Tight steric closure at the intracellular activation gate of a voltage-gated K(+) channel

In voltage-gated K(+) channels (Kv), an intracellular gate regulates access from the cytoplasm to the pore by organic channel blockers and by chemical modifiers. But is ion flow itself controlled instead by constriction of the narrow selectivity filter near the extracellular surface? We find that the intracellular gate of Kv channels is capable of regulating access even by the small cations Cd(2+) and Ag(+). It can also exclude small neutral or negatively charged molecules, indicating that the gate operates by steric exclusion rather than electrostatically. Just intracellular to the gated region, channel closure does not restrict access even to very large reagents. Either these Kv channels have a broader inner entrance than seen in the KcsA crystal, even in the closed state, or the region is highly flexible (but nevertheless remains very securely closed nearby).

Blocker state dependence and trapping in hyperpolarization-activated cation channels: evidence for an intracellular activation gate

Hyperpolarization-activated cation currents (I(h)) are key determinants of repetitive electrical activity in heart and nerve cells. The bradycardic agent ZD7288 is a selective blocker of these currents. We studied the mechanism for ZD7288 blockade of cloned I(h) channels in excised inside-out patches. ZD7288 blockade of the mammalian mHCN1 channel appeared to require opening of the channel, but strong hyperpolarization disfavored blockade. The steepness of this voltage-dependent effect (an apparent valence of approximately 4) makes it unlikely to arise solely from a direct effect of voltage on blocker binding. Instead, it probably indicates a differential affinity of the blocker for different channel conformations. Similar properties were seen for ZD7288 blockade of the sea urchin homologue of I(h) channels (SPIH), but some of the blockade was irreversible. To explore the molecular basis for the difference in reversibility, we constructed chimeric channels from mHCN1 and SPIH and localized the structural determinant for the reversibility to three residues in the S6 region likely to line the pore. Using a triple point mutant in S6, we also revealed the trapping of ZD7288 by the closing of the channel. Overall, the observations led us to hypothesize that the residues responsible for ZD7288 block of I(h) channels are located in the pore lining, and are guarded by an intracellular activation gate of the channel.

Blocker protection in the pore of a voltage-gated K+ channel and its structural implications

The structure of the bacterial potassium channel KcsA has provided a framework for understanding the related voltage-gated potassium channels (Kv channels) that are used for signalling in neurons. Opening and closing of these Kv channels (gating) occurs at the intracellular entrance to the pore, and this is also the site at which many open channel blockers affect Kv channels. To learn more about the sites of blocker binding and about the structure of the open Kv channel, we investigated here the ability of blockers to protect against chemical modification of cysteines introduced at sites in transmembrane segment S6, which contributes to the intracellular entrance. Within the intracellular half of S6 we found an abrupt cessation of protection for both large and small blockers that is inconsistent with the narrow 'inner pore' seen in the KcsA structure. These and other results are most readily explained by supposing that the structure of Kv channels differs from that of the non-voltage-gated bacterial channel by the introduction of a sharp bend in the inner (S6) helices. This bend would occur at a Pro-X-Pro sequence that is highly conserved in Kv channels, near the site of activation gating.

The bacterial K+ channel structure and its implications for neuronal channels

Voltage-gated K+ (Kv) channels play a central role in generating action potentials and rhythmic patterns, as well as in dendritic signal processing in neurons. Recently, the first structure of a member of the K+ channel family was solved. Although this channel is from bacteria and has a streamlined body plan with no voltage gating, it establishes the architecture of the functional core of the voltage-gated (K+) channels and their relatives. This architecture explains the crucial features of ion permeation and blockade, and gives some strong hints about gating. The bacterial K+ channel structure is the central piece in a puzzle; it remains to be seen how it will fit together with other domains of the Kv channels, with auxiliary subunits, and with other signal transduction molecules.

The activation gate of a voltage-gated K+ channel can be trapped in the open state by an intersubunit metal bridge

Voltage-activated K+ channels are integral membrane proteins containing a potassium-selective transmembrane pore gated by changes in the membrane potential. This activation gating (opening) occurs in milliseconds and involves a gate at the cytoplasmic side of the pore. We found that substituting cysteine at a particular position in the last transmembrane region (S6) of the homotetrameric Shaker K+ channel creates metal binding sites at which Cd2+ ions can bind with high affinity. The bound Cd2+ ions form a bridge between the introduced cysteine in one channel subunit and a native histidine in another subunit, and the bridge traps the gate in the open state. These results suggest that gating involves a rearrangement of the intersubunit contacts at the intracellular end of S6. The recently solved structure of a bacterial K+ channel shows that the S6 homologs cross in a bundle, leaving an aperture at the bundle crossing. In the context of this structure, the metal ions form a bridge between a cysteine above the bundle crossing and a histidine below the bundle crossing in a neighboring subunit. Our results suggest that gating occurs at the bundle crossing, possibly through a change in the conformation of the bundle itself.

The moving parts of voltage-gated ion channels

Ion channels, like many other proteins, have moving parts that perform useful functions. The channel proteins contain an aqueous, ion-selective pore that crosses the plasma membrane, and they use a number of distinct ‘gating’ mechanisms to open and close this pore in response to biological stimuli such as the binding of a ligand or a change in the transmembrane voltage.

This review is written at a watershed in our understanding of ion channels.

1. INTRODUCTION 240

1.1 Basic structure of voltage-activated channels 241

1.2 What are the physical motions of the channel protein during gating? 243

1.3 Gating involves several distinct mechanisms of activation and inactivation 246

2. ACTIVATION GATING 246

2.1 Early evidence for an activation gate at the intracellular mouth 247

2.1.1 Open channel blockade 247

2.1.2 The ‘ foot-in-the-door’ effect 249

2.1.3 Trapping of blockers behind closed activation gates 249

2.2 Site-directed mutagenesis and the difficulty of inferring structural roles from functional effects 250

2.3 State-dependent cysteine modification as a reporter of position and motion 251

2.4 Localization of activation gating 254

2.4.1 The trapping cavity 254

2.4.2 The activation gate 255

2.4.3 Is there more than one site of activation gating? 258

3. INACTIVATION GATING 259

3.1 Ball-and-chain (N-type) inactivation 261

3.1.1 Nature of the ‘ball’ – a tethered blocking particle 262

3.1.2 The ball receptor 263

3.1.3 The chain 264

3.1.4 Variations on the N-type inactivation theme: multiple balls, foreign balls, anti-balls 265

3.2 C-type inactivation 266

3.2.1 C-type inactivation and the outer mouth of the K+channel 266

3.2.2 The selectivity filter participates in C-type inactivation 267

3.2.3 A consistent structural picture of C-type inactivation 269

3.3 By what mechanism do other voltage-gated channels inactivate? 272

4. THE VOLTAGE SENSOR 273

4.1 Quantitative principles of voltage-dependent gating 276

4.2 S4 (and its neighbours) as the principal voltage sensor 277

4.2.1 Mutational effects on voltage-dependence and charge movement 277

4.2.2 Evidence for the translocation of S4 279

4.2.3 Real-time monitoring of S4motion by fluorescence 282

4.3 Coupling between the voltage sensor and gating 283

5. CONCLUSION 284

6. ACKNOWLEDGEMENTS 287

7. REFERENCES 287

Gated access to the pore of a voltage-dependent K+ channel

Voltage-activated K+ channels are integral membrane proteins that open or close a K(+)-selective pore in response to changes in transmembrane voltage. Although the S4 region of these channels has been implicated as the voltage sensor, little is known about how opening and closing of the pore is accomplished. We explored the gating process by introducing cysteines at various positions thought to lie in or near the pore of the Shaker K+ channel, and by testing their ability to be chemically modified. We found a series of positions in the S6 transmembrane region that react rapidly with water-soluble thiol reagents in the open state but not the closed state. An open-channel blocker can protect several of these cysteines, showing that they lie in the ion-conducting pore. At two of these sites, Cd2+ ions bind to the cysteines without affecting the energetics of gating; at a third site, Cd2+ binding holds the channel open. The results suggest that these channels open and close by the movement of an intracellular gate, distinct from the selectivity filter, that regulates access to the pore.

Trapping of organic blockers by closing of voltage-dependent K+ channels: evidence for a trap door mechanism of activation gating

Small organic molecules, like quaternary ammonium compounds, have long been used to probe both the permeation and gating of voltage-dependent K+ channels. For most K+ channels, intracellularly applied quaternary ammonium (QA) compounds such as tetraethylammonium (TEA) and decyltriethylammonium (C10) behave primarily as open channel blockers: they can enter the channel only when it is open, and they must dissociate before the channel can close. In some cases, it is possible to force the channel to close with a QA blocker still bound, with the result that the blocker is "trapped." Armstrong (J. Gen. Physiol. 58:413-437) found that at very negative voltages, squid axon K+ channels exhibited a slow phase of recovery from QA blockade consistent with such trapping. In our studies on the cloned Shaker channel, we find that wild-type channels can trap neither TEA nor C10, but channels with a point mutation in S6 can trap either compound very efficiently. The trapping occurs with very little change in the energetics of channel gating, suggesting that in these channels the gate may function as a trap door or hinged lid that occludes access from the intracellular solution to the blocker site and to the narrow ion-selective pore.

Defective "pacemaker" current (Ih) in a zebrafish mutant with a slow heart rate

At a cellular level, cardiac pacemaking, which sets the rate and rhythm of the heartbeat, is produced by the slow membrane depolarization that occurs between action potentials. Several ionic currents could account for this pacemaker potential, but their relative prominence is controversial, and it is not known which ones actually play a pacemaking role in vivo. To correlate currents in individual heart cells with the rhythmic properties of the intact heart, we have examined slow mo (smo), a recessive mutation we discovered in the zebrafish Danio rerio. This mutation causes a reduced heart rate in the embryo, a property we can quantitate because the embryo is transparent. We developed methods for culture of cardiocytes from zebrafish embryos and found that, even in culture, cells from smo continue to beat relatively slowly. By patch-clamp analysis, we discovered that a large repertoire of cardiac currents noted in other species are present in these cultured cells, including sodium, T-type, and L-type calcium and several potassium currents, all of which appear normal in the mutant. The only abnormality appears to be in a hyperpolarization-activated inward current with the properties of Ih, a current described previously in the nervous system, pacemaker, and other cardiac tissue. smo cardiomyocytes have a reduction in Ih that appears to result from severe diminution of one kinetic component of the Ih current. This provides strong evidence that Ih is an important contributor to the pacemaking behavior of the intact heart.

Two functionally distinct subsites for the binding of internal blockers to the pore of voltage-activated K+ channels

Many blockers of Na+ and K+ channels act by blocking the pore from the intracellular side. For Shaker K+ channels, such intracellular blockers vary in their functional effect on slow (C-type) inactivation: Some blockers interfere with C-type inactivation, whereas others do not. These functional differences can be explained by supposing that there are two overlapping "subsites" for blocker binding, only one of which inhibits C-type inactivation through an allosteric effect. We find that the ability to bind to these subsites depends on specific structural characteristics of the blockers, and correlates with the effect of mutations in two distinct regions of the channel protein. These interactions are important because they affect the ability of blockers to produce use-dependent inhibition.

N-type inactivation and the S4-S5 region of the Shaker K+ channel

The intracellular segment of the Shaker K+ channel between transmembrane domains S4 and S5 has been proposed to form at least part of the receptor for the tethered N-type inactivation "ball." We used the approach of cysteine substitution mutagenesis and chemical modification to test the importance of this region in N-type inactivation. We studied N-type inactivation or the block by a soluble inactivation peptide ("ball peptide") before and after chemical modification by methanethiosulfonate reagents. Particularly at position 391, chemical modification altered specifically the kinetics of ball peptide binding without altering other biophysical properties of the channel. Results with reagents that attach different charged groups at 391 C suggested that there are both electrostatic and steric interactions between this site and the ball peptide. These findings identify this site to be in or near the receptor site for the inactivation ball. At many of the other positions studied, modification noticeably inhibited channel current. The accessible cysteines varied in the state-dependence of their modification, with five- to tenfold changes in reactions rate depending on the gating state of the channel.

Dynamic rearrangement of the outer mouth of a K+ channel during gating

With prolonged stimulation, voltage-activated K+ channels close by a gating process called inactivation. This inactivation gating can occur by two distinct molecular mechanisms: N-type, in which a tethered particle blocks the intracellular mouth of the pore, and C-type, which involves a closure of the external mouth. The functional motion involved in C-type inactivation was studied by introducing cysteine residues at the outer mouth of Shaker K+ channels through mutagenesis, and by measuring state-dependent changes in accessibility to chemical modification. Modification of three adjacent residues in the outer mouth was 130-10,000-fold faster in the C-type inactivated state than in the closed state. At one position, state-dependent bridging or crosslinking between subunits was also possible. These results give a consistent picture in which C-type inactivation promotes a local rearrangement and constriction of the channel at the outer mouth.

The inward rectification mechanism of the HERG cardiac potassium channel

A human genetic defect associated with 'long Q-T syndrome', an abnormality of cardiac rhythm involving the repolarization of the action potential, was recently found to lie in the HERG gene, which codes for a potassium channel. The HERG K+ channel is unusual in that it seems to have the architectural plan of the depolarization-activated K+ channel family (six putative transmembrane segments), yet it exhibits rectification like that of the inward-rectifying K+ channels, a family with different molecular structure (two transmembrane segments). We have studied HERG channels expressed in mammalian cells and find that this inward rectification arises from a rapid and voltage-dependent inactivation process that reduces conductance at positive voltages. The inactivation gating mechanism resembles that of C-type inactivation, often considered to be the 'slow inactivation' mechanism of other K+ channels. The characteristics of this gating suggest a specific role for this channel in the normal suppression of arrhythmias.

Use-dependent blockers and exit rate of the last ion from the multi-ion pore of a K+ channel

Quaternary ammonium blockers inhibit many voltage-activated potassium (K+) channels from the intracellular side. When applied to Drosophila Shaker potassium channels expressed in mammalian cells, these rapidly reversible blockers produced use-dependent inhibition through an unusual mechanism--they promoted an intrinsic conformational change known as C-type inactivation, from which recovery is slow. The blockers did so by cutting off potassium ion flow to a site in the pore, which then emptied at a rate of 10(5) ions per second. This slow rate probably reflected the departure of the last ion from the multi-ion pore: Permeation of ions (at 10(7) per second) occurs rapidly because of ion-ion repulsion, but the last ion to leave would experience no such repulsion.

On the use of thiol-modifying agents to determine channel topology

A powerful tool in the study of cloned ion channels is the combined use of site-directed mutagenesis and chemical modification. Site-directed mutagenesis is used to introduce new cysteine residues at specific positions in a channel protein, and chemical modification by thiol-specific reagents is then used to assess the exposure of the introduced cysteins. This method has been used to assess secondary structure, membrane topology and conformational changes. We report that one commonly used, charged reagent (MTSEA; aminoethyl methanethiosulfonate) can cross the membrane quite readily. We also find that other reagents that are quite membrane-impermeant can cross the membrane when patches are electrically leaky. Both of these undesired effects can be controlled by the use of a thiol scavenger. These findings argue for caution in the use of modifying reagents to determine the membrane topology of channels and other membrane proteins.

Modulation of K+ current by frequency and external [K+]: a tale of two inactivation mechanisms

Voltage-activated K+ currents and their inactivation properties are important for controlling frequency-dependent signaling in neurons and other excitable cells. Two distinct molecular mechanisms for K+ channel inactivation have been described: N-type, which involves rapid occlusion of the open channel by an intracellular tethered blocker, and C-type, which involves a slower change at the extracellular mouth of the pore. We find that frequency-dependent cumulative inactivation of Shaker channels is very sensitive to changes of extracellular [K+] in the physiological range, with much more inactivation at low [K+]out, and that it results from the interaction of N- and C-type inactivation. N-type inactivation enhances C-type inactivation by two mechanisms. First, it inhibits outward K+ flux, which normally fills an external ion site and thus prevents C-type inactivation. Second, it keeps the channel's activation gate open even after repolarization, allowing C-type inactivation to occur for a prolonged period.

A discrete site for general anesthetics on a postsynaptic receptor

General anesthetics depress central nervous system excitability via a mechanism that probably involves effects on synaptic ion channels, but the fundamental molecular nature of the site where they act is unknown. Although the importance of hydrophobicity for general anesthetic drug potency has long been established, it remains uncertain whether these "nonspecific" drugs act on membrane proteins directly or by modification of the physical properties of the lipid membrane or the lipid-protein interface. We find that specific mutations in the acetylcholine receptor pore-forming M2 domains enhance the sensitivity of the receptor to the general anesthetics isoflurane, hexanol, and octanol, suggesting that these agents act by binding directly to a discrete protein site at or near these residues. The sensitivity of the receptor to block by general anesthetics increases with increased hydrophobicity of these residues, demonstrating that hydrophobic forces dominate the interaction of drugs with their protein site. Furthermore, octanol inhibits both wild-type and mutant nicotinic acetylcholine receptors preferentially after channel opening, which is consistent with a mechanism where drugs bind within the receptor's pore. Similar sites on postsynaptic ion channels in brain may represent general anesthetic targets for modulating consciousness.

Covalent modification of engineered cysteines in the nicotinic acetylcholine receptor agonist-binding domain inhibits receptor activation

We constructed and characterized a series of nicotinic receptor mutants with a cysteine substituted for one of the amino acid residues in the alpha-subunit between positions 183 and 198. This region of the receptor is known to participate in agonist binding and channel activation. The goal of this 'cysteine scanning mutagenesis' is to introduce the reactivity of a free thiol group into functionally important protein domains; modification of the introduced cysteines can then be used to probe the structure and function of the targeted region. Mutants were examined by coexpression with the beta-, gamma- and delta-subunits in Xenopus oocytes using two-microelectrode voltage clamp recording. Twelve of fourteen mutants expressed receptors with properties comparable with the wild-type, including sensitivity to reduction by dithiothreitol (DTT). This indicates that introduction of an additional cysteine within this region of the receptor did not interfere with formation of the native disulphide between alpha Cys-192 and alpha Cys-193. Only one mutation, alpha Y198C, caused dramatic changes in the EC50 for acetylcholine (ACh) and in the sensitivity to DTT. We then examined the effects of the thiol modification and found two mutants, alpha H186C and alpha V188C, that showed significant decreases in responsiveness to ACh after exposure to methylmethanethiosulphonate (MMTS). Dose-response measurements show that exposure of alpha H186C mutants to MMTS causes a shift in apparent agonist affinity without changing the peak response, and this is not reversible by DTT. In contrast, the MMTS-treated alpha V188C mutants show changes in both apparent affinity and peak response which are readily reversed by DTT. Together, our data show that these two nearby residues occupy markedly different environments relative to the contact points for ACh. They also demonstrate that cysteine-substitution mutagenesis can be successfully applied to protein domains that include functionally important disulphides.

Visual identification of individual transfected cells for electrophysiology using antibody-coated beads

Electrophysiological study of transiently transfected cells requires the identification of individual cells that express the protein of interest. We describe a simple, quick and inexpensive method for visually identifying cells that have been co-transfected with an expression plasmid for a lymphocyte surface antigen (CD8-alpha). Transfected cells are incubated briefly with polystyrene microspheres (4.5 microns diameter) that have been precoated with antibody to CD8. Cells expressing CD8 on their surface are decorated with many beads and are thus readily distinguishable from untransfected cells. Beads already coated with antibody are available commercially. The method takes less than five minutes and requires no reagent preparation or special equipment for visualization of the beads.

An engineered cysteine in the external mouth of a K+ channel allows inactivation to be modulated by metal binding

Substitution of a cysteine in the extracellular mouth of the pore of the Shaker-delta K+ channel permits allosteric inhibition of the channel by Zn2+ or Cd2+ ions at micromolar concentrations. Cd2+ binds weakly to the open state but drives the channel into the slow (C-type) inactivated state, which has a Kd for Cd2+ of approximately 0.2 microM. There is a 45,000-fold increase in affinity when the channel changes from open to inactivated. These results indicate that C-type inactivation involves a structural change in the external mouth of the pore. This structural change is reflected in the T449C mutant as state-dependent metal affinity, which may result either from a change in proximity of the introduced cysteine residues of the four subunits or from a change of the exposure of this residue on the surface of the protein.

Cysteines in the Shaker K+ channel are not essential for channel activity or zinc modulation

We investigated whether the cysteine residues in Shaker potassium (K+) channels are essential for activation and inactivation gating or for modulation of activation gating by external zinc (Zn2+). Mutants of the Shaker K+ channel were prepared in which all seven cysteine residues were replaced (C-less). These changes were made in both wild-type Shaker H4 channels and in a deletion mutant (delta 6-46) lacking N-type ("fast") inactivation. Replacement of all cysteines left most functional properties of the K+ currents unaltered. The most noticeable difference between the C-less and wild-type currents was the faster C-type inactivation of the C-less channel which could be attributed largely to the mutation of Cys462. This is consistent with the effects of previously reported mutations of nearby residues in the S6 region. There were also small changes in the activation gating of C-less currents. Modulation by external Zn2+ of the voltage dependence and rate of activation gating is preserved in the C-less channels, indicating that none of the cysteines in the Shaker K+ channel plays an important role in Zn2+ modulation.

Conductance mutations of the nicotinic acetylcholine receptor do not act by a simple electrostatic mechanism

Fixed negative charges in many cation channels raise the single-channel conductance, apparently by an electrostatic mechanism: their effects are accentuated in solutions of low ionic strength and attenuated at high ionic strength. The charges of specific amino acids near the ends of the proposed pore-lining M2 segment of the nicotinic acetylcholine receptor, termed the extracellular and cytoplasmic rings, have recently been shown to influence the single-channel K+ conductance (Imoto, K., C. Busch, B. Sakmann, M. Mishina, T. Konno, J. Nakai, H. Bujo, Y. Mori, K. Fukuda and S. Numa. 1988. Nature 335:645-648). We examined whether these charges might act by a direct electrostatic effect on the energy of ions in the pore, rather than indirectly by inducing a structural change. To this end, we measured the conductances of charge mutants over a range of K+ concentrations (ionic strengths). As expected, we found that negative charge mutations raise the conductance, and positive charge mutations lower it. The effects of cytoplasmic-ring mutations are accentuated at low ionic strength, but they are not completely attenuated at high ionic strength. The effects of extracellular-ring mutations are independent of ionic strength. These results are inconsistent with the simplest electrostatic model. We suggest a modified model that qualitatively accounts for the data.

The internal quaternary ammonium receptor site of Shaker potassium channels

Quaternary ammonium (QA) compounds inhibit K+ conductance by entering and occluding the open pore of voltage-activated K+ channels. We characterized the effects of a series of alkyl-triethylammonium blockers on the Shaker K+ channel and tested them on a series of site-directed mutants of the channel protein in order to define the structural features of the binding sites. We found that mutations in two regions of the channel protein, the pore (P) region and the last transmembrane sequence (S6), appear to alter QA binding, not through their effects on gating but perhaps through direct effects on the binding site. Several mutations in the P region affect tetraethylammonium binding but have minimal effects on longer blockers, suggesting that the hydrophobic tail contributes to binding in a nonadditive fashion. Binding of the longer blockers can be affected by varying the hydrophobicity of 1 residue within S6 by site-specific substitution, in a manner consistent with a direct hydrophobic interaction between the side chain at this site and the alkyl chains of the blocker.

Ion effects on gating of the Ca(2+)-activated K+ channel correlate with occupancy of the pore

We studied the effects of permeant ions on the gating of the large conductance Ca(2+)-activated K+ channel from rat skeletal muscle. Rb+ blockade of inward K+ current caused an increase in the open probability as though Rb+ occupancy of the pore interferes with channel closing. In support of this hypothesis, we directly measured the occupancy of the pore by the impermeant ion Cs+ and found that it strongly correlates with its effect on gating. This is consistent with the "foot-in-the-door" model of gating, which states that channels cannot close with an ion in the pore. However, because Rb+ and Cs+ not only slow the closing rate (as predicted by the model), but also speed the opening rate, our results are more consistent with a modified version of the model in which the channel can indeed close while occupied, but the occupancy destabilizes the closed state. Increasing the occupancy of the pore by the addition of other permeant (K+ and Tl+) and impermeant (tetraethylammonium) ions did not affect the open probability. To account for this disparity, we used a two-site permeation model in which only one of the sites influenced gating. Occupancy of this "gating site" interferes with channel closing and hastens opening. Ions that directly or indirectly increase the occupancy of this site will increase the open probability.

A novel K+ channel with unique localizations in mammalian brain: molecular cloning and characterization

Using a cDNA library prepared from circumvallate papillae of rat tongue, we have identified, cloned, and sequenced a novel K+ channel, designated cdrk. The cdrk channel appears to be a member of the Shab subfamily, most closely resembling drk1. Electrophysiologic analysis of expressed cdrk channels reveals delayed rectifier properties similar to those of drk1 channels. Localizations of cdrk mRNA in rat brain and peripheral tissues, assessed by in situ hybridization and Northern blot analysis, differ from any other reported K+ channels. In the brain cdrk mRNA is most concentrated in granule cells of the olfactory bulb and cerebellum. In peripheral tissues, mRNAs for cdrk and drk1 are reciprocally localized, indicating that the K+ channel properties contributed by mammalian Shab homologs may be important in a variety of excitable tissues.

The inactivation gate of the Shaker K+ channel behaves like an open-channel blocker

Following voltage-dependent activation, Drosophila Shaker K+ channels enter a nonconducting, inactivated state. This process has been proposed to occur by a "ball-and-chain" mechanism, in which the N-terminus of the protein behaves like a blocker tethered to the cytoplasmic side of the channel and directly occludes the pore to cause inactivation. To complement the ample evidence for the involvement of the N-terminus, we sought evidence that it blocks the pore directly. We found that inactivation exhibits several distinctive properties of pore blockade. First, recovery was speeded by increased external K+ concentrations, just as blockade can be relieved by trans-permeant ions. Second, single-channel experiments show that the channel reopens from the inactivated state upon repolarization. These openings were usually required for recovery, as though the blocking particle must exit the pore before the channel can close.

Mutations affecting agonist sensitivity of the nicotinic acetylcholine receptor

The nicotinic acetylcholine receptor (AChR) is a pentameric transmembrane protein (alpha 2 beta gamma delta) that binds the neurotransmitter acetylcholine (ACh) and transduces this binding into the opening of a cation selective channel. The agonist, competitive antagonist, and snake toxin binding functions of the AChR are associated with the alpha subunit (Kao et al., 1984; Tzartos and Changeux, 1984; Wilson et al., 1985; Kao and Karlin, 1986; Pederson et al., 1986). We used site-directed mutagenesis and expression of AChR in Xenopus oocytes to identify amino acid residues critical for ligand binding and channel activation. Several mutations in the alpha subunit sequence were constructed based on information from sequence homology and from previous biochemical (Barkas et al., 1987; Dennis et al., 1988; Middleton and Cohen, 1990) and spectroscopic (Pearce and Hawrot, 1990; Pearce et al., 1990) studies. We have identified one mutation, Tyr190 to Phe (Y190F), that had a dramatic effect on ligand binding and channel activation. These mutant channels required more than 50-fold higher concentrations of ACh for channel activation than did wild type channels. This functional change is largely accounted for by a comparable shift in the agonist binding affinity, as assessed by the ability of ACh to compete with alpha-bungarotoxin binding. Other mutations at nearby conserved positions of the alpha subunit (H186F, P194S, Y198F) produce less dramatic changes in channel properties. Our results demonstrate that ligand binding and channel gating are separable properties of the receptor protein, and that Tyr190 appears to play a specific role in the receptor site for acetylcholine.

Tetraethylammonium blockade distinguishes two inactivation mechanisms in voltage-activated K+ channels

Voltage-activated K+ channels are a family of closely related membrane proteins that differ in their gating behavior, conductance, and pharmacology. A prominent and physiologically important difference among K+ channels is their rate of inactivation. Inactivation rates range from milliseconds to seconds, and K+ channels with different inactivation properties have very different effects on signal integration and repetitive firing properties of neurons. The cloned Shaker B (H4) potassium channel is an example of a K+ channel that inactivates in a few milliseconds. Recent experiments have shown that removal of an N-terminal region of the Shaker protein by site-directed deletion practically abolishes this fast inactivation, but the modified channel does still inactivate during a prolonged depolarization lasting many seconds. Here we report that this remnant inactivation must occur by a distinct mechanism from the rapid inactivation of the wild-type Shaker channel. Like the inactivation of another K+ channel [Grissmer, S. & Calahan, M. (1989) Biophys. J. 55, 203-206], this slow inactivation is retarded by the application of a channel blocker, tetraethylammonium, to the extracellular side of the channel. By contrast, the fast inactivation of the wild-type Shaker channel is sensitive only to intracellular application of tetraethylammonium. Intracellular tetraethylammonium slows down the fast inactivation process, as though it competes with the binding of the inactivation particle.

Mutations affecting internal TEA blockade identify the probable pore-forming region of a K+ channel

The active site of voltage-activated potassium channels is a transmembrane aqueous pore that permits ions to permeate the cell membrane in a rapid yet highly selective manner. A useful probe for the pore of potassium-selective channels is the organic ion tetraethylammonium (TEA), which binds with millimolar affinity to the intracellular opening of the pore and blocks potassium current. In the potassium channel encoded by the Drosophila Shaker gene, an amino acid residue that specifically affects the affinity for intracellular TEA has now been identified by site-directed mutagenesis. This residue is in the middle of a conserved stretch of 18 amino acids that separates two locations that are both near the external opening of the pore. These findings suggest that this conserved region is intimately involved in the formation of the ion conduction pore of voltage-activated potassium channels. Further, a stretch of only eight amino acid residues must traverse 80 percent of the transmembrane electric potential difference.

Mutations affecting TEA blockade and ion permeation in voltage-activated K+ channels

Voltage-dependent ion channels are responsible for electrical signaling in neurons and other cells. The main classes of voltage-dependent channels (sodium-, calcium-, and potassium-selective channels) have closely related molecular structures. For one member of this superfamily, the transiently voltage-activated Shaker H4 potassium channel, specific amino acid residues have now been identified that affect channel blockade by the small ion tetraethylammonium, as well as the conduction of ions through the pore. Furthermore, variation at one of these amino acid positions among naturally occurring potassium channels may account for most of their differences in sensitivity to tetraethylammonium.

Human cardiac sodium channels expressed in Xenopus oocytes

We report the expression of voltage-dependent Na+ channels in Xenopus oocytes injected with total RNA isolated from explanted human hearts. The expressed channels demonstrate characteristic voltage-dependent gating, inhibition by tetrodotoxin, and selectivity for Na+. Oocytes injected with sterile water or intentionally degraded RNA had no similar channel activity. The antiarrhythmic agent lidocaine (20 microM) inhibits current flow through the channel in a voltage-dependent fashion. Na+ channels expressed by injection of human cardiac RNA into Xenopus oocytes qualitatively resemble channels in the native tissue.

Expression of Torpedo nicotinic acetylcholine receptor subunits in yeast is enhanced by use of yeast signal sequences

We have produced the four subunits of the nicotinic acetylcholine receptor of Torpedo californica, an integral membrane protein, in the yeast Saccharomyces cerevisiae. Two of the subunits (alpha and delta) were readily produced from their cDNAs after simply subcloning them into a yeast shuttle vector adjacent to a yeast promoter. The other two protein subunits (beta and gamma) were not produced by this strategy, although the amounts of mRNA produced from these expression constructs are similar to those for alpha and delta. Replacing the DNA coding for the normal N-terminal signal sequences for the beta and gamma subunits with DNA coding for the signal sequence of yeast invertase results in successful protein synthesis. The yeast signal sequence allows these subunits to be translocated across the membrane of the endoplasmic reticulum and to be glycosylated. The appropriate final size of the subunit proteins suggests that the yeast signal sequence has been properly cleaved after translocation.

Sodium channels from human brain RNA expressed in Xenopus oocytes. Basic electrophysiologic characteristics and their modification by diphenylhydantoin

We describe the expression and characterization of sodium channels from human brain RNA in the Xenopus oocyte. The expressed channel, studied by whole-cell voltage clamp, reveals characteristic selectivity for sodium as the permeant ion, voltage-dependent gating, and block by nanomolar concentrations of tetrodotoxin. Such channels are not seen in control oocytes injected with solvent only. The anticonvulsant diphenylhydantoin (DPH) inhibits the expressed channel in a voltage- and use-dependent manner, much like the effect seen in primary mammalian neuronal preparations. The inhibition of the expressed human sodium channel by DPH can be described by models previously developed to explain block of Na channels by local anesthetics. The preferential block of Na channels during depolarization helps explain the selectivity of DPH for neurons involved in seizure activity.

Permeation in potassium channels: implications for channel structure

The SR K+ channel is a single-ion channel with a tunnel that is not very selective, while the DR and CaK channels are both more selective, multi-ion channels. The permeation mechanisms of the three channels are probably most systematically distinguished by the length of their tunnels; the SR has the shortest and the DR the longest. Although different in their mechanisms of activation, the DR and CaK channels have very similar permeation characteristics, down to the details of selectivity and blockade. The longer tunnel and reduced conductance (perhaps a result of the extra tunnel length) of the DR K+ channel are the main differences. The selectivity of the rate-limiting barriers and the binding sites within the channels, however, are strikingly similar. A successful potassium channel must satisfy two criteria: It must let potassium ions through and not much else, and it must let many potassium ions through. To be selective the channel must have a narrow selectivity filter, so that an ion must shed some of its waters of hydration to pass through. Sodium ions are excluded because they are more reluctant to lose their water, and they are not adequately compensated for this loss by interaction with the selectivity filter. To carry a large current the narrow region must be short, with wide antechambers to reduce the diffusional access resistance (48). Energetically, the channel must strike a balance. There must be enough binding energy to compensate the ions for their lost hydration energy, so that the energy barrier to permeation is small. If the channel binds the ion too tightly, however, the ion will not be able to exit, and the current will be small. Some of the shared properties of different potassium channels are probably consequences of these requirements; others may be incidental to function, suggesting a common origin. Barium ions have almost exactly the same radius as potassium ions but twice the charge, so it is perhaps not surprising that barium can block any potassium channel by binding where potassium does, but too tightly. It seems more surprising that blockade by TEA+ and other quaternary ammonium ions is also well conserved. All three of the potassium channels considered here have a mouth that binds QA ions and that has a nearby hydrophobic pocket; the frog DR and the CaK channels also have a TEA+-specific site on the opposite side. The QA site might not be an obligatory feature of potassium channels, but rather a conserved evolutionary vestige.(ABSTRACT TRUNCATED AT 400 WORDS)

Relief of Na+ block of Ca2+-activated K+ channels by external cations

The flickery block of single Ca2+-activated K+ channels that is produced by internally applied Na+ can be relieved by millimolar concentrations of external K+. This effect of K+ on the kinetics of Na+ block was studied by the method of amplitude distribution analysis described in the companion paper (Yellen, G., 1984b, J. Gen. Physiol., 84:157-186). It appears that K+ relieves block by increasing the exit rate of the blocking ion from the channel, not by competitively slowing its entrance rate. This suggests that a K ion that enters the channel from the outside can expel the blocking Na ion, which entered the channel from the inside. Cs+, which cannot carry current through the channel, and Rb+, which carries a reduced current through the channel, are just as effective as K+ in relieving the block by internal Na+. The kinetics of block by internal nonyltriethylammonium (C9) are unaffected by the presence of these ions in the external bathing solution.

Ionic permeation and blockade in Ca2+-activated K+ channels of bovine chromaffin cells

Single channel currents through Ca2+-activated K+ channels of bovine chromaffin cells were measured to determine the effects of small ions on permeation through the channel. The channel selects strongly for K+ over Na+ and Cs+, and Rb+ carries a smaller current through the channel than K+. Tetraethylammonium ion (TEA+) blocks channel currents when applied to either side of the membrane; it is effective at lower concentrations when applied externally. Millimolar concentrations of internal Na+ reduce the average current through the channel and produce large fluctuations (flicker) in the open channel currents. This flickery block is analyzed by a new method, amplitude distribution analysis, which can measure block and unblock rates in the microsecond time range even though individual blocking events are not time-resolved by the recording system. The analysis shows that the rate of block by Na+ is very voltage dependent, but the unblock rate is voltage independent. These results can be explained easily by supposing that current flow through the channel is diffusion limited, a hypothesis consistent with the large magnitude of the single channel current.