点击显示 收起
【关键词】 Allosteric
In general, the M2 subtype of muscarinic acetylcholine receptors has the highest sensitivity for allosteric modulators and the M5 subtype the lowest. The M2/M5 selectivity of some structurally diverse allosteric agents is known to be completely explained by M2177Tyr and M2423Thr in receptors whose orthosteric site is occupied by the conventional ligand N-methylscopolamine (NMS). This study explored the role of the conserved M2422Trp and the adjacent M2423Thr in the binding of alkane-bisammonio type modulators, gallamine, and diallylcaracurine V. Experiments were performed with human M2 or M5 receptors or mutants thereof. It was found that M2422Trp and M2423Thr independently influenced allosteric agent binding. The presence of M2423Thr may enhance the affinity of binding, depending on the allosteric agent, either directly or indirectly (by avoiding sterical hindrance through its M5 counterpart 478His). Replacement of M2422Trp and of the corresponding M5477Trp by alanine revealed a pronounced contribution of these epitopes to subtype independent baseline affinity in NMS-bound and NMS-free receptors for all agents except diallylcaracurine V. In a few instances, this tryptophan also influenced cooperativity and subtype selectivity. Docking simulations using a three-dimensional M2 receptor model revealed that the aromatic rings of M2177Tyr and M2422Trp, in a concerted action, might fix one of the aromatic moieties of alkane-bisammonio compounds between them. Thus, M2422Trp and the spatially adjacent M2177Tyr, as well as M2423Thr, form a cluster of amino acids within the allosteric binding cleft that is pivotal for both M2/M5 subtype selectivity and baseline affinity of allosteric agents.All five subtypes of muscarinic acetylcholine receptors contain an allosteric site apart from the orthosteric site that is addressed by acetylcholine and conventional muscarinic agonists and antagonists. Binding of an allosteric modulator allows formation of ternary complexes consisting of the allosteric agent, the orthosteric ligand, and the receptor protein. Through ternary complex formation, allosteric agents may evoke particular actions that cannot be induced by orthosteric ligands alone and that may have therapeutic potential. For instance, allosteric modulators may increase the binding of orthosteric agonists or antagonists (positive cooperativity) or they may inhibit orthosteric ligand binding (negative cooperativity). In either case, the magnitude of the cooperativity will define an intrinsic limit on the magnitude of the positive or negative effect, in marked contrast to the unconstrained action of orthosteric agonists and antagonists. It is also possible for allosteric modulators to leave orthosteric ligand binding unchanged (neutral cooperativity) while nevertheless changing the kinetics of binding (Ellis, 1997; Christopoulos and Kenakin, 2002; Krejí et al., 2004; Soudijn et al., 2004; Birdsall and Lazareno, 2005; Wess, 2005). Finally, in addition to modulating orthosteric ligand binding properties, allosteric agents also may modulate agonist induced intrinsic efficacy (Zahn et al., 2002). A better understanding of the molecular topology and mechanisms of allosteric agent binding and action will help to design new allosteric agents with improved properties and will lead to a better insight into the principles of muscarinic receptor function. The M2 subtype of muscarinic receptors generally displays highest affinity for allosteric modulators, whereas the M5 subtype has lowest sensitivity. A rather good insight into the allosteric binding area has now been achieved by combining three strategies [i.e., development of allosteric agents with high affinity and selectivity for M2 receptors that presumably fit tightly in a fixed position at the receptor protein (Mohr et al., 2003), receptor mutagenesis starting from M2/M5 chimeric receptor constructs to identify essential epitopes for allosteric agent binding (Ellis et al., 1993; Gnagey et al., 1999; Buller et al., 2002; Huang et al., 2005), and generation of a three-dimensional M2 receptor model based on the crystal structure of the inactive bovine rhodopsin (Jöhren and Höltje, 2002; Voigtländer et al., 2003)]. This approach has allowed the visualization of different binding topologies for typical and atypical allosteric agents (Tränkle et al., 2005; Wess, 2005). However, the mode by which certain epitopes affect binding affinity of allosteric agents is still in question. For instance, an amino acid may directly serve as a docking point or alternatively constitute a sterical hindrance, or it may indirectly contribute to ligand binding by governing the conformation of amino acid strands that contain a relevant point of attachment. We have found that two amino acids are sufficient to account completely for the 100-fold M2/M5 selectivity of structurally different allosteric agents (Voigtländer et al., 2003). These amino acids are M2177Tyr and M2423Thr, corresponding to the M5 amino acids 184Gln and 478His. The receptor model suggested M2423Thr to be a direct docking point for caracurine V-type agents. For alkane-bisammonio-type compounds such as W84 (Fig. 1), however, the model suggested an indirect influence, in that M2423Thr induces a favorable spatial adjustment of the adjacent M2422Trp for its interaction with one of the phthalimide moieties of W84. The involvement of M2422Trp is supported by a broad mutagenesis study by Matsui et al. (1995), who found the corresponding tryptophan of M1 to be relevant to the binding of the allosteric agent gallamine. Therefore, we set out to clarify the role of this conserved tryptophan and its neighboring M2423Thr or M5478His. As a major outcome of this study, we found that this tryptophan is of crucial importance for M2/M5 subtype independent baseline affinity of alkane-bisammonio type allosteric modulators and of gallamine. Thus, an epitope of high relevance to M2/M5 subtype independent baseline affinity of allosteric agents has been discovered. Furthermore, we found that this tryptophan in certain instances may provide subtype selectivity for allosteric agents or may modulate their cooperativity with an orthosteric antagonist. Taken together, the findings reveal that relevant epitopes for subtype dependent and subtype independent allosteric agent binding are clustered in close spatial proximity at the junction between the allosteric and the orthosteric binding areas of muscarinic acetylcholine receptors.
Fig. 1. Structures of the applied allosteric agents.
Materials
Atropine sulfate, gallamine triethiodide, and N-methylscopolamine bromide were obtained from Sigma-Aldrich (Steinheim, Germany). W84 is commercially available from Tocris Cookson Inc. (St. Louis, MO). Dimethyl-W84 (Tränkle et al., 1998), naphmethonium (Muth et al., 2003) and diallylcaracurine V (Zlotos et al., 2000) were generously provided by Prof. Dr. Ulrike Holzgrabe, Dr. Mathias Muth, and Dr. Darius P. Zlotos (Institute of Pharmacy, University of Würzburg, Würzburg, Germany). The orthosteric radioligand [3H]NMS ([3H]N-methylscopolamine chloride, 81 Ci/mmol) was purchased from NEN-DuPont (Homburg, Germany).
Mutagenesis and Expression
Point-mutated M2 and M5 receptors were generated as described previously (Voigtländer et al., 2003; Huang et al., 2005) using the QuikChange site-directed mutagenesis kit (Stratagene, Amsterdam, The Netherlands). pCD plasmids, including wild-type genes for human M2 and M5 muscarinic receptors, served as templates in polymerase chain reaction. Mutations were inserted by addition of synthetic oligonucleotide primers containing the required triplet changes (Sigma Genosys, Steinheim, Germany). Primers were elongated during temperature cycling by a high-fidelity DNA polymerase. After amplification over 12 to 16 cycles, the parental DNA was digested by a methylation-specific endonuclease, thus selecting the mutated plasmids. The polymerase chain reaction products were transformed into supercompetent Escherichia coli cells. Bacteria were plated on ampicillin-containing agar and incubated overnight. Clones with the putative mutation were isolated and grown up in 500 ml of Luria-Bertani medium (Sigma-Aldrich). Plasmids were extracted using the Plasmid Maxi kit (QIAGEN, Hilden, Germany) and mutation was confirmed by sequencing. Thereafter, wild-type or mutant plasmid-DNA was transiently transfected into COS-7 cells using PolyFect transfection reagent (QIAGEN).
Cell Culture and Membrane Preparation
COS-7 cells were cultured at 37°C under humidified air supplemented with 5% CO2 in Dulbecco's modified Eagle's medium containing 1% penicillin-streptomycin and 10% fetal bovine serum. One day before transfection, cells were grown in 10-cm dishes by seeding 1.6 x 106 cells per dish. Forty-eight hours after transfection, cells were harvested and homogenized in 5 mM sodium/potassium phosphate buffer (PB; 4 mM Na2HPO4 and 1 mM KH2PO4), pH 7.4. After centrifugation, membranes were resuspended in 5 mM PB and stored in aliquots at -80°C.
Dissociation Binding Assays
Two-point [3H]NMS dissociation experiments were performed in 5 mM PB, pH 7.4, at 23°C as described previously (Voigtländer et al., 2003; Huang et al., 2005). Membranes with the respective wild-type or mutant M2 or M5 receptor were prelabeled with 1 nM radioligand for 30 min. Measurement of [3H]NMS dissociation was started by the addition of 3 µM atropine, with or without varying concentrations of allosteric modulator. After an appropriate time interval, depending on the [3H]NMS dissociation half-time of the respective receptor, dissociation was terminated by filtration using a Brandel cell harvester and GMF2 glass microfiber sheets (Sartorius, Göttingen, Germany), saturated with 0.1% polyethylenimine. Filtration was immediately followed by two rinses with ice-cold 40 mM PB. Filter-bound radioactivity was detected by liquid scintillation counting. Nonspecific binding of [3H]NMS was determined in the presence of 3 µM atropine.
Equilibrium Binding Assays
[3H]NMS equilibrium binding was measured in the presence of 0.2 nM radioligand and the indicated concentrations of the allosteric agent, as described previously (Voigtländer et al., 2003). Experiments were conducted in 5 mM PB, pH 7.4, at 23°C, and incubation was carried on until equilibrium binding was achieved. The incubation time sufficient to attain equilibrium was calculated according to equation 31 in Lazareno and Birdsall (1995).
The affinity of the orthosteric probe [3H]NMS for the wild-type and mutant receptors was determined in homologous competition experiments. For this purpose, membranes were incubated with 0.2 nM radioligand and unlabeled NMS (1 pM to 30 nM) for the same period, as applied in the equilibrium binding experiment with the allosteric agent. Filtration and quantification of membrane-bound radioligand were carried out as described above.
Data Analysis
Data from dissociation experiments were converted into apparent rate constants assuming monoexponential decay. The apparent rate constant obtained in the absence of allosteric modulator served as control, and the rate constants determined in the presence of each concentration of allosteric agent were expressed as a percentage of that control value. This percentage was then plotted against the log concentration of the modulator. Curve fitting by nonlinear regression analysis was based on a four-parameter logistic function as described by Tränkle and Mohr (1997). The resulting concentration-effect curves for the allosteric delay of [3H]NMS dissociation reflect the binding of the allosteric agent to the NMS-occupied receptors. When there was no significant difference (F-test, p 0.05) between the top plateau of the curve and the control value k-1 = 100%, or between the bottom plateau of the curve and k-1 = 0%, plateaus were fixed at 100% and 0%, respectively. The negative logarithm of the concentration inducing a half-maximal reduction of the NMS dissociation rate, pEC0.5,diss, indicates the affinity of the allosteric agent for the NMS-occupied receptor.
[3H]NMS equilibrium binding data were analyzed according to the ternary complex model of allosteric interactions (Stockton et al., 1983; Ehlert, 1988) using the following equation
(1)
L and A indicate the concentration of the orthosteric ligand (L) and the allosteric agent (A), respectively. KL and KA are the equilibrium dissociation constants for the binding of [3H]NMS and the allosteric agent, respectively, to the free receptor. The negative log values, pKL and pKA, reflect the corresponding binding affinities. is the factor of cooperativity and serves as a measure of the magnitude and direction of the interaction between L and A. In two cases of steep curves, equation 2 from Tränkle et al. (2003) was applied; this equation was derived from Lazareno and Birdsall (1995) and contains the slope factor as a variable. As a consequence of the allosteric delay of [3H]NMS dissociation, rather long incubation periods can be required to attain equilibrium binding of [3H]NMS, especially in the case of M5 receptors (control half-time of [3H]NMS dissociation approximately 2 h). W84 is known to be sensitive to spontaneous hydrolytic cleavage (t1/2 = 11 h; Schulz, 1998); the concentration values indicated for W84 represent effective concentrations calculated on the basis of the half-time of the decay and the applied incubation time. In the case of almost neutral cooperativity ( 1), [3H]NMS equilibrium binding remains nearly unchanged under the influence of increasing concentrations of the allosteric modulator. Under this condition, curve fitting with eq. 1 does not work. Because the validity of the cooperativity model can be assumed (e.g., Ellis, 1997; Tränkle et al., 1998; Raasch et al., 2002) the affinity of the allosteric agent to the NMS-occupied receptors pEC0.5,diss is identical to p( · KA). Therefore, we replaced KA in eq. 1 with EC0.5,diss/ (Raasch et al., 2002). EC0.5,diss was derived from preceding dissociation experiments, and curve fitting yielded the cooperativity factor , which then served to calculate KA = EC0.5,diss/. Nonlinear regression analysis was performed using the Prism program (vers. 3.02; GraphPad Software, San Diego, CA).
Three-Dimensional Modeling and Docking Simulations
Homology Modeling. As template for the model, the latest X-ray structure of bovine rhodopsin (Protein Data Bank accession 1U19; Okada et al., 2004) with a resolution of 2.2 Å was used. The sequences of the human M2 receptor and the bovine rhodopsin were extracted from SwissProt [codes P08172 (Bonner et al., 1987) and P02699 (Nathans and Hogness, 1983), respectively]. The sequence alignment was carried out with ClustalW (Thompson et al., 1994) as well as based on the pinpoints identified by Baldwin et al. (1997). Transmembrane regions of the M2 receptor were detected using several secondary structure prediction methods. The extracellular and intracellular loops and the N terminus, respectively, were created by the application of a loop search routine based on an -carbon distance-matrix as implemented in the Homology module of Insight II 2000 (Accelrys, San Diego, CA). The three-dimensional coordinates for the C terminus were added in analogy to the X-ray structure of bovine rhodopsin.
Receptor-Ligand Complex. The orthosteric ligand N-methylscopolamine and the allosteric agent W84 were manually docked. The W84 conformation used in the docking procedure was the one that had been detected as the most favorable one in the course of a dynamic conformational search in aqueous environment (Voigtländer et al., 2003). Two steps were used to reach a good starting geometry for the following molecular dynamics simulation. First, the free volume located between the helices and the extracellular loops was calculated using the program SURFNET (Laskowski, 1995). In the second step, the large cavities were analyzed using GRID interaction fields (Goodford, 1985). Different GRID probes were applied to mimic the functional groups present in the ligands; afterward, the ligands were manually docked according to the favorable positions detected by GRID.
Molecular Dynamics Simulation. For further refinement and structure validation, molecular dynamics simulations were carried out using the software package GROMACS (Lindahl et al., 2001). For this purpose, the model was embedded in a phospholipid bilayer with aqueous phases containing Na+ and Cl- as counter-ions both extra- and intracellularly. Schlegel et al. (2005) recently described this procedure. Position restraints were initially set on the ternary complex with a force of 5000 kJ · mol-1 · nm-2 to equilibrate the membrane and solvent molecules. Afterward the position restraints were slowly reduced in 10 steps of 100 ps each until an unconstrained dynamics simulation over a period of 2500 ps was carried out. Frames were written out for every 2 ps.
Model Quality Check. The molecular dynamics simulation was checked for equilibration of the protein and its stability. The intramolecular interaction energy of the protein and the root-mean-square deviation of the entire and helical backbone and of W84, respectively, were taken as measures for complex equilibration. The program g_cluster implemented in GROMACS was applied to attain a representative structure from the trajectory after equilibration. This structure was minimized using the steepest descent algorithm and the protein geometry was checked with PROCHECK (Laskowski et al., 1993).
Fig. 2. Schematic presentation of the second and third outer loop (o2, o3) of the M2 receptor along with the flanking -helical regions. The pictograph in between shows the respective positions of the loops in the M2 receptor protein. The adjacent -helical domains were set according to the M2 receptor model and referred to as transmembrane domains (TM). Amino acids mutated in this study are highlighted.
In this study, we aimed at characterizing the role for allosteric action of two conserved tryptophans located in the essential region of the allosteric binding domain of the M2 receptor and the possible interplay with the spatially adjacent amino acids M2177Tyr and M2423Thr (Fig. 2). To probe the influence of these M2 and the respective M5 amino acids on the binding of allosteric modulators, these residues were either replaced by the corresponding amino acids of the counterpart receptor subtype (M2/M5 or M5/M2 substitution) or by alanine as a neutral residue. M2/M5 substitution and M5/M2 substitution, respectively, provide insight into the contribution of an amino acid of interest relative to its counterpart amino acid to the high affinity of the allosteric agents at M2 compared with M5, whereas substitution against alanine serves to elucidate the role of the amino acid by itself and was applied throughout in the case of conserved residues.
We first investigated the effect of point mutations on the formation of ternary complexes, which is the main characteristic of allosteric interactions. For this purpose, we employed dissociation binding experiments, using [3H]N-methylscopolamine ([3H]NMS) as an orthosteric ligand. The observed alteration, generally a delay, of the [3H]NMS dissociation rate is strictly indicative of an allosteric action, because it results from an interaction of the modulator with receptors whose orthosteric site is occupied by [3H]NMS. The half-times of [3H]NMS dissociation at the diverse wild-type and mutant receptors in the absence of allosteric agents are shown in Table 1. Dissociation half-times at the M2 receptors ranged from 5.6 min to 56 min. At M5, dissociation was much slower than at M2, which is in accordance with data presented previously (Ellis et al., 1993; Buller et al., 2002; Voigtländer et al., 2003; Huang et al., 2005). Mutation-induced changes of the half-times are mentioned in the subsequent paragraphs.
TABLE 1 [3H]NMS binding parameters in wild-type and mutant M2 and M5 receptors
Data represent mean values ± S.E. of three or more independent experiments or mean values of two independent experiments along with the single values, respectively.
M2422Trp and the Corresponding M5477Trp Confer Sensitivity to W84 Independent of Whether the Adjacent Downstream Amino Acid Is Threonine. The previous M2 receptor model predicted the conserved residue M2422Trp to interact directly with W84, whereas the adjacent amino acid M2423Thr relative to the corresponding M5478His was suspected to be essential for the proper orientation of the side chain of this tryptophan (Voigtländer et al., 2003). If so, replacement of M2422Trp by alanine should reduce the potency of W84. Moreover, substitution of M2423Thr by the corresponding M5478His, which was suspected to be unfavorable, should diminish the role of M2422Trp for the binding of W84. In the M5 receptor, however, mutation of M5477Trp should have less effect on the binding of W84 if the absence of a neighboring threonine causes an unfavorable conformation of 477Trp in M5. To check this concept, we generated mutant M2 and M5 receptors. The effects of the substitutions in M2 on the allosteric potency of W84 are illustrated in Fig. 3. As anticipated, mutation of M2422Trp to alanine led to a pronounced loss of potency of W84, compared with M2 wild type; the pEC0.5,diss-value at the M2422WA mutant was reduced by -1.51 log units (Table 2). The double mutant M2422WA+423TH displayed a further decrease of sensitivity for W84, compared with the single mutant M2423TH. This finding was unexpected, because the absence of M2423Thr was thought to induce an unfavorable orientation of the adjacent M2422Trp and thus to diminish its contribution to the potency of W84. It is noteworthy that this unexpected further reduction of potency (-1.47 log units relative to M2423TH) was almost equal to the loss of sensitivity observed at the M2422WA mutant relative to M2 (-1.51 log units). The effects of both single mutations were additive and thus seem to be independent of each other. At M5 (data compiled in Table 2), substitution of the corresponding tryptophan 477Trp by alanine resulted in an impressive loss of potency of W84 by -2.17 log units. This was also unexpected, because the presence of M5478His instead of threonine was thought to force M5477Trp into an unfavorable orientation. As shown previously (Voigtländer et al., 2003), replacement of the neighboring M5478His by the corresponding threonine of M2 increases W84's potency by 0.77 log units compared with M5. Additional mutation of 477Trp in the M5478HT mutant reduced the potency of W84 by -0.78 log units relative to M5; compared with the M5478HT mutant, affinity was diminished by -1.55 log units. This value is similar to the affinity shift found in M5477WA relative to M5 wild type (-2.17 log units). Taken together, the findings indicate that the conserved tryptophans M2422Trp and M5477Trp, respectively, are important for the binding of W84 irrespective of whether a neighboring threonine is present or not. Thus, in contrast to the previous concept, M2423Thr does not seem to induce a proper orientation of the adjacent 422Trp for W84 binding.
Fig. 3. Role of M2422Trp in the absence and presence of M2423Thr for the interaction of W84 with the NMS-occupied M2 receptor. Ordinate, apparent rate constant of [3H]NMS dissociation, expressed as percentage of the control value. Abscissa, log concentration of W84. Indicated are mean values ± S.E. of three to four separate experiments. The curve without data points is replotted from Voigtländer et al. (2003).
TABLE 2 Potencies (pEC0.5,diss) of W84 to retard [3H]NMS dissociation from the indicated wild-type and mutant receptors
pEC0.5,diss indicate the difference between the pEC0.5,diss values for the mutant and for the respective wild-type receptor. Values are means ± S.E. of three to five separate experiments.
M5478His Hinders NMS Binding Kinetics. The aforementioned data showed that the loss of potency of W84 observed at the M2423TH mutant is not caused by an impaired orientation of the adjacent tryptophan. To further elucidate the role of this residue, 423 in M2 and 478 in M5, respectively, we introduced alanine instead of the respective M2 and M5 amino acid. The binding parameters of [3H]NMS for wild-type and mutant receptors are given in Table 1. Whereas replacement of M2423Thr by the corresponding amino acid of M5, histidine, led to a 3-fold increase of the [3H]NMS dissociation half-time, insertion of alanine instead of threonine had only a minor effect on the [3H]NMS dissociation. Conversely, substitution of the corresponding residue 478His in M5 by either alanine or threonine reduced the NMS dissociation half-time by approximately 2- to 3-fold. Because none of these mutations had much effect on the binding affinity of [3H]NMS (pKD values in Table 1), compared with the respective wild-type receptor, it can be concluded that the association of NMS at these mutants was affected to the same extent as the dissociation. In conclusion, 423Thr is located near the entrance of the orthosteric site of M2 but is itself rather unimportant for the kinetics of [3H]NMS binding at M2, whereas histidine at the corresponding position either in M2 or in M5 seems to be an obstacle for the passage of NMS to and from the orthosteric binding pocket.
Low Potency Binding of W84 at the M2423TH Mutant and at M5 Wild Type Is Due to a Detrimental Effect of Histidine. Because M5478His has been shown in the preceding paragraph to slow NMS binding kinetics and is part of the allosteric site, we hypothesized that the histidine at this position might be an obstacle for the binding of W84 to the NMS-liganded receptor. Concentration-effect curves, displaying the allosteric interaction of W84 with the different receptor mutants, are shown in Fig. 4. It has been reported previously (Voigtländer et al., 2003) that insertion of histidine in M2 instead of 423Thr lowers the potency of W84 (-0.79 log units; Fig. 4, Table 2). The present findings show that replacement of M2423Thr by alanine does not affect the potency of W84 compared with the wild-type receptor. The corresponding mutations in M5, M5478HA and M5478HT, both resulted in an equal increase in sensitivity for W84 (0.70 and 0.77 log units). We conclude that the beneficial role of M2423Thr for the M2/M5 selectivity of W84 is not an "active" contribution to W84 binding but "merely" reflects the replacement of an affinity lowering amino acid (i.e., M5478His).
Fig. 4. Role of M2423Thr and the corresponding M5478His for the interaction of W84 with NMS-liganded M2 and M5 receptors. For sake of comparison, effects of substitution of M2423Thr and M5478His by the corresponding amino acids of M5 or M2, respectively, are also shown. Indicated are mean values ± S.E. of three to four separate experiments. Curves without data points are replotted from Voigtländer et al. (2003).
In the Case of Diallylcaracurine V, M2423Thr Enhances Potency Directly. In the case of diallylcaracurine V, the former M2 receptor model proposed a direct interaction between M2423Thr and the allosteric agent (Voigtländer et al., 2003). In that case, replacement of this M2 residue by alanine should result in a loss of the allosteric potency. Figure 5 displays the sensitivities of the different receptors for diallylcaracurine V. Replacement of M2423Thr by either alanine or histidine reduced the modulator's potency to the same extent (-0.66 and -0.68 log units, respectively; Table 3). The corresponding M5 mutants M5478HA and M5478HT exhibited significantly higher sensitivity for diallylcaracurine V than the wild-type receptor. It is noteworthy that diallylcaracurine V benefits to a greater extent from the replacement of M5478His by threonine (0.98 log units) than by alanine (0.58 log units). These findings support the notion that M2423Thr has a direct beneficial effect for the interaction of diallylcaracurine V with the allosteric site.
Fig. 5. Role of M2423Thr and the corresponding M5478His for the interaction of diallylcaracurine V with NMS-liganded M2 and M5 receptors. For sake of comparison effects of substitution of M2423Thr and M5478His by the corresponding amino acids of M5 or M2, respectively, are also shown. Indicated are mean values ± S.E. of three to four separate experiments. Curves without data points are replotted from Voigtländer et al. (2003).
TABLE 3 Potencies (pEC0.5,diss) of diallylcaracurine V to retard [3H]NMS dissociation from the indicated wild-type and mutant receptors
pEC0.5,diss indicates the difference between the pEC0.5,diss values for the mutant and for the respective wild-type receptor. Values are means ± S.E. of three to five separate experiments.
Kinetics of NMS Binding at M2 Are Accelerated by the Conserved Tryptophan M2427Trp, Whereas the Binding of W84 Does not Depend on This Residue. In the M2 receptor, there is a second conserved tryptophan next to M2422Trp at position 427 that might be involved in the binding of W84. To check this assumption, we introduced alanine instead of M2427Trp. We also created the double mutant M2422WA+427WA. Table 1 indicates the consequences of these mutations on the binding of [3H]NMS in the absence of an allosteric agent. NMS dissociation was remarkably retarded by the substitution of M2427Trp by alanine (5-fold), whereas the mutation of M2422Trp to alanine increased the dissociation half-time only slightly (2-fold). Both effects were nearly additive, in that the NMS dissociation half-time was increased by a factor of 8 at the double mutant. The pKD-value of [3H]NMS was reduced at most by a factor of 3 by the aforementioned mutations (Table 1). Thus we conclude that [3H]NMS association at these receptors was also retarded. Therefore, M2427Trp and, to a minor degree, M2422Trp seem to be important for speeding up the access of NMS to and its egress from the orthosteric binding pocket of the M2 receptor.
In the case of W84, as mentioned above, substitution of M2422Trp by alanine led to a pronounced loss of the modulator's allosteric potency. In contrast to this, replacement of M2427Trp by alanine did not change the potency of the modulator significantly compared with the wild-type receptor (Fig. 6, top; Table 4, left). Therefore, the double mutant M2422WA+427WA displayed the same reduction of sensitivity for W84 as the single mutant M2422WA. Hence, only M2422Trp conveys allosteric potency to W84, whereas M2427Trp does not seem to be involved in the binding of the modulator (although it does alter the cooperativity between W84 and NMS quite strikingly, as discussed below).
Fig. 6. Allosteric effects of W84 at the indicated wild type and point mutated M2 receptors. Top, concentration-effect curves for the allosteric delay of [3H]NMS dissociation. Bottom, effects on the equilibrium binding of [3H]NMS. Ordinate, specific [3H]NMS binding as a percentage of the control value. Abscissa, log concentration of W84. Experiments were conducted and analyzed as described under Materials and Methods. Indicated are means ± S.E. of three to five separate experiments. Data points for lower concentrations at M2 wild type and M2427W A were all on the control level and are not shown for the sake of identical concentration ranges in both panels.
TABLE 4 Interaction of the indicated allosteric agents with wild-type M2 receptors and related mutants
Minus log values of the factors of cooperativity (p) are indicated in parentheses on the right of the table. Indicated are mean values ± S.E. of three to ten separate experiments with duplicated values.
M2422Trp Plays a Crucial Role for Other Allosteric Modulators at the NMS-Occupied M2 Receptor. We also explored the role of both conserved tryptophans in M2 receptors for the other allosteric agents (Fig. 1). We chose two additional representatives of alkane-bisammonio type agents [i.e., dimethyl-W84, which has been developed as a radioalloster (Tränkle et al., 1998), and naphmethonium, which is an enhancer of NMS binding (Muth et al., 2003)]. Furthermore, we included gallamine, which is an archetypal allosteric agent (Clark and Mitchelson, 1976; Stockton et al., 1983), and diallylcaracurine V (Zlotos et al., 2000). Allosteric potencies of the modulators, expressed as pEC0.5,diss, are compiled in Table 4, left. The alkane-bisammonio type agents (dimethyl-W84 and naphmethonium) and gallamine were quite similar to W84 in terms of their patterns of epitope dependence at M2. That is, replacement of M2422Trp by alanine was always deleterious for the binding of the modulators, whereas mutation of M2427Trp had no effect. However, in contrast with W84, there was actually an increase in potency toward diallylcaracurine V when either M2422Trp or both tryptophans were substituted by alanine (Table 4, left). This gain in potency was caused mainly by the replacement of M2422Trp. It is intriguing that diallylcaracurine V seems to be exceptional in two aspects: 1) it depends directly on M2423Thr, and 2) it is negatively influenced by M2422Trp. Taken together, the amino acid 422Trp is critical, either in a positive or in a negative fashion, for allosteric agent binding at the NMS-occupied M2 receptor; depending on the agent, it may provide affinity or impair binding.
422Trp Is Often Critical for Allosteric Agent Binding in NMS-Free M2 Receptors. The results presented above refer to the formation of ternary complexes (i.e., the ability of allosteric test compounds to interact with NMS-occupied receptors). To gain insight into the role of the two conserved tryptophans for allosteric agent binding at NMS-free receptors, we carried out equilibrium binding experiments with [3H]NMS. These experiments also reveal the type of cooperativity between the allosteric agents and the orthosteric probe [3H]NMS. pKD values, indicating the binding affinity of [3H]NMS at the diverse wild-type and mutant receptors in the absence of allosteric modulator, are compiled in Table 1. Binding of [3H]NMS was hardly affected by mutations in M2 or by mutations in M5. The wild-type pKD was maximally changed by a factor of 3; this effect was found in receptors that contained the M2422WA mutant. Figure 6, bottom, shows the effect of W84 on the equilibrium binding of [3H]NMS at M2 and at the tryptophan M2 receptor mutants. Affinity values (pKA) for the allosteric agents at the unoccupied receptors and measures of cooperativity with [3H]NMS (p) were derived from curve fitting based on the ternary complex model of allosteric interactions as outlined in Materials and Methods and are displayed in Table 4, right. W84 revealed a negative cooperativity with [3H]NMS at M2 and in higher concentrations at the M2422WA mutant, too (Fig. 6, bottom). Thus, replacement of M2422Trp by alanine led to a pronounced reduction of the binding affinity of W84 for the NMS-free receptors (Table 4, right). When M2427Trp was replaced by alanine, [3H]NMS equilibrium binding remained unaffected by W84 at concentrations which fully inhibited the dissociation kinetics of [3H]NMS binding (Fig. 6, top). Thus, the lack of effect of W84 on [3H]NMS equilibrium binding reflects neutral cooperativity between these ligands. The change from negative (M2 wild type) to neutral cooperativity at the M2427WA mutant is accompanied by a small loss of affinity of W84 to the NMS-free receptor mutant relative to the NMS-free wild-type receptor (Table 4, pKA values, right).
The affinity of naphmethonium, an enhancer of the binding of [3H]NMS at M2 (curves not shown), was considerably reduced at the M2422WA mutant, whereas the mutant M2427WA showed the same sensitivity for naphmethonium as the NMS-free wild-type M2 receptor (Table 4, right). Naphmethonium continued to be positively cooperative with NMS at both singly and doubly mutated receptors.
Gallamine revealed a significantly reduced binding affinity both at the NMS-free M2422WA mutant and at the M2427WA mutant relative to M2, although the effects were not additive (Table 4, right). The underlying [3H]NMS inhibition curves (not shown) were rather steep (curve slopes for M2422WA, nH = -2.46 ± 0.38; for M2427WA, -2.33 ± 0.30; means ± S.E., n = 4-5). p-values indicate a strong negative cooperativity between gallamine and [3H]NMS at any M2 receptors tested in this study. We have no explanation why the sum of p and pKA was often considerably different from pEC0.5,diss; further investigation is needed to understand this phenomenon. In the case of diallylcaracurine V, there was no significant change of binding affinity at the NMS-free receptor mutants relative to the M2 wild-type receptor. Taken together, the role of the tryptophans at NMS-free M2 receptors is more complex than at the NMS-liganded M2 receptor, but in general M2422Trp is more important for allosteric agent binding than M2427Trp.
In M5 Receptors, the Tryptophan Corresponding to M2422Trp Also Plays a Key Role for Allosteric Agent Action. Because M2422Trp and M2427Trp are conserved among all five muscarinic receptor subtypes, we aimed to find out in how far the corresponding tryptophans M5477Trp and M5482Trp are also involved in the action of allosteric agents at this subtype. Therefore, we replaced these tryptophans by alanine. [3H]NMS binding parameters under control conditions are given in Table 1. Similar to the results obtained with M2 receptors, the binding affinity toward NMS was hardly changed; with the double mutant M5477WA+482WA, a small loss of affinity by a factor of 2 was seen. However, unlike M2, the kinetics of NMS dissociation remained almost unchanged; only in M2482WA, k-1 was increased by a factor of 1.4 relative to M5 wild type. It was remarkable that the double mutation, which slowed dissociation of NMS by 8-fold in M2, had no effect at all in M5. With regard to allosteric agent binding, we first investigated the influence of both residues at NMS-occupied receptors. As mentioned above, replacement of M5477Trp by alanine resulted in a pronounced reduction of potency of W84 (Fig. 7, top). Furthermore, we found a pronounced loss of allosteric efficacy of W84; i.e., dissociation of the orthosteric ligand could hardly be inhibited, even in the presence of highest concentrations of the allosteric modulator (lower plateau at 87%; maximum effects found in this study are compiled in Supplemental Table 1 online). This suggests that the egress of NMS from the ternary complex is considerably facilitated when the aromatic residue of M5477Trp is absent. Essentially the same findings were made with the double mutant, whereas the M5482WA mutant had the same sensitivity for W84 as the M5 wild type receptor. The allosteric actions of dimethyl-W84 and naphmethonium on [3H]NMS dissociation were likewise affected by mutation of the two tryptophans (Table 5, left). However, at the M5477WA mutant, both modulators revealed a more pronounced efficacy than W84 to inhibit NMS dissociation with maximum reduction of the apparent k-1 value to a level of 73% (dimethyl-W84) and 57% (naphmethonium) (curves not shown). It is noteworthy that gallamine had no visible effect up to a concentration of 3 mM on the NMS dissociation at this mutant. Furthermore, the M5482WA mutant exhibited significantly lower sensitivity for gallamine than the wild-type receptor. Diallylcaracurine V showed no change of potency to retard [3H]NMS dissociation at the diverse M5 receptor mutants.
Fig. 7. Allosteric effects of W84 at the indicated wild type and mutant M5 receptors. Top, concentration-effect curves for the allosteric delay of [3H]NMS dissociation. Bottom, effects on the equilibrium binding of [3H]NMS. Indicated are means ± S.E. of three to five separate experiments. Data points for lower concentrations at M5482WA and M5477W A+482W A were all on the control level and are not shown for the sake of identical concentration ranges in both panels.
TABLE 5 Interaction of the indicated allosteric agents with wild type M5 receptors and related mutants
Minus log values of the factors of cooperativity (p) are indicated in parentheses on the right of the table. Indicated are mean values ± S.E. of three to ten separate experiments with duplicated values.
In equilibrium binding experiments, W84 had no effect on the binding of [3H]NMS at M5 wild type and at the M5482WA mutant; i.e., the cooperativity was neutral (Fig. 7, bottom). In contrast, replacement of M5477Trp by alanine resulted in a pronounced inhibition of NMS equilibrium binding by W84; i.e., the cooperativity was negative. The pKA values shown in Table 5 reveal that the NMS-free mutant receptor M5477WA and the double mutant, but not M5482WA, displayed significantly lower sensitivity for W84 than M5 wild type, but this loss of sensitivity was less pronounced than at the NMS-occupied receptor. Therefore, W84 exhibited strong negative cooperativity at these mutants (Table 5, p-values, right). In the case of gallamine, only substitution of 477Trp affected the agent's binding affinity to the NMS-free M5 receptors (Table 5, right). The effect of gallamine on [3H]NMS equilibrium binding (i.e., the extent of negative cooperativity) was left almost unaffected by the diverse mutations. Diallylcaracurine V again hardly revealed any affinity-shift at the mutants compared with M5 wild type (Table 5, right), but a change toward weaker negative cooperativity was noted when M5477Trp was substituted.
Taken together, in the NMS-bound as well as in the NMS-free M5 receptor, 477Trp is important for the receptor's sensitivity to allosteric agents.
The High-Potency Interaction between W84 and M2 Is Mediated to a Great Extent by M2177Tyr and M2422Trp. Preceding studies (Buller et al., 2002; Voigtländer et al., 2003; Huang et al., 2005) suggested that the subtype-selective amino acid M2177Tyr interacts directly via - interaction with W84. Here, we identified the conserved M2422Trp as a second amino acid that contributes to a major extent to the allosteric potency of W84. To determine whether the contribution of both amino acids for the action of W84 is additive, we generated the double mutant M2177YQ+422WA and measured the effect of W84 on [3H]NMS-liganded receptors (Fig. 8). As mentioned above, both single mutants M2177YQ (Voigtländer et al., 2003; pEC0.5,diss, 6.20 ± 0.05) and M2422WA showed clearly reduced sensitivities for W84 compared with M2 wild type. Combined replacement of both residues led to a further decrease of the potency of W84 (pEC0.5,diss, 4.97 ± 0.04). The effects of the single mutants were almost additive. Hence, approximately 2.5 log units of allosteric potency of W84 at M2 can be attributed to only two amino acids (i.e., M2177Tyr and M2422Trp).
Fig. 8. Additive contribution of M2177Tyr and the conserved M2422Trp to the allosteric potency of W84. Indicated are means ± S.E. of three to five independent experiments. The curve without data points is taken from Voigtländer et al. (2003).
Molecular Modeling of the W84 Docking to the M2 Receptor. To gain more insight into the topography of allosteric agent binding to the M2 receptor, a three-dimensional model of the receptor protein was built (Jöhren and Höltje, 2002; Voigtländer et al., 2003). The availability of an actual and revised bovine rhodopsin X-ray structure showing higher resolution (Okada et al., 2004) as well as progress in computational techniques made it necessary to generate a new and improved M2 receptor model. In contrast to the preceding model, the receptor is now embedded in a phospholipid bilayer composed of dipalmitoylphosphatidylcholine with aqueous phases containing sodium ions and chloride ions as counter-ions extra- and intracellularly (Fig. 9). The new model was not only in good agreement with previous results but also seemed to be more suitable to explain experimental data reported previously (Voigtländer et al., 2003). Because the template of this model, the X-ray structure of bovine rhodopsin, appears in its inactive state, the resulting geometry of the homology model should also be in this state and should thus be appropriate to simulate an antagonist-(NMS-) liganded receptor. As described previously (Voigtländer et al., 2003; Tränkle et al., 2005) the orthosteric binding site is located between the transmembrane helices TM3-TM6 and is connected through a narrow channel with the allosteric binding site which is formed by the extracellular loops.
Fig. 9. Representative structure from the molecular dynamics simulation. Side view of the M2 receptor in its membrane environment. Protein: blue ribbon. The orthosteric ligand NMS and the allosteric agent W84 (above NMS) are shown as solid surfaces. W84/NMS: carbon, orange. Phospholipid bilayer: carbon, green. Aqueous phase: water, blue; sodium, green; chloride, magenta. Color code of the other atoms: hydrogen, white; nitrogen, blue; oxygen, red; phosphorus, magenta.
To model the ternary complex characteristic for allosteric interactions, NMS was placed in the orthosteric site. To obtain a realistic low-energy conformation for the highly flexible W84, molecular dynamics simulations were carried out in an aqueous environment as reported previously (Voigtländer et al., 2003). The conformation resulting from the dynamical treatment of W84 was docked to the NMS-occupied receptor. The system showed a stable geometry during the molecular dynamics simulation. The representative structure from the molecular dynamics simulation was of good quality. The Ramachandran plot, calculated by PROCHECK, showed no disallowed residues near the orthosteric or allosteric binding site. The position of the orthosteric ligand NMS is consistent with the interaction fields detected by the different probes implemented in GRID. W84 fills the cavity (Fig. 10A, white transparent silhouette) between the extracellular loops to a great extent. It is remarkable that the phthalimide group diving down into the ligand binding cavity of the receptor (Fig. 9) is positioned between Tyr177 and Trp422 (Fig. 10A), almost in a sandwich-like manner. This position would allow the formation of - interactions between the aromatic residues of the two amino acids and the phthalimide moiety of W84. Hence, an interaction of W84 with both residues at the same time is possible. The narrow corridor between the orthosteric site and the allosteric site is closed and NMS (partially visible in Fig. 10A, bottom) cannot dissociate from the M2 receptor.
Fig. 10. A, M2 wild type: interaction between W84 and the adjacent amino acids M2177Tyr, M2422Trp, and M2423Thr; view from the top of the receptor protein occupied by W84 and NMS (partially visible at the bottom of the figure). Helices, red; o2, cyan; disulfide bridge, magenta; other, gray. W84/NMS: carbon, orange; amino acids: carbon, green. Other atoms: hydrogen, white; nitrogen, blue; oxygen, red. Volumes of the binding sites: white, transparent silhouette. Note the sandwich-like arrangement of M2177Tyr, the phthalimide residue of W84 and M2422Trp. B, M2423T H mutant: Color code as above; 423His: all atoms yellow. Volumes of the binding sites: white, transparent silhousette. Note the unfavorable spatial orientation of histidine in M2423T H.
In some instances, even maximally effective concentrations of an allosteric agent did not induce complete inhibition of [3H]NMS dissociation (bottom k-1 significantly higher than k-1 = 0%). This phenomenon is known for some allosteric agents such as obidoxime and hexamethonium even at wild-type receptors (e.g., Tränkle and Mohr, 1997; Tränkle et al., 2005). We used the double mutant M2422WA+423TH, in which W84 reduced the probability of [3H]NMS-dissociation at maximum only 4-fold (bottom k-1 = 24.7%; see Supplemental Table 1 online), to find out whether the model gives a clue about why [3H]NMS dissociation is not completely prevented. A molecular dynamics simulation revealed a higher flexibility of the outer loops, especially o3, and a movement of W84 away from the critical cluster of amino acids. The receptor did not stabilize during the relatively long 6000-ps simulation. A free volume (in the sense of a channel allowing escape of NMS from the W84-occupied receptor) was initially formed but did not persist permanently during the simulation.
Figure 10A shows that M2423Thr lines W84 lying in the allosteric cavity. To gain more insight into the postulated detrimental effect of the mutation M2423TH on W84 binding (see Fig. 4), M2423Thr was replaced in the model by histidine. As shown in Fig. 10B, the histidine residue protrudes into the allosteric binding cavity yielding a reduction of space. Thus, M5478His seems to decrease the binding affinity of W84 sterically. Furthermore, the slowing of the [3H]NMS binding kinetics by histidine (see Table 1) could be nicely explained by its protrusion into the passage used by NMS to reach and leave the orthosteric site. In conclusion, the results of the docking and molecular dynamics simulation carried out with the current three-dimensional model correspond very well with the experimental findings from the binding assays in the wild-type M2 and the mutant M2 receptors.
Previous studies using the pronounced M2/M5 receptor subtype selectivity of muscarinic allosteric modulators have identified a domain near the entrance of the orthosteric ligand binding cavity that is essential for the high-affinity binding of various allosteric modulators to muscarinic M2 relative to M5 receptors (Ellis and Seidenberg, 2000; Buller et al., 2002; Jöhren and Höltje, 2002; Voigtländer et al., 2003; Huang et al., 2005; Tränkle et al., 2005). This domain is lined by the second outer loop (o2) containing M2177Tyr and parts of the third outer loop at the beginning of the seventh transmembrane region (o3/TM7) containing M2423Thr (Fig. 2) (Ellis and Seidenberg, 2000; Jöhren and Höltje, 2002). The M2/M5 mutagenesis approach, however, cannot give direct insight into the contribution of conserved amino acids to allosteric agent binding that provide subtype independent "baseline" affinity for allosteric agents. Lipophilic pockets lying apart from the above-mentioned essential domain (i.e., o2 and o3/TM7) and containing highly conserved residues were speculated to provide such baseline affinity (Voigtländer et al., 2003). In the present study, however, we have identified a conserved amino acid, 422Trp in M2 and 477Trp in M5, respectively, that plays a key role for the binding of allosteric agents and that is a key component in a pivotal cluster of amino acids additionally to M2177Tyr and M2423Thr, instead of being located outside the essential domain. Furthermore, we provide evidence that the adjacent M2423Thr may sometimes be beneficial for allosteric agent affinity, not because of a positive effect of its own but by lack of a negative spatial influence, as exerted by the corresponding histidine of M5.
The molecular dynamics simulations of the docking of W84 to the allosteric site of M2 receptors, whose orthosteric site is blocked by an antagonist, indicate that one of its lateral phthalimido groups is enclosed by the aromatic residues of M2177Tyr and M2422Trp in a sandwich-like manner (Fig. 10A). Forming - interactions, these aromatic amino acids, like a pair of tongs, fix the aromatic phthalimido group between them. According to this model, the phthalimide moiety of W84 achieves both baseline affinity by 422Trp and M2/M5 subtype selectivity by 177Tyr at the same time. The contributions of these amino acids to W84 affinity were nearly additive. In this context, it may be mentioned that the substitution of M2177Tyr either by glutamine of M5 or by alanine (Huang et al., 2005) yielded almost the same loss of W84 binding affinity. The additivity of the contributions of M2177Tyr and M2422Trp to W84 binding affinity suggests that both amino acids interact independent of each other with the allosteric agent. Substitution of M2422Trp by alanine reduced binding affinity of the alkane-bisammonio compounds by approximately 1.5 log units and of gallamine by approximately 1 log unit. This suggests that the contribution to affinity of this residue depends on the type of allosteric agent. In the case of diallylcaracurine V, M2422Trp is even detrimental for this agent's affinity. In any case, these findings support the concept of a central role of M2422Trp for allosteric agent binding. In addition, in the NMS-occupied M5 receptor, the corresponding tryptophan (M5477Trp) is important for the binding of all applied alkane-bisammonio compounds. In the case of gallamine, the lack of effect on [3H]NMS dissociation indicates either a pronounced loss of gallamine's affinity or a complete loss of gallamine's efficacy to inhibit [3H]NMS dissociation. In any case, this tryptophan is essential for gallamine's action at M5. With diallylcaracurine V, however, M5477Trp did not affect affinity, in contrast to the corresponding M2422Trp.
Most of the allosteric modulators tested in this study displayed a pronounced sensitivity to mutation of both M2422Trp and M5477Trp at the NMS-occupied receptor and also at the NMS-free receptor. In the search for the binding location of gallamine at the M1 receptor, Matsui et al. (1995) mutated more than 20 residues in the extracellular region of the receptor and found that replacement of the corresponding M1400Trp by alanine reduced gallamine's affinity by approximately 10-fold in NMS-occupied and NMS-free receptors. These results are quite similar to our findings at M2422Trp and M5477Trp. The similarity suggests that this conserved tryptophan may contribute to the baseline affinities of many allosteric agents for the muscarinic receptor family (diallylcaracurine V is certainly one exception to this rule).
Closer inspection of our findings reveals that this tryptophan, although conserved among subtypes, may not only provide subtype-independent baseline affinity but also contribute to subtype selectivity. For example, in the binding of both W84 and gallamine to NMS-free receptors, substitution of M5477Trp by alanine resulted in a lesser reduction of affinity than did replacement of M2422Trp by alanine. Therefore, this tryptophan seems to contribute to the M2 preference of some allosteric agents. On the contrary, the subtype selectivity of diallylcaracurine V at the NMS-occupied receptor is reduced as a result of the detrimental effect of M2422Trp, because this ligand is insensitive to mutation of M5477Trp. In other words, the M2/M5 selectivity of diallylcaracurine V amounts to 2.4 log units in wild-type receptors, but to 2.9 log units in the corresponding M2422/M5477Trp mutants.
Furthermore, our findings reveal that M2422Trp, as well as its counterpart M5477Trp, may regulate cooperativity between allosteric and orthosteric ligands. Substitution of M5477Trp by alanine resulted in a switch from neutral to strongly negative cooperativity between W84 and NMS (Tables 4 and 5, p values). Moreover, in the case of diallylcaracurine V, mutation of M2422Trp and the corresponding M5477Trp yielded enhanced positive cooperativity at M2 and tempered negative cooperativity at M5, respectively.
In contrast to M2422Trp and its counterpart M5477Trp, the downstream conserved tryptophans M2427Trp and M5482Trp generally have no relevance for the binding and action of allosteric agents. Exceptions were encountered with the M2427Trp Ala mutant that clearly reduced the affinity of gallamine for the free receptor and that switched cooperativity between W84 and NMS from negative to neutral.
Taken together, the conserved tryptophan in position 422 of M2 and 477 of M5 located at the junction of o3 and the beginning of TM7 has to be taken into account when the epitope dependence of cooperative interactions is analyzed in muscarinic receptors. Using the approach of site-directed mutagenesis and the construction of hybrid receptors Krejí and Tuek (2001) showed that the second and the third outer loop of the M2 receptor protein are important for the cooperative interaction of alcuronium and gallamine with [3H]NMS. Voigtländer et al. (2003) found a large influence of M2177Tyr and M2423Thr on the cooperativity between NMS and diallylcaracurine V and a smaller effect in the case of dimethyl-W84. Studying the allosteric interactions of strychnine-like modulators, Jakubik et al. (2005) reported an important role for the M2 sequence of the third outer loop and especially of three amino acids (M2419Asp, M2421Val, and M2423Thr) that are adjacent to M2422Trp. We now provide evidence for an involvement of this tryptophan and its counterpart M5477Trp in the allosteric actions of several muscarinic allosteric modulators.
It is remarkable that the conserved tryptophan at position 422 of M2 is also involved in the binding of orthosteric ligands. We found that substitution of the tryptophan by alanine reduced the affinity of NMS for M2 receptors approximately 3-fold and slowed NMS dissociation by approximately 2-fold. In M5, however, the corresponding mutation did not reduce NMS affinity and hardly affected NMS binding kinetics. For M1, Matsui et al. (1995) showed that replacement of this tryptophan by alanine reduced [3H]NMS dissociation by a factor of 2 and hardly affected NMS binding affinity, but reduced acetylcholine binding affinity by a factor of 10. The new M2 model suggests that M2422Trp and M2423Thr are situated at the bottom of the allosteric binding cavity and concomitantly at the top of the orthosteric binding pocket. According to the M2 receptor model, both conserved tryptophans, M2422Trp and in particular M2427Trp, may frame the pathway leading from the external surface of the receptor protein with its allosteric binding cleft to the orthosteric binding pocket and thereby facilitate the passage of orthosteric agents. Because mutation of the corresponding tryptophans in M5 did not change NMS binding affinity and kinetics, the M5 receptor protein seems to possess a quite different overall conformation of this region compared with M2.
In conclusion, the present study demonstrates that the conserved tryptophan at position 422 in M2 is important for receptor interactions of structurally diverse allosteric agents in M2 and M5 receptors. Depending on the allosteric agent and the receptor, this epitope is important for baseline affinity, subtype selectivity, and cooperativity. Together with spatially closely related epitopes of o2 and o3/TM7, such as the M2/M5 selectivity providing epitopes M2177Tyr and M2423Thr, this tryptophan forms a pivotal docking point complex for allosteric agents.
Acknowledgements
It is gratefully indicated that dimethyl-W84, naphmethonium, and diallylcaracurine V were synthesized and provided by Prof. Dr. Ulrike Holzgrabe, Dr. Mathias Muth, and Dr. Darius Paul Zlotos, Institute of Pharmaceutical Chemistry, University of Würzburg, Germany. We are grateful to Dr. Christian Tränkle, Department of Pharmacology and Toxicology, Institute of Pharmacy, University of Bonn (Bonn, Germany), for valuable help with the data analysis.
ABBREVIATIONS: M2, M2 subtype of the muscarinic acetylcholine receptor; M5, M5 subtype of the muscarinic acetylcholine receptor; NMS, N-methylscopolamine; o2, second outer (extracellular) loop of the receptor; o3, third outer (extracellular) loop of the receptor; PB, sodium-potassium phosphate buffer (5 mM, pH 7.4); W84, hexamethylene-bis-[dimethyl-(3-phthalimidopropyl)ammonium]dibromide.
The online version of this article (available at http://molpharm.aspetjournals.org) contains supplemental material.
【参考文献】
Baldwin JM, Schertler GFX, and Unger VM (1997) An alpha carbon template for the transmembrane helices of the rhodopsin family of G-protein-coupled receptors. J Mol Biol 272: 144-164.
Birdsall NJ and Lazareno S (2005) Allosterism at muscarinic receptors: ligands and mechanisms. Mini Rev Med Chem 5: 523-543.
Bonner TI, Buckley NJ, Young AC, and Brann MR (1987) Identification of a family of muscarinic acetylcholine receptor genes. Science (Wash DC) 237: 527-532.[Abstract/Free Full Text]
Buller S, Zlotos DP, Mohr K, and Ellis J (2002) Allosteric site on muscarinic acetylcholine receptors: a single amino acid in transmembrane region 7 is critical to the subtype selectivities of caracurine V derivatives and alkane-bisammonium ligands. Mol Pharmacol 61: 160-168.[Abstract/Free Full Text]
Christopoulos A and Kenakin T (2002) G protein-coupled receptor allosterism and complexing. Pharmacol Rev 54: 323-374.[Abstract/Free Full Text]
Clark AL and Mitchelson F (1976) The inhibitory effect of gallamine on muscarinic receptors. Br J Pharmacol 58: 323-331.
Ehlert FJ (1988) Estimation of the affinities of allosteric ligands using radioligand binding and pharmacological null methods. Mol Pharmacol 33: 187-194.
Ellis J, Seidenberg M, and Brann MR (1993) Use of chimeric muscarinic receptors to investigate epitopes involved in allosteric interactions. Mol Pharmacol 44: 583-588.
Ellis J (1997) Allosteric binding sites on muscarinic receptors. Drug Dev Res 40: 193-204.
Ellis J and Seidenberg M (2000) Interactions of alcuronium, TMB-8 and other allosteric ligands with muscarinic acetylcholine receptors: studies with chimeric receptors. Mol Pharmacol 58: 1451-1460.
Gnagey AL, Seidenberg M, and Ellis J (1999) Site-directed mutagenesis reveals two epitopes involved in the subtype selectivity of the allosteric interactions of gallamine at muscarinic acetylcholine receptors. Mol Pharmacol 56: 1245-1253.[Abstract/Free Full Text]
Goodford PJ (1985) A computational procedure for determining energetically favorable binding sites on biologically important macromolecules. J Med Chem 28: 849-857.
Huang XP, Prilla S, Mohr K, and Ellis J (2005) Critical amino acid residues of the common allosteric site on the M2 muscarinic acetylcholine receptor: more similarities than differences between the structurally divergent agents gallamine and bis(ammonio)alkane-type hexamethylene-bis-[dimethyl-(3-phthalimidopropyl)ammonium]dibromide. Mol Pharmacol 68: 769-778.[Abstract/Free Full Text]
Jakubík J, Krejí and Doleal V (2005) Asparagine, valine and threonine in the third extracellular loop of muscarinic receptor have essential roles in the positive cooperativity of strychnine-like allosteric modulators. J Pharmacol Exp Ther 313: 688-696.[Abstract/Free Full Text]
J?hren K and H?ltje HD (2002) A model of the human M2 muscarinic acetylcholine receptor. J Comput Aided Mol Des 16: 795-801.
Krejí A and Tuek S (2001) Changes of cooperativity between N-methylscopolamine and allosteric modulators alcuronium and gallamine induced by mutations of external loops of muscarinic M3 receptors. Mol Pharmacol 60: 761-767.[Abstract/Free Full Text]
Krejí A, Michal P, Jakubík J, ín J and Doleal V (2004) Regulation of signal transduction at M2 muscarinic receptor. Physiol Res 53(Suppl 1): S131-S140.
Laskowski RA (1995) SURFNET: A program for visualizing molecular surfaces, cavities and intermolecular interactions. J Mol Graph 13: 323-328.
Laskowski RA, Macarthur MW, Moss DS, and Thornton JM (1993) PROCHECK: A program to check the stereochemical quality of protein structures. J Appl Cryst 26: 283-291.
Lazareno S and Birdsall NJ (1995) Detection, quantitation and verification of allosteric interactions of agents with labeled and unlabeled ligands at G protein-coupled receptors: interactions of strychnine and acetylcholine at muscarinic receptors. Mol Pharmacol 48: 362-378.
Lindahl E, Hess B, and van der Spoel D (2001) GROMACS 3.0: a package for molecular simulation and trajectory analysis. J Mol Model 7: 306-317.
Matsui H, Lazareno S, and Birdsall NJ (1995) Probing of the location of the allosteric site on M1 muscarinic receptors by site-directed mutagenesis. Mol Pharmacol 47: 88-98.
Mohr K, Tr?nkle C, and Holzgrabe U (2003) Structure/activity relationships of M2 muscarinic allosteric modulators. Recept Channels 9: 229-240.
Muth M, Bender W, Scharfenstein O, Holzgrabe U, Balatková E, Tr?nkle C, and Mohr K (2003) Systematic development of high affinity bis(ammonio)alkane-type allosteric enhancers of muscarinic ligand binding. J Med Chem 46: 1031-1040.
Nathans J and Hogness DS (1983) Isolation, sequence-analysis and intron exon arrangement of the gene encoding bovine rhodopsin. Cell 34: 807-814.
Okada T, Sugihara M, Bondar AN, Elstner M, Entel P, and Buss V (2004) The retinal conformation and its environment in rhodopsin in light of a new 2.2 ? crystal structure. J Mol Biol 342: 571-583.
Raasch A, Scharfenstein O, Tr?nkle C, Holzgrabe U, and Mohr K (2002) Elevation of ligand binding to muscarinic M2 acetylcholine receptors by bis(ammonio)alkane-type allosteric modulators. J Med Chem 45: 3809-3812.
Schlegel B, Sippl W, and Holtje HD (2005) Molecular dynamics simulations of bovine rhodopsin: influence of protonation states and different membrane-mimicking environments. J Mol Model 12: 49-64.
Schulz U (1998) Antagonismus-Untersuchungen zur Topographie der Bindung allosterischer Modulatoren an muskarinischen Acetylcholinrezeptoren des Subtyps M2. Doctoral Thesis, Faculty of Mathematics and Natural Sciences, University of Bonn.
Soudijn W, Van WI, and IJzerman AP (2004) Allosteric modulation of G protein-coupled receptors: perspectives and recent developments. Drug Discov Today 9: 752-758.
Stockton JM, Birdsall NJ, Burgen AS, and Hulme EC (1983) Modification of the binding properties of muscarinic receptors by gallamine. Mol Pharmacol 23: 551-557.
Thompson JD, Higgins DG, and Gibson TJ (1994) CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673-4680.[Abstract/Free Full Text]
Tr?nkle C, Dittmann A, Schulz U, Weyand O, Buller S, Johren K, Heller E, Birdsall NJ, Holzgrabe U, Ellis J, et al. (2005) Atypical muscarinic allosteric modulation: cooperativity between modulators and their atypical binding topology in muscarinic M2 and M2/M5 chimeric receptors. Mol Pharmacol 68: 1597-1610.[Abstract/Free Full Text]
Tr?nkle C, Mies-Klomfass E, Botero Cid MH, Holzgrabe U, and Mohr K (1998) Identification of a [3H]ligand for the common allosteric site of muscarinic acetylcholine M2 receptors. Mol Pharmacol 54: 139-145.[Abstract/Free Full Text]
Tr?nkle C and Mohr K (1997) Divergent modes of action among cationic allosteric modulators of muscarinic M2 receptors. Mol Pharmacol 51: 674-682.[Abstract/Free Full Text]
Tr?nkle C, Weyand O, Voigtl?nder U, Mynett A, Lazareno S, Birdsall NJ, and Mohr K (2003) Interactions of orthosteric and allosteric ligands with [3H]dimethyl-W84 at the common allosteric site of muscarinic M2 receptors. Mol Pharmacol 64: 180-190.[Abstract/Free Full Text]
Voigtl?nder U, J?hren K, Mohr M, Raasch A, Tr?nkle C, Buller S, Ellis J, H?ltje HD, and Mohr K (2003) Allosteric site on muscarinic acetylcholine receptors: identification of two amino acids in the muscarinic M2 receptor that account entirely for the M2/M5 subtype selectivities of some structurally diverse allosteric ligands in N-methylscopolamine-occupied receptors. Mol Pharmacol 64: 21-31.[Abstract/Free Full Text]
Wess J (2005) Allosteric binding sites on muscarinic acetylcholine receptors. Mol Pharmacol 68: 1506-1509.[Abstract/Free Full Text]
Zahn K, Eckstein N, Tr?nkle C, Sadée W, and Mohr K (2002) Allosteric modulation of muscarinic receptor signaling: alcuronium-induced conversion of pilocarpine from an agonist into an antagonist. J Pharmacol Exp Ther 301: 720-728.[Abstract/Free Full Text]
Zlotos DP, Buller S, Tr?nkle C, and Mohr K (2000) Bisquaternary caracurine V derivatives as allosteric modulators of ligand binding to M2 acetylcholine receptors. Bioorg Med Chem Lett 10: 2529-2532.
作者单位：Department of Pharmacology and Toxicology, Institute of Pharmacy, Rheinische Friedrich-Wilhelms-Universit?t Bonn, Germany (S.P., K.M.); Institute of Pharmaceutical and Medicinal Chemistry, Heinrich-Heine-Universit?t Düsseldorf, Germany (J.S., H.-D.H.); and Departments of Psychiatry and Pharmacology,