点击显示 收起
【关键词】 Mutational
The 2nd outer loop (o2) of muscarinic acetylcholine receptors (mAChRs) contains a highly conserved cysteine residue that is believed to participate in a disulfide bond and is flanked on either side by epitopes that are critical to the binding of many muscarinic allosteric modulators. We determined the allosteric binding parameters of the modulators gallamine, W84, and tetrahydroaminoacridine (THA) at M2 and M3 mAChRs in which these cysteine residues had been mutated to alanines. THA is known to bind to mAChRs with a strong positive homotropic cooperativity (a Hill slope of approximately 2) that implies that it must interact with multiple allosteric sites. The disulfide cysteine mutations in M2 receptors reduced the allosteric potencies of the tested modulators as if the critical adjacent residue (Tyr177) itself had been mutated. However, in M3 receptors, the disulfide cysteine mutations had no effect on the potencies of gallamine or W84 and even increased the potency of THA. It was most interesting that the strong, positive, homotropic interactions of THA at both M2 and M3 receptors were markedly reduced by the cysteine mutations. In addition, gallamine also displayed positive homotropic cooperativity in its interactions with M3 receptors (but not M2 receptors), and this cooperativity was not evident in the cysteine mutants. Thus, it seems that these cysteine residues play a role in linking cooperating allosteric sites, although it is not currently possible to say whether these multiple sites lie within one receptor or on two linked receptors of a dimer or higher order oligomer.
A growing number of G protein-coupled receptors (GPCRs) are known to possess extracellular allosteric binding sites that are topographically distinct from, but conformationally linked to, the orthosteric binding site for the endogenous agonist and its competitive ligands (Christopoulos and Kenakin, 2002). Binding of the allosteric modulator triggers conformational changes in the receptor that in turn modify the orthosteric binding site and/or receptor function. Allosteric modulators can have several favorable advantages as therapeutic agents over traditional orthosteric ligands, such as better subtype selectivity and safety and the ability to preserve spatiotemporal patterning in the CNS (Christopoulos, 2002). Therefore, they represent a new trend for drug design and development for many GPCRs, especially those for which highly selective orthosteric agonists or antagonists are not available yet, such as the muscarinic acetylcholine receptors (mAChRs). The family of mAChRs is one of the most intensively studied model systems for allosteric modulation, especially at the molecular level. A dramatic demonstration of this modulation is the retardation of the dissociation rate of orthosteric ligands, such as [3H]N-methyl scopolamine (NMS), by allosteric modulators, such as the prototypical mAChR allosteric modulators gallamine and W84 and the atypical modulator tetrahydroaminoacridine (THA/tacrine). THA is considered atypical among muscarinic allosteric modulators in that it slows the dissociation rate of [3H]NMS with a steep concentration-response curve that has a Hill slope of approximately 2 at all five subtypes of mAChRs (Potter et al., 1989; Ellis and Seidenberg, 2000), suggesting that it interacts with multiple allosteric sites that are cooperatively linked in a positive homotropic interaction. The molecular mechanism underlying this unusual allosteric interaction has recently been investigated further (Tränkle et al., 2003, 2005).
Four extracellular cysteine residues are highly conserved in the mAChR family (Fig. 1) and in most GPCRs. Two of these have been reported to form a disulfide bond between the top of the third transmembrane domain and the second outer (o2) loop of the M1 receptor (Curtis et al., 1989; Kurtenbach et al., 1990), as do the corresponding cysteine residues in rhodopsin (Karnik and Khorana, 1990; Palczewski et al., 2000). The other two cysteines are located in the short third outer (o3) loop of the receptor. These extracellular cysteine residues have been systematically characterized in site-directed mutagenesis studies in rat M1 (Savarese et al., 1992) and rat M3 receptors (Zeng and Wess, 1999; Zeng et al., 1999), as well as in many other GPCRs, for their roles in surface expression, orthosteric binding, and receptor functions. The disulfide cysteine residues (but not the o3 loop cysteine residues) have consistently been found to be involved in receptor folding, surface expression, and maintaining the structure of the orthosteric binding site. In addition, the disulfide cysteine residues in M3 receptors have been reported to be crucial to receptor dimerization, because mutations at these residues eliminated M3 homodimerization (Zeng and Wess 1999).
Fig. 1. Schematic presentations of the human M2 muscarinic acetylcholine receptor and partial sequence alignment of the muscarinic acetylcholine receptor family around the four conserved extracellular cysteine residues. The transmembrane domain boundaries were set by manual alignment of the sequences to bovine rhodopsin, whose crystal structure has been solved (Palczewski et al., 2000).
It has been suggested that the o2 loop might fold back, because of constraints imposed by the conserved disulfide bond, to form a lid structure on top of the orthosteric binding pocket and might even participate in orthosteric ligand binding for many rhodopsin-like GPCRs (Palczewski et al., 2000; Shi and Javitch, 2004). Indeed, the o2 loop of the complement factor 5a receptor seems to play a critical role in receptor activation, because many mutations in the loop cause constitutive activity (Klco et al., 2005). The Cys176 residue in the o2 loop of the hM2 receptor is right in the middle of two epitopes that are important for allosteric binding (Fig. 1). Previous chimeric and site-directed mutagenesis studies have indicated that the acidic EDGE motif (172-175) and residue Tyr177 are crucial for the high-affinity binding of a number of typical mAChR allosteric modulators (including gallamine and W84) to M2 receptors (Leppik et al., 1994; Gnagey et al., 1999; Voigtländer et al., 2003; Huang et al., 2005). In addition, that region of the receptor is also important for THA binding (Ellis and Seidenberg, 2000; Tränkle et al., 2005; Huang and Ellis, manuscript in preparation). In the present study, we have replaced the four extracellular cysteine residues with alanines, in pairs, and we have examined the effects of these mutations on the allosteric binding of gallamine, THA, and W84 in M2 and M3 receptors. We found that mutation of the cysteine residues involved in the disulfide bond did alter the potencies of these allosteric ligands in some cases, but the most striking finding was that positive homotropic interactions of the allosteric ligands were markedly reduced or eliminated.
Materials. Atropine sulfate, gallamine triethiodide, tacrine hydrochloride, and polyethylenimine were purchased from Sigma (St. Louis, MO). W84 is from Tocris Cookson Inc. (Ellisville, MO). [3H]NMS (81 Ci/mmol) was ordered from PerkinElmer Life and Analytical Sciences (Boston, MA). Glass fiber filters and all other inorganic chemicals are all from VWR International, Inc. (Bridgeport, NJ). The modified rat M3' plasmid (rM3') was kindly provided by Dr. J. Wess (National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD). The modification to the rat M3 receptor includes a 1xHA tag at the N terminus, mutations of Asp to Gln at positions of 6, 15, 41, 48, and 52 to delete N-glycosylation sites, and a large deletion (Ala274 to Lys469) of the central portion of the third intracellular loop (Zeng et al., 1999).
Site-Directed Mutagenesis, Receptor Expression, and Membrane Preparation. All mutations and modifications to muscarinic receptors were carried out using the QuikChange method (Stratagene, La Jolla, CA) in the pcD vector (Okayama and Berg, 1983; Bonner et al., 1987). Primer synthesis and sequence confirmation were carried out in the Core Facility of Hershey Medical Center (Hershey, PA). Plasmid purification kits from QIAGEN (Valencia, CA) were used for all plasmid purification. Removal of potential N-terminal N-glycosylation sites (Asn to Asp) of the human M2 (hM2) receptor at positions 2, 3, 6, and 9 and insertion of one hemagglutinin (HA) epitope (YPYDVPDYA) were designed into a pair of long primers. Primers are shown in Table 1.
TABLE 1 Primers
Receptor constructs were transiently transfected in COS-7 cells using PolyFect Reagent (QIAGEN). Approximately 48 h after transfection, cells were scraped into 5 mM sodium/potassium/phosphate buffer (PB, pH 7.4; 4 mM Na2HPO4 and 1 mM K2HPO4) and homogenized in the PB on ice with a mechanical Bio Homogenizer from Biospec Products, Inc. (Bartlesville, OK). After centrifugation at 50,000g for 30 min at 4°C, the membrane pellet was resuspended with a glass homogenizer in ice-cold 5 mM PB. The membrane suspension was either used immediately or stored in aliquots at -70°C.
[3H]NMS Saturation Binding Assays. Saturation binding assays were carried out in 5 mM PB, pH 7.4, in a 25°C water bath as described previously (Huang et al., 2005). In brief, membranes were incubated with six concentrations of [3H]NMS (ranging from 3 to 3000 pM), in duplicate, in a final volume of 1 ml for 30 min. In most assays, 2 to 5 µg of protein per assay tube were used; however, the low expression of the disulfide mutants necessitated much higher amounts (30-100 µg of protein). The incubation was terminated by filtration onto #32 glass-fiber filters (Whatman Schleicher and Schuell, Keene, NH) pretreated with ice-cold 0.1% polyethylenimine solution and followed with two quick washes with 5 ml of ice-cold 40 mM PB, pH 7.4. Radioactivity from membranes trapped on filter discs was determined by liquid scintillation counting. Data were fitted to a one-site hyperbolic binding curve, using Prism 4.03 (GraphPad Software, Inc., San Diego, CA). Nonspecific binding was determined in the presence of 3 µM atropine at each concentration point.
[3H]NMS Dissociation Assays. [3H]NMS dissociation assays were set up and conducted as described previously (Ellis and Seidenberg, 2000; Buller et al., 2002; Huang et al., 2005) in 5 mM PB, pH 7.4, in a 25°C water bath. In brief, membranes (similar amounts to those quoted in the saturation assays, above) were first labeled with [3H]NMS for 30 min in 1 ml. Dissociation of [3H]NMS was then initiated by the addition of atropine (3 µM in a final volume of 2 ml) with or without serial concentrations of allosteric modulators. The dissociation process was stopped by filtration after a period of time, which is typically set between 2 and 3 times the standard half-time of [3H]NMS dissociation (determined in the presence of atropine but in the absence of any allosteric modulator). Radioactivity was counted as above.
Data Analysis. Dissociation assays were set up such that the delay of [3H]NMS dissociation had to be mediated by the binding of the allosteric modulator to an allosteric site, distinct from the orthosteric site at which [3H]NMS binds, because orthosteric binding sites were prelabeled and then blocked by the high concentration of atropine during the dissociation phase. Dissociation was assumed to follow a mono-exponential time course. The standard dissociation rate constant (k0) was determined in the presence of atropine alone, and the apparent dissociation rate constant (kobs) was determined in the presence of both atropine and the allosteric modulator. The ratios of kobs/k0 were then plotted against the logarithms of the concentrations of allosteric modulator (including the basal value, obtained in the absence of allosteric ligand). The variable slope model uses the equation kobs/k0 = Bottom + [(Top - Bottom)/{1 + 10nH(logEC50 - X)}], where X is the logarithm of the concentration of allosteric modulator and Top and Bottom refer to the upper and lower plateaus of the sigmoidal curve, respectively. When the best-fit Hill slope (nH) has an absolute value greater than 1, a positive homotropic allosteric interaction between two or more binding sites is indicated. In the "standard slope" model, nH was constrained to unity. This concentration-response curve corresponds to the occupancy of the [3H]NMS-bound receptor by the allosteric modulator; the concentration at which allosteric modulators reduce the kobs/k0 ratio to 50% of the maximal effect is defined as potency (EC50). For cases in which the standard-slope model is appropriate, the potency (EC50) is also the apparent binding affinity of the allosteric modulator at the [3H]NMS-occupied receptor, Kapp (often presented as pKapp; Ellis and Seidenberg, 1992; Lazareno and Birdsall, 1995). Statistical tests were carried out between data sets to determine whether changes in the Hill slope that were induced by mutation were statistically significant relative to the Hill slope of the relevant parental receptor; if the mutation did produce a significant change, then tests were carried out within data sets to determine whether the standard or variable slope model was the most appropriate fit. In all cases, statistical significance was set at the p < 0.05 level, based on an F test, using the built-in features of Prism 4.03. The program also estimates standard errors that give an indication of the precision with which the parameter values are known, but these estimates are not suitable for calculating P values. The F test, based on global curve-fitting, is the most robust and appropriate method for testing the boundaries of parameter values (Motulsky and Christopoulos, 2003).
In this study, we targeted four extracellular cysteine residues, Cys96, Cys176, Cys413, and Cys416 of hM2 receptors, and the corresponding cysteine residues in hM3 and rM3 receptors (Fig. 1). Because the first two cysteine residues form a disulfide bond that is highly conserved in the majority of GPCRs (and therefore referred to as the disulfide cysteine residues in this study) and the last two cysteine residues are separated by only two amino acids, we decided to mutate them in pairs rather than singly. This approach could eliminate the potential complication that remaining unpaired cysteine residues might form an abnormal disulfide bond that could compromise our interpretation. A double cysteine mutant in the rM3' receptor was created to make the construct that was identical to the one reported previously by Dr. Wess and his colleagues (Zeng and Wess, 1999; Zeng et al., 1999). Anticipating future characterizations of some of the mutants, and to maintain similarity to the rM3' constructs, we also inserted an HA epitope in the hM2 receptor after the initiating Met residue and at the same time deleted potential N-glycosylation sites in the N terminus at position of 2, 3, 6, and 9 (Fig. 1). For consistency, this modified receptor was named hM2'. Such modifications at the N termini of M3 receptors (Zeng et al., 1999) and M2 receptors (van Koppen and Nathanson, 1990) do not affect surface expression, ligand binding, or receptor functions.
Disulfide Cysteine Residues, but Not o3 Cysteine Residues, Are Important for Surface Expression and [3H]NMS Binding. Mutations of the disulfide cysteine residues in the mutant hM2' C96A, C176A dramatically reduced cell surface expression levels and also reduced the binding affinity for [3H]NMS compared with the parental control (i.e., hM2'; see Table 2). Likewise, corresponding mutations in hM3 (C141A, C221A) reduced surface expression by more than 90% and substantially reduced [3H]NMS binding affinity. In addition, we also constructed and tested the corresponding double-cysteine mutations in rM3' receptors (C140A, C220A). Consistent with their counterparts in hM2 and hM3, as well as previously reported numbers (Zeng and Wess, 1999; Zeng et al., 1999), expression levels and [3H]NMS binding affinity of rM3' C140A, C220A were also greatly reduced. The degree of reduction of affinity for [3H]NMS for the rM3' C140A, C220A mutant was smaller in our study than reported previously (Zeng et al., 1999), possibly because of different assay conditions. In contrast to the dramatic effects produced by mutations of the disulfide cysteine residues, receptors with mutations at the two cysteine residues in the short o3 loop of hM2 (C413A,C416A) retained expression levels and [3H]NMS affinity comparable with those of the hM2 wild-type receptors. Likewise, the hM3 mutant with corresponding mutations in the o3 loop (C517A, C520A) also had expression levels and [3H]NMS affinity comparable with those of the hM3 wild type receptor. N-terminal modifications (HA tagging and deletion of N-glycosylation sites) of hM2 receptors did not alter receptor expression or [3H]NMS binding affinity very much, consistent with previous results from similar modifications in the rM3 receptor (Zeng et al., 1999).
TABLE 2 Binding properties of M2 and M3 receptors
The hM2' construct is the hM2 receptor with 1x HA tagging after the initiating Met residue and deletion of potential N-glycosylation sites (see Materials and Methods). The rM3' construct is the rM3 receptor with 1x HA tagging after the initiating Met residue, deletion of potential N-glycosylation sites, and a large deletion from Ala274 to Lys469 in the center portion of the third intracellular loop (Zeng et al., 1999). The t1/2 values represent the mean ± S.E.M. from three or more assays and are the standard half times of [3H]NMS dissociation in the presence of atropine. Bmax and pKd values are the mean ± S.E.M. from two or more assays.
N-Terminal Modifications Have No Effect on Allosteric Binding. Because several mutant receptors in this study contain modifications (HA tagging and deletion of potential N-glycosylation sites) at the N terminus, we first performed control assays to determine whether such modifications themselves affected the allosteric binding of the tested modulators, THA, gallamine, and W84. As indicated in Fig. 2D and Tables 3 and 4, allosteric binding at hM2 receptors was not significantly affected by HA tagging or by deletion of N-glycosylation sites. In addition, the N-terminal modified rM3 receptor (rM3'), had essentially the same allosteric binding profiles as hM3 receptors without modifications, except for a somewhat slower dissociation rate with rM3' receptors than with hM3 receptors (Table 2). This difference in [3H]NMS dissociation rates did not seem to be due to the N-terminal modifications, because unmodified rM3 receptors also had slower rates (t1/2 = 61 min) than hM3 receptors, in agreement with previous studies (Ellis et al., 1993; Gnagey et al., 1999).
Fig. 2. Effects of disulfide cysteine mutations on allosteric binding of tetrahydroaminoacridine (THA) in hM2 receptors (A), N-terminally modified rM3 receptors (B), hM3 receptors (C), and N-terminally modified hM2 receptors, under typical conditions and low percentage labeling with a lower concentration of [3H]NMS (D). The M2 curve from A is shown as the broken line in D. Allosteric modulation of the [3H]NMS dissociation rate was determined as described under Materials and Methods and data are shown with the best-fit curves (see Table 3 for parameters and statistical data). Points represent the means ± S.E.M. from three to five independent assays, each conducted in duplicate.
TABLE 3 Allosteric properties of THA at M2 and M3 receptors
The parameters pKapp and Hill slope (nH) represent best-fit values ± S.E. from global curve fitting to the combined data of three or more assays. Statistical significance was determined by F test, either between or within data sets (see Materials and Methods). The between-set P values shown are for comparisons with the immediate parental receptor; for example, hM2' was compared with hM2, while hM2'C96A, C176A was compared with hM2'. When nH was found to be significantly (P < 0.05) different from the parental receptor (or when there was no parental receptor), a global F test was also carried out to determine whether nH differed from the standard (unit) slope.
TABLE 4 Allosteric properties of gallamine and W84 at M2 and M3 receptors
The parameters pKapp and Hill slope (nH) represent best-fit values ± S.E. from global curve fitting to the combined data of three or more assays. Statistical significance was determined by F test, either between or within data sets (see Materials and Methods). The between-set P values shown are for comparisons with the immediate parental receptor; for example, hM2' was compared with hM2, whereas hM2'C96A, C176A was compared with hM2'. When nH was found to be significantly (P < 0.05) different from the parental receptor (or when there was no parental receptor), a global F test was also carried out to determine whether nH differed from the standard (unit) slope.
Subtype-Specific Effects of Disulfide Cysteine Mutations on Allosteric Binding Properties of THA. The atypical mAChR allosteric modulator THA showed steep dose-response curves in inhibiting [3H]NMS dissociation with variable slopes of approximately 2, indicating that positive homotropic allosteric interaction was present at both M2 and M3 receptors (Table 3). The mutations C96A and C176A in the hM2' receptor significantly reduced the slope factor for the THA curve to a value that was no longer significantly different from 1; that is, in this mutated construct, there was no evidence for positive homotropic cooperativity in the interaction of THA. In addition, the mutation caused a reduction in the potency of THA of approximately 6-fold (Fig. 2A). The corresponding mutations in hM3 (C141A, C221A) and rM3' (C140A, C220A) produced similar effects on the slope factors for the THA curves (Fig. 2, B and C, Table 3); for these M3 mutants, the best-fit slope factors suggested negative cooperativities (nH < 1), but were not significantly less than unity. In contrast to the result with the hM2 mutant, THA potency was significantly increased by the double mutation in both hM3 and rM3'.
As we have noted above, these cysteine mutants showed reduced binding affinity for [3H]NMS, and would therefore be labeled at far below saturation with 1 to 2 nM [3H]NMS. To determine whether nonsaturation labeling at these mutants might be responsible for the observed effects, we chose concentrations [3H]NMS that would label less than 10% of the total available binding sites at the hM2 and rM3' receptors and examined the effects of THA on [3H]NMS dissociation. As indicated in Fig. 2D and Table 3, results obtained with this low percentage labeling did not differ significantly from those obtained with saturation labeling.
These cysteine mutations also enhanced the rate of dissociation of [3H]NMS from the orthosteric binding site. Dissociation from the hM2' C96A, C176A mutant was approximately twice as fast as that from the parental hM2' receptor; hM3 C141A, C221A was approximately 6 times faster than hM3; and the rM3' C140A, C220A mutant was approximately 13 times faster than its parent, rM3' (Table 2).
Subtype Specific Effects of Disulfide Cysteine Mutations on Allosteric Binding Properties of Gallamine and W84. The double-cysteine mutations in the hM2' receptor (C96A and C176A) reduced the potency of gallamine somewhat (approximately 2.5-fold) but reduced W84 potency by approximately 70-fold (Fig. 3A and Table 4). In contrast, however, the potency of W84 was not changed by the double-cysteine mutation in the hM3 receptor (C141A and C221A) or in the rM3' receptor (C140A, C220A; Fig. 3B, 3C and Table 4). The data for both gallamine and W84 were better fitted to the standard slope model at the hM2' receptor, and the mutations of the disulfide cysteines did not significantly alter that slope. It is noteworthy that the data for gallamine at both hM3 and rM3' receptors were significantly better fitted to the variable slope model, with Hill slopes of 1.21 and 1.34 (Fig. 3, B and C, and Table 4), respectively. That is, the Hill slopes for gallamine at both of these M3 receptors indicated positive homotropic allosteric interactions. Double-cysteine mutations of C141A and C221A in hM3 or C140A and C220A in rM3' resulted in curves with significantly lower slopes; moreover, these slopes were not significantly different from unity (Fig. 3B and Table 4). The slope factors for W84 were not significantly different from unity at either hM3 or rM3' (Table 4); eliminating the disulfide cysteines did not significantly alter these slopes.
Fig. 3. Effects of disulfide cysteine mutations on allosteric binding of gallamine and W84 in hM2 receptors (A), hM3 receptors (B), and N-terminally modified rM3 receptors (C). [3H]NMS dissociation assays were conducted and data were analyzed as in Fig. 2 and best-fit parameters are presented in Table 4. Points represent the mean ± S.E.M. from 3 to 10 independent assays, each conducted in duplicate.
Mutations of o3 Cysteine Residues Do Not Alter Binding Parameters. For comparison, we also examined the allosteric effects of gallamine, W84, and THA in receptors with double-cysteine mutations in the o3 loop, hM2 C413A, C416A, and hM3 C517A, C520A. As indicated in Fig. 4 and Table 4, these mutations were quite benign. That is, there were no significant changes in potencies, and THA and gallamine displayed the same positive homotropic allosteric features at these mutant receptors as they did at the corresponding wild-type receptors. In addition, the dissociation rate of [3H]NMS was not altered by the cysteine mutations in the o3 loop (Table 2).
Fig. 4. Effects of cysteine mutations in the o3 loop on allosteric binding of gallamine, W84, and THA in the hM2 receptor (A) and hM3 receptor (B). [3H]NMS dissociation assays were conducted and data were analyzed as in Fig. 2 and best-fit parameters are presented in Tables 3 and 4. Points represent the mean ± S.E.M. from three to six independent assays, each in duplicate.
Phe222 in the o2 Loop of the hM3 Receptor Is Important for the Binding of Allosteric Modulators. Tyr177 in the hM2 receptor is known to be more important for the allosteric binding of W84 than for that of gallamine, participating in - interactions with W84 (Voigtländer et al., 2003; Huang et al., 2005; Prilla et al., 2006). This residue is immediately adjacent to Cys176. Therefore, mutations of C176A and C96A that eliminate the conserved disulfide bond could potentially change the orientation of Tyr 177 in hM2 receptors. An aromatic residue is conserved at the corresponding position of every mAChR except hM5, which has the lowest binding affinities for gallamine or W84 of any muscarinic subtype. In M3 receptors, the corresponding residue is Phe222 (Fig. 1). Mutations of C141A and C221A in the hM3 receptor, however, had no effect on the potency of either gallamine or W84. This observation suggested that Phe222 might not be important for allosteric binding. To directly examine the role of hM3 Phe222 in allosteric binding, we replaced it with either Tyr (conserving the aromatic ring) or Ala. HM3 F222A and hM3 F222Y were expressed at 1.87 and 5.81 mol/mg, with [3H]NMS binding affinities (pKd) of 10.16 and 10.47, respectively. As indicated in Fig. 5 and Table 5, replacing Phe222 with Ala had similar effects on gallamine, THA, and W84, reducing allosteric potencies by 2- to 4-fold. Replacing Phe222 with Tyr slightly increased the potencies of gallamine and W84 by 2- to 4-fold but had no effect on the potency of THA. In addition to these modest effects on the potencies of allosteric binding, mutations at Phe222 also changed the Hill slopes associated with THA binding. The slope factor for THA at the F222Y mutant was increased somewhat (although not significantly), whereas the F222A mutant receptor retained homotropic interactions in THA binding but with a significantly reduced Hill slope (Fig. 5C, Table 5). Both mutations decreased the slope factors for gallamine to near unity, although statistical significance was achieved only for the F222A mutation and only for gallamine (Fig. 5A and Table 5).
Fig. 5. Effect of Phe222 mutations in hM3 receptors on allosteric binding of gallamine (A), W84 (B), and THA (C). Binding curves for each allosteric modulator at the hM3 wild type receptor from Fig. 2 and Fig. 3 are shown in broken lines in each panel. [3H]NMS dissociation assays were conducted and data were analyzed as in Fig. 2 and best-fit parameters are presented in Table 5. Points represent the mean ± S.E.M. from 3 to 10 independent assays, each conducted in duplicate.
TABLE 5 Allosteric properties of hM3 receptors with mutations at Phe222
The parameters pKapp and Hill slope (nH) represent best-fit values ± S.E. from global curve fitting to the combined data of three or more assays. Data for wild-type hM3 receptors were included for comparison. Statistical significance was determined by F test, either between or within data sets (see Materials and Methods). The between-set P values shown are for comparisons with the immediate parental receptor, hM3. When nH was found to be significantly (P < 0.05) different from the parental receptor (or when there was no parental receptor), a global F test was also carried out to determine whether nH differed from the standard (unit) slope.
The hM2 receptor contains two epitopes in the o2 loop (the EDGE motif and Tyr177) that are important for allosteric binding. Each is adjacent to the Cys176 residue that links to Cys96 at the top of TM III. In this study, we examined allosteric interactions in mutants of M2 and M3 receptors that lacked this disulfide link.
The disulfide cysteine mutations in hM2' receptors reduced affinity for W84 by approximately 70-fold, slightly more than the 50-fold reduction we observed previously with the single Y177Q mutation. The same mutations also reduced gallamine potency by approximately 2-fold, equivalent to that of the Y177Q mutation (Huang et al., 2005). This big difference between W84 and gallamine is consistent with the fact that Tyr177 is much more important for W84 binding than for gallamine binding, because Tyr177 participates in a unique - interaction with W84 (Voigtländer et al., 2003; Huang et al., 2005; Prilla et al., 2006). Thus, it seems that the degrees of reduction in potency of these two modulators after the elimination of the disulfide bond could be attributed mainly to changes in the contributions of the Tyr177 residue. This suggests that the disulfide bond helps to constrain the o2 loop into a proper conformation for allosteric binding, especially for the binding of W84. Recent studies have suggested that the aromatic ring of W84 is sandwiched between Tyr177 and Trp422 (Prilla et al., 2006), which would require a very specific orientation of these two residues. On the other hand, the binding of gallamine is much more sensitive to acidic residues in the EDGE region on the other side of Cys176 (Huang et al., 2005); it seems reasonable to assume that this ionic interaction with gallamine would still be available, given the probable increase in the flexibility of the o2 loop that would result from the elimination of the disulfide attachment. Indeed, an increase in the flexibility of the o2 loop may be inferred from the doubling of the rate of dissociation of [3H]NMS from the disulfide cysteine mutant, relative to its parent receptor, hM2'. Elimination of the corresponding disulfide bond by cysteine mutation in bovine rhodopsin also confers greater conformational flexibility (Davidson et al., 1994).
The elimination of the disulfide bond in hM2' reduced the potency (EC50) toward THA by approximately 4-fold, slightly more than the reduction we have observed from the single Y177Q mutation in hM2 (X.-P. Huang and J. Ellis, manuscript in preparation). More interesting than the change in potency, however, is the finding that the Hill slope for the interaction of THA with the mutant receptor is not significantly different from 1. It seems that this mutation either interferes with the communication between the THA binding sites or occludes one of the binding sites. Elimination of the disulfide bond in M3 receptors produced results that were quite similar to those seen with the hM2' receptor. That is, both hM3 and rM3' exhibited robust positive cooperativity (Hill slope of approximately 2) that was abolished by the mutation (to Hill slopes not significantly different from 1). These mutations accelerated the rate of dissociation of [3H]NMS even more so than those in hM2', again suggesting that enhanced flexibility in the o2 loop relieved constraints on the access of NMS to the orthosteric site. However, there were also several differences between the M2 and M3 receptors. For one, gallamine showed significant positive cooperativity at both hM3 and rM3', although it was not as robust (Hill slope estimated at 1.21 at hM3 and 1.34 at rM3') as that seen with THA. As with THA, M3 mutants that lacked the disulfide cysteines did not exhibit positive cooperativity in the action of gallamine. This is the first report of homotropic cooperativity for gallamine at any muscarinic receptor subtype in this type of assay. However, it has been known for some time that gallamine exhibits a biphasic regulation of the dissociation of quinuclidinyl benzilate from M2 receptors. Low concentrations of gallamine accelerate the dissociation of this orthosteric antagonist, whereas higher concentrations slow the dissociation, implying actions at two different allosteric sites (Ellis and Seidenberg, 1989). These complex interactions seen with gallamine suggest that the differentiation between typical and atypical ligands may be quantitative rather than qualitative and that it may take a particular assay or condition to reveal the "atypical" characteristics of a given ligand.
In contrast to the 70-fold reduction in potency toward W84 that was caused by the elimination of the disulfide bond in the hM2' mutant, both the hM3 mutant and the rM3' mutant had essentially the same potencies for W84 as their respective parent receptors. This difference led us to investigate the contribution of the Phe222 residue in hM3 to the potencies of the allosteric ligands. Our results with hM3F222A and hM3F222Y mutants can be compared with the corresponding hM2 mutants that we have reported previously (Huang et al., 2005). Gallamine and W84 show similar qualitative relative potencies in the two subtypes, Tyr > Phe > Ala, but the magnitudes are significantly different. The loss of the aromatic quality (F222A) produces a significantly greater loss in affinity toward W84 in the M2 background than in the M3 background. This suggests that features of the M3 receptor structure may prevent the participation of the Phe residue in the - sandwich that has been suggested for hM2 (Prilla et al., 2006). However, this Phe residue does seem to play a role in modulating the cooperativity between allosteric sites. Eliminating the aromatic quality altogether (hM3F222A) eliminates the cooperativity toward gallamine and reduces it partially, but significantly, in the case of THA.
It should be noted that all of the allosteric assays in the present article were carried out using receptors that were prelabeled with [3H]NMS (i.e., dissociation assays, see Materials and Methods). Therefore, we cannot know how much the mutations may have altered the potencies of the interactions between these allosteric ligands and the free receptor. However, this approach has the advantage that we can be certain that we are measuring cooperativity (and changes in cooperativity) between sites that are necessarily allosteric to the NMS binding site. These assays also support the suggestion that mutations at the disulfide cysteines enhance the flexibility of the o2 region of the receptor. That is, at wild-type receptors, the dissociation of [3H]NMS seems to be completely prevented by high concentrations of the allosteric ligands, but, at receptors lacking the disulfide cysteines, a significant rate of dissociation remains even at saturating concentrations (for example, see Fig. 3B).
The cooperativity shown by THA raises the question of where the multiple sites are located and how they relate to the sites for other muscarinic allosteric modulators. One way to approach this question is to determine whether two modulators with different efficacies interact competitively (Ellis and Seidenberg, 1992; Waelbroeck, 1994). Using this technique, a number of modulators have been shown to act at a common site (Ellis and Seidenberg, 1992, 2000; Tränkle and Mohr, 1997), whereas others seem to bind at different sites (Lazareno et al., 2000, 2002). Therefore, there seem to be at least two different allosteric binding sites per muscarinic receptor. If THA interacts with two cooperative sites, are they on the same receptor, or are they the result of dimerization of receptors? Recent molecular modeling studies have suggested that, because of its small size, two molecules of THA can fit simultaneously into the allosteric space of the M2 receptor, one in the more typical space between o2 and o3, and the other closer to o1 and the N terminus. However, the experimental data that accompany that model revealed additional complexities that make it impossible to exclude the explanation that two sites are linked via the interface between two receptors. Indeed, at this point, the two explanations are not incompatible; there could be four sites involved, two on each receptor (Tränkle et al., 2005). Mechanistic, molecular, and biophysical studies have demonstrated that GPCRs, including M2 and M3 muscarinic receptors, can form dimers and even higher order oligomers (Park et al., 2004).
It is noteworthy that Zeng and Wess (1999) have reported Cys140 and Cys220 to be involved in covalent homodimerization of the same rM3' receptor construct that we employed in the present studies; the mutations C140A and C220A eliminated the covalent link between the receptors. However, the mutant receptors were still able to form noncovalent complexes that were identified by immunoprecipitation, even though expression was greatly reduced (to a similar level as in our studies; Zeng and Wess, 1999; Zeng et al., 1999). This finding is consistent with other studies that have detected dimerization of muscarinic receptors at low expression levels in live cells, using bioluminescence resonance energy transfer (Goin and Nathanson, 2006). Indeed, it is likely that many GPCRs dimerize within the endoplasmic reticulum, even before transport to the plasma membrane (Bulenger et al., 2005). In any event, these disulfide cysteine residues are located approximately at the junction of the two allosteric pockets predicted by the molecular model (Tränkle et al., 2005), making it difficult to choose between the possibilities that the cysteine mutations disrupted an intrareceptor communication between the two pockets or that they disrupted communication between binding sites on two linked receptors. It is expected that future studies will determine which mechanism applies.
ABBREVIATIONS: GPCR, G protein-coupled receptor; NMS, N-methyl scopolamine; W84, hexamethylene-bis-[3-phthalimidopropyl]ammonium) dibromide; THA, tetrahydroaminoacridine; o2, the second outer (extracellular) loop of the receptor; o3, the third outer (extracellular) loop of the receptor; mAChR, muscarinic acetylcholine receptor; HA, hemagglutinin; PB, sodium/potassium/phosphate buffer; h, human; r, rat; rM3', N-terminal-modified rat M3 receptor.
【参考文献】
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]
Bulenger S, Marullo S, and Bouvier M (2005) Emerging role of homo- and heterodimerization in G-protein-coupled receptor biosynthesis and maturation. Trends Pharmacol Sci 26: 131-137.
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 (2002) Allosteric binding sites on cell-surface receptors: novel targets for drug discovery. Nat Rev Drug Discov 1: 198-210.
Christopoulos A and Kenakin T (2002) G protein-coupled receptor allosterism and complexing. Pharmacol Rev 54: 323-374.[Abstract/Free Full Text]
Curtis CA, Wheatley M, Bansal S, Birdsall NJ, Eveleigh P, Pedder EK, Poyner D, and Hulme EC (1989) Propylbenzilylcholine mustard labels an acidic residue in transmembrane helix 3 of the muscarinic receptor. J Biol Chem 264: 489-495.[Abstract/Free Full Text]
Davidson FF, Loewen PC, and Khorana HG (1994) Structure and function in rhodopsin: replacement by alanine of cysteine residues 110 and 187, components of a conserved disulfide bond in rhodopsin, affects the light-activated metarhodopsin II state. Proc Natl Acad Sci USA 91: 4029-4033.[Abstract/Free Full Text]
Ellis J and Seidenberg M (1989) Gallamine exerts biphasic allosteric effects at muscarinic receptors. Mol Pharmacol 35: 173-176.
Ellis J and Seidenberg M (1992) Two allosteric modulators interact at a common site on cardiac muscarinic receptors. Mol Pharmacol 42: 638-641.
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.
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.
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]
Goin JC and Nathanson NM (2006) Quantitative analysis of muscarinic acetylcholine receptor homo- and heterodimerization in live cells: regulation of receptor down-regulation by heterodimerization. J Biol Chem 281: 5416-5425.[Abstract/Free Full Text]
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]
Karnik SS and Khorana HG (1990) Assembly of functional rhodopsin requires a disulfide bond between cysteine residues 110 and 187. J Biol Chem 265: 17520-17524.[Abstract/Free Full Text]
Klco JM, Wiegand CB, Narzinski K, and Baranski TJ (2005) Essential role for the second extracellular loop in C5a receptor activation. Nat Struct Mol Biol 12: 320-326.
Kurtenbach E, Curtis CA, Pedder EK, Aitken A, Harris AC, and Hulme EC (1990) Muscarinic acetylcholine receptors. Peptide sequencing identifies residues involved in antagonist binding and disulfide bond formation. J Biol Chem 265: 13702-13708.[Abstract/Free Full Text]
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.
Lazareno S, Popham A, and Birdsall NJ (2000) Allosteric interactions of staurosporine and other indolocarbazoles with N-[methyl-3H]scopolamine and acetylcholine at muscarinic receptor subtypes: identification of a second allosteric site. Mol Pharmacol 58: 194-207.[Abstract/Free Full Text]
Lazareno S, Popham A, and Birdsall NJ (2002) Analogs of WIN 62,577 define a second allosteric site on muscarinic receptors. Mol Pharmacol 62: 1492-1505.[Abstract/Free Full Text]
Leppik RA, Miller RC, Eck M, and Paquet JL (1994) Role of acidic amino acids in the allosteric modulation by gallamine of antagonist binding at the m2 muscarinic acetylcholine receptor. Mol Pharmacol 45: 983-990.
Motulsky HJ and Christopoulos A (2004) Fitting Models to Biological Data Using Linear and Nonlinear Regression. A Practical Guide to Curve Fitting. Oxford University Press, New York.
Okayama H and Berg P (1983) A cDNA cloning vector that permits expression of cDNA inserts in mammalian cells. Mol Cell Biol 3: 280-289.[Abstract/Free Full Text]
Palczewski K, Kumasaka T, Hori T, Behnke CA, Motoshima H, Fox BA, Le Trong I, Teller DC, Okada T, Stenkamp RE, et al. (2000) Crystal structure of rhodopsin: A G protein-coupled receptor. Science (Wash DC) 289: 739-745.[Abstract/Free Full Text]
Park PS, Filipek S, Wells JW, and Palczewski K (2004) Oligomerization of G protein-coupled receptors: past, present, and future. Biochem 43: 15643-15656.
Potter LT, Ferrendelli CA, Hanchett HE, Hollifield MA, and Lorenzi MV (1989) Tetrahydroaminoacridine and other allosteric antagonists of hippocampal M1 muscarine receptors. Mol Pharmacol 35: 652-660.[Abstract/Free Full Text]
Prilla S, Schrobang J, Ellis J, H?ltje HD, and Mohr K (2006) Allosteric interactions with muscarinic acetylcholine receptors: complex role of the conserved tryptophan M 4222Trp in a critical cluster of amino acids for baseline affinity, subtype selectivity, and cooperativity. Mol Pharmacol 70: 181-193.[Abstract/Free Full Text]
Savarese TM, Wang CD, and Fraser CM (1992) Site-directed mutagenesis of the rat m1 muscarinic acetylcholine receptor. Role of conserved cysteines in receptor function. J Biol Chem 267: 11439-11448.[Abstract/Free Full Text]
Shi L and Javitch JA (2004) The second extracellular loop of the dopamine D2 receptor lines the binding-site crevice. Proc Natl Acad Sci USA 101: 440-445.[Abstract/Free Full Text]
Tr?nkle C, Dittmann A, Schulz U, Weyand O, Buller S, J?hren 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 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]
van Koppen CJ and Nathanson NM (1990) Site-directed mutagenesis of the m2 muscarinic acetylcholine receptor. Analysis of the role of N-glycosylation in receptor expression and function. J Biol Chem 265: 20887-20892.[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]
Waelbroeck M (1994) Identification of drugs competing with d-tubocuarine for an allosteric site on cardiac muscarinic receptors. Mol Pharmacol 46: 685-692.
Zeng FY, Soldner A, Sch?neberg T, and Wess J (1999) Conserved extracellular cysteine pair in the M3 muscarinic acetylcholine receptor is essential for proper receptor cell surface localization but not for G protein coupling. J Neurochem 72: 2404-2414.
Zeng FY and Wess J (1999) Identification and molecular characterization of m3 muscarinic receptor dimers. J Biol Chem 274: 19487-19497.[Abstract/Free Full Text]
作者单位：Departments of Psychiatry and Pharmacology, Hershey Medical Center, Penn State University College of Medicine, Hershey, Pennsylvania