|
|
|
|
|
|
|
||
ARTICLE |
Correspondence to Carol Deutsch: cjd{at}mail.med.upenn.edu
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Furthermore, the allosteric coupling among different gates of the same ion channel provides a means for enhancing the subtleties and range of ion channel function.
Voltage-gated potassium (Kv) channels in the Shaker subfamily have three well-studied gates: an activation gate and two types of inactivation gates. These two inactivation gates, which typically prevent ion flow through depolarized channels, correspond to processes originally designated N-type and C-type inactivation. In general, N-type inactivation is the faster of the two, leading to the name "slow inactivation" for the latter. The activation gate is formed by a four-helix bundle at the cytoplasmic end of the pore, specifically by the C-terminal region of the sixth transmembrane segment (S6) of each of the four subunits of Kv channels (Liu et al., 1997; del Camino et al., 2000; del Camino and Yellen, 2001). The fast inactivation gate, lacking in many Kv channels, is composed of the cytosolic N terminus, which snakes up into the cavity of the channel and blocks conductance (Hoshi et al., 1990; Choi et al., 1991; Demo and Yellen, 1991; Zhou et al., 2001a). Virtually all Kv channels possess a slow inactivation gate that closes by a cooperative rearrangement of regions in the outer mouth of the pore and selectivity filter in response to prolonged depolarization (Hoshi et al., 1991; Ogielska et al., 1995; Panyi et al., 1995; Liu et al., 1996; Loots and Isacoff, 1998).
In the simplest depiction, these gates can each assume either of two positions, open or closed. Thus, there are four composite gating states to consider, as shown in Fig. 1 A.
For convenience, we refer to the slow inactivation gate simply as the inactivation gate. By this we mean a region in (or near) the selectivity filter that operationally functions as a gate. We assign names to the four composite gating states according to standard terminology, where C and O represent the closed and open conformations of the activation gate, respectively, and I indicates that the inactivation gate is closed (Fig. 1, A and B). The only completely open (i.e., conducting) state in this depiction is state O, because closure of either gate prevents ion flux. Note that this representation focuses on the gates and ignores the many conformational states that are known to be associated with each of these composite states, especially the many conformations the channel can assume when the activation gate is closed. In the diagram in Fig. 1 B, left-to-right movement (C
O or CIOI) is opening of the activation gate in response to a depolarization, and movement from top to bottom (CCI or OOI) is closing of the inactivation gate.
|
), the way in which these gates are energetically and kinetically coupled is less well understood. It is known that N- and C-type inactivation coexist and that N-type can speed C-type inactivation (Baukrowitz and Yellen, 1995), but the relationship between the activation gate and the slow inactivation gate has not been elucidated. What is the order and timing of movement of these two gates after a change of membrane potential? Does the state of one gate affect the movement of the other? If these two gates operate independently, then the time- and voltage-dependent state of the activation gate will be oblivious to whether the inactivation gate is open or closed. To test this hypothesis, we investigated the status of the activation gate during entry into and recovery from slow inactivation. Our findings indicate that the rate of activation gate movement depends on the state of the inactivation gate. Closure of the inactivation gate slows closing of the activation gate at negative voltages and speeds its opening. Moreover, we show that hyperpolarization closes the activation gate long before a channel recovers from inactivation.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
DNA Clones and Site-directed Mutagenesis
Modified Shaker-IR in a GW1-CMV mammalian expression plasmid, under the control of a highly expressing Kozak consensus promoter sequence (Kozak, 1991), was provided by R. Horn (Thomas Jefferson University, Philadelphia, PA) (Ding and Horn, 2002). This construct includes a deletion of amino acids 6–46 to remove N-type inactivation and C301S and C308S point mutations (Holmgren et al., 1996). These two mutations are necessary to exclude possible methanethiosulfonate ethyltrimethylammonium (MTSET; Toronto Research Corp.) modification of endogenous Shaker cysteines and accompanying functional effects. Amino acid substitutions at position 449 (T449A) and 474 (V474C) were introduced using a site-directed mutagenesis kit (QuikChange; Stratagene). Mutants were sequenced at the University of Pennsylvania School of Medicine DNA Sequencing Facility using a 16-capillary sequencing apparatus (ABI 3100; Applied Biosystems, Inc.) with BigDye Taq FS Terminator (version 3.1; Applied Biosystems, Inc.) chemistry. A Calcium Phosphate Transfection Kit (Invitrogen) was used to cotransfect CD8 carried in a EBO-pcD vector (Margolskee et al., 1988, 1993) with the Shaker-IR T449A/V474C construct using 4 or 10 µg of CD8 or Shaker-IR DNA, respectively, per 100-mm dish of tsA201 cells. Transfected cells were replated onto 35-mm polystyrene cell culture dishes (Corning) pretreated with poly–L-ornithine (Sigma-Aldrich) to improve cell adhesion for excising patches. 12–36 h after transfection, current was recorded from transfected cells, which were identified by decoration with anti-CD8 antibody-coated Dynabeads (Dynal) as described previously (Margolskee et al., 1993; Jurman et al., 1994).
Electrophysiology
Standard methods were used to record currents in inside-out patches. Data were acquired using a HEKA EPC-9 amplifier, digitized with an ITC-16 analogue-to-digital converter (HEKA Elektronik, GmbH). The filter frequency was set to less than or equal to half the sampling frequency. Pipettes pulled from lead-free 8520 glass (Warner Instruments) were 8–9 M
and coated with R-6101 elastomer (Essex Group, Inc.) and fire polished. Adjustments for bath-pipette liquid junction potentials were made before current recording. Typical current amplitudes were 300–600 pA at +50 mV test potential, allowing the recording of macroscopic currents. Only those cells with a steady-state current <5% of the peak current were used in these experiments. Inside-out currents are plotted with reversed polarity for easier comparison with conventional whole-cell currents. In general, all holding potentials were –120, and voltage errors were <3 mV after series resistance compensation. All experiments were performed at room temperature (20–24°C). Data were analyzed using Pulse + PulseFit software (version 8.77; HEKA Elektronik, Dr. Schulze, GmbH). Reported errors are SEM.
Solutions
The standard intracellular (bath) solution contained 105 mM KF, 35 mM KCl, 10 mM EGTA, and 10 mM HEPES titrated to pH 7.36–7.38 with KOH for a final concentration of 160–165 mM K+ and an osmolarity of 285–295 mOsm. For Cs liberation experiments, the standard intracellular solution was identical except that it contained 105 CsF and 35 CsCl instead of the corresponding K+ salts and was titrated with CsOH. Standard extracellular (pipette) solution was as follows: 150 mM NaCl, 2 mM KCl, 1.5 mM CaCl2, 1 mM MgCl2, and 10 mM HEPES, pH 7.38. The osmolarity of the solution was 290 mOsm, and pH was titrated to 7.36–7.38 with NaOH. MTSET solution was made fresh in intracellular (bath) solutions from 100 mM stocks in water, which were stored at –80°C. MTSET stock was freshly diluted into the bath solution and loaded immediately in the perfusion apparatus before the start of MTSET application. Fresh MTSET solution was loaded into the perfusion apparatus every 8 min.
Characterization of the Perfusion Kinetics
A Warner Instruments SF-77A Perfusion Fast-Step system with three-barrel square glass (700 µM internal diameter) was used for rapid solution exchange. Inside-out patches were perfused with standard intracellular solutions at a rate of 0.5 ml/min. Fig. 2 A illustrates the calibration of the solution exchange.
The patch was positioned in the center barrel and exposed to the bath solution (140 mM K+, unfilled portion of bar) at the holding potential. The patch was depolarized from –120 mV to +50 mV for 600 ms, and the barrel was simultaneously switched to a low K+ solution (50 mM K+, hatched portion of bar). The duration of the low K+ exposure was 300 ms. The solution exchange is determined by two factors: a delay (d), followed by a rapid exponential decay characterized by the exchange time constant (
e). The time required for complete solution exchange was calculated as d + 4e. Fig. 2 B shows the distribution of d and e in a box plot. The medians of d and e were 33 and 4.13 ms, respectively. Characteristics of solution exchange were determined for each patch, and values were used to calculate the correct cumulative modification time for MTSET.
|
Analysis of Data
The cumulative modification time for the nth pulse is
 | (1) |
e are defined in Fig. 2. A single exponential function was fit to the data points, and the modification rate (
) was determined. In the simplest case, this rate is proportional to three terms, as follows:
 | (2) |
 | (3) |
O, OICI, and CIOI transitions. For CO and CIOI transitions, the value of Eq. 3 was calculated using PX(t) = (1–e–t/)4 and PX,max = 1, where is derived from the current traces elicited at +50 mV. For OICI transitions, the value of Eq. 3 was calculated using an exponential decay function for POI(t) based on determination of ßI (see Fig. 5).
Recovery Measurements
To measure recovery, a standard two-step voltage protocol was used. Unless otherwise stated, voltage steps were to +50 mV. The first step, used to inactivate channels and measure initial peak current amplitude, was at least fivefold longer than the channel's time constant for inactivation. After a recovery period defined as the interpulse interval (ipi) at –120 mV, the second voltage step was applied and peak recovered current amplitude was measured. No leak subtraction was used. Fractional recovery (FR) was calculated as (I2 – Iss1)/(I1 – Iss1), where I1 and I2 represent the peak current amplitudes during the first and second depolarizing step, respectively, whereas Iss1 represents the steady-state current at the end of the first depolarization. To determine the time course of recovery from inactivation (see Fig. 3 E), ipi was varied between 0 and 60 s. For the liberation experiments (see Figs. 6 and 7), the ipi was set to give 90% recovery (ipi 
21 s; FR 0.9) when K+ was present continuously (see Fig. 6, protocol 1). This procedure for determining ipi was made for every cell before Cs+ trapping and liberation to avoid complicating effects of K+ accumulation on recovery kinetics (Levy and Deutsch, 1996). To ensure complete restoration of current between recovery measurements and maintain viable patches, patches were maintained at –100 mV in 
for 
2 min after the final depolarization in intracellular K+ for a given recovery protocol. Statistical significance of differences in mean FR was determined using a one-way ANOVA on all the data, followed by an all-pairwise multiple comparison (Student-Newman- Keuls method).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
130 ms at +50 mV (Fig. 3 A).
Residue 474 faces into the water-filled cavity between the inactivation and activation gates, and accessibility of 474C to hydrophilic modifying reagents reliably tracks the status of the activation gate (Liu et al., 1997). This double mutant was studied in inside-out patches from tsA201 cells and exposed to intracellular MTSET to obtain modification rate constants in the open state, at depolarized voltages, and when the channel is inactivated, at both depolarized and hyperpolarized voltages.
|
O transitions in the noninactivated and inactivated channel. A Na-based extracellular solution was used in the pipette and a KF/KCl solution in the bath (intracellular solution). The patch was held at –120 mV and depolarized to test potentials ranging from –80 to +70 mV in 10-mV increments every 45 s (Fig. 3 A). The voltage dependence of the open probability (PO) was determined from the peak currents, as described in the legend to Fig. 3 B. The midpoint of the PO vs voltage curve is –50.5 ± 0.8 mV (n = 3). This is in reasonable agreement with V474C Shaker-IR, which has an activation midpoint of approximately –58 mV (Liu et al., 1997). A step to +50 mV thus ensures that all channels open quickly and that the PO at this voltage is maximal. By –80 mV, the open probability is close to zero. Similar solutions and configuration were used to determine the voltage dependence of steady-state inactivation (Fig. 3 C; see legend for details); the midpoint is –72.9 ± 2.1 mV (n = 3). Therefore, at voltages negative to –80 mV, all channels are in a noninactivated conformation at steady state. Thus, holding potentials in the range of –80 to –120 mV ensure a high probability that channels contain a closed activation gate and an open inactivation gate (Fig. 1 A, composite state C).
In addition to these thermodynamic properties, two kinetic parameters were required to set conditions for this study: closing kinetics of the activation gate (OC; i.e., deactivation) at negative voltages and recovery kinetics from slow inactivation. The former was determined by standard tail current analysis (a step to the indicated voltage after a brief 7-ms depolarization to +50 mV). A fit of the decaying tail current to a single exponential function gave deactivation time constants (d) of 0.59 ± 0.05 ms (n = 5) and 1.03 ± 0.07 ms (n = 15) at –120 and –105 mV, respectively (Fig. 3 D). These time constants are comparable to those obtained from Shaker-IR (Zagotta et al., 1994). At voltages more hyperpolarized than –80 mV, a channel with a closed activation gate will not reopen. Therefore, in this voltage range the rate of deactivation represents the rate of closure of the activation gate. Recovery at –120 mV, after complete slow inactivation at +50 mV (for 1.5 s), was measured using a standard two-pulse protocol (Levy and Deutsch, 1996) to give a recovery time constant (rec) of 10.6 ± 0.5 s (n = 4; Fig. 3 E). These electrophysiological parameters were used to devise protocols to determine (a) when the activation gate closes after repolarization of the inactivated channel, (b) whether the closed inactivation gate stabilizes the activation gate in a particular configuration, and (c) whether inactivation alters the closing and opening rates of the activation gate. The results of these determinations will reflect whether movement of the activation gate is coupled to the state of the inactivation gate.
Time of Activation Gate Closure in the Inactivated Channel
To assess the open/closed status of the activation gate, we measured the kinetics of modification of V474C, which is only accessible to intracellular cysteine reagents when the activation gate is open (Liu et al., 1997). MTSET modification of V474C renders the channel nonconducting. Inside-out patches were exposed to MTSET during or after appropriate protocols designed to capture channels predominantly in one composite state only, either O, C, OI, or CI. In all cases, the kinetics of solution changes were determined for each patch (see Materials and methods), and this exchange time was used to calculate the correct cumulative modification time for MTSET. In general, solution changes were 90% complete within 20 ms (Fig. 2 and Materials and methods). The rate constant for modification of the open channel was determined by repeated brief depolarizations to +50 mV (Fig. 4 A) in the continuous presence of 200 µM of MTSET.
The 7-ms duration of the depolarizing pulse is sufficient to completely activate channels but is too short to induce inactivation (Fig. 4 A). A decrease of the peak currents occurs upon repetitive pulsing in the presence of MTSET (Fig. 4 A). The peak currents for each pulse were determined and normalized to the peak current of the first pulse and plotted as a function of cumulative modification time (Fig. 4 B). The modification rate is a product of the probability of being in the MTSET-modifiable state, kmod (the second order modification rate constant for that state), and the MTSET concentration (see Eq. 2 in Materials and methods). The probability of being in the MTSET-modifiable state changes with time, as manifested in the gradual activation of the current (Fig. 4 A). Thus, the integral of this probability must be used to calculate the modification rate constant (see Eq. 3 in Materials and methods). A single exponential function was fit to the data to give a kmod for the open state of 1.28 x 105 M–1s–1. The average of four replicate experiments is 1.31 x 105 ± 0.19 x 105 M–1s–1 (n = 4). These values are consistent with those obtained for V474C in the Shaker-IR open channel (Liu et al., 1997).
|
rec 10 s at –120 mV; Fig. 3 E); i.e., the channels are still inactivated, but not yet recovered. Then MTSET was applied for 500 ms. This yielded no substantial modification; i.e., there was no decrement in peak current (Fig. 4, C and D). These results suggest that the activation gate closes within 400 ms upon return of the voltage to –120 mV. We suggest that this is the CI state of the channel, which cannot be modified because access of MTSET to 474C is prevented by closure of the activation gate.
Our second evaluation of inactivated channels was designed to study the OI state and was therefore performed at a depolarized voltage. We determined kmod at +50 mV, after 800 ms of depolarization at +50 mV, which inactivated the channels completely (residual steady-state current was 2.2 ± 0.2% of the peak current; n = 45). Fig. 4 E shows that exposure to MTSET during successive depolarizing pulses to +50 mV decreases peak outward current. Normalized current plotted as a function of cumulative modification time was well fit by a single exponential to give a kmod of 9.13 x 103 M–1s–1 (Fig. 4 F), only 7% of the rate for the open state. The average kmod was 9.31 x 103 ± 0.99 x 103 M–1s–1 (n = 5). This reduced rate of modification could reflect reduced accessibility/reactivity of V474C in the inactivated channel, or it could represent a small contamination of the channel pool with open channels. This latter possibility was eliminated by the following experiment and analysis. We determined the modification rate for two distinctly different probabilities of the channel being inactivated (POI). An 800-ms depolarization to +50 mV is sufficient to inactivate 98% of channels (inact is 130.4 ± 7.8 ms at +50 mV; n = 45; Fig. 4 G) before repolarization. Thus, an 800-ms depolarization produces a POI of 1 (Fig. 4, G and H), whereas an 80-ms depolarization produces a POI of 0.52 (Fig. 4, I and J). For both cases, we exposed patches to MTSET coincident with repolarization to –80 mV; however, in the latter case (POI = 0.52) a brief (5 ms) repolarization to –120 mV was used before MTSET application to close all open but noninactivated channels. This protocol prevents contamination of the results with modification of the open, noninactivated channels. Patches were exposed to MTSET for 400 ms at –80 mV, a voltage at which kmod is maximal and closed channels do not open (Fig. 3 B). The rates of modification are 13.41 x 103 (Fig. 4 H) and 7.14 x 103 M–1s–1(Fig. 4 J), respectively, a ratio of 0.53. These results demonstrate that the modification rates are proportional to POI and that a contaminating contribution from open, noninactivated channels does not occur. The reduced modification rate for the inactivated channel reflects reduced accessibility/reactivity of V474C. These results also suggest that the measured modification rate in Fig. 4 (E and F) reflects modification of the inactivated channel alone (i.e., the OI state). This OI state was entered from the open, noninactivated state of the channel at +50 mV. The modification rate for the OI state at –80 mV (Fig. 4 H) is larger than that obtained at +50 mV (Fig. 4 F). Because both the voltage dependence of [MTSET] in the pore and of activation gate opening would each predict a smaller modification rate at –80 mV, the observed increase likely reflects voltage-dependent conformations of the inactivated channel. No such additional voltage dependence exists for noninactivated channels (Liu et al., 1997).
To confirm that we are measuring modification rates of the OI state, we investigated reopening of the channel from the CI state. A transition from CI to OI should produce, at steady state, the same modification rate at +50 mV as channels that have entered the inactivated state directly from the open state. Provided that the final POI is the same in both cases and that the opening kinetics are comparable, regardless of the pathway, kmod should be the same. To test this hypothesis, we used the reopening protocol shown in Fig. 4 K. After a repolarization period of 400 ms at –120 mV (all channel activation gates have closed; Fig. 3 D), a step to +50 mV for 500 ms, with concomitant MTSET exposure, produces a kmod of 8.76 x 103 M–1s–1. The average kmod was 8.59 x 103 ± 0.88 x 103 M–1s–1 (n = 4), in close agreement with that obtained after a direct OOI transition during the initial depolarization to +50 mV (P > 0.59; Fig. 4 F). In contrast, when channels were first accumulated in the CI state, at –120 mV, and then triggered with a step to –80 mV to equilibrate between CI and OI states, the modification rate was zero; i.e., the activation gate does not open.
Closing Kinetics of the Activation Gate Are Coupled to the State of the Inactivation Gate
In noninactivated channels, the activation gate closes rapidly, with a time constant of 
1 ms at –120 mV (Fig. 3 D). However, we do not know the kinetics for closure of this gate in inactivated channels. The rate of activation gate closure in inactivated channels was measured using the following strategy. As depicted in Fig. 5 (A and D)
, complete inactivation of the channels was achieved by a depolarizing pulse to +50 for 800 ms, followed by pulses of varying duration at –120 or –105 mV before MTSET application. During this time at a hyperpolarized voltage, the activation gate may close. Closure of the activation gate within the duration of the hyperpolarizing pulse prevents modification during subsequent exposure to MTSET that is coincident with a step to –80 mV, a voltage at which CI channels do not open. Thus, a channel whose activation gate has closed at –120 or –105 mV in the allotted time of the pulse will remain closed at –80 mV. In contrast, a channel whose activation gate did not close within the allotted time will be modified by MTSET at –80 mV. Thus, the overall modification rate will be proportional to the fraction of channels remaining in the OI state after the hyperpolarizing pulse. That is, as activation gates close, POI will decrease and therefore, so will the overall modification rate. For example, a 20-ms hyperpolarizing pulse to –120 mV before MTSET application gives an overall rate constant of 5.1 x 103 M–1s–1 (Fig. 5 B) and an average of 5.56 x 103 ± 0.78 x 103 M–1s–1 (n = 3), a value that is significantly smaller (P < 0.012) than that obtained in the absence of the hyperpolarizing pulse (Fig. 4 H). This indicates that the activation gate of a substantial fraction of the channels closes within 20 ms at –120 mV. When this determination is made for various pulse durations at –120 mV and plotted as a function of pulse duration, an exponentially decaying curve is obtained (Fig. 5 C). The time constant for this decay, 23 ms, is the reciprocal of the closing rate constant, ßI, for inactivated channels. A similar determination at –105 mV (Fig. 5, E and F) yields a time constant of 63.3 ms. Both of these closing time constants are much larger than the corresponding deactivation time constants obtained from the tail currents for noninactivated channels (Fig. 3 D). To determine the closing rate for the activation gate at –80 mV, patches were returned to –80 mV after complete inactivation at +50 mV and exposed to MTSET after variable periods of time at –80 mV. For example, channels modified at –80 mV for 400 ms with a 400-ms delay after the end of the voltage step to +50 mV show an average rate constant of 7.79 x 103 ± 1.04 x 103 M–1s–1 (n = 3; Fig. 5 H). The dependence of kmod at –80 mV on duration at –80 mV before MTSET modification reveals a closing time constant of 785 ms, which is, again, substantially slower (100-fold) than deactivation of noninactivated channel. Two conclusions are apparent. First, as with noninactivated channels, the deactivation kinetics are voltage dependent, slowing with increasing depolarization. Second, a closed inactivation gate slows closing of the activation gate.
|
O or CIOI, the steady-state value of PO or POI will be the ratio 
/( + ß) or I/(I + ßI), respectively. These determinations are simplistic for both inactivated and noninactivated channels, in part because a two-state gating model is invalid for most Kv channels. Moreover, because POI is a function of both opening and closing rates, a determination of ßI alone (Fig. 5) is insufficient to determine POI. Opening kinetics must also be considered. The challenge here is to measure the opening of the activation gate of inactivated channels that do not produce measurable currents. To overcome this obstacle, we chose a strategy previously introduced to measure liberation of a trapped cation in inactivated Shaker-IR channels as an indicator of the opening of the activation gate (Ray and Deutsch, 2006). This method uses a multistep protocol and relies on trapping an intracellular cesium ion (Cs+) in the inactivated channel upon repolarization after a long depolarization. A trapped Cs+ ion in the inactivated channel causes the channel to recover slowly (measured by a third depolarizing pulse). If the trapped Cs+ is liberated before recovery, e.g., by opening the activation gate in the presence of intracellular K+ (second pulse), then the channel will recover at a rate commensurate with the recovery rate when a K+ ion is trapped in the channel (Ray and Deutsch, 2006). That is, the liberated Cs+ ion is replaced by a K+ ion. Thus, recovery kinetics can reveal whether a depolarization is long enough in duration for the activation gate to open and liberate a Cs+ ion.
Protocol 4 (Fig. 6) shows the liberation protocol, including the initial depolarizing pulse (a step to +50 mV for 800 ms), followed in the presence of intracellular Cs+ by a step to –120 mV, which traps the Cs+ in the channel.
We then applied the liberation pulse, a step to the indicated voltage for various durations, which liberates Cs+, replacing it with K+. A subsequent depolarization to +50 mV (the final pulse, 21 s later; see Materials and methods) is used to measure the FR. At –120 mV, in the presence of only, the FR is 0.88 ± 0.01 (n = 16; Fig. 6, protocol 1). In 
, FR is 0.64 ± 0.02 (n = 17; protocol 2). Whether is present continuously or a trapped Cs+ is liberated by a voltage step to +50 mV for 7 ms, recovery is virtually identical: FR is 0.90 ± 0.02 (n = 3; protocol 3) and 0.88 ± 0.03 (n = 4; protocol 4), respectively. At +50 mV the Cs+ is completely liberated within 7 ms (Fig. 6 and Fig. 7 A).
|
|
). Moreover, these first-latency kinetics describe the time course of liberation, assuming that the Cs+-K+ exchange is rapid when the activation gate is open (see Discussion in Ray and Deutsch, 2006). Therefore, the superposition in Fig. 7 A indicates that at +50 mV inactivation has no effect on opening kinetics of the activation gate. At more hyperpolarized voltages, however, where the closing rate (ß or ßI) is not negligible, the kinetics of activation gate opening (i.e., PO(t) or POI(t)) will in general be faster than the first-latency opening kinetics of the activation gate after a depolarization (see legend to Fig. 7). The first-latency kinetics is not a function of closing rates, whereas the activation kinetics includes eigenvalues that are functions of the sum of rates both entering and leaving the open state and is thus faster. To intuit this, consider the simple two-state model depicted in Fig. 1. The first latency rate (eigenvalue) is and the activation rate is + ß, which is greater than alone. Therefore, at sufficiently hyperpolarized voltages, where PO(
) is <1, a comparison of activation and liberation kinetics has further significance. Specifically, if the rate of liberation is faster than or equal to that of activation, then the CIOI transition (opening rate in the inactivated channel) must be faster than the CO transition (opening rate in the noninactivated channel).
At –50 mV, where PO is 0.5 at steady state for noninactivated channels (Fig. 3 B), Cs+ liberation kinetics were indistinguishable from those of channel activation (Fig. 7 B), as at +50 mV (Fig. 7 A). This result shows that the rate of activation gate opening in inactivated channels is greater than in noninactivated channels at –50 mV. We extended these measurements to more hyperpolarized voltages. Initially, we assessed FR using long durations (e.g., 300 ms) for the liberation pulses of –55, –60, and –70 mV (Fig. 7 D). The FR decreased monotonically from –55 to –70 mV; in the latter case, no channels reopened within 300 ms to liberate the trapped Cs+. At –60 mV, the liberation kinetics were too slow to measure accurately (unpublished data). The activation kinetics at –60 mV of noninactivated channels, however, displayed a time constant of 14 ms, clearly faster than the liberation kinetics. A comparison of the two sets of kinetics was nevertheless possible using a slightly less negative voltage. At –55 mV, liberation kinetics of inactivated channels were clearly slower than activation kinetics of noninactivated channels (Fig. 7 C). Moreover, both activation kinetics and liberation kinetics are slower at –55 mV than at –50 mV, in part reflecting the voltage dependence of activation gate opening. The discrepancy between the kinetics of liberation and activation shown in Fig. 7 C does not, unfortunately, offer a conclusive interpretation about activation gate opening at this voltage. As mentioned earlier, it is only when liberation kinetics are comparable to or faster than activation kinetics that we can make an unambiguous comparison of activation gate opening for inactivated and noninactivated channels. Nevertheless, we have demonstrated that the closing rate of the activation gate is decreased by inactivation (Fig. 5) and that the opening rate at –50 mV is increased by inactivation (Fig. 7 B). The composite results therefore indicate that a closed inactivation gate energetically favors an open activation gate and that inactivation alters both the closing and opening rates of the activation gate.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Starkus et al., 1998; Kiss et al., 1999; Wang et al., 2000a) and even that multiple inactivated states exist, including one that is only permeable to Na+ and through which an inactivated channel must move to deactivate and recover (the "R" state; Wang et al., 2000a). These observations led Fedida and co-workers to use Na+ tail currents as a readout for the inactivated state of Kv1.5 channel (Wang et al., 2000b; Wang and Fedida, 2001; Wang and Fedida, 2002; Kurata et al., 2004). However, the R state may not represent the inactivated state. It may only exist when K+ ions normally bound to sites in the selectivity filter are replaced with Na+ ions, which occurs under unusual ionic conditions that may themselves distort the selectivity filter (Zhou et al., 2001b). Moreover, the kinetics of the Na+ tails are difficult to interpret unambiguously. They reflect changes in macroscopic conductance and could be caused by changes in single-channel conductance, selectivity, and/or gating at the selectivity filter, and not to changes in activation gating. Despite these attempts to characterize the inactivated states, several questions remained unanswered. For example, precisely when does the activation gate close when an inactivated channel is hyperpolarized? Is movement of the activation gate coupled to inactivation gating?
To more directly and unambiguously address these issues, we chose an accessibility assay, similar to that first used by Yellen and co-workers to locate the activation gate in the lower portion of S6 (Liu et al., 1997; del Camino and Yellen, 2001). The results displayed in Fig. 4 constitute evidence for the existence of CI and OI states, i.e., states in which the activation gate is closed (CI) or open (OI) in the inactivated channel. Within 400 ms of returning completely inactivated channels to a hyperpolarized voltage, no modification of V474C occurs (Fig. 4). This is the CI state. The activation gate has closed, yet the channel is still inactivated and does not completely recover until 50 s later. Because the time constant for closure of the activation gate is 23 ms in an inactivated channel at –120 mV (Fig. 5), 500-fold faster than the recovery rate at this voltage, the rate-limiting step for recovery is not closure of the activation gate. The existence of an inactivated channel with an open activation gate (i.e., the OI state) is clearly demonstrated by the rate of MTSET modification at +50 mV after a prolonged depolarization and decrease of the outward ionic current to 2% of the peak amplitude. The activation gate remains open despite the lack of current. This is the OI state. Our results are consistent with those of Kurata et al. (Kurata et al., 2004), who showed that the N-terminal fast inactivation peptide and TEA can access the central cavity when the channel is slow inactivated at positive membrane potentials. Our clear demonstration of a voltage-dependent OICI equilibrium suggests that the pathway for recovery at hyperpolarized voltages is from OI to CI to C.
The lower kmod for the OI state (9.31 x 103 M–1s–1) versus the open state (1.31 x 105 M–1s–1), both at +50 mV, could reflect a different accessibility/environment for V474C in the cavity of an inactivated channel than in the open channel, each state having an open activation gate. Alternatively, the nonzero value of kmod could reflect a small fraction of open channels that are modified because flickering between inactivated (OI) and noninactivated (O) channels, each with open activation gates, leads to the inactivation gate being open 2% of the time at steady state. We favor the former hypothesis, because under conditions in which there are no noninactivated channels (i.e., –80 mV), we still observe a kmod that is substantially different from the modification rate for the open (O) state. This argues strongly that the lower modification rate likely reflects an altered conformation of the cavity in the inactivated state compared with the cavity in the open state. Such an altered conformation could be very slight and yet have a profound effect. An altered conformation of the inactivated state is consistent with the block of Kv1.3 by small-molecule inhibitors that bind to the pore at intracellular sites and are specific for the slow-inactivated state (Hanner et al., 2001).
The activation pathway includes several steps and preopen closed states of the channel, thus complicating a simple analysis of activation gate opening kinetics. Closure of the activation gate, however, may be considered as a single transition from the open state to a preopen closed state. This rate is ß (or ßI). Two considerations preclude the use of a simple analysis of the kinetics to determine open probabilities, i.e., POI = I/(I + ßI). First, as stated in the Results, activation is a multistep process. Second, we cannot precisely determine both the opening and the closing kinetics of the activation gate in the inactivated channel at the same voltage. Nonetheless, we may draw a valid conclusion: closing of the inactivation gate stabilizes the activation gate in the open position. At hyperpolarized voltages the activation kinetics are faster (Cs+ liberation experiment) and closing kinetics are slower (MTSET experiment) relative to the noninactivated channel; therefore, the POI must increase relative to PO at hyperpolarized voltages.
What is the mechanism for stabilization of the open state in the inactivated channel? Thus far, we do not have an atomic level structure of the inactivated state of a voltage-gated K+ channel. However, several structural studies of KcsA have suggested that the conformation of the inactivated channel includes altered ion occupancy in the selectivity filter of the pore (Lenaeus et al., 2005), rearrangement of residues in the outer mouth/turret (Liu et al., 1996; Cordero-Morales et al., 2006), and gross distortions of the selectivity filter itself (Zhou et al., 2001b). Whether any of these pertain to the Shaker-IR inactivated channel and provide clues to the mechanism by which the inactivation gate alters activation kinetics is not yet known. However, functional experiments provide clues. Thermodynamic mutant cycle analyses demonstrate that gating-sensitive residues that are as much as 15Å apart in S6 are coupled to each other (Yifrach and MacKinnon, 2002). These residues cluster mainly in the bundle-crossing region of the activation gate and the upstream region where S6 contacts the pore helix. Yifrach and MacKinnon suggest that pore opening involves simultaneous changes in both regions and that S6 moves as a rigid body. These features could underlie the crosstalk we observe between activation and inactivation gates. Moreover, because movement of the activation gate is coupled to voltage-sensor movement (Sigworth, 1994; Bezanilla, 2000), our demonstration of coupling between the inactivation and activation gates predicts that inactivation gate closure should slow the return of voltage sensors to their resting states at a hyperpolarized voltage. That is, inactivation should cause charge immobilization in gating current measurements. Studies by Olcese et al. confirm these predictions (Olcese et al., 1997).
Regardless of the mechanism by which these two gates are coupled, the cross talk between them will surely affect cell excitability because inactivation (slow and fast) governs the amount of available K+ current in a cell. It is this conductance that contributes critically to the resting membrane potential, the shape and frequency of the action potential, and repolarization of the membrane in the final phases of an action potential, all of which influence other membrane potential–dependent processes (e.g., neurotransmitter release; Dodson and Forsythe, 2004). Thus, slow inactivation, including both entry into and exit from the inactivated state, will regulate cell function in at least two ways: one that is direct and one that modulates excitability through its coupled effects on activation gating. Both activation and slow inactivation are commonly found in voltage-gated channels and play very important physiological roles in nonexcitable as well as excitable cells. It is therefore likely that the cross talk we have demonstrated between activation and inactivation gates for Shaker-IR is widespread and critical.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by National Institutes of Health grants GM 069837 and NS 052665, and Hungarian Ministry of Health ETT 068/2006. G. Panyi is a Bolyai Fellow.
Angus Nairn served as editor.
Submitted: 3 August 2006
Accepted: 20 September 2006
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Uysal, V. Vasquez, V. Tereshko, K. Esaki, F. A. Fellouse, S. S. Sidhu, S. Koide, E. Perozo, and A. Kossiakoff Crystal structure of full-length KcsA in its closed conformation PNAS, April 21, 2009; 106(16): 6644 - 6649. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Dougherty, J. A. De Santiago-Castillo, and M. Covarrubias Gating Charge Immobilization in Kv4.2 Channels: The Basis of Closed-State Inactivation J. Gen. Physiol., February 25, 2008; 131(3): 257 - 273. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Sadovsky and O. Yifrach Principles underlying energetic coupling along an allosteric communication trajectory of a voltage-activated K+ channel PNAS, December 11, 2007; 104(50): 19813 - 19818. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Chakrapani, J. F Cordero-Morales, and E. Perozo A Quantitative Description of KcsA Gating I: Macroscopic Currents J. Gen. Physiol., October 29, 2007; 130(5): 465 - 478. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Panyi and C. Deutsch Probing the Cavity of the Slow Inactivated Conformation of Shaker Potassium Channels J. Gen. Physiol., April 30, 2007; 129(5): 403 - 418. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. W. Claydon, M. Vaid, S. Rezazadeh, D. C.H. Kwan, S. J. Kehl, and D. Fedida A Direct Demonstration of Closed-State Inactivation of K+ Channels at Low pH J. Gen. Physiol., April 30, 2007; 129(5): 437 - 455. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
) was calculated as I/I–120, where I is the peak current evoked by 7-ms depolarization to +50 mV from a prepulse potential applied for 60 s, whereas I–120 is the peak current evoked by identical depolarization from the holding potential of –120 mV. The line shows the best fit Boltzmann function. The mean values ± SEM of the V1/2 and slope factor are shown (n = 3). (D) Deactivation kinetics of the current. In the protocol used, tail currents were evoked by stepping to the indicated membrane potentials after a 7-ms depolarization to +50 mV. The deactivation time constant (
characterizing the OI
