Interregulation of Proton-gated Na+ Channel 3 and Cystic Fibrosis Transmembrane Conductance Regulator*

Proton-gated Na+ channels (ASIC) are new members of the epithelial sodium channel/degenerin gene family. ASIC3 mRNA has been detected in the homogenate of pulmonary tissues. However, whether ASIC3 is expressed in the apical membranes of lung epithelial cells and whether it regulates cystic fibrosis transmembrane conductance regulator (CFTR) function are not known at the present time. Using reverse transcription-PCR, we found that the ASIC3 mRNA was expressed in the human airway mucosal gland (Calu-3) and human airway epithelial (16HBE14o) cells. Indirect immunofluorescence microscopy revealed that ASIC3 was co-segregated with CFTR in the apical membranes of Calu-3 cells. Proton-gated, amiloride-sensitive short circuit Na+ currents were recorded across Calu-3 monolayers mounted in an Ussing chamber. In whole-cell patch clamp studies, activation of CFTR channels with cAMP reduced proton-gated Na+ current in Calu-3 cells from -154 ± 28 to -33 ± 16 pA (n = 5, p < 0.05) at -100 mV. On the other hand, cAMP-activated CFTR activity was significantly inhibited following constitutive activation of putative ASIC3 at pH 6.0. Immunoassays showed that both ASIC3 and CFTR proteins were expressed and co-immunoprecipitated mutually in Calu-3 cells. Similar results were obtained in human embryonic kidney 293T cells following transient co-transfection of ASIC3 and CFTR. Our results indicate that putative CFTR and ASIC3 channels functionally interact with each other, possibly via an intermolecular association. Because acidic luminal fluid in the cystic fibrosis airway and lung tends to stimulate ASIC3 channel expression and activity, the interaction of ASIC3 and CFTR may contribute to defective salt and fluid transepithelial transport in the cystic fibrotic pulmonary system.

Increased reabsorption of Na ϩ ions via apically located Na ϩ channels is a major pathophysiological feature in cystic fibrosis (CF) 2 epithelia (1,2). The molecular basis of the hyperactive amiloride-sensitive Na ϩ channels is not clear. Several cation channels, including epithelial Na ϩ channels (ENaC), cyclic nucleotide-gated Na ϩ channels, non-selective cation channels, and acid-sensitive ion channels (ASIC), are detected in lung epithelia. Amiloride inhibits these apically located Na ϩ channels with varying efficacies. CFTR down-regulates amiloridesensitive ␣␤␥ ENaC activity in Xenopus oocytes (3)(4)(5)(6)(7)(8)(9)(10)(11). These results are in agreement with observations from normal and CF tissues (12)(13)(14)(15)(16)(17). On the other hand, ENaC up-regulates the CFTR current by up to 6-fold, and this process is not Na ϩ -dependent (7,8,18,19). Mutagenesis studies have identified the first nucleotide binding domain of CFTR and Cl Ϫ transport as being required for CFTR-mediated down-regulation of ␣␤␥ ENaC (3,11), whereas the cytosolic C termini of ␤ and ␥ ENaC subunits are involved in the intermolecular regulation between CFTR and ␣␤␥ ENaC (7,8). In contrast to the extensive studies of the ENaC channels and their interactions with CFTR in the last three decades, little is known about the contribution of other apical cation channels of normal and CF epithelia on CFTR function.
A new branch of the ENaC/degenerin family, the protongated Na ϩ channels (ASIC1a, -1b, -2a, -2b, -3, and -4), has been cloned from neural tissue (20 -22). ASIC3 activity and expression are regulated by low extracellular pH, divalent cations, inflammation, and ischemia, all of which are commonly featured in CF lungs (20,22,23). Distribution of ASIC and ENaC overlaps in both epithelial and neural tissue. For example, in the human testis, expression of CFTR, ASIC3, and all four ENaC subunits have been detected (24 -26). Analysis of CF fluid collected from epithelial organs (e.g. airway, pancreas, intestine, and testis) confirmed that CF epithelial secretions, in addition to being thicker and having less volume, have an abnormal ionic composition and an acidic pH (27)(28)(29)(30)(31)(32). The condensate from the exhaled breath of CF patients is also more acidic (pH 5.3 versus 6.1) (33). Although ASIC3 mRNA has been detected in a number of epithelial organs (34 -36), its cellular expression and physiological role in the lung have not been studied.
To test the hypothesis that ASIC3 mRNA and proteins are expressed in the apical membrane of pulmonary epithelial cells and contribute to elevated Na ϩ channel activity in CF epithelia, we studied the expression of ASIC3 channels and possible interaction with CFTR in Calu-3 cells. Our results confirm that ASIC3 and ASIC2 mRNAs were expressed in Calu-3 cells, human airway, pancreatic, and colonic epithelial cells. ASIC3 and CFTR co-segregated in the apical membrane of polarized Calu-3 monolayers. Co-immunoprecipitation of native and heterologously expressed CFTR and ASIC3 proteins suggests that these two proteins may interregulate each other via a physical association. ASIC3 channels functionally contribute to transepithelial Na ϩ transport, which is down-regulated by CFTR in Calu-3 cells.

EXPERIMENTAL PROCEDURES
Reverse Transcription (RT)-PCR-Total RNA was isolated from Calu-3 (submucosal cells), 16HBE14o (bronchial cells), CFPAC (pancreatic cells), and T84 (colonic cells) using TRIzol reagent (Invitrogen) according to the manufacturer's instructions. RNA concentration was assessed by spectrophotometry at 260 nm. RNA quality was confirmed by running 5 g of total RNA from each preparation on a 1.2% denatured agarose gel. The first strand cDNAs were synthesized using the Superscript TM first strand synthesis system for RT-PCR kit (Invitrogen). For RT-PCR analysis of ASIC3 expression, the forward and reverse primers were 5Ј-TAT GAG ACC GTG GAG CAG-3Ј and 5Ј-TGT GTG ACA AGG TAG CAG G-3Ј, which correspond to nucleotides 1279 -1296 and 1610 -1592, respectively, of ASIC3 (GenBank TM accession number AF057711) (37). PCR was performed in a Mastercycler gradient thermocycler (Eppendorf Scientific). Two microliters of the first strand product was used as a template in each 50-l PCR reaction, and PCR conditions were optimized using the PCR Optimizer TM kit (Invitrogen). The cycling parameters consisted of one cycle of 95°C for 3 min and then 30 cycles of 95°C for 0.5 min, 58°C for 0.5 min, and 72°C for 1.5 min followed by a single 10-min cycle at 72°C for extension. RT-PCR products were electrophoresed on a 2% agarose gel using 100-bp PCR markers (Promega) as a standard to determine the molecular size. Products of the predicted molecular size were excised and purified from the gel using the QIAquick gel extraction kit (Qiagen) and subcloned into the pCR-2.1 vector (Invitrogen). Positive clones were selected by blue/white screening and followed by digestion with EcoRI (Promega) to verify incorporation of the insert of the correct size. PCR products in the pCR2.1 vector were verified by automated DNA sequencing (DNA Sequencing Facility, Iowa State University).
Immunofluorescent Staining-Calu-3 cells were grown as polarized monolayers on 12-mm tissue culture inserts (Costar). Three percent formaldehyde-fixed cells were stained by incubating with the primary antibodies (rabbit polyclonal anti-ASIC3 antibody from Neuromics diluted 1:50 in phosphatebuffered saline with 5% bovine serum albumin (blocking buffer) and mouse monoclonal anti-CFTR antibody from Chemicon 1:100 in blocking buffer) for 2 h at room temperature. The samples were rinsed in phosphate-buffered saline following blocking buffer incubation and exposed to the secondary antibodies (1:1000; Molecular Probes, Eugene, OR) for 2 h at room temperature. The samples were mounted and imaged using an Olympus 1 ϫ 70 epifluorescence microscope. Negative controls consisted of substituting non-immune IgG for the primary antibody. For imaging side views of the Calu-3 monolayers, the filters were folded sharply, and the cells at the folded edge were photographed.
Oocyte Expression and Voltage Clamp Assay-cRNA for ASIC3 and CFTR were prepared as previously described (8).
Defolliculated oocytes were cytosolically injected with a 1:1 ratio of ASIC3 and CFTR cRNA (25 ng of cRNA in 50 nl of RNase-free water/oocyte) and incubated in half-strength L-15 medium. Whole-cell ASIC3 and CFTR currents were measured by the two-electrode voltage clamp technique 2 days after injection (8).
Deparaffinized human lung tissue sections were provided by University of Alabama at Birmingham Human Tissue Procurement. Sections from both CF patients and healthy controls were incubated with anti-ASIC3 antibody and fluorescein isothiocyanate-conjugated IgG using similar protocols as described above for Calu-3 monolayers.
Delivery of superfusates at various pH values were controlled by the SF-77B perfusion fast step system (Warner) and CLAMPEX version 7.1. Solutions from the square perfusion tubing with a size of 1 ϫ 1 mm covered the clamped cell completely. Approximately 1 ms was measured to switch perfusate tubing from one to another in a chamber of 0.1 ml.
Ussing Chamber Studies-Calu-3 monolayers were mounted into Ussing chambers (Jim's Instrument Manufacturing, Inc., Iowa City, IA) and bathed with solution consisting of (in mM): Na ϩ 145, K ϩ 5, Ca 2ϩ 1.2, Mg 2ϩ 1.2, gluconic acid 125, HEPES 10, glucose 10 (pH 7.4). Bath solution in the apical chamber contained 10 mM mannitol in place of glucose to minimize the contribution of the Na ϩ -glucose co-transporter to the Na ϩ transport. Bath solutions were continuously bubbled with 100% N 2 , combined with both HPO 4 Ϫ -free and HCO 3 Ϫ -free medium to eliminate pH buffering and the activity of the HCO 3 Ϫ -dependent transporter/exchangers (pH 7.4). Bath solutions were maintained at 37°C. The monolayers were voltage-clamped at 0 mV. Transepithelial short circuit current (I sc ) and resistance (R TE ) were simultaneously measured using 5-mV pulses every 20 s. Data were recorded for off-line analysis using the Acquire and Analyze program combined with an epithelial voltage clamp amplifier VCC MC8 (Physiologic Instruments, Inc., San Diego, CA). After stabilization of the basal I sc , the pH of the apical or basolateral solution was adjusted to 6.8 by 1 M HCl. A pH of 6.8 was chosen, because more acidic pH values would have altered gap junction and possibly the intracellular pH (40). Clinically, pH 6.8 is within the range of acidic CF luminal fluid (29,41,42). Moreover, Verkman and co-workers (32) have recently reported that the pH of mucus stimulated with pilocarpine is from ϳ6.6 to ϳ6.9. The pH values of bath solutions were concurrently monitored by a digital pH meter during Ussing chamber experiments (Corning).
Western Blotting and Co-immunoprecipitation-Calu-3 cells were split and plated, per 75-mm flask, with minimum essential medium supplemented with 10% fetal bovine serum (Cellgro), 1% glutamine, and 1% strep and penicillin for 48 h. The cells were lysed in 0.7 ml of lysis buffer (1% Triton X-100, 50 mM Tris buffer pH 7.5) with a freshly added protein inhibitor mixture tablet (44). Immunoprecipitation and Western blotting assays were performed as described previously (44,45). Equivalency of loading for Western blot analysis and immunoprecipitation was ensured by performing protein assays (BCA; Pierce) and using equal amounts of protein for each group. Briefly, anti-CFTR monoclonal antibody (University of Alabama at Birmingham) or anti-ASIC3 polyclonal antibody (Chemicon) was added into the cell lysates of the protein samples and rotated at 4°C overnight. For co-immunoprecipitation, the protein complexes were immunoprecipitated by protein G-Sepharose beads (Amersham Biosciences), and the samples were loaded and run in either 6 or 8% SDS-PAGE gel. The precipitates were transferred to a polyvinylidene difluoride membrane and immunoblotted by different antibodies. All blots were developed by the Enhanced Chemiluminescence kit (Pierce).
For co-immunoprecipitation in human embryonic kidney 293T (HEK293T) cells, Lipofectamine 2000 (Invitrogen) was used to transiently transfect ASIC3 and CFTR following the manufacturer's instructions. Cell protein was purified for CFTR and ASIC3 expression 2 days post-transfection. The amount of plasmid DNA used per 35-mm dish was 2 g for CFTR in pcDNA3 and 1 g for rat ASIC3 in pCI per 35-mm dish. Cells were lysed in 1% Triton X-100 in phosphate-buffered saline and a mixture of protease inhibitors (Roche Applied Science). CFTR was immunoprecipitated using 1 g of anti-C-terminal CFTR monoclonal antibody cross-linked to protein A/G-agarose beads (Santa Cruz Biotechnology). After incubating the lysates with cross-linked antibody, the beads were pelleted and washed three times in phosphatebuffered saline containing 1% Triton X-100, and the samples were processed for Western blotting using the monoclonal anti-CFTR antibody (Chemicon) or polyclonal anti-ASIC3 antibody (Neuromics).
Data Analysis-Proton-activated Na ϩ current was calculated as the difference between the basal current and sustained current at acidic pH. The EC 50 value of external pH was computed by fitting the dose-response curve with the Hill equation where I Na represents the proton-activated current, I max the maximal proton-activated current, pH o the extraoocyte pH, EC 50 the value of pH o that results in half of the maximal current of the ASIC3 channel, and n the Hill coefficient.
All average results are presented as mean Ϯ S.E. One-way analysis of variance computation combined with the Bonfer-roni test was used to analyze data with unequal variance between each group. A probability level of 0.05 was considered significant.

ASIC3 mRNA in CFTR-expressing Human Epithelial Cells-
The expression of ASIC3 in airway and other epithelial cells has not been examined previously. Our RT-PCR results revealed that ASIC3 mRNA was present in CFTR-expressing human epithelial cell lines Calu-3 (submucosal cells), 16HBE14o (bronchial cells), CFPAC (pancreatic cells), and T84 (colonic cells) (Fig. 1). However, ASIC1 (585 bp) was not detected in Calu-3 cells. Interestingly, expression of ASIC2 mRNA was also detected in these epithelial cells. The RT-PCR products (585 bp for ASIC1, 503 bp for ASIC2, and 351 bp for ASIC3) from each cell line were ligated into the pCR2.1 vector and sequenced. BLAST analysis revealed that sequences of the RT-PCR products were identical to corresponding human ASIC1, -2, and -3.
ASIC3 Protein Expression in Human CF Lung Tissues-Because ASIC channels are regulated by protons and inflammatory mediators (46,47), the expression of ASIC3 in CF lung may be enhanced by acidic luminal fluid and infection. To examine the expression of ASIC3 in normal and CF lung tissues, we immunolabeled human lung tissue sections with a specific anti-ASIC3 antibody. As shown in Fig. 2, ASIC3 was expressed in the pneumocytes of human lung alveoli. In CF lung, the normal alveolar structure was altered by the aggregation of inflammatory cells and propagation of non-flat abnormal alveolar cells (Fig. 2, E and F). ASIC3 expression, as shown by the brightness of the green color, is greater in CF lung sections due to inflam-mation than that in non-CF tissues. Immunolabeling of ASIC3 was not detected when the anti-ASIC3 antibody was neutralized with ASIC3 immunopeptide (Fig. 2, A and D).
Co-segregation of ASIC3 and CFTR in the Apical Membrane of Calu-3 Monolayers-To examine the subcellular location of ASIC3 and its possible co-expression with CFTR, Calu-3 monolayers grown on filters and cultured with an air-liquid interface were incubated with specific anti-ASIC3 (green channel) and anti-CFTR (red channel) antibodies. Lateral views of the monolayers show that ASIC3 is located in the apical membrane of most Calu-3 cells (Fig. 2, A and D). In agreement with previous observations (38), CFTR was also observed in the apical membrane of Calu-3 cells (Fig. 2, B and E). Merged images indicate that ASIC3 and CFTR are co-segregated with each other within the apical membrane domain of Calu-3 cells (Fig. 2, C  and F). Apical labeling of the monolayers was not observed when non-immune rabbit and mouse IgG were substituted for the anti-ASIC3 and -CFTR antibodies, respectively. To verify these immunostaining results, Western blot assay was used to determine the ASIC3 and CFTR protein in Calu-3 cells (Fig. 3, B and C). A 75-kDa protein was shown with an identical size as in dorsal root ganglia (Fig. 3B). As previously reported (38), CFTR protein is expressed in Calu-3 cells abundantly (Fig. 3C).
ASIC3-type Channel Activity in Calu-3 Monolayers-If ASIC3 functions as a Na ϩ channel, the channel should be activated by lowering apical pH. To test this hypothesis, we bathed Calu-3 cells with HPO 3 Ϫ -free, HCO 3 Ϫ -free, Cl Ϫ -free medium on both sides and bubbled them with pure gaseous nitrogen. Lowering the apical pH from 7.5 to 6.8 resulted in a transient followed by a sustained activation of an apical Na ϩ current (Fig. 4). The proton activation was repeatable. In the presence of amiloride (100 M), an increase of apical proton concentration did not elicit acid-sensitive sustained I sc (Fig. 4).
Putative Proton-gated Whole-cell Na ϩ Current in Calu-3 Cells-To further characterize the putative ASIC3 channels in individual Calu-3 cells, we applied a series of external pH values from 4.0 to 8.5 after the whole-cell configuration of the patch clamp assay was established (Fig. 5). In 27 of 32 cells, a transiently activated current and a sustained current were activated by protons, whereas alkali pH (8.5) inhibited basal current slightly (5A). An EC 50 value of 5.2 for proton activation was determined (5.2 Ϯ 0.1, n ϭ 6) by fitting the dose-response curve with the Hill equation (Fig.  5B). Multiple sites were required for proton activation (Hill coefficient, 12 Ϯ 3), which supports previous publications on this topic (20,22). Seventy percent of this proton-  gated current was abolished in Na ϩ -free medium. The gating pattern of proton-dependent Na ϩ current and the EC 50 value for proton activation in Calu-3 cells were identical to those of cloned ASIC3 but not ASIC1a and -2a channels (48). The large sustained current, which tends to become stable, is a hallmark of homo-and heteromeric ASIC3, which is not observed in other ASIC channels.
Down-regulation of Putative ASIC3 Na ϩ Channels by cAMPactivated CFTR in Calu-3 Cells-To investigate the effect of CFTR activation on the proton-gated ASIC3-type Na ϩ current in Calu-3 cells, proton-gated Na ϩ currents from a group of cells in the absence of forskolin were compared with those obtained from a parallel group in the presence of 10 M forskolin. After CFTR activity reached its maximum in whole-cell patch clamp mode and was stabilized at neutral pH (7.5), pH 6.0 was applied to activate native ASIC3 (Fig. 5C). The proton-gated sustained Na ϩ current in Calu-3 cells in the presence of forskolin was Ϫ33 Ϯ 16 pA (n ϭ 5), which was markedly less than the controls in the absence of forskolin (Ϫ154 Ϯ 28 pA, n ϭ 5, p Ͻ 0.01).
To exclude the direct effect of forskolin on ASIC3 activity, we expressed ASIC3 in oocytes and measured whole-cell currents using the two-electrode voltage clamp technique before and after the addition of forskolin at pH 6.0. The ASIC3 current at Ϫ60 mV was not altered by forskolin (Fig. 5D). These results were consistent with our previous observation that elevation of cAMP does not have an effect on ␣␤␥ ENaC and ASIC1a/2a channels expressed in oocytes (8,39).
Down-regulation of cAMP-activated CFTR Channels by Activated Native ASIC3 in Calu-3 Cells-Similar to the observations described by other groups (49), 10 M forskolin activated both inward and outward CFTR currents in Calu-3 cells (Fig.  6A). Because ASIC3 is not inactivated at neutral pH (Fig. 4), pH 6.0 was applied continuously to test the ASIC3-mediated regulation of the CFTR channel. The CFTR current decreased significantly from Ϫ1056.89 Ϯ 113 to Ϫ743.58 Ϯ 76 pA (n ϭ 5, p Ͻ 0.05) following ASIC3 activation (Fig. 6B). Because ASIC3 was not functionally expressed in each Calu-3 cell, cells were screened with pH 6.0 before forskolin application.
Acidic luminal pH does not regulate the native cAMP-activated Cl Ϫ conductance in sweat gland ducts (50). To examine whether acidic pH influences CFTR activity, CFTR was expressed in oocytes, and the bath solution was switched from ND-96 at pH 7.5 to 6.0 (Fig. 6, C and D). The CFTR current was not markedly reduced (n ϭ 6, p Ͼ 0.05), supportive of results observed in the sweat glands (50).
Co-immunoprecipitation of ASIC3 and CFTR-To further investigate a possible association between CFTR and ASIC3 channel proteins, CFTR was immunoprecipitated from Calu-3  Average sustained currents at Ϫ100 mV were normalized to the maximal current level at pH 4.0 and then were plotted against the applied pH. The line was created by fitting the dose-response curve with the Hill equation to calculate EC 50 for H ϩ activation (n ϭ 6). C, effect of CFTR activation on ASIC3 whole-cell activity in Calu-3 cells. ASIC3 current at Ϫ100 mV activated at pH 6.0 was compared before and after the application of 10 M forskolin to activate CFTR. See Fig. 6A for the time course of CFTR activation by forskolin. D, representative current trace showing the effect of forskolin on ASIC3 activity expressed in an oocyte. ASIC3 mRNA was microinjected into the cytosol of oocytes. The arrow points out the starting time for forskolin application. Holding potential, Ϫ60 mV. ASIC3-associated current (downward) was activated by pH 6.0 in an interval of 1 min. ASIC3 current was monitored in the presence of forskolin (10 M) for 30 min.
cell lysates with a monoclonal anti-CFTR antibody. The immunoprecipitate was next probed with ASIC3 antibody (Fig. 7A). Conversely, the total Calu-3 cell protein was precipitated with ASIC3 antibody and then probed with CFTR antibody (Fig.  7B). However, following washout of antibodies with stripping buffer (Pierce), we reprobed the same membranes with antibody and corresponding immunopeptide; no bands were shown even after 24 h of exposure (Fig. 7C). Our results showed that CFTR and ASIC3 proteins co-immunoprecipitated each other. To corroborate these results, we transiently co-transfected ASIC3 and CFTR in HEK293T cells. The ASIC3 protein could be detected in HEK293T cells only after transient transfection, indicating that parental HEK293T cells did not endogenously express ASIC3 (Fig. 7, D and E). In co-immunoprecipitation experiments, we could specifically and reproducibly co-immunoprecipitate ASIC3 using an anti-CFTR antibody (Fig. 7F). In the absence of transfected CFTR, ASIC3 was not detected in the immunoprecipitates. The lower CFTR expression level in HEK293T cells (Fig. 7D, right lane (both)) is most likely due to the downregulation of CFTR by ASIC3.

DISCUSSION
The hypotheses that ASIC3 may be expressed in the apical membrane of pulmonary epithelial cells and that ASIC3 may interregulate with CFTR have been tested. Our studies show that ASIC3 is expressed in Calu-3 cells at the mRNA and protein levels. ASIC3 is co-segregated with CFTR in the apical membrane of Calu-3 cell monolayers, a finding verified by the immunoblotting assays. Na ϩ current in Calu-3 cells is activated in a pH-dependent manner, and putative ASIC3 channel activity is downregulated by CFTR activation. ASIC3 immunoprecipitates with CFTR in Calu-3 and transfected HEK293T cells, suggesting that a physical protein-protein interaction between these two proteins may contribute to the functional interregulation of native ASIC3 and CFTR channels in Calu-3 cells.
Expression of ASIC3 in Lung Epithelial Cells-Both ENaC and ASIC channels are regulated by extracellular pH (20,51). To examine the role of endogenous ASIC3 in the absence of ENaC, we chose Calu-3 cells, because only ␣ ENaC is detected using RT-PCR (43). Amiloride-sensitive Na ϩ current has not been documented in Calu-3 cells (52), supportive of the fact that even heterologously overexpressed ␣ ENaC alone in oocytes only has onehundredth the channel activity of ␣␤␥ ENaC (23). To our knowledge, this is the first study to show the ASIC3 expression at the mRNA and protein levels in lung epithelial cells. The unique gating kinetics by protons, as shown by patch clamp studies, is consistent with those of heterologously expressed ASIC3 homomers (but not ASIC2 channels) (34,35,48). The presence of the K ϩ channel inhibitor (triethanolamine) in the pipette solution combined with the inhibitory effect of acidic pH on the I sc when applied to the basolateral membrane of Calu-3 monolayers indicate that the proton-gated currents in Calu-3 cells are not carried by other alkali metal ions. Any HCO 3 Ϫ -coupled cation transport system is expected to be eliminated in HCO 3 Ϫ -free medium and in the absence of CO 2 (53)(54)(55). Most importantly, the proton-activated I sc and whole-cell current are sensitive to amiloride, a hallmark of the ENaC/degenerin superfamily (23).
The nature of the I sc peak current in Calu-3 is unknown. Obviously, the kinetics of the peak current is not completely consistent with the heterologously expressed ASIC3 channel (34,35,48). Diversities in their biophysical properties and regulation have been reported for other epithelial channels. For example, cloned ␣␤␥ ENaC is not acutely regulated by cAMP (23). Also, amiloride sensitivity, ion selectivity, and single channel conductance are different from native epithelial Na ϩ channels (23). In addition, the putative CFTR channels in Calu-3 cells (but not heterologously expressed CFTR in oocytes) exhibit a basal, glibenclamide-inhibitable Cl Ϫ current before the addition of cAMP (4,7,8,38). The mechanisms for the differences between cloned and native epithelial cation and anion channels are not clear. It has been proposed that culture FIGURE 6. Regulation of CFTR current level by activated ASIC3. A, both outward (at ϩ60 mV) and inward currents (at Ϫ120 mV) in Calu-3 cells were monitored. Application of forskolin and pH 6.0 are indicated by the arrow and line, respectively. B, average current level. Forskolin-activated CFTR activity at neutral pH 7.5 (Neutral). CFTR activity was then significantly reduced at pH 6.0 (Acidic). n ϭ 5; **, p Ͻ 0.05 compared with basal level and ##, with that at neutral pH. C and D, effects of acidic pH 6.0 on CFTR currents in oocytes. CFTRexpressing oocytes were superfused with forskolin at neutral (7.5) and acidic (6.0) pH. Acidic pH 6.0 did not markedly affect the CFTR current amplitude (n ϭ 5, p Ͼ 0.05). DECEMBER  Ϫ -mediated electrical ion transport, we cannot exclude the possibility that the peak current is mixed with Cl Ϫ and HCO 3 Ϫ current (55). Even after we removed Cl Ϫ , HCO 3 Ϫ , and HPO 3 Ϫ from the extracellular medium, these anions remained in cytosol. We thus focused on sustained current to avoid possible contribution of non-ASIC3 channel current.

ASIC3 Interacts with CFTR
Voltage-dependent H ϩ channels have also been found in other pulmonary epithelia, but not in Calu-3 cells (56). Similar to the apical K ϩ channel, this H ϩ channel is activated by intracellular (but not extracellular) protons. Although some epithelial Cl Ϫ channels, such as ClC2, are regulated by extracellular pH, Ba 2ϩ (5 mM) included in the bath solution should completely inhibit ClC2 activity (57)(58)(59). Collectively, our results strongly suggest that ASIC3 is functionally expressed in the apical membrane of the Calu-3 cell line.
Regulation of ASIC3 in Epithelia-Co-immunoprecipitation of heterologously expressed ASIC3 and CFTR together with the co-segregation of CFTR and ASIC3 in the apical membrane of Calu-3 cell monolayers suggests that the two proteins are associated in a macromolecular complex and may directly interact with each other. A reciprocal interaction between homomeric CFTR and ENaC has been reported (4,7,8); CFTR down-regulates ␣␤␥ ENaC, whereas ␣␤␥ ENaC up-regulates CFTR. In contrast, CFTR up-regulates heteromeric ASIC1a/2a channels (39). The functional interaction between ASIC3 and CFTR in Calu-3 cells differs from these reports in either case.
Potential Physiologic and Pathologic Relevance of ASIC3 in CF Epithelia-A more acidic pH has been detected in CF epithelial tissues, including airway surface fluid (41), gastrointestinal secretions (60), pancreatic juices (28,29), and epididymal semen (42). The mechanisms underlying the acidic CF luminal fluid are not completely clear. The possible events contributing to the low pH in airway and mucosal secretions include 1) defective HCO 3 Ϫ secretion via CFTR and other anion transporters/exchangers, such as putative anion transporter 1 (PAT-1) and diarrhea anion exchanger (DRA); 2) leukocyte inflammation; 3) raised metabolic acids and proton efflux; and 4) abnormal NH 3 ϩ secretion. During a bronchospastic asthmatic attack, the pH of the airway surface liquid can be as low as 5.2 (61), and acidic luminal pH has been implicated in both cough and reduced ciliary beating. Acidic pH also compromises epithelial integrity in the lung and may exacerbate chronic lung infection (62). Luminal liquid acidification in CF may, therefore, represent a critical pathological step in CF lung infection. Under an acidic microenvironment, epithelial ASIC3 is activated and strongly impairs CFTR function (32). Because both ENaC and ASIC3 are amiloride-sensitive, it is difficult to distinguish the corresponding contributions of these two Na ϩ channels to the hyperactive Na ϩ reabsorption. We propose that salt absorption through overactive ASIC3 represents an important participant in the CF sodium transport defect (Fig. 8). Silent ASIC3 channels at a neutral pH may be strongly activated in acidic CF epithelia. This . Co-immunoprecipitation of native ASIC3 and CFTR proteins in Calu-3 cells. A, CFTR precipitates with ASIC3. Calu-3 cell lysates were immunoprecipitated (IP) with anti-CFTR monoclonal antibody in 8% SDS-PAGE gel, and the blot was then probed with anti-ASIC3 polyclonal antibody (IB). B, ASIC3 co-immunoprecipitates with CFTR. Cell lysates were immunoprecipitated with ASIC3 antibody in 6% SDS-PAGE gel, and the blot was probed with CFTR antibody. C, reprobing the same blot as shown in A with ASIC3 immunopeptide and antibody (ASIC3 ϩ Ag) after washing out antibodies with stripping buffer. These experiments were repeated three times with identical results. D, CFTR expression in transfected HEK293T cells. CFTR was not detected in HEK293T cells transfected with ASIC3 alone (ASIC3) but with CFTR or co-transfected with CFTR and ASIC3 (both). E, ASIC3 expression in transfected HEK293T cells. ASIC3 protein is only detected by Western blotting (IB) in cells transfected with ASIC3 (ASIC3 and both). F, co-immunoprecipitation of CFTR and ASIC3. ASIC3 specifically co-precipitates with CFTR (both) only in cells co-transfected with CFTR and ASIC3 (both). n ϭ 3. FIGURE 8. Proposed working model for CF epithelia. In normal epithelium (left panel), luminal fluid has a neutral pH. ASIC3 is silent and down-regulated by functional CFTR channels. In a CF epithelium (right panel), the luminal fluid is acidic and up-regulating ASIC3. Stimulatory (ϩ) and inhibitory (Ϫ) effects are enhanced (thicker arrows), reduced (thinner arrows), or even completely shut down (lines removed) in CF epithelium.