U-type inactivation of Kv3.1 and Shaker potassium channels.

We previously concluded that the Kv2.1 K(+) channel inactivates preferentially from partially activated closed states. We report here that the Kv3.1 channel also exhibits two key features of this inactivation mechanism: a U-shaped voltage dependence measured at 10 s and stronger inactivation with repetitive pulses than with a single long depolarization. More surprisingly, slow inactivation of the Kv1 Shaker K(+) channel (Shaker B Delta 6--46) also has a U-shaped voltage dependence for 10-s depolarizations. The time and voltage dependence of recovery from inactivation reveals two distinct components for Shaker. Strong depolarizations favor inactivation that is reduced by K(o)(+) or by partial block by TEA(o), as previously reported for slow inactivation of Shaker. However, depolarizations near 0 mV favor inactivation that recovers rapidly, with strong voltage dependence (as for Kv2.1 and 3.1). The fraction of channels that recover rapidly is increased in TEA(o) or high K(o)(+). We introduce the term U-type inactivation for the mechanism that is dominant in Kv2.1 and Kv3.1. U-type inactivation also makes a major but previously unrecognized contribution to slow inactivation of Shaker.     

Membrane stretch accelerates activation and slow inactivation in Shaker channels with S3-S4 linker deletions

At low P(open)(V) Shaker exhibits pronounced stretch-activation. Possible explanations for Shaker's sensitivity to tension include 1) Shaker channels are sufficiently distensible that stretch produces novel channel states and 2) Shaker channels expand in the plane of the membrane during voltage gating. For channels expressed in oocytes, we compared effects of patch stretch on Shaker and mutants that retain their voltage-gating ability but activate sluggishly because all or most of the S3-S4 linker has been deleted. Deletants had 10, 5, or 0 amino acid (aa) linkers, whereas wild-type is 31 aa. In deletants, though activation is exceptionally slow, slow inactivation is exceptionally quick; the resulting kinetic match was a bonus that allowed effects of stretch to be followed simultaneously in both processes. With the intact linker, an approximately 3 orders of magnitude mismatch in the two processes makes this impracticable. Standard stretch stimuli increased the rates and extent of activation by about the same degree in wild type and deletants, with effects especially pronounced near the foot of G(V). In deletants (where slow inactivation is strongly coupled to activation) stretch also accelerated slow inactivation. Maximum conductances were unaffected by stretch in all variants. In ramp clamp dose experiments, near-lytic patch stretch acted, for all variants, like a approximately 10 mV hyperpolarizing shift. These results suggested that, whether basal rates were high (wild type) or low (deletants), stretch acted by facilitating voltage-dependent activation. Channel activity was therefore simulated with/without "tension," tension being simulated via rate changes at voltage-dependent closed-closed transitions that might involve in-plane expansion (explanation 2). Simulated Delta P(open) arising from approximately 2 kT of "mechanical gating energy" mimicked experimental effects seen with comfortably sub-lytic stretch.     

Voltage dependence of slow inactivation in Shaker potassium channels results from changes in relative K(+) and Na(+) permeabilities

Time constants of slow inactivation were investigated in NH(2)-terminal deleted Shaker potassium channels using macro-patch recordings from Xenopus oocytes. Slow inactivation is voltage insensitive in physiological solutions or in simple experimental solutions such as K(+)(o)//K(+)(i) or Na(+)(o)//K(+)(i). However, when [Na(+)](i) is increased while [K(+)](i) is reduced, voltage sensitivity appears in the slow inactivation rates at positive potentials. In such solutions, the I-V curves show a region of negative slope conductance between approximately 0 and +60 mV, with strongly increased outward current at more positive voltages, yielding an N-shaped curvature. These changes in peak outward currents are associated with marked changes in the dominant slow inactivation time constant from approximately 1.5 s at potentials less than approximately +60 mV to approximately 30 ms at more than +150 mV. Since slow inactivation in Shaker channels is extremely sensitive to the concentrations and species of permeant ions, more rapid entry into slow inactivated state(s) might indicate decreased K(+) permeation and increased Na(+) permeation at positive potentials. However, the N-shaped I-V curve becomes fully developed before the onset of significant slow inactivation, indicating that this N-shaped I-V does not arise from permeability changes associated with entry into slow inactivated states. Thus, changes in the relative contributions of K(+) and Na(+) ions to outward currents could arise either: (a) from depletions of [K(+)](i) sufficient to permit increased Na(+) permeation, or (b) from voltage-dependent changes in K(+) and Na(+) permeabilities. Our results rule out the first of these mechanisms. Furthermore, effects of changing [K(+)](i) and [K(+)](o) on ramp I-V waveforms suggest that applied potential directly affects relative permeation by K(+) and Na(+) ions. Therefore, we conclude that the voltage sensitivity of slow inactivation rates arises indirectly as a result of voltage-dependent changes in the ion occupancy of these channels, and demonstrate that simple barrier models can predict such voltage-dependent changes in relative permeabilities.     

Steady-state availability of sodium channels. Interactions between activation and slow inactivation.

Changes in holding potential (Vh), affect both gating charge (the Q(Vh) curve) and peak ionic current (the F(Vh) curve) seen at positive test potentials. Careful comparison of the Q(Vh) and F(Vh) distributions indicates that these curves are similar, having two slopes (approximately 2.5e for Vh from -115 to -90 mV and approximately 4e for Vh from -90 to -65 mV) and very negative midpoints (approximately -86 mV). Thus, gating charge movement and channel availability appear closely coupled under fully-equilibrated conditions. The time course by which channels approach equilibration was explored using depolarizing prepulses of increasing duration. The high slope component seen in the F(Vh) and Q(Vh) curves is not evident following short depolarizing prepulses in which the prepulse duration approximately corresponds to the settling time for fast inactivation. Increasing the prepulse duration to 10 ms or longer reveals the high slope, and left-shifts the midpoint to more negative voltages, towards the F(Vh) and Q(Vh) distributions. These results indicate that a separate slow-moving voltage sensor affects the channels at prepulse durations greater than 10 ms. Charge movement and channel availability remain closely coupled as equilibrium is approached using depolarizing pulses of increasing durations. Both measures are 50% complete by 50 ms at a prepulse potential of -70 mV, with proportionately faster onset rates when the prepulse potential is more depolarized. By contrast, charge movement and channel availability dissociate during recovery from prolonged depolarizations. Recovery of gating charge is considerably faster than recovery of sodium ionic current after equilibration at depolarized potentials. Recovery of gating charge at -140 mV, is 65% complete within approximately 100 ms, whereas less than 30% of ionic current has recovered by this time. Thus, charge movement and channel availability appear to be uncoupled during recovery, although both rates remain voltage sensitive. These data suggest that channels remain inactivated due to a separate process operating in parallel with the fast gating charge. We demonstrate that this behavior can be simulated by a model in which the fast charge movement associated with channel activation is electrostatically-coupled to a separate slow voltage sensor responsible for the slow inactivation of channel conductance.     

An electrostatic interaction between TEA and an introduced pore aromatic drives spring-in-the-door inactivation in Shaker potassium channels

Slow inactivation of Kv1 channels involves conformational changes near the selectivity filter. We examine such changes in Shaker channels lacking fast inactivation by considering the consequences of mutating two residues, T449 just external to the selectivity filter and V438 in the pore helix near the bottom of the selectivity filter. Single mutant T449F channels with the native V438 inactivate very slowly, and the canonical foot-in-the-door effect of extracellular tetraethylammonium (TEA) is not only absent, but the time course of slow inactivation is accelerated by TEA. The V438A mutation dramatically speeds inactivation in T449F channels, and TEA slows inactivation exactly as predicted by the foot-in-the-door model. We propose that TEA has this effect on V438A/T449F channels because the V438A mutation produces allosteric consequences within the selectivity filter and may reorient the aromatic ring at position 449. We investigated the possibility that the blocker promotes the collapse of the outer vestibule (spring-in-the-door) in single mutant T449F channels by an electrostatic attraction between a cationic TEA and the quadrupole moments of the four aromatic rings. To test this idea, we used in vivo nonsense suppression to serially fluorinate the introduced aromatic ring at the 449 position, a manipulation that withdraws electrons from the aromatic face with little effect on the shape, net charge, or hydrophobicity of the aromatic ring. Progressive fluorination causes monotonically enhanced rates of inactivation. In further agreement with our working hypothesis, increasing fluorination of the aromatic gradually transforms the TEA effect from spring-in-the-door to foot-in-the-door. We further substantiate our electrostatic hypothesis by quantum mechanical calculations.     

Membrane tension accelerates rate-limiting voltage-dependent activation and slow inactivation steps in a Shaker channel

A classical voltage-sensitive channel is tension sensitive—the kinetics of Shaker and S3–S4 linker deletion mutants change with membrane stretch (Tabarean, I.V., and C.E. Morris. 2002. Biophys. J. 82:2982–2994.). Does stretch distort the channel protein, producing novel channel states, or, more interestingly, are existing transitions inherently tension sensitive? We examined stretch and voltage dependence of mutant 5aa, whose ultra-simple activation (Gonzalez, C., E. Rosenman, F. Bezanilla, O. Alvarez, and R. Latorre. 2000. J. Gen. Physiol. 115:193–208.) and temporally matched activation and slow inactivation were ideal for these studies. We focused on macroscopic patch current parameters related to elementary channel transitions: maximum slope and delay of current rise, and time constant of current decline. Stretch altered the magnitude of these parameters, but not, or minimally, their voltage dependence. Maximum slope and delay versus voltage with and without stretch as well as current rising phases were well described by expressions derived for an irreversible four-step activation model, indicating there is no separate stretch-activated opening pathway. This model, with slow inactivation added, explains most of our data. From this we infer that the voltage-dependent activation path is inherently stretch sensitive. Simulated currents for schemes with additional activation steps were compared against datasets; this showed that generally, additional complexity was not called for. Because the voltage sensitivities of activation and inactivation differ, it was not possible to substitute depolarization for stretch so as to produce the same overall P_O time course. What we found, however, was that at a given voltage, stretch-accelerated current rise and decline almost identically—normalized current traces with and without stretch could be matched by a rescaling of time. Rate-limitation of the current falling phase by activation was ruled out. We hypothesize, therefore, that stretch-induced bilayer decompression facilitates an in-plane expansion of the protein in both activation and inactivation. Dynamic structural models of this class of channels will need to take into account the inherent mechanosensitivity of voltage-dependent gating.     

Structural requirements for rapid inactivation and voltage dependence in splice variants of lobster Shaker potassium channels

We studied the properties of currents generated in Xenopus oocytes by nine splice variants of the spiny lobster Shaker gene. These isoforms differ in their amino termini and in the P-loop region of the pore. Both the voltage dependence and kinetic properties of the currents varied significantly, depending on which amino terminus was present. A cluster of net positive charges at the N-terminus was not necessary for rapid inactivation: negatively charged N-termini also inactivated rapidly. There was no obvious correlation between N-terminus length and inactivation rate. These N-terminal effects were additive with a separate set of voltage and kinetic properties controlled by the two alternative P-loop exons.     

Voltage-dependent potassium channels: minK and Shaker families

In the last 4 years, the molecular identity of several types of voltage-dependent potassium channels has been discovered. These include channels that terminate action potentials and control repetitive neuronal firing, as well as channels whose biological role is not yet understood. The majority of these are encoded by genes related to the Drosophila Shaker gene. The large number of genes comprising the Shaker gene family, coupled with the existence of different channels that result from alternatively spliced messages from the same gene, provide both vertebrates and invertebrates with a wide selection of channels whose voltage-dependence and kinetics can be tailored to the needs of a specific cell. Mutagenesis experiments on such channels are providing new information on those regions of the protein that govern essential aspects of channel activity, such as gating by voltage and ion permeation. Another gene, unrelated to the Shaker family, encodes a voltage-dependent potassium channel that activates much more slowly than the Shaker channels. This has been termed the MinK channel.     

Gating currents in Shaker K+ channels. Implications for activation and inactivation models.

We have studied ionic and gating currents in mutant and wild-type Shaker K+ channels to investigate the mechanisms of channel activation and the relationship between the voltage sensor of the channel and its inactivation particle. The turn on of the gating current shows a rising phase, indicating that the hypothetical identical activation subunits are not independent. Hyperpolarizing prepulses indicate that most of the voltage-dependence occurs in the transitions between closed states. The open-to-closed transition is voltage independent, as suggested by the presence of a rising phase in the off gating currents. In Shaker channels showing fast inactivation, the off gating charge is partially immobilized as a result of depolarizing pulses that elicit inactivation. In mutant channels lacking inactivation, the charge is recovered quickly at the end of the pulse. Internal TEA mimics the inactivation particle in its behavior but the charge immobilization is established faster and is complete. We conclude that the activation mechanism cannot be due to the movement of identical independent gating subunits, each undergoing first order transitions, and that the inactivation particle is responsible for charge immobilization in this channel.     

Electrostatics and the gating pore of Shaker potassium channels

Various experiments have suggested that the S4 segment in voltage-dependent Na(+) and K(+) channels is in contact with a solvent-accessible cavity. We explore the consequences of the existence of such a cavity through the electrostatic effects on the gating currents of Shaker K(+) channels under conditions of reduced ionic strength S. We observe that approximately 10-fold reductions of intracellular S produce reductions of the measured gating charge of approximately 10%. These effects continue at even lower values of S. The reduction of gating charge when S is reduced by 10-fold at the extracellular surface is much smaller (approximately 2%). Shifts of the Q(V) curve because of a reduced S are small (<10 mV in size), which is consistent with very little fixed surface charge. Continuum electrostatic calculations show that the S effects on gating charge can be explained by the alteration of the local potential in an intracellular conical cavity of 20-24-A depth and 12-A aperture, and a smaller extracellular cavity of 3-A depth and the same aperture. In this case, the attenuation of the membrane potential at low S leads to reduction of the apparent gating charge. We suggest that this cavity is made by a bundle of transmembrane helices, and that the gating charge movement occurs by translocation of charged residues across a thin septum of approximately 3-7 A thickness.     

Mechanisms of maurotoxin action on Shaker potassium channels

Maurotoxin (alpha-KTx6.2) is a toxin derived from the Tunisian chactoid scorpion Scorpio maurus palmatus, and it is a member of a new family of toxins that contain four disulfide bridges (, Eur. J. Biochem. 254:468-479). We investigated the mechanism of the maurotoxin action on voltage-gated K(+) channels expressed in Xenopus oocytes. Maurotoxin blocks the channels in a voltage-dependent manner, with its efficacy increasing with greater hyperpolarization. We show that an amino acid residue in the external mouth of the channel pore segment that is known to be involved in the actions of other peptide toxins is also involved in maurotoxin's interaction with the channel. We conclude that, despite the unusual disulfide bridge pattern, the mechanism of the maurotoxin action is similar to those of other K(+) channel toxins with only three disulfide bridges.     

Atomic proximity between S4 segment and pore domain in Shaker potassium channels

A recently proposed model for voltage-dependent activation in K+ channels, largely influenced by the KvAP X-ray structure, suggests that S4 is located at the periphery of the channel and moves through the lipid bilayer upon depolarization. To investigate the physical distance between S4 and the pore domain in functional channels in a native membrane environment, we engineered pairs of cysteines, one each in S4 and the pore of Shaker channels, and identified two instances of spontaneous intersubunit disulfide bond formation, between R362C/A419C and R362C/F416C. After reduction, these cysteine pairs bound Cd2+ with high affinity, verifying that the residues are in atomic proximity. Molecular modeling based on the MthK structure revealed a single position for S4 that was consistent with our results and many other experimental constraints. The model predicts that S4 is located in the groove between pore domains from different subunits, rather than at the periphery of the protein.     

Energetics of Shaker K channels block by inactivation peptides

A synthetic peptide of the NH2-terminal inactivation domain of the ShB channel blocks Shaker channels which have an NH2-terminal deletion and mimics many of the characteristics of the intramolecular inactivation reaction. To investigate the role of electrostatic interactions in both peptide block and the inactivation process we measured the kinetics of block of macroscopic currents recorded from the intact ShB channel, and from ShB delta 6-46 channels in the presence of peptides, at different ionic strengths. The rate of inactivation and the association rate constants (k(on)) for the ShB peptides decreased with increasing ionic strength. k(on) for a more positively charged peptide was more steeply dependent on ionic strength consistent with a simple electrostatic mechanism of enhanced diffusion. This suggests that a rate limiting step in the inactivation process is the diffusion of the NH2-terminal domain towards the pore. The dissociation rates (k(off)) were insensitive to ionic strength. The temperature dependence of k(on) for the ShB peptide was very high, (Q10 = 5.0 +/- 0.58), whereas k(off) was relatively temperature insensitive (Q10 approximately 1.1). The results suggest that at higher temperatures the proportion of time either the peptide or channel spends in the correct conformation for binding is increased. There were two components to the time course of recovery from block by the ShB peptide, indicating two distinct blocked states, one of which has similar kinetics and dependence on external K+ concentration as the inactivated state of ShB. The other is voltage- dependent and at -120 mV is very unstable. Increasing the net charge on the peptide did not increase sensitivity to knock-off by external K+. We propose that the free peptide, having fewer constraints than the tethered NH2-terminal domain binds to a similar site on the channel in at least two different conformations.     

Hydropathic analysis and comparison of KcsA and Shaker potassium channels

The similarity in structure of potassium (K(+)) channels from different families has been revealed by only recently available crystallographic 3D structural data. The hydropathic analysis presented in this work illuminates whether homologous residues perform the same functions in channels that use different gating mechanisms. We calculated and compared the hydropathic profiles of two K(+) channels, KcsA and Kv1.2 (the latter a member of the Shaker family), at their pore-forming domain. Quantitative information describing important interactions stabilizing the protein beyond obvious secondary-structure elements was extracted from the analysis and applied as a template for subsequent molecular-dynamics (MD) analyses. For example, two key groups of interactions, defining the turns that connect the transmembrane helices and responsible for the orientation of the pore helix, were identified. Our results also indicate that Asp(80) and Asp(379) play a similar role in stabilizing the P-loop of KcsA and Kv1.2, respectively, but to significantly different extents.     

Thermodynamic coupling between activation and inactivation gating in potassium channels revealed by free energy molecular dynamics simulations

The amount of ionic current flowing through K(+) channels is determined by the interplay between two separate time-dependent processes: activation and inactivation gating. Activation is concerned with the stimulus-dependent opening of the main intracellular gate, whereas inactivation is a spontaneous conformational transition of the selectivity filter toward a nonconductive state occurring on a variety of timescales. A recent analysis of multiple x-ray structures of open and partially open KcsA channels revealed the mechanism by which movements of the inner activation gate, formed by the inner helices from the four subunits of the pore domain, bias the conformational changes at the selectivity filter toward a nonconductive inactivated state. This analysis highlighted the important role of Phe103, a residue located along the inner helix, near the hinge position associated with the opening of the intracellular gate. In the present study, we use free energy perturbation molecular dynamics simulations (FEP/MD) to quantitatively elucidate the thermodynamic basis for the coupling between the intracellular gate and the selectivity filter. The results of the FEP/MD calculations are in good agreement with experiments, and further analysis of the repulsive, van der Waals dispersive, and electrostatic free energy contributions reveals that the energetic basis underlying the absence of inactivation in the F103A mutation in KcsA is the absence of the unfavorable steric interaction occurring with the large Ile100 side chain in a neighboring subunit when the intracellular gate is open and the selectivity filter is in a conductive conformation. Macroscopic current analysis shows that the I100A mutant indeed relieves inactivation in KcsA, but to a lesser extent than the F103A mutant.     

A molecular link between activation and inactivation of sodium channels

A pair of tyrosine residues, located on the cytoplasmic linker between the third and fourth domains of human heart sodium channels, plays a critical role in the kinetics and voltage dependence of inactivation. Substitution of these residues by glutamine (Y1494Y1495/QQ), but not phenylalanine, nearly eliminates the voltage dependence of the inactivation time constant measured from the decay of macroscopic current after a depolarization. The voltage dependence of steady state inactivation and recovery from inactivation is also decreased in YY/QQ channels. A characteristic feature of the coupling between activation and inactivation in sodium channels is a delay in development of inactivation after a depolarization. Such a delay is seen in wild-type but is abbreviated in YY/QQ channels at -30 mV. The macroscopic kinetics of activation are faster and less voltage dependent in the mutant at voltages more negative than -20 mV. Deactivation kinetics, by contrast, are not significantly different between mutant and wild-type channels at voltages more negative than -70 mV. Single-channel measurements show that the latencies for a channel to open after a depolarization are shorter and less voltage dependent in YY/QQ than in wild-type channels; however the peak open probability is not significantly affected in YY/QQ channels. These data demonstrate that rate constants involved in both activation and inactivation are altered in YY/QQ channels. These tyrosines are required for a normal coupling between activation voltage sensors and the inactivation gate. This coupling insures that the macroscopic inactivation rate is slow at negative voltages and accelerated at more positive voltages. Disruption of the coupling in YY/QQ alters the microscopic rates of both activation and inactivation.     

On the activation-inactivation coupling in Shaker potassium channels.

The 'ball-and-chain' model suggests the existence of a negative site which may attract the positively charged inactivation ball to occlude the pore when the channel is in the open state. For Shaker K+ channels, we propose that the state-dependent negative site be tryptophan-435, which becomes negatively charged after receiving an electron from tyrosine-445. The kinetic scheme for the channel's activation-inactivation coupling as derived from the YW-gated model resembles a successful 'scheme 8' proposed by Zagotta and Aldrich. Our model suggests that the final rapid voltage-independent transition to the open state is due to the deprotonation of tyrosine-445.     

Interaction between fast and slow inactivation in Skm1 sodium channels.

Rat skeletal muscle (Skm1) sodium channel alpha and beta 1 subunits were coexpressed in Xenopus oocytes, and resulting sodium currents were recorded from on-cell macropatches. First, the kinetics and steady-state probability of both fast and slow inactivation in Skm1 wild type (WT) sodium channels were characterized. Next, we confirmed that mutation of IFM to QQQ (IFM1303QQQ) in the DIII-IV 'inactivation loop' completely removed fast inactivation at all voltages. This mutation was then used to characterize Skm1 slow inactivation without the presence of fast inactivation. The major findings of this paper are as follows: 1) Even with complete removal of fast inactivation by the IFM1303QQQ mutation, slow inactivation remains intact. 2) In WT channels, approximately 20% of channels fail to slow-inactivate after fast-inactivating, even at very positive potentials. 3) Selective removal of fast inactivation by IFM1303QQQ allows slow inactivation to occur more quickly and completely than in WT. We conclude that fast inactivation reduces the probability of subsequent slow inactivation.     

Voltage-dependent gating of Shaker A-type potassium channels in Drosophila muscle

The voltage-dependent gating mechanism of A1-type potassium channels coded for by the Shaker locus of Drosophila was studied using macroscopic and single-channel recording techniques on embryonic myotubes in primary culture. From a kinetic analysis of data from single A1 channels, we have concluded that all of the molecular transitions after first opening, including the inactivation transition, are voltage independent and therefore not associated with charge movement through the membrane. In contrast, at least some of the activation transitions leading to first opening are considerably voltage dependent and account for all of the voltage dependence seen in the macroscopic currents. This mechanism is similar in many ways to that of vertebrate neuronal voltage-sensitive sodium channels, and together with the sequence similarities in the S4 region suggests a conserved mechanism for voltage-dependent gating among channels with different selectivities. By testing independent and coupled models for activation and inactivation we have determined that the final opening transition and inactivation are not likely to arise from the independent action of multiple subunits, each with simple gating transitions, but rather come about through their aggregate properties. A partially coupled model accurately reproduces all of the single-channel and macroscopic data. This model will provide a framework on which to organize and understand alterations in gating that occur in Shaker variants and mutants.