FS, Cai HM, Kartner N, Grinstein S, Riordan JR, and Lukacs GL. Functional expression and apical localization of the cystic fibrosis transmembrane conductance regulator in MDCK I cells. Biochem J 322: 259-265, 1997.
Morris RG and Schafer JA. cAMP increases density of ENaC subunits in the apical membrane of MDCK cells in direct proportion to amiloride-sensitive Na + transport. J Gen Physiol 120: 71-85, 2002.
Morris RG, Tousson A, Benos DJ, and Schafer JA. Microtubule disruption inhibits AVT-stimulated Cl - secretion but not Na + reabsorption in A6 cells. Am J Physiol Renal Physiol 274: F300-F314, 1998.
Nilius B and Droogmans G. Amazing chloride channels: an overview. Acta Physiol Scand 177: 119-147, 2003.
Ostedgaard LS, Shasby DM, and Welsh MJ. Staphylococcus aureus -toxin permeabilizes the basolateral membrane of a Cl - -secreting epithelium. Am J Physiol Lung Cell Mol Physiol 263: L104-L112, 1992.
Petersen KU and Reuss L. Cyclic AMP-induced chloride permeability in the apical membrane of Necturus gallbladder epithelium. J Gen Physiol 81: 705-729, 1983.
Reddy MM, Light MJ, and Quinton PM. Activation of the epithelial Na + channel (ENaC) require CFTR Cl - channel function. Nature 402: 301-304, 1999.
Reddy MM and Quinton PM. Functional interaction of CFTR and ENaC in sweat glands. Pflügers Arch 445: 499-503, 2003.
Robertson MA and Foskett JK. Na + transport pathways in secretory acinar cells: membrane cross talk mediated by [Cl - ] i. Am J Physiol Cell Physiol 267: C146-C156, 1994.
Rotin D, Kanelis V, and Schild L. Trafficking and cell surface stability of ENaC. Am J Physiol Renal Physiol 281: F391-F399, 2001.
Rubera I, Tauc M, Bidet M, Poujeol C, Cuiller B, Watrin A, Touret N, and Poujeol P. Chloride currents in primary cultures of rabbit proximal and distal convoluted tubules. Am J Physiol Renal Physiol 275: F651-F663, 1998.
Shorofsky SR, Field M, and Fozzard HA. Mechanism of Cl secretion in canine trachea: changes in intracellular chloride activity with secretion. J Membr Biol 81: 1-8, 1984.
Simmons NL. Chloride secretion stimulated by prostaglandin E 1 and by forskolin in a canine renal epithelial cell line. J Physiol 432: 459-472, 1991.
Stutts MJ, Rossier BC, and Boucher RC. Cystic fibrosis transmembrane conductance regulator inverts protein kinase A-mediated regulation of epithelial sodium channel single channel kinetics. J Biol Chem 272: 14037-14040, 1997.
Treharne KJ, Riemen CE, Marshall LJ, Muimo R, and Mehta A. Nucleoside diphosphate kinase-a component of the [Na + ]- and [Cl - ]-sensitive phosphorylation cascade in human and murine airway epithelium. Pflügers Arch 443, Suppl 1: S97-S102, 2001.
Valeva A, Weisser A, Walker B, Kehoe M, Bayley H, Bhakdi S, and Palmer M. Molecular architecture of a toxin pore: a 15-residue sequence lines the transmembrane channel of staphylococcal -toxin. EMBO J 15: 1857-1864, 1996.
Verrey F, Groscurth P, and Bolliger U. Cytoskeletal disruption in A6 kidney cells: impact on endo/exocytosis and NaCl transport regulation by antidiuretic hormone. J Membr Biol 145: 193-204, 1995.
作者单位：Department of Physiology and Biophysics and Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama 35294