Initial Coupling of Binding to Gating Mediated by Conserved Residues in the Muscle Nicotinic Receptor
http://www.100md.com
《普通生理学杂志》
Receptor Biology Laboratory, Department of Physiology and Biomedical Engineering, Mayo Clinic College of Medicine, Rochester, MN 55905
We examined functional consequences of intrasubunit contacts in the nicotinic receptor subunit using single channel kinetic analysis, site-directed mutagenesis, and structural modeling. At the periphery of the ACh binding site, our structural model shows that side chains of the conserved residues K145, D200, and Y190 converge to form putative electrostatic interactions. Structurally conservative mutations of each residue profoundly impair gating of the receptor channel, primarily by slowing the rate of channel opening. The combined mutations D200N and K145Q impair channel gating to the same extent as either single mutation, while K145E counteracts the impaired gating due to D200K, further suggesting electrostatic interaction between these residues. Interpreted in light of the crystal structure of acetylcholine binding protein (AChBP) with bound carbamylcholine (CCh), the results suggest in the absence of ACh, K145 and D200 form a salt bridge associated with the closed state of the channel. When ACh binds, Y190 moves toward the center of the binding cleft to stabilize the agonist, and its aromatic hydroxyl group approaches K145, which in turn loosens its contact with D200. The positional changes of K145 and D200 are proposed to initiate the cascade of perturbations that opens the receptor channel: the first perturbation is of -strand 7, which harbors K145 and is part of the signature Cys-loop, and the second is of -strand 10, which harbors D200 and connects to the M1 domain. Thus, interplay between these three conserved residues relays the initial conformational change from the ACh binding site toward the ion channel.
Abbreviations used in this paper: AChBP, acetylcholine binding protein; Btx, bungarotoxin; CCh, carbamylcholine.
INTRODUCTION
Activation of post-synaptic receptors is accomplished by allosteric communication between the neurotransmitter binding site and the distal ion channel. The local conformational change due to neurotransmitter binding is amplified in a cascade that culminates in the global conformational change that opens the channel. For the Cys-loop superfamily of post-synaptic receptors, atomic-scale insight into the neurotransmitter binding site emerged from the x-ray structure of the homologous acetylcholine binding protein (AChBP; Brejc et al., 2001) and from homology models derived from it (Le Novere et al., 2002; Molles et al., 2002; Schapira et al., 2002; Sine et al., 2002a). Additionally a 4- resolution structural model of the binding, channel, and cytoplasmic domains was generated from electron microscopy of two-dimensional arrays of Torpedo receptors (Miyazawa et al., 2003; Unwin, 2005). At the interface between binding and pore domains, a network of loops was shown to couple neurotransmitter binding to channel gating (Kash et al., 2003; Bouzat et al., 2004; Chakrapani et al., 2004), but coupling structures near the neurotransmitter binding site remain unknown.
In the adult muscle nicotinic receptor, the two ACh binding sites are formed at interfaces between an subunit and a neighboring or subunit where multiple recognition domains from each subunit converge (for reviews see Corringer et al., 2000; Karlin, 2002; Sine, 2002). The subunit contributes recognition domains A–C, while the non- subunit contributes domains D–G. Conserved aromatic residues from domains A–D form an aromatic cage that coordinates the positively charged agonist (Celie et al., 2004), whereas nonconserved residues in domains D–G endow the two binding sites with selectivity for agonists and competitive antagonists (Sine, 2002). Domains C and F are the most peripheral of the recognition domains, and of these, domain C exhibits the greatest displacement on binding the agonist (Celie et al., 2004; Gao et al., 2005).
To identify structures that propagate the local conformational change elicited by ACh away from the binding site, we examined our homology model of the receptor ligand binding domain (Sine et al., 2002a). We looked for conserved residues at the periphery of the binding site near the mobile domain C and found a cluster of three conserved residues, K145, D200, and Y190, whose side chains potentially interact through electrostatic forces. Our patch clamp recordings show that structurally conservative mutations of each residue profoundly impair gating of the channel. The common contributions to channel gating and direct linkages to the binding site and binding-pore interface suggest interplay among these three residues initiates the allosteric cascade from the binding site to the channel gate.
MATERIALS AND METHODS
Construction of Wild-type and Mutant AChRs
Human , , , and subunit cDNAs were obtained as previously described (Ohno et al., 1996) and subcloned into the mammalian expression vector pRBG4 (Lee et al., 1991). Site-directed mutations were introduced using the Quick-Change site-directed mutagenesis kit (Stratagene). The presence of each mutation and the absence of unwanted mutation were determined by sequencing the entire cDNA insert.
Mammalian Cell Expression
BOSC 23 cells (Pear et al., 1993; Wang et al., 2000), a variant of HEK 293 clonal fibroblasts, were transfected with mutant or wild-type cDNAs using calcium phosphate precipitation as previously described (Lee and Sine, 2004). Patch clamp and [125I]-bungarotoxin (Btx) binding measurements were made 2 and 3 d after transfection, respectively.
Patch-clamp Recordings
Recordings were obtained in the cell-attached configuration (Hamill et al., 1981) at a membrane potential of –70 mV and a temperature of 21°C. Bath and pipette solutions contained 142 mM KCl, 5.4 mM NaCl, 1.8 mM CaCl2, 1.7 mM MgCl2, 10 mM HEPES (pH 7.4). Single channel currents were recorded using an Axopatch 200B (Axon Instruments Inc.). Data were obtained from two to four different patches for each ACh concentration. Recordings were accepted for analysis only when the frequency of occurrence of clusters of events was low enough to be sure they originated from a single receptor channel. Currents were low pass filtered at 100 kHz and recorded to hard disk at 200 kHz using the program Acquire (Bruxton Co.), using the option that allows high bandwidth sampling of periods of channel activity while keeping track of the intervening quiescent periods.
Single-channel Kinetic Analysis
Digitized current signals were analyzed as recently described (Lee and Sine, 2004) using a 10-kHz digital Gaussian filter, cubic spline interpolation of the signal, a dead time of 10 μs, the half-amplitude threshold for detection and correction of the measured dwell times at threshold for the effects of the Gaussian filter (Colquhoun and Sigworth, 1983). Dwell time histograms were constructed using a logarithmic abscissa and square root ordinate and fitted to the sum of exponential components by maximum likelihood using the program TACFit (Bruxton Co.). Clusters of events corresponding to a single channel were identified as a series of closely spaced openings preceded and followed by closed intervals greater than a specified duration. This duration was taken as the point of intersection of the closed time component that depended on ACh concentration with the succeeding closed time component due to desensitization. For analyses that included desensitized states, the closed duration for defining clusters was taken as the point of intersection of the closed time due to intermediate onset desensitization and the succeeding component determined by slow desensitization and the number of channels in the patch. Clusters of events with five or more openings and cluster mean open probabilities and cluster mean closed times within two standard deviations of the respective means were accepted for further analysis. Kinetic analysis was conducted simultaneously on data obtained across a range of ACh concentrations, termed global analysis, using MIL software (Qub suite, State University of New York, Buffalo), which corrects for missed events and gives error estimates of the fitted parameters (Qin et al., 1996). A dead time of 22 μs was imposed on all datasets. For each type of receptor studied, global analysis included data from two to four patches for each ACh concentration, ACh concentrations spaced at half-log unit intervals of the concentration, and a 10- to 333-fold range of the ACh concentration.
ACh Binding Measurements
The total number of [125I]-Btx binding sites on the cell surface of transfected BOSC 23 cells and ACh competition against the initial rate of [125I]-Btx were determined as previously described (Sine, 1993). ACh competition measurements were analyzed according to the following form of the Hill equation:
where Y is fractional occupancy by ACh, nH is the Hill coefficient, and Kapp is the apparent dissociation constant.
RESULTS
Triad of Conserved Residues at the Periphery of the ACh Binding Site
Structures of AChBP and our homology model of the nicotinic receptor ligand binding domain show that recognition domain C is strategically positioned to transmit local conformational change due to ACh binding to structures linked to the ion channel (Fig. 1 B). Domain C comprises the loop spanning -strands 9 and 10 of the subunit and links directly to the M1 transmembrane domain via -strand 10. It also overlays the Cys-loop, which articulates with the linker joining the M2 and M3 transmembrane domains (Kash et al., 2003; Bouzat et al., 2004; Unwin, 2005). Inspection of the AChBP structure, with and without carbamylcholine (CCh) bound, reveals two salt bridges that have counterparts in our structural model of the receptor ligand binding domain. At binding sites without CCh bound, a salt bridge forms between K139 and D194 of AChBP, whereas at binding sites with CCh bound, a salt bridge forms between K139 and Y185 (Celie et al., 2004; Fig. 1 C). In our structural model of the receptor ligand binding domain, residues equivalent to these, K145, D200, Y190, are close together in three-dimensional space and are conserved across all 1 subunits (Figs. 1, A and D), suggesting key structural or functional contributions. Thus the x-ray structure of AChBP, our structural model of the receptor ligand binding domain and residue conservation together suggest that K145, D200, and Y190 are candidates for transmitting local conformational change at the ACh binding site to structures linked to the ion channel.
Single Channel Kinetics of the Adult Human Receptor
To provide a frame of reference for assessing the effects of receptor mutations, we recorded an independent set of ACh-evoked single channel currents for the adult human receptor under experimental conditions identical to those used to study the mutant receptors. We coexpressed , , , and subunits in BOSC 23 cells and recorded single channel currents using the cell-attached configuration of the patch clamp with defined concentrations of ACh in the patch pipette (see MATERIALS AND METHODS). Currents recorded at desensitizing concentrations of ACh consist of a series of clusters of closely spaced current pulses separated by prolonged quiescent periods (Sakmann et al., 1980; Sine and Steinbach, 1987; Colquhoun and Ogden, 1988). Channel events within each cluster correspond to activation of a single receptor channel, and exhibit an ACh-dependent shortening of closed periods between successive openings, with little ACh dependence of the open periods (Fig. 2). The corresponding closed dwell time histograms show a shift from long to short time of a major component of closings as the ACh concentration is increased, whereas the major component of openings is concentration independent.
To estimate rate constants for elementary steps underlying activation of a single receptor channel, we analyzed the global set of dwell times according to kinetic schemes (Fig. 3). Scheme I depicts receptor activation as sequential binding of ACh to the resting, closed receptor, followed by opening and then block by ACh of the doubly occupied open receptor. Scheme II contains all of Scheme I but includes two desensitized states connected to the open state (Elenes and Auerbach, 2002; Lee and Sine, 2004). Scheme II is used to quantitatively account for the small fraction of closings due to fast and intermediate onset desensitization that are short enough to be indistinguishable from closings due to receptor activation. A third scheme depicting independent binding of ACh to each site, although able to describe data from the wild-type receptor (Hatton et al., 2003; Lee and Sine, 2004), is not applied here because it did not permit adequate definition of all the rate constants for the mutant receptors examined in this study.
Maximum likelihood fitting based on Scheme I yields probability density functions that describe the global set of closed and open dwell times for the adult human receptor (MATERIALS AND METHODS; Fig. 2). Rate constants underlying ACh binding, channel opening, and ACh blocking steps are all well defined by the analysis, and show rapid and efficient opening of the doubly occupied receptor, distinguishable association and dissociation rate constants for each binding site, and low affinity block of the open channel by ACh (Table I). The fitted rate constants are within the range of values previously reported for the adult human receptor (Ohno et al., 1996; Wang et al., 1999; Hatton et al., 2003; Lee and Sine, 2004).
Maximum likelihood fitting based on Scheme II, which incorporates brief and intermediate duration desensitized states, also describes the distributions of closed and open dwell times, and yields essentially the same rate constants for binding and gating steps obtained using Scheme I (Table I). The similar results obtained with Schemes I and II show that the presence of closings due to desensitization does not affect the estimates of rate constants for receptor activation when Scheme I is applied to the wild-type human receptor (Lee and Sine, 2004).
Functional Contribution of K145
K145 is a potential link between the binding site and the ion channel because it is proximal to the binding site, it is physically linked to the Cys-loop, it is proximal to electron-rich side chains that link to M1 via -strand 10, and it is conserved across 1 subunits. In the agonist-free structure of AChBP, the residue equivalent to K145, K139, forms a salt bridge with D194, whereas in the CCh-bound structure, K139 forms a salt bridge with the key binding site tyrosine, Y185 (Celie et al., 2004). To investigate the functional contribution of K145, we generated a series of mutations of K145, coexpressed each with complementary , , and subunits, and recorded ACh-evoked single channel currents.
Substitution of Ala for K145 profoundly impairs receptor activation at all concentrations of ACh, with activation episodes due to a single receptor consisting of predominantly single openings separated by prolonged closings (compare Figs. 2 and 4). The corresponding closed time histograms span a wide range of times at all ACh concentrations examined, and the major peak of closings shows an ACh-dependent shift toward shorter times. Maximum likelihood fitting of Scheme I to the global dataset for K145A well describes the closed and open dwell time histograms (Fig. 4). The fitted rate constants reveal nearly a 200-fold slowing of the channel opening rate constant and no change of the channel closing rate constant (Table I). For K145A, adequate definition of the rate constants underlying ACh binding steps required the approximation that the set of elementary association and dissociation rate constants are equal at each binding site. Presumably this approximation was required because information on the first binding step was limited: clusters of events corresponding to a single receptor could only be identified at ACh concentrations >30 μM, which may exceed the dissociation constant for ACh binding to the first binding site, and brief, mono-liganded openings were not present. Nevertheless, our analysis suggests that the rate constants underlying ACh binding are only modestly affected by K145A (Table I). Thus substitution of Ala for K145 at the periphery of the ACh binding site predominantly impairs elementary steps underlying channel gating.
Functional Consequences of K145Q
Because the Lys to Ala substitution represents a substantial structural change, we examined functional consequences of the more structurally conservative mutation K145Q. The Gln substitution neutralizes the positive charge but maintains polarity and moderate size of the side chain. Receptors containing K145Q show impaired channel gating similar to K145A, but individual activation episodes show three kinetic modes distinguished by low, intermediate, and high open probability (Popen; Fig. 5). These kinetic modes were observed only at high concentrations of ACh (300 μM to 1 mM), and although they were present at all high concentrations, the contributions of each mode varied from patch to patch. We therefore describe detection of the three kinetic modes for one patch to exemplify our treatment of the overall data for K145Q.
We identified clusters of events corresponding to a single channel using our standard cluster discrimination method (MATERIALS AND METHODS), and generated histograms of mean cluster Popen and mean cluster closed time (C) for K145Q (Fig. 5; Milone et al., 1998). The distribution of Popen shows a major and a minor peak, while the distribution of C shows a major peak and a secondary tail of long C. Analysis of cluster properties reveals that distinct portions of the two distributions are correlated: the peak of high Popen corresponds to clusters with short C (filled bars), and the tail with long C corresponds to clusters with low Popen (open bars). The remainder of the distributions consists of clusters with intermediate Popen and C (gray bars). Clusters from wild-type receptors subjected to the same analysis show a predominant mode of high Popen and short C.
We then separated clusters of events for each of the kinetic modes for K145Q receptors, and combined one mode at a time with the kinetically homogeneous data obtained at lower ACh concentrations. Maximum likelihood fitting showed that only clusters with intermediate Popen and C were compatible with data obtained at lower ACh concentrations. Fitting Scheme I to the intermediate mode dataset (30 μM to 1 mM ACh) shows that substitution of Gln for K145 profoundly impairs the efficiency of channel gating; the corresponding probability density functions adequately describe the distributions of closed and open dwell times (Fig. 6). The effects of K145Q on channel gating are similar to those of K145A, with the channel opening rate constant slowed 60-fold and the channel closing rate constant relatively unchanged (Table I). Adequate definition of rate constants underlying ACh binding steps again required the approximation that the set of elementary association and dissociation rate constants are equal at each binding site. Given this approximation, the analysis suggests K145Q has relatively little effect on agonist binding (Table I). Thus, located at the periphery of the ACh binding site, K145 contributes primarily to gating of the receptor channel, fulfilling an important criterion as a transduction element.
Functional Contribution of D200
D200 is a potential electrostatic partner of K145, and in the mouse receptor was found to contribute predominantly to channel gating (Akk et al., 1996). To determine whether D200 contributes to channel gating in the human muscle receptor, and further, whether its functional contribution mirrors that of K145, we recorded currents from adult human receptors harboring the mutation D200N. Like mutations of K145, D200N profoundly impairs receptor activation at all concentrations of ACh; individual clusters consist of predominantly single openings interspersed by long closings (Fig. 7). Maximum likelihood fitting according to Scheme I reveals a predominant effect on channel gating, with the channel opening rate constant slowed nearly 70-fold and the channel closing rate constant slightly increased (Table I). As observed for mutations of K145, adequate definition of rate constants underlying ACh binding steps required the approximation that the set of elementary association and dissociation rate constants are equal at each binding site. The rate constants underlying ACh binding are relatively unaffected by D200N, as observed for the same mutation in the mouse receptor (Table I). Thus the consequences of D200N for channel gating are highly similar between mouse and human receptors, and, moreover, mirror the consequences of mutations of K145. The similar effects of mutations of K145 and D200 on channel gating and proximity of the two side chains in our structural model collectively suggest these residues are interdependent and constitute part of an intermediate link between binding and gating domains.
Interdependence of K145 and D200
To gain additional evidence for interdependence of K145 and D200, we combined the mutations K145Q and D200N into a single subunit and recorded ACh-evoked single channel currents from the resulting double mutant receptors. The current traces again show impaired receptor activation at all ACh concentrations (Fig. 8), with predominantly single openings separated by long closings, similar to either single mutation (compare Figs. 6, 7, and 8). Unlike receptors containing K145Q alone, only a single kinetic mode was distinguished for the double mutant receptor, simplifying the analysis. Maximum likelihood fitting of Scheme I to the global set of dwell times reveals a 60-fold slowing of the channel opening rate constant and little change of the channel closing rate constant when compared with the wild-type receptor (Table I). These results are similar to those for either single mutation, showing qualitatively that K145Q and D200N are interdependent in contributing to channel gating. Double mutant cycles analysis (Horovitz and Fersht, 1990) applied to the set of gating equilibrium constants for wild-type, K145Q, D200N, and double mutant receptors yields a free energy of interaction of –2.6 kcal/mol. Rate constants underlying ACh binding steps are altered twofold or less by the double mutation, similar to D200N alone, but this conclusion remains provisional because the assumption of equivalent binding sites was required to obtain well-defined rate constants. Thus, the largely selective perturbation of channel gating by the double mutation quantitatively mirrors that for either single mutation, further suggesting K145 and D200 are interdependent in coupling agonist binding to channel gating.
Charge Reversal of K145 and D200
To test for electrostatic interaction between K145 and D200, we generated the single mutants K145E and D200K and the corresponding double mutant and recorded ACh-evoked single channel currents from the resulting mutant receptors (Fig. 9). The single mutation D200K slows the channel opening rate constant 400-fold and increases the channel closing rate constant nearly fourfold, yielding a 1,400-fold decrease of the channel gating equilibrium constant (Table I). For D200K, well-defined single receptor clusters could not be discerned at ACh concentrations <1 mM, preventing estimation of rate constants underlying ACh binding. Consequences of the single mutation K145E mirror those of other mutations of K145; compared with the wild-type receptor, the channel opening rate constant slows 100-fold, the channel closing rate constant slows almost twofold, and the ACh dissociation rate constants are only modestly affected. The rate constants for ACh association, on the other hand, slow three- to fivefold compared with other mutations of K145 (Table I).
Compared with D200K alone, restoring the local negative charge by combining K145E markedly increases the frequency of channel openings within single receptor clusters (Fig. 9), and counteracts the slowing of the channel opening rate constant and the increase of the channel closing rate constant (Table I). The extent to which the gating rate constants are restored is only partial, perhaps because residues flanking each member of the putative salt bridge are not compatible with the reversed charge. The net result of charge reversal is a 14-fold increase of the channel gating equilibrium constant compared with D200K alone. Double mutant cycles analysis applied to the set of four gating equilibrium constants yields a free energy of interaction of –4 kcal/mol. Further, the slow association of ACh caused by K145E alone is restored to the range of values observed for other mutations of K145 (Table I).
Finally, we examined the single mutation K145D and found that it prevented expression of functional channels and cell-surface -Btx binding sites. However, K145D is rescued by the charge reversal mutation D200K. Full kinetic analysis of the double mutation reveals an eightfold increase of the channel gating equilibrium constant compared with D200K alone, due to a twofold increase of the channel opening rate constant and a fourfold slowing of the channel closing rate constant (Table I). Changes in rate constants underlying ACh binding remain unknown because these could not be measured for either single mutation. However, the ACh association rate constants for the double mutation increase twofold compared with K145E alone, while the ACh dissociation rate constants slow twofold compared with K145E. The overall results from charge-neutralization and charge-reversal experiments provide strong evidence that K145 and D200 interact through electrostatic forces to couple agonist binding to channel gating.
Functional Contributions of Y190 and T202
Our structural model predicts that two more residues are proximal to K145 and D200: Y190 and T202 (Fig. 1). We therefore generated structurally conservative mutations of each residue, Y190F and T202V, and recorded single channel currents from the resulting mutant receptors. For receptors containing Y190F, the ACh concentration had to be increased to 3 mM to obtain well-defined clusters of single receptor currents; the corresponding intracluster events consist of single brief openings separated by prolonged closings (Fig. 10). The availability of data at only a single high concentration of ACh limited kinetic fitting to the subset of doubly occupied gating steps in Scheme I, but nevertheless reveals a 1,000-fold slowing of the channel opening rate constant with little change in the channel closing rate constant (Table I). These effects of Y190F on channel gating are analogous to those observed for the same mutation in the fetal mouse muscle receptor (Chen et al., 1995), but the 800-fold decrease of the gating equilibrium constant for the adult human receptor exceeds the 200-fold decrease observed for the fetal mouse receptor.
The kinetics of receptor activation are only modestly affected by T202V. Rate constants for elementary binding or gating steps change twofold or less compared with the wild-type receptor (Table I). Thus despite its conservation across 1 subunits and proximity to D200 and K145, T202 does not contribute substantially to either ACh binding or channel gating.
Interdependence of K145 and Y190
The agonist-bound conformation of AChBP shows that K139 and Y185 form a salt bridge (Celie et al., 2004). To determine whether residues in the receptor equivalent to these, K145 and Y190, are interdependent, we combined K145Q and Y190F into a single subunit and recorded ACh-evoked single channel currents from the resulting double mutant receptors. The current traces again show impaired receptor activation at 1 mM ACh, with brief openings separated by long closings, similar to either single mutation (Fig. 10). Unlike receptors containing K145Q alone, only a single kinetic mode could be distinguished for the double mutant receptor. Further, well-defined clusters of single receptor openings could only be discerned at ACh concentrations from 100 μM to 1 mM, restricting the analysis to the channel gating rate constants. Maximum likelihood fitting shows that compared with the wild-type receptor, the double mutation slows the channel opening rate constant 200-fold without affecting the channel closing rate constant (Table I). Double mutant cycles analysis applied to the set of gating equilibrium constants for wild-type, K145Q, Y190F, and double mutant receptors yields a free energy of interaction of –2.9 kcal/mol. Thus K145Q and Y190F are interdependent in coupling agonist binding to channel gating.
Pairwise Interdependence Assessed by Steady-state Binding of ACh
Most mutations that affect receptor function alter steady-state binding of ACh (examples include Chen et al., 1995; Ohno et al., 1996; Sine et al., 2002b). Although measurements of steady-state ACh binding contain contributions from closed, open, and desensitized states of the receptor, each of which differs in affinity for the agonist (Katz and Thesleff, 1957; Sine and Taylor, 1979; Jackson, 1989), binding measurements can be used as an overall index of receptor function. We therefore sought further evidence for pairwise interdependence by measuring steady-state binding of ACh to receptors containing single and double mutations. We first studied the charge neutral mutations K145Q and D200N and the corresponding double mutation. Relative to the wild-type receptor, the single mutations K145Q and D200N increase the apparent dissociation constant for ACh (Kapp) 10- and 26-fold, respectively, qualitatively consistent with expectations from impaired channel gating observed by single channel recording (Fig. 11; Table II). The double mutation shows a virtually identical binding profile to that of D200N alone, and double mutant cycles analysis applied to Kapp yields an interaction free energy of 1 kcal/mol for this charge neutralization experiment. Like Kapp, interaction free energy based on Kapp arises from multiple processes: ACh binding, desensitization, and channel gating. One of these processes is channel gating, which contributes 2.6 kcal/mol, but the other processes likely contribute. We next measured steady-state ACh binding for the charge reversal mutations K145E and D200K and the corresponding double mutation. The single mutations K145E and D200K increase Kapp 100- and 860-fold, respectively, again qualitatively consistent with profoundly impaired channel gating (Fig. 11; Table II). However, charge reversal by the double mutation strongly counteracts the effects of either single mutation, yielding an overall 20-fold increase of Kapp compared with the wild-type receptor. Double mutant cycles analysis applied to Kapp for this charge reversal experiment reveals an interaction free energy of 4.8 kcal/mol, which is similar to the contribution of 4 kcal/mol observed for channel gating alone. The mutation K145D does not produce functional receptors, but when combined with D200K, yields a Kapp coincident with that obtained for K145E plus D200K (Fig. 11; Table II).
The double mutation containing K145Q and Y190F failed to bind -Btx, preventing assessment of interdependence by steady-state ACh binding. Nevertheless the overall results provide strong evidence for electrostatic interaction between K145 and D200 and interdependence of K145 and Y190. A three-way electrostatic interplay of charged residues thus emerges as a compelling candidate for an initial link between ligand binding and channel gating domains owing to the proximity of the triad to the ACh binding site, interdependence of residues in the triad, connections to the Cys-loop and M1 domain, and the predominant contributions of all three residues to channel gating.
Subunit Specificity of the Mutations
The three residues studied here are unique to and conserved across 1 subunits. However, subunits also contain conserved Asp and Lys at positions equivalent to D200 and K145, respectively. To determine whether this residue pair in the subunit contributes to receptor function, we generated the mutations K147Q and D214N, coexpressed each with complementary , , and subunits, and measured steady-state ACh binding. Neither mutation affects Kapp for ACh binding (Table II), showing that if these two residues form a salt bridge, it is unlikely to contribute to receptor function.
DISCUSSION
We show that three conserved residues at the periphery of the ACh binding site, K145, D200, and Y190, contribute profoundly to gating of the receptor channel. One of these residues, Y190, contributes to the aromatic cage that coordinates the quaternary ammonium moiety of the agonist (Celie et al., 2004), and its hydroxyl group projects away from the binding site toward the other two members of the triad (Fig. 1). Of the other two residues, K145 is the most likely contact for Y190, but this contact corresponds to the agonist-bound conformation. Evidence for this contact comes from the AChBP crystal structure with CCh bound, which shows a salt bridge between the equivalent residues K139 and Y185 (Celie et al., 2004), and from the interdependent contributions of K145 and Y190 to gating shown here. Support for electrostatic contact between K145 and D200 comes from the AChBP crystal structure without bound agonist, which shows a salt bridge between the equivalent Lys and Asp residues (Brejc et al., 2001), and from our data demonstrating interdependence of K145 and D200 in contributing to channel gating.
Our findings combine with inter-atomic distances in AChBP to suggest how interplay among these three residues couples local conformational change due to agonist binding to gating of the channel. Our inferences draw from two sets of atomic coordinates of AChBP with bound CCh deposited in the protein data bank (PDB code 1UV6): a pentameric unit with one molecule of CCh bound and another unit with two molecules of CCh bound. Inspection of the three subunit interfaces to which CCh is bound shows that recognition domain C is drawn inward toward the center of the binding site, whereas inspection of one of the unoccupied interfaces shows that domain C is extended outward from the site. Molecular dynamics simulation of AChBP and measurements of intrinsic tryptophan fluorescence reveal similar changes in the conformations of domain C with and without bound agonist (Gao et al., 2005). In the following mechanistic interpretation, we refer to these conformations as agonist-free and agonist-bound.
To illustrate our structural interpretation, the agonist-free and -bound conformations of recognition domain C of AChBP are compared (Fig. 12). The agonist-bound conformation is that of the crystal structure of AChBP with bound CCh (Celie et al., 2004). The agonist-free conformation is that generated by prolonged molecular dynamics simulation of AChBP (Gao et al., 2005), which is similar to the extended conformation of a subunit of AChBP at an unoccupied interface (see PDB entry 1UV6). ACh is shown docked in the orientation obtained by computational docking to the HEPES-bound crystal structure of AChBP (Brejc et al., 2001; Gao et al., 2005). In the agonist-free conformation, the Tyr at the position equivalent to Y190 is 8 away from the Lys equivalent to K145, while residues equivalent to K145 and D200 form a salt bridge. In the agonist-bound conformation, the Tyr equivalent to Y190 moves to within 2–3 of the Lys equivalent to K145, drawing recognition domain C inward, and allowing the residue equivalent to D200 to move away from the Tyr/Lys pair. Thus by analogy to AChBP, in the resting state of the receptor, K145 and D200 are proposed to pair through electrostatic forces and Y190 is out of register (Fig. 11). However when the agonist binds, Y190 is drawn toward the K145/D200 pair and, through electrostatic forces, pulls K145 away from D200, allowing both residues to relax to new positions. These local inter-residue displacements could propagate to the channel via -strands 7 and 10 of the ligand binding domain. -Strand 7 harbors K145 and forms part of the signature Cys-loop, which interacts with the linker spanning the M2 and M3 transmembrane domains (Kash et al., 2003; Bouzat et al., 2004), while -strand 10 contains D200 and links directly to the M1 transmembrane domain.
Although all three residues of the triad contribute profoundly to channel gating, each has a distinct role in linking the binding site to the channel. Y190 is the clear link to the ACh binding site, as shown by site-directed labeling (Abramson et al., 1989) and its location in the aromatic cage that coordinates the agonist (Celie et al., 2004). Its function is to help draw recognition domain C toward the center of the binding site when the agonist is bound and, in the process, form a salt bridge with K145. Accordingly, removal of the aromatic hydroxyl group by mutation to Phe would prevent formation of the salt bridge and disrupt coupling of binding to gating. Likewise, substitution of Gln, Ala, or Glu for K145 would remove the electrostatic partner for Y190, also disrupting coupling. By analogy to AChBP in the agonist-free conformation, D200 is proposed to form a salt bridge with K145, positioning K145 for contact with the incoming Y190 and maintaining -strands 7 and 10 in their closed state conformations. When the agonist binds, Y190 draws K145 away from D200, allowing -strands 7 and 10 to move to their active conformations. Substitution of Asn for D200 would prevent stabilization of K145 for the incoming Y190, and break the link between -strand 10 and the binding site, disrupting coupling.
One of the main lines of evidence that the triad identified here is an initial link in the binding to gating cascade is that mutation of each residue profoundly impairs gating of the channel. However two additional observations require explanation. First, the allosteric link between binding and gating domains is not completely disrupted by any of the mutations, as ACh can still trigger opening of the channel, although with greatly reduced efficiency. The most likely explanation is there are additional allosteric links that remain intact in each mutant receptor and account for the residual ACh-evoked currents. Second, the extent to which gating is disrupted differs by as much as 10-fold among the three mutations. If each member of the triad is essential for coupling, the extent to which gating is disrupted should be the same following mutation of each member. However in addition to coupling binding to gating, the triad may also contribute to the allosteric equilibrium between resting and open channel states in the absence of agonist (Jackson, 1989). This contribution to the allosteric set point likely involves dynamic interplay among the three residues, such that changes in the set point differ when one of the members is structurally altered. Thus for each mutation, the gating equilibrium for di-liganded receptors could be altered by differing amounts, owing to different contributions to the allosteric set point.
Other residues that contribute to the allosteric set point have been identified in the ligand binding domains of members of the Cys-loop receptor superfamily. In the nicotinic receptor 1 subunit, mutation of Asp 97 markedly increases the frequency of channel opening events in the absence of ACh (Chakrapani et al., 2003). In the GABAA receptor 1 subunit, mutation of Tyr 107 increases the resting membrane conductance in a manner that was sensitive to the specific GABA receptor inhibitor picrotoxin (Torres and Weiss, 2002). Similarly in the GABAA receptor 2 subunit, mutation of Glu 155 increases the resting conductance in a picrotoxin-sensitive manner (Newell et al., 2004). In all three cases, the allosteric set point was increased in favor of the open channel state, whereas mutation of any residue of the triad in the present study impedes opening of the channel. Thus among functionally key residues in the ligand binding domain of the receptor, a balance is likely established between residues that promote and those that impair channel gating.
The triad of residues identified here is found in all species of nicotinic receptor 1 subunits, suggesting they contribute to agonist binding transduction in all muscle type nicotinic receptors (for subunit sequences see Le Novere and Changeux, 2001). Similarly, the triad is also present in all neuronal nicotinic subunits (2–10), except 5, which contains Asp at the position equivalent to Y190, and 10, which contains Thr at the position equivalent to K145. However, neither of these exceptions likely forms the principal face of the binding site; 5 appears to be a structural subunit, requiring both neuronal and subunits to form functional receptors (Ramirez-Latorre et al., 1996), and 10 requires coexpression of the 9 subunit (Sgard et al., 2002). On the other hand, the triad is not present in GABAA, glycine, and 5-HT receptors. Thus among the Cys-loop superfamily of receptors, the initial coupling mechanism is likely to be highly similar for muscle and neuronal nicotinic receptors, but the more evolutionary distant receptors evolved different structures to relay the initial consequences of agonist binding. This structural divergence among receptors activated by different neurotransmitters is not surprising because the different agonist structures inherently require different chemical counterparts in the binding site.
In summary, our observations lead to identification of a triad of conserved residues as an initial link that couples agonist binding to channel gating. All three residues of the triad localize to the same three-dimensional region of our structural model of the receptor ligand binding domain, and all three residues are conserved across 1 subunits. Also, equivalent residues in AChBP change conformation when the agonist is bound. Finally, structurally conservative mutations of each residue predominantly impair channel gating, and two pairs within the triad, K145/D200 and K145/Y190, are interdependent in contributing to gating. Our findings suggest future studies aimed at identifying additional links that couple agonist binding to channel gating.
ACKNOWLEDGMENTS
We thank Dr. Hai-Long Wang for the homology model and Fan Gao for the coordinates used to make Fig. 12.
This work was supported by National Institutes of Health grant NS31744 to S.M. Sine.
Olaf S. Andersen served as editor.
Submitted: 9 March 2005
Accepted: 5 May 2005
REFERENCES
Abramson, S.N., Y. Li, P. Culver, and P. Taylor. 1989. An analog of lophotoxin reacts covalently with tyrosine 190 in the subunit of the nicotinic receptor. J. Biol. Chem. 264:12666–12672.
Akk, G., S. Sine, and A. Auerbach. 1996. Binding sites contribute unequally to the gating of mouse nicotinic D200N acetylcholine receptors. J. Physiol. 496:185–196.
Bouzat, C., F. Gumilar, G. Spitzmaul, H.-L. Wang, D. Rayes, S.B. Hansen, P. Taylor, and S.M. Sine. 2004. Coupling of agonist binding to channel gating in an ACh-binding protein linked to an ion channel. Nature. 430:896–900.
Brejc, K., W. van Dijk, R. Klassen, M. Schuurmans, J. van der Oost, A. Smit, and T. Sixma. 2001. Crystal structure of an ACh-binding protein reveals the ligand-binding domain of nicotinic receptors. Nature. 411:269–276.
Celie, P., S. van Rossum-Fikkert, W. van Dijk, K. Brejc, A. Smit, and T. Sixma. 2004. Nicotine and carbamylcholine binding to nicotinic acetylcholine receptors as studied in AChBP crystal structures. Neuron. 41:907–914.
Chakrapani, S., T.D. Bailey, and A. Auerbach. 2003. The role of loop 5 in acetylcholine receptor channel gating. J. Gen. Physiol. 122:521–539.
Chakrapani, S., T.D. Bailey, and A. Auerbach. 2004. Gating dynamics of the acetylcholine receptor extracellular domain. J. Gen. Physiol. 123:341–356.
Chen, J., Y. Zhang, G. Akk, S. Sine, and A. Auerbach. 1995. Activation kinetics of recombinant mouse nicotinic acetylcholine receptors with mutations at subunit residue tyrosine 190. Biophys. J. 69:849–859.
Colquhoun, D., and D.C. Ogden. 1988. Activation of ion channels in the frog end-plate by high concentrations of acetylcholine. J. Physiol. 395:131–159.
Colquhoun, D., and F. Sigworth. 1983. Fitting and statistical analysis of single channel records. Single Channel Recording. B. Sakmann and E. Neher, editors. Plenum Publishing Corp., New York. 191–264.
Corringer, J.P., N. LeNovere, and J.P. Changeux. 2000. Nicotinic receptors at the amino acid level. Annu. Rev. of Toxicol. 40:431–458.
Elenes, S., and A. Auerbach. 2002. Desensitization of diliganded mouse muscle nicotinic acetylcholine receptor channels. J. Physiol. 541:367–383.
Gao, F., T. Burghardt, N. Bren, S.B. Hansen, R. Henchman, P. Taylor, J.A. McCammon, and S.M. Sine. 2005. Acetylcholine-mediated conformational changes in acetylcholine-binding protein revealed by simulation and intrinsic tryptophan fluorescence. J. Biol. Chem. 280:8443–8451.
Hamill, O.P., A. Marty, E. Neher, B. Sakmann, and F.J. Sigworth. 1981. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch. 391:85–100.
Hatton, C.J., C. Shelley, M. Brydson, D. Beeson, and D. Colquhoun. 2003. Properties of the human muscle nicotinic receptor, and of the slow-channel myasthenic syndrome mutant L221F, inferred from maximum likelihood fits. J. Physiol. 547:729–760.
Horovitz, A., and A.R. Fersht. 1990. Strategy for analysing the co-operativity of intramolecular interactions in peptides and proteins. J. Mol. Biol. 214:613–617.
Jackson, M.B. 1989. Perfection of a synaptic receptor: kinetics and energetics of the acetylcholine receptor. Proc. Natl. Acad. Sci. USA. 86:2199–2203.
Karlin, A. 2002. The emerging structure of the nicotinic acetylcholine receptors. Nat. Rev. Neurosci. 3:102–114.
Kash, T., A. Jenkins, J. Kelly, J. Trudell, and N.L. Harrison. 2003. Coupling of agonist binding to channel gating in the GABAA receptor. Nature. 421:272–275.
Katz, B., and S. Thesleff. 1957. A study of desensitization produced by acetylcholine at the motor end-plate. J. Physiol. 138:63–80.
Le Novere, N., and J.P. Changeux. 2001. LGICdb: the ligand-gated ion channel database. Nucleic Acids Res. 29:294–295.
Le Novere, N., T. Grutter, and J.P. Changeux. 2002. Models of the extra-cellular domain of the nicotinic receptors and of agonist- and Ca2+-binding sites. Proc. Natl. Acad. Sci. USA. 99:3210–3215.
Lee, B.S., R.B. Gunn, and R.R. Kopito. 1991. Functional differences among nonerythroid anion exchangers expressed in a transfected human cell line. J. Biol. Chem. 266:11448–11454.
Lee, W.Y., and S.M. Sine. 2004. Invariant aspartic acid in muscle nicotinic receptor contributes selectively to the kinetics of agonist binding. J. Gen. Physiol. 124:555–567.
Milone, M., H.L. Wang, K. Ohno, R. Prince, T. Fukudome, X.M. Shen, J.M. Brengman, R.C. Griggs, S.M. Sine, and A.G. Engel. 1998. Mode switching kinetics produced by a naturally occurring mutation in the cytoplasmic loop of the human acetylcholine receptor epsilon subunit. Neuron. 20:575–588.
Miyazawa, A., Y. Fujiyoshi, and N. Unwin. 2003. Structure and gating mechanism of the acetylcholine receptor pore. Nature. 423:949–955.
Molles, B.E., I. Tsigelny, P. Nguyen, S. Gao, S.M. Sine, and P. Taylor. 2002. Residues in the subunit of the nicotinic acetylcholine receptor interact to confer selectivity of Waglerin-1 for the – subunit interface site. Biochemistry. 41:7895–7906.
Newell, J.G., R.A. McDevitt, and C. Czajkowski. 2004. Mutation of glutamate 155 of the GABA(A) receptor 2 subunit produces a spontaneously open channel: a trigger for channel activation. J. Neurosci. 24:11226–11235.
Ohno, K., H.-L. Wang, M. Milone, N. Bren, J.M. Brengman, S. Nakano, P. Quiram, J.N. Pruitt, S.M. Sine, and E.G. Engel. 1996. Congential myasthenic syndrome caused by decreased agonist binding affinity due to a mutation in the acetylcholine receptor subunit. Neuron. 17:157–170.
Pear, W.S., G.P. Nolan, M.L. Scott, and D. Baltimore. 1993. Production of high-titer helper-free retroviruses by transient transfection. Proc. Natl. Acad. Sci. USA. 90:8392–8396.
Schapira, M., R. Abagyan, and M. Totrov. 2002. Structural model of nicotinic acetylcholine receptor isotypes bound to acetylcholine and nicotine. BMC Struct. Biol. 2:1.
Qin, F., A. Auerbach, and F. Sachs. 1996. Estimating single channel kinetic parameters from idealized patch clamp data containing missed events. Biophys. J. 70:264–280.
Ramirez-Latorre, J., C.R. Yu, X. Qu, F. Perin, A. Karlin, and L. Role. 1996. Functional contributions of -5 subunit to neuronal receptors. Nature. 380:347–351.
Sakmann, B., J. Patlak, and E. Neher. 1980. Single acetylcholine-activated channels show burst-kinetics in presence of desensitizing concentrations of agonist. Nature. 286:71–73.
Sgard, F., E. Charpantier, S. Bertrand, N. Walker, D. Caput, D. Graham, D. Bertrand, and F. Besnard. 2002. A novel human nicotinic receptor subunit, 10, that confers functionality to the 9-subunit. Mol. Pharmacol. 61:150–159.
Sine, S.M., and J.H. Steinbach. 1987. Activation of acetylcholine receptors in clonal BC3H-1 cells by high concentrations of agonist. J. Physiol. 385:325–359.
Sine, S.M. 1993. Molecular dissection of subunit interfaces in the acetylcholine receptor: identification of residues that determine curare selectivity. Proc. Natl. Acad. Sci. USA. 90:9436–9440.
Sine, S.M., and P. Taylor. 1979. Functional consequences of agonist-mediated state transitions in the cholinergic receptor. J. Biol. Chem. 254:3315–3325.
Sine, S.M. 2002. The nicotinic receptor ligand binding domain. J. Neurobiol. 53:431–446.
Sine, S.M., H.L. Wang, and N. Bren. 2002a. Lysine scanning mutagenesis delineates structural model of the nicotinic receptor ligand binding domain. J. Biol. Chem. 277:29210–29223.
Sine, S.M., X.M. Shen, H.L. Wang, K. Ohno, W.Y. Lee, A. Tsujino, J. Brengmann, N. Bren, J. Vajsar, and A.G. Engel. 2002b. Naturally occurring mutations at the acetylcholine receptor binding site independently alter ACh binding and channel gating. J. Gen. Physiol. 120:483–496.
Torres, V.I., and D.S. Weiss. 2002. Identification of a tyrosine in the agonist binding site of the homomeric rho1 -aminobutyric acid (GABA) receptor that, when mutated, produces spontaneous opening. J. Biol. Chem. 277:43741–43748.
Unwin, N. 2005. Refined structure of the nicotinic acetylcholine receptor at 4 resolution. J. Mol. Biol. 346:967–989.
Wang, H.-L., M. Milone, K. Ohno, X.-M. Shen, A. Tsujuno, A. Paola, P. Tonali, J. Brengman, A.G. Engel, and S.M. Sine. 1999. Acetylcholine receptor M3 domain: stereochemical and volume contributions to channel gating. Nat. Neurosci. 2:226–233.
Wang, H.-L., K. Ohno, M. Milone, J.M. Brengman, A. Evoli, A.P. Batocchi, L.T. Middleton, K. Christodoulou, A.G. Engel, and S.M. Sine. 2000. Fundamental gating mechanism of nicotinic receptor channel revealed by mutation causing a congenital myasthenic syndrome. J. Gen. Physiol. 116:449–462.(Nuriya Mukhtasimova, Chri)
We examined functional consequences of intrasubunit contacts in the nicotinic receptor subunit using single channel kinetic analysis, site-directed mutagenesis, and structural modeling. At the periphery of the ACh binding site, our structural model shows that side chains of the conserved residues K145, D200, and Y190 converge to form putative electrostatic interactions. Structurally conservative mutations of each residue profoundly impair gating of the receptor channel, primarily by slowing the rate of channel opening. The combined mutations D200N and K145Q impair channel gating to the same extent as either single mutation, while K145E counteracts the impaired gating due to D200K, further suggesting electrostatic interaction between these residues. Interpreted in light of the crystal structure of acetylcholine binding protein (AChBP) with bound carbamylcholine (CCh), the results suggest in the absence of ACh, K145 and D200 form a salt bridge associated with the closed state of the channel. When ACh binds, Y190 moves toward the center of the binding cleft to stabilize the agonist, and its aromatic hydroxyl group approaches K145, which in turn loosens its contact with D200. The positional changes of K145 and D200 are proposed to initiate the cascade of perturbations that opens the receptor channel: the first perturbation is of -strand 7, which harbors K145 and is part of the signature Cys-loop, and the second is of -strand 10, which harbors D200 and connects to the M1 domain. Thus, interplay between these three conserved residues relays the initial conformational change from the ACh binding site toward the ion channel.
Abbreviations used in this paper: AChBP, acetylcholine binding protein; Btx, bungarotoxin; CCh, carbamylcholine.
INTRODUCTION
Activation of post-synaptic receptors is accomplished by allosteric communication between the neurotransmitter binding site and the distal ion channel. The local conformational change due to neurotransmitter binding is amplified in a cascade that culminates in the global conformational change that opens the channel. For the Cys-loop superfamily of post-synaptic receptors, atomic-scale insight into the neurotransmitter binding site emerged from the x-ray structure of the homologous acetylcholine binding protein (AChBP; Brejc et al., 2001) and from homology models derived from it (Le Novere et al., 2002; Molles et al., 2002; Schapira et al., 2002; Sine et al., 2002a). Additionally a 4- resolution structural model of the binding, channel, and cytoplasmic domains was generated from electron microscopy of two-dimensional arrays of Torpedo receptors (Miyazawa et al., 2003; Unwin, 2005). At the interface between binding and pore domains, a network of loops was shown to couple neurotransmitter binding to channel gating (Kash et al., 2003; Bouzat et al., 2004; Chakrapani et al., 2004), but coupling structures near the neurotransmitter binding site remain unknown.
In the adult muscle nicotinic receptor, the two ACh binding sites are formed at interfaces between an subunit and a neighboring or subunit where multiple recognition domains from each subunit converge (for reviews see Corringer et al., 2000; Karlin, 2002; Sine, 2002). The subunit contributes recognition domains A–C, while the non- subunit contributes domains D–G. Conserved aromatic residues from domains A–D form an aromatic cage that coordinates the positively charged agonist (Celie et al., 2004), whereas nonconserved residues in domains D–G endow the two binding sites with selectivity for agonists and competitive antagonists (Sine, 2002). Domains C and F are the most peripheral of the recognition domains, and of these, domain C exhibits the greatest displacement on binding the agonist (Celie et al., 2004; Gao et al., 2005).
To identify structures that propagate the local conformational change elicited by ACh away from the binding site, we examined our homology model of the receptor ligand binding domain (Sine et al., 2002a). We looked for conserved residues at the periphery of the binding site near the mobile domain C and found a cluster of three conserved residues, K145, D200, and Y190, whose side chains potentially interact through electrostatic forces. Our patch clamp recordings show that structurally conservative mutations of each residue profoundly impair gating of the channel. The common contributions to channel gating and direct linkages to the binding site and binding-pore interface suggest interplay among these three residues initiates the allosteric cascade from the binding site to the channel gate.
MATERIALS AND METHODS
Construction of Wild-type and Mutant AChRs
Human , , , and subunit cDNAs were obtained as previously described (Ohno et al., 1996) and subcloned into the mammalian expression vector pRBG4 (Lee et al., 1991). Site-directed mutations were introduced using the Quick-Change site-directed mutagenesis kit (Stratagene). The presence of each mutation and the absence of unwanted mutation were determined by sequencing the entire cDNA insert.
Mammalian Cell Expression
BOSC 23 cells (Pear et al., 1993; Wang et al., 2000), a variant of HEK 293 clonal fibroblasts, were transfected with mutant or wild-type cDNAs using calcium phosphate precipitation as previously described (Lee and Sine, 2004). Patch clamp and [125I]-bungarotoxin (Btx) binding measurements were made 2 and 3 d after transfection, respectively.
Patch-clamp Recordings
Recordings were obtained in the cell-attached configuration (Hamill et al., 1981) at a membrane potential of –70 mV and a temperature of 21°C. Bath and pipette solutions contained 142 mM KCl, 5.4 mM NaCl, 1.8 mM CaCl2, 1.7 mM MgCl2, 10 mM HEPES (pH 7.4). Single channel currents were recorded using an Axopatch 200B (Axon Instruments Inc.). Data were obtained from two to four different patches for each ACh concentration. Recordings were accepted for analysis only when the frequency of occurrence of clusters of events was low enough to be sure they originated from a single receptor channel. Currents were low pass filtered at 100 kHz and recorded to hard disk at 200 kHz using the program Acquire (Bruxton Co.), using the option that allows high bandwidth sampling of periods of channel activity while keeping track of the intervening quiescent periods.
Single-channel Kinetic Analysis
Digitized current signals were analyzed as recently described (Lee and Sine, 2004) using a 10-kHz digital Gaussian filter, cubic spline interpolation of the signal, a dead time of 10 μs, the half-amplitude threshold for detection and correction of the measured dwell times at threshold for the effects of the Gaussian filter (Colquhoun and Sigworth, 1983). Dwell time histograms were constructed using a logarithmic abscissa and square root ordinate and fitted to the sum of exponential components by maximum likelihood using the program TACFit (Bruxton Co.). Clusters of events corresponding to a single channel were identified as a series of closely spaced openings preceded and followed by closed intervals greater than a specified duration. This duration was taken as the point of intersection of the closed time component that depended on ACh concentration with the succeeding closed time component due to desensitization. For analyses that included desensitized states, the closed duration for defining clusters was taken as the point of intersection of the closed time due to intermediate onset desensitization and the succeeding component determined by slow desensitization and the number of channels in the patch. Clusters of events with five or more openings and cluster mean open probabilities and cluster mean closed times within two standard deviations of the respective means were accepted for further analysis. Kinetic analysis was conducted simultaneously on data obtained across a range of ACh concentrations, termed global analysis, using MIL software (Qub suite, State University of New York, Buffalo), which corrects for missed events and gives error estimates of the fitted parameters (Qin et al., 1996). A dead time of 22 μs was imposed on all datasets. For each type of receptor studied, global analysis included data from two to four patches for each ACh concentration, ACh concentrations spaced at half-log unit intervals of the concentration, and a 10- to 333-fold range of the ACh concentration.
ACh Binding Measurements
The total number of [125I]-Btx binding sites on the cell surface of transfected BOSC 23 cells and ACh competition against the initial rate of [125I]-Btx were determined as previously described (Sine, 1993). ACh competition measurements were analyzed according to the following form of the Hill equation:
where Y is fractional occupancy by ACh, nH is the Hill coefficient, and Kapp is the apparent dissociation constant.
RESULTS
Triad of Conserved Residues at the Periphery of the ACh Binding Site
Structures of AChBP and our homology model of the nicotinic receptor ligand binding domain show that recognition domain C is strategically positioned to transmit local conformational change due to ACh binding to structures linked to the ion channel (Fig. 1 B). Domain C comprises the loop spanning -strands 9 and 10 of the subunit and links directly to the M1 transmembrane domain via -strand 10. It also overlays the Cys-loop, which articulates with the linker joining the M2 and M3 transmembrane domains (Kash et al., 2003; Bouzat et al., 2004; Unwin, 2005). Inspection of the AChBP structure, with and without carbamylcholine (CCh) bound, reveals two salt bridges that have counterparts in our structural model of the receptor ligand binding domain. At binding sites without CCh bound, a salt bridge forms between K139 and D194 of AChBP, whereas at binding sites with CCh bound, a salt bridge forms between K139 and Y185 (Celie et al., 2004; Fig. 1 C). In our structural model of the receptor ligand binding domain, residues equivalent to these, K145, D200, Y190, are close together in three-dimensional space and are conserved across all 1 subunits (Figs. 1, A and D), suggesting key structural or functional contributions. Thus the x-ray structure of AChBP, our structural model of the receptor ligand binding domain and residue conservation together suggest that K145, D200, and Y190 are candidates for transmitting local conformational change at the ACh binding site to structures linked to the ion channel.
Single Channel Kinetics of the Adult Human Receptor
To provide a frame of reference for assessing the effects of receptor mutations, we recorded an independent set of ACh-evoked single channel currents for the adult human receptor under experimental conditions identical to those used to study the mutant receptors. We coexpressed , , , and subunits in BOSC 23 cells and recorded single channel currents using the cell-attached configuration of the patch clamp with defined concentrations of ACh in the patch pipette (see MATERIALS AND METHODS). Currents recorded at desensitizing concentrations of ACh consist of a series of clusters of closely spaced current pulses separated by prolonged quiescent periods (Sakmann et al., 1980; Sine and Steinbach, 1987; Colquhoun and Ogden, 1988). Channel events within each cluster correspond to activation of a single receptor channel, and exhibit an ACh-dependent shortening of closed periods between successive openings, with little ACh dependence of the open periods (Fig. 2). The corresponding closed dwell time histograms show a shift from long to short time of a major component of closings as the ACh concentration is increased, whereas the major component of openings is concentration independent.
To estimate rate constants for elementary steps underlying activation of a single receptor channel, we analyzed the global set of dwell times according to kinetic schemes (Fig. 3). Scheme I depicts receptor activation as sequential binding of ACh to the resting, closed receptor, followed by opening and then block by ACh of the doubly occupied open receptor. Scheme II contains all of Scheme I but includes two desensitized states connected to the open state (Elenes and Auerbach, 2002; Lee and Sine, 2004). Scheme II is used to quantitatively account for the small fraction of closings due to fast and intermediate onset desensitization that are short enough to be indistinguishable from closings due to receptor activation. A third scheme depicting independent binding of ACh to each site, although able to describe data from the wild-type receptor (Hatton et al., 2003; Lee and Sine, 2004), is not applied here because it did not permit adequate definition of all the rate constants for the mutant receptors examined in this study.
Maximum likelihood fitting based on Scheme I yields probability density functions that describe the global set of closed and open dwell times for the adult human receptor (MATERIALS AND METHODS; Fig. 2). Rate constants underlying ACh binding, channel opening, and ACh blocking steps are all well defined by the analysis, and show rapid and efficient opening of the doubly occupied receptor, distinguishable association and dissociation rate constants for each binding site, and low affinity block of the open channel by ACh (Table I). The fitted rate constants are within the range of values previously reported for the adult human receptor (Ohno et al., 1996; Wang et al., 1999; Hatton et al., 2003; Lee and Sine, 2004).
Maximum likelihood fitting based on Scheme II, which incorporates brief and intermediate duration desensitized states, also describes the distributions of closed and open dwell times, and yields essentially the same rate constants for binding and gating steps obtained using Scheme I (Table I). The similar results obtained with Schemes I and II show that the presence of closings due to desensitization does not affect the estimates of rate constants for receptor activation when Scheme I is applied to the wild-type human receptor (Lee and Sine, 2004).
Functional Contribution of K145
K145 is a potential link between the binding site and the ion channel because it is proximal to the binding site, it is physically linked to the Cys-loop, it is proximal to electron-rich side chains that link to M1 via -strand 10, and it is conserved across 1 subunits. In the agonist-free structure of AChBP, the residue equivalent to K145, K139, forms a salt bridge with D194, whereas in the CCh-bound structure, K139 forms a salt bridge with the key binding site tyrosine, Y185 (Celie et al., 2004). To investigate the functional contribution of K145, we generated a series of mutations of K145, coexpressed each with complementary , , and subunits, and recorded ACh-evoked single channel currents.
Substitution of Ala for K145 profoundly impairs receptor activation at all concentrations of ACh, with activation episodes due to a single receptor consisting of predominantly single openings separated by prolonged closings (compare Figs. 2 and 4). The corresponding closed time histograms span a wide range of times at all ACh concentrations examined, and the major peak of closings shows an ACh-dependent shift toward shorter times. Maximum likelihood fitting of Scheme I to the global dataset for K145A well describes the closed and open dwell time histograms (Fig. 4). The fitted rate constants reveal nearly a 200-fold slowing of the channel opening rate constant and no change of the channel closing rate constant (Table I). For K145A, adequate definition of the rate constants underlying ACh binding steps required the approximation that the set of elementary association and dissociation rate constants are equal at each binding site. Presumably this approximation was required because information on the first binding step was limited: clusters of events corresponding to a single receptor could only be identified at ACh concentrations >30 μM, which may exceed the dissociation constant for ACh binding to the first binding site, and brief, mono-liganded openings were not present. Nevertheless, our analysis suggests that the rate constants underlying ACh binding are only modestly affected by K145A (Table I). Thus substitution of Ala for K145 at the periphery of the ACh binding site predominantly impairs elementary steps underlying channel gating.
Functional Consequences of K145Q
Because the Lys to Ala substitution represents a substantial structural change, we examined functional consequences of the more structurally conservative mutation K145Q. The Gln substitution neutralizes the positive charge but maintains polarity and moderate size of the side chain. Receptors containing K145Q show impaired channel gating similar to K145A, but individual activation episodes show three kinetic modes distinguished by low, intermediate, and high open probability (Popen; Fig. 5). These kinetic modes were observed only at high concentrations of ACh (300 μM to 1 mM), and although they were present at all high concentrations, the contributions of each mode varied from patch to patch. We therefore describe detection of the three kinetic modes for one patch to exemplify our treatment of the overall data for K145Q.
We identified clusters of events corresponding to a single channel using our standard cluster discrimination method (MATERIALS AND METHODS), and generated histograms of mean cluster Popen and mean cluster closed time (C) for K145Q (Fig. 5; Milone et al., 1998). The distribution of Popen shows a major and a minor peak, while the distribution of C shows a major peak and a secondary tail of long C. Analysis of cluster properties reveals that distinct portions of the two distributions are correlated: the peak of high Popen corresponds to clusters with short C (filled bars), and the tail with long C corresponds to clusters with low Popen (open bars). The remainder of the distributions consists of clusters with intermediate Popen and C (gray bars). Clusters from wild-type receptors subjected to the same analysis show a predominant mode of high Popen and short C.
We then separated clusters of events for each of the kinetic modes for K145Q receptors, and combined one mode at a time with the kinetically homogeneous data obtained at lower ACh concentrations. Maximum likelihood fitting showed that only clusters with intermediate Popen and C were compatible with data obtained at lower ACh concentrations. Fitting Scheme I to the intermediate mode dataset (30 μM to 1 mM ACh) shows that substitution of Gln for K145 profoundly impairs the efficiency of channel gating; the corresponding probability density functions adequately describe the distributions of closed and open dwell times (Fig. 6). The effects of K145Q on channel gating are similar to those of K145A, with the channel opening rate constant slowed 60-fold and the channel closing rate constant relatively unchanged (Table I). Adequate definition of rate constants underlying ACh binding steps again required the approximation that the set of elementary association and dissociation rate constants are equal at each binding site. Given this approximation, the analysis suggests K145Q has relatively little effect on agonist binding (Table I). Thus, located at the periphery of the ACh binding site, K145 contributes primarily to gating of the receptor channel, fulfilling an important criterion as a transduction element.
Functional Contribution of D200
D200 is a potential electrostatic partner of K145, and in the mouse receptor was found to contribute predominantly to channel gating (Akk et al., 1996). To determine whether D200 contributes to channel gating in the human muscle receptor, and further, whether its functional contribution mirrors that of K145, we recorded currents from adult human receptors harboring the mutation D200N. Like mutations of K145, D200N profoundly impairs receptor activation at all concentrations of ACh; individual clusters consist of predominantly single openings interspersed by long closings (Fig. 7). Maximum likelihood fitting according to Scheme I reveals a predominant effect on channel gating, with the channel opening rate constant slowed nearly 70-fold and the channel closing rate constant slightly increased (Table I). As observed for mutations of K145, adequate definition of rate constants underlying ACh binding steps required the approximation that the set of elementary association and dissociation rate constants are equal at each binding site. The rate constants underlying ACh binding are relatively unaffected by D200N, as observed for the same mutation in the mouse receptor (Table I). Thus the consequences of D200N for channel gating are highly similar between mouse and human receptors, and, moreover, mirror the consequences of mutations of K145. The similar effects of mutations of K145 and D200 on channel gating and proximity of the two side chains in our structural model collectively suggest these residues are interdependent and constitute part of an intermediate link between binding and gating domains.
Interdependence of K145 and D200
To gain additional evidence for interdependence of K145 and D200, we combined the mutations K145Q and D200N into a single subunit and recorded ACh-evoked single channel currents from the resulting double mutant receptors. The current traces again show impaired receptor activation at all ACh concentrations (Fig. 8), with predominantly single openings separated by long closings, similar to either single mutation (compare Figs. 6, 7, and 8). Unlike receptors containing K145Q alone, only a single kinetic mode was distinguished for the double mutant receptor, simplifying the analysis. Maximum likelihood fitting of Scheme I to the global set of dwell times reveals a 60-fold slowing of the channel opening rate constant and little change of the channel closing rate constant when compared with the wild-type receptor (Table I). These results are similar to those for either single mutation, showing qualitatively that K145Q and D200N are interdependent in contributing to channel gating. Double mutant cycles analysis (Horovitz and Fersht, 1990) applied to the set of gating equilibrium constants for wild-type, K145Q, D200N, and double mutant receptors yields a free energy of interaction of –2.6 kcal/mol. Rate constants underlying ACh binding steps are altered twofold or less by the double mutation, similar to D200N alone, but this conclusion remains provisional because the assumption of equivalent binding sites was required to obtain well-defined rate constants. Thus, the largely selective perturbation of channel gating by the double mutation quantitatively mirrors that for either single mutation, further suggesting K145 and D200 are interdependent in coupling agonist binding to channel gating.
Charge Reversal of K145 and D200
To test for electrostatic interaction between K145 and D200, we generated the single mutants K145E and D200K and the corresponding double mutant and recorded ACh-evoked single channel currents from the resulting mutant receptors (Fig. 9). The single mutation D200K slows the channel opening rate constant 400-fold and increases the channel closing rate constant nearly fourfold, yielding a 1,400-fold decrease of the channel gating equilibrium constant (Table I). For D200K, well-defined single receptor clusters could not be discerned at ACh concentrations <1 mM, preventing estimation of rate constants underlying ACh binding. Consequences of the single mutation K145E mirror those of other mutations of K145; compared with the wild-type receptor, the channel opening rate constant slows 100-fold, the channel closing rate constant slows almost twofold, and the ACh dissociation rate constants are only modestly affected. The rate constants for ACh association, on the other hand, slow three- to fivefold compared with other mutations of K145 (Table I).
Compared with D200K alone, restoring the local negative charge by combining K145E markedly increases the frequency of channel openings within single receptor clusters (Fig. 9), and counteracts the slowing of the channel opening rate constant and the increase of the channel closing rate constant (Table I). The extent to which the gating rate constants are restored is only partial, perhaps because residues flanking each member of the putative salt bridge are not compatible with the reversed charge. The net result of charge reversal is a 14-fold increase of the channel gating equilibrium constant compared with D200K alone. Double mutant cycles analysis applied to the set of four gating equilibrium constants yields a free energy of interaction of –4 kcal/mol. Further, the slow association of ACh caused by K145E alone is restored to the range of values observed for other mutations of K145 (Table I).
Finally, we examined the single mutation K145D and found that it prevented expression of functional channels and cell-surface -Btx binding sites. However, K145D is rescued by the charge reversal mutation D200K. Full kinetic analysis of the double mutation reveals an eightfold increase of the channel gating equilibrium constant compared with D200K alone, due to a twofold increase of the channel opening rate constant and a fourfold slowing of the channel closing rate constant (Table I). Changes in rate constants underlying ACh binding remain unknown because these could not be measured for either single mutation. However, the ACh association rate constants for the double mutation increase twofold compared with K145E alone, while the ACh dissociation rate constants slow twofold compared with K145E. The overall results from charge-neutralization and charge-reversal experiments provide strong evidence that K145 and D200 interact through electrostatic forces to couple agonist binding to channel gating.
Functional Contributions of Y190 and T202
Our structural model predicts that two more residues are proximal to K145 and D200: Y190 and T202 (Fig. 1). We therefore generated structurally conservative mutations of each residue, Y190F and T202V, and recorded single channel currents from the resulting mutant receptors. For receptors containing Y190F, the ACh concentration had to be increased to 3 mM to obtain well-defined clusters of single receptor currents; the corresponding intracluster events consist of single brief openings separated by prolonged closings (Fig. 10). The availability of data at only a single high concentration of ACh limited kinetic fitting to the subset of doubly occupied gating steps in Scheme I, but nevertheless reveals a 1,000-fold slowing of the channel opening rate constant with little change in the channel closing rate constant (Table I). These effects of Y190F on channel gating are analogous to those observed for the same mutation in the fetal mouse muscle receptor (Chen et al., 1995), but the 800-fold decrease of the gating equilibrium constant for the adult human receptor exceeds the 200-fold decrease observed for the fetal mouse receptor.
The kinetics of receptor activation are only modestly affected by T202V. Rate constants for elementary binding or gating steps change twofold or less compared with the wild-type receptor (Table I). Thus despite its conservation across 1 subunits and proximity to D200 and K145, T202 does not contribute substantially to either ACh binding or channel gating.
Interdependence of K145 and Y190
The agonist-bound conformation of AChBP shows that K139 and Y185 form a salt bridge (Celie et al., 2004). To determine whether residues in the receptor equivalent to these, K145 and Y190, are interdependent, we combined K145Q and Y190F into a single subunit and recorded ACh-evoked single channel currents from the resulting double mutant receptors. The current traces again show impaired receptor activation at 1 mM ACh, with brief openings separated by long closings, similar to either single mutation (Fig. 10). Unlike receptors containing K145Q alone, only a single kinetic mode could be distinguished for the double mutant receptor. Further, well-defined clusters of single receptor openings could only be discerned at ACh concentrations from 100 μM to 1 mM, restricting the analysis to the channel gating rate constants. Maximum likelihood fitting shows that compared with the wild-type receptor, the double mutation slows the channel opening rate constant 200-fold without affecting the channel closing rate constant (Table I). Double mutant cycles analysis applied to the set of gating equilibrium constants for wild-type, K145Q, Y190F, and double mutant receptors yields a free energy of interaction of –2.9 kcal/mol. Thus K145Q and Y190F are interdependent in coupling agonist binding to channel gating.
Pairwise Interdependence Assessed by Steady-state Binding of ACh
Most mutations that affect receptor function alter steady-state binding of ACh (examples include Chen et al., 1995; Ohno et al., 1996; Sine et al., 2002b). Although measurements of steady-state ACh binding contain contributions from closed, open, and desensitized states of the receptor, each of which differs in affinity for the agonist (Katz and Thesleff, 1957; Sine and Taylor, 1979; Jackson, 1989), binding measurements can be used as an overall index of receptor function. We therefore sought further evidence for pairwise interdependence by measuring steady-state binding of ACh to receptors containing single and double mutations. We first studied the charge neutral mutations K145Q and D200N and the corresponding double mutation. Relative to the wild-type receptor, the single mutations K145Q and D200N increase the apparent dissociation constant for ACh (Kapp) 10- and 26-fold, respectively, qualitatively consistent with expectations from impaired channel gating observed by single channel recording (Fig. 11; Table II). The double mutation shows a virtually identical binding profile to that of D200N alone, and double mutant cycles analysis applied to Kapp yields an interaction free energy of 1 kcal/mol for this charge neutralization experiment. Like Kapp, interaction free energy based on Kapp arises from multiple processes: ACh binding, desensitization, and channel gating. One of these processes is channel gating, which contributes 2.6 kcal/mol, but the other processes likely contribute. We next measured steady-state ACh binding for the charge reversal mutations K145E and D200K and the corresponding double mutation. The single mutations K145E and D200K increase Kapp 100- and 860-fold, respectively, again qualitatively consistent with profoundly impaired channel gating (Fig. 11; Table II). However, charge reversal by the double mutation strongly counteracts the effects of either single mutation, yielding an overall 20-fold increase of Kapp compared with the wild-type receptor. Double mutant cycles analysis applied to Kapp for this charge reversal experiment reveals an interaction free energy of 4.8 kcal/mol, which is similar to the contribution of 4 kcal/mol observed for channel gating alone. The mutation K145D does not produce functional receptors, but when combined with D200K, yields a Kapp coincident with that obtained for K145E plus D200K (Fig. 11; Table II).
The double mutation containing K145Q and Y190F failed to bind -Btx, preventing assessment of interdependence by steady-state ACh binding. Nevertheless the overall results provide strong evidence for electrostatic interaction between K145 and D200 and interdependence of K145 and Y190. A three-way electrostatic interplay of charged residues thus emerges as a compelling candidate for an initial link between ligand binding and channel gating domains owing to the proximity of the triad to the ACh binding site, interdependence of residues in the triad, connections to the Cys-loop and M1 domain, and the predominant contributions of all three residues to channel gating.
Subunit Specificity of the Mutations
The three residues studied here are unique to and conserved across 1 subunits. However, subunits also contain conserved Asp and Lys at positions equivalent to D200 and K145, respectively. To determine whether this residue pair in the subunit contributes to receptor function, we generated the mutations K147Q and D214N, coexpressed each with complementary , , and subunits, and measured steady-state ACh binding. Neither mutation affects Kapp for ACh binding (Table II), showing that if these two residues form a salt bridge, it is unlikely to contribute to receptor function.
DISCUSSION
We show that three conserved residues at the periphery of the ACh binding site, K145, D200, and Y190, contribute profoundly to gating of the receptor channel. One of these residues, Y190, contributes to the aromatic cage that coordinates the quaternary ammonium moiety of the agonist (Celie et al., 2004), and its hydroxyl group projects away from the binding site toward the other two members of the triad (Fig. 1). Of the other two residues, K145 is the most likely contact for Y190, but this contact corresponds to the agonist-bound conformation. Evidence for this contact comes from the AChBP crystal structure with CCh bound, which shows a salt bridge between the equivalent residues K139 and Y185 (Celie et al., 2004), and from the interdependent contributions of K145 and Y190 to gating shown here. Support for electrostatic contact between K145 and D200 comes from the AChBP crystal structure without bound agonist, which shows a salt bridge between the equivalent Lys and Asp residues (Brejc et al., 2001), and from our data demonstrating interdependence of K145 and D200 in contributing to channel gating.
Our findings combine with inter-atomic distances in AChBP to suggest how interplay among these three residues couples local conformational change due to agonist binding to gating of the channel. Our inferences draw from two sets of atomic coordinates of AChBP with bound CCh deposited in the protein data bank (PDB code 1UV6): a pentameric unit with one molecule of CCh bound and another unit with two molecules of CCh bound. Inspection of the three subunit interfaces to which CCh is bound shows that recognition domain C is drawn inward toward the center of the binding site, whereas inspection of one of the unoccupied interfaces shows that domain C is extended outward from the site. Molecular dynamics simulation of AChBP and measurements of intrinsic tryptophan fluorescence reveal similar changes in the conformations of domain C with and without bound agonist (Gao et al., 2005). In the following mechanistic interpretation, we refer to these conformations as agonist-free and agonist-bound.
To illustrate our structural interpretation, the agonist-free and -bound conformations of recognition domain C of AChBP are compared (Fig. 12). The agonist-bound conformation is that of the crystal structure of AChBP with bound CCh (Celie et al., 2004). The agonist-free conformation is that generated by prolonged molecular dynamics simulation of AChBP (Gao et al., 2005), which is similar to the extended conformation of a subunit of AChBP at an unoccupied interface (see PDB entry 1UV6). ACh is shown docked in the orientation obtained by computational docking to the HEPES-bound crystal structure of AChBP (Brejc et al., 2001; Gao et al., 2005). In the agonist-free conformation, the Tyr at the position equivalent to Y190 is 8 away from the Lys equivalent to K145, while residues equivalent to K145 and D200 form a salt bridge. In the agonist-bound conformation, the Tyr equivalent to Y190 moves to within 2–3 of the Lys equivalent to K145, drawing recognition domain C inward, and allowing the residue equivalent to D200 to move away from the Tyr/Lys pair. Thus by analogy to AChBP, in the resting state of the receptor, K145 and D200 are proposed to pair through electrostatic forces and Y190 is out of register (Fig. 11). However when the agonist binds, Y190 is drawn toward the K145/D200 pair and, through electrostatic forces, pulls K145 away from D200, allowing both residues to relax to new positions. These local inter-residue displacements could propagate to the channel via -strands 7 and 10 of the ligand binding domain. -Strand 7 harbors K145 and forms part of the signature Cys-loop, which interacts with the linker spanning the M2 and M3 transmembrane domains (Kash et al., 2003; Bouzat et al., 2004), while -strand 10 contains D200 and links directly to the M1 transmembrane domain.
Although all three residues of the triad contribute profoundly to channel gating, each has a distinct role in linking the binding site to the channel. Y190 is the clear link to the ACh binding site, as shown by site-directed labeling (Abramson et al., 1989) and its location in the aromatic cage that coordinates the agonist (Celie et al., 2004). Its function is to help draw recognition domain C toward the center of the binding site when the agonist is bound and, in the process, form a salt bridge with K145. Accordingly, removal of the aromatic hydroxyl group by mutation to Phe would prevent formation of the salt bridge and disrupt coupling of binding to gating. Likewise, substitution of Gln, Ala, or Glu for K145 would remove the electrostatic partner for Y190, also disrupting coupling. By analogy to AChBP in the agonist-free conformation, D200 is proposed to form a salt bridge with K145, positioning K145 for contact with the incoming Y190 and maintaining -strands 7 and 10 in their closed state conformations. When the agonist binds, Y190 draws K145 away from D200, allowing -strands 7 and 10 to move to their active conformations. Substitution of Asn for D200 would prevent stabilization of K145 for the incoming Y190, and break the link between -strand 10 and the binding site, disrupting coupling.
One of the main lines of evidence that the triad identified here is an initial link in the binding to gating cascade is that mutation of each residue profoundly impairs gating of the channel. However two additional observations require explanation. First, the allosteric link between binding and gating domains is not completely disrupted by any of the mutations, as ACh can still trigger opening of the channel, although with greatly reduced efficiency. The most likely explanation is there are additional allosteric links that remain intact in each mutant receptor and account for the residual ACh-evoked currents. Second, the extent to which gating is disrupted differs by as much as 10-fold among the three mutations. If each member of the triad is essential for coupling, the extent to which gating is disrupted should be the same following mutation of each member. However in addition to coupling binding to gating, the triad may also contribute to the allosteric equilibrium between resting and open channel states in the absence of agonist (Jackson, 1989). This contribution to the allosteric set point likely involves dynamic interplay among the three residues, such that changes in the set point differ when one of the members is structurally altered. Thus for each mutation, the gating equilibrium for di-liganded receptors could be altered by differing amounts, owing to different contributions to the allosteric set point.
Other residues that contribute to the allosteric set point have been identified in the ligand binding domains of members of the Cys-loop receptor superfamily. In the nicotinic receptor 1 subunit, mutation of Asp 97 markedly increases the frequency of channel opening events in the absence of ACh (Chakrapani et al., 2003). In the GABAA receptor 1 subunit, mutation of Tyr 107 increases the resting membrane conductance in a manner that was sensitive to the specific GABA receptor inhibitor picrotoxin (Torres and Weiss, 2002). Similarly in the GABAA receptor 2 subunit, mutation of Glu 155 increases the resting conductance in a picrotoxin-sensitive manner (Newell et al., 2004). In all three cases, the allosteric set point was increased in favor of the open channel state, whereas mutation of any residue of the triad in the present study impedes opening of the channel. Thus among functionally key residues in the ligand binding domain of the receptor, a balance is likely established between residues that promote and those that impair channel gating.
The triad of residues identified here is found in all species of nicotinic receptor 1 subunits, suggesting they contribute to agonist binding transduction in all muscle type nicotinic receptors (for subunit sequences see Le Novere and Changeux, 2001). Similarly, the triad is also present in all neuronal nicotinic subunits (2–10), except 5, which contains Asp at the position equivalent to Y190, and 10, which contains Thr at the position equivalent to K145. However, neither of these exceptions likely forms the principal face of the binding site; 5 appears to be a structural subunit, requiring both neuronal and subunits to form functional receptors (Ramirez-Latorre et al., 1996), and 10 requires coexpression of the 9 subunit (Sgard et al., 2002). On the other hand, the triad is not present in GABAA, glycine, and 5-HT receptors. Thus among the Cys-loop superfamily of receptors, the initial coupling mechanism is likely to be highly similar for muscle and neuronal nicotinic receptors, but the more evolutionary distant receptors evolved different structures to relay the initial consequences of agonist binding. This structural divergence among receptors activated by different neurotransmitters is not surprising because the different agonist structures inherently require different chemical counterparts in the binding site.
In summary, our observations lead to identification of a triad of conserved residues as an initial link that couples agonist binding to channel gating. All three residues of the triad localize to the same three-dimensional region of our structural model of the receptor ligand binding domain, and all three residues are conserved across 1 subunits. Also, equivalent residues in AChBP change conformation when the agonist is bound. Finally, structurally conservative mutations of each residue predominantly impair channel gating, and two pairs within the triad, K145/D200 and K145/Y190, are interdependent in contributing to gating. Our findings suggest future studies aimed at identifying additional links that couple agonist binding to channel gating.
ACKNOWLEDGMENTS
We thank Dr. Hai-Long Wang for the homology model and Fan Gao for the coordinates used to make Fig. 12.
This work was supported by National Institutes of Health grant NS31744 to S.M. Sine.
Olaf S. Andersen served as editor.
Submitted: 9 March 2005
Accepted: 5 May 2005
REFERENCES
Abramson, S.N., Y. Li, P. Culver, and P. Taylor. 1989. An analog of lophotoxin reacts covalently with tyrosine 190 in the subunit of the nicotinic receptor. J. Biol. Chem. 264:12666–12672.
Akk, G., S. Sine, and A. Auerbach. 1996. Binding sites contribute unequally to the gating of mouse nicotinic D200N acetylcholine receptors. J. Physiol. 496:185–196.
Bouzat, C., F. Gumilar, G. Spitzmaul, H.-L. Wang, D. Rayes, S.B. Hansen, P. Taylor, and S.M. Sine. 2004. Coupling of agonist binding to channel gating in an ACh-binding protein linked to an ion channel. Nature. 430:896–900.
Brejc, K., W. van Dijk, R. Klassen, M. Schuurmans, J. van der Oost, A. Smit, and T. Sixma. 2001. Crystal structure of an ACh-binding protein reveals the ligand-binding domain of nicotinic receptors. Nature. 411:269–276.
Celie, P., S. van Rossum-Fikkert, W. van Dijk, K. Brejc, A. Smit, and T. Sixma. 2004. Nicotine and carbamylcholine binding to nicotinic acetylcholine receptors as studied in AChBP crystal structures. Neuron. 41:907–914.
Chakrapani, S., T.D. Bailey, and A. Auerbach. 2003. The role of loop 5 in acetylcholine receptor channel gating. J. Gen. Physiol. 122:521–539.
Chakrapani, S., T.D. Bailey, and A. Auerbach. 2004. Gating dynamics of the acetylcholine receptor extracellular domain. J. Gen. Physiol. 123:341–356.
Chen, J., Y. Zhang, G. Akk, S. Sine, and A. Auerbach. 1995. Activation kinetics of recombinant mouse nicotinic acetylcholine receptors with mutations at subunit residue tyrosine 190. Biophys. J. 69:849–859.
Colquhoun, D., and D.C. Ogden. 1988. Activation of ion channels in the frog end-plate by high concentrations of acetylcholine. J. Physiol. 395:131–159.
Colquhoun, D., and F. Sigworth. 1983. Fitting and statistical analysis of single channel records. Single Channel Recording. B. Sakmann and E. Neher, editors. Plenum Publishing Corp., New York. 191–264.
Corringer, J.P., N. LeNovere, and J.P. Changeux. 2000. Nicotinic receptors at the amino acid level. Annu. Rev. of Toxicol. 40:431–458.
Elenes, S., and A. Auerbach. 2002. Desensitization of diliganded mouse muscle nicotinic acetylcholine receptor channels. J. Physiol. 541:367–383.
Gao, F., T. Burghardt, N. Bren, S.B. Hansen, R. Henchman, P. Taylor, J.A. McCammon, and S.M. Sine. 2005. Acetylcholine-mediated conformational changes in acetylcholine-binding protein revealed by simulation and intrinsic tryptophan fluorescence. J. Biol. Chem. 280:8443–8451.
Hamill, O.P., A. Marty, E. Neher, B. Sakmann, and F.J. Sigworth. 1981. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch. 391:85–100.
Hatton, C.J., C. Shelley, M. Brydson, D. Beeson, and D. Colquhoun. 2003. Properties of the human muscle nicotinic receptor, and of the slow-channel myasthenic syndrome mutant L221F, inferred from maximum likelihood fits. J. Physiol. 547:729–760.
Horovitz, A., and A.R. Fersht. 1990. Strategy for analysing the co-operativity of intramolecular interactions in peptides and proteins. J. Mol. Biol. 214:613–617.
Jackson, M.B. 1989. Perfection of a synaptic receptor: kinetics and energetics of the acetylcholine receptor. Proc. Natl. Acad. Sci. USA. 86:2199–2203.
Karlin, A. 2002. The emerging structure of the nicotinic acetylcholine receptors. Nat. Rev. Neurosci. 3:102–114.
Kash, T., A. Jenkins, J. Kelly, J. Trudell, and N.L. Harrison. 2003. Coupling of agonist binding to channel gating in the GABAA receptor. Nature. 421:272–275.
Katz, B., and S. Thesleff. 1957. A study of desensitization produced by acetylcholine at the motor end-plate. J. Physiol. 138:63–80.
Le Novere, N., and J.P. Changeux. 2001. LGICdb: the ligand-gated ion channel database. Nucleic Acids Res. 29:294–295.
Le Novere, N., T. Grutter, and J.P. Changeux. 2002. Models of the extra-cellular domain of the nicotinic receptors and of agonist- and Ca2+-binding sites. Proc. Natl. Acad. Sci. USA. 99:3210–3215.
Lee, B.S., R.B. Gunn, and R.R. Kopito. 1991. Functional differences among nonerythroid anion exchangers expressed in a transfected human cell line. J. Biol. Chem. 266:11448–11454.
Lee, W.Y., and S.M. Sine. 2004. Invariant aspartic acid in muscle nicotinic receptor contributes selectively to the kinetics of agonist binding. J. Gen. Physiol. 124:555–567.
Milone, M., H.L. Wang, K. Ohno, R. Prince, T. Fukudome, X.M. Shen, J.M. Brengman, R.C. Griggs, S.M. Sine, and A.G. Engel. 1998. Mode switching kinetics produced by a naturally occurring mutation in the cytoplasmic loop of the human acetylcholine receptor epsilon subunit. Neuron. 20:575–588.
Miyazawa, A., Y. Fujiyoshi, and N. Unwin. 2003. Structure and gating mechanism of the acetylcholine receptor pore. Nature. 423:949–955.
Molles, B.E., I. Tsigelny, P. Nguyen, S. Gao, S.M. Sine, and P. Taylor. 2002. Residues in the subunit of the nicotinic acetylcholine receptor interact to confer selectivity of Waglerin-1 for the – subunit interface site. Biochemistry. 41:7895–7906.
Newell, J.G., R.A. McDevitt, and C. Czajkowski. 2004. Mutation of glutamate 155 of the GABA(A) receptor 2 subunit produces a spontaneously open channel: a trigger for channel activation. J. Neurosci. 24:11226–11235.
Ohno, K., H.-L. Wang, M. Milone, N. Bren, J.M. Brengman, S. Nakano, P. Quiram, J.N. Pruitt, S.M. Sine, and E.G. Engel. 1996. Congential myasthenic syndrome caused by decreased agonist binding affinity due to a mutation in the acetylcholine receptor subunit. Neuron. 17:157–170.
Pear, W.S., G.P. Nolan, M.L. Scott, and D. Baltimore. 1993. Production of high-titer helper-free retroviruses by transient transfection. Proc. Natl. Acad. Sci. USA. 90:8392–8396.
Schapira, M., R. Abagyan, and M. Totrov. 2002. Structural model of nicotinic acetylcholine receptor isotypes bound to acetylcholine and nicotine. BMC Struct. Biol. 2:1.
Qin, F., A. Auerbach, and F. Sachs. 1996. Estimating single channel kinetic parameters from idealized patch clamp data containing missed events. Biophys. J. 70:264–280.
Ramirez-Latorre, J., C.R. Yu, X. Qu, F. Perin, A. Karlin, and L. Role. 1996. Functional contributions of -5 subunit to neuronal receptors. Nature. 380:347–351.
Sakmann, B., J. Patlak, and E. Neher. 1980. Single acetylcholine-activated channels show burst-kinetics in presence of desensitizing concentrations of agonist. Nature. 286:71–73.
Sgard, F., E. Charpantier, S. Bertrand, N. Walker, D. Caput, D. Graham, D. Bertrand, and F. Besnard. 2002. A novel human nicotinic receptor subunit, 10, that confers functionality to the 9-subunit. Mol. Pharmacol. 61:150–159.
Sine, S.M., and J.H. Steinbach. 1987. Activation of acetylcholine receptors in clonal BC3H-1 cells by high concentrations of agonist. J. Physiol. 385:325–359.
Sine, S.M. 1993. Molecular dissection of subunit interfaces in the acetylcholine receptor: identification of residues that determine curare selectivity. Proc. Natl. Acad. Sci. USA. 90:9436–9440.
Sine, S.M., and P. Taylor. 1979. Functional consequences of agonist-mediated state transitions in the cholinergic receptor. J. Biol. Chem. 254:3315–3325.
Sine, S.M. 2002. The nicotinic receptor ligand binding domain. J. Neurobiol. 53:431–446.
Sine, S.M., H.L. Wang, and N. Bren. 2002a. Lysine scanning mutagenesis delineates structural model of the nicotinic receptor ligand binding domain. J. Biol. Chem. 277:29210–29223.
Sine, S.M., X.M. Shen, H.L. Wang, K. Ohno, W.Y. Lee, A. Tsujino, J. Brengmann, N. Bren, J. Vajsar, and A.G. Engel. 2002b. Naturally occurring mutations at the acetylcholine receptor binding site independently alter ACh binding and channel gating. J. Gen. Physiol. 120:483–496.
Torres, V.I., and D.S. Weiss. 2002. Identification of a tyrosine in the agonist binding site of the homomeric rho1 -aminobutyric acid (GABA) receptor that, when mutated, produces spontaneous opening. J. Biol. Chem. 277:43741–43748.
Unwin, N. 2005. Refined structure of the nicotinic acetylcholine receptor at 4 resolution. J. Mol. Biol. 346:967–989.
Wang, H.-L., M. Milone, K. Ohno, X.-M. Shen, A. Tsujuno, A. Paola, P. Tonali, J. Brengman, A.G. Engel, and S.M. Sine. 1999. Acetylcholine receptor M3 domain: stereochemical and volume contributions to channel gating. Nat. Neurosci. 2:226–233.
Wang, H.-L., K. Ohno, M. Milone, J.M. Brengman, A. Evoli, A.P. Batocchi, L.T. Middleton, K. Christodoulou, A.G. Engel, and S.M. Sine. 2000. Fundamental gating mechanism of nicotinic receptor channel revealed by mutation causing a congenital myasthenic syndrome. J. Gen. Physiol. 116:449–462.(Nuriya Mukhtasimova, Chri)