Epithelial sodium channel activity in detergent-resistant membrane

Am J Physiol Renal Physiol 284: F182–F188, 2003. First published September 11, 2002; 10.1152/ajprenal.00216.2002.

Epithelial sodium channel activity in detergent-resistant membrane microdomains ´ DE ´ RIQUE MIES, AND SARAH SARIBAN-SOHRABY VADIM G. SHLYONSKY, FRE Laboratory of Physiology and Physiopathology, Universite´ Libre de Bruxelles, 1070 Brussels, Belgium Submitted 6 June 2002; accepted in final form 4 September 2002

Shlyonsky, Vadim G., Fre´de´rique Mies, and Sarah Sariban-Sohraby. Epithelial sodium channel activity in detergent-resistant membrane microdomains. Am J Physiol Renal Physiol 284: F182–F188, 2003. First published September 11, 2002; 10.1152/ajprenal.00216.2002.—The activity of epithelial Na⫹ selective channels is modulated by various factors, with growing evidence that membrane lipids also participate in the regulation. In the present study, Triton X-100 extracts of whole cells and of apical membrane-enriched preparations from cultured A6 renal epithelial cells were floated on continuous-sucrose-density gradients. Na⫹ channel protein, probed by immunostaining of Western blots, was detected in the high-density fractions of the gradients (between 18 and 30% sucrose), which contain the detergentsoluble material but also in the lighter, detergent-resistant 16% sucrose fraction. Single amiloride-sensitive Na⫹ channel activity, recorded after incorporation of reconstituted proteoliposomes into lipid bilayers, was exclusively localized in the 16% sucrose fraction. In accordance with other studies, highand low-density fractions of sucrose gradients likely represent membrane domains with different lipid contents. However, exposure of the cells to cholesterol-depleting or sphingomyelin-depleting agents did not affect transepithelial Na⫹ current, single-Na⫹ channel activity, or the expression of Na⫹ channel protein. This is the first reconstitution study of native epithelial Na⫹ channels, which suggests that functional channels are compartmentalized in discrete domains within the plane of the apical cell membrane. sodium reabsorption; A6 cells; amiloride; lipid bilayers ⫹

AMILORIDE-SENSITIVE NA

CHANNELS mediate vectorial transport of Na across reabsorbing epithelia including renal distal and collecting tubules, distal colon, and lungs. Their function is essential in salt and water homeostasis, including the regulation of blood volume and pressure. These channels, located in the apical cell membrane of polarized epithelia, consist of heterooligomeric complexes comprising several proteins required for full activity and hormone responsiveness (1, 8, 12, 24, 32). A number of intracellular factors and signaling pathways regulate these channels, but the influence of the composition and biophysical properties of the cell membrane itself on the number and location of functional native channel complexes remains unclear. ⫹

Address for reprint requests and other correspondence: S. SaribanSohraby, Campus Erasme, CP 604, 808, route de Lennik, 1070 Brussels, Belgium (E-mail: [email protected]). F182

The heterogeneous distribution of lipid in the cell membranes of epithelia, first described by Simons and Van Meer (31), leads to the formation of lipid microdomains that resist detergent solubilization. A number of studies have been done to characterize the composition of microdomains and reveal that specific proteins, including ion channels, tend to localize in them in a functional way (17, 21). Furthermore, it was recently demonstrated that lipid modifications of microdomains alone are sufficient to confer specific sublocalization of active proteins (34). Here, we report the reconstitution of functional amiloride-sensitive Na⫹ channels obtained from cultured renal epithelial cells (A6) into artificial planar lipid bilayer membranes. When apical membrane-enriched extracts are subjected to sucrose density centrifugation, active Na⫹ channels float in a low-density, detergent-insoluble fraction whereas channel protein found in the detergent-soluble fractions is inactive, indicating that the association with native lipids directly contributes to the regulation of Na⫹ movement through the channel. EXPERIMENTAL PROCEDURES

Cell culture. A6 cells (American Type Culture Collection derived originally from Xenopus laevis) were maintained in culture on plastic flasks in DMEM-F-12 growth medium, adapted for amphibian tissue culture by a 20% dilution with distilled water and supplemented with 5% FBS (HiClone). For biochemical work, cells were plated on 100-cm2 homemade structures with porous supports (HAWP, Millipore) and harvested after 10 days in culture. Maximum and stable values of transepithelial Na⫹ transport and electrical parameters are observed at this time, indicating that apical Na⫹ channels are functional (27). Transepithelial measurements of voltage and resistance were performed on 0.33-cm2 structures (Costar) using an EVOM volt-ohmmeter (World Precision Instruments). The corresponding amiloride-inhibitable Na⫹ current was calculated from these values. Whole cell Triton X-100 extracts. Cells were scraped from porous supports in MOPS-buffered saline (MBS; 25 mM MES, 150 mM NaCl, 1 mM PMSF, pH 6.5) and homogenized. Samples were allowed to solubilize for 1 h on ice in the

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.

0363-6127/03 $5.00 Copyright © 2003 the American Physiological Society

http://www.ajprenal.org

NA⫹ CHANNELS IN EPITHELIAL MICRODOMAINS

presence of 1% Triton X-100. Sucrose was then added to a final concentration of 40%. Apical membrane-enriched Triton X-100 extracts. Cells were scraped in MBS, homogenized, and centrifuged at low speed (5,500 g) for 10 min. Supernatants were then centrifuged at 28,000 g for 1 h. The apical membrane-enriched pellets (cf. Ref. 27) were recovered, resuspended in MBS, and solubilized for 1 h on ice in the presence of 1% Triton X-100. Floatation on sucrose density gradients. Samples were placed at the bottom of linear 5–30% sucrose gradients prepared in MBS without Triton X-100 and centrifuged to equilibrium in a Beckman SW41 rotor at 39,000 rpm for 18 h at 4°C. Gradient fractions of 600 ␮l were collected from the top and snap-frozen. Typically, ⬃20 fractions were recovered, in which sucrose concentration was measured by refractometry. SDS-PAGE and immunoblotting. Aliquots of sucrose gradient fractions were subjected to 7.5% SDS-PAGE under reducing conditions and transferred to nitrocellulose. Immunoblotting was performed in Tris-buffered saline with 5% powered low-fat milk and 0.1% Tween 20. Na⫹ channels were probed with a polyclonal rabbit antibody at a final concentration of 4.5 ␮g/ml. The antibody (a gift from Dr. T. Kleyman) was raised against a portion of the extracellular loop of the ␣-subunit of the cloned epithelial Na⫹ channel from A6 cells and was previously shown to recognize native Na⫹ channels in A6 cells (35). Reactive proteins were detected using a 1:5,000 dilution of alkaline phosphatase-conjugated goat anti-rabbit IgG and the Renaissance chemiluminescence reagent (NEN). Caveolin was probed using a commercial antibody raised against human caveolin-1 (Santa Cruz Biotechnology, Santa Cruz, CA), at a 1:200 dilution. Proteoliposome reconstitution. For functional studies, aliquots (100 ␮l) from each sucrose gradient fraction were added to 100 ␮g of dried palmitoyl-oleoyl-phosphatidylcholine (POPC; Avanti Polar Lipids). Detergent was removed by incubating the samples overnight at 4°C with BioBeads equilibrated with MBS without Triton X-100. Proteoliposomes were recovered by decanting. Before their reconstitution into liposomes, the fractions obtained from membrane material were concentrated six times. Planar lipid bilayer experiments and analysis. Bilayer membranes were formed at room temperature by passing a bubble from a pipette tip prewettted with a membraneforming solution of POPC (25 mg/ml in n-octane) over a 150-␮m-diameter aperture drilled in a 50-␮m-thick wall of a delrin cup containing symmetrical 200 mM Na-gluconate solutions. Currents were measured using a conventional current-to-voltage converter based on an OPA-101 (Burr-Brown, Tucson, AZ) operational amplifier with a 1-G⍀ feedback resistor. The current-to-voltage converter was connected to the trans compartment (0.8 ml) of a bilayer chamber using an Ag-AgCl electrode and 3 M KCl-3% agar bridge. Thus the trans side was a virtual ground. The cis compartment (0.6 ml) was connected to a voltage source. Membrane formation was monitored by the increase in capacitive current to triangle pulses from a function generator. Only membranes with a capacitance of 150–200 pF and a basal conductance of ⬍10 pS were considered satisfactory for experimentation. The trans side was held at ⫺40 mV until the appearance of channel activity. The incorporation protocol consisted of consecutive additions of 1–2 ␮l of proteoliposome suspension to the trans chamber, under constant stirring, to a maximum of 5 ␮l. Twenty minutes of stirring were allowed between the additions. Fusion events occurred after 20–60 min. Currents AJP-Renal Physiol • VOL

F183

were monitored on an oscilloscope and/or a computer screen. Current records were low-pass filtered at 200 Hz through an 8-pole Bessel filter (900 LPF, Frequency Devices, Haverhill, MA) before acquisition at the rate of 1 kHz using a TL-1 DMA interface and Axotape 1.2 software (Axon Instruments). The unitary currents were determined from current amplitude histograms constructed from events lists generated by WinASCD software (Laboratorium voor Fysiologie, KU Leuven, Belgium). The relative Na⫹/K⫹ permeability was calculated from the reversal potentials measured after switching of solutions to bi-ionic conditions, using the Goldman-Hodgkin-Katz equation. RESULTS

As a first approach, we used whole cell, Triton X-100treated samples to look for the presence of Na⫹ channel protein and for single Na⫹ channel activity recovered from each of the sucrose density gradient fractions. With the use of an antibody shown previously to recognize native Na⫹ channel complexes in A6 cells (15), protein was detected by immunoblotting in gradient fractions corresponding to sucrose concentrations of 16 and 18–28.5% (Fig. 1A). Channel protein was not detected in the lighter fractions (5–15% sucrose). Identical volumes of each fraction were used, and it is obvious that channel protein becomes more abundant toward the bottom of the gradient, which contains the detergent-soluble material. In functional studies, we used bathing solutions consisting simply of buffered Na-gluconate, so we would detect only Na⫹ channels. In agreement with the pattern of protein distribution, Na⫹ channel activity from reconstituted proteoliposomes into lipid bilayers was not detected in fractions with ⬍16% sucrose. The fusion of reconstituted liposomes from higher density gradient fractions (from 18 to 28.5% sucrose) into planar lipid bilayers resulted in the appearance of big integral currents. To discern single-channel events, each fraction was diluted with liposomes made of POPC in MBS plus 7.2% sucrose and subjected to freeze-thawing on ice. However, no amiloride-sensitive Na⫹ current was recorded (n ⫽ 15, data not shown). Bilayer incorporation of proteoliposomes from a single fraction (16% sucrose) resulted in the appearance of Na⫹ channel activity typical of native in situ as well as reconstituted channels from A6 cells (Fig. 1B) (9, 29). Based on these data on whole cell extracts, we proceeded with the study of apical membrane-containing fractions from A6 cells. These membrane preparations have been extensively studied previously. Although they contain only 3% of the total cell protein, they are enriched 10-fold in apical membrane markers; active Na⫹ channels are also exclusively located in them (27). Similar to the observations in whole cell extracts, Na⫹ channel protein was detected by immunoblot analysis in fractions of sucrose densities of 16% and 18.5–30% and was more abundant in the heavy-density pellet, which contains the solubilized protein. Figure 2 shows

284 • JANUARY 2003 •

www.ajprenal.org

F184


Fig. 1. A: distribution of A6 Na⫹ channel protein from whole cell extracts in sucrose density gradient fractions. Samples were treated with 1% Triton X-100, floated on a 5–30% linear sucrose gradient, and separated on 7.5% polyacrylamide SDS gels. Samples (25 ␮l; i.e., 4% vol/vol of each fraction) were loaded per lane. Immunoblotting with an anti-␣-Xenopus laevis epithelial Na⫹ channel (ENaC) antibody shows the distribution of A6 Na⫹ channel protein in the different sucrose density fractions. In these preliminary experiments, the blots were cut and only the region where part of the channel protein was most likely to be found was probed, i.e., the 90- to 100-kDa region (35). B: single-channel records of ENaC activity from the 16% sucrose fraction of whole cell Triton X-100 extract. Top trace: bilayers were bathed with symmetrical solutions containing 200 mM Na-gluconate, 10 mM HEPES-Tris, pH 7.4. Bottom trace: 1 ␮M amiloride was added to both compartments of the bilayer chamber. Records were digitally filtered at 100 Hz using pCLAMP software. Horizontal markers at the right indicate closed state of the channel. Holding potential was ⫺100 mV.

the specific protein pattern obtained by immunoblotting of the 16% sucrose membrane-enriched fraction. The 97-kDa protein was again detected, along with a very faint band of 150 kDa relative mass (Mr). Similar Mr polypeptides were previously identified as components of the native Na⫹ channel (3, 15, 28, 35) and as related to its Na⫹ transport function (13, 26).

Fig. 2. A6 Na⫹ protein from apical membrane extract. Immunoblotting of the 16% sucrose gradient fraction with anti-␣-X. laevis ENaC antibody shows specific bands at Mr of 150 (very faint), 97, and 50 kDa. These bands were not detected when the blots were incubated with nonimmune serum in place of the primary antibody (data not shown). Due to loss during the membrane preparation and despite concentration of the samples on microconcentrators (Microcon, Amicon), protein concentration is 4.6 times less than in the corresponding lane of Fig. 1.

AJP-Renal Physiol • VOL

We also observed a 50-kDa band, previously identified with this anti-X. laevis epithelial Na⫹ channel (ENaC) antibody (15). Functionally, only the 16% sucrose fraction contained active amiloride-sensitive Na⫹ channels. In particular, similar to the observation in whole cell extracts, no amiloride-sensitive channel activity was found in the pellet. This confirms previous observations that once Na⫹ channel proteins are solubilized in Triton X-100, they lose their amiloride-binding and Na⫹-transporting properties (27). A representative example of the amiloride-sensitive Na⫹ channel activity from the 16% sucrose fraction is shown in the top trace of Fig. 3. Of nine successful incorporations of protein from 16% sucrose fractions, five bilayers contained single Na⫹ channels, and the remainder contained multiple channels. These channels had linear current voltage (I/V) relationships, with the slope conductances averaging 10.1 ⫾ 0.8 pS in the 200 mM Na-gluconate solution (Fig. 3B, open symbols, n ⫽ 9). Open times varied from tens of milliseconds to several seconds. The nature of the observed channels was confirmed by their inhibition by amiloride, which reduced the channel open probability by 90% at a concentration of 1 ␮M (Fig. 3A, bottom trace, n ⫽ 7). Amiloride-sensitive Na⫹ channels were the only channels observed in the 16% sucrose fraction in Na-gluconate buffer. We examined ion selectivity for two channels under bi-ionic conditions. Ion selectivity was calculated from the values of the reversal potentials (23.1 and 25.9 mV) that yielded a PNa⫹/PK⫹ selectivity coefficient of ⬃2.5 (Fig. 3B, closed symbols). This low selectivity was reported pre-

284 • JANUARY 2003 •

www.ajprenal.org


Fig. 3. Single-channel records of epithelial Na⫹ channel activity from the 16% sucrose fraction of apical membrane preparations. A: bilayers were bathed with symmetrical solutions containing 200 mM Na-gluconate, 10 mM HEPES-Tris, pH 7.4 (top trace), and 1 ␮M amiloride was added to both compartments of the bilayer chamber (bottom trace). Records were digitally filtered at 100 Hz using pCLAMP software. Horizontal markers at the right indicate closed state of the channel. Holding potential was ⫹100 mV. B: current-voltage (I/V) relationships in symmetrical and bi-ionic conditions. Data points represent single-channel current amplitudes in symmetrical solutions (200 mM Na-gluconate, open symbols, n ⫽ 9) and after the switch to bi-ionic conditions (200 mM Na-gluconate/200 mM K-gluconate, trans/cis, closed symbols, n ⫽ 2). In symmetrical solutions, the line represents linear regression of the data. In bi-ionic conditions, approximation to the GoldmanHodgkin-Katz current equation was drawn using SigmaPlot software.

viously for the native unsolubilized Na⫹ channel from A6 cells incorporated into planar lipid bilayers (29).1 To test the influence of lipid environment on channel function, we tried to reconstitute Na⫹ channels into bilayers composed of PC-cholesterol-sphingomyelin (1: 1:1 molar ratio), a lipid composition resembling that found in membrane microdomains shown in other systems to contain active proteins (see DISCUSSION). We did not observe the appearance of channel activity from 1 The low selectivity of reconstituted channels that we observed differs from the higher selectivity of endogenous channels found in epithelia studied by the path-clamp technique (7, 9). Recently, Jovov et al. (14) addressed these discrepancies, showing that the addition of the cytoskeletal protein actin to reconstituted rat ENaC activity resulted in the increase in Na⫹/K⫹ ion selectivity of the channel. We cannot exclude that Triton X-100 treatment and/or differential centrifigation affect cytoskeletal proteins from the channel complex, which, in turn, could result in the decrease in channel ion selectivity.


F185

membrane-enriched fractions in bilayers of this composition. However, this kind of highly packed bilayer may limit the fusion of proteoliposomes. Agents that modify the lipid composition of membrane microdomains have been shown to alter the function of associated proteins (21). To begin characterizing the lipid surroundings of the active channels, we tested the effects of fumonisin, an inhibitor of the biosynthetic pathway of sphingomyelin (16), and of cyclodextrin, an agent used to deplete the cells of cholesterol (4), on Na⫹ channel expression, transepithelial amiloride-sensitive Na⫹ currents, and single Na⫹ channel activity. A6 cells were exposed to 25 ␮M fumonisin and 5 mM 2-hydroxypropyl-␤-cyclodextrin before membrane preparation and sucrose floatation. Membrane lipid modification was assessed by the expression of caveolin, a glycolsylphosphatidylinositol (GPI)-anchored protein present in microdomains enriched in cholesterol and sphygomyelin (5). As shown in Fig. 4A, caveolin was detected in control membranes but not in membranes prepared from cells exposed to the lipid-modifying agents. By contrast, Na⫹ channel protein was still detected in fractions of 16 and 18–30% sucrose density (Fig. 4B). Amiloride-sensitive transepithelial Na⫹ currents were not affected by treatment of the cells with either fumonisin (25 ␮M) or cyclodextrin (5 mM) (Fig. 5A). When both agents were added together, currents dropped after 2 h of incubation. However, this drop could be attributed to a 30.3 ⫾ 6.5% drop in transepithelial resistance, indicating an effect of these drugs on the tight junctions rather than on the Na⫹ channels. This was confirmed by reconstitution of single-channel activity in lipid bilayers. Amiloride-sensitive channel activity was found only in the 16% sucrose density fraction containing membrane material prepared from cells exposed to fumonisin and cyclodextrin, with conductance and open probability similar to control channels (Fig. 5B). These data suggest that active Na⫹ channels cluster in membrane microdomains that do not depend on cholesterol and sphingomyelin.

Fig. 4. Effects of lipid-modifying agents on membrane proteins. A6 cells were exposed to 25 ␮M fumonisin for 68 h, together with 5 mM cyclodextrin in the last 2 h before membrane preparation. A: immunoblot of caveolin control (100 ␮g protein; lane 1) and treated membranes (125 ␮g protein; lane 2). B: immunoblot of Na⫹ channel protein isolated from membranes of treated A6 cells and floated on a 5–30% sucrose gradient. The 90- to 100-kDa region is shown. Sucrose densities are indicated at the bottom.

284 • JANUARY 2003 •

www.ajprenal.org

F186


Fig. 5. Effects of lipid-modifying agents on Na⫹ transport function. A: amiloride-sensitive transepithelial measurements of Na⫹ currents on A6 cells. Confluent monolayers were incubated with 5 mM cyclodextrin in water, 25 ␮M fumonisin in DMSO, or a combination of both for the times indicated. Control wells received both vehicles. Normalized data of short-ciruit current (ISC) calculated with respect to control values (5.33 ⫾ 0.21 ␮A/cm2) are shown. *P ⫽ 0.002 (n ⫽ 3, Student’s t-test). Cyclodextrin (10 mM) was also tested as well as fumonisin for up to 68 h without further change (data not shown). B: single-channel I/V relationships in symmetrical 200 mM Nagluconate. Data points represent single-channel current amplitudes. E, Control (␥ ⫽ 10.1 ⫾ 0.8 pS); {, experimental (␥ ⫽ 9.3 ⫾ 0.9) (P ⫽ 0.2, NS).

of membrane domains to solubilization in nonionic detergent conferred by an enriched lipid content results in a high buoyancy (5), which is consistent with our observation of channel protein in the low-density region of the sucrose gradient. The preservation of native protein-lipid interactions is important for the biological activity of the extracted proteins. In this respect, we previously found that solubilization of native A6 Na⫹ channels with cationic, anionic, or nonionic detergents destroys transport activity, while extracts obtained with a zwitterionic detergent contained functional Na⫹ channels (27). The zwitterionic detergents, which are the most efficient detergents in extracting active protein, have been shown to produce the highest solubilized lipid/protein ratios (2), while all hydrophobic detergents such as Tritons extract little protein-associated lipid, and their use results in poorly active (2) or inactive protein (11). Our data are consistent with the idea that native Na⫹ channels must be closely associated with native lipids in the membrane to sustain functional activity. Because this activity was observed with the cloned ␣-ENaC subunit alone as well as with all combinations of ␣- with ␤- and ␥-subunits (22), the absence of channel activity in heavier gradient fractions can be attributed to disruption of this association by Triton rather than to a change in subunit composition. In this regard, it was recently reported that in sucrose gradient fractions from A6 cells, ␣-, ␤-, and ␥-subunits of native channels were found to be associated.2 Although protein-lipid interactions are potentially important for regulating native Na⫹ channel function (18, 33), the results obtained with cloned ENaCs are variable. For example, expression of ENaC was not observed in membrane fractions resistant to Triton X-100 solubilization in MDCK cells (10), which contain microdomains (19), whereas in transfected COS-7 and HEK-293 cells, ENaC was shown to be transformed from a Triton X-100-soluble form in the endoplasmic reticulum to a Triton X-100-insoluble form during trafficking to the cell surface (23). Because these cells presumably transport Na⫹, these are apparently contradictory findings. However, Na⫹ channel function was not evaluated in either of those studies. The influence of the lipid environment on the function of membrane proteins was highlighted by the discovery of lipid rafts, which are particular membrane microdomains enriched in cholesterol and sphingomyelin. These liquid-ordered regions are insoluble in nonionic detergent (17). Such membrane regions have

DISCUSSION

The present study combines biochemical and functional analyses of native Na⫹ channel protein in Triton X-100-soluble and -insoluble fractions obtained from membranes of A6 renal epithelial cells in culture and shows that amiloride-sensitive Na⫹ channels, recorded in artificial bilayer membranes, are restricted to Triton X-100-resistant membrane microdomains. Resistance AJP-Renal Physiol • VOL

2 Recently, Hill WG, An B, and Johnson JP (Originally published August 6, 200210.1074/jbc.C200309200. J Biol Chem 277: 33541– 33544, 2002) have shown the discontinuous sucrose gradient centrifugation fractionation pattern of Na⫹ channels endogenously expressed in A6 cells in the absence of Triton X-100. ␣-, ␤-, and ␥-subunits of the native channels were found to localize in cholesterol-enriched regions of high buoyancy and to migrate in the sucrose gradient similarly to caveolin, suggesting raft localization. However, function was not tested in these fractions.

284 • JANUARY 2003 •

www.ajprenal.org


proven important for clustering of active proteins (25, 30) but also ion channels. In this regard, targeting of a functional isoform of the voltage-gated K⫹ channel Kv1.5 to distinct lipid rafts of transfected mouse L-cells was shown (20, 21). In that study, the soluble material represented mostly overexpressed intracellular channels. This is consistent with our observations that Na⫹ channel protein is abundantly present at the bottom of the gradients obtained with whole cell extracts. Because the amount of this soluble material is greatly reduced in the apical membrane-enriched preparations, it is likely to represent intracellular proteins, possibly in an immature form (15, 32). In the apical membrane extracts, the amount of channel protein in detergent-resistant fractions represents only a small part of the total membrane pool of these proteins. Quantitative analysis of transepithelial current vs. the number of protein molecules at the membrane led Firsov et al. (6) to suggest the presence of two pools of conducting channels at the apical membrane, a large pool of merely silent channels and a small pool of activated channels. Our results could provide a mechanistic interpretation for these observations. Compartmentalization of active Na⫹ channels within discrete regions of the apical cell membrane could explain the very low probability of finding channel activity in native or cultured cells using the patch-clamp method (7, 9) and could also reconcile the diverging biophysical properties of this channel in different tissues, species, or artificial membranes. Our negative results with agents known to modify the amount of cholesterol and sphingomyelin in the cell membrane indicate that the microdomains surrounding active Na⫹ channels in epithelia may be different from the classic rafts that consist of these molecules. In support of this conclusion, it was recently shown that some proteins cluster in membrane microdomains that do not depend on cholesterol and sphingomyelin content (34). In this regard, it was shown that at ambient temperature saturated PC alone can form lipid domains that are Triton insoluble (17). Regardless of whether specialized membrane regions or simply a close protein-lipid association is required for functional activity, future studies will attempt to define the lipid composition of the membrane fraction containing active Na⫹ channels as well as the effect of different lipid environments on native Na⫹ channel behavior to characterize further the regulatory role of the native membrane environment on Na⫹ transport function. We thank Dr. T. R. Kleyman for the generous gift of the anti␣xENaC antibody and Nancy Leclercq for technical assistance. V. G. Shlyonsky was the recipient of a “Subside a` savant” from the Universite´ Libre de Bruxelles. This research was supported by CER funds from Universite´ Libre de Bruxelles (to S. Sariban-Sohraby).


F187

REFERENCES 1. Alvarez de la Rosa D, Canessa CM, Fyfe GK, and Zhang P. Structure and regulation of amiloride-sensitive sodium channels. Annu Rev Physiol 62: 573–594, 2000. 2. Banerjee P, Joo JB, Buse JT, and Dawson G. Differential solubilization of lipids along with membrane proteins by different classes of detergents. Chem Phys Lipids 77: 65–78, 1995. 3. Benos DJ, Saccomani G, and Sariban-Sohraby S. The epithelial sodium channel subunit number and location of the amiloride binding site. J Biol Chem 262: 10613–10618, 1987. 4. Brown DA and London E. Structure and function of sphingolipid- and cholesterol-rich membrane rafts. J Biol Chem 275: 17221–17224, 2000. 5. Brown DA and Rose JK. Sorting of GPI-anchored proteins to glycolipid-enriched membrane subdomains during transport to the apical cell surface. Cell 68: 533–544, 1992. 6. Firsov D, Schild L, Gautschi I, Me´ rillat AM, Schneeberger E, and Rossier BC. Cell surface expression of the epithelial Na channel and a mutant causing Liddle syndrome: a quantitative approach. Proc Natl Acad Sci USA 93: 15370–15375, 1996. 7. Frings S, Purves RD, and Macknight ADC. Single-channel recordings from the apical membrane of the toad urinary bladder epithelial cell. J Membr Biol 106: 157–172, 1988. 8. Garty H and Palmer LG. Epithelial sodium channels: function, structure, and regulation. Physiol Rev 77: 359–396, 1997. 9. Hamilton KL and Eaton DC. Single-channel recordings from amiloride-sensitive epithelial sodium channel. Am J Physiol Cell Physiol 249: C200–C207, 1985. 10. Hanwell D, Ishikawa T, Saleki R, and Rotin D. Trafficking and cell surface stability of the epithelial Na⫹ channel expressed in epithelial Madin-Darby canine kidney cells. J Biol Chem 277: 9772–9779, 2002. 11. Hollerer-Beitz G and Heinemann SH. Influence of detergents on the function of cloned potassium channels. Receptors Channels 5: 61–78, 1998. 12. Horisberger JD. Amiloride-sensitive Na channels. Curr Opin Cell Biol 10: 443–449, 1998. 13. Ismailov II, McDuffie JH, Sariban-Sohraby S, Johnnson JP, and Benos DJ. Carboxyl methylation activates purified renal amiloride-sensitive Na⫹ channels in planar lipid bilayers. J Biol Chem 269: 22193–22197, 1994. 14. Jovov B, Tousson A, Ji HL, Keeton D, Shlyonsky V, Ripoll PJ, Fuller CM, and Benos DJ. Regulation of epithelial Na⫹ channels by actin in planar lipid bilayers and in the Xenopus oocyte expression system. J Biol Chem 274: 37845–37854, 1999. 15. Kleyman TR, Zuckerman JB, Middleton P, McNulty KA, Hu X, An B, Eaton DC, and Smith PR. Cell surface expression and turnover of the ␣-subunit of the epithelial sodium channel. Am J Physiol Renal Physiol 281: F213–F221, 2001. 16. Ledesma MD, Simons K, and Dotti CG. Neuronal polarity: essential role of protein-lipid complexes in axonal sorting. Proc Natl Acad Sci USA 95: 3966–3971, 1998. 17. London E and Brown DA. Insolubility of lipids in Triton X-100: physical origin and relationship to sphingolipid/cholesterol membrane domains (rafts). Biochim Biophys Acta 1508: 182–195, 2000. 18. Ma HP, Saxena S, and Warnock DG. Anionic phospholipids regulate native and expressed epithelial sodium channel (ENaC). J Biol Chem 277: 7641–7644, 2002. 19. Mamdouh Z, Giocondi MC, and Le Grimellec C. In situ determination of intracellular membrane physical state heterogeneity in renal epithelial cells using fluorescence ratio microscopy. Eur Biophys J 27: 341–351, 1998. 20. Martens JR, Navarro-Polanco R, Coppock EA, Nishiyama A, Parshley L, Grobaski TD, and Tamkun MM. Differential targeting of Shaker-like potassium channels to lipid rafts. J Biol Chem 275: 7443–7446, 2000. 21. Martens JR, Sakamoto N, Sullivan SA, Grobaski TD, and Tamkun MM. Isoform-specific localization of voltage-gated K⫹ channels to distinct lipid raft populations. Targeting of Kv15 to caveolae. J Biol Chem 276: 8409–8414, 2001.

284 • JANUARY 2003 •

www.ajprenal.org

F188


22. McNicholas CM and Canessa CM. Diversity of channels generated by different combinations of epithelial sodium channel subunits. J Gen Physiol 109: 681–692, 1997. 23. Prince LS and Welsh MJ. Effect of subunit composition and Liddle’s syndrome mutations on biosynthesis of ENaC. Am J Physiol Cell Physiol 276: C1346–C1351, 1999. 24. Rotin D. Regulation of the epithelial sodium channel (ENaC) by accessory proteins. Curr Opin Nephrol Hypertens 9: 529–534, 2000. 25. Rozelle AL, Machesky LM, Yamamoto M, Driessens MH, Insall RH, Roth MG, Luby-Phelps K, Marriott G, Hall A, and Yin HL. Phosphatidylinositol 4,5-bisphosphate induces actin-based movement of raft-enriched vesicles through WASPArp2/3. Curr Biol 10: 311–320, 2000. 26. Sariban-Sohraby S, Abramow M, and Fisher RS. Singlechannel behavior of a purified epithelial Na⫹ channel subunit that binds amiloride. Am J Physiol Cell Physiol 263: C1111– C1117, 1992. 27. Sariban-Sohraby S and Benos DJ. Detergent solubilization, functional reconstitution, and partial purification of epithelial amiloride-binding protein. Biochemistry 25: 4639–4646, 1986. 28. Sariban-Sohraby S, Fisher RS, and Abramow M. Aldosterone-induced and GTP-stimulated methylation of a 90-kDa


29.

30. 31. 32. 33. 34. 35.

polypeptide in the apical membrane of A6 epithelia. J Biol Chem 268: 26613–26617, 1993. Sariban-Sohraby S, Latorre R, Burg M, Olans L, and Benos DJ. Amiloride-sensitive epithelial Na⫹ channels reconstituted into planar lipid bilayer membranes. Nature 308: 80–82, 1984. Simons K and Toomre D. Lipid rafts and signal transduction. Nat Rev Mol Cell Biol 1: 31–39, 2000. Simons K and Van Meer G. Lipid sorting in epithelial cells. Biochemistry 27: 6197–6202, 1988. Valentijn JA, Fyfe GK, and Canessa CM. Biosynthesis and processing of epithelial sodium channels in Xenopus oocytes. J Biol Chem 273: 30344–30351, 1998. Yue G, Malik B, Yue G, and Eaton DC. Phosphatidylinositol 4,5-bisphosphate (PIP2) stimulates epithelial sodium channel activity in A6 cells. J Biol Chem 277: 11965–11969, 2002. Zacharias DA, Violin JD, Newton AC, Tsien RY. Partitioning of lipid-modified monomeric GFPs into membrane microdomains of live cells. Science 296: 913–916, 2002. Zuckerman JB, Chen X, Jacobs JD, Hu B, Kleyman TR, and Smith PR. Association of the epithelial sodium channel with Apx and alpha-spectrin in A6 renal epithelial cells. J Biol Chem 274: 23286–23295, 1999.

284 • JANUARY 2003 •

www.ajprenal.org