Introduction

Summary Introduction Materials & Methods Results Discussion References

It has long been known that the affinity of muscarinic receptors for their agonists and antagonists can be diminished by compounds acting allosterically (for reviews, see Refs. 1-3). It was discovered, however, that the affinity of cardiac muscarinic receptors for the muscarinic antagonist NMS can be increased by the neuromuscular blocker alcuronium (4, 5). Subsequently, it was shown that the positive allosteric action of alcuronium on the binding of NMS is specific for the M₂ and M₄ muscarinic receptor subtypes (Ref. 6, but see Refs. 7 and 8 for conflicting data concerning the M₃ and M₄ subtypes) and that it also applies to the binding of atropine, N-methylpiperidinyl benzilate, and several other muscarinic antagonists (9, 10). Other compounds exerting positive allosteric effects on the binding of muscarinic antagonists were discovered recently, such as strychnine (11, 12), ()-eburnamonine (13), fangicholine and tetrandrine (7), and 9-methoxy--lapachone (8). It is obvious that it is important to know whether the allosteric modulators are also able to increase the affinity of muscarinic receptors for their agonists.

We investigated the effects of alcuronium and four compounds with structural similarities to alcuronium (strychnine, brucine, vincamine, and ()-eburnamonine) on the affinities of muscarinic receptors of the M₁-M₄ subtypes for acetylcholine and 11 other agonists. We discovered that the affinities for the agonists can be allosterically enhanced by most of the compounds tested, although the direction of the allosteric interaction (positive or negative) and its extent vary depending on receptor subtype and the nature of the agonist and allosteric modulator. To measure the affinities of receptors for the unlabeled agonists tested, we used a procedure devised by Proöka (10)¹ in which the allosterically induced changes in the affinity for the agonist are reflected by changes in the binding of a labeled classic antagonist ([³H]NMS), which can be easily determined by the filtration method.

	Materials and Methods

Summary Introduction Materials & Methods Results Discussion References

Reagents. [³H]NMS (80 Ci/mmol) was from DuPont-New England Nuclear (Dreieich, Germany). Arecaidine propargyl ester was from Cookson Chemicals (Southampton, UK). Acetylcholine iodide, arecoline, bethanechol, brucine, carbachol, ()-eburnamonine, furmethide, methylfurmethide, oxotremorine, pilocarpine, strychnine, and vincamine were from Sigma Chemical (St. Louis, MO). McN-A-343 and oxotremorine-M were from Research Biochemicals (Natick, MA). Alcuronium was kindly provided by Hoffmann-La Roche (Basel, Switzerland), and pentylthio-TZTP was provided by Dr. P. Sauerberg (Copenhagen, Denmark). Pentylthio-TZTP is a thio analogue of the M₁-selective muscarinic agonist xanomeline (14-16).

Cells and cell membranes. Experiments were performed on membranes of CHO cell lines stably transfected with human genes for individual M₁-M₄ subtypes of muscarinic receptors (17). Cells were grown as described (6) in plastic dishes without coating in Dulbecco's modified Eagle's medium with 10% fetal calf serum and 0.005% geneticin. Seven days after subculturing, they were released by mild trypsinization, suspended in Dulbecco's modified Eagle's medium, and centrifuged for 3 min at 300 × g. The sedimented cells were resuspended in homogenization medium composed of 136 mM NaCl, 5 mM KCl, 1 mM MgSO₄, 1 mM Na-phosphate buffer, pH 7.4, and 10 mM Na-HEPES buffer, pH 7.4. They were washed twice through centrifugation (3 min at 300 × g) and resuspension in fresh homogenization medium and then homogenized with an Ultra-Turrax homogenizer (Janke and Kunkel, Staufen, Germany). The homogenate was centrifuged for 10 min at 600 × g, and the sediment was resuspended and recentrifuged. The supernatants from this and the previous centrifugation (containing the cell membranes) were combined, and their concentration was adjusted to correspond to 10 million cells of the starting suspension/1 ml. Then, they were kept frozen at 20°. On the day of the experiment, they were thawed and centrifuged for 15 min at 60,000 × g; the sediment was washed twice through resuspension in the homogenization medium and recentrifugation.

Radioligand binding experiments. Radioligand binding and its changes in the presence of competitive and allosteric ligands were examined essentially as previously described (6, 18-20). Membranes corresponding to 600,000 cells were incubated at 25° in a final volume of 0.8 ml for the time periods indicated below. The composition of the incubation medium corresponded with that of the homogenization medium, supplemented with 0.5 mM GTP in all binding experiments and with the ligands as indicated; GTP was included to induce the dissociation of G proteins from receptors. The incubation was terminated by filtration on Whatman GF/C glass-fiber filters in a Brandel cell harvester.

Data calculations. The B_max and K_d values for the binding of [³H]NMS (K_X) and the K_d values for the binding of agonists (K_L) and the allosteric modulators (K_A), as well as the Hill slope factors (n_H) for the competition between the agonists and [³H]NMS and the coefficient of cooperativity between [³H]NMS and the allosteric modulators [; determined and used as proposed by Ehlert (21)], were computed by nonlinear regression as previously described (6, 18, 19). Allosteric interaction between the binding of the allosteric ligands (A) and the unlabeled agonists (L) was evaluated from changes in the binding of [³H]NMS occurring in the presence of a fixed concentration of the agonist and of increasing concentrations of the allosteric ligand; it was characterized by the cooperativity coefficient . The cooperativity coefficient  is analogous to Ehlert's (21) coefficient  and corresponds to the ratio of the K_d value for the binding of an agonist to receptors occupied by the allosteric ligand and the K_d value for the binding of the same agonist to free receptors. Correspondingly,  < 1 in the case of positive cooperativity and  > 1 in the case of negative cooperativity between the allosteric ligand and the agonist. In the current study, we reserve the use of the coefficient  for the description of the cooperativity between the allosteric agents and the radiolabeled marker substance [³H]NMS.

(1)

(2)

(3)

(4)

	Results

Summary Introduction Materials & Methods Results Discussion References

Saturation binding experiments with [³H]NMS. K_d values for [³H]NMS binding to the M₁-M₄ receptor subtypes varied in the range of 103-148 pM, and B_max values were in the range of 85-143 fmol/10⁶ cells. A summary of the data is provided in Table 1.

Binding competition between agonists and [³H]NMS. Affinities of agonists for the M₁-M₄ receptor subtypes were determined according to the ability of the agonist to inhibit [³H]NMS binding in the presence of 0.5 mM GTP; in this way, the low affinity binding of agonists was measured. The computed pK_i and n_H values are summarized in Table 2. The Hill slopes were not significantly different from unity for any agonist. The affinities of acetylcholine and its analogues bethanechol and carbachol were in the range of 29-98 µM; those of furmethide and methylfurmethide were in the range of 12-78 µM; those of arecoline and arecaidine propargyl ester were in the range of 0.4-5.8 µM; those of oxotremorine and its derivatives (oxotremorine M and McN-A-343) were in the range of 3.3-17.0 µM; and those of pilocarpine were in the range of 6-12 µM. Pentylthio-TZTP displayed affinities in the range of 2-12 nM and was the only agonist with marked subtype selectivity (M₁/M₂ = 5.4, M₄/M₂ = 6.5). The M₁/M₂ selectivity of arecaidine propargyl ester was 4.4-fold, whereas all the other differences in the affinities for individual subtypes (including those concerning McN-A-343) were <3-fold.

Effects of allosteric ligands on the binding of [³H]NMS. Five allosteric ligands were investigated, and all of them have been found to have both positive and negative effects on the binding of [³H]NMS, depending on individual muscarinic receptor subtypes. These effects are shown (see Fig. 2, filled symbols), and the computed K_A and  values are summarized in Table 3. Increases in the binding of [³H]NMS were found with alcuronium and ()-eburnamonine on the M₂ and M₄ subtypes; with strychnine and brucine on the M₁, M₂, and M₄ subtypes; and with vincamine on the M₄ subtype.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   [3H]NMS binding to membranes of CHO cells expressing the human m1-4 (hm1-4) muscarinic receptor genes and incubated in the presence of increasing concentrations of the allosteric ligands alcuronium (, ), strychnine (, ), brucine (, ]), vincamine (, ), or eburnamonine (, ) and in the absence (filled symbols) or presence (open symbols) of acetylcholine. Abscissa, log10 of the concentration (M) of the allosteric ligand. Ordinate, [3H]NMS binding expressed as percent of the binding in the absence of acetylcholine and the allosteric ligands. The concentration of [3H]NMS was as described in the text. The concentration of acetylcholine was 100 µM in experiments with the M1 and M2 receptor subtypes and 50 µM in experiments with the M3 and M4 receptor subtypes. Symbols are mean ± standard error of three or four experiments, with incubations performed in triplicates. Curves, correspondence to eq. 3. Computed from these curves were the  coefficients describing the allosteric interactions between acetylcholine and the allosteric ligands; they have been listed in Table 4.

View this table: [in this window] [in a new window]   TABLE 3 Affinities of allosteric ligands for the M1-M4 muscarinic receptor subtypes in CHO cell membranes and cooperative interaction between the binding of allosteric ligands and [3H]NMS Affinities are expressed as Kd values for the binding of allosteric ligands to the receptors (KA; µM), and magnitudes of the allosteric interactions are described by the cooperativity coefficients . Data have been computed from measurements of the binding of [3H]NMS in the presence of increasing concentrations of the allosteric ligands and are mean ± standard error of four or five experiments with triplicate incubations.

	Alcuronium, M2 > M3 = M4 > M1
	Strychnine, M3 > M1 = M2 > M4
	Brucine, M4> M1 > M2 > M3
	Vincamine, M3 = M2 > M1 > M4
	()-Eburnmonine, M3 > M1 > M4 > M2

[³H]NMS binding in the simultaneous presence of an agonist and an allosteric modulator: Model binding curves. Fig. 1 shows simulated curves describing the expected changes in the binding of [³H]NMS in the presence of a fixed concentration of the orthosteric ligand L and of varying concentrations of the allosteric ligand A, provided there is positive cooperativity between the binding of [³H]NMS and A ( = 0.1; Fig. 1A) or negative cooperativity ( = 10; Fig. 1B) and the cooperativity between the binding of L and A is positive ( = 0.1), absent ( = 1.0), or negative ( = 10). K_NMS was taken as equal to 100 pM, K_A and K_L as equal to 1 µM, and the concentration of L as equal to K_L. The continuous curves have been computed according to eq. 3. The dashed curves represent the model of [³H]NMS binding described by eq. 4, which assumes that the binding of L to the orthosteric site sterically prevents the binding of A to the allosteric site and that the binding of A to the allosteric site sterically prevents the binding of L to the orthosteric site.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Binding of [3H]NMS in the presence of an allosteric ligand A at the concentrations (abscissa) and in the presence (open symbols) or absence (filled symbols) of an unlabeled orthosteric ligand L at a fixed concentration. Continuous curves, correspondence to eq. 3. Dotted curves, correspondence to eq. 4. The assumptions were (a) the cooperativity between A and [3H]NMS is either positive ( = 0.1; left) or negative ( = 10; right), (b) the concentration of [3H]NMS corresponds to KNMS and equals 100 pM, (c) the cooperativity between A and L is either positive ( = 0.1), neutral ( = 1), or negative ( = 10 or 100); values of  are indicated near the curves, (d) the concentration of L is 1 µM and equals KL, (e) KA = 1 µM, and (e) the binding of L to the orthosteric binding site prevents the binding of A to the allosteric binding site, and vice versa (squares and dotted lines). Abscissa, log10 of the concentration of A (M). Ordinate, [3H]NMS binding expressed as percent of the binding in the absence of A and L.

[³H]NMS binding in the simultaneous presence of an agonist and an allosteric modulator: Actual measurements. To illustrate the type of data obtained when [³H]NMS binding was measured in the presence of varying concentrations of the allosteric ligand alone or in the presence of varying concentrations of the allosteric ligand and of a fixed concentration of an agonist, Fig. 2 shows the results of experiments with [³H]NMS binding to the M₁-M₄ receptors in the absence or presence of acetylcholine and in the presence of varying concentrations of alcuronium, strychnine, ()-eburnamonine, vincamine, and brucine. Experiments of the type shown in Fig. 2 permitted us to evaluate the allosteric interaction between the unlabeled allosteric ligand and the unlabeled agonist. This was done by fitting eq. (3) to the data obtained. Values of the cooperativity coefficients are summarized in Table 4.

View this table: [in this window] [in a new window]   TABLE 4 Factors of cooperativity () between the binding of muscarinc agonists and allosteric ligands to the M1-M4 subtypes of muscarinic receptors in the membranes of CHO cells Values of  were computed from the results of experiments with the binding of [3H]NMS in the presence of a fixed concentration of the agonist and of increasing concentrations of the allosteric ligand, according to eq. 3. Data are mean of three or four experiments with triplicate incubations. Relative errors were <10% for each value. No data have been included for the interactions between alcuronium and oxotremorine and its analogues on the M1, M2, and M4 receptor subtypes because a good fit could not be obtained with eq. 3 for any  value conceivable.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   [3H]NMS binding to membranes of CHO cells expressing the human m1-4 (hm1-4) muscarinic receptor genes and incubated in the presence of increasing concentrations of alcuronium (Alc) and a stable concentration of muscarinic agonists McN-A-343 (), oxotremorine (), or oxotremorine-M () or in the absence of any agonist (). The concentrations of [3H]NMS were as indicated in the text. The concentrations of agonists used in experiments on the M1, M3, and M4 subtypes were 8, 5, and 5 µM for McN-A-343, oxotremorine, and oxotremorine-M, respectively. On the M2 subtype, the respective concentrations were 10, 8, and 8 µM. Symbols are mean ± standard error of four experiments with incubations in triplicates. Abscissa, log10 of the concentration of alcuronium. Ordinate, [3H]NMS binding expressed as percent of the binding in the absence of alcuronium and agonists. The curves for the binding in the presence of agonists correspond to eq. 4, assuming that simultaneous binding of alcuronium to the allosteric site and of oxotremorine, oxotremorine-M, and McN-A-343 to the orthosteric site is not possible for steric reasons on the M1, M3, and M4 receptor subtypes.

	Discussion

Summary Introduction Materials & Methods Results Discussion References

The main finding of the current study consists in the discovery of positive effects of allosteric modulators on the affinity of muscarinic receptors for muscarinic agonists, including acetylcholine. Only negative effects of allosteric modulators on the affinities of muscarinic receptors for agonists have been reported (22-25). The discovery is important for several reasons: (a) It suggests that drugs may be developed with the potential for increasing the efficiency of cholinergic (muscarinic) synaptic transmission in a very physiological way, by enhancing the sensitivity of postsynaptic receptors toward their natural agonist. (b) It points to a new way to achieve subtype- and tissue-specific activation of muscarinic receptors. Although the subtype selectivity of most muscarinic agonists is low, suitable allosteric modulators should permit the selective enhancement of the effect of a nonselective muscarinic agonist on a single receptor subtype and suppression of it on the other receptor subtypes (e.g., ()-eburnamonine-induced changes in the affinities for acetylcholine). (c) The enhancement of synaptic transmission by allosteric modulators is likely to be extremely safe and completely activity dependent. The effects of allosteric ligands have clear-cut inherent pharmacodynamic limits because they alter the K_d values for agonists and antagonists within a defined range. This should help to avoid receptor overstimulation (by allosteric enhancers) and overinhibition (by allosteric inhibitors). It is also important that the allosteric enhancers will act only when the presynaptic nerve terminal is active and the neurotransmitter is released, because they must bind to receptors simultaneously with the neurotransmitter to have an effect (26, 27). The question of whether allosteric enhancers will affect receptor desensitization requires specific investigation.

It seems unlikely that the five allosteric modulators we examined will be used as allosteric enhancers in medical practice, mainly because their affinities for receptors are low (K_A = 0.7 × 10⁶ to 10⁴ × 10⁶; Table 3) and because some of them have additional undesirable effects. It is not difficult to imagine, however, that chemically related compounds with higher potency and allosteric efficacy and with no toxicity will be discovered, which will prove suitable for modulation of muscarinic receptor function in medical practice. It seems impossible to say at present which features of the modulator, receptor, and agonist determine the direction (positive or negative) of the cooperative interaction; we can only offer several comments relevant to this point:

Receptor subtype. Literature on the negative allosteric effects of neuromuscular blockers on muscarinic receptors (28) suggests that the M₂ receptor subtype is most prone to allosteric modulation. Observations that the allosteric enhancement of the binding of NMS by alcuronium (4, 5, 7, 8) and strychnine (11, 12) is strongest on the M₂ receptors and that alcuronium has the highest affinity for the M₂ receptors (6) led to anticipation that the positive allosteric effects on the binding of agonists would also mainly occur on the M₂ receptor subtype. The present data do not support such expectation. Allosteric enhancement of agonist binding could be revealed on all four receptor subtypes examined. The affinity for the physiological agonist acetylcholine was allosterically enhanced on the M₁ and M₃ subtypes by brucine and on the M₂ and M₄ subtypes by ()-eburnamonine.

Affinity for the allosteric modulator. In the membranes of transfected CHO cells, the values of K_d for the binding of alcuronium to muscarinic receptor subtypes were 9.33, 0.62, 1.56, 1.31, and 22.0 µM on the M₁, M₂, M₃, M₄ and M₅, receptor subtypes, respectively (6), and the cooperativity between alcuronium and NMS was positive on the M₂ and M₄ and negative on the M₁, M₃, and M₅ subtypes. These data suggested that the effects of an allosteric ligand might be generally positive on those receptor subtypes with which the ligand associates with a high affinity. Inspection of Tables 3 and 4 indicates, however, that such generalization is not substantiated by data obtained with the other allosteric modulators, neither in relation to their cooperativity with the antagonist NMS nor in relation to their cooperativity with the agonists. Although strychnine had the highest affinity for the M₃ receptors, its cooperativity with [³H]NMS binding was positive on the M₁, M₂, and M₄ receptors. Similarly, positive cooperativity between ()-eburnamonine and [³H]NMS was observed on the two receptor subtypes for which the affinity of ()-eburnamonine was the lowest. No generally valid correlation is evident between the rank orders of the affinities of allosteric modulators for receptors and the rank orders of the  factors computed for their interactions with individual agonists. For example, brucine has a 100-fold higher affinity for the M₄ than for the M₃ subtype, but its cooperativity with most agonists is similar on both subtypes.

Affinity for the agonist. This factor is difficult to consider with the current data because the experiments were performed in the presence of 0.5 mM GTP, which eliminated the high affinity binding of agonists. Under these conditions, no consistent correlation is apparent between the K_L values of agonists in Table 2 and the relevant cooperativity factors  in Table 4. For complete analysis, it will be necessary to explore the cooperativity between agonists and allosteric modulators under conditions in which the receptor/G protein coupling occurs in a normal way, but this will be demanding in view of the complexity of the system and of recent data indicating that the allosteric modulators themselves modify the receptor/G protein interactions (29).