Amiloride-sensitive epithelial Na+ channels (ENaCs) are subject to modulation by many factors. Recent data have also linked the N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) machinery to this regulation of ENaC, but the molecular mechanisms that underlie this modulation are poorly understood. In this study, we demonstrate that syntaxin 1A physically interacts with ENaC and functionally regulates ENaC activity. Syntaxin 1A was able to coimmunoprecipitate in vitro-translated γ-ENaC, but not α- or β-ENaC. Also, using antibodies raised against α-, β-, or γ-ENaC, we detected syntaxin 1A in immunoprecipitates from Madin-Darby canine kidney cells stably transfected with αβγ-ENaC. In bilayers, syntaxin 1A inhibited ENaC, and this syntaxin 1A modulation of ENaC activity was eliminated by truncations of cytoplasmic domains of the ENaC subunits. Our findings provide evidence for a direct physical interaction between ENaC and syntaxin 1A and suggest involvement of ENaC's cytoplasmic domains in functional modulation of ENaC activity by syntaxin 1A.
the amiloride-sensitive epithelial Na+ channel (ENaC) is a member of the degenerin/ENaC family of ion channels. This class of ion channel fulfills a key role in Na+ handling. Its improper functioning has been implicated in several diseases, including salt-sensitive hypertension (Liddle's syndrome), salt-wasting syndrome (pseudohypoaldosteronism type I), pulmonary edema, and cystic fibrosis (3, 8, 33, 42). Strict regulation of ENaC occurs through a wide variety of hormonal and nonhormonal mechanisms (1, 16). Recent studies have implicated the N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) machinery in the regulation of ENaC (35, 37), but mechanisms by which this modulation is achieved remain obscure (34, 36, 38). The observation that syntaxin 1A increased surface epifluorescence of ENaC but decreased channel activity led to the suggestion that factors other than surface expression are involved in the regulation of ENaC (37). Alternatively, Qi et al. (35) ascribed the decrease in ENaC current to the ability of syntaxin 1A to reduce Na+ channel plasma membrane density by syntaxin 1A's interference with insertion of ENaC into the plasma membrane.
In this study, we explored the molecular events involved in syntaxin-ENaC interactions. We found that 1) syntaxin 1A coimmunoprecipitated in vitro translated γ-ENaC, but not α- or β-ENaC; 2) antibodies raised against α-, β-, or γ-ENaC detected syntaxin 1A in immunoprecipitates from Madin-Darby canine kidney (MDCK) cells stably transfected with αβγ-ENaC; 3) syntaxin 1A had no effect on channels formed by α-ENaC alone; 4) the functional effects of syntaxin 1A on ENaC activity formed by αβ-, αγ-, and αβγ-ENaC required the intact NH2- and COOH-terminal domains of α-ENaC and intact β- or γ-ENaC; and 5) in the αβγ-ENaC complex, β- and γ-ENaC were interchangeable with regard to syntaxin 1A inhibition. In conclusion, our findings provide evidence for direct interaction between syntaxin 1A and ENaC and the involvement of the cytoplasmic domains of ENaC in the functional regulation of ENaC by syntaxin 1A.
MATERIALS AND METHODS
Full-length rat ENaCs were the kind gifts of Drs. Cecilia Canessa (Yale University) and Bernard Rossier (Universitat de Lausanne). The cDNA for human syntaxin 1A was the kind gift of Dr. Kevin Kirk (University of Alabama at Birmingham) (32). cRNAs were made from these DNA samples. DNA samples were in vitro transcribed using the T3 mMessage mMachine kits (Ambion). The quality and lengths of the cRNAs were confirmed by denaturing formaldehyde-agarose gel electrophoresis. The concentration of each RNA was determined by UV spectrophotometry (wavelength = 260 nm).
Preparation of syntaxin 1A protein.
Syntaxin 1A glutathione S-transferase (GST) fusion protein was prepared as previously described (6).
MDCK cell culture.
We grew stable transfects of MDCK cells expressing all three ENaC subunits (courtesy of Dr. James Schafer, University of Alabama at Birmingham) on poly-l-lysine-coated semipermeable supports (24-mm diameter, 4-μm pore size; Transwell, Costar) in DMEM containing 10% fetal bovine serum. This medium also contained 1% penicillin-streptomycin, G-418 (800 μg/ml), hygromycin (300 μg/ml), and puromycin (5 μg/ml). To induce the expression of ENaC, the medium was supplemented with 2 μM dexamethasone and 2 mM sodium butyrate 24 h before the cells were harvested. Under these conditions and when raised on permeable supports, these cells typically achieve a transepithelial resistance of 1 kΩ·cm2 (30).
The antibodies against the individual subunits of ENaC were previously characterized and are specific for the respective subunits (9). The syntaxin 1A antibody was specific, in that it did not cross-react with syntaxin 2, 3, or 4 (32).
In vitro transcription and translation.
cDNAs were transcribed and translated in vitro using the TNT transcription/translation system (Promega) without dog pancreas microsomes, as previously described (25). To test for protein-protein interaction between different ENaC subunits and syntaxin 1A, we translated these constructs with radioactive or nonradioactive methionine, immunopurified them, and reconstituted them in different combinations in proteoliposomes as previously described (25). Proteoliposomes were solubilized in immunoprecipitation buffer that contained 50 mM Tris (pH 7.5), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, and 0.1% SDS. All immunoprecipitation reactions were carried out in this buffer. The antibodies directed against nonlabeled proteins were used, and coimmunoprecipitated radioactively labeled proteins were detected using SDS-PAGE and autoradiography.
Immunoprecipitation, gel electrophoresis, and Western blotting.
Immunoprecipitation and coimmunoprecipitation from MDCK cell lysate were performed using a protein A immunoprecipitation kit (Pierce) according to the manufacturer's instructions. Briefly, each affinity-purified rat ENaC subunit antibody (50 μg) was bound and cross linked to protein A beads. After they were washed extensively, the beads were added to MDCK cell lysate (0.3 ml) and incubated overnight at 4°C. Immunoprecipitated proteins were eluted with elution buffer and subjected to electrophoresis and Western blotting. Proteins were run on 8% SDS-PAGE minigels for ∼1 h at 200 V in a Minisubcell apparatus (Bio-Rad, Hercules, CA). Gels were transferred onto polyvinylidene fluoride, treated with 5% nonfat dry milk-Tris-buffered saline-Tween, and probed with the anti-syntaxin 1A antibodies. Secondary antibody was conjugated to horseradish peroxidase, and visualization was performed with chemiluminescence reagents (Amersham Pharmacia Biotech, Piscataway, NJ). Controls included substitution of nonimmune rabbit IgG for primary antibodies.
Planar lipid bilayer experiments were conducted as previously described (22). Symmetrical bathing solutions were used in all experiments. This solution contained 100 mM NaCl + 10 mM MOPS-Tris buffer (pH 7.4). Phospholipids were obtained from Avanti Polar Lipids (Alabaster, AL). Filter-sterilized MilliQ water was used in all experiments. Data analysis was performed as described previously (22). In vitro translation and reconstitution of ENaC proteins were performed as previously described (2, 5). Control liposomes were prepared from a mock in vitro translation mixture lacking ENaC. The truncation mutants of ENaC subunits used in this study have been described previously (4, 5). The statistical significance of the differences between properties of ENaCs in the presence and absence of syntaxin 1A was estimated using a standard two-sample t-test for comparison of the means in two independent groups (48).
To provide biochemical evidence of a direct interaction between ENaC and syntaxin 1A, coimmunoprecipitations were performed using in vitro translated individual ENaC subunits and syntaxin 1A (Fig. 1). In addition, these experiments were designed to define which subunits of ENaC interact with syntaxin 1A. Our results show that syntaxin 1A can coimmunoprecipitate only the γ-subunit of ENaC, and not the α- or β-subunit (Fig. 1A). The same results were obtained if in vitro translation of proteins was performed in the presence of dog pancreas microsomes (data not shown). Figure 1B shows that syntaxin 1A can coimmunoprecipitate radiolabeled α-, β-, and γ-ENaC when they are in combination with their conjugate partner in a macromolecular complex. However, syntaxin 1A can coimmunoprecipitate the other ENaC subunits only when the γ-subunit is present. Thus protein-protein interactions between syntaxin and the γ-ENaC subunit are likely.
To provide more direct evidence of the physical interaction between ENaC and syntaxin 1A in an intact epithelial cell, α-, β-, and γ-ENaC stably transfected MDCK cells were grown to confluence, a total cell lysate was prepared, and an immunoprecipitate was made using antibodies raised against α-, β-, or γ-ENaC. The immunoprecipitate was transferred to nitrocellulose and probed for syntaxin 1A (Fig. 2). Syntaxin 1A was detected in each of the lanes.
Effects of syntaxin 1A on wild-type ENaCs in bilayers.
Next, we tested the effects of syntaxin 1A on ENaCs in bilayers. A GST fusion protein of syntaxin 1A lacking the COOH-terminal membrane anchor (GST-Syn1AΔC) inhibited the single-channel activity of αβγ-ENaC (Fig. 3A). GST-Syn1AΔC inhibited αβγ-ENaC in a dose-dependent manner (Fig. 3B), with Ki = 0.26 ± 0.50 μM. In a control set of experiments using the fusion protein GST-syntaxin 1A-(1–194) (lacking the H3 domain), we observed no functional effect (data not shown).
Biochemically, syntaxin 1A was able to immunoprecipitate only the in vitro translated γ-subunit of ENaC, and not the α- or β-subunits (Fig. 1). Also, syntaxin 1A was detected in immunoprecipitates from MDCK cells stably transfected with αβγ-ENaC using antibodies raised against α-, β-, or γ-ENaC (Fig. 2). These observations prompted us to study the effects of syntaxin 1A on channels made of different ENaC subunits. We described previously in detail the single-channel properties of channels made of wild-type (WT) α-, αβ-, αγ-, and αβγ-ENaCs and found that they were indistinguishable from each other (4, 5, 23). Figure 4 depicts the single-channel recordings of WT α-, αβ-, αγ-, and αβγ-ENaCs in the presence of GST-Syn1AΔC. Addition of GST-Syn1AΔC to the bilayer bathing solution inhibited open probability of αβ-, αγ-, and αβγ-ENaCs but had no effect on α-ENaC (Table 1). These experiments show the importance of the β- and γ-subunits in the effects of syntaxin in the modulation of ENaC. However, the role of the α-subunit in the effects of syntaxin cannot be excluded a priori, because only α-ENaC is capable of forming a channel by itself (10, 21).
Effects of syntaxin 1A on ENaCs with truncated cytoplasmic domains.
Next, we explored the role of cytoplasmic domains of individual subunits in the effects of syntaxin 1A on ENaCs. To investigate the involvement of cytoplasmic domains of ENaC in mediating syntaxin 1A effects on ENaCs, we performed a series of experiments with the NH2- or COOH-terminal intracellular domains of ENaCs deleted (with WT and/or each other).
Effects of NH2-terminal truncations of ENaC subunits.
We showed previously that the NH2-terminal truncation of the α-subunit drastically changes channel behavior, resulting in “acceleration” of channel kinetics, whereas NH2-terminal truncation of the β- and γ-subunits had no apparent effect on channel activity (4, 5). Channels with NH2-terminal truncation of α-ENaC were resistant to the inhibitory effects of GST-Syn1AΔC (Fig. 5, Table 2). The NH2-terminal truncation of β- and γ-subunits (with α-ENaC intact) abolished the inhibitory effects of GST-Syn1AΔC on the heterodimers (αβΔN- and αγΔN-ENaC; Table 2), but not on the heterotrimers (αβΔNγ- and αβγΔN-ENaC, Fig. 5, Table 2); however, the channel activity of the heterotrimer with NH2-terminal truncation of both subunits (αβΔNγΔN-ENaC; Fig. 5, Table 2) was not affected by GST-Syn1AΔC. These results underscore the necessity of the NH2-terminal domain of α-ENaC and the NH2 terminus of β- or γ-ENaC in the effects of syntaxin 1A on the heteromeric channel complex.
Effects of COOH-terminal truncations of ENaC subunits.
COOH-terminal truncation of the α-subunit did not alter the channel behavior of ENaC, whereas COOH-terminal truncation of the β- and γ-subunits converted the channels to continuously open channels (5, 15, 23, 39), a situation analogous to that in Liddle's disease. The outcome of the syntaxin-ENaC experiments with COOH-terminal truncations of subunits was similar to that with NH2-terminal truncation of subunits. A COOH-terminal truncation of α-ENaC effectively abolished the inhibitory effects of GST-Syn1AΔC on ENaC (Fig. 6, Table 3). COOH-terminal truncation of β- or γ-subunits (with α-ENaC intact) in heterotrimeric channels (αβΔCγ- and αβγΔC-ENaC; Fig. 6, Table 3) did not interfere with the inhibitory effects of GST-Syn1AΔC. On the other hand, the heterodimeric channels with these truncations (αβΔC- and αγΔC-ENaC; Table 3) were resistant to the inhibitory influence of GST-Syn1AΔC. Also, the inhibition by GST-Syn1AΔC was absent when the heterotrimeric channel contained COOH terminally truncated β- and γ-subunits (αβΔCγΔC-ENaC; Fig. 6, Table 3). The results of these sets of experiments emphasize the importance of the COOH-terminal domains of α- and β-ENaC or γ-ENaC in the effects of syntaxin 1A on channel activity.
ENaC-syntaxin 1A interaction.
The results of our study indicate that syntaxin physically interacts and functionally regulates ENaC. We found that syntaxin 1A coimmunoprecipitated with in vitro translated γ-ENaC, but not α- or β-ENaC. These observations are in agreement with those of Qi et al. (35), who reported a physical association between γ-ENaC and syntaxin 1A, and Saxena et al. (37), who found that the COOH terminus of the β-subunit does not interact with syntaxin 1A. Also, we were able to detect syntaxin 1A using antibodies raised against α-, β-, or γ-ENaC in immunoprecipitates from MDCK cells stably transfected with αβγ-ENaC. Recently, Condliffe et al. (12) found that syntaxin 1A was able to interact with the COOH termini of all three ENaC subunits, but not with the NH2 termini of any ENaC subunits. It is possible that even though syntaxin 1A is able to interact physically with any particular subunit in vitro (or in the cells), the unveiling of the functional effects probably requires the participation of more than one subunit.
Syntaxin 1A effects and WT ENaCs.
Any biochemical association and subsequent functional effects of syntaxin 1A on ENaC should involve α-ENaC. This assumption is based on the findings that only α-ENaC is capable of forming a channel by itself; neither β- nor γ-ENaC by itself or together is capable of forming a channel without the α-subunit (10, 21), but these subunits are very important targets of different regulatory inputs (1, 16). In bilayers, syntaxin 1A lacked any effect on α-ENaC, but the combination of α-ENaC with β- or γ-ENaC (αβ and αγ) or both (αβγ) rendered the channel sensitive to syntaxin 1A. In these experiments, we used the GST fusion protein of syntaxin that lacked the transmembrane domain. In a separate set of experiments (data not shown), we evaluated the effects of the target (t)-SNARE heterodimer (kind gift of Dr. T. Weber, Mount Sinai School of Medicine, New York, NY) composed of mouse His6-25-kDa synaptosome-associated protein (SNAP-25) and rat syntaxin 1A (containing the COOH-terminal membrane anchor), full-length syntaxin 1A alone, and the combination of full-length syntaxin 1A and SNAP-23 (a nonneuronal homolog of SNAP-25). These proteins produced an inhibition of ENaC activity that was qualitatively comparable to the effects of GST-Syn1AΔC, but a quantitative comparison is complicated by the fact that preparation of these proteins involves their reconstitution into proteoliposomes. The use of proteoliposomes can lead to different local concentrations of these proteins. Also, the coexpression of syntaxin 1A with SNAPs could lead to coordinate (14) and reciprocal (50) regulation of ion channel activity. These scenarios could potentially lead to underestimation of the role of the transmembrane domain of intact syntaxin in its functional effects on ENaC, which is thought to influence the syntaxin protein-protein interactions (26). This circumstance prompted us to restrict our functional experiments to GST-Syn1AΔC to simplify interpretation of our data.
Syntaxin 1A effects and different ENaC subunits.
In bilayers, we found that the inhibitory influence of syntaxin 1A on different ENaCs was reversed by truncations of cytoplasmic domains of ENaC. Our findings show the importance of all the ENaC subunits in the effects of syntaxin in modulating ENaC activity.
Other groups also reported regulation of ENaC by syntaxin 1A (12, 35, 37). Saxena et al. (37) found that the COOH terminus of the β-subunit does not interact with syntaxin 1A. Also, syntaxin 1A unexpectedly increased surface epifluorescence of ENaC; however, the activity of ENaC was decreased, prompting the authors to suggest that factors other than surface expression can regulate ENaC function (37). On the other hand, Qi et al. (35) reported a physical association between the γ-subunit and syntaxin 1A and attributed the decrease in ENaC current to syntaxin 1A's ability to reduce the number of Na+ channels in the plasma membrane resulting from interference of syntaxin 1A with ENaC insertion into the plasma membrane. Recent findings by Condliffe et al. (12) indicate that syntaxin 1A interacts with the COOH termini of all three ENaC subunits, but not with the NH2 termini of any ENaC subunits. In this respect, our bilayer data correlate and contrast with the findings of Condliffe et al. Our results suggest the involvement of the NH2 termini of all three subunits in the effects of syntaxin 1A in regulating ENaC, whereas the findings of Condliffe et al. suggest that the NH2 termini are not involved. The reasons for differences between our findings and those of Condliffe et al. are not clear but could be a reflection of the following factors. First, of course, the most obvious difference lies in the systems used (in vitro translation and artificial planar lipid bilayers vs. oocytes and A6 cells). Condliffe et al. showed some differences between their findings and the data of Saxena et al. (even though the same systems were employed). A second reason might be the source of the GST fusion proteins used in coimmunoprecipitation experiments (prepared from bacteria vs. in vitro translation). Third, the environment in which ENaC finds itself in our system, or any other model system, differs from its native environment. Indeed, a growing number of studies have indicated that the lipid environment has a role in ion channel function. We cannot rule out the possibility that the lipid environment may affect the functioning of the ENaC itself and/or its association with any interacting protein(s). This possibility is underscored by the findings that some phospholipids may prevent rundown of ENaC (28) and the existence of endogenously expressed ENaC subunits in cholesterol-enriched membrane microdomains known as lipid rafts (19). In bilayers, syntaxin 1A did not affect the activity of α-ENaC itself, but the COOH and NH2 termini of α-ENaC were required for the effect of syntaxin 1A in heteromeric channels. Moreover, COOH and NH2 termini of the β- and γ-subunits (with α-ENaC intact) were necessary for the inhibitory effects of syntaxin 1A to occur on the heterodimers (Tables 2 and 3), but not on the heterotrimers (with truncation only in β- or γ-ENaC, with α-ENaC intact; Tables 2 and 3). Finally, the heterotrimer with NH2-terminal (or COOH-terminal) truncation of both subunits (Tables 2 and 3) was not affected by syntaxin 1A. Even though the NH2-terminal domain of α-ENaC was proposed as the location for the channel gate (4, 5, 18), its overall role, as well as that of β- and γ-ENaC in channel functioning, remains obscure. This fact is also underscored by the role of the NH2 terminus in the effects of syntaxin 1A on ENaC. We propose that the absence of the syntaxin 1A effect on ENaC may reflect improper cooperative assembly of ENaC as a consequence of NH2-terminal truncation of ENaC subunits by analogy with the role of the NH2 terminus of other ion channels in assembly (7, 20, 27, 40, 41, 44, 49).
The COOH-terminal domain of ENaC is a known site of interaction of kinases, Nedd4, and cytoskeletal elements (13, 17, 45, 46). Similarly, if the COOH-terminal domain of α-ENaC is a binding site of syntaxin 1A, then an intact β- or γ-ENaC will be needed for this interaction to take place (with the assumption that α-ENaC remains intact). Under this scenario, the β- and γ-subunits may be interchangeable in the complex with respect to syntaxin interaction. This involvement of the cytoplasmic domain(s) of ENaC in mediating syntaxin 1A effects may reflect the generalized phenomenon of interplay between syntaxin 1A and ion channels (11, 31), which is, in turn, dependent on many contributing factors (12, 24, 29, 43, 47).
The results of our study indicate the physical and functional interaction between syntaxin 1A and ENaC and suggest that this interaction depends on the cytoplasmic domains of ENaC. If extrapolated broadly, our findings offer a potential physiological explanation for the ENaC-syntaxin 1A interaction: a means to deliver an inactive channel to the plasma membrane.
This work was supported by National Institutes of Health Grants DK-37206 (D. J. Benos), GM-56827 (E. R. Chapman), and MH-61876 (E. R. Chapman) and by the Milwaukee Foundation (E. R. Chapman). E. R. Chapman is a Pew Scholar in the Biomedical Sciences.
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.
- Copyright © 2004 the American Physiological Society