Anionic phospholipids regulate native and expressed epithelial sodium channel (ENaC).

Using patch clamp techniques, we found that the epithelial sodium channel (ENaC) activity in the apical membrane of A6 distal nephron cells showed a sudden rundown beginning at 4 min after forming the inside-out configuration. This sudden rundown was prevented by addition of anionic phospholipids such as phosphatidylinositol 4,5-bisphosphate (PIP(2)), phosphatidylinositol 3,4,5-trisphosphate (PIP(3)), and phosphatidylserine (PS) to the "cytoplasmic" bath. Conversely, chelation of endogenous PIP(2) with anti-PIP(2) antibody, hydrolysis of PIP(2) with either exogenous phospholipase C (PLC) or activation of endogenous PLC by extracellular ATP, or application of the positively charged molecule, poly-L-lysine, accelerated channel rundown. However, neutral phosphatidylcholine had no effect on ENaC activity. By two-electrode voltage clamp recordings, we demonstrated that PIP(2) and PIP(3) significantly increased amiloride-sensitive current in Xenopus oocytes injected with cRNAs of rat alpha-, beta-, and gamma-ENaC. However, PIP(2) and PIP(3) did not affect surface expression of ENaC, indicating that PIP(2) and PIP(3) regulate ENaC at the level of the inner plasma membrane through a mechanism that is independent of ENaC trafficking. These data suggest that anionic phospholipids may mediate the regulation of ENaC by PLC- or phosphoinositide 3-kinase-coupled receptors.

Using patch clamp techniques, we found that the epithelial sodium channel (ENaC) activity in the apical membrane of A6 distal nephron cells showed a sudden rundown beginning at 4 min after forming the inside-out configuration. This sudden rundown was prevented by addition of anionic phospholipids such as phosphatidylinositol 4,5-bisphosphate (PIP 2 ), phosphatidylinositol 3,4,5-trisphosphate (PIP 3 ), and phosphatidylserine (PS) to the "cytoplasmic" bath. Conversely, chelation of endogenous PIP 2 with anti-PIP 2 antibody, hydrolysis of PIP 2 with either exogenous phospholipase C (PLC) or activation of endogenous PLC by extracellular ATP, or application of the positively charged molecule, poly-Llysine, accelerated channel rundown. However, neutral phosphatidylcholine had no effect on ENaC activity. By two-electrode voltage clamp recordings, we demonstrated that PIP 2 and PIP 3 significantly increased amiloride-sensitive current in Xenopus oocytes injected with cRNAs of rat ␣-, ␤-, and ␥-ENaC. However, PIP 2 and PIP 3 did not affect surface expression of ENaC, indicating that PIP 2 and PIP 3 regulate ENaC at the level of the inner plasma membrane through a mechanism that is independent of ENaC trafficking. These data suggest that anionic phospholipids may mediate the regulation of ENaC by PLC-or phosphoinositide 3-kinase-coupled receptors.
The phospholipid compositions of the two lipid bilayer leaflets of the plasma membrane are strikingly different. Anionic phospholipids are normally located in the inner leaflet to form a negatively charged surface. However, whether the phospholipid asymmetry affects the function of membrane proteins remains largely unknown. Previous studies have shown that one of the anionic phospholipids, phosphatidylinositol 4,5bisphosphate (PIP 2 ), 1 regulates Na ϩ -Ca 2ϩ exchangers and ATP-sensitive potassium (K ATP ) channels (1,2). Convincing evidence suggests that PIP 2 directly interacts with the proximal COOH terminus of inward-rectifier K ϩ channels (3). Not only PIP 2 , but also other negatively charged phospholipids such as phosphatidylinositol 4-phosphate (PI-4-P) and phosphatidylinositol 3,4,5-trisphosphate (PIP 3 ), regulate K ATP channels (4,5). A model for the regulation of K ATP channels by anionic phospholipids has been proposed, which argues that the negatively charged head group of PIP 2 , PI-4-P, or PIP 3 locks the positively charged carboxyl terminus of K ATP channels at a certain position, resulting in the failure of ATP binding to the terminus (6). This model raises an interesting question: can anionic phospholipids interact with the positively charged cytoplasmic termini of other ion channels?
The epithelial sodium channel (ENaC) plays a very important role in regulating total body Na ϩ homeostasis. Recent studies suggest that PIP 2 stimulates ENaC in A6 cells (7) and that a decrease in PIP 2 concentration may account for the inhibition of ENaC by luminal purinergic P 2Y receptors (8). It is known that ENaC consists of three subunits designated ␣, ␤, and ␥ (9). By examining the first 50 amino acids of the NH 2terminal tails of ␣-, ␤-, and ␥-ENaC, we found that the NH 2terminal tails of ␤and ␥-ENaC, but not of ␣-ENaC, contain significant numbers of positive charges. In fact, the P 3 geKiKaKiKKnL 15 sequence in the ␥ subunit NH 2 terminus is very similar to the pleckstrin homology domain in PLC-␦ 1 (10). We hypothesize that these positive charges might interact with anionic phospholipids of the inner leaflet of the plasma membrane to modulate ENaC activity. Previous studies have shown that deletion of the NH 2 -terminal tails of ␤-ENaC (⌬2-49) and ␥-ENaC (⌬2-53), but not the NH 2 -terminal tail of ␣-ENaC (⌬2-46), dramatically reduces ENaC activity (11), suggesting that the positively charged NH 2 -terminal tails of ␤and ␥-ENaC play an important role in regulating ENaC activity. It has been shown that positively charged poly-L-lysine partially reversed the effect of PIP 2 on K ATP channels (12), indicating that positively charged agents may compete with the positively charged COOH terminus of K ATP channels for binding to PIP 2 . Similarly, the positively charged NH 2 -terminal tails of ␤and ␥-ENaC could be physically "locked" by negatively charged phospholipids to the inner surface of the plasma membrane. This putative interaction may account for the role of the NH 2terminal tails of ␤and ␥-ENaC in regulation ENaC activity (11). Therefore, the present study aims to determine whether anionic phospholipids such as PIP 2 , PIP 3 , and PS regulate ENaC activity.

EXPERIMENTAL PROCEDURES
Cell Culture-A6 distal nephron cells were purchased from American Type Culture Collection (Rockville, MD) at passage 68. The cells were cultured in a plastic flask in a modified NCTC-109 medium (Invitrogen) with 10% fetal bovine serum (Invitrogen) and 1.5 M aldosterone (Sigma) at 26°C and 4% CO 2 . Cells from passages 72-82 were removed from the flasks and plated on permeable supports attached to Snapwell inserts from Corning Costar Co. The permeable supports were coated with rat-tail collagen according to the protocol that is used by Corning Costar Co. The cells were cultured on permeable supports for 10 -14 days before patch clamp recordings, as we reported previously (13).
Chemicals and Solutions-Most chemicals, including phosphatidylinositol-specific PLC, adenosine 5Ј-triphosphate, PS, phosphatidylcho-* This work was supported by a National Kidney Foundation Young Investigator Award (to H.-P. M.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Patch Clamp Inside-Out Recordings-Immediately before use, a Snapwell insert was thoroughly washed with NaCl bath solution (see "Chemicals and Solutions") and transferred into the patch chamber mounted in the stage of a Leitz inverted microscope. Using patch clamp techniques, inside-out recordings were established on the apical membrane of A6 cells with polished micropipettes with tip resistance of 2.5-5 megaohms. Under the above culture conditions, a patch seal (seal resistance Ͼ 20 gigaohms) was usually formed after releasing positive pressure in the patch pipette or after applying a slightly negative pressure. After establishing the cell-attached mode, only patches containing channel activity without base-line drift were used for experiments. Before forming inside-out patches, NaCl bath solution in the patch chamber was replaced with KCl bath solution. Single-channel currents were obtained with ϩ40-mV applied pipette potential (i.e. V m ϭ Ϫ40 mV), filtered at 1 kHz, and recorded on video tapes with a modified Sony PCM video converter (Vetter Instruments). Before digitization with pClamp 8 software (Axon Instruments), single-channel records were low-pass filtered at 100 Hz. The total numbers of functional channels (N) in the patch were estimate by observing the number of peaks detected on the current amplitude histograms. As a measure of channel activity, NP o (number of channels ϫ the open probability, P o ) was calculated by using at least 2 min of a single-channel record as we described previously (13). Experiments were conducted at 22-23°C.
Two-electrode Voltage Clamp Recordings-Oocytes were excised from adult female Xenopus frogs and treated with collagenase. Stage V-VI oocytes were injected with cRNAs for wild-type ␣-, ␤-, and ␥-ENaC subunits and then were incubated at 18°C in modified Leibovitz medium. Electrophysiological recordings were performed 24 -48 h after the injections using two microelectrodes filled with 3 mM KCl and inserted into the oocyte, as we described previously (14). A voltage step protocol from Ϫ120 to ϩ40 mV in increments of 20 mV was used. Between voltage steps the membrane was voltage-clamped at a holding potential of Ϫ40 mV. The macroscopic ENaC currents were verified by application of 10 M amiloride to the bath. The net amiloride-sensitive currents were used to represent ENaC activity. After recording control ENaC currents, the oocytes were taken out of the chamber and injected with 1 l of H 2 O, PIP 2 (30 M), or PIP 3 (30 M), respectively. Phosphatase inhibitor mixture (2 M) was included in each injection. 30 min after these injections, amiloride-sensitive currents were re-measured in these oocytes and compared with the currents before these injections.
Evaluation of ENaC Surface Expression by Confocal Microscopy-Using confocal microscopy, the surface expression of ENaC after each experimental manipulation was evaluated; rat ␤ and ␥-ENaC subunits were tagged in the extracellular loops with the FLAG TM epitope (DYK-DDDDK), which can be recognized by M2 monoclonal antibody, as described previously (15). The FLAG-tagged ␤-, ␥-, and ␣-ENaC cRNAs were injected into Xenopus oocytes. Fluorescent imaging analysis of the expression level by confocal microscopy was carried out, as we described previously (16). The oocytes were then secondarily injected with H 2 O, PIP 2 , or PIP 3 as described above. The effects of PIP 2 and PIP 3 on ENaC surface expression were evaluated using the confocal fluorescent imaging methods.
Statistical Analysis-A paired t test or analysis of variance for multiple comparisons was used for statistical analysis, as we described previously (13). A p value less than 0.05 was considered significant.

RESULTS AND DISCUSSION
A Decrease in PIP 2 Concentration Appears to Mediate Inhibition of ENaC by the P2Y 2 Receptor-The G protein-coupled P2Y 2 receptor is expressed in renal epithelial cells (17). We have found that ATP inhibits ENaC via a PLC-dependent pathway in A6 cells (8). It is well known that activation of PLC hydrolyzes PIP 2 to generate IP 3 and diacylglycerol and subsequently mobilizes [Ca 2ϩ ] i . However, recent studies suggest that inhibition of Na ϩ absorption by the P2Y 2 receptor occurs independently of an increase in [Ca 2ϩ ] i (18). Therefore, we hypothesize that a decrease in PIP 2 concentration might mediate the P2Y 2 receptor-induced inhibition of ENaC. To test this hypoth-esis, inside-out patch experiments were performed as shown in Fig. 1. We found that ENaC activity in inside-out patches was stable for the initial 4 min. However, a sudden rundown occurred during the period from 4 to 5 min. Interestingly, the channel rundown was clearly prevented when the "cytoplasmic" bath contained 5 M PIP 2 . In contrast, application of 100 nM anti-PIP 2 antibody to the cytoplasmic bath to chelate endogenous PIP 2 significantly accelerated the rundown process. Application of exogenous PLC (0.5 unit/ml) to the cytoplasmic bath, which could hydrolyze PIP 2 , also accelerated the rundown. Furthermore, application of 100 M ATP in the patch pipette, which presumably activates endogenous PLC, reduced ENaC activity as we recently observed in cell-attached patches (8). The initial values of NP o in the ATP experiments were much lower than the values in other group experiments. We argue that the inhibition of ENaC already occurred before forming the inside-out configuration, because the effect of ATP occurred when the patch pipette was attached to the cell membrane. These data suggest that a decrease in PIP 2 concentration at the inner membrane leaflet appears to mediate the inhibition of ENaC by the P2Y 2 receptor.

An Increase in PIP 3 Concentration May Mediate Stimulation of ENaC by Corticoid Receptors and
Insulin-It is known that K ATP channels are not only regulated by PIP 2 , but also by PIP 3 (4). However, the role of PIP 3 has been neglected, because the plasma membrane does not contain PIP 3 under normal conditions. Nevertheless, PIP 3 can be generated by activation of phosphoinositide 3-kinase (PI 3-kinase). Interestingly, recent studies have shown that both aldosterone and insulin enhance Na ϩ transport by activating PI 3-kinase in A6 cells and that inhibition of PI 3-kinase will block their stimulatory effect on ENaC activity (19 -21). To test whether PIP 3 could affect ENaC activity, the inside-out patch configuration was used. Consist- Regulation of ENaC by Anionic Phospholipids 7642 ent with the results as shown in Fig. 1, ENaC activity in inside-out patches was steady during the initial 4 -5 min before a sudden rundown occurred. In contrast, the channel activity was maintained without rundown when the cytoplasmic bath contained 5 M PIP 3 (Fig. 2). Because the concentration of PIP 3 is elevated in response to aldosterone (19), the effect of PIP 3 on ENaC activity may account in part for the regulation of ENaC by aldosterone. To test whether other anionic phospholipids can also regulate ENaC activity, inside-out patches were examined when negatively charged PS (20 M) was applied to the cytoplasmic bath. Similar to the effect of PIP 2 and PIP 3 , anionic PS also prevented ENaC rundown, suggesting that the effect of phospholipids on ENaC activity may be related to their anionic composition. To test whether negative charges are important for the effect of PIP 2 , PIP 3 , and PS on ENaC activity, the effect of the positively charged molecule, poly-L-lysine, on ENaC activity was examined. It appears that addition of poly-L-lysine (10 g/ml) accelerated ENaC rundown. However, neutral PC had no effect on ENaC activity (Fig. 2). These data suggest that an increase in PIP 3 concentration may mediate stimulation of ENaC by corticoid receptors and insulin at the level of interaction of the ENaC complex with the inner leaflet of the plasma membrane.
PIP 2 and PIP 3 Enhance ENaC Current in Xenopus Oocytes-In addition to A6 cells that natively express ENaC when conditioned by aldosterone, the Xenopus oocyte system was also used to test the role of PIP 2 and PIP 3 in regulating ENaC activity. Using two-electrode voltage clamp techniques exogenously expressed ENaC activity was evaluated with amiloride-sensitive currents following injection of rat ␣-, ␤-, and ␥-ENaC cRNAs. Amiloride-sensitive currents were compared in the same oocyte before and 30 min after injection of equal volume of H 2 O (as a control), PIP 2 (30 M), or PIP 3 (30 M), respectively. Amiloride-sensitive currents were not changed in the oocytes injected with H 2 O (Ϫ728 Ϯ 84 nA versus Ϫ750 Ϯ 72 nA; n ϭ 7). In contrast, amiloride-sensitive currents were increased, from Ϫ797 Ϯ 40 nA to Ϫ1091 Ϯ 69 nA (p Ͻ 0.001; n ϭ 14) after injection with PIP 2 and from Ϫ733 Ϯ 31 nA to Ϫ1077 Ϯ 73 nA (p Ͻ 0.001; n ϭ 10) after injection with PIP 3 . To further determine whether the increase in amiloride-sensitive currents were related to ENaC trafficking, the density of ENaC on the surface of the plasma membrane was evaluated with confocal surface labeling techniques. The oocytes that expressed FLAG TM -tagged ENaC were injected with equal volume of H 2 O (as a control), PIP 2 (30 M), or PIP 3 (30 M), respectively. Fluorescent labeling was carried out 30 min after these injections. The data demonstrated that there was no difference in ENaC surface density between each group of oocytes injected with H 2 O, PIP 2 , or PIP 3 (Fig. 3), indicating that PIP 2 and PIP 3 up-regulate ENaC through a mechanism that appears to be independent of ENaC trafficking as it affects the density of surface expression. Although PIP 2 , PIP 3 , and PS failed to enhance ENaC activity in A6 cells, but only maintain the channel activity, it is likely that the stimulatory effect of anionic phospholipids on ENaC activity may be already saturated in A6 cells, which are continuously cultured in the presence of aldosterone. Further experiments will address this hypothesis by using ENaC-expressing renal epithelial cells cultured either in the absence or in the presence of aldosterone. Presumably, without prestimulation by aldosterone, anionic phospholipids will increase the low basal level of ENaC activity in such cells.  A, summary plots of amiloride-sensitive currents before (blank bars) and after injections with H 2 O, PIP 2 , or PIP 3 (hatched bars). Oocytes used in this set of experiments were injected with cRNAs encoding rat ␣-, ␤-, and ␥-ENaC subunits. Two-electrode voltage clamp recordings were carried out during the period of 24 -48 h after the injection. Amiloride-sensitive currents were the currents at a potential of Ϫ100 mV under control conditions subtracted by the currents after addition of 10 M amiloride. After initial recordings of amiloride-sensitive currents as a control, the oocytes were then injected with H 2 O, PIP 2 , or PIP 3 . The secondary recordings were performed 30 min after these injections. B, no obvious change in ENaC surface expression after injection of PIP 2 or PIP 3 . Oocytes used in this set of experiments were injected with cRNAs encoding FLAG-tagged rat ␣-, ␤-, and ␥-ENaC subunits. After ENaC was significantly expressed (24 -48 h after injection with cRNAs of ENaC), the oocytes were then injected with H 2 O, PIP 2 , or PIP 3 . Fluorescent labeling was performed 30 min after the secondary injections.

Regulation of ENaC by Anionic Phospholipids 7643
Conclusion and Potential Significance-We (8) and others (18) have recently demonstrated that stimulation of the P 2Y family, probably the P2Y 2 receptor, inhibits ENaC activity in A6 distal nephron cells and amiloride-sensitive I sc in mouse cortical collecting duct principal cells via a pathway that appears to occur independently of an increase in [Ca 2ϩ ] i . The present study demonstrates that anionic phospholipids activate endogenously expressed ENaC in A6 cells and exogenously expressed ENaC in Xenopus oocytes. Since both chelation of endogenous PIP 2 with anti-PIP 2 antibody and hydrolysis of endogenous PIP 2 with exogenous PLC or extracellular ATP that presumably activates endogenous PLC could reduce ENaC activity (Fig. 1), a decrease in PIP 2 concentration in the inner leaflet of the plasma membrane may explain inhibition of ENaC by the P2Y 2 Receptor. In addition, we have found that PIP 3 also regulates ENaC activity (Fig. 2). With the recognition of the role of PIP 3 and PI 3-kinase in the responses to aldosterone and insulin on ENaC activity in A6 cells (19 -21), and the recent recognition of other phosphatidylinositol kinases (22), the role of anionic phospholipids in the tonic regulation of ENaC activity at the level of the plasma membrane may well be of general importance. The response to aldosterone is pleotropic and involves sgk kinase as well as changes in PI 3-kinase (23)(24)(25)(26). It appears that aldosterone-mediated activation of sgk kinase rapidly stimulates translocation of ENaC to the apical membrane (27), while the experiments described in the legend to Fig. 3, using anionic phospholipids as the putative downstream effectors of the aldosterone response, demonstrate activation of ENaC in situ rather than recruitment or translocation of ENaC complexes to the plasma membrane in the oocyte system. Therefore, an increase in PIP 3 concentration in the inner plasma membrane may account in part for the stimulatory effects of aldosterone and insulin on ENaC activity at the level of the inner leaflet of the plasma membrane.