Abstract
Epithelial Na+ channels (ENaC) coexist with a family of ATP-gated ion channels known as P2X receptors in the renal collecting duct. Although ENaC is itself insensitive to extracellular ATP, tubular perfusion of ATP can modify the activity of ENaC. To investigate a possible regulatory relationship between P2X receptors and ENaC, coexpression studies were performed in Xenopus oocytes. ENaC generated a persistent inward Na+ current that was sensitive to the channel blocker amiloride (Iam-s). Exogenous ATP transiently activated all cloned isoforms of P2X receptors, which in some cases irreversibly inhibited Iam-s. The degree of inhibition depended on the P2X receptor subtype present. Activation of P2X2, P2X2/6, P2X4, and P2X4/6 receptor subtypes inhibited Iam-s, whereas activation of P2X1, P2X3, and P2X5 receptors had no significant effect. The degree of inhibition of Iam-s correlated positively with the amount of ionic charge conducted by P2X receptor subtypes. ENaC inhibition required Na+ influx through Iam-s-inhibiting P2X ion channels but also Ca2+ influx through P2X4 and P2X4/6 ion channels. P2X-mediated inhibition of Iam-s was found to be due to retrieval of ENaC from the plasma membrane. Maximum amplitudes of ATP-evoked P2X-mediated currents (IATP) were significantly increased for P2X2, P2X2/6, and P2X5 receptor subtypes after coexpression of ENaC. The increase in IATP was due to increased levels of plasma membrane–bound P2X receptor protein, suggesting that ENaC modulates protein trafficking. In summary, ENaC was downregulated by the activation of P2X2, P2X2/6, P2X4, and P2X4/6 receptors. Conversely, ENaC increased the plasma membrane expression of P2X2, P2X2/6, and P2X5 receptors.
The amiloride-sensitive epithelial sodium channel (ENaC) largely determines the extent of Na+ reabsorption from tubular fluid along the distal nephron and consequently is crucial for the maintenance of whole-body Na+ balance and regulation of extracellular fluid volume (1). The physiologic importance of ENaC is exemplified by several inheritable human mutations of the genes encoding ENaC subunits that lead to abnormal Na+ balance and altered BP (2). There has been rapid progress in understanding the regulation of ENaC activity since the constituent subunits were cloned (3). ENaC is composed of three nonidentical but homologous subunits (α, β, and γ) that form a tetrameric structure with a stoichiometry of two α-, one β-, and one γ-subunit(s) (4). Expression of these ENaC subunits in Xenopus laevis oocytes results in a persistent Na+ conductance pathway that corresponds to the biophysical and pharmacologic properties of native epithelial Na+ channels present on the apical membrane of the collecting duct (CD) (4).
It is well documented that several hormones can regulate ENaC activity (most notably, aldosterone and vasopressin). In addition, since the advent of cloning and expression studies, it has been demonstrated that ENaC is regulated by a number of intracellular signaling pathways activated by intra- and extracellular factors (e.g., Na+ itself, free Ca2+, Cl−, H+), second messengers (cAMP and protein kinases), and regulatory proteins (G proteins, Nedd4, certain proteases, and actin filaments) (4–6). These signals control ENaC activity via a number of diverse mechanisms, including changes in the kinetics of channel opening and closing, channel synthesis, regulation of intracellular channel trafficking to the plasma membrane, membrane insertion, and channel endocytosis (7). It is interesting that a number of factors that are known to regulate ENaC activity—such as changes in intracellular cations, changes in cAMP levels, and the activation of regulatory G proteins—can be modulated by nucleotide receptors (P2 receptors) (8).
There are two classes of P2 receptors: P2X and P2Y (8). The P2X class consists of seven cloned subunits, which form seven trimeric nucleotide-gated homomeric ion channels (P2X1 to 7) and five established heteromeric ion-channel assemblies (P2X1/2,1/5,2/3,2/6,4/6). P2X receptors are permeable to Na+, K+, Ca2+, and, in some cases, Cl− and on activation cause cell depolarization (9–12). The P2Y class consists of eight G protein–coupled nucleotide-activated receptors (P2Y1,2,4,6,11 to 14),which can mobilize intracellular Ca2+, alter intracellular cAMP levels, or release their regulatory βγ subunits (8). Both P2X and P2Y receptor subtypes are expressed in the nephron (13), and both the variety and the abundance of P2X receptors are highest in the distal nephron. Multiple P2X receptors (P2X1 to P2X7) have been molecularly identified in cell models of the distal nephron (14,15), and our own studies in native rat tissue have immunolocalized P2X4, P2X5, and P2X6 receptor subunits to the distal tubule and CD principal cells (16), where they may co-localize with ENaC in the apical membrane.
Extracellular ATP can modify ENaC activity, although ENaC itself is insensitive to extracellular ATP (17). Consequently, it has been proposed that ATP must act indirectly to regulate ENaC by activating cell surface P2 receptors (16,18). This theory has been supported by a number of studies demonstrating P2 receptor–mediated inhibition of sodium reabsorption in native CD epithelia (19,20) and renal epithelial cell lines (14,21,22). ATP-evoked inhibition of Na+ reabsorption in the distal nephron is usually thought to be the consequence of stimulation of P2Y receptors (23). However, two separate studies have inferred that apical P2X receptors may mediate ENaC inhibition. First, a study on mIMCD-K2 cells (an immortalized CD cell line) described inhibition of Na+ absorption (and stimulation of Cl− secretion) after activation of P2X3 and P2X4 receptors (14). Second, we have proposed P2X receptor–mediated inhibition of ENaC in an in vivo microperfusion study of the rat CD, in which 22Na+ recovery was measured to assess Na+ reabsorption after P2 receptor activation (19). We showed that a nonhydrolyzable analogue of ATP, ATPγS, inhibited Na+ reabsorption in Na+-restricted animals, whereas other nonhydrolyzable analogues had no measurable effect on 22Na+ uptake. Investigations into the selectivity and the potency of this series of analogues on recombinant P2 receptors corresponding to those found in the rat CD, expressed in Xenopus oocytes, suggested that P2X4/6 receptors mediated the inhibition of Na+ reabsorption (24).
Furthermore, it has been demonstrated that ENaC can stimulate an increase in the surface expression of certain ion channels in the kidney, namely cystic fibrosis transmembrane conductance regulator (CFTR) chloride channels and ROMK1 potassium channels (25). Thus, the possibility exists that ENaC may also increase expression levels of P2X receptors.
The molecular mechanisms by which P2X receptors may regulate ENaC activity and, conversely, the mechanisms by which ENaC might influence P2X receptor activity are unknown. These issues have been addressed in the present study by performing coexpression experiments in the Xenopus laevis oocyte recombinant model, using two-electrode voltage-clamp recordings in combination with measurements of surface expression of ENaC and P2X ion channels.
Materials and Methods
Oocyte Preparation
Xenopus laevis frogs were anesthetized in tricaine (0.4% wt/vol) and killed by decapitation, and their ovarian lobes were removed surgically. Stage V and VI oocytes were defolliculated by a two-step process involving collagenase treatment (type IA, 1.5 mg/ml in a Ca2+-free Ringer solution, for 2 h) followed by mechanical stripping away of the follicular layer by fine forceps. Defolliculated oocytes were injected with 50 nl of a cRNA/sterile water mixture or cRNA/cRNA combination (50%/50%) in the case of heteromeric assemblies (total cRNA quantities: rat α-, β-, γ-ENaC, 10 ng of each subunit; rat P2X1, P2X3, P2X4, P2X5, P2X6, 50 ng; rat P2X2, 20 ng; rat P2X2/6 and rat P2X4/6, 50 ng of each subunit). Injected oocytes were incubated for 24 h at 18°C in a low-Na+ modified Barth’s solution that contained (in mM) 1 NaCl, 60 N-methyl-D-glutamine-Cl (NMDG), 40 KCl, 0.8 MgSO4, 0.3 Ca(NO3)2, 0.4 CaCl2, and 10 HEPES-NMDG (pH 7.4) and was supplemented with 50 μg/L gentamicin sulfate, then kept at 4°C for up to 5 d until used in electrophysiologic experiments. All experiments conformed to local and British Home Office regulations with regard to the Animal Scientific Procedures Act (1986).
Two-Electrode Voltage Clamp
Oocytes were studied using the two-electrode voltage-clamp technique. Oocytes were clamped at a holding potential (Vh) of −100 mV, with two exceptions: (1) In initial experiments to determine the IC50 value for amiloride (the concentration of amiloride that blocked ENaC-mediated currents by 50%), in which the Vh was clamped at −10 mV in the absence of amiloride (20 min), to avoid channel rundown, and −100 mV while recording in the presence of amiloride (40 s; Figure 1A); and (2) in experiments to determine the effects of changing cytosolic Cl− levels on P2X-mediated inhibition of amiloride-sensitive currents (Iam-s; by incubating oocytes in low extracellular Cl− Ringer solution supplemented with 10 μM amiloride [low [Cl−]int] for 72 h), in which the Vh was clamped at −30 mV because of the poor health of the oocytes. The voltage-recording and current-recording microelectrodes (1 to 5 MΩ tip resistance) were filled with 3.0 M KCl. Oocytes were superfused with a Ringer solution (10 ml/min, 18°C) that contained (in mM) 100 NaCl, 3.0 KCl, 10 HEPES, and 1.8 CaCl2 (pH 7.4), by a gravity-fed continuous flow system that allowed rapid drug addition and washout. In ion-substitution experiments, Na+, Cl−, or Ca2+ concentrations of the Ringer solution were altered as stated in the text and in Table 1. The Iam-s was determined by subtracting the value measured in the absence of 10 μM amiloride (90 s) from that measured in the presence of amiloride in the bathing Ringer solution, with a rest time of 20 min after each removal of amiloride. ATP-activated currents (IATP) were recorded after addition of agonist for 90 s or until the current reached a peak, then washed for a period of 20 min for all of the P2X receptors tested. Electrophysiologic data were stored on a computer using an MP100 WSW interface (Biopac Systems, Goleta, CA) and analyzed (including area-under-the-curve calculations) using the software package Acqknowledge III (Biopac Systems).
Amiloride-sensitive epithelial Na+ channel (ENaC)-mediated currents are inhibited by the activation of ATP-gated P2X2 receptors. (A) Inhibition curve for amiloride blockade of ENaC-mediated currents in oocytes that expressed ENaC (n = 3). (B) Typical trace showing consistent ENaC-mediated amiloride-sensitive currents (Iam-s) evoked by the removal of IC100 concentration of amiloride (10 μM). Oocytes were voltage clamped at −100 mV (holding potential [Vh]), and 20 min was left between evoked currents. (C) A typical trace showing an oocyte coexpressing ENaC and P2X2 receptors, in which application of ATP (30 μM) for 120 s in the presence of amiloride (10 μM) evokes a P2X2-mediated current (ATP-activated current [IATP]). Subsequent removal of amiloride results in an inhibition of ENaC-mediated currents (Iam-s), irreversible over a 2-h period.
Solutions used for oocyte current measurements
Detection of Cell Surface Expression by Chemiluminescence
For binding experiments, α, β, and γ rat ENaC subunits were tagged, as described by Firsov et al. (26), with the FLAG reporter octapeptide (FLAG sequence: DYKDDDDK; D, aspartate; Y, tyrosine; K, lysine) in the extracellular loop directly above the COOH-terminal of the second transmembrane segment of each subunit. Oocytes were incubated for 30 min in Ringer solution supplemented with 1% BSA at 4°C to block nonspecific binding of the antibodies. Subsequently, oocytes were incubated for 45 min at 4°C with 2 μg/ml mouse anti-FLAG (M2) peroxidase (horseradish peroxidase) conjugated antibody (Sigma-Aldrich, St. Louis, MO) in 1% BSA/Ringer. Oocytes then were washed, initially in 1% BSA/Ringer (45 min at 4°C), then in Ringer solution without BSA (45 min at 4°C). Individual oocytes were placed in 200 μl of Visualizer Western Blot Detection solution (Upstate Cell Signaling Solutions, Dundee, UK) for 10 s, and chemiluminescence was quantified using a Fluor-S Multiimager (Bio-Rad, Hemel Hempstead, UK) by integrating the signal over a period of 30 s. Results are given in relative light units (RLU) as a function of measured density of chemiluminescence after the correction for background chemiluminescence. Nonspecific binding was determined from parallel assays of noninjected oocytes.
Western Blot Analysis
Oocytes were homogenized (25 oocytes per experimental group) in buffer that contained (in mM) 10 HEPES, 1 EDTA, 0.5 PMSF (pH 7.4), and 6.25 × 103 IU/L Aprotinin, using an Ultra Turrax homogenizer (Janke & Kunkel, Staufen, Germany) at half speed for 20 s. Homogenates were centrifuged for 10 min at 3000 rpm and the supernatant was re-centrifuged at 13,000 rpm for 30 min. The pellet was resuspended in buffer, and high-speed centrifugation was repeated. The resulting pellet was resuspended in 25 μl of buffer, and the final protein concentration was determined using the Bradford method (27). All steps were performed at 4°C.
Oocyte membrane preparations (15 μg of protein) were then solubilized in Laemmli sample buffer that contained 2% (wt/vol) SDS and 3% (wt/vol) dithiothreitol and electrophoresed on a 10% SDS polyacrylamide gel using 25 mM Tris, 192 mM glycine, and 0.1% (wt/vol) SDS running buffer (pH 8.3) at 20 mA per gel. The proteins were transferred to nitrocellulose membranes by electrophoretic blotting for 1 h at a constant current of 1 mA/cm2 (Transblot semidry transfer cell; Bio-Rad) using transfer buffer that contained 10% (vol/vol) methanol, 25 mM Tris, 192 mM glycine, and 0.1% (wt/vol) SDS (pH 8.2 to 8.4). Nonspecific protein-binding sites were then blocked with PBS-T (pH 7.4) that contained 0.1% Tween 20 and 6% (wt/vol) fat-free milk powder for 1 h at room temperature. The membranes were incubated with rabbit anti-rat P2X2 antibody (Alomone, Jerusalem, Israel) at 1:1000 in PBS-T for 16 h at 4°C. The membranes were then washed with PBS-T and incubated with a swine anti-rabbit IgG antibody conjugated to horseradish peroxidase (1:2000 dilution; Amersham Biosciences, Piscataway, NJ) for 1 h at room temperature and finally washed again with PBS-T (2 × 15 min). Bound antibodies were detected using a Visualizer Western Blot Detection Kit (Upstate Cell Signaling Solutions) and visualized and quantified (for 30 s) using a Fluor-S Multiimager (Bio-Rad). Band density was expressed in arbitrary units. All Western blot experiments were repeated using four different batches of oocytes.
Statistical Analyses
All data are presented as mean values ± SEM; n indicates the number of oocytes from different frogs; significance was evaluated by t test (Instat v2.05a; GraphPad Software, San Diego, CA) with P < 0.05 considered significant.
Results
ENaC and P2X Receptor Coexpression
Heterotetrameric ENaC expressed in defolliculated oocytes carried a persistent inward Na+ current that was amiloride sensitive (Iam-s). Amiloride (100 pM to 10 μM) concentration-dependently blocked Iam-s with an IC50 value of 29.0 ± 14.9 nM (n = 3) at −100 mV, and full blockade occurred at ≤10 μM (Figure 1A). The effect of amiloride was rapidly reversed on washout. To avoid ENaC rundown in response to cytosolic Na+ accumulation, oocytes were used in all further experiments with 10 μM amiloride present. Brief removal of amiloride blockade (40 s) at 20-min intervals resulted in large (≥3 μA) reproducible Iam-s (Figure 1B).
ENaC was coexpressed with P2X receptors to investigate whether activation of P2X receptors by extracellular ATP had an effect on Iam-s. Figure 1C shows a typical trace in which ENaC was coexpressed with homotrimeric ATP-gated P2X2 receptors. Activation of the P2X2 receptor by ATP (30 μM, approximately EC80) elicited a large (approximately 4 μA), slowly desensitizing, inward cationic membrane current (IATP) in the presence of 10 μM amiloride. Activation of the P2X2 receptor 5 min before the removal of amiloride resulted in an approximately 80% decrease in the amplitude of Iam-s. The inhibition in Iam-s after P2X2 receptor activation did not change over a period of 2 h and therefore was considered irreversible. When expressed alone, ENaC was found to be insensitive to extracellular ATP (100 μM; maximal Iam-s values pre- and post-ATP challenge were 3428 ± 511 and 3354 ± 543 nA, respectively), and all P2X receptors were found to be insensitive to extracellular amiloride (10 μM).
P2X-Mediated Inhibition of Iam-s
Homomeric and heteromeric P2X receptors were coexpressed with ENaC and tested for their ability to inhibit Iam-s. Activation of homomeric P2X2 and P2X4 receptors (by 100 and 30 μM ATP, respectively) resulted in a significant inhibition of Iam-s (by 80 ± 6%, n = 3, P < 0.01 and 52 ± 5%, n = 3, P < 0.01; respectively; Figure 2A). Previously characterized (pharmacologically and biophysically) heteromeric P2X receptor assemblies P2X2/6 and P2X4/6 (9), when activated by 100 μM ATP, also significantly inhibited Iam-s (by 72 ± 3%, n = 3, P < 0.01 and 45 ± 2%, n = 3, P < 0.01; respectively; Figure 2A). In contrast, activation of coexpressed P2X1, P2X3, P2X5, and P2X6 receptors had no significant effect on the amplitude of Iam-s. The rank order of relative efficacy of P2X receptors (activated by an EC100 concentration of ATP) to inhibit Iam-s was P2X2 ≥ P2X2/6 > P2X4 = P2X4/6.
Selectivity of P2X-mediated inhibition of Iam-s. (A) Histogram comparing the percentage inhibition of Iam-s after activation of coexpressed P2X receptors by extracellular ATP (by an EC100 concentration of ATP). Data are expressed as a percentage of control Iam-s measured before activation of P2X receptors (n = 3). Iam-s was evoked by the removal of amiloride (10 μM) blockade at 20-min intervals. P2X receptors were activated 5 min before the subsequent Iam-s was evoked. (B) Typical P2X receptor–mediated currents evoked by 100 μM ATP (Vh = −100 mV). The largest and slowest desensitizing P2X-mediated currents correlated with the P2X receptors capable of inhibiting Iam-s (as shown in A). (C) The two distinct linear relationships between the percentage inhibition of Iam-s after the activation of P2X2 (and P2X2/6) and P2X4 (and P2X4/6) receptors by ATP (as shown in A; 5 min before the removal of 10 μM amiloride) versus the area under curve or charge transfer of P2X-mediated IATP. A range of ATP concentrations was used (100 nM to 100 μM), with each concentration of ATP repeated on at least three cells. For all four P2X receptors tested, inhibition of Iam-s could not be achieved beyond approximately 90% inhibition.
P2X receptor subtypes present strikingly different activation and inactivation properties (9). Homomeric and heteromeric assemblies of P2X1 and P2X3 receptors are defined as rapidly inactivating, P2X5 and P2X6 homomeric assemblies as poorly activating, and homomeric and heteromeric P2X2 and P2X4 assemblies as slowly inactivating (Figure 2B). During electrophysiologic recordings, it became clear that only the slowly inactivating receptor subtypes (P2X2, P2X4, P2X2/6, and P2X4/6) that also mediated the largest amplitude currents resulted in the inhibition of Iam-s. Therefore, we investigated the correlation between Iam-s inhibition and the size of IATP for these receptor subtypes. For a range of ATP concentrations (100 nM to 100 μM; one concentration per cell repeated on at least three cells), a linear relationship was found between the area under the curve (I × t = Q, or charge transfer) of IATP and the percentage inhibition of Iam-s after P2X2, P2X4, P2X2/6, or P2X4/6 receptor activation (Figure 2C). For all four P2X receptors, inhibition of Iam-s could not be achieved beyond approximately 90% inhibition, irrespective of any further increase in IATP charge transfer. The slope of the linear relationship for the P2X2 receptor (0.8 ± 0.1, n = 3) and its heteromer (0.8 ± 0.2, n = 3) was less than that of the P2X4 receptor (1.7 ± 0.2, n = 3) and its respective heteromer (1.7 ± 0.2, n = 3), suggesting different molecular mechanisms of IATP-induced Iam-s inhibition.
P2X2-Mediated Mechanism of Iam-s Inhibition
Previous studies have demonstrated that an increase in intracellular Na+ causes feedback inhibition of ENaC (5), thought to be triggered by the binding of Na+ to a putative intracellular Na+ receptor/binding site that is G protein linked and blocked by 5-N-dimethyl-amiloride (DMA) (28). In addition, it has been reported that P2 receptor–mediated activation of 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS)–sensitive calcium-activated Cl− channels (CaCC), which are endogenous to Xenopus oocytes, results in the inhibition of Iam-s (29). With these observations in mind, the molecular mechanism of P2X-mediated inhibition of Iam-s was investigated pharmacologically.
We found that when P2X2 receptors were coexpressed with ENaC and activated by 100 μM ATP (EC100), decreasing the extracellular Na+ concentration (Δ[Na+]ext) from 100 to 80 mM led to a significant decrease (by 24 ± 4%, n = 3, P < 0.05) in the percentage inhibition of Iam-s with no change in the amplitude of IATP (Figure 3A). Cytosolic injection of DMA (5 nl, 100 μM) into the oocyte 30 min before electrophysiologic recording completely abolished IATP-mediated inhibition of Iam-s (96 ± 4% decrease in inhibition, n = 3, P < 0.001), with no change in the amplitude of IATP. Similarly, cytosolic injection of 50 μg/ml pertussis toxin (PTX), a G protein inhibitor, into the oocyte 24 h before recording also abolished the P2X2-mediated inhibition of Iam-s (97 ± 3% decrease in inhibition, n = 3, P < 0.001), with no change in the amplitude of IATP. Perfusion with 100 μM wortmannin, a phosphatidylinositol 3-kinase inhibitor, had the same effect (data not shown). Similar reductions in Iam-s inhibition with Δ[Na+]ext, DMA, and PTX were found for the P2X2 heteromeric assembly P2X2/6 (data not shown). However, applying 100 μM DIDS extracellularly, lowering extracellular concentrations of Cl− acutely, incubating oocytes in low Cl− for 72 h to reduce the cytosolic levels of Cl− (low [Cl−]int), or removing extracellular Ca2+ had no significant effect on the degree of P2X2- or P2X2/6-mediated inhibition of Iam-s (or the amplitude of IATP).
Mechanisms of P2X-mediated inhibition of ENaC currents. (A) Histogram showing the percentage inhibition of Iam-s as a result of ATP-evoked P2X2 receptor activation, in a number of experimental conditions (n = 3 for each experimental condition). (B) As in A, histogram showing the effect of various experimental conditions on the percentage inhibition of Iam-s after P2X receptor activation, this time in oocytes that coexpressed ENaC and P2X4 receptors (n = 3 for each condition). (C) A typical trace demonstrating that P2X4-mediated inhibition of Iam-s is abolished in the presence of the calcium-activated Cl− channel blocker 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS; 100 μM) and the injection of dimethyl amiloride (DMA; 100 μM) into oocytes 30 min before experimentation. PTX, pertussis toxin.
P2X4-Mediated Mechanisms of Iam-s Inhibition
When P2X4 receptors were coexpressed with ENaC and activated by 30 μM ATP (EC100), changing [Na+]ext from 100 to 80 mM resulted in a significant decrease (21 ± 2%, n = 3, P < 0.05) in the P2X-mediated inhibition of Iam-s (Figure 3B). In contrast to the P2X2 and P2X2/6 receptor experiments, DMA reduced P2X4-mediated inhibition by only 47 ± 6% (n = 3, P < 0.01). Acutely decreasing [Cl−]ext had no effect, whereas incubation in low Cl− for 72 h (low [Cl−]int), removal of [Ca2+]ext, or addition of the calcium-activated Cl− channel blocker DIDS significantly reduced the inhibition of Iam-s, by 53 ± 8% (n = 3, P < 0.01), 45 ± 5% (n = 3, P < 0.01), and 47 ± 3% (n = 3, P < 0.01), respectively. As with P2X2 and P2X2/6, PTX abolished the P2X4-mediated inhibition of Iam-s (n = 3). DIDS and zero extracellular Ca2+ (zero [Ca2+]ext) independently and additively reduced P2X4-mediated inhibition of Iam-s by approximately 50% without significantly altering the amplitude of IATP (data not shown). Given that DIDS, low [Cl−]int, and DMA independently reduced P2X4-mediated inhibition of Iam-s by approximately 50%, their effects were also tested additively. Figure 3C (a representative trace from n = 3) shows that the cytosolic injection of DMA (100 μM) and the perfusion of DIDS (100 μM) abolished P2X4-mediated inhibition of Iam-s. Similarly, injection of DMA and incubation of oocytes in low Cl− for 72 h (low [Cl−]int) also abolished the P2X4-mediated inhibition of Iam-s. Furthermore, the application of DIDS to oocytes with low [Cl−]int resulted in no further change in the reduction in inhibition of Iam-s. Collectively, these data suggest two independent mechanisms of inhibition: (1) an increase in [Na+]int (as with the P2X2 receptor) and (2) an increase in [Ca2+]int leading to the activation of CaCC (and probable Cl− efflux). Similar reductions in inhibitory effects on Iam-s were observed with Δ[Na+]ext, DMA, zero [Ca2+]ext, DIDS, DMA + DIDS, and PTX when using the P2X4/6 receptor (data not shown).
Effect of Outward IATP on Iam-s
To confirm that inhibition of Iam-s was due to an increase in intracellular Ca2+ and/or Na+ carried by IATP, the Vh during application of ATP was increased from −100 to 50 mV, making IATP an outward current. Iam-s was still recorded at −100 mV. For both P2X2 and P2X4 receptors, activation by a maximal concentration of ATP at Vh 50 mV had no significant effect on Iam-s (n = 3). The area under the curve (and peak amplitude) for outward IATP (Vh = 50 mV) was 238 ± 42 × 10−10 Q (498 ± 72 nA) and 183 ± 17 × 10−10 Q (323 ± 89 nA) for P2X2- and P2X4-mediated currents, respectively (data not shown). Figure 2C confirms that the charge transfer of these outward currents, had they been inward currents, would have been sufficient to have inhibited Iam-s significantly.
Time Course of P2X-Mediated Inhibition of Iam-s
For both P2X2 and P2X4 receptors, activation (and subsequent return to baseline, i.e., deactivation of IATP) by extracellular ATP (100 and 30 μM, respectively), either immediately before or 5 min before the removal of amiloride, did not result in a significant change in the percentage inhibition of Iam-s (immediately before 82 ± 6 and 52 ± 9% inhibition of Iam-s; 5 min before 80 ± 6 and 52 ± 5% inhibition of Iam-s, for P2X2 and P2X4 receptors, respectively), indicating that P2X-mediated inhibition of Iam-s was immediate. Further analysis of the time course of P2X-mediated inhibition of Iam-s by investigating the decay constant for IATP during the removal of amiloride blockade proved impossible because of the additive large whole-cell currents and the whole-cell voltage clamping limitations of the system.
Current/Voltage Relationships
Current/voltage (I/V) relationships for whole-cell currents were investigated for IATP and Iam-s. In oocytes in which ENaC was expressed alone, the I/V curve for Iam-s showed inward rectification with a reversal potential (Erev) of 10 ± 1 mV (n = 3; Figure 4A). Given an extracellular concentration of Na+ of approximately 100 mM, the Erev suggests partial Na+ loading of the oocytes. Application of ATP (100 μM) did not change the I/V curve for Iam-s. IATP was negligible (<1 nA in amplitude) as expected, because Xenopus oocytes lack functional endogenous P2 receptors (30).
Average current/voltage (I/V) plots of amiloride-sensitive whole-cell currents (Iam-s) and ATP-sensitive whole-cell currents (IATP) measured in oocytes that expressed ENaC alone (A; n = 3) or coexpressed ENaC and P2X2 receptors (B; n = 3) or ENaC and P2X4 receptors (C; n = 3). Voltage pulse protocols were performed using step changes of the clamp potential from −90 to −130 mV up to 50 mV in 20-mV increments. Iam-s was determined by subtracting the whole-cell current trace recorded in the presence of amiloride (10 μM) from that recorded in the absence of amiloride. Iam-s was measured before ATP (100 μM) exposure (▪) and 20 min later after ATP exposure (▴). IATP was measured during application of 100 μM ATP when IATP reached maximum amplitude (▾).
In oocytes that coexpressed ENaC and P2X2 receptors, Iam-s (before the activation of P2X2 receptors) exhibited a similar inward rectification and Erev to that found in oocytes that expressed ENaC alone (Erev = 10 ± 2 mV, n = 3; Figure 4B). After an ATP challenge and activation of P2X2 receptors, Iam-s was found to be smaller in amplitude, yet the Erev value (7 ± 2 mV, n = 3) and inward rectification were found to be unchanged. The 3-mV shift (NS) in the mean Erev is compatible with a small change in the electrochemical Na+ gradient as a result of an increase in intracellular Na+. The IATP I/V curve showed inward rectification and an Erev of −10 ± 1 mV (n = 3), as expected for either a poorly selective cation channel that conducts Na+ and K+ or one that conducts Cl−. Because removal of extracellular Cl− had no effect on P2X2-mediated inhibition of Iam-s, the latter is unlikely (Figure 3A).
Figure 4C shows the I/V relationships for Iam-s and IATP in oocytes that coexpressed ENaC and P2X4 receptors. The curves for Iam-s, both before and after P2X4 receptor activation, were inwardly rectifying with Erev values of 10 ± 2 mV (n = 3) and 8 ± 2 mV (n = 3), respectively. After P2X4 receptor activation, the amplitude of Iam-s decreased. The P2X4-mediated IATP I/V curve was inwardly rectifying with an Erev value of 31 ± 4 mV (n = 3). This indicates a poorly selective cation channel permeable to Na+ and Ca2+, fitting with the pharmacologic findings shown in Figure 3B. Zero [Ca2+]ext resulted in a negative shift of the Erev to −8 ± 3 mV (n = 3), confirming that the P2X4 receptor is additionally Ca2+ permeable (data not shown).
FLAG-Tagged ENaC: Receptor Trafficking
ENaC protein density in the plasma membrane was investigated using FLAG-tagged ENaC coexpressed with P2X receptors in Xenopus oocytes. Membrane chemiluminescence of bound anti-FLAG horseradish peroxidase–conjugated antibody was measured in intact oocytes before and after ATP exposure. Oocytes that expressed ENaC-FLAG and oocytes that coexpressed ENaC-FLAG and P2X2 receptors showed chemiluminescence levels of 9.3 ± 2.0 (n = 11) and 7.1 ± 1.0 RLU (n = 11), respectively (NS). Addition of ATP (100 μM for 90 s) to oocytes that coexpressed ENaC and P2X2 receptors resulted in a significant fourfold decrease in membrane chemiluminescence (1.7 ± 1.0 RLU, n = 11, P < 0.01; Figure 5). A similar decrease (3.5-fold) in membrane chemiluminescence (n = 3) was also measured in oocytes that coexpressed P2X4 receptors and ENaC-FLAG after ATP exposure (data not shown). This suggests the removal of ENaC from the oocyte membrane after activation of the coexpressed P2X2 and P2X4 receptors by ATP.
Activation of P2X2 receptors by extracellular ATP decreases surface expression of ENaC-FLAG. A chemiluminescence assay was used to detect surface expression of extracellular FLAG-tagged ENaC in oocytes that expressed ENaC-FLAG alone or coexpressed P2X2/ENaC-FLAG. Oocytes that expressed ENaC alone served as controls for nonspecific chemiluminescence. Single oocyte chemiluminescence was measured from 11 oocytes per group, and mean values are expressed in relative light units (RLU). FLAG sequence: DYKDDDDK.
The amplitude of Iam-s in oocytes that expressed the FLAG-tagged ENaC subunits (ENaC-FLAG) was 3.9 ± 1.1 μA (n = 10 oocytes; Vh = −100 mV), which was not significantly different from that in oocytes that expressed nontagged ENaC subunits (3.7 ± 0.8 μA, n = 50 oocytes).
Maximum Current Amplitudes
During investigation of the effects of IATP on Iam-s, it became apparent that the maximal amplitude of P2X2 receptor–mediated IATP was much greater (approximately 2.5-fold) when P2X2 receptors were coexpressed with ENaC. We therefore investigated whether coexpression could affect the amplitude of maximal ENaC- or P2X receptor–mediated currents over a 5-d period (kept at 18°C) after the cytosolic injection of cRNA.
When ENaC was expressed alone, maximal Iam-s was 3.5 ± 0.5 (n = 3), 3.7 ± 0.4 (n = 3), and 3.1 ± 0.2 μA at day 1, day 2, and day 3 after injection, respectively. Iam-s decreased significantly by approximately 50% on day 5 to 1.2 ± 0.2 μA (n = 3, P < 0.01; Figure 6A), without affecting the slope of the inhibition curve or IC50 value for amiloride (data not shown). In contrast, when P2X receptors were expressed alone, maximal IATP remained relatively constant over the 5-d period (Figure 6B).
ENaC coexpression increases the maximum amplitude of IATP. (A) Amplitude of Iam-s evoked by the removal of amiloride (10 μM), in oocytes that expressed ENaC alone measured at 24-h intervals for 5 d after cRNA injection (n = 3 for each data point). (B) Amplitude of P2X-mediated currents evoked by a maximally stimulating concentration of ATP (IATP) in oocytes that expressed a series of P2X receptors alone, measured at 24-h intervals for 5 d after cRNA injection (n = 3 for each data point). (C) Amplitude of ENaC-mediated Iam-s in oocytes that coexpressed ENaC and members of a series of P2X receptors, measured at 24-h intervals for 5 d after cRNA injection (n = 3 in each case). Expressed P2X receptors were not activated. (D) Amplitude of maximal IATP in oocytes that coexpressed ENaC and P2X receptors whose maximal amplitude significantly altered in the presence of ENaC (n = 3 in each case). Coexpressed ENaC were blocked continuously by 10 μM amiloride. *P < 0.01. (E) Amplitude of maximal IATP in oocytes that coexpressed ENaC and P2X receptors whose maximal amplitude was significantly altered in the presence of ENaC compared with amplitude of maximal IATP in oocytes that expressed these P2X receptors alone (taken from B and D).
Coexpression of ENaC with P2X receptors but the latter not activated did not significantly change the trend of maximal Iam-s over the 5-d period (Figure 6C). Conversely, for certain P2X receptors that were coexpressed with ENaC but the latter blocked by amiloride (10 μM), the amplitude of IATP significantly increased over the first 3 d after injection. This increase in maximal IATP was evident for P2X2 (n = 3), P2X2/6 (n = 3), and P2X5 (n = 3) receptors (P < 0.01 in each case; Figure 6, D and E). No change in the mean maximal amplitude of IATP was evident on day 4 or 5. Furthermore, no change in the ATP concentration-response curves or in the EC100 concentration was found (data not shown).
ENaC Coexpression Changes P2X2 Membrane–Bound Protein Levels
Western blots using oocyte total membrane fractions were performed using a specific P2X2 receptor antibody at days 1, 3, and 5 after ENaC and P2X2 receptor cRNA injection to investigate whether there was an increase in P2X2 receptor protein in oocytes that coexpressed ENaC. Figure 7A shows a representative Western blot. No band is evident for oocytes that expressed ENaC alone on days 1, 3, and 5, signifying that the P2X2 receptor antibody did not cross-react with ENaC proteins. On days 1, 3, and 5, for oocytes that expressed P2X2 receptors alone, a 56-kD protein band is seen. Comparison of these 56-kD bands with those observed for oocytes that coexpressed ENaC and P2X2 receptors reveals a considerable increase in the density of the P2X2 receptor band on days 1 and 3 but not 5. Band intensity was quantified and expressed as a percentage difference in P2X2 protein intensity for oocytes that expressed P2X2 alone and those that coexpressed ENaC/P2X2. There were increases of 262 ± 136% (n = 4, P < 0.01) and 408 ± 275% (n = 4, P < 0.01) in P2X2 protein intensity on days 1 and 3, respectively, whereas on day 5 there was no increase in intensity (Figure 7B).
ENaC increase membrane expression of P2X2 receptors. (A) Western blot analysis of P2X2 receptor protein in oocyte membrane homogenates for oocytes that expressed ENaC alone (lane 1) and P2X2 receptors alone (lane 2) and coexpressed ENaC/P2X2 receptors (lane 3) at 1, 3, and 5 d after cRNA injection. (B) Density of western blot bands were quantified and expressed as the percentage difference in P2X2 protein intensity for oocytes that expressed P2X2 alone and those that coexpressed ENaC/P2X2 (n = 4).
ENaC coexpression may increase the expression of P2X receptors (and consequently the amplitude of maximal IATP) by increasing the delivery of receptor subunits to the cell surface or by inhibiting receptor retrieval. To assess the rate of retrieval (endocytosis), we inhibited delivery of new channels to the plasma membrane by incubating oocytes in 18 μM brefeldin A (BFA) for 4 h, 2 d after injection with cRNA. BFA significantly decreased maximal IATP amplitude for P2X2 (n = 3), P2X2/6 (n = 3), and P2X5 (n = 3) receptors in the absence of ENaC by 57 ± 12 (P < 0.01), 61 ± 15 (P < 0.01), and 59 ± 17% (P < 0.01), respectively. In contrast, BFA did not alter IATP amplitude for P2X2 (n = 3), P2X2/6 (n = 3), and P2X5 (n = 3) receptors in the presence of ENaC (by 9 ± 2%, 4 ± 5%, and 11 ± 7%, respectively). Given that BFA inhibits delivery of new P2X channels to the plasma membrane and that maximal IATP increases when P2X2, P2X2/6, and P2X5 receptors are coexpressed with ENaC, this suggests that the rate of P2X receptor retrieval is affected when ENaC and certain P2X receptors are coexpressed in Xenopus oocytes. When ENaC were expressed alone, BFA significantly decreased the amplitude of Iam-s by 37 ± 9% (n = 4, P < 0.05); when ENaC were coexpressed with P2X2, P2X2/6, and P2X5 receptors, BFA similarly decreased the amplitude of Iam-s by 23 ± 7% (n = 3, P < 0.05), 32 ± 8% (n = 3, P < 0.05), and 34 ± 7% (n = 3, P < 0.05), respectively.
Discussion
It is known that the activity of ENaC in the distal nephron of the kidney is modified by the tubular perfusion of ATP or its analogues (19), even though ENaC itself is insensitive to extracellular ATP (17), and the mechanism of this inhibition is unresolved. In the present study, we have shown that activation of recombinant rat P2X2, P2X4, P2X2/6, and P2X4/6 receptor ion channels by extracellular ATP can inhibit amiloride-sensitive ENaC currents. This inhibition is due to a reduction in surface expression of ENaC, which is mediated by Na+ influx and, in the case of P2X4 and P2X4/6 receptors, also by Ca2+ influx. Furthermore, we have shown that increasing expression levels of recombinant ENaC can result in a concomitant increase in coexpressed rat P2X2, P2X2/6, and P2X5 receptors as a result of reduced retrieval from the plasma membrane (summarized in Figure 8).
Potential regulatory interdependence of P2X receptors and ENaC. Schematic demonstrating that increased expression of ENaC (1) may result in the concomitant upregulation of P2X2, P2X2/6, and P2X5 receptors (by decreased retrieval; 2). Activation of P2X2, P2X4, P2X2/6, and/or P2X4/6 receptor ion channels by endogenous extracellular ATP may result in an influx of Na+ and Ca2+ (3) and the subsequent downregulation of ENaC (4).
Na+-Dependent Mechanism of P2X Receptor–Mediated ENaC Downregulation
It was demonstrated previously that downregulation of ENaC occurs when [Na+]int is increased during Na+ reabsorption, an effect known as feedback inhibition (5,31). Although an increase in intracellular Na+ is the trigger for ENaC downregulation, it has been demonstrated that this phenomenon is not mediated by a direct effect of Na+ on the ENaC protein (5). Several cellular mechanisms have been proposed to explain feedback inhibition, of which the activation of the ubiquitin ligase Nedd4 seems particularly important (5,6,28,32,33). Hubner et al. (6) and Komwatana et al. (34) proposed that intracellular Na+ binds initially to a putative Na+ receptor, which, once activated, initiates a Nedd4-dependent signaling cascade involving G proteins and results in ubiquitination and retrieval of ENaC (34). In accordance with this proposed mechanism, we found that IATP-mediated inhibition of Iam-s was completely abolished by DMA and by PTX. It therefore is plausible that the Nedd4 cascade plays a key role in the observed P2X-mediated inhibition of ENaC. Thus, P2X-mediated inhibition of ENaC may be achieved (entirely, in the case of P2X2 and P2X2/6, or partly, in the case of P2X4 and P2X4/6 receptors) via Na+ feedback inhibition.
In previous studies, Na+ feedback inhibition (in the absence of P2X receptor expression) was reported to occur at a much slower rate in Xenopus oocytes than we observed during P2X receptor stimulation (5,6). In both the present and previous studies, when ENaC (in the absence of amiloride) was exposed to high extracellular Na+ concentrations (>100 mM), it took >25 min to achieve a 50% inhibition of ENaC-mediated Na+ current (Vh = −100 mV) (7). In contrast, after P2X2 and P2X2/6 receptor activation, we found a rapid inhibition of Iam-s (up to 90% within 5 min; Figure 2C). Thus, it seems that the influx of Na+ through P2X2 and P2X2/6 channels (and in part P2X4 and P2X4/6 channels) results in removal of ENaC from the plasma membrane at a much faster rate than does the influx of Na+ through ENaC itself. Consequently, the mechanism by which P2X receptors inhibit ENaC is unlikely to be “classical” Na+-mediated feedback inhibition. It is tempting to speculate that P2X receptor ion channels (or at least P2X2, P2X4, P2X2/6, and P2X4/6 receptors) may be in close proximity to ENaC, in such a way that activation of the P2X receptor results in a localized delivery of sodium ions to a site on or near ENaC that is particularly sensitive to increases in [Na+]int.
Ca2+-Dependent Mechanism of P2X Receptor–Mediated ENaC Downregulation
A Ca2+-dependent reduction in surface expression of ENaC was demonstrated previously in cultured epithelial cells in which downregulation of ENaC occurred after a P2Y receptor–mediated increase in [Ca2+]int (14,20,35–37). However, this effect can be discounted in the present study because Xenopus oocytes are devoid of endogenous P2Y receptors (29). A study in cultured mIMCD-K2 cells (an immortalized mouse CD cell line in which P2X1, P2X3, and P2X4 receptors are implicated in ENaC inhibition) is in keeping with our findings; this demonstrated Ca2+-dependent inhibition of Na+ reabsorption and stimulation of Cl− secretion (after P2 receptor activation) (14,35). Also in accordance with our observations, downregulation of ENaC has been associated with the activation of apical CaCC in mouse endometrial epithelial cells (36). In the latter study, removal of extracellular Ca2+ or application of DIDS blocked inhibition of P2 receptor–mediated Na+ absorption (and induction of Cl− secretion). Evidence suggests that changes in cytosolic Cl− concentration may mediate the inhibition of ENaC (31). However, to date, it remains controversial as to whether the inhibition of ENaC mediated by Cl− currents (including CaCC currents) is (1) nonspecific and simply due to changes in intracellular Cl− (31,38) or (2) specific to CFTR activation and the subsequent influx of Cl− (39). The latter seems improbable given our recent recombinant findings. The exact mechanism by which activation of CaCC can result in the inhibition of ENaC or indeed the physiologic significance of Cl− secretion-linked inhibition of ENaC has yet to be determined.
Mechanism of ENaC-Mediated P2X Receptor Upregulation
Our data demonstrate that an ENaC-mediated increase in P2X2, P2X2/6, and P2X5 receptor whole-cell currents is due to an increase in P2X ion channel membrane expression. The number of ion channels expressed in the plasma membrane is determined by the balance between channel insertion into the membrane and retrieval from the plasma membrane. Our experiments using BFA (an inhibitor of channel delivery to the plasma membrane) demonstrate that the rate of P2X receptor retrieval is affected when ENaC is coexpressed with certain P2X receptors in Xenopus oocytes. This is in contrast to ENaC-mediated upregulation of ROMK1 (CFTR dependently) and CFTR, in which an increase in whole-cell currents in Xenopus oocytes was shown to be due to increased insertion into the plasma membrane (25). The ENaC-mediated increase in CFTR resulted in a two- to six-fold increase in cAMP-activated CFTR Cl− currents, a magnitude not dissimilar to the increase that we observed for P2X2 receptor currents. However, in the absence of single-channel data, we cannot discount the possibility that an increase in channel open probability could contribute to this effect, a possibility also raised in the ROMK1 and CFTR study (25).
The fact that only three specific P2X receptor subtypes were upregulated in response to increasing ENaC levels suggests that the P2X/ENaC interaction was specific and did not involve a nonspecific effect of ENaC on P2X protein degradation. It could be hypothesized that the mechanism of this linkage may involve a direct protein–protein interaction. There is already some evidence for direct physical interaction between P2X receptors and other ion channels. Intracellular cross-talk and physical interactions have been demonstrated between the P2X2 receptor and the 5-HT3 serotonin-gated ion channel in myenteric neurons, Xenopus oocytes, and HEK293 cells (40), where concurrent activation of the ion channels produced nonadditive currents and the channels co-immunoprecipitated and associated in clusters. Furthermore, in a recent study, it was demonstrated that activation of P2X receptors can lead to rapid cytoskeletal remodeling involving intracellular structural proteins (β-actin, α-actinin 4, laminin α3, and integrin β2) and membrane signaling proteins (phosphatidylinositol 4-kinase and receptor protein tyrosine phosphatase β) (41). It is known that β-actin can directly regulate ENaC by modifying Ca2+-dependent feedback inhibition (42). Consequently, protein–protein coupling may be responsible for the interactions between certain P2X receptors and ENaC.
Conclusion
In conclusion, these basic findings demonstrate a functional link between certain P2X receptor ion channels and ENaC. This interaction may provide an important mechanism for the coordinated regulation of Na+ transport in epithelia that express P2X receptors, including the distal nephron of the kidney.
Acknowledgments
This work was supported by the Les Clark Fund; St. Peter’s Trust for Kidney, Bladder and Prostate Research (UK); and the British Heart Foundation. S.S.W. is a recipient of a British Heart Foundation Research Fellowship.
Part of this study was presented at the American Society of Nephrology Renal Week, November 12 to 17, 2003, San Diego, CA (17,43) and the joint Experimental Biology and International Union of Physiologic Sciences meeting, March 31 to April 5, 2005, San Diego, CA (44).
We are grateful to Mary Rahman for molecular biology assistance; Dr. Chrystalla Orphanides for editorial assistance; Prof. Bernard Rossier for providing the three subunits of rat ENaC cDNA; Dr. Dmitri Firsov for providing three epitope-tagged subunits of rat ENaC cDNA; Dr. Gary Buell for providing rat P2X1 cDNA; Dr. David Julius for providing rat P2X2 cDNA; and Dr. Xuenong Bo and Prof. Alan North for providing rat P2X3, P2X4, P2X5, and P2X6 cDNA.
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
- © 2005 American Society of Nephrology