Introduction

Top Summary Introduction Procedures Results Discussion References

It has been known for some time that the binding properties of muscarinic acetylcholine receptors may be modulated by agents acting allosterically (for a review, see Tuek and Proka, 1995) and that the allosteric modulators not only diminish (Clark and Mitchelson, 1976; Stockton et al., 1983), but also enhance the affinity of the receptors for their orthosteric ligands (Tuek et al., 1990; Proka and Tuek, 1994; Guo et al., 1995; Jakubík et al., 1995b, 1997; Lazareno and Birdsall, 1995). We found recently that several allosteric modulators of muscarinic receptors (i.e., alcuronium, gallamine, and strychnine) not only modify the affinity of the receptors for the agonists and antagonists but also influence the interaction between the receptors and the G proteins (Jakubík et al., 1996). In experiments on CHO cell lines stably transfected with individual subtypes of muscarinic receptors, allosteric modulators had agonist-like effects on the synthesis of cAMP and of inositol phosphates in the cells. These effects could not be blocked by the orthosteric muscarinic antagonist QNB and could not be induced in cells that had not been transfected with the genes for muscarinic receptors. In certain cases, the direction of the effect of allosteric modulators on the synthesis of second messengers varied depending on the density of muscarinic receptors in the cell line used.

In the current work, we investigated whether the interaction between muscarinic receptors and a G protein also can be affected by the allosteric modulator alcuronium in a simplified system consisting of artificial lipid membranes incorporating purified M₂ receptors and purified G_o proteins and whether it depends on the stoichiometric ratio between the densities of the M₂ receptors and the G_o proteins (R/G_o ratio). The effects of carbachol (an orthosteric agonist) and atropine (an orthosteric antagonist) were investigated in parallel with those of alcuronium. Receptor-mediated increases in the binding of [³⁵S]GTPS served as the indicator of G protein activation. The results that we obtained indicate that empty receptors can themselves activate the G_o proteins and that there are substantial differences among the effects of the agonist, the allosteric modulator, and the antagonist. The effects of the allosteric modulator vary depending on the R/G_o ratio. The "active state" of muscarinic receptors induced by the allosteric modulator is different from the "active state" induced by the orthosteric agonist.

	Experimental Procedures

Top Summary Introduction Procedures Results Discussion References

Materials. Agarose (Sepharose 4-B) and DEAE-Sephadex were from Pharmacia Fine Chemicals (Uppsala, Sweden), Ultrogel AcA-34 was from Sepracor/IBF (Villeneuve la Garenne, France). Hydroxylapatite was from Sigma Chemical (St. Louis, MO). DEAE-Toyopearl 650S was from Toyo Soda MFG (Tokyo, Japan). Digitonin, phosphatidylcholine (from egg yolk), and phosphatidylinositol (L--phosphatidyl-D-myo-inositol-4-phosphate from bovine brain) were from Wako Pure Chemicals (Osaka, Japan). [³H]QNB (79 Ci/mmol) and [³⁵S]GTPS (1 Ci/µmol) were from DuPont-NEN (Boston, MA). ABT was synthesized (Haga and Haga, 1983), and ABT-agarose gel (Haga and Haga, 1985) and heptylamine Sepharose (Shaltiel, 1974) were prepared as described.

Purification of muscarinic M₂ receptors. Receptors were obtained by solubilization of membranes of Sf9 cells transiently expressing human m2 gene (Rinken et al., 1994; Guo et al., 1995). The membranes were treated with 1% digitonin and 0.1% sodium cholate, and muscarinic receptors were isolated by single-step affinity chromatography on ABT-agarose. After elution with atropine, they were concentrated on hydroxylapatite columns as described previously (Haga and Haga, 1985). The purity of the final preparation was checked by sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

Isolation of G proteins. Crude membranes were prepared from porcine brains and their integral proteins were solubilized with 1% sodium cholate. G_o and G_i proteins were isolated together by chromatography on DEAE-Sephacel, Ultrogel AcA 34, and heptylamine-Sepharose (Haga et al., 1986). DEAE-Toyopearl columns were used to separate the G_o from the G_i protein as described previously (Haga et al., 1986), except that 0.7% 3-[(3-cholamidopropyl)dimethylammonio]propanesulfonate was substituted for Lubrol PX in all solutions applied. The purity of G_o protein was confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blots (negative reaction with monoclonal antibodies against G_i, G_q, G₁₁, G₁₆, and G_s, kindly provided by Dr. J. Novotný).

Reconstitution of M₂ receptors and G proteins in lipid vesicles. The procedure was modified from that described by Shiozaki and Haga (1992). The lipid mixture was obtained by mixing 4 volumes of a solution of cholesterol hemisuccinate (10 mg/ml in methanol), 48 volumes of a solution of phosphatidyl choline (20 mg/ml in chloroform), and 48 volumes of a solution of phosphatidyl inositol (20 mg/ml in chloroform). The solvents were evaporated with a stream of nitrogen so as to obtain a thin layer of lipids on the wall of the test tube. Then, the lipids were emulsified to form vesicles in a solution of 20 mM K-HEPES, pH 8.0, 1 mM EDTA, 160 mM NaCl, and 1% sodium cholate by 30-min sonication at 4°. Vesicles were mixed with a suitable amount of purified M₂ receptors and separated on a column with 2 ml of Sephadex G-50. A suitable amount of G_o protein was added to them, and the mixture was incubated on ice for 60 min, after which it was diluted as requested.

Kinetics of [³⁵S]GTPS binding. For measurements of [³⁵S]GTPS binding to G proteins in reconstituted vesicles, the vesicles were preincubated at 30° in 150 µl of a medium consisting of 10 mM MgCl₂, 1 mM dithiothreitol, and the investigated muscarinic ligand. After 15 min (or 30 min, where indicated), [³⁵S]GTPS and GDP were added to final concentrations of 50 nM and 1 µM, respectively, and the incubation was continued in a final volume of 200 µl as needed. The binding was arrested by the addition of 500 µl of a stopping solution consisting of 50 mM Tris·HCl, pH 8.0, 100 mM NaCl, 5 mM MgCl₂, and 0.1 mM GTP, and the radioactivity bound to vesicles was separated on Whatman GF/B filters. In experiments designed to compare the effect of various concentrations of the same muscarinic ligand on [³⁵S]GTPS binding (i.e., to obtain concentration-response curves), [³⁵S]GTPS association was arrested at a suitable time during the phase of linear association, namely after 10 min in experiments with most ligands, after 5 min in experiments with carbachol at the 50:100 R/G_o ratio, or after 30 min in experiments with atropine and alcuronium at the 1:100 and 10:100 R/G_o ratios.

Treatment of data. Data in figures and tables are mean ± standard error of three experiments performed with incubation in duplicate. Values of K_d (equilibrium dissociation constant for the binding of a ligand to a single population of binding sites or to an undefined number of populations), K_low, and K_high (equilibrium dissociation constants for the binding of a ligand to two populations of binding sites displaying low and high affinity, respectively), f_high (fraction of binding sites displaying high affinity and expressed as percentage of the total number of the binding sites), K_i (equilibrium dissociation constant of an inhibitor), B_max (maximum binding capacity), EC₅₀ (concentration of a ligand producing one half of maximum effect), and k_obs (apparent association rate constant) were computed by nonlinear regression according to accepted conventions (Limbird, 1986; Hulme, 1992; Jakubík and Tuek, 1994a, 1994b; Jakubík et al., 1995a, 1995b). Apparent association rate constants (k_obs) for the binding of [³⁵S]GTPS to G proteins in liposomes were obtained by fitting the equation where B_t is binding at time t, and B_eq is binding at equilibrium.

View this table: [in this window] [in a new window]   TABLE 1 [35S]GTPS binding to reconstituted vesicles containing M2 receptors and Go proteins at different concentration ratios, with the concentration of Go held constant Values were derived from the experiments shown in Figs. 1 and 3. They are mean ± standard error of three experiments performed with incubations in duplicate.

View this table: [in this window] [in a new window]   TABLE 2 [35S]GTPS binding to reconstituted vesicles containing M2 receptors and Go proteins at a fixed molar ratio of 10:100 but at different final concentrations in liposomal membranes Values were derived from the experiments shown in Figs. 4 and 5. They are mean ± standard error of three experiments performed with incubations in duplicate.

	Results

Top Summary Introduction Procedures Results Discussion References

Effects of carbachol, atropine, and alcuronium on [³⁵S]GTPS association with reconstituted vesicles at different R/G_o. The time course of the association of [³⁵S]GTPS with G proteins in vesicles with three different R/G_o ratios is shown in Fig. 1. Characteristics of the vesicles used and the numerical values computed from experiments in Fig. 1 are summarized in Table 1. The effects of muscarinic ligands on the rates of [³⁵S]GTPS association were investigated in additional separate experiments on a broader scale of R/G_o ratios (Fig. 2).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Effects of 1 mM carbachol, 10 µM alcuronium, and 100 nM atropine on the time course of [35S]GTPS binding to lipid vesicles containing the M2 receptors and the Go protein at nominal ratios of 1:100, 10:100, and 50:100. Abscissa, time after the addition of [35S]GTPS. Ordinate, fmol of [35S]GTPS bound/incubation tube. Note that the content of Go/incubation tube was the same in all three graphs, whereas the content of receptors increased (left graph < middle graph < right graph). See Table 1 for numerical characteristics of the incubation conditions and of the results.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Initial velocity of [35S]GTPS binding to reconstituted vesicles with different R/Go ratios. Vesicles were preincubated in the control medium () or in the presence of 100 µM carbachol (), 10 µM alcuronium (), or 0.1 µM atropine (). After 30 min, [35S]GTPgS (50 nM) was added simultaneously with 1 µM GDP, and [35S]GTPS binding was determined at time 0 and at three time points within the linear phase of radioligand association. Abscissa, log R/Go ratio. Ordinate, rate of [35S]GTPS association (fmol/tube/min).

1

1

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Concentration-response curves for the effects of carbachol, alcuronium, and atropine on the binding of [35S]GTPS to lipid vesicles containing the M2 receptors and the Go protein at the ratios of 1:100, 10:100, and 50:100. Abscissa, log10 of the concentration of the muscarinic ligand (M). Ordinate, fold change in the amount of [35S]GTPS bound, compared with the situation in the absence of any muscarinic ligand. See the text for the duration of incubation and Table 1 for numerical characteristics of the incubation conditions and of the results.

Effects of carbachol, alcuronium, and atropine on [³⁵S]GTPS association with reconstituted vesicles at different densities of receptors and G_o proteins. Data in the previous section were obtained in experiments in which the concentration of the G_o proteins in the reconstituted vesicles was kept constant and the concentration of receptors was varied. The observed variations in the binding rates were presented as a consequence of changes in the ratio of receptors to G proteins. It might be argued that the rates of [³⁵S]GTPS association were influenced by changes in receptor densities in the membranes rather than by changes in R/G_o ratio. To check such an alternative interpretation, experiments were performed in which the R/G_o ratio was kept at a constant level of 10:100 and the total amounts of receptors and G_o proteins per incubation tube also were kept constant, but the amounts of the lipid mixture (and thus the numbers of vesicles and the areas of the membranes available for receptor and G protein incorporation) used for reconstitution were different. Three different sets of tubes were used. In those with the "1 × Basic" densities, the densities of receptors and G_o proteins in liposomal membranes were the same as they had been in the experiments described in Figs. 1-3 and in Table 1 using the 10:100 R/G_o ratio. In the tubes designated as "0.1 × Basic" or "5 × Basic," the densities of receptors and G_o proteins were 10-fold lower or 5-fold higher, respectively, than for the "1 × Basic" concentrations. The R/G_o ratio always remained at the 10:100 level.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Effects of 1 mM carbachol, 10 µM alcuronium, and 100 nM atropine on the time course of [35S]GTPS binding to lipid vesicles containing the M2 receptors and the Go protein at the same ratio of 10:100 but at three different concentrations. Reconstituted vesicles had been obtained in such a way that the same amounts of receptors and 10-fold higher amounts of Go protein had been incorporated into different volumes of the vesicle emulsion. Consequently, the densities of both receptors and Go were 10- and 50-fold higher (middle and right, respectively, compared with the left). Under the "1 Basic" conditions (middle), the densities of receptors and Go proteins corresponded to those in the 10:100 samples used in Figs. 1 and 3. See Table 2 for numerical characteristics of the incubation conditions and of the results.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Concentration-response curves for the effects of carbachol, alcuronium, and atropine on the binding of [35S]GTPS to lipid vesicles reconstituted with the M2 receptors and the Go protein at a fixed ratio of 10:100 and at three different receptor and Go protein concentrations. Abscissa, log10 of the concentration (M) of muscarinic ligand. Ordinate, fold change of the amount of [35S]GTPS bound compared with the situation in the absence of the muscarinic ligand. The arrangement of experiments was the same as in Fig. 4. See the text for the duration of incubation and Table 2 for numerical characteristics of the incubation conditions and of the results.

Inhibition of [³H]QNB binding to reconstituted vesicles by carbachol. It generally is assumed that receptors that are associated with a G protein display higher affinities for their agonists and that this can be seen in agonist-versus-antagonist competition binding curves. To obtain information about this aspect of our system, we performed competition binding experiments (carbachol versus [³H]QNB) on vesicles in which the R/G_o ratios were 1:100, 10:100, 50:100, or 10:0. As shown in Fig. 6 (left), the curves describing the inhibition of [³H]QNB binding by carbachol differed in their steepness and their left-to-right position depending on the R/G_o ratio. The higher the relative density of receptors (compared with that of G proteins), the more to the right and the steeper were the binding curves, suggesting that the proportion of receptors coupled to G_o ([RG]/[R_total]) was the highest in the vesicles with the 1:100 R/G_o ratio and the lowest in the vesicles without G proteins. The fraction of the high affinity binding sites (presumably reflecting the proportion of receptor/G protein complexes) computed from the binding data diminished from 73% at the 1:100 R/G_o ratio to 16% at the 50:100 R/G_o ratio.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Inhibition of [3H]QNB binding to reconstituted vesicles by carbachol. Abscissa, log10 of the concentration (M) of carbachol. Ordinate, [3H]QNB binding in the presence of carbachol, expressed as percentage of the binding in its absence. Left, inhibition of [3H]QNB binding to vesicles with different R/Go ratios. These vesicles corresponded to those used in Figs. 1 and 3 and Table 1. Right, inhibition of [3H]QNB binding to vesicles with a constant R/Go ratio of 10:100 but with three different concentrations of receptors and Go proteins, corresponding to those used in Figs. 3 and 4 and Table 2. The "1 Basic" concentrations of receptors and Go proteins (right) corresponded to the concentrations in the samples denoted as 10:100 (left). See Table 3 for computed characteristics of carbachol binding.

	Discussion

Top Summary Introduction Procedures Results Discussion References

Artificial membranes reconstituted with purified G protein-coupled receptors and G proteins have been fruitfully used for investigations of receptor/G protein interactions (for a review, see Birnbaumer and Birnbaumer, 1995). Although many aspects of these interactions remain controversial (Chidiac and Wells, 1992), two points have been generally accepted:

First, agonist-liganded receptors diminish the affinity of G proteins for GDP and stimulate the dissociation of GDP from them (Brandt and Ross, 1986; Tota et al., 1987; Florio and Sternweis, 1989; Haga et al., 1989; Shiozaki and Haga, 1992).

Second, agonist-liganded receptors stimulate the binding of GTP or its unhydrolyzable analogue [³⁵S]GTPS to G proteins, and this is mainly due to the receptor-induced decrease in the binding of GDP. In studies of reconstituted systems containing G_i and G_o, the addition of GDP was necessary to reveal the effect of agonists on [³⁵S]GTPS binding (Florio and Sternweis, 1989; Ikegaya et al., 1990; Shiozaki and Haga, 1992; but see Kurose et al., 1986, and Freissmuth et al., 1991a), whereas the addition of GDP was not necessary in systems containing G_s (Cerione et al., 1985; Brandt and Ross, 1986; Freissmuth et al., 1991b) or G_q (Nakamura et al., 1995). The meaning of these differences is not clear, and their analysis is difficult in view of the tight binding of GDP to G proteins (Ferguson et al., 1986).

Our data offer several new insights with regard to G_o protein activation by unliganded muscarinic receptors and by receptors associated with an agonist, antagonist, or allosteric modulator, to the importance of the R/G_o ratio, and to the likely spacial organization of receptors and G proteins in liposomal membranes. The following are the most important findings.

Empty muscarinic receptors increase the apparent rate of [³⁵S]GTPS binding to G_o proteins. When present at a high concentration (R/G_o ratio = 300:100), empty receptors were able to activate the G_o proteins to the same degree as the agonist-liganded receptors (Fig. 2). Similar effects of empty A₁ adenosine and -adrenergic receptors have been noted (Schütz and Freissmuth, 1992).

Atropine prevents or diminishes the agonist-independent stimulatory effects of muscarinic receptors on [³⁵S]GTPS binding. However, increases in the rate of [³⁵S]GTPS binding could be observed even in the presence of atropine if the R/G_o ratio was raised above 50:100 (Fig. 2). It seems that the effect of atropine can be viewed as an increase of the EC₅₀ concentration of receptors required for G_o activation.

Carbachol increases the apparent rate of [³⁵S]GTPS binding to G_o proteins and its effect varies with the R/G_o ratio. Although the enhancement of [³⁵S]GTPS binding by muscarinic agonist was expected (Kurose et al., 1986; Florio and Sternweis, 1989; Hilf et al., 1989; Shiozaki and Haga, 1992; Lazareno and Birdsall, 1993; Nakamura et al., 1995), several features of the current data merit attention:

Alcuronium affects the binding of [³⁵S]GTPS to G_o proteins in two directions. Compared with the binding that occurs in the absence of alcuronium, the binding of [³⁵S]GTPS in the presence of alcuronium is enhanced at low R/G_o ratios (<10:100) and diminished at high R/G_o ratios (>10:100; Figs. 1-3). This point is discussed further.

The effects of carbachol, atropine, and alcuronium on [³⁵S]GTPS binding are all receptor mediated. The three ligands had no effect when receptors were absent from the system (Table 1).

The kinetics of [³⁵S]GTPS association with the vesicles are affected very little by changes in the density of receptors and G_o proteins as long as the R/G_o ratio remains constant. The most likely explanation of relevant observations (Fig. 4 and Table 2) is that muscarinic receptors and G_o proteins are not uniformly dispersed in liposomal membranes but rather are clustered in microdomains in which their concentrations remain nearly constant despite variations in the amounts of lipidic membranes in the system. The agglomeration of signaling proteins in microdomains of cell surface membranes has been highlighted by Neubig (1994; see also Huang et al., 1997).

Increases in R/G_o ratios bring about decreases of the fraction of receptors displaying high affinity for the agonist carbachol. The higher the R/G_o ratio, the lower proportion of receptor molecules could be expected to be associated with a G protein. Data in Fig. 6 (left) and Table 3 thus support the notion that receptors associated with G proteins (RG complexes) have a higher affinity for the agonist than free receptors. When, however, the densities of receptors and G proteins in liposomal membranes were manipulated by altering the amount of lipid membranes (Fig. 6, right; Table 3), no change occurred in the fraction of receptors with a high affinity for the agonist. As pointed out, such behavior of the system can be explained on the assumption that the receptors and G proteins are clustered within microdomains in which their densities remain constant.