HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 36, 27947-27956, September 8, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/36/27947    most recent M002848200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ji, H.-L. Articles by Benos, D. J. Search for Related Content PubMed PubMed Citation Articles by Ji, H.-L. Articles by Benos, D. J. Social Bookmarking                      What's this?

The Cytosolic Termini of the - and -ENaC Subunits Are Involved in the Functional Interactions between Cystic Fibrosis Transmembrane Conductance Regulator and Epithelial Sodium Channel^*

Hong-Long Ji, Michael L. Chalfant^§, Biljana Jovov, Jason P. Lockhart, Suzanne B. Parker, Catherine M. Fuller, Bruce A. Stanton^§, and Dale J. Benos^¶

From the  Department of Physiology and Biophysics, University of Alabama at Birmingham, Birmingham, Alabama 35294-0005 and the ^§ Department of Physiology, Dartmouth Medical School, Hanover, New Hampshire 03755

Received for publication, April 4, 2000, and in revised form, May 2, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Epithelial sodium channel (ENaC) and cystic fibrosis transmembrane conductance regulator (CFTR) are co-localized in the apical membrane of many epithelia. These channels are essential for electrolyte and water secretion and/or reabsorption. In cystic fibrosis airway epithelia, a hyperactivated epithelial Na⁺ conductance operates in parallel with defective Cl secretion. Several groups have shown that CFTR down-regulates ENaC activity, but the mechanisms and the regulation of CFTR by ENaC are unknown. To test the hypothesis that ENaC and CFTR regulate each other, and to identify the region(s) of ENaC involved in the interaction between CFTR and ENaC, rENaC and its mutants were co-expressed with CFTR in Xenopus oocytes. Whole cell macroscopic sodium currents revealed that wild type (wt) -rENaC-induced Na⁺ current was inhibited by co-expression of CFTR, and further inhibited when CFTR was activated with a cAMP-raising mixture (CKT). Conversely, -rENaC stimulated CFTR-mediated Cl currents up to ~6-fold. Deletion mutations in the intracellular tails of the three rENaC subunits suggested that the carboxyl terminus of the  subunit was required both for the down-regulation of ENaC by activated CFTR and the up-regulation of CFTR by ENaC. However, both the carboxyl terminus of the  subunit and the amino terminus of the  subunit were essential for the down-regulation of rENaC by unstimulated CFTR. Interestingly, down-regulation of rENaC by activated CFTR was Cl-dependent, while stimulation of CFTR by rENaC was not dependent on either cytoplasmic Na⁺ or a depolarized membrane potential. In summary, there appear to be at least two different sites in ENaC involved in the intermolecular interaction between CFTR and ENaC.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Epithelial sodium channels (ENaC)¹ are the main pathway for sodium absorption and routinely co-localize with CFTR in the apical membrane of many polarized epithelial cells, including airway, gastrointestinal, and renal cells (1). Studies of the pathogenesis of cystic fibrosis in freshly isolated CF airway epithelium demonstrated that ENaC is hyperactivated in the absence of CFTR (2, 3), indicating that CFTR may down-regulate ENaC. This idea was supported from co-expression studies of CFTR and ENaC in Xenopus oocytes (4-6), fibroblast cells (7), and planar lipid bilayers (8-10).

Elimination of the inhibitory effect on ENaC by CFTR has been proposed to account for the observed hyperabsorption of Na⁺ in the airway and gastrointestinal epithelia of CF patients that lack functional CFTR. The simplest explanation would be a direct physical protein-protein interaction between CFTR and ENaC. Kunzelmann's group (4) identified a possible protein-protein interaction between CFTR and ENaC using the yeast two-hybrid analysis system. These authors concluded that CFTR physically interacted with the COOH terminus of -rENaC. However, Berdiev et al. (10) reported that CFTR could not down-regulate channels comprised only of -rENaC, and that the - and/or -ENaC subunits were required for CFTR's functional interaction with ENaC. Nonetheless, CFTR's ability to down-regulate wild-type -rENaC was lost if COOH-terminal-truncated -rENaC replaced wild-type -rENaC. In addition, Schreiber et al. (11) found that the first nucleotide-binding domain of CFTR was essential in the inhibition of the epithelial sodium conductance. However, few studies have examined the influence of ENaC on CFTR activity (12).

In the present study, truncation and missense mutations in the cytoplasmic NH₂- and COOH-terminal regions of -, -, and -rENaC subunits of rENaC were co-expressed with CFTR in Xenopus oocytes to identify domains within ENaC that are essential for the molecular interactions between ENaC and CFTR. Electrophysiological studies of macroscopic currents showed that CFTR activity is up-regulated by rENaC co-expression, and rENaC current is inhibited by both quiescent (i.e. non-conducting) and activated CFTR. Studies in which the cytosolic tails of rENaC subunits were truncated demonstrated that the carboxyl terminus of the -rENaC subunit was required for the functional interaction between activated CFTR and rENaC, while the amino termini of the  and  subunits were critical for the down-regulation of rENaC by quiescent CFTR. Therefore, at least two distinct mechanisms are involved in the intermolecular regulation of CFTR and rENaC.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

cRNA Preparation for Oocyte Injection and Confocal Microscopy-- The cDNA constructs of wt -, -, and -rENaC cloned from rat colon and the Liddle's mutants were the kind gifts from Drs. Cecilia Canessa and Bernard Rossier (13-16). Deletions of the cytoplasmic NH₂ termini of -, -, and -rENaC were made by PCR-based mutagenesis as described previously (17).

Enhanced green fluorescent protein (EGFP)-rENaC constructs were prepared as described before (18). In brief, and pEGFP- rENaC were constructed by excising the full-length subunits from pSport/ rENaC or pSport/ rENaC with SalI/KpnI, and ligating the excised fragment into SalI/KpnI digested pEGFP-Cl or C2 (CLONTECH, Palo Alto, CA). Two pre-existing stop codons in the  subunit were removed using a synthesized sense primer corresponding to nucleotides 1-20 and an antisense primer corresponding to nucleotides 514-543. PCR was performed, and the product was subcloned into pCR 2.1. The fusion cDNA EGFP rENaC was constructed by digesting the subcloned PCR product with SalI/BsmBI and ligating the gel purified 316-base pair PCR fragment into digested cDNA. EGFP--rENaC then was subcloned into pcDNA3.1 (InVitrogen) using NheI and KpnI. pcDNA3.1 was digested with NheI and treated with calf intestinal alkaline phosphatase to prevent self-ligation. The sequence of both strands was confirmed by ABI PRISM dye terminator cycle sequencing. cDNAs of full-length, truncated, and missense mutations of rENaC were transcribed in vitro by using T7 transcription kits (Ambion, Austin, TX), following the manufacturer's instructions.

Two Electrode Voltage Clamp of Oocytes-- The protocol for channel expression in oocytes was the same as described previously (19). Briefly, female Xenopus laevis, purchased from Xenopus Express (Burley Hill, FL), were anesthetized in 0.5% ice/tricaine (Sigma) solution. Ovary lobes were harvested through a small abdominal incision. Healthy oocytes at maturation stage V and VI were identified following 2-3 h rotating digestion with 3 mg/ml collagenase (Roche Molecular Biochemicals, Indianapolis, IN) in Ca²⁺-free OR-2 medium (in mM: 82.5 NaCl, 2.4 KCl, 1.8 MgCl₂, and 5 HEPES, pH 7.4). Oocytes were incubated in half-strength Leibovitz medium (L15) overnight before cRNA microinjection. Using a nano-microinjector (Drummond Science Co., Broomall, PA), 2.5 ng of cRNA of CFTR, or 12.5 ng of cRNA of each subunit of -rENaC or their combination in a volume of 50 nl, was injected into each oocyte.

Conventional two-electrode voltage clamp was used to record the macroscopic currents associated with rENaCs and CFTR. Oocytes were impaled with two microelectrodes filled with 3 M KCl. The tip resistance was 0.5-2.0 M. Oocytes were voltage clamped at a holding potential of 30 mV using pCLAMP 6.0.4 software (Axon Instruments, Burlingame, CA) through a TEA-200 voltage clamp amplifier (Dagan Corp., Minneapolis, MN). Two reference electrodes were connected to the bath by 3 M KCl, 3% agar bridges. Oocytes were continuously superfused at a rate of 3 ml/min in a small holding chamber with a volume of approximately 100 µl. The normal superfusate was Ringer's solution (in mM: 110 NaCl, 2.0 KCl, 0.2 CaCl₂, 1.0 MgCl₂, 5.0 HEPES, pH 7.4). Both inward and outward currents were monitored at holding potentials of 100 and +40 mV, respectively, at an interval of 10-30 s. The stepped I-V curves were collected when the rENaC-induced currents were stable. The perfusing system was driven by gravity, and the superfusates were switched from Ringer's solution to those of the desired composition. All experiments were conducted at room temperature (22 °C).

The same set of oocytes and the same batch of cRNAs were used to study the interaction of CFTR and rENaCs. Oocytes were transferred to the recording chamber 48 or 72 h post-cRNA injection, and superfused with Ringer's solution at room temperature. For oocytes only expressing rENaCs, rENaC current was calculated as the amiloride-sensitive Na⁺ current by subtracting the amiloride-resistant current from the total current measured in the absence of amiloride (10 µM). For oocytes only expressing CFTR, the differences between the currents recorded in the absence and in the presence of a cAMP-stimulating mixture (CKT: 0.2 mM isobutylmethylxanthine, 0.2 mM cAMP, 10 µM forskolin) were computed as CFTR-associated Cl currents. For oocytes co-expressing CFTR and rENaC, the protocol we used was first to add 10 µM amiloride to obtain the maximal rENaC current, followed by extensive washing to remove the amiloride. Then, oocytes were either treated with CKT or the combination of CKT and amiloride until the total current stabilized. To test the specificity of the interaction between CFTR and rENaCs, the endogenous Ca²⁺-activated Cl (CaCC) conductance of the oocyte was used as a control to see if wt -rENaC had any effect on its properties. A third microelectrode filled with 5 mM CaCl₂ was impaled into ENaC-expressing oocytes, and 25 nl (approximately 100 µM CaCl₂) was injected into the oocytes. Then the basal currents for CaCC or rENaC were recorded. Conversely, the pore-forming toxin palytoxin was employed as a substitute for rENaC to increase intracellular Na⁺ concentration and depolarize the resting membrane potential, to examine if these changes produced any effects on CFTR.

Immunofluorescence Assay-- The protocol for fluorescence image acquisition has been described previously (17, 20). Briefly, healthy oocytes were injected with one group of the following cRNAs: EGFP-CFTR; EGFP-CFTR + -rENaC; -rENaC + EGFP--rENaC + EGFP--rENaC; -rENaC + EGFP- rENaC + EGFP--rENaC + CFTR. The total injected cRNA of CFTR and -rENaC, the time for processing, and the application of CKT were identical to those for the voltage-clamp experiments. Water-injected oocytes were used as a control for background fluorescence.

In order to identify the plasma membrane, eggs were incubated for 1 h with EZ-Link Sulfo-NHS-LC-Biotin (10 mg/ml, Pierce, Rockford, IL). Labeling was carried out in ND-48 solution containing 10 µg/ml Texas Red-conjugated streptavidin (Molecular Probes, Eugene, OR) in the dark for 2 h at 4 °C with gentle agitation. The biotinylation reaction was stopped with glycine (100 mM). After washing in L-15 medium, images were acquired using a Leica DMIRBE TCSNT Laser Confocal Microscope (Leica, Germany) with a 10 × dry objective, and equipped with an acousto-optical turntable filter and three detector channels. The 488-nm argon laser line and the 568 nm krypton laser line excited the GFP and Texas Red, respectively. XY scans were obtained at a 1024 × 1024 resolution format at approximately the mid-section of each oocyte.

The distribution of EGFP-CFTR and EGFP-rENaC fluorescence present at the membrane surface was quantitated in randomly acquired, confocal xy sections at approximately the mid-section of each oocyte (18, 21). Using TSCNT software (Leica, Germany), a box (~500 µm²) was drawn over the area to be measured, and the mean pixel intensity (on a scale of 0-225) within the boxed region was determined. Twenty sections of each oocyte membrane were recorded and averaged for each cell. Between 5 and 10 oocytes per condition were used in the analysis. Using pixel intensity histograms, the EGFP fluorescence and means were evaluated to determine the relative expression of EGFP-CFTR or EGFP-rENaC before and after the addition of CKT.

In Vitro Transcription and Translation-- cDNAs were transcribed and translated in vitro using the TnT transcription/translation system from Promega, in the presence of canine pancreatic microsomes (22). The cDNAs encoding full-length, epitope-tagged -, -, and -rENaC subunits are described elsewhere (22, 23). CFTR in the pTM-1 vector was the generous gift of Dr. S. Cheng (Genzyme Corp., Cambridge, MA). The synthesized proteins were analyzed by SDS-polyacrylmide gel electrophoresis (PAGE) and autoradiography, or reconstituted into proteoliposomes and immunoprecipitated. To test for protein-protein interactions between different rENaC subunits and CFTR, the rENaC subunits and CFTR were translated either with radioactive methionine or with non-radioactive methionine. HA--rENaC and HA--rENaC epitope-tagged constructs were used. The same amount of each in vitro translated rENaC subunit immunopurified and reconstituted in proteoliposomes was then mixed with the same amount of in vitro translated CFTR immunopurified and reconstituted in proteoliposomes, and subjected to coprecipitation. Antibodies directed against nonlabeled protein were used, and the presence of co-precipitated, radioactively labeled protein was detected using autoradiography. The anti-HA and M2 antibodies were used in the following concentrations: 5 µg/ml anti-M2 antibody, 2 µg/ml anti-HA antibody. Rabbit anti-CFTR antibodies were raised against a fusion protein corresponding to the first nucleotide-binding domain of CFTR. All immunoprecipitations, SDS-PAGE, and Western blots were performed as described previously (22, 23).

Chemicals-- Leibovitz-15 powdered solution, heated inactivated horse serum, and gentamicin reagent solution were obtained from Life Technologies, Inc. (Grand Island, NY). One-half strength L-15 medium consisted of 50% concentration of L-15, 15 mM HEPES, 5% horse serum, and 1 mg/ml gentamycin, pH 7.5). Amiloride (RBI, Natick, MA) was prepared as a 5 mM stock solution in dimethyl sulfoxide/water (1:1, vol/vol). 8-Chlorophenylthio-cAMP (8-CPT-cAMP) was purchased from Roche Molecular Biochemicals, and stocked at 200 mM in water. Forskolin, isobutylmethylxanthine, and palytoxin were from Sigma and stored in dimethyl sulfoxide (10 and 500 mM, respectively) at 200 °C. All other reagents were from Sigma.

Statistics-- Current magnitudes were measured by using the Clampfit program (Axon Instruments, Burlingame, CA). The means of more than 100 data points at each voltage (episode) were acquired. The current data recorded from different oocytes were averaged and presented as mean ± S.E., where n stands for the number of oocytes. Significant statistical differences between each group were taken as p < 0.05, tested by one-way ANOVA or paired Student's t test for parallel groups, or from data obtained from identical oocytes.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Expression of rENaC and/or CFTR, and Their Interactions-- More than 90% of the inward current recorded in oocytes expressing -rENaC was inhibited by 10 µM amiloride, consistent with the high amiloride sensitivity of -rENaC (Fig. 1A). The slightly inward rectified I-V curve and a reversal potential of approximately +10 mV indicated that the amiloride-sensitive current was carried by sodium via exogenously expressed -rENaC (Fig. 1B). The depolarized resting membrane potential in the oocytes expressing wt -rENaC resulted from the accumulated intracellular Na⁺ loading that occurred during the interval between RNA injection and current recording, and was +8.35 ± 1.05 mV (n = 78), significantly greater than that of control oocytes injected with water (42.5 ± 0.8 mV, n = 45, p < 0.001). In oocytes also expressing CFTR, CFTR current could be elicited by application of a cAMP-elevating mixture (CKT), and had a reversal potential of 25.0 ± 0.9 mV (n = 25; Fig. 1B). This resting membrane potential was close to that of oocytes injected with water (37.9 ± 4.1 mV, n = 30, p > 0.05). The sensitivity of CFTR to CKT, the measured reversal potential, and the inability of these currents to be inhibited by 100 µM DIDS (not shown), are all features of the CFTR channel (see Ref. 24).

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Co-expression of CFTR and -rENaC affect sodium and chloride currents. A, CKT inhibits the amiloride-sensitive Na currents in oocytes co-expressing CFTR and ENaC (+40 mV, closed circles; and 100 mV, closed squares). B, mean I-V relationships of amiloride-sensitive rENaC currents in (n = 22) and CKT-activated CFTR currents in oocytes co-expressing rENaC and CFTR (n = 25). The reversal potentials of rENaC and CFTR are +10 and 25 mV, respectively. C, mean rENaC currents are down-regulated in oocytes co-expressing CFTR and rENaC. CKT decreases ENaC current. D, ENaC enhances the CFTR current (Vm of 100 mV).

The cross-talk between currents of CFTR and -rENaC was assessed from identical oocytes co-expressing CFTR and -rENaC, and compared with a control group expressing CFTR alone. The amiloride-sensitive Na⁺ current was detected before and during activation of CFTR with CKT, as shown in Fig. 1A. Fig. 1C shows that rENaC current was reduced when CFTR was present, and the significant inhibition of -rENaC by CFTR could be divided into two parts. First, non-activated CFTR inhibited ENaC current by 54% (Fig. 1C). This current was then further reduced an additional 83% by CFTR activated by CKT (Fig. 1C). Conversely, there was a 6-fold increment in the CFTR-mediated Cl currents in oocytes co-expressing CFTR and -rENaC (Fig. 1D). The magnitude of CFTR current in oocytes co-expressing CFTR and -rENaC was 8654.2 ± 890.2 nA (n = 84), as compared with 1405.1 ± 358.2 nA in oocytes expressing CFTR alone, (n = 81).

To test the specificity of the stimulatory effect of rENaC on CFTR, the effect of rENaC on an endogenous oocyte CaCC was examined. As shown in Fig. 2B, this endogenous CaCC, activated by cytosolic injection of 100 µM Ca²⁺ (Fig. 2A), was almost completely suppressed in oocytes expressing rENaC (Fig. 2, B-D). The outwardly rectified I-V relationship, with a reversal potential of about 20 mV (Fig. 2C), was identical to that of the endogenous oocyte CaCC documented previously (25). Because ENaC co-expression had opposite effects on CaCC as compared with CFTR, it is unlikely that the up-regulation of CFTR by rENaC could be due to nonspecific effects. CKT had no effect on the amiloride-sensitive current (Fig. 2E). This result is different than that reported for rENaC following transfection into mammalian cells (7, 26, 27), but is consistent with our (8, 28) and others (4, 29) observations in oocytes.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   rENaC inhibits CaCC current. A, endogenous Ca2+-activated Cl current (CaCC) in Xenopus oocytes is activated by 100 µM Ca2+, injection (in 25 nl) into the cytoplasm. B, -rENaC eliminated activation of endogenous CaCC to increased cytosolic Ca2+. C, I-V relationships show that the outwardly rectified CaCC conductance (closed circles) has a reversal potential of approximately 20 mV. D, CaCC current at 100 mV. **, p < 0.001. E, rENaC current in the presence or absence of CKT at a holding potential of 100 mV.

Lack of Effect of Palytoxin on CFTR and rENaC Currents-- Overexpression of rENaC in oocytes produces an increase in cytoplasmic sodium and a depolarization of the resting membrane potential (14, 16, 25). Thus, we tested the possibility that an increase in intracellular Na⁺ concentration with a concomitant depolarization of the membrane potential could up-regulate CFTR currents. Oocytes expressing CFTR alone were treated with 2 nM palytoxin, a venom that mimics rENaC expression by acting as a Na⁺ ionophore and inhibitor of the Na⁺/K⁺-ATPase (30). The Na⁺ current induced by palytoxin was activated gradually (Fig. 3A), and the resting membrane potential depolarized to 1.4 ± 0.6 from 27.2 ± 5.9 mV (n = 8). In oocytes expressing CFTR (Fig. 3B), palytoxin did not significantly affect CFTR current (Fig. 3B). Similar results were obtained following injection of 100 mM Na⁺ into CFTR-expressing oocytes (data not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Changes in intracellular sodium and membrane voltage do not stimulate CFTR currents. A, palytoxin (PTX) enhanced Na+ current and depolarized the resting membrane potential (data not shown) in oocytes expressing CFTR, but had no effect on CKT-stimulated CFTR. B, PTX had no effect on CFTR chloride current (Vm = 100 mV).

Is Cl Conductance Required for CFTR Down-regulation of ENaC?-- We next tested the hypothesis that CFTR can more effectively down-regulate ENaC when conducting Cl. This was investigated by means of Cl depletion (oocytes incubated >l h with Cl-free Ringer's solution as the superfusate). As shown in Fig. 4, CFTR inhibited the rENaC current. However, CFTR current was not inhibited further by CKT in Cl depleted oocytes. Therefore, we conclude that Cl does not have to be transported by CFTR in order for CFTR to down-regulate ENaC.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Cl is required for CKT-activated CFTR to inhibit ENaC. A, the sodium current trace in an oocyte expressing -rENaC and CFTR after Cl depletion (incubated with Cl-free Ringer's solution 1 h at room temperature prior to recording) perfused with Cl-free Ringer's medium. B, mean rENaC currents at 100 mV in Cl-depleted oocytes with Cl-free superfusate. *, p < 0.05 compared with control.

Co-expression of Truncated and Missense Mutations of rENaC with CFTR-- Although down-regulation of ENaC by CFTR was first reported more than 10 years ago (see Refs. 1 and 31), the mechanisms underlying this phenomenon are unknown. Yeast two-hybrid analysis revealed that there was a physical protein-protein interaction between CFTR and the COOH terminus of -rENaC (4), but it has not been confirmed in vivo (11) or functionally tested. Additionally, down-regulation of rENaC by CFTR was observed in planar lipid bilayers, a cell-free model system utilizing purified CFTR and rENaC proteins, implying that cross-regulation of CFTR and rENaC was due to protein-protein interaction (8, 9).

The intracellular tails of rENaC subunits contain consensus sequences for phosphorylation (32), endocytosis (17), a PY motif for Nedd4-regulated endocytosis (33), and sites for interaction with other proteins such as cytoskeletal elements (34) and syntaxin (35, 36). Similarily, most of the regulatory elements of CFTR are located in the cytoplasmic regions, i.e. the regulatory or R domain for phosphorylation and gating (37), the nucleotide-binding domain (11), and the NH₂ terminus for syntaxin interactions (38, 39). Therefore, to identify the amino acid regions of the cytoplasmic termini of -, -, and -rENaC subunits critical for the intermolecular interaction of CFTR and rENaC, truncation mutants missing the cytosolic carboxyl or NH₂ termini, and missense mutations in the cytosolic carboxyl termini, were co-expressed with wt CFTR in oocytes and the resulting Na⁺ and Cl macroscopic currents examined. If the up-regulation of CFTR by ENaC was due to a physical interaction between the carboxyl terminus of the -rENaC subunit and CFTR (4), this up-regulation should be abolished by the co-expression of the carboxyl terminus truncation of -rENaC subunit and CFTR. However, truncation of the COOH-terminal of -rENaC (_R613X) did not prevent ENaC up-regulation of CFTR (Fig. 5). Likewise, a series of NH₂-terminal truncations of -rENaC did not abolish ENaCs stimulatory effect on CFTR, but the magnitude of CFTR up-regulation was significantly attenuated (Fig. 5). NH₂-terminal truncation of -rENaC, or NH₂- or COOH-terminal truncation of -rENaC had no influence on ENaCs ability to up-regulate CFTR (Fig. 5). Only COOH-terminal truncation of -rENaC resulted in an ENaC that failed to up-regulate CFTR activity (Fig. 5).

View larger version (31K): [in this window] [in a new window]   Fig. 5.   CFTR currents in oocytes co-expressing truncated and missense rENaC mutants with CFTR. The normalized CFTR currents with or without co-expression of -rENaC and its truncated and missense mutants. *, p < 0.05; **, p < 0.001

Down-regulation of rENaC by CFTR was also evaluated by measuring and comparing the amiloride-sensitive Na⁺ currents in oocytes co-expressing different ENaC mutants and wt CFTR. Fig. 6A shows that, as for the wild-type ENaC channel, macroscopic Na⁺ currents produced by expression of _2-67-rENaC, _2-109--rENaC, and _R613X-ENaC were inhibited by both quiescent and activated CFTR, although in the case of _R613X and _2-109, the magnitude of the rENaC current was less, perhaps due to inefficient channel processing (Fig. 6A). Trucation of the -rENaC subunit (_R564X) had no effect on the rENaC current with or without CKT activation (Fig. 6B). However, in _Y618A-rENaC-expressing oocytes, CFTR effectively down-regulated the ENaC current. In contrast, the -rENaC NH₂ terminus truncation mutant (_2-53-rENaC) was stimulated by CFTR co-expression, but was still down-regulated by activated CFTR (Fig. 6C), whereas _R564X and _Y628A responded to CFTR and CFTR + mixture similarly as did their wild-type counterparts.

View larger version (26K): [in this window] [in a new window]   Fig. 6.   ENaC Na+ currents in oocytes expressing CFTR and mutant ENaC. Co-expression of CFTR inhibits wt and mutants of -rENaC (A). *, p < 0.05, but CKT application does not affect the current associated with the COOH-terminal truncated -rENaC (B). Reversely, CFTR co-expression up-regulates the NH2-terminal truncated -rENaC (C).

Distribution of rENaC and CFTR-- Because cAMP regulates CFTR trafficking to the membrane surface (5, 40, 41), and because Nedd4-regulated endocytosis occurred through a PY-motif located in the carboxyl terminus of the -rENaC (33), it is possible that down-regulation of rENaC by CFTR or up-regulation of CFTR by ENaC was due to alterations or re-distribution of these channel proteins between the plasma membrane and an intracellular pool. GFP, as a reporter protein, was tagged directly onto the - and -rENaC subunits, or onto CFTR directly, and visualized by confocal microscopy (Fig. 7). GFP labeling of either the - and -rENaC subunits or CFTR had no effect on the absolute magnitude of either I_Na or I_Cl (data not shown). In oocytes expressing GFP--rENaC, bright GFP fluorescence was seen along the plasma membrane (Fig. 7, middle panel, rows A-D). There was no visual or measured difference in the GFP brightness between oocytes expressing rENaC alone (rows A and B) and those co-expressing untagged CFTR (rows C and D). Elevation of cytosolic cAMP level by mixture did not change the surface expression of rENaC-GFP (row D). These results are summarized in Table I. No green fluorescence was observed at the oocyte surface or under the plasma membrane in water-injected oocytes (data not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Representative confocal images of oocytes injected with EGFP-labeled rENaC only (rows A and B) or EGFP-labeled rENaC plus CFTR (rows C and D). Proteins in the plasma membrane were labeled with biotin, which was detected with streptavidin-conjugated Texas Red (images in the left columns of all rows). GFP-labeled rENaC is shown in the middle panels, and the overlaid images in the right column show the co-location of GFP-rENaC. The images in rows B and D were taken 5 min after exposing the oocytes to CKT. The bar length represents 40 µm.

Although the mixture-induced activation of CFTR in cultured mammalian epithelia was not due to detectable re-localization of channel proteins from the cytosol to the surface (18), we could not rule out that the up-regulation of CFTR by co-expressed rENaC may have resulted from an increment in surface CFTR expression, as has been reported in other systems including oocytes (5, 40, 41). Surface expression of GFP-CFTR both in oocytes expressing CFTR alone (Fig. 8, rows A and B) and co-expressing rENaC (Fig. 8, rows C and D) was enhanced by 25-50% following addition of the cAMP-elevating mixture (n = 5 and 10 for CFTR and CFTR + ENaC expressing oocytes, respectively). Compared with oocytes injected with GFP-CFTR alone (Fig. 8, row A), there was a 56 ± 15% increase in GFP-CFTR surface fluorescence in oocytes co-expressing untagged rENaC (n = 9; Fig. 8, row C). Addition of mixture to CFTR/ENaC expressing oocytes again produced an additional 50% increase in CFTR surface fluorescence (n = 10; Fig. 8, row D). Thus, while cAMP did not affect apical surface fluorescence of ENaC, it did for CFTR (see Table I). Also, there was significantly more CFTR present at the apical surface of oocytes co-expressing ENaC compared with CFTR-expressing oocytes.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Representative confocal images of oocytes injected with GFP-labeled CFTR without (rows A and B) or with (rows C and D) rENaC. The identical series of images as indicated in the legend to Fig. 7 were generated. Images presented in rows B and D were acquired 5 min post-treatment with CKT. Experiment was repeated four times with similar results.

Protein-Protein Interaction of CFTR and rENaC-- The co-localization of CFTR and rENaC (Figs. 7 and 8) and the functional identification of the carboxyl terminus of the -ENaC subunit and the amino terminus of -ENaC subunit as potential domains of ENaC-CFTR interaction, led us to investigate the hypothesis that the functional cross-talk between these two channel proteins results from physical interaction of CFTR and rENaC. In the first series of experiments, wt -rENaC, containing an HA tag, or wt CFTR were in vitro translated separately, the reaction products individually reconstituted into proteoliposomes, and immunoprecipitated with HA or CFTR antibodies (Fig. 9). When an HA antibody, or an anti-CFTR antibody, was used to immunoprecipitate radioactively labeled -rENaC from the solubulized proteoliposomes, only antibodies that recognized HA-tagged ENaC precipitated -rENaC (see left side of Fig. 9). The first lane on each side of the figure shows an autoradiograph of the in vitro translated products of either -rENaC (left side) or CFTR (right side). Thus, immunoprecipitating with an anti-CFTR antibody did not immunoprecipitate -rENaC. Conversely, in vitro translated CFTR could only be immunoprecipitated by anti-CFTR antibodies, not by the anti-HA antibody. These results demonstrate the specificity of the antibody probes.

View larger version (56K): [in this window] [in a new window]   Fig. 9.   Analysis of anti--rENaC (HA) and anti-CFTR antibodies for cross-reactivity. In order to use anti-CFTR and anti-HA antibodies for co-immunoprecipitation, we first tested if these antibodies were cross-reactive. The -rENaC subunits containing the HA tag were used in these experiments. Panel A, in vitro translated -rENaC was subjected to immunoprecipitation using anti-HA or anti-CFTR antibodies. Only anti-HA antibodies precipitated -rENaC, demonstrating non-cross-reactivity of anti-CFTR antibodies with -rENaC subunit. Panel B, in vitro translated CFTR was subjected to immunoprecipitation using anti-CFTR or anti-HA antibodies. Only anti-CFTR antibodies precipitated CFTR.

We next prepared proteoliposomes containing in vitro translated radioactively labeled or non-labeled -rENaC and CFTR, and incubated them overnight prior to coimmunoprecipitation. When anti-CFTR antibodies were used for immunoprecipitation, -rENaC was detected (middle lane of left side of Fig. 10). Likewise, when the anti-HA antibody was used, CFTR in vitro translated products were brought down (middle lane of right side of Fig. 10). Secondary IgG antibodies did not immunoprecipitate any protein from the proteoliposome mixture. These results show that -rENaC and CFTR interact in a complex within proteoliposomes made from in vitro translated protein.

View larger version (54K): [in this window] [in a new window]   Fig. 10.   Analysis of interaction between -rENaC and CFTR using coprecipitation. Panel A, the -rENaC subunit was translated in vitro in the presence of radioactive methionine (Lane 1) and mixed with in vitro translated, nonradioactive CFTR. Precipitation with anti-CFTR antibodies revealed co-precipitation of -rENaC subunit. Precipitation with non-immune IgG was used as a negative control. Panel B, CFTR was translated in vitro in the presence of radioactive methionine (Lane 1) and mixed with in vitro translated nonradioactive -rENaC (HA-tagged). Precipitation from this mixture with non-immune IgG was used as a negative control. Precipitation with anti--rENaC antibodies (HA) revealed co-precipitation of CFTR.

Following the same protocol, we performed co-immunoprecipitation experiments from combinations of different in vitro translated rENaC subunits and truncated constructs and CFTR. The results of these immunoprecipitation experiments are summarized in Table II. Wild type - and -rENaC can be co-immunoprecipitated with CFTR, whereas the wild type -rENaC subunit cannot. A COOH-terminal truncated -rENaC or -rENaC can also be co-immunoprecipitated with CFTR. However, NH₂-terminal deletions of -rENaC or -rENaC cannot. These results suggest that the NH₂ termini of the - and -rENaC subunits are sites of contact for protein-protein interaction between rENaC and CFTR.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The hyperabsorption of sodium in native CF airway and intestinal epithelia and in cultured renal A6 cells is closely paralleled by the lack of functional CFTR (1, 42, 43). Evidence obtained from heterologous systems, amiloride inhibitable transepithelial electrical potentials, and in short circuit current measurements is consistent with the interpretation that CFTR can down-regulate ENaC. However, the mechanisms underlying the down-regulation of ENaC by CFTR remain unknown. Although noted by some laboratories (5, 12), activation by ENaC of CFTR currents have largely been neglected. Accordingly, human CFTR and rENaC were co-expressed in Xenopus oocytes to determine the functional relationships between them. Inhibition of rENaC was produced not only by activated CFTR, consistent with the observations of other groups (5, 6, 11, 29, 44), but also rENaC reversibly facilitated cAMP-dependent CFTR activation.

The major finding reported in this work is that there is a coordinate influence of CFTR on ENaC and rENaC on CFTR when heterologously expressed in Xenopus oocytes. While CFTR can down-regulate ENaC activity, ENaC can stimulate the activity of CFTR up to 6-fold. We show that this up-regulation of CFTR by ENaC is produced in part by an increase in surface density of CFTR. We found that the carboxyl terminus of the -rENaC subunit was involved in both the down-regulation of ENaC by CFTR and the up-regulation of CFTR by ENaC. However, both the carboxyl terminus of -rENaC and the amino terminus of r-rENaC were essential for CFTR to down-regulate ENaC. Co-immunoprecipitation experiments revealed a biochemical interaction between ENaC and CFTR, but no clear correlation between the regions of biochemical interaction and functional interaction emerged.

The proposed intracellular regulatory fragments of CFTR are in the first nucleotide-binding domain (NBD1) (11). Mutation of NBD1 abolishes the inhibitory effects of CFTR on ENaC co-expressed in oocytes. We therefore performed mutagenesis studies on rENaC to locate the subunit(s) that might mediate this effect. The electrophysiological results revealed that the cytoplasmic carboxyl terminus of the  subunit is closely linked to the intermolecular regulation between rENaC and activated CFTR. Furthermore, the corresponding subunit for the interaction of non-activated CFTR and rENaC might be mediated by the cytoplasmic NH₂ terminus of  subunit and the carboxyl terminus of  subunit. Additionally, the observation that rENaC down-regulated the activity of an endogenous CaCC of the oocyte confirmed the specificity of the interaction between ENaC and CFTR. The loss of the down-regulation of rENaC by cAMP-activated CFTR following intraoocyte Cl depletion implied a requirement for Cl. The lack of response of rENaC expressed in oocytes to cAMP was different from that observed in transfected mammalian cells (7, 26).

The observation that there was little or no inhibition of CFTR with increasing [Na⁺]_i and decreasing V_m with palytoxin demonstrated a lack of effect of [Na⁺]_i and V_m on CFTR activity. Combined with the results of intracellular Na⁺ injection, we conclude that the up-regulation of CFTR by ENaC was not due to increased cytoplasmic Na⁺ or a depolarized membrane potential. The equivalent stimulation of CFTR by the Liddle's mutants (_Y618A, _Y628A, and _R574X), which produce a constitutively activated Na⁺ conductance, and the NH₂-terminal -mutant (_2-109), which has undetectable current, also support the conclusion that elevated [Na⁺]_i and depolarization failed to activate CFTR. Even though there was a greater than 2-fold enhancement of surface fluorescence of GFP-labeled CFTR in oocytes co-expressing ENaC (Fig. 8), and if this could be translated into a 2-3-fold increase in number of CFTR channels at the surface, this is not enough to account for the greater than 4-fold increase in macroscopic current measured under these conditions (Fig. 10). Moreover, the lack of a correlation between mRNA, conductance of CFTR, and severity of disease (44) supports the hypothesis that the ENaC influences CFTR in ways other than simply increasing CFTR number. This model is not unique and has been used to interpret the constitutively activated Na⁺ conductance in Liddle's syndrome (46, 47). Electrophysiological recording further suggested that CFTR channel number per patch and the mean open time were increased in oocytes co-expressing rENaC (48). The potential pathophysiololgical relevance of ENaC's ability to activate CFTR remains to be determined in vivo.

Functional coupling of inhibition of rENaC with stimulation of CFTR suggests that cAMP-dependent protein kinase phosphorylation could be involved in the interaction between two channels. The effects of cAMP (26, 32), and actin (49) on ENaC and CFTR predicts that the functional activation of CFTR by phosphorylation is critical for the secondary inactivation of ENaC by activated CFTR. Whether or not this cAMP-regulated system requires the presence of associated proteins binding to both channels is unknown. Associated proteins with possible regulatory functions include the syntaxin family (35, 36), G proteins (8), cAMP-response element-binding protein (50), and NHERF (51). It is also conceivable that protein kinase C, which regulates both CFTR and ENaC, might be involved in the cross-talk because protein kinase C regulatory sites are located in the carboxyl terminus of -ENaC and the intracellular R domain of CFTR (52). It is presently unclear whether or not the up-regulation of CFTR by ENaC controlled via the same signaling pathway.

Because there is a PY domain in the carboxyl terminus of each ENaC subunit, it is possible that Nedd4-regulated internalization (33) could contribute to the down-regulation of ENaC activity by CFTR. But an increment in the surface channel density of ENaC was not evident in oocytes not co-expressing CFTR (Fig. 7). Moreover, these confocal imaging experiments are consistent with the earlier patch-clamp experiments of Stutts et al. (26) who reported that co-expression of CFTR with ENaC in mouse fibroblasts did not change the average channel number of ENaC observed per patch. Therefore, it is unlikely that CFTR influences ENaC activity by modulating endocytosis through Nedd4. Overexpression of CFTR would bring about multiple alterations in oocytes (for example, endogenous K channel activation), and further experiments need to be undertaken to explore the mechanisms of CFTR and ENaC interaction.

Co-immunoprecipitation experiments showed that COOH-terminal deleted -rENaC could interact with CFTR, and that -rENaC, when expressed without its - and -rENaC partners, could not (Table II). The lack of concordance between the functional and biochemical studies concerning the subunit domains of ENaC involved in the interactions between CFTR and ENaC reveal the complexity of these interactions. This complexity is underscored by the observation that in human sweat ducts, ENaC activity increases rather than decreases in the presence of CFTR (53). While the functional data suggest that the carboxyl terminus of the -ENaC subunit and amino terminus of the -ENaC subunit are potential domains of ENaC-CFTR interaction, co-immunoprecipitation results suggest that NH₂ termini of - and -rENaC are involved in this interaction as well. However, there are several differences in experimental conditions for functional expression in oocytes and for in vitro coprecipitation. Oocyte expression was performed expressing all three rENaC subunits and CFTR simultaneously, permitting oligomerization of rENaC subunits (see Refs. 54 and 55). Co-precipitation experiments tested the interaction between each individual ENaC subunit and CFTR separately. Immunoprecipitation experiments in oocytes proved to be difficult, even using epitope-tagged ENaC constructs, undoubtedly due to relatively low expression levels. Moreover, functional studies in oocytes can include other intermediate proteins in ENaC-CFTR interaction, such as those interacting with PDZ domains or cytoskeletal elements (56-58). Thus, our results suggest that there may be "tethering" domains that hold the channels together so that other "regulatory" domains can interact.

In conclusion, co-expression of CFTR and rENaC in Xenopus oocytes, and the up-regulation of CFTR activity was associated closely with a decrease in parallel in ENaC conductance. The cytoplasmic carboxyl terminus of the -ENaC subunit appears to be required for the functional cross-talk between ENaC and CFTR, and the cytosolic amino termini of both the - and -ENaC subunits are essential for the down-regulation of ENaC by CFTR co-expression.

	ACKNOWLEDGEMENTS

We thank Dr. Albert Tousson (Cell Biology, University of Alabama) and Dr. Anne Lynn Langloh for their kind help in using confocal microscope, and Dr. Pierre-Jean Ripoll for making the HA-tagged ENaC subunit constructs. We also thank Dr. Thomas R. Kleyman for providing access to his unpublished work. We thank Cathy Guy and Elaine Dean for providing excellent administrative support. We greatly appreciate the discussions with Dr. Ernest M. Wright (Physiology, UCLA) and Jian Fu (Physiology, University of Alabama).

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants DK45881, DK51067, DK53090 and the Cystic Fibrosis Foundation.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Dept. of Physiology and Biophysics, University of Alabama at Birmingham, 1918 University Blvd., MCLM 704, Birmingham, AL 35294-0005. Tel.: 205-934-6220; Fax: 205-934-1445; E-mail:Benos@physiology.uab.edu.

Published, JBC Papers in Press, May 19, 2000, DOI 10.1074/jbc.M002848200

	ABBREVIATIONS

The abbreviations used are: ENaC, epithelial sodium channel; CF, cystic fibrosis; CFTR, cystic fibrosis transmembrane conductance regulator; CKT, cAMP-stimulating mixture; CaCC, calcium-activated chloride channel; GFP, green fluorescent protein; DIDS, 4,4'-diisothiocyanostilbene-2,2'-disulfonic acid; PCR, polymerase chain reaction; HA, hemagglutinin; PAGE, polyacrylamide gel electrophoresis.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES