|
|
|
|
|
|
|
||
Gating Competence of Constitutively Open CLC-0 Mutants Revealed by the Interaction with a Small Organic Inhibitor
Address correspondence to Michael Pusch, Istituto di Biofisica, Sezione di Genova, CNR, via de Marini, 6, I-16149 Genova, Italy. Fax: (139) 0106475 500; email: pusch{at}icb.ge.cnr.it
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
4 µM). A three-state linear model with an open state, a low-affinity and a high-affinity CPA-bound state can quantitatively describe steady-state and kinetic properties of the CPA block. The parameters of the model and additional mutagenesis suggest that the high-affinity CPA-bound state is similar to the closed configuration of the protopore gate of wild-type CLC-0. In the E166A mutant the glutamate side chain that occludes the permeation pathway is absent. Thus, if gating consists only in movement of this side-chain the mutant E166A should not be able to assume a closed conformation. It may thus be that fast gating in CLC-0 is more complex than anticipated from the bacterial structures.
Key Words: CLC • chloride channel • clofibric acid • chloride dependence • mutation
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The determination of the three-dimensional structure of bacterial CLC homologues (Dutzler et al., 2002) marked a breakthrough of the understanding of CLC channels. Unfortunately, bacterial homologues could not be successfully studied electrophysiologically (Maduke et al., 1999). For the "prototype" Torpedo channel, CLC-0 (Jentsch et al., 1990), pioneering studies showed that it is functionally a dimer with two identical but independent permeation pathways (Miller and White, 1984), a model that was fully confirmed by mutagenesis (Ludewig et al., 1996; Middleton et al., 1996) and a low-resolution structure (Mindell et al., 2001), and then verified by the high-resolution structure of the bacterial StCLC (Dutzler et al., 2002). CLC channels are voltage dependent. However, the voltage dependence is not caused by the movement of a charged segment of the protein, as is the case for cation channels (Sigworth, 1994; Jiang et al., 2003a,b), but rather by the movement of a permeant anion that is accompanying channel opening in an obligate manner (Pusch et al., 1995; Chen and Miller, 1996). A highly conserved glutamate residue blocks the exit of a crystallographically resolved Cl- ion in StCLC and Dutzler et al. (2002) proposed that the structure corresponds to a closed channel configuration and that the glutamate could be the sensor of the extracellular Cl-. Recently Dutzler et al. (2003) succeeded in obtaining crystal structures of an E. coli CLC channel (EriC) in which the critical glutamate was mutated to alanine (E148A) and glutamine (E148Q). The structures of both mutants apparently represent an open channel with three closely located anion binding sites, with the outermost corresponding to the location of the negatively charged E148 side-chain in the closed channel structure. The suggestion that the structures of the mutated channels represent an open state was supported by functional data obtained with mutants of the glutamate in vertebrate CLC channels that lead to a constitutively open phenotype (Fahlke et al., 1997; Schmidt-Rose and Jentsch, 1997; Dutzler et al., 2003; Estévez et al., 2003). Thus, it appears that the single protopore gating of CLC channels involves only a very small conformational rearrangement.
In the present paper we investigated mutations of the critical glutamate and neighboring residues in the prototype Torpedo CLC-0 channel. In particular, we studied the block of the constitutively open mutants E166A and E166S by intracellular p-chlorophenoxy-acetic acid (CPA). The binding site of CPA and of another Cl- channel blocker, 9-AC, was recently mapped in the muscle CLC-1 channel to a binding pocket (Estévez et al., 2003) that partially overlaps with the central Cl- ion binding site seen in the crystal structure of EriC (Dutzler et al., 2002, 2003). The critical glutamate (E232 in CLC-1) was proposed to form part of the CPA binding site based on a tenfold increased CPA affinity of the mutant E232C (Estévez et al., 2003). Estévez et al. (2003) suggested that the identified binding pocket corresponds to a structure of the closed channel because CPA and 9-AC exhibit a strongly state-dependent block being much more potent at negative voltages, where channels are closed (Pusch et al., 2001; Accardi and Pusch, 2003; Estévez et al., 2003). However, the above mentioned, newly resolved, and presumably open bacterial CLC structures (Dutzler et al., 2003) suggest on the contrary that the open and closed conformations display only minimal structural differences.
These predictions can be tested using the small organic molecule CPA as a tool. CPA has a drastically different affinity for the closed and for the open state with the open channel block being fast and of low affinity and the closed channel block of relatively high affinity (Accardi and Pusch, 2003). It is thus a priori expected that CPA block of the constitutively open mutant E166A corresponds to the low-affinity, open-channel block of WT CLC-0. In contrast, we find that CPA exerts a strongly voltage-dependent high-affinity CPA block reminiscent of the closed state block of WT CLC-0. Further analysis of the CPA block of E166A and E166S in CLC-0 shows that these constitutively open mutants are able to assume a conformation that closely resembles the closed state of WT CLC-0 and that this conformation is strongly stabilized by CPA. This finding has important implications for the mechanism of gating of CLC-0.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). All constructs were in the PTLN vector (Lorenz et al., 1996). cRNA was transcribed and injected in Xenopus oocytes as described (Accardi and Pusch, 2000).
Electrophysiology
Currents were recorded using the two-electrode voltage-clamp methods and excised patch-clamp recording as described (Pusch et al., 2000). For whole-oocyte voltage-clamp the bath solution contained (in mM) 100 NaCl (or 100 NaI), 4 MgSO4, 5 HEPES, pH 7.3, and the holding potential was chosen close to the resting membrane potential (-30 to -50 mV). For patch-clamping, the intracellular solution contained (mM): 100 N-methyl-D-glucamine (NMDG)-Cl, 2 MgCl2, 10 HEPES, 2 EGTA, pH 7.3, whereas the standard extracellular solution contained 100 NMDG-Cl, 5 MgCl2, 10 HEPES, pH 7.3. In the "low Cl-" extracellular solution 90 mM Cl- was replaced by the impermeable glutamate. All substances were purchased from Sigma-Aldrich. CPA was applied to the internal side of the membrane by inserting the patch pipette into perfusion tubes of 0.5-mm diameter. Patch-clamp data were recorded using an EPC-7 amplifier (HEKA) and the Pulse acquisition program (HEKA), and were analyzed using custom software. The holding potential in patch clamp measurements was 0 mV throughout.
Nonstationary Noise Analysis
Noise analysis was performed as described (Pusch et al., 1994; Saviane et al., 1999). Briefly, 50–100 identical pulses to various test potentials were applied and the mean response, I, was calculated. The variance, 
2, was calculated from the averaged squared difference of consecutive traces. Background variance at 0 mV was subtracted and the variance-mean plot was fitted by 2 = iI - I2/N, with the single channel current, i, and the number of channels, N, as free parameters. The maximal open-probability, pmax, was estimated by
 |
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The glutamate shown in bold was mutated in CLC-0 and first studied using two-electrode voltage-clamp. For comparison, representative traces for WT CLC-0 are shown in Fig. 1 B. All mutants drastically affect gating, whereas none eliminated the characteristic block by extracellular iodide (Fig. 1, thick traces) or significantly altered the ion selectivity sequence among halides (unpublished data), indicating that the glutamate is not important for ion selectivity. Changing the glutamate to serine (unpublished data) or alanine (Fig. 1 C) completely abolished any gating relaxation, resulting in an almost linear, ohmic phenotype. In contrast, the charge-conserving E166D mutation significantly slowed channel opening (Fig. 1 D).
|
15 mM (reduction of [Cl-]ext in two-electrode voltage clamp and of [Cl-]ext or [Cl-]int in excised patch clamp measurements) only shifted the current reversal potential and reduced the current amplitude (unpublished data), as expected for an open, Cl--selective pore (Hille, 2001).
|
; Estévez et al., 2003) to the intracellular face of the membrane (Fig. 2, B and C). CPA strongly inhibited the currents, especially at negative voltages, and the inhibition developed with concentration- and voltage-dependent kinetics. We evaluated first the steady-state inhibition as a function of the concentration of CPA ([CPA]) (Fig. 3
A). The steady-state inhibition could be well described by a simple Michaelis-Menten relationship of the form
 | (1) |
4 µM and thus 200-fold larger than that of CLC-0 (Accardi and Pusch, 2003; Estévez et al., 2003).
|
) the CPA block of mutant E166A was not much different at negative voltages in a solution with a lower intracellular chloride concentration, whereas for positive voltages the reduction of [Cl-]int led to an increase of the apparent affinity by about a factor of three (Fig. 4)
.
|
with voltage-dependent opening rate-constants 
and ß (Hanke and Miller, 1983; Chen and Miller, 1996; Pusch et al., 2001). The lack of time, voltage, and [Cl-]ext dependence of currents mediated by E166A channels could be caused by various reasons: (a) state C could be unstable (i.e., >> ß); (b) and ß could have become voltage and [Cl-]ext independent; (c) state C could be conducting in the mutant, with a conductance similar to that of WT or a combination of these possibilities.
The latter possibility (c) is suggested by the finding of Dutzler et al. (2002) that the glutamate residue appears to block the passage of Cl- ions and by the hypothesis that the crystal structure of StCLC represents the closed conformation. Removing the "barrier" in the mutant could render the regular closed state a conductive open state. The recent "open" structures of mutated E. coli CLC (Dutzler et al., 2003) further strengthen the viability of this possibility.
The following experiments are aimed at a distinction of these three possibilities. Stationary and nonstationary noise analysis can provide substantial insight into the microscopic properties of ion channels (Conti, 1984) and recordings can be acquired at high bandwidth and with relatively ease. In particular, possibility (b) (voltage-independent and ß) can easily be tested using nonstationary noise analysis if state C remains a nonconducting closed state. In fact, if channels fluctuated between open and closed states with a certain voltage-independent open probability, pO = /( + ß), they would generate a stationary and measurable variance of magnitude 2 = N x i2 x pO x (1 - pO), where N is the number of channels and i the single-channel current amplitude.
For an almost constitutively open channel like E166A, nonstationary noise analysis cannot be applied in a useful manner. However, we can exploit the strong voltage dependence of the inhibition by CPA to obtain information about the open probability and the single-channel current in the absence of CPA by performing noise analysis in the presence of a small concentration of CPA. This leads to current relaxations upon voltage-steps, whereas we know that at large positive voltages only a very small steady-state inhibition is present. Such a nonstationary noise analysis is illustrated in Fig. 5
that shows the results in the presence of CPA (A, 1 mM; B and C, 0.1 mM) at 80 mV (A and B) and -100 mV (C). The left panels show the mean response (top panel) and the variance (bottom panel) calculated from several identical stimulations as described in MATERIALS AND METHODS. The right panels show the variance, 2, plotted versus the mean current, I, (symbols) together with a parabolic fit (lines) that yields estimates for the single channel current, i, the number of channels, N, and the maximal open probability, pmax (see MATERIALS AND METHODS). The estimated value of the single-channel current, i, at 80 mV is similar in 0.1 and 1 mM CPA (0.42 and 0.37 pA, respectively). The single-channel conductance estimated from these values and those measured at negative voltages is similar to that of WT CLC-0 (Bauer et al., 1991; Chen and Miller, 1996; Ludewig et al., 1997b; Lin et al., 1999). Most importantly, the pmax values at 80 mV (achieved at the end of the pulse) and also at -100 mV (corresponding to the initial points of the pulse) are close to one. For the experiments shown in Fig. 5, the values are: pmax = 0.88 at 80 mV with 1 mM CPA, pmax = 0.95 at 80 mV with 0.1 mM CPA, and pmax = 0.94 at -100 mV with 0.1 mM CPA. The smaller value of pmax in 1 mM CPA compared with that measured in 0.1 mM is in quantitative agreement with the block seen in 1 mM CPA at 80 mV that amounts to 0.07 (Fig. 3 A). Hence, an open probability of 0.88 in 1 mM CPA would scale to 0.94 in 0.1 mM CPA, almost exactly as observed experimentally. Thus, from the noise analysis an upper estimate for the probability of occurrence of a closed state, as in possibility (b), is 5%. These considerations do not take other noise sources into account, in particular open-channel noise, a prominent feature of CLC-0 (Ludewig et al., 1997b). From the noise analysis we can practically exclude possibility (b) that the opening and closing transitions have become voltage independent and that the channel spends a significant amount of time in a closed state. Using noise analysis alone it is, however difficult to distinguish between the possibilities (a) and (c) or a combination of these with possibility (b). Schematically, possibility (a) is represented by a single kinetic state, O, whereas (c) can possibly be schematized by
|
and ß have not completely lost their voltage dependence, the absence of macroscopic-gating relaxations in response to voltage steps excludes the possibility that that the conductance of the two states is significantly different, in agreement with the results from the noise analysis. Additional insight into the gating behavior of E166A channels was obtained by registrations from patches that contained a small number of channels. An example is shown in Fig. 6 A, which shows superimposed traces recorded at -100 mV in the presence of 0.1 mM CPA (black trace) and without CPA (gray trace). With CPA the channels show a clear apparent gating behavior that is remarkably similar to regular single protopore gating. Without CPA the channels remain mostly in the conductive state, while occasionally gating transitions are observed, as for example at the end of the stretch illustrated in Fig. 6 A. Often, but not always, these transitions corresponded to a conductance change of about double the size of the unitary size seen with CPA (e.g., last transition of gray trace in Fig. 6 A, see arrow), and thus they probably represent gating transitions of the common gate. Even though we could not detect voltage-dependent common gate transitions using macroscopic recordings (unpublished data), it cannot be excluded that such transitions occur in a voltage-independent manner. At 5 mM CPA (Fig. 6, B and C), the overall open probability is drastically reduced and the openings at -100 mV are much shorter than those seen with 0.1 mM CPA. At this concentration the apparent conductance level of the openings was also significantly reduced compared with the ones seen with 0.1 mM CPA (Fig. 6 D). This reduction of the conductance most likely represents a fast open channel block (see below). At 60 mV the channels stay open most of the time in the absence of CPA (Fig. 6 E) and also with 0.1 mM CPA present at the intracellular side (Fig. 6 F). Occasional, slow time-scale gating transitions are seen also at 60 mV (unpublished data). Addition of 5 mM CPA leads to a flickery appearance of the current (Fig. 6 G), probably representing unresolved closure events.
|
In particular, we investigate what the minimal requirements for a quantitative description of the properties of CPA block are.
Assuming a single kinetic state, O, and considering the simplest case in which CPA is a voltage-dependent open channel blocker that binds to and dissociates from the channel with the experimentally observed kinetics, as in the scheme
the apparent KD is the "true" open channel KD given by KD = koff/kon. Scheme III
|
-1, of the inhibition process is described by
 |
-1 is a linear function of [CPA]. In contrast to this, the experimental values for -1 (Fig. 8
A, symbols) show a clear saturation at large [CPA].
|
|
-1, it is reasonable to hypothesize that drug-bound E166A channels are also able to close, in agreement with previous studies of the block of CLC-0 by CPA and related substances (Pusch et al., 2001), resulting in the extended scheme
Here, drug-bound blocked channels (OB) are able to undergo a conformational change to a closed configuration (CB), described by the closing rate constant ß and the opening rate constant . Since the relaxations in the presence of CPA are well described by a single-exponential function and as the open channel block of CLC-0 by CPA is a fast process (Accardi and Pusch, 2003), we can suppose that CPA dissociation is much faster than the gating transitions, described by and ß. This assumption is also supported by the reduction of the apparent single channel amplitude seen with 5 mM CPA at -100 mV (Fig. 6 D). With this simplification, states O and OB are in fast equilibrium and inhibition by CPA is characterized by an apparent dissociation constant
 | (2) |
-1, depends on [CPA] as
 | (3) |
We fitted Eq. 3 to the experimental values for -1 (Fig. 8 A, lines) to estimate , ß, and the open channel KD. Although ß appears to be voltage independent (Fig. 8 B, squares), the opening rate, (Fig. 8 B, circles), depends exponentially on voltage with an apparent valence of 0.58 (Fig. 8 B, line). The estimated open channel KD (Fig. 3 B, diamonds) is much larger than the apparent KDapp at negative voltages. As an internal check of consistency, the parameters obtained by the fit of Eq. 3 can be used to calculate KDapp according to Eq. 2. The values (Fig. 3 B, open squares) are in good agreement with those found directly from the steady-state inhibition. Thus, the extended Scheme IV
|
; Estévez et al., 2003). Its value is in the millimolar range and it has only a slight voltage dependence. Also the reduction of the single channel current amplitude seen in 5 mM CPA (reduction of the amplitude by 25% corresponding to an apparent KD of 15 mM, Fig. 6 D) is qualitatively consistent with the open channel KD of 5 mM at -100 mV found from the kinetic analysis (see Fig. 3, triangles). Even though the values differ by a factor of 3, the order of magnitude is correct and it must be kept in mind that the values in Fig. 3 for the open channel KD have been obtained in a rather indirect way.
Furthermore, the voltage dependence of is very similar to that of the CLC-0 protopore opening rate at positive voltages (apparent valence z = 0.58 found here compared with z = 0.7 for the Cl--dependent component of the opening rate of CLC-0; Chen and Miller, 1996) suggesting that the opening step of CPA-bound mutant channels is analogous to the regular conformational channel opening. The analogy is not perfect as, at difference to WT gating, the closing rate constant, ß, appears to be voltage independent. Altogether, Scheme IV is in very good quantitative agreement with the CPA block and results in the prediction that CPA stabilizes a conformation of the channel that corresponds to the regular closed state of WT CLC-0.
Even though Scheme IV is sufficient for a quantitative description of the CPA block it assumes inherently that the closed conformation is possible only with CPA bound and that CPA cannot bind directly to or dissociate from the closed conformation. These two properties are probably an oversimplification. Similar to what has been found for derivatives of CPA (Pusch et al., 2001) and also for CPA itself (Accardi and Pusch, 2003) a quantitative description of CPA (and CPB) block of CLC-0 pores requires the possibility that the drug binds to open and also to closed channels, though with much lower affinity and faster kinetics to the open than to the closed state. Furthermore, CPA stabilizes the closed state rendering the opening of drug-bound closed channels more difficult (Pusch et al., 2002). Adopting these findings to the gating Scheme I
|
with inhibition constants for the closed state KDC = koffC/konC and for the open state KDO = koffO/konO. Scheme V
|
+ ß or koffC (or both) are much larger than ' + ß' the scheme can be reduced to an effective three-state system with effective rate-constants connecting states CB and O. This scheme, which also predicts for example single exponential relaxation kinetics if the open-channel block is fast, cannot be easily distinguished from the simpler Scheme IV that lacks a direct connection between states CB and O. However, a common feature of all these schemes that are quantitatively compatible with the data (Schemes IV–VI) is the presence of an additional state, CB, with a relatively high CPA affinity.
As stated above, the parameters obtained by fitting Scheme IV to the kinetic data revealed a striking resemblance of the transition CB
OB to the regular CLC-0 protopore gating, suggesting that the relaxations seen in the presence of CPA at least partially reflect a conformational change that resembles regular channel gating, and that therefore state CB is analogous to the regular closed state of WT pores.
Formally, however, Scheme IV is completely equivalent to a model in which the high-affinity state is another open state
in which the rate constants and ß do not reflect a gating transition but for example a movement of CPA from a more superficial site to a deeper site.
To further test if the -ß transitions reflect regular gating transitions we changed additional parameters that are known to interfere with fast gating of CLC-0. First, we investigated the effect of reducing the extracellular Cl- concentration ([Cl-]ext) in the presence of 0.1 mM intracellular CPA. It can be seen that reducing [Cl-]ext from 110 to 20 mM leads to a rightward shift of the apparent open-probability by 30 mV (Fig. 9)
, similar to what is seen for the regular protopore gating of CLC-0 (Pusch et al., 1995; Chen and Miller, 1996). A similar shift might be expected, however, also for the two-open-state Scheme VI
|
|
; Ludewig et al., 1996, 1997a). The mutations were introduced in the background of E166A and in the absence of CPA displayed a voltage-independent phenotype, however, with an outwardly rectifying current voltage relationship (unpublished data). In the presence of CPA both mutations strongly slow down the onset of and the recovery from inhibition (Fig. 10)
with K519E having a larger effect.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, similarly to what happens in CLC-1 (Fahlke et al., 1997; Schmidt-Rose and Jentsch, 1997; Estévez et al., 2003). Using noise-analysis and recordings with a small number of channels we showed that the "open" phenotype of these mutations represents indeed an almost permanently conductive state, in agreement with similar findings of Dutzler et al. (2003). Infrequent gating transitions were observed for E166A that represent possible transitions of the common gate. It is, of course, difficult to definitively rule out the occurrence of closures of the regular single protopore gate. Altogether, however, the data presented here suggest that such gating transitions are practically insignificant.
In the crystal structure of the E148A mutant of the E. coli CLC EriC, the position of the missing negatively charged side-chain is occupied by a chloride ion while the rest of the structure is practically identical to that of the WT (Dutzler et al., 2003). Consequently, Dutzler et al. (2003) concluded that the single protopore gate mainly involves a reorientation of the Glu side-chain with almost no additional conformational change and that the alanine mutant corresponds to a permanently open channel. Here, we used the small molecule CPA as a tool to further test this hypothesis.
CPA can be used to probe gating transitions because it presents a strongly state-dependent binding. Open channel block is fast and of low affinity, while closed channels are inhibited with much higher affinity (Accardi and Pusch, 2003). CPA block of WT CLC-0 could be quantitatively described by a model similar to one that had been developed for the inhibition by the related compound CPB (Pusch et al., 2001) (Scheme V).
If the mutant E166A indeed represents a permanently open channel, its CPA-blocking properties are expected to correspond to that of the open-channel block of WT CLC-0, i.e., block should be of low-affinity, of fast kinetics, and of slight voltage dependence. The opposite was found here: CPA block is even stronger than the closed channel block of WT CLC-0, it is relatively strongly voltage dependent, and displays slow kinetics. How can this discrepancy be explained?
A linear three-state model with a fast low-affinity open-channel block and a subsequent high-affinity state could quantitatively account for the voltage and concentration dependence of steady-state inhibition and inhibition kinetics. Schemes IV and VI are formally equivalent but differ significantly in the interpretation of the transition between the two CPA-bound states. In Scheme IV these transitions represent a conformational change that might correspond to a closing of the channel, analogous to the block mechanism of WT CLC-0. In contrast, in Scheme VI the transition between the low- and the high-affinity states reflects a rearrangement of the CPA molecule.
For the following reasons we favor Scheme IV over Scheme VI. First, and most importantly, Scheme VI does not explain why the relaxation kinetics seen in the presence of CPA change in parallel to the normal gating kinetics when additional mutations are introduced (K519E or S123T), whereas Scheme IV naturally explains this coincidence because the relaxations seen in the presence of CPA actually reflect the regular gating. Second, no evidence for a high affinity open channel block was obtained for WT CLC-0 (Accardi and Pusch, 2003) and no evidence for two binding sites arose from the study by Estévez et al. (2003), who located the 9-AC and CPA binding site close to the central Cl- ion binding site (Estévez et al., 2003). Lastly, the voltage dependence of the "opening" rate is similar to that which has been described for the regular opening rate of CLC-0 (Chen and Miller, 1996). On the other hand, two other findings seemingly contradict Scheme IV: first, intracellular chloride has only a small effect on the apparent CPA affinity of the mutant, whereas the regular closing rate increases with smaller [Cl-]int (Chen and Miller, 1996). Second, the "closing" rate, ß, is found to be voltage independent here while the regular closing rate increases at negative voltages with an apparent valence of approximately -0.35 (Chen and Miller, 1996). In order for Scheme IV to be valid it must therefore be required that the closing transition with a bound CPA is differently sensitive to the intracellular chloride concentration and voltage compared with the regular closing rate. We believe that such a different behavior of the closing transition with a bound CPA compared with "no CPA" is not completely unexpected because it might well be that the presence of a CPA molecule inside the channel could alter the movement of chloride ions that is normally associated with channel closure and thereby alter chloride dependence and voltage dependence.
However, we cannot strictly rule out either of the two models.
Scheme IV, i.e., the ability of mutant E166A to undergo a conformational change that corresponds to a channel closure, is in contradiction to the predictions of Dutzler et al. (2003), according to which the mutant is not able to undergo a conformational closure because it lacks the relevant negatively charged side-chain of the glutamate residue. Thus, it might be that fast gating of CLC-0 involves additional conformational changes that are not revealed by the bacterial structures.
Alternatively, CPA block of E166A does not involve a conformational change as in Scheme VI. In that case the striking analogies of the transition between low- and high-affinity states to the regular gating transitions raise the suspicion that also the regular CLC-0 fast-gating kinetics do not reflect a conformational change but the movement of an ion within the channel protein.
Based on the differential effects of several pore mutations on CPA block of CLC-0 Accardi and Pusch (2003) concluded that opening of CLC-0 involves a conformational change in addition to the side-chain movement of glutamate 166, thus favoring the "conformational" Scheme IV.
|
|
|
FOOTNOTES |
|---|
|
ACKNOWLEDGMENTS |
|---|
This work was supported by grants from Telethon Italy (grant 1079) and the Italian Research Ministry (FIRB RBAU01PJMS). S. Traverso receives a CNR doctoral fellowship.
David C. Gadsby served as editor.
Submitted: 3 January 2003
Accepted: 21 July 2003
|
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:
|
|
 |
|
  J. Lisal and M. Maduke Proton-coupled gating in chloride channels Phil Trans R Soc B, January 27, 2009; 364(1514): 181 - 187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X.-D. Zhang, P.-Y. Tseng, W.-P. Yu, and T.-Y. Chen Blocking Pore-open Mutants of CLC-0 by Amphiphilic Blockers J. Gen. Physiol., December 29, 2008; 133(1): 43 - 58. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Jayaram, A. Accardi, F. Wu, C. Williams, and C. Miller Ion permeation through a Cl--selective channel designed from a CLC Cl-/H+ exchanger PNAS, August 12, 2008; 105(32): 11194 - 11199. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Zifarelli, A. R. Murgia, P. Soliani, and M. Pusch Intracellular Proton Regulation of ClC-0 J. Gen. Physiol., July 1, 2008; 132(1): 185 - 198. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Traverso, G. Zifarelli, R. Aiello, and M. Pusch Proton Sensing of CLC-0 Mutant E166D J. Gen. Physiol., December 27, 2005; 127(1): 51 - 66. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Babini and M. Pusch A Two-Holed Story: Structural Secrets About ClC Proteins Become Unraveled? Physiology, October 1, 2004; 19(5): 293 - 299. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Accardi, L. Kolmakova-Partensky, C. Williams, and C. Miller Ionic Currents Mediated by a Prokaryotic Homologue of CLC Cl- Channels J. Gen. Physiol., January 26, 2004; 123(2): 109 - 119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. I. Niemeyer, L P. Cid, L. Zuniga, M. Catalan, and F. V Sepulveda A conserved pore-lining glutamate as a voltage- and chloride-dependent gate in the ClC-2 chloride channel J. Physiol., December 15, 2003; 553(3): 873 - 879. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
