Extracellular Zn2+ activates epithelial Na+ channels by eliminating Na+ self-inhibition.

Inhibition of epithelial Na(+) channel (ENaC) activity by high concentrations of extracellular Na(+) is referred to as Na(+) self-inhibition. We investigated the effects of external Zn(2+) on whole cell Na(+) currents and on the Na(+) self-inhibition response in Xenopus oocytes expressing mouse alphabetagamma ENaC. Na(+) self-inhibition was examined by analyzing inward current decay from a peak current to a steady-state current following a fast switching of a low Na(+) (1 mm) bath solution to a high Na(+) (110 mm) solution. Our results indicate that external Zn(2+) rapidly and reversibly activates ENaC in a dose-dependent manner with an estimated EC(50) of 2 microm. External Zn(2+) in the high Na(+) bath also prevents or reverses Na(+) self-inhibition with similar affinity. Zn(2+) activation is dependent on extracellular Na(+) concentration and is absent in ENaCs containing gammaH239 mutations that eliminate Na(+) self-inhibition and in alphaS580Cbetagamma following covalent modification by a sulfhydryl-reactive reagent that locks the channels in a fully open state. In contrast, external Ni(2+) inhibition of ENaC currents appears to be additive to Na(+) self-inhibition when Ni(2+) is present in the high Na(+) bath. Pretreatment of oocytes with Ni(2+) in a low Na(+) bath also prevents the current decay following a switch to a high Na(+) bath but rendered the currents below the control steady-state level measured in the absence of Ni(2+) pretreatment. Our results suggest that external Zn(2+) activates ENaC by relieving the channel from Na(+) self-inhibition, and that external Ni(2+) mimics or masks Na(+) self-inhibition.

Inhibition of epithelial Na ؉ channel (ENaC) activity by high concentrations of extracellular Na ؉ is referred to as Na ؉ self-inhibition. We investigated the effects of external Zn 2؉ on whole cell Na ؉ currents and on the Na ؉ self-inhibition response in Xenopus oocytes expressing mouse ␣␤␥ ENaC. Na ؉ self-inhibition was examined by analyzing inward current decay from a peak current to a steady-state current following a fast switching of a low Na ؉ (1 mM) bath solution to a high Na ؉ (110 mM) solution.
Our results indicate that external Zn 2؉ rapidly and reversibly activates ENaC in a dose-dependent manner with an estimated EC 50 of 2 M. External Zn 2؉ in the high Na ؉ bath also prevents or reverses Na ؉ self-inhibition with similar affinity. Zn 2؉ activation is dependent on extracellular Na ؉ concentration and is absent in ENaCs containing ␥H239 mutations that eliminate Na ؉ self-inhibition and in ␣S580C␤␥ following covalent modification by a sulfhydryl-reactive reagent that locks the channels in a fully open state. In contrast, external Ni 2؉ inhibition of ENaC currents appears to be additive to Na ؉ self-inhibition when Ni 2؉ is present in the high Na ؉ bath. Pretreatment of oocytes with Ni 2؉ in a low Na ؉ bath also prevents the current decay following a switch to a high Na ؉ bath but rendered the currents below the control steady-state level measured in the absence of Ni 2؉ pretreatment. Our results suggest that external Zn 2؉ activates ENaC by relieving the channel from Na ؉ self-inhibition, and that external Ni 2؉ mimics or masks Na ؉ self-inhibition.
Epithelial Na ϩ channels (ENaC) 1 mediate Na ϩ transport across apical membranes of high resistance epithelia. The regulation of Na ϩ transport via ENaC has an important role in the maintenance of extracellular fluid volume homeostasis and the control of blood pressure in humans (1). Alterations in channel activity have been associated with several disorders, including Liddle's syndrome, pseudohypoaldosteronism type 1, and cystic fibrosis (2). ENaC activity is regulated by a variety of both intracellular and extracellular factors, including selected hormones, cations, enzymes, and other channel proteins (3,4). Na ϩ exhibits two types of inhibitory effects on ENaC activity: self-inhibition and feedback inhibition that are due to increases in either extracellular or intracellular Na ϩ concentration, respectively. These regulatory phenomena have been proposed to provide a mechanism to prevent sudden or excessive increases in intracellular Na ϩ concentration (5,6).
Although most studies on Na ϩ self-inhibition have utilized native Na ϩ -transporting tissues, including frog skin, toad urinary bladder, and kidney collecting tubule, Na ϩ self-inhibition has also been observed in Xenopus oocytes expressing ENaCs (7)(8)(9). Extracellular cations have been reported to affect ENaC activity in native tissues that may reflect changes in Na ϩ selfinhibition (3,5). We and others previously reported that external Ni 2ϩ blocks whole cell currents of ␣␤␥ ENaCs expressed in oocytes (10,11). In contrast, external Ni 2ϩ is reported to stimulate Na ϩ currents in A6 cells by relieving ENaC from Na ϩ selfinhibition (12). Recent work suggested that external Zn 2ϩ is a voltage-dependent blocker of ENaC (13). In this report, we examined the effects of external Zn 2ϩ on amiloride-sensitive Na ϩ currents in oocytes expressing ␣␤␥ mouse ENaC (mENaC), and the effects of external Zn 2ϩ and Ni 2ϩ on Na ϩ self-inhibition of ␣␤␥ mENaC. We report that external Zn 2ϩ reversibly activates ENaC by eliminating Na ϩ self-inhibition, whereas external Ni 2ϩ appears to mimic Na ϩ self-inhibition.

EXPERIMENTAL PROCEDURES
Site-directed Mutagenesis-All ENaC clones used in this study are mouse ENaC subunits whose cDNAs were inserted into pBluescript SK-(Stratagene, La Jolla, CA) (14). Point mutations were generated previously by using a PCR-based method (11).
ENaC Expression and Two-electrode Voltage Clamp-ENaC expression in Xenopus oocytes and two-electrode voltage clamp were performed as previously reported (11). Stage V and VI oocytes free of follicle cell layers were injected with 1-4 ng of cRNA for each mENaC subunit per oocyte and incubated at 18°C in modified Barth's saline (MBS, 88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO 3 , 15 mM HEPES, 0.3 mM Ca (NO 3 ) 2 , 0.41 mM CaCl 2 , 0.82 mM MgSO 4 , 10 g/ml sodium penicillin, 10 g/ml streptomycin sulfate, 100 g/ml gentamycin sulfate, pH 7.4). All experiments were performed at room temperature (20 -24°C). Oocytes were continuously clamped at Ϫ60 or Ϫ100 mV in most experiments. Current-voltage relationships were determined by clamping oocytes at holding potentials in the range of Ϫ140 to 60 mV in 20-mV increments.
The responses of Na ϩ self-inhibition were examined as previously reported (7,9). A current decay from a peak current to a relatively steady-state current was considered the response for Na ϩ self-inhibition. The current decay was initiated by rapidly replacing a low Na ϩ bath solution (NaCl-1: containing 1 mM NaCl, 109 mM N-methyl-Dglucamine, 2 mM KCl, 2 mM CaCl 2 , 10 mM HEPES, pH 7.4) with a high Na ϩ bath solution (NaCl-110: containing 110 mM NaCl, 2 mM KCl, 2 mM CaCl 2 , 10 mM HEPES, pH 7.4). Rapid solution exchange was performed with a 6-channel Teflon valve perfusion system from Warner Instruments (Hamden, CT). At the end of an experiment, 10 M amiloride was added to the bath to obtain the amiloride-insensitive current. Whole cell currents in the presence of 10 M amiloride were generally less than 200 nA at Ϫ60 or Ϫ100 mV. Results from oocytes that showed unusually large amiloride-insensitive currents (Ͼ5% of total currents) were discarded to minimize current contamination from endogenous channels and membrane leak. The Na ϩ self-inhibition response was described with two parameters, a time constant () and the ratio of the steady-state current (I ss ) and the peak current (I peak ) (see Ref. 9).
The time course of Na ϩ self-inhibition and the effects of Zn 2ϩ and Ni 2ϩ on this process were analyzed as previously described (9). Briefly, the first 40 s of current decay (or increase) were fitted with an exponential equation by Clampfit 9.0 (Axon Instruments Inc.). The concentration at which half-maximal effects were observed (EC 50 ) were estimated by non-linear least square curve fitting of the dose response data with the Hill equation: R ϭ C n /(C n ϩ EC 50 n ), in which R is the relative response, C is the concentration, and n is the Hill coefficient.
Michaelis constants (K m ) for Na ϩ concentration-current relationships were obtained by a best fitting of the data according to the following equation with non-linear least square curve fitting: I ϭ V max ⅐ C/(C ϩ K m ), in which I is the relative I peak or I ss , and C refers to the Na ϩ concentration used to initiate self-inhibition. The apparent inhibitory constant (K i ) of Na ϩ self-inhibition was estimated from a best fitting of the data with the equation: I ss /I peak ϭ K i n /(C n ϩ K i n ) in which C and n are the concentration of Na ϩ and Hill coefficient, respectively. Superpure NiCl 2 and ZnCl 2 (Ͼ99.999%) were purchased from Sigma-Aldrich and dissolved at 1 M in water and diluted to the desired concentrations in bath solutions. The addition of NiCl 2 or ZnCl 2 to the bath solutions at the highest concentrations (1 mM for NiCl 2 and 5 mM for ZnCl 2 ) used in this study did not alter the pH of the solutions or form precipitates.
Statistical Analysis-Data are presented as mean Ϯ S.E. Significance comparisons between groups were performed using Student's t test. Curve fittings were performed with Clampfit 9.0 (Axon Instruments Inc., Union City, CA) and Origin Pro 7.0 (OriginLab Corp., Northampton, MA).

External Zn 2ϩ Activates ␣␤␥ mENaC Expressed in Xenopus
Oocytes-The effect of extracellular Zn 2ϩ on ␣␤␥ mENaC expressed in Xenopus oocytes was examined by comparing amiloride-sensitive Na ϩ currents prior to and following the addition of ZnCl 2 into the bath solution. Oocytes expressing wild type (WT) ␣␤␥ mENaC were clamped from a holding potential equaling the measured membrane potential to a series of voltages from Ϫ140 to 60 mV in 20-mV increments. Whole cell Na ϩ currents were recorded prior to and 60 s following the addition of varying concentrations of ZnCl 2 to a bath solution containing 110 mM Na ϩ (NaCl-110). 100 M ZnCl 2 increased amiloridesensitive Na ϩ currents by 1.66 Ϯ 0.06-fold ( Fig. 1, A-C). External Zn 2ϩ did not alter the current-voltage relationship, indicating a lack of voltage dependence of Zn 2ϩ activation (Fig.  1B). The Zn 2ϩ stimulation of whole cell currents was completely reversed following washout of Zn 2ϩ , with currents returning to the original level prior to Zn 2ϩ addition ( Fig. 1, A, B, and E). The Zn 2ϩ dose response was analyzed with increasing concentrations of ZnCl 2 (10 Ϫ8 , 10 Ϫ7 , 10 Ϫ6 , 10 Ϫ5 , 10 Ϫ4 , 10 Ϫ3 , and 5 ϫ 10 Ϫ3 M). Stimulatory effects on the amiloride-sensitive Na ϩ currents measured at Ϫ60 mV were observed in the range of 10 Ϫ6 to 10 Ϫ3 M in a dose-dependent manner with a maximal activation at 10 Ϫ4 M. Currents in the presence of Zn 2ϩ at concentrations higher than 100 M tended to return to the basal level, and 5 mM Zn 2ϩ did not affect whole cell currents (Fig. 1C). To exclude the possibility of an influence of a prior Zn 2ϩ exposure on channel activity in the presence of 5 mM Zn 2ϩ , whole cell currents were measured prior to and following 5 mM Zn 2ϩ in oocytes that had not been previously exposed to Zn 2ϩ . In agreement with results presented in Fig. 1C, the current ratio in the presence of 5 mM Zn 2ϩ , relative to current prior to Zn 2ϩ addition, was 1.00 Ϯ 0.06 (n ϭ 7). The apparent affinity for Zn 2ϩ activation of ENaC currents was estimated by analyzing dose-response data with Zn 2ϩ concentrations varying from 10 Ϫ8 to 10 Ϫ4 M (Fig. 1D). The EC 50 , Hill coefficient, and correlation coefficient from a best fitting using Hill equation were 1.99 Ϯ 0.28 M (n ϭ 10), 0.87 Ϯ 0.10 (n ϭ 10), and 0.97 Ϯ 0.01 (n ϭ 10), respectively.
To examine the time course of the Zn 2ϩ effect, experiments were performed with oocytes that were continuously voltageclamped at either Ϫ60 or Ϫ100 mV prior to and following application of external Zn 2ϩ . A typical recording is shown in Fig. 1E. The stimulatory effect was rapid with a time constant of 4.9 Ϯ 0.4 s (n ϭ 6). The ratio of the whole cell current in the presence of 100 M Zn 2ϩ , relative to the current prior to Zn 2ϩ addition, was 1.78 Ϯ 0.09 (n ϭ 6). The current in the presence of Zn 2ϩ returned to basal level following washout and was blocked completely by 10 M amiloride. Similar magnitude of stimulation of the current was observed with repetitive applications of Zn 2ϩ without an obvious decline in its effect.
Xenopus oocytes express several types of endogenous channels that may conduct Na ϩ (15)(16)(17)(18). To exclude the possibility that the observed increase in whole cell currents by extracellular Zn 2ϩ was due to activation of an endogenous channel, we examined the effect of 100 M ZnCl 2 on the whole cell currents in H 2 O-injected oocytes. The currents measured from six oocytes at Ϫ100 mV in NaCl-110 NaCl-110 with 100 M ZnCl 2 , and NaCl-110 with 10 M amiloride were Ϫ66.7 Ϯ 27.9, Ϫ50.0 Ϯ 12.9, and Ϫ83.2 Ϯ 16.7 nA (p Ͼ 0.05), respectively, suggesting that 100 M external Zn 2ϩ does not activate endogenous currents. Furthermore, no changes in whole cell currents were observed in oocytes expressing ␣␤␥ mENaC when 10 M ZnCl 2 was externally applied in the presence of 10 M amiloride. The relative currents (i.e. currents normalized to values in the absence of amiloride) in the presence of amiloride alone and amiloride with Zn 2ϩ were 1.9 Ϯ 0.6 and 1.7 Ϯ 0.3% (n ϭ 6), respectively (p Ͼ 0.05). The results suggest that the stimulatory effect of external Zn 2ϩ on currents in oocytes expressing ENaC results from activation of ENaC. These findings differ from the reported blocking effect of external Zn 2ϩ in oocytes expressing ␣␤␥ rat ENaCs and in A6 cells that express endogenous ENaC (13).
External Zn 2ϩ Eliminates Na ϩ Self-inhibition-The stimulation of the amiloride-sensitive whole cell currents by extracellular Zn 2ϩ may result from an increase in unitary current, open probability, or number of active channels in oocyte membranes. Because the extracellular Na ϩ concentration was constant and Zn 2ϩ did not change the reversal potentials ( Fig. 1B), it is unlikely that Zn 2ϩ had a significant effect on the driving forces for generating the whole cell currents. The rapid time course of Zn 2ϩ activation and recovery following washout of Zn 2ϩ suggest a direct effect of Zn 2ϩ on single channel properties rather than a change in surface channel density. Several extracellular divalent cations are known to stimulate Na ϩ transport in model epithelia such as frog skin and toad bladder, possibly through interfering with Na ϩ self-inhibition (5). To investigate the mechanism of ENaC activation by external Zn 2ϩ , we examined whether extracellular Zn 2ϩ altered the Na ϩ self-inhibition response of ␣␤␥ mENaC expressed in Xenopus oocytes. Na ϩ self-inhibition was studied by monitoring changes in whole cell Na ϩ currents measured at either Ϫ60 or Ϫ100 mV by two-electrode voltage clamp during a rapid increase in extracellular Na ϩ concentration. We previously reported that ␣␤␥ mENaCs exhibit a Na ϩ self-inhibition response that is similar to that of rat, human, and Xenopus ␣␤␥ ENaC (7-9). The effects of external Zn 2ϩ on Na ϩ self-inhibition were studied with two sets of experiments. First, we examined the effect of the presence of Zn 2ϩ in the high Na ϩ bath solution (NaCl-110). A typical response for Na ϩ self-inhibition of ␣␤␥ mENaC is shown in Fig. 2A. The current reached a maximal level (termed as I peak ) and declined to a relatively steady level (termed as I ss ) following a fast solution exchange that increased the extracellular Na ϩ concentration from 1 to 110 mM. The current decay reflects the Na ϩ self-inhibition response. After successful observation of a typical response of Na ϩ self-inhibition in oocytes expressing ␣␤␥ mENaC, we examined the self-inhibition response when the low Na ϩ solution (NaCl-1) was replaced by the NaCl-110 solution containing 100 M ZnCl 2 . No current decay was observed, and the current remained at a level slightly higher than the I peak observed in the previous test, as shown in Fig. 2A. The Zn 2ϩ -activated currents were completely blocked by 10 M amiloride or by reducing bath Na ϩ to 1 mM (Fig. 2, A  and B), indicating that all the currents in the presence of Zn 2ϩ were conducted through ENaC. These results indicate that external Zn 2ϩ at 100 M completely prevented Na ϩ self-inhibition. This Zn 2ϩ effect was also reversible, because washout of Zn 2ϩ quickly restored the typical self-inhibition response ( Fig.  2A). We also examined whether the addition of Zn 2ϩ was able to reverse current inhibition by extracellular Na ϩ by adding 100 M Zn 2ϩ to NaCl-110 when the currents reached a steadystate level following the Na ϩ self-inhibition response. As seen in Fig. 2C, Zn 2ϩ increased the current to a level near the I peak observed in the absence of Zn 2ϩ , and washout of Zn 2ϩ returned the current to the inhibited level. Therefore, external Zn 2ϩ not only prevents Na ϩ self-inhibition, it also reverses Na ϩ self-inhibition.
In the second set of experiments, we examined whether pretreatment of oocytes expressing ␣␤␥ mENaC with Zn 2ϩ affected Na ϩ self-inhibition. Oocytes were perfused with NaCl-1 containing 100 M ZnCl 2 and then perfused with NaCl-110 without ZnCl 2 . A current decay was observed with an elevated I peak and an unchanged I ss compared with the values in the control test (Fig. 2D). Although Zn 2ϩ added at the same time as high concentration of Na ϩ was sufficient to prevent Na ϩ from causing the self-inhibition, Zn 2ϩ pretreatment was unable to prevent Na ϩ self-inhibition (Fig. 2).
To estimate the apparent affinity for Zn 2ϩ elimination of Relative currents are the ratios of amiloride-sensitive Na ϩ currents measured at Ϫ60 mV in the presence of increasing concentrations of ZnCl 2 in NaCl-110, relative to the amiloride-sensitive currents measured prior to addition of ZnCl 2 . D, re-plot of the Zn 2ϩ dose-response data in C to determine the half-maximal effective concentration (EC 50 ). The relative responses represent the increases in current in response to 10 Ϫ8 , 10 Ϫ7 , 10 Ϫ6 , 10 Ϫ5 , or 10 Ϫ4 M Zn 2ϩ , normalized to the current increase observed with 10 Ϫ4 M Zn 2ϩ at which maximal Zn 2ϩ activation was observed. The curve is a best fit of the data by non-linear least-square fitting with a Hill equation: Response ϭ C n /(C n ϩ EC 50 n ), where "Response" represents the relative response and "n" is the Hill coefficient. The fitting parameters are: EC 50 , 1.74 M; Hill coefficient, 0.77, and correlation coefficient (R 2 ), 0.9954. E, time course of Zn 2ϩ activation. An ␣␤␥ mENaC-expressing oocyte was bathed in NaCl-110 solution as indicated by the black bar and continuously clamped at Ϫ60 mV. The gray bar indicates the period of time when the oocyte was perfused with NaCl-110 containing 100 M ZnCl 2 . The arrow indicates the time when 10 M amiloride was introduced into the oocyte chamber. By convention, the negative values represent inward whole cell currents. The trace is representative of six experiments. Na ϩ self-inhibition, we examined Na ϩ self-inhibition responses in the presence of increasing concentrations of ZnCl 2 in the NaCl-110 solution (Table I). A typical experiment is shown in Fig. 3A. The effect of external Zn 2ϩ on preventing Na ϩ selfinhibition was dose-dependent ( Fig. 3B and Table I). The esti-mated EC 50 for Zn 2ϩ was 1.3 M by a best fitting of the doseresponse relationship (Fig. 3B, inset). The EC 50 value was almost identical to that obtained from analysis of the Zn 2ϩ dose-response curve for activation of ENaC currents (Fig. 1D), suggesting a link between ENaC activation and loss of Na ϩ self-inhibition by external Zn 2ϩ .
The Effect of External Zn 2ϩ on ENaC Is Dependent on the Extracellular Na ϩ Concentration-External Zn 2ϩ may eliminate Na ϩ self-inhibition by three possible mechanisms: (i) interfering with Na ϩ binding to a "receptor," (ii) preventing conformational changes induced by Na ϩ binding, or (iii) locking ENaC in a fully open state that is insensitive to regulation by extracellular Na ϩ . Na ϩ self-inhibition is considered a low affinity event with an estimated inhibitory constant of Ͼ100 mM (7,9). To determine whether Zn 2ϩ and Na ϩ are binding to a common site, we analyzed the changes in the Na ϩ concentration-current relationship by low concentrations of Zn 2ϩ . The Na ϩ concentration-current relationships for the peak and steady-state currents were examined in the presence of 1 and 10 M of ZnCl 2 . In the presence of 1 and 10 M Zn 2ϩ , the estimated K m values for I ss were significantly higher than the values obtained in the absence of Zn 2ϩ (p Ͻ 0.01 or 0.001), whereas the K m values for I peak were not significantly different from the control value (Table II). The apparent inhibitory constants (K i ) for Na ϩ self-inhibition in the presence of 1 or 10 M Zn 2ϩ were also significantly higher than the apparent K i in the absence of Zn 2ϩ . As shown in Fig. 4A, Zn 2ϩ shifted the relationship between I ss /I peak and Na ϩ concentrations to the right. A complete analysis of the shift is precluded due to limitations of further increases of the extracellular Na ϩ concentration. These data suggest that Zn 2ϩ and Na ϩ may compete for a common binding site.
Our results indicate that external Zn 2ϩ activates ENaC by eliminating Na ϩ self-inhibition, a process that is dependent on the extracellular Na ϩ concentration. If extracellular Na ϩ concentration is below the minimal concentration causing Na ϩ self-inhibition, Zn 2ϩ activation should be abolished. We examined the effect of Zn 2ϩ on ENaC currents in oocytes that were expressing ␣␤␥ mENaCs and bathed in a low Na ϩ concentration solution (10 mM). External Zn 2ϩ did not significantly alter amiloride-sensitive Na ϩ currents in oocytes that had been incubated in either regular Na ϩ MBS (88 mM) or low Na ϩ MBS (10 mM) following cRNA injections (Fig. 4B). The MBS solution with a low Na ϩ concentration was used to prevent Na ϩ loading of oocytes in the period following the injection of ENaC cRNAs and preceding the voltage clamp experiments, as Amuzescu et al. had reported Zn 2ϩ -dependent block of ENaC in oocytes that were maintained in a low Na ϩ concentration bath prior to functional assays (13). We did not observe Zn 2ϩ -dependent inhibition of ENaC currents. The effects of external Zn 2ϩ were also examined in studies performed with the membrane voltage held at Ϫ100 mV while continuously monitoring the whole cell current. Increasing bath Na ϩ concentration from 1 to 10 mM led to no obvious current decay, as we previously reported (9). The addition and washout of 10 M Zn 2ϩ did not significantly affect the currents following an increase of the bath Na ϩ concentration from 1 mM to 10 mM (Fig. 4, C and D).
Mutations at ␥H239 That Disrupt Na ϩ Self-inhibition Diminish Zn 2ϩ Activation-We recently reported that mutations of a His residue within the extracellular loop (ECL) of ␥ mENaC (␥H239R, ␥H239D, and ␥H239C) eliminated Na ϩ selfinhibition (9). If the stimulatory effect of external Zn 2ϩ depends on the presence of Na ϩ self-inhibition, Zn 2ϩ should have no effect on channels with ␥H239 mutations. No current decay was observed in oocytes expressing ␣␤␥H239R following a fast increase in bath Na ϩ concentration to 110 mM in the absence or showing the effect of Zn 2ϩ pretreatment on Na ϩ self-inhibition. Following a typical self-inhibition response, the l mM Na ϩ solution was replaced by a 100 M ZnCl 2 , 1 mM Na ϩ solution. The bath Na ϩ concentration was rapidly increased to 110 mM by replacing 100 M ZnCl 2 , NaCl-1 with Zn 2ϩ -free NaCl-110. After the current reached a steady state, 100 M Zn 2ϩ was added to the NaCl-110 solution. At the end of the recording, the oocyte was perfused with NaCl-110 containing 10 M amiloride.
presence of 100 M Zn 2ϩ (Fig. 5A). In experiments using the same protocol as in Fig. 1C for WT ENaCs, whole cell amiloride-sensitive Na ϩ currents in oocytes expressing ␣␤␥H239R also remained constant in the presence of increasing concentrations of external Zn 2ϩ (Fig. 5B). These data suggest that external Zn 2ϩ activation of ENaC depends on the presence of Na ϩ self-inhibition. In contrast to the effect of ␥H239 mutations, we previously reported that ␣H282D␤␥ channels show an enhanced Na ϩ self-inhibition response (9). Addition of Zn 2ϩ to the NaCl-110 solution blunted, but did not eliminate Na ϩ selfinhibition in oocytes expressing ␣H282D␤␥ mENaCs (Fig. 5C). The I ss was much higher than the I ss observed in the absence of Zn 2ϩ . Addition of 100 M ZnCl 2 into high Na ϩ bath significantly increased the I ss following a self-inhibition response but did not fully restore the current to the I peak (Fig. 5D). These results suggest that external Zn 2ϩ results in a "partial" activation of ␣H282D␤␥ mENaCs, in contrast to a "full" activation of WT.
External Zn 2ϩ Has No Effect on ENaCs with a High Open Probability-We previously reported that ␣S580C␤␥ channels following modification by external MTSET have a high open probability and do not exhibit Na ϩ self-inhibition (9,19). If the stimulatory effect of external Zn 2ϩ on ENaC currents is due to an increase in channel open probability that would occur with a loss of Na ϩ self-inhibition, the effect of Zn 2ϩ on ENaC currents should be abolished in oocytes expressing ␣S580C␤␥ following MTSET treatment. Fig. 6 shows a typical experiment we performed to test this possibility. Addition of 100 M ZnCl 2 to the bath solution significantly increased whole cell currents in oocytes expressing ␣S580C␤␥. External application of 1 mM MTSET caused an irreversible increase in the inward current and removed the typical response for Na ϩ self-inhibition, as we previously reported (9). Subsequent application of 100 M ZnCl 2 resulted in a very minimal change in the current. These results demonstrate that external Zn 2ϩ has a minimal effect on channels with a high intrinsic open probability and suggest that Zn 2ϩ activation of ENaC reflects an increase in P o .
External Ni 2ϩ and Na ϩ Self-inhibition-We and others reported that external Ni 2ϩ is a blocker of mouse and rat ␣␤␥ ENaC expressed in Xenopus oocytes (10,11). In contrast, transepithelial Na ϩ transport in A6 cells was stimulated by external Ni 2ϩ , which was proposed to compete with extracellular Na ϩ and relieve ENaC from self-inhibition (12). We examined the effect of external Ni 2ϩ on Na ϩ self-inhibition of ␣␤␥ mENaC in oocytes. When 1 mM Ni 2ϩ was added only to the high Na ϩ bath solution, the current decline was faster and deeper than the control response in the absence of Ni 2ϩ (Fig. 7A and Table I). Furthermore, the peak and steady-state currents were lower than the currents observed in the absence of Ni 2ϩ . The enhanced current decay following the change to a high Na ϩ bath solution that contained Ni 2ϩ may reflect additive effects of The time constants () for Na ϩ self-inhibition were obtained by exponential fitting of the current decay at the clamping voltage of Ϫ60 mV following a rapid increase of bath Na ϩ concentrations from 1 to 110 mM. Both I peak and I ss are amiloride-sensitive inward Na ϩ currents expressed as negative values by convention and are not compared due to different batches of oocytes used in the experiments. Fitting for 100 M Zn 2ϩ and 1 mM Ni 2ϩ pretreatment failed to generate a time constant due to a absence of current decay. Zn 2ϩ and Ni 2ϩ were added to either the NaCl-110 solution or the NaCl-1 solution (pretreatment).

FIG. 3. Dose-dependent effect of Zn 2؉ on Na ؉ self-inhibition.
A, a typical recording shows Na ϩ self-inhibition responses in the presence of 1, 10, or 100 M ZnCl 2 in NaCl-110. A control Na ϩ self-inhibition response was performed before the Zn 2ϩ effect was examined. Na ϩ concentrations are indicated as open (1 mM) or black bars (110 mM). The trace is representative of experiments performed in six oocytes. B, dose-response curve of Zn 2ϩ effect on Na ϩ self-inhibition. The maximal inward currents measured following quick increase of bath Na ϩ from 1 to 110 mM are referred to as the peak current (I peak ), and the current measured at 40 s after the I peak measurement was referred to as the steady state current (I ss ). The I ss /I peak was used as a parameter to describe the magnitude of Na ϩ self-inhibition. A value of 1 indicates a lack of self-inhibition. The inset shows a best fitting of the mean relative response versus Zn 2ϩ concentrations with the Hill equation to estimate the EC 50 . The relative response was the difference between the mean I ss /I peak values from B in the presence and absence of Zn 2ϩ that was normalized to the difference between the mean I ss /I peak values in the presence of 100 M Zn 2ϩ (which eliminated Na ϩ self-inhibition) and the absence of Zn 2ϩ . The fitting parameters are shown in the inset. Na ϩ self-inhibition and Ni 2ϩ inhibition. In contrast to Zn 2ϩ , external Ni 2ϩ did not prevent or reverse Na ϩ self-inhibition. The I ss in NaCl-110 was further blocked by 1 mM NiCl 2 (Fig.  7B). Interestingly, the relative difference (dashed arrow in Fig.  7A) between the initial I peak prior to Ni 2ϩ application and the I ss in the presence of Ni 2ϩ was similar in magnitude to the relative difference (dashed arrow in Fig. 7B) between the I peak prior to Ni 2ϩ and the current following a sequential inhibition by high Na ϩ and Ni 2ϩ . Additional experiments were performed with oocytes expressing ␣␤␥ mENaC treated with 1 mM NiCl 2 in the low (1 mM) Na ϩ bath for 1 min. The bath was then changed to NaCl-110 in the absence of NiCl 2 . Almost no current decay was observed, and the steady-state current was reduced to a level lower than the I ss before the Ni 2ϩ pretreatment. The typical Na ϩ self-inhibition response was gradually restored after washing out Ni 2ϩ for a few minutes (Fig. 7C). It appears that Na ϩ self-inhibition is "masked" by Ni 2ϩ inhibition, which reflects a decrease in ENaC P o (10,11).
We previously observed that ␣H282D␤␥ was not blocked by external Ni 2ϩ but showed enhanced Na ϩ self-inhibition (9,11). Interestingly, Ni 2ϩ added in NaCl-110 had no effect on the Na ϩ self-inhibition response of ␣H282D␤␥ (Fig. 8A), suggesting that Ni 2ϩ inhibition and Na ϩ self-inhibition may be separate events that share a final pathway leading to a reduced channel P o . When 1 mM Ni 2ϩ was added only to the low Na ϩ bath, a slower (, 7.2 Ϯ 1.0 s, n ϭ 5, compared with 3.4 Ϯ 0.4 s without Ni 2ϩ pretreatment, n ϭ 5, p Ͻ 0.05) and smaller (I ss /I peak , 0.47 Ϯ 0.03, n ϭ 5, compared with 0.27 Ϯ 0.03 without Ni 2ϩ pretreatment, n ϭ 5, p Ͻ 0.01) response of Na ϩ self-inhibition was observed (Fig. 8A). The smaller magnitude of self-inhibition  4. Dependence of the Zn 2؉ effect on the Na ؉ concentration. Zn 2ϩ effects were examined in oocytes expressing ␣␤␥ mENaCs. A, the effects of 1 and 10 M ZnCl 2 on the I ss /I peak representing the amplitude of the Na ϩ self-inhibition responses initiated by different Na ϩ concentrations. Numbers of oocytes used in the experiments were 7, 8, and 7 for the 0 Zn 2ϩ , 1 M Zn 2ϩ , and 10 M Zn 2ϩ groups, respectively. B, the dose response of Zn 2ϩ on amiloride-sensitive Na ϩ currents measured with a 10 mM Na ϩ bath solution. Oocytes injected with WT ␣␤␥ mENaC cRNAs were kept in either regular MBS with 88 mM Na ϩ (open circle) or low Na ϩ MBS with 10 mM Na ϩ and 78 mM N-methyl-D-glucamine (open square). A clamping voltage of Ϫ100 mV was used in this experiment. Relative currents have the same meaning as in Fig. 1C. The Zn 2ϩ effects were examined in a low Na ϩ bath solution (10 mM). Amiloride (100 M) was used to define amiloride-insensitive currents. Numbers of observations were 5 for the MBS group and 14 for the 10 mM Na ϩ MBS group. C, a representative recording showing the effect of 10 M Zn 2ϩ on the steady-state current in a 10 mM Na ϩ bath solution. The oocyte was clamped to Ϫ100 mV. The trace is representative of five experiments. D, a recording showing the effect of 10 M Zn 2ϩ on Na ϩ self-inhibition initiated by 10 mM Na ϩ . The oocyte was clamped to Ϫ100 mV. After a control self-inhibition response to 10 mM Na ϩ , ZnCl 2 (10 M) was included in the 10 mM Na ϩ solution for the second test. During the third test, NaCl-1 was rapidly changed to NaCl-110 to verify that the channels exhibit a typical Na ϩ self-inhibition response. In C and D, open, dashed, and solid bars indicate 1, 10, or 110 mM Na ϩ concentrations in the bath, respectively. The gray bar indicates the period in the presence of 10 M ZnCl 2 in the 10 mM Na ϩ bath solution.
was obviously due to a reduction of I peak , because the I ss was not affected by Ni 2ϩ pretreatment. A reduction in the outward current in the low Na ϩ bath (Fig. 8A) was also observed. These data suggest that Ni 2ϩ inhibits ␣H282D␤␥ channels when the extracellular Na ϩ concentration is low (no self-inhibition) but does not inhibit ␣H282D␤␥ when the extracellular Na ϩ concentration is high (strong self-inhibition). Mutations of ␥H239 led to a loss of Na ϩ self-inhibition and a partial loss of Ni 2ϩ inhibition of ENaC (Fig. 5 and Refs. 9 and 11). Although ab␥H239D does not exhibit Na ϩ self-inhibition or Zn 2ϩ activation, the addition of 1 mM Ni 2ϩ to the NaCl-110 bath resulted in a modest current decay (Fig. 8B). DISCUSSION Our results demonstrate that external Zn 2ϩ stimulates amiloride-sensitive Na ϩ currents in oocytes expressing ␣␤␥ mENaC in a dose-dependent manner in association with a loss of Na ϩ self-inhibition. The stimulatory effect of Zn 2ϩ depends on the extracellular Na ϩ concentration, a key factor in the Na ϩ self-inhibition response. Extracellular divalent cations, including Zn 2ϩ , Cd 2ϩ , Cu 2ϩ , and Ni 2ϩ , have been previously shown to stimulate amiloride-sensitive Na ϩ currents in Na ϩ -transporting epithelial tissues such as frog skin and toad urinary bladder (5,20). Although external Ni 2ϩ is a blocker of mouse and rat ENaCs expressed in Xenopus oocytes (10,11), extracellular Ni 2ϩ was found to stimulate amiloride-sensitive short-circuit current in A6 cells by competitively interfering with Na ϩ selfinhibition (12). The differences in the response to external Ni 2ϩ may reflect species differences in the structures that mediate Ni 2ϩ binding and Na ϩ self-inhibition. We anticipate that future studies will address whether other members of ENaC/DEG family are also regulated by external transition metals. For example, extracellular Zn 2ϩ facilitates H ϩ activation of the acid sensing ionic channels that are homologous to ENaCs (21).
The stimulatory effect of external Zn 2ϩ on ENaC currents that we observed contrasts with the voltage-dependent block of the endogenous and expressed ENaCs by external Zn 2ϩ that was recently reported by Amuzescu and co-workers (13). It is not clear why opposite effects of Zn 2ϩ on ENaC activity were observed. The stimulatory effect of external Zn 2ϩ on ENaC is dependent on the magnitude of Na ϩ self-inhibition and on the extracellular Na ϩ concentration (Fig. 4). External Zn 2ϩ only stimulates ENaC currents measured in bath Na ϩ concentration greater than the minimal concentration required for Na ϩ self-inhibition (ϳ10 mM). The blocking effect of Zn 2ϩ on ENaC currents reported by Amuzescu et al. was observed in oocytes bathed in a solution containing 10 mM Na ϩ (13) at Zn 2ϩ concentrations of 1-10 M. We have not observed an inhibitory effect of extracellular Zn 2ϩ in the concentration range of 10 nM to 1 mM when oocytes were bathed in a solution containing 10 mM Na ϩ (Fig. 4B). Amuzescu et al. also reported that 100 M Zn 2ϩ stimulated ␣␤␥ rENaC, in agreement with our findings (13). Although ENaCs derived from different species were used in these two studies, it is unlikely that the conflicting results are due to species differences, because rat and mouse ENaCs have Ͼ95% amino acid identity (14). Furthermore, we observed similar activation of the whole cell Na ϩ currents and elimination of Na ϩ self-inhibition by external Zn 2ϩ in oocytes expressing human ␣␤␥ ENaCs. 2 Our results are also consistent with a previous report (22) demonstrating a lack of an effect of external Zn 2ϩ at the concentrations up to 1 mM on inward Li ϩ currents in oocytes expressing ␣␤␥ rENaC and bathed in a low Li ϩ (20 mM) bath.
Our results suggest that external Zn 2ϩ activates ENaC by reversing or preventing Na ϩ self-inhibition, although the mechanism by which Zn 2ϩ alters the Na ϩ self-inhibition re-2 S. Sheng, C. J. Perry, and T. R. Kleyman, unpublished data.
FIG. 5. Zn 2؉ activation is absent in oocytes expressing ␣␤␥ mENaCs with ␥H239 mutations. Open, solid, and gray bars have the same meanings as in other figures. Application of 10 M amiloride is indicated by an arrow. All experiments in this figure have been repeated at least six times, and the results are indistinguishable. A, lack of effect of 100 M Zn 2ϩ on the Na ϩ self-inhibition response in an oocyte expressing ␣␤␥H239R. B, dose-response curve for external Zn 2ϩ effect on amiloride-sensitive Na ϩ currents measured at Ϫ60 mV and normalized to control values in oocytes expressing ␣␤␥H239R. Experiments were performed with the same protocol as for Fig. 1C. C, a recording with a protocol similar to A in an oocyte expressing ␣H282D␤␥ mENaC. D, a representative recording shows the effect of 100 M Zn 2ϩ on the Na ϩ current following a Na ϩ self-inhibition response in an oocyte expressing ␣H282D␤␥. sponse is unclear. Na ϩ self-inhibition appears to be dependent on an external Na ϩ sensor or receptor. Zn 2ϩ may interfere with Na ϩ binding to its receptor by competing with Na ϩ for the same binding site or blocking Na ϩ access to the site. If so, the conformational changes in association with Na ϩ binding that result in a decrease in channel P o do not occur when Zn 2ϩ is bound to this site. Alternatively, Zn 2ϩ binding to the channel may interfere with the conformational changes that are induced by Na ϩ binding to an external site and are required for Na ϩ self-inhibition. It is also possible that Zn 2ϩ increases ENaC P o through a mechanism that is distinct from Na ϩ selfinhibition. Several observations support the possibility that Zn 2ϩ and Na ϩ interact with the channel at a common external site, including (i) the dependence of Zn 2ϩ activation on the external Na ϩ concentration, (ii) the rapid activation of ENaC currents ( of 5 s) by external Zn 2ϩ and the rapid reversal of this effect following the removal of Zn 2ϩ , and (iii) a Zn 2ϩ -dependent rightward shift in Na ϩ concentration versus I ss /I peak curves.
Alkali metal ions such as Na ϩ favor oxygen atoms as a binding ligand, whereas Zn 2ϩ is an intermediate divalent cation that prefers nitrogen atoms from imidazoles and sulfur atoms from Cys side chains as binding ligands. Because the coordination chemistries of Zn 2ϩ and Na ϩ differ, the presence of a common (or overlapping) binding site for both Zn 2ϩ and Na ϩ would impose certain constraints, such that some residues participate in the coordination of both Zn 2ϩ and Na ϩ , whereas other residues would participate in coordination of either Zn 2ϩ or Na ϩ .
We recently reported that a His residue within the extracellular loop of ␥mENaC (H239) is required for Na ϩ self-inhibition, because substitutions of ␥H239 with Arg, Asp, or Cys eliminated Na ϩ self-inhibition (9). Our current study suggests that ␥H239 is also required for Zn 2ϩ activation of ENaC. Based on these observations, we propose that ␥H239 may provide a binding ligand for both Na ϩ and Zn 2ϩ . A direct -interaction between Na ϩ and an imidazole nitrogen atom has been observed in resolved structures (23). A Na ϩ is coordinated by three imidazole nitrogen atoms and one water oxygen atom in the structure of a mannose-binding protein (PDB ID, 1BCH, see Fig. 9A). Given the fact that the most preferred coordination numbers of Na ϩ are 6 and 8, we speculate that Na ϩ may be coordinated by a number of different ligands, including the ␥H239 nitrogen, as well as oxygen atoms from other residues or from the solvent.
Based on our current observations and previous reports, we present two working models to illustrate two possible mechanisms for Na ϩ self-inhibition and the Zn 2ϩ effect on self-inhibition (Fig. 9, B and C). In both models, Na ϩ binds to a site within the extracellular allosteric regulatory sites (EARS) of the ENaC ECLs that we previously proposed (9) and induces local conformational changes that are transmitted to a putative gate located at the outer pore of the channel (19,24,25). Working model 1 was generated to emphasize an overlapping binding site for Na ϩ and Zn 2ϩ within the EARS. Residue ␥H239 is proposed to provide coordination for both Na ϩ and Zn 2ϩ based on our observation of the loss of both Na ϩ selfinhibition and Zn 2ϩ activation of channels with mutations of FIG. 7. Effects of external Ni 2؉ on Na ؉ self-inhibition of ␣␤␥ mENaC. Oocytes were injected with cRNAs for WT ␣␤␥ mENaC subunits and clamped at Ϫ60 mV. External applications of 1 mM NiCl 2 are indicated by bold dashed lines. Black arrows indicate the addition of 10 M amiloride. All recordings are representative of at least six independent observations. A, a representative recording showing the effect of Ni 2ϩ added in NaCl-110 on Na ϩ self-inhibition. Following a typical Na ϩ self-inhibition response, NaCl-1 was quickly replaced by NaCl-110 containing 1 mM NiCl 2 . The dashed arrow shows the difference between the control I peak and the I ss in the presence of 1 mM NiCl 2 , which represents the total inhibition by the simultaneous application of Ni 2ϩ and 110 mM Na ϩ . B, a recording showing the inhibitory effect of Ni 2ϩ on the steady-state current. NiCl 2 was added after the Na ϩ current reached a steady state and was then washed out to show the poor reversibility of Ni 2ϩ block. The dashed arrow indicates the difference between I peak in the absence of 1 mM NiCl 2 and the current measured after addition of 1 mM NiCl 2 into the NaCl-110 bath, which represents the inhibition due to sequential application of high Na ϩ and Ni 2ϩ . C, a trace showing the effect of Ni 2ϩ pretreatment on Na ϩ self-inhibition. The oocyte was bathed in NaCl-1 containing 1 mM NiCl 2 for 1 min and then perfused with Ni 2ϩ -free NaCl-110. Following washout of Ni 2ϩ , two consecutive responses for Na ϩ self-inhibition were tested to show recovery from the Ni 2ϩ effect.
␥H239. The role of ␣H282 in Na ϩ and Zn 2ϩ binding is not clear, because mutation of ␣H282 to an Arg, Asp, Cys, or Trp residue result in enhanced Na ϩ self-inhibition (9), and the magnitude of Zn 2ϩ activation of ␣H282 mutants was similar to wild type ENaC. The Na ϩ binding site is shown to contain several oxygen atoms from unspecified residues or solvent, because Na ϩ coordination often involves multiple oxygen atoms. Other binding ligands for Zn 2ϩ are also included in this model based on a favorable Zn 2ϩ coordination pattern and on previous work on Na ϩ self-inhibition. Some Zn 2ϩ coordination shells involve a number of different ligands, including nitrogen, sulfur, and perhaps oxygen atoms. In addition to the imidazole nitrogen of ␥H239, other moieties such as -SH groups from Cys residues that are abundant within the ECLs of ENaC subunits may participate in forming a high affinity binding site for Zn 2ϩ . Reagents that are capable of reacting with sulfhydryl groups, as well as extracellular cations that are often coordinated by Cys (Cd 2ϩ , Zn 2ϩ , and Cu 2ϩ ), have been shown to stimulate Na ϩ transport across model epithelia, a response that likely reflects a loss of Na ϩ self-inhibition (5). According to this model, Zn 2ϩ binding would prevent or reverse Na ϩ binding and thus eliminate Na ϩ -induced conformational changes that ultimately lead to a P o reduction. The second model (Fig. 9C) was generated to show an alternative mechanism of the Zn 2ϩ effect on Na ϩ self-inhibition. Residue ␥H239 is proposed to reside in a putative gate, which swings in to close the pore during channel closing and not to interact directly with either Zn 2ϩ or Na ϩ . In this model, Na ϩ and Zn 2ϩ bind to their sites within the EARS that may or may not overlap. The conformational changes (shown as magenta arrows) induced by Na ϩ binding are passed to the gate and promote the gate to close the outer pore. In contrast, Zn 2ϩ binding induces the opposite conformational changes (shown as green arrows) within the EARS that hold the gate in a location to keep the pore open. Thus Zn 2ϩ prevents or reverses Na ϩ -induced conformational changes that promote channel closing and eliminates Na ϩ self-inhibition. Consistent with this possibility, external Zn 2ϩ did not activate currents in oocytes expressing ␣S580C␤␥ following MTSET treatment, which locks the channels in a fully open state.
We and others previously reported that external Ni 2ϩ inhibits whole cell amiloride-sensitive Na ϩ currents due to a decrease in ENaC open probability (10,11). Although Ni 2ϩ and Zn 2ϩ share several properties such as size, charge, and binding ligands, they exhibit opposing effects on Na ϩ self-inhibition. Pretreatment of oocytes expressing ␣␤␥ mENaC with 1 mM Ni 2ϩ in the low Na ϩ bath solution prevented the typical current decay following an increase in bath Na ϩ concentration, and the steady-state currents were below the I ss levels measured prior to Ni 2ϩ pretreatment and following Ni 2ϩ washout (Fig. 7C). A more rapid and deeper Na ϩ self-inhibition response was observed when Ni 2ϩ was present only in the NaCl-110 solution. This response likely reflects additive inhibitory effects of Na ϩ and Ni 2ϩ . Although Zn 2ϩ activation appears to be dependent on Na ϩ self-inhibition, the effect of Ni 2ϩ on ENaC activity is independent of Na ϩ self-inhibition, because Ni 2ϩ inhibits ENaC activity when added to either the NaCl-1 or NaCl-110 solution. We previously reported that ␣H282D␤␥ is not blocked by Ni 2ϩ (11). In agreement with our previous observations, the addition of Ni 2ϩ to the NaCl-110 solution did not alter the Na ϩ self-inhibition response of ␣H282D␤␥ (9). It is possible that Ni 2ϩ binds to a site that overlaps with Zn 2ϩ and Na ϩ binding sites. Other investigators have suggested that Ni 2ϩ competes with Na ϩ for a common binding site (12). Our results suggest that Ni 2ϩ -bound ENaC has a low open probability (11) regardless of the external Na ϩ concentration and that the Na ϩ selfinhibition response is "masked" by Ni 2ϩ . Na ϩ self-inhibition and Ni 2ϩ inhibition of ENaC may share a common mechanism.
Epithelial Na ϩ channels are expressed in the distal nephron, distal colon, and within both the airway and alveolae. The concentrations of Na ϩ in the distal nephron where ENaCs are expressed are highly variable. For example, human urine Na ϩ concentrations vary in a wide range from Ͻ10 mM to Ͼ100 mM and are altered by changes in dietary Na ϩ intake as well as by changes in extracellular fluid volume (26). Luminal Na ϩ concentrations in rat late distal tubules are reported in the range of 24 -48 mM (27). Colonic fluid Na ϩ concentrations are likely variable as well. Recent studies suggest that Na ϩ concentrations in distal airway surface liquid are high (Ͼ100 mM) (28). Na ϩ concentrations in the lumen of ENaC-expressing tissues are frequently above the minimal concentration (ϳ30 mM) at which Na ϩ self-inhibition is observed, suggesting that Na ϩ self-inhibition may participate in the physiological regulation of transepithelial Na ϩ transport (6). These observations, together with our data indicating that external Zn 2ϩ at concentrations in the low micromolar range enhances Na ϩ influx via ENaC by relieving Na ϩ self-inhibition, suggest that extracellular Zn 2ϩ may be a physiological regulator of ENaC. Urinary Zn 2ϩ concentrations of ϳ5 M have been reported in humans (29,30), and given the low concentration of proteins and amino acids in urine that could potentially bind Zn 2ϩ , free urinary Zn 2ϩ concentrations likely reach levels that affect ENaC activity. Because increases in rates of urinary Zn 2ϩ excretion may be associated with increases in ENaC activity in collecting ducts by relieving Na ϩ self-inhibition, we propose that increases in the urinary concentration of Zn 2ϩ may increase the risk of developing salt-sensitive hypertension. Angiotensinconverting enzyme has a key role in blood pressure regulation FIG. 8. Effects of external Ni 2؉ on the Na ؉ self-inhibition response of mutant ENaCs. Oocytes were injected with cRNAs for one mutant subunit with other two WT subunits and clamped at Ϫ60 mV. A, a representative recording showing effects of Ni 2ϩ on Na ϩ selfinhibition in an oocyte expressing ␣H282D␤␥. NiCl 2 (1 mM) was applied in either the high or low Na ϩ bath solution as indicated by a dashed line. The tracing is representative of five experiments. B, a recording showing responses of Na ϩ self-inhibition prior to, in the presence of, and following washout of 1 mM NiCl 2 in NaCl-110. The tracing is representative of six experiments. Amiloride at 10 M was added as indicated by an arrow. and is a Zn 2ϩ -dependent protease (31). An association between increases in erythrocyte Zn 2ϩ and essential hypertension has been reported (32). A relationship between urinary Zn 2ϩ excretion and blood pressure has not been established, although studies with small numbers of patients have suggested an association between increases in urinary Zn 2ϩ excretion and hypertension (33,34).
In summary, our results demonstrate that extracellular Zn 2ϩ activates mouse ␣␤␥ ENaCs expressed in oocytes bathed in high Na ϩ by eliminating Na ϩ self-inhibition. Our previous results and work from other investigators suggest that ENaC gating is regulated in an allosteric manner by extracellular cues (5,9,11). These observations raise the possibility that ENaC is a ligand-gated channel, similar to ATP-sensitive inward rectifier K ϩ channels (K ATP ) (35), cyclic nucleotide-gated channels (36), and glutamate receptor ion channels (37). Extracellular Na ϩ would serve as a ligand whose binding to ENaC reduces channel open probability, and Zn 2ϩ prevents or reverses Na ϩ self-inhibition as a potential physiological regulator or ligand. FIG. 9. Working models for the mechanisms of Na ؉ self-inhibition and Zn 2؉ activation of ENaC. A, the Na ϩ coordination shell is displayed in ball-and-cylinder mode from coordinates of a structure of mannose-binding protein (PDB ID, 1BCH). The colors are cyan for carbon, red for oxygen, and blue for nitrogen. A Na ϩ is shown as a magenta sphere, and a water molecule is shown as a red sphere. The Na ϩ is coordinated by a water molecule, and three ␦-nitrogen atoms from histidine residues that are located in three helices from three identical peptide chains A, B, and C. B, working model 1 illustrating overlapping putative Na ϩ and Zn 2ϩ binding sites. ENaC pore is shown in blue, and a bound Na ϩ is displayed as a magenta sphere with putative coordinating oxygen atoms in red. The yellow and green portions represent parts of the extracellular loops of ENaC subunits with an emphasis on the hypothesized extracellular allosteric regulatory site (EARS). A putative Zn 2ϩ binding site depicted as a green dashed circle is formed by two imidazole nitrogen atoms from ␥H239 and an unidentified histidine, a sulfur atom from a unidentified cysteine residue, and an oxygen atom from the backbone or side chain of an unidentified residue. The Na ϩ binding site depicted as a magenta dashed circle is constituted by one imidazole nitrogen atom from ␥H239 and several oxygen atoms from solvent or unidentified residues. The blue arrows indicate potential movements of the outer pore in ENaC gating as suggested by us and other investigators (19,24,25). The green arrows indicate possible movements that transmit the local conformational changes induced by Na ϩ binding to its receptor onto a putative gate. C, working model 2 showing ␥H239 as a key residue in a putative gate that is in a location distinct from the Na ϩ and Zn 2ϩ binding sites. The blue and yellow portions represent the pore and the EARS within ECL of ENaC, respectively. A putative gate within ENaC is shown as a black rod whose movement during channel closing is indicated as a black arched arrow. In the model, Zn 2ϩ and Na ϩ binding sites are proposed within the EARS and may or may not overlap. The directions of conformational changes induced by Na ϩ and Zn 2ϩ are shown as magenta and green arrows, respectively. Part of the EARS is proposed to interact with the putative gate during channel gating.