Gating of the bacterial sodium channel, NaChBac: voltage-dependent charge movement and gating currents

The bacterial sodium channel, NaChBac, from Bacillus halodurans provides an excellent model to study structure-function relationships of voltage-gated ion channels. It can be expressed in mammalian cells for functional studies as well as in bacterial cultures as starting material for protein purification for fine biochemical and biophysical studies. Macroscopic functional properties of NaChBac have been described previously (Ren, D., B. Navarro, H. Xu, L. Yue, Q. Shi, and D.E. Clapham. 2001. Science. 294:2372-2375). In this study, we report gating current properties of NaChBac expressed in COS-1 cells. Upon depolarization of the membrane, gating currents appeared as upward inflections preceding the ionic currents. Gating currents were detectable at -90 mV while holding at -150 mV. Charge-voltage (Q-V) curves showed sigmoidal dependence on voltage with gating charge saturating at -10 mV. Charge movement was shifted by -22 mV relative to the conductance-voltage curve, indicating the presence of more than one closed state. Consistent with this was the Cole-Moore shift of 533 micros observed for a change in preconditioning voltage from -160 to -80 mV. The total gating charge was estimated to be 16 elementary charges per channel. Charge immobilization caused by prolonged depolarization was also observed; Q-V curves were shifted by approximately -60 mV to hyperpolarized potentials when cells were held at 0 mV. The kinetic properties of NaChBac were simulated by simultaneous fit of sodium currents at various voltages to a sequential kinetic model. Gating current kinetics predicted from ionic current experiments resembled the experimental data, indicating that gating currents are coupled to activation of NaChBac and confirming the assertion that this channel undergoes several transitions between closed states before channel opening. The results indicate that NaChBac has several closed states with voltage-dependent transitions between them realized by translocation of gating charge that causes activation of the channel.     

Effects of bound Ts3 on voltage dependence of sodium channel transitions to and from inactivation and energetics of its unbinding

Recently, we proposed a quantitative model to explain the molecular mechanism of action of the Tityus serrulatus Ts3 α-toxin on sodium channels. In this model, the toxin acts as a stop that prevents the segment S4 of domain IV from reaching its outermost position, thus impairing the normal fast inactivation without affecting activation. In the present work, we analyze the predictions of the proposed model with regard to the voltage-dependent transitions to and from inactivation. Our results show that the recovery from inactivation was significantly faster in Ts3-bound channels and that there was no significant voltage dependence. The transition to inactivated state from open state in Ts3-modified channels presented a small but significant voltage dependence, which may derive from an intrinsic voltage dependence of inactivation or by a short movement of IVS4 in the presence of bound Ts3. We also studied the thermodynamic parameters of the voltage-dependent displacement of Ts3 from its binding site. We have observed that the activation energy to remove the toxin is 27 kJ/mol, part of which derives from the imposed depolarizing potential and the movement of an equivalent electrical charge of 0.54 c ₀. These results support the proposed model.     

N type rapid inactivation in human Kv1.4 channels: functional role of a putative C-terminal helix

Voltage gated potassium channels are tetrameric membrane proteins, which have a central role in cellular excitability. Human Kv1.4 channels open on membrane depolarization and inactivate rapidly by a ‘ball and chain’ mechanism whose molecular determinants have been mapped to the cytoplasmic N terminus of the channel. Here we show that the other terminal end of the channel also plays a role in channel inactivation. Swapping the C-terminal residues of hKv1.4 with those from two non-inactivating channels (hKv1.1 and hKv1.2) affects the rates of inactivation, as well as the recovery of the channel from the inactivated state. Secondary structure predictions of the hKv1.4 sequence reveal a helical structure at its distal C-terminal. Complete removal or partial disruption of this helical region results in channels with remarkably slowed inactivation kinetics. The ionic selectivity and voltage-dependence of channel opening were similar to hKv1.4, indicative of an unperturbed channel pore. These results demonstrate that fast inactivation is modulated by structural elements in the C-terminus, suggesting that the process involves the concerted action of the N- and C-termini.     

The role of intracellular sodium in the regulation of NMDA-receptor-mediated channel activity and toxicity

Sodium (Na⁺) is the major cation in extracellular space and, with its entry into cells, may act as a critical intracellular second messenger that regulates many cellular functions. Through our investigations of mechanisms underlying the activity-dependent regulation of N-methyl-d-aspartate (NMDA) receptors, we recently characterized intracellular Na⁺ as a possible signaling factor common to processes underlying the upregulation of NMDA receptors by non-NMDA glutamate channels, voltage-gated Na⁺ channels, and remote NMDA receptors. Furthermore, although Ca²⁺ influx during the activation of NMDA receptors acts as a negative feedback mechanism that downregulates NMDA receptor activity, Na⁺ influx provides an essential positive feedback mechanism to overcome Ca²⁺-induced inhibition, thereby potentiating both NMDA receptor activity and inward Ca²⁺ flow. NMDA receptors may be recruited to cause excitoxicity through a Na⁺-dependent mechanism. Therefore, the further characterization of mechanisms underlying the regulation of NMDA receptors by intracellular Na⁺ is essential to understanding activity-dependent neuroplasticity in the nervous system.     

Domain III S4 in closed-state fast inactivation: Insights from a periodic paralysis mutation

Heterologous expression of sodium channel mutations in hypokalemic periodic paralysis reveals 2 variants on channel dysfunction. Charge-reducing mutations of voltage sensing S4 arginine residues alter channel gating as typically studied with expression in mammalian cells. These mutations also produce leak currents through the voltage sensor module, as typically studied with expression in Xenopus oocytes. DIIIS4 mutations at R3 in the skeletal muscle sodium channel produce gating defects and omega current consistent with the phenotype of reduced excitability. Here, we confirm DIIIS4 R3C gating defects in the oocyte expression system for fast inactivation and its recovery. We provide novel data for the effects of the cysteine mutation on voltage sensor movement, to further our understanding of sodium channel defects in hypokalemic periodic paralysis. Gating charge movement and its remobilization are selectively altered by the mutation at hyperpolarized membrane potential, as expected with reduced serum potassium.     

A physical model of sodium channel gating

Most current models of membrane ion channel gating are abstract compartmental models consisting of many undefined states connected by rate constants arbitrarily assigned to fit the known kinetics. In this paper is described a model with states that are defined in terms of physically plausible real systems which is capable of describing accurately most of the static and dynamic properties measured for the sodium channel of the squid axon. The model has two components. The Q-system consists of charges and dipoles that can move in response to an electric field applied across the membrane. It would contain and may compose the gating charge that is known to transfer prior to channel opening. The N-system consists of a charged group or dipole that is constrained to move only in the plane of the membrane and thus does not interact directly with the trans-membrane electric field but can interact electrostatically with the Q-system. The N-system has only two states, its resting state (channel closed) and its excited state (channel open) and its response time is very short in comparison with that of the Q-system. On depolarizing the membrane the the N-system will not make a transition to its open state until a critical amount of Q-charge transfer has occurred. Using only four adjustable parameters that are fully determined by fitting the equilibrium properties of the model to those of the sodium channel in the squid axon, the model is then able to describe with some accuracy the kinetics of channel opening and closing and includes the Cole and Moore delay. In addition to these predictions of the behaviour of assemblies of channels the model predicts some of the individual channel properties measured by patch clamp techniques.     

Uncoupling of gating charge movement and closure of the ion pore during recovery from inactivation in the Kv1.5 channel

Both wild-type (WT) and nonconducting W472F mutant (NCM) Kv1.5 channels are able to conduct Na(+) in their inactivated states when K(+) is absent. Replacement of K(+) with Na(+) or NMG(+) allows rapid and complete inactivation in both WT and W472F mutant channels upon depolarization, and on return to negative potentials, transition of inactivated channels to closed-inactivated states is the first step in the recovery of the channels from inactivation. The time constant for immobilized gating charge recovery at -100 mV was 11.1 +/- 0.4 ms (n = 10) and increased to 19.0 +/- 1.6 ms (n = 3) when NMG(+)(o) was replaced by Na(+)(o). However, the decay of the Na(+) tail currents through inactivated channels at -100 mV had a time constant of 129 +/- 26 ms (n = 18), much slower than the time required for gating charge recovery. Further experiments revealed that the voltage-dependence of gating charge recovery and of the decay of Na(+) tail currents did not match over a 60 mV range of repolarization potentials. A faster recovery of gating charge than pore closure was also observed in WT Kv1.5 channels. These results provide evidence that the recovery of the gating elements is uncoupled from that of the pore in Na(+)-conducting inactivated channels. The dissociation of the gating charge movements and the pore closure could also be observed in the presence of symmetrical Na(+) but not symmetrical Cs(+). This difference probably stems from the difference in the respective abilities of the two ions to limit inactivation to the P-type state or prevent it altogether.     

The Na+ channel inactivation gate is a molecular complex: a novel role of the COOH-terminal domain

Electrical activity in nerve, skeletal muscle, and heart requires finely tuned activity of voltage-gated Na+ channels that open and then enter a nonconducting inactivated state upon depolarization. Inactivation occurs when the gate, the cytoplasmic loop linking domains III and IV of the alpha subunit, occludes the open pore. Subtle destabilization of inactivation by mutation is causally associated with diverse human disease. Here we show for the first time that the inactivation gate is a molecular complex consisting of the III-IV loop and the COOH terminus (C-T), which is necessary to stabilize the closed gate and minimize channel reopening. When this interaction is disrupted by mutation, inactivation is destabilized allowing a small, but important, fraction of channels to reopen, conduct inward current, and delay cellular repolarization. Thus, our results demonstrate for the first time that physiologically crucial stabilization of inactivation of the Na+ channel requires complex interactions of intracellular structures and indicate a novel structural role of the C-T domain in this process.     

FGF13 modulates the gating properties of the cardiac sodium channel Nav1.5 in an isoform-specific manner

FGF13 (FHF2), the major fibroblast growth factor homologous factor (FHF) in rodent heart, directly binds to the C-terminus of the main cardiac sodium channel, Na_V1.5. Knockdown of FGF13 in cardiomyocytes induces slowed ventricular conduction by altering Na_V1.5 function. FGF13 has five splice variants, each of which possess the same core region and C terminus but differing in their respective N termini. Whether and how these alternatively spliced N termini impart isoform-specific regulation of Na_V1.5, however, has not been reported. Here, we exploited a heterologous expression to explore the specific modulatory effects of FGF13 splice variants FGF13S, FGF13U and FGF13YV on Na_V1.5 function. We found these three splice variants differentially modulated Na_V1.5 current density. Although steady-state activation was unaltered by any of the FGF13 isoforms (compared to control cells expressing Nav1.5 but not expressing FGF13), open-state fast inactivation and closed-state fast inactivation were markedly slowed, steady-state availability was significantly shifted toward the depolarizing direction, and the window current was increased by each of FGF13 isoforms. Most strikingly, FGF13S hastened the rate of Na_V1.5 entry into the slow inactivation state and induced a dramatic slowing of recovery from inactivation, which caused a large decrease in current after either low or high frequency stimulation. Overall, these data showed the diversity of the roles of the FGF13 N-termini in Na_V1.5 channel modulation and suggested the importance of isoform-specific regulation.     

Computing transient gating charge movement of voltage-dependent ion channels

The opening of voltage-gated sodium, potassium, and calcium ion channels has a steep relationship with voltage. In response to changes in the transmembrane voltage, structural movements of an ion channel that precede channel opening generate a capacitative gating current. The net gating charge displacement due to membrane depolarization is an index of the voltage sensitivity of the ion channel activation process. Understanding the molecular basis of voltage-dependent gating of ion channels requires the measurement and computation of the gating charge, Q. We derive a simple and accurate semianalytic approach to computing the voltage dependence of transient gating charge movement (Q–V relationship) of discrete Markov state models of ion channels using matrix methods. This approach allows rapid computation of Q–V curves for finite and infinite length step depolarizations and is consistent with experimentally measured transient gating charge. This computational approach was applied to Shaker potassium channel gating, including the impact of inactivating particles on potassium channel gating currents.     

Movement of voltage sensor S4 in domain 4 is tightly coupled to sodium channel fast inactivation and gating charge immobilization.

The highly charged transmembrane segments in each of the four homologous domains (S4D1-S4D4) represent the principal voltage sensors for sodium channel gating. Hitherto, the existence of a functional specialization of the four voltage sensors with regard to the control of the different gating modes, i.e., activation, deactivation, and inactivation, is problematic, most likely due to a functional coupling between the different domains. However, recent experimental data indicate that the voltage sensor in domain 4 (S4D4) plays a unique role in sodium channel fast inactivation. The correlation of fast inactivation and the movement of the S4D4 voltage sensor in rat brain IIA sodium channels was examined by site-directed mutagenesis of the central arginine residues to histidine and by analysis of both ionic and gating currents using a high expression system in Xenopus oocytes and an optimized two-electrode voltage clamp. Mutation R1635H shifts the steady state inactivation to more hyperpolarizing potentials and drastically increases the recovery time constant, thereby indicating a stabilized inactivated state. In contrast, R1638H shifts the steady state inactivation to more depolarizing potentials and strongly increases the inactivation time constant, thereby suggesting a preferred open state occupancy. The double mutant R1635/1638H shows intermediate effects on inactivation. In contrast, the activation kinetics are not significantly influenced by any of the mutations. Gating current immobilization is markedly decreased in R1635H and R1635/1638H but only moderately in R1638H. The time courses of recovery from inactivation and immobilization correlate well in wild-type and mutant channels, suggesting an intimate coupling of these two processes that is maintained in the mutations. These results demonstrate that S4D4 is one of the immobilized voltage sensors during the manifestation of the inactivated state. Moreover, the presented data strongly suggest that S4D4 is involved in the control of fast inactivation.     

Allosteric voltage gating of potassium channels II. Mslo channel gating charge movement in the absence of Ca(2+).

Large-conductance Ca(2+)-activated K(+) channels can be activated by membrane voltage in the absence of Ca(2+) binding, indicating that these channels contain an intrinsic voltage sensor. The properties of this voltage sensor and its relationship to channel activation were examined by studying gating charge movement from mSlo Ca(2+)-activated K(+) channels in the virtual absence of Ca(2+) (<1 nM). Charge movement was measured in response to voltage steps or sinusoidal voltage commands. The charge-voltage relationship (Q-V) is shallower and shifted to more negative voltages than the voltage-dependent open probability (G-V). Both ON and OFF gating currents evoked by brief (0.5-ms) voltage pulses appear to decay rapidly (tau(ON) = 60 microseconds at +200 mV, tau(OFF) = 16 microseconds at -80 mV). However, Q(OFF) increases slowly with pulse duration, indicating that a large fraction of ON charge develops with a time course comparable to that of I(K) activation. The slow onset of this gating charge prevents its detection as a component of I(gON), although it represents approximately 40% of the total charge moved at +140 mV. The decay of I(gOFF) is slowed after depolarizations that open mSlo channels. Yet, the majority of open channel charge relaxation is too rapid to be limited by channel closing. These results can be understood in terms of the allosteric voltage-gating scheme developed in the preceding paper (Horrigan, F.T., J. Cui, and R.W. Aldrich. 1999. J. Gen. Physiol. 114:277-304). The model contains five open (O) and five closed (C) states arranged in parallel, and the kinetic and steady-state properties of mSlo gating currents exhibit multiple components associated with C-C, O-O, and C-O transitions.     

The gating currents of sodium channels: Pore-population-size effects

Sodium-channel behavior has been modeled in order to determine the answer to the following question: How large must a population of “on-off” Sodium pores be before the inherently random behavior of the individual channels becomes smoothed to yield the expected gating current-conductance relationships which would be predicted from an infinite pore array? Results of this analysis show that for the “opening” situation, an excellent fit was obtained whenever more than about 10 pores were considered. Significant discrepanciesd were observed in the “Closeing” situation, however, for pore arrays of 50 or less. Marked hysteresis is apparent in the behavior of small pore populations.     

The early phase of sodium channel gating current in the squid giant axon

A fast component of displacement current which accompanies the sodium channel gating current has been recorded from the membrane of the giant axon of the squid Loligo forbesii. This component is characterized by relaxation time constants typically shorter than 25 µs. The charge displaced accounts for about 10% (or 2 nC/cm²) of the total displacement charge attributed to voltage-dependent sodium channels. Using a low noise, wide-band voltage clamp system and specially designed voltage step protocols we could demonstrate that this component: (i) is not a recording artifact; (ii) is kinetically independent from the sodium channel activation and inactivation processes; (iii) can account for a significant fraction of the initial amplitude of recorded displacement current and (iv) has a steady state charge transfer which saturates for membrane potentials above + 20 mV and below – 100 mV This component can be modelled as a single step transition using the Eyring-Boltzmann formalism with a quantal charge of 1 e^– and an asymmetrical energy barrier. Furthermore, if it were associated with the squid sodium channel, our data would suggest one fast transition per channel. A possible role as a sodium channel activation trigger, which would still be consistent with kinetic independence, is discussed. Despite uncertainties about its origin, the property of kinetic independence allows subtraction of this component from the total displacement current to reveal a rising phase in the early time course of the remaining current. This will have to be taken into account when modelling the voltage-dependent sodium channel.     

The gating charge pathway of an epilepsy-associated potassium channel accommodates chemical ligands

Voltage-gated potassium (Kv) channels derive their voltage sensitivity from movement of gating charges in voltage-sensor domains (VSDs). The gating charges translocate through a physical pathway in the VSD to open or close the channel. Previous studies showed that the gating charge pathways of Shaker and Kv1.2-2.1 chimeric channels are occluded, forming the structural basis for the focused electric field and gating charge transfer center. Here, we show that the gating charge pathway of the voltage-gated KCNQ2 potassium channel, activity reduction of which causes epilepsy, can accommodate various small molecule ligands. Combining mutagenesis, molecular simulation and electrophysiological recording, a binding model for the probe activator, ztz240, in the gating charge pathway was defined. This information was used to establish a docking-based virtual screening assay targeting the defined ligand-binding pocket. Nine activators with five new chemotypes were identified, and in vivo experiments showed that three ligands binding to the gating charge pathway exhibit significant anti-epilepsy activity. Identification of various novel activators by virtual screening targeting the pocket supports the presence of a ligand-binding site in the gating charge pathway. The capability of the gating charge pathway to accommodate small molecule ligands offers new insights into the gating charge pathway of the therapeutically relevant KCNQ2 channel.     

Tracking voltage-dependent conformational changes in skeletal muscle sodium channel during activation

The primary voltage sensor of the sodium channel is comprised of four positively charged S4 segments that mainly differ in the number of charged residues and are expected to contribute differentially to the gating process. To understand their kinetic and steady-state behavior, the fluorescence signals from the sites proximal to each of the four S4 segments of a rat skeletal muscle sodium channel were monitored simultaneously with either gating or ionic currents. At least one of the kinetic components of fluorescence from every S4 segment correlates with movement of gating charge. The fast kinetic component of fluorescence from sites S216C (S4 domain I), S660C (S4 domain II), and L1115C (S4 domain III) is comparable to the fast component of gating currents. In contrast, the fast component of fluorescence from the site S1436C (S4 domain IV) correlates with the slow component of gating. In all the cases, the slow component of fluorescence does not have any apparent correlation with charge movement. The fluorescence signals from sites reflecting the movement of S4s in the first three domains initiate simultaneously, whereas the fluorescence signals from the site S1436C exhibit a lag phase. These results suggest that the voltage-dependent movement of S4 domain IV is a later step in the activation sequence. Analysis of equilibrium and kinetic properties of fluorescence over activation voltage range indicate that S4 domain III is likely to move at most hyperpolarized potentials, whereas the S4s in domain I and domain II move at more depolarized potentials. The kinetics of fluorescence changes from sites near S4-DIV are slower than the activation time constants, suggesting that the voltage-dependent movement of S4-DIV may not be a prerequisite for channel opening. These experiments allow us to map structural features onto the kinetic landscape of a sodium channel during activation.     

The quantal gating charge of sodium channel inactivation

Using a very low noise voltage clamp technique it has been possible to record from the squid giant axon a slow component of gating current (I _g ) during the inactivation phase of the macroscopic sodium current (I _Na ) which was hitherto buried in the baseline noise. In order to examine whether this slowI _g contains gating charge that originates from transitions between the open (O) and the inactivated (I) states, which would indicate a true voltage dependence of inactivation, or whether other transitions contribute charge to slowI _g , a new model independent analysis termed isochronic plot analysis has been developed. From a direct correlation ofI _g and the time derivative of the sodium conductance dg _Na/d the condition when only O-I transitions occur is detected. Then the ratio of the two signals is constant and a straight line appears in an isochronic plot ofI _g vs. dg _Na/d . Its slope does not depend on voltage or time and corresponds to the quantal gating charge of the O-I transition (q _h ) divided by the single channel ionic conductance (). This condition was found at voltages above – 10 mV up to + 40 mV and a figure of 1.21e ^– was obtained forq _h at temperatures of 5 and 15°C. At lower voltages additional charge from other transitions, e.g. closed to open, is displaced during macroscopic inactivation. This means that conventional Eyring rate analysis of the inactivation time constant _h is only valid above – 10 mV and here the figure forq _h was confirmed also from this analysis. It is further shown that most of the present controversies surrounding the voltage dependence of inactivation can be clarified. The validity of the isochronic plot analysis has been confirmed using simulated gating and ionic currents.Abbreviations I _g gating current - I _Na sodium ionic current - g _Na macroscopic sodium conductance - single channel conductance - C, O, I closed, open, inactivated state occupancy of channels - g _h quantal charge displaced in a single O-I transition of Na channel - e ^– equivalent electron charge - h index referring to inactivation process - S _l limiting slope in isochronic plot see Eq.(3) - fractional distance, see Fig. 4 and (4, 5) - TMA tetramethylammonium - TTX tetrodotoxin - Tris tris(hydroxymethyl)aminomethane - HEPES N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid     

Neutralization of gating charges in domain II of the sodium channel alpha subunit enhances voltage-sensor trapping by a beta-scorpion toxin

beta-Scorpion toxins shift the voltage dependence of activation of sodium channels to more negative membrane potentials, but only after a strong depolarizing prepulse to fully activate the channels. Their receptor site includes the S3-S4 loop at the extracellular end of the S4 voltage sensor in domain II of the alpha subunit. Here, we probe the role of gating charges in the IIS4 segment in beta-scorpion toxin action by mutagenesis and functional analysis of the resulting mutant sodium channels. Neutralization of the positively charged amino acid residues in the IIS4 segment by mutation to glutamine shifts the voltage dependence of channel activation to more positive membrane potentials and reduces the steepness of voltage-dependent gating, which is consistent with the presumed role of these residues as gating charges. Surprisingly, neutralization of the gating charges at the outer end of the IIS4 segment by the mutations R850Q, R850C, R853Q, and R853C markedly enhances beta-scorpion toxin action, whereas mutations R856Q, K859Q, and K862Q have no effect. In contrast to wild-type, the beta-scorpion toxin Css IV causes a negative shift of the voltage dependence of activation of mutants R853Q and R853C without a depolarizing prepulse at holding potentials from -80 to -140 mV. Reaction of mutant R853C with 2-aminoethyl methanethiosulfonate causes a positive shift of the voltage dependence of activation and restores the requirement for a depolarizing prepulse for Css IV action. Enhancement of sodium channel activation by Css IV causes large tail currents upon repolarization, indicating slowed deactivation of the IIS4 voltage sensor by the bound toxin. Our results are consistent with a voltage-sensor-trapping model in which the beta-scorpion toxin traps the IIS4 voltage sensor in its activated position as it moves outward in response to depolarization and holds it there, slowing its inward movement on deactivation and enhancing subsequent channel activation. Evidently, neutralization of R850 and R853 removes kinetic barriers to binding of the IIS4 segment by Css IV, and thereby enhances toxin-induced channel activation.     

Extracellular sodium interacts with the HERG channel at an outer pore site

Most voltage-gated K(+) currents are relatively insensitive to extracellular Na(+) (Na(+)(o)), but Na(+)(o) potently inhibits outward human ether-a-go-go-related gene (HERG)-encoded K(+) channel current (Numaguchi, H., J.P. Johnson, Jr., C.I. Petersen, and J.R. Balser. 2000. Nat. Neurosci. 3:429-30). We studied wild-type (WT) and mutant HERG currents and used two strategic probes, intracellular Na(+) (Na(+)(i)) and extracellular Ba(2+) (Ba(2+)(o)), to define a site where Na(+)(o) interacts with HERG. Currents were recorded from transfected Chinese hamster ovary (CHO-K1) cells using the whole-cell voltage clamp technique. Inhibition of WT HERG by Na(+)(o) was not strongly dependent on the voltage during activating pulses. Three point mutants in the P-loop region (S624A, S624T, S631A) with intact K(+) selectivity and impaired inactivation each had reduced sensitivity to inhibition by Na(+)(o). Quantitatively similar effects of Na(+)(i) to inhibit HERG current were seen in the WT and S624A channels. As S624A has impaired Na(+)(o) sensitivity, this result suggested that Na(+)(o) and Na(+)(i) act at different sites. Extracellular Ba(2+) (Ba(2+)(o)) blocks K(+) channel pores, and thereby serves as a useful probe of K(+) channel structure. HERG channel inactivation promotes relief of Ba(2+) block (Weerapura, M., S. Nattel, M. Courtemanche, D. Doern, N. Ethier, and T. Hebert. 2000. J. Physiol. 526:265-278). We used this feature of HERG inactivation to distinguish between simple allosteric and pore-occluding models of Na(+)(o) action. A remote allosteric model predicts that Na(+)(o) will speed relief of Ba(2+)(o) block by promoting inactivation. Instead, Na(+)(o) slowed Ba(2+) egress and Ba(2+) relieved Na(+)(o) inhibition, consistent with Na(+)(o) binding to an outer pore site. The apparent affinities of the outer pore for Na(+)(o) and K(+)(o) as measured by slowing of Ba(2+) egress were compatible with competition between the two ions for the channel pore in their physiological concentration ranges. We also examined the role of the HERG closed state in Na(+)(o) inhibition. Na(+)(o) inhibition was inversely related to pulsing frequency in the WT channel, but not in the pore mutant S624A.