Los subtipos M3 y M2 del mAchR fueron localizados en fracciones de membranas plasmáticas de músculo liso traqueal de bovino (MLTB). Los mAchRs nativos de MLTB fueron estudiados usando enlazamiento de [3H]QNB. La hetereogeneidad del receptor fue expresada para agonistas y antagonistas muscarínicos siendo los estados de alta afinidad sensibles a GTPgS. Mostramos que los efectos de GTPgS parecen ser mediados por proteínas Gi/o sensibles a PTX. Se detectó la hetereogeneidad del mAchR hacia 4-DAMP (M3) y pirenzepina (M1). Sin embargo, los antagonistas subtipo M2 como methoctramina y AF-DX 116 no mostraron tales respuestas en la presencia o ausencia de GTPgS. A través de la alquilación con 4-DAMP del subtipo M3 del mAchR, prevaleció el subtipo M2 el cual exhibió un enlazamiento dependiente de GTPgS parecido a un agonista mientras el enlazamiento de los antagonistas fue insensible al GTPgS. Por lo anterior, la heterogeneidad a antagonistas del receptor descrita aquí es debida al subtipo M3 del mAchR, el cual muestra dos estados de afinidad siendo uno regulado por proteínas Gi/o.

Palabras Clave: Agonistas muscarínicos, Antagonistas muscarínicos, Pertussis toxin, GTPgS, Músculo liso traqueal.

The M3 and M2 mAchR subtypes were located in plasma membranes fractions from bovine tracheal smooth muscle (BTSM). Native mAchRs from BTSM were studied by using [3H]QNB binding. Receptor heterogeneity was expressed for muscarinic agonists and antagonists being the high affinity states sensitive to GTPgS. We showed that GTPgS effects seem to be mediated by PTX sensitive Gi/o proteins. Antagonist mAchR heterogeneity towards 4-DAMP (M3) and pirenzepine (M1) was detected. However, M2 antagonists as methoctramine and AF-DX 116 did not show such responses either in the presence or absence of GTPgS. Through 4-DAMP alkylation of the M3 mAchR subtype, a M2 subtype prevailed, which exhibits GTPgS-dependent agonist binding whereas antagonist binding was GTPgS-insensitive. Thus, antagonists binding receptor heterogeneity here described is due to the M3 mAchR subtype showing two affinity states, one being regulated by Gi/o-proteins.

Alkylation of plasma membranes with 4-DAMP mustard was performed in excess of methoctramine to protect the M2 mAchR subtype as described elsewhere (Reddy et al., 1995) with some modifications. Briefly, the aziridinium ion was generated from incubation of the M3-selective irreversible receptor antagonist 4-DAMP mustard at 37° for 30 min. Then, washed plasma membrane fractions were exposed to freshly prepared 4-DAMP mustard (40 nM) aziridinium ion for 60 min, in the presence of methoctramine (1 mM). Thus, the reaction mixture proceeded for 10 minutes in the presence of dithiotreitol (1 mM) and the membranes recovered after centrifugation at 150,000g for 30 min. Finally, the membrane sediment was washed and suspended in 20 mM Tris-HCl (pH 7.2)- buffer and centrifuged four times at a 150,000g for 30 min.

The [3H]QNB displacement studies were performed as described previously (Misle et al., 1994; Misle et al., 1995). Briefly, [3H]QNB binding assay was started by adding membrane protein (2-5 mg) in low ionic strength buffer as 66 mM Tris-HCl (pH 7.8) and increasing concentration of unlabeled drug plus L-[3H]QNB (0.625 nM) to a final volume of 120 ml. After one hour of incubation at 37°, the incubation mixture was placed onto a pre-centrifuged Sephadex G-50 column (3 ml) equilibrated with 0.25 M sucrose-5 mM Tris-HCl (pH 7.8) and immediately centrifuged at 700g for 1.5 min to remove free [3H]QNB. The column effluent containing 95-98% of the bound [3H]QNB was transferred to vials containing Aquasol®. Radioactivity was measured in a RackBeta liquid scintillation counter LKB, Wallac 1214/1219) and all the samples were counted with approximately the same efficiency. Specific binding was calculated by subtracting the non-specific binding (which was less than 1% of total binding measured with 1 mM atropine sulfate), from the total binding (Fields et al., 1978; Misle et al., 1995). In all binding experiments, no more than 5% of the fixed radioligand concentration was allowed to bind to the membranes. Similar amounts of active receptors were employed in the agonist/antagonist binding displacement experiments.

Activation of Pertussis toxin and the ADP-ribosylation reaction were performed as described elsewhere (Woolkalis et al., 1986) and modified by us (Lippo de et al., 1995). Briefly, the pre-activation reaction was performed in 50 mM HEPES-NaOH, pH 8.0; containing: 0.1 mg/ml BSA, 20 mM DTT, 0.125 % SDS and 20-25 mg of toxin in a volume of 100 ml at 30°. The ADP-ribosylation reaction was carried out in a volume of 1 ml in the presence of 50 mM MOPS, pH 8.0, 1 mM ATP, 0.1 mM GTP, 5 mM MgCl2, 1 mM thymidine and 1 mM NAD+ for 1 h at 30°. The ADP-ribosylation reaction was started by the addition of washed plasma membranes fraction (1 mg). To stop the reaction and remove the ions and nucleotides from the ADP-ribosylation step, aliquots of the reaction mixture (0.3 ml) was applied to a pre-packed (3 ml) Sephadex G-50 column. The last pre-packed column was equilibrated with 0.3 M sucrose-20 mM Tris-HCl buffer, pH 7.6 and centrifuged for 70 sec to obtain the membrane material at the void volume of the column. Samples membranes were run also without PTX in the ADP ribosylation assay to serve as control (-PTX). All treated membrane material was kept on ice and the protein content was determined before the determination of [3H]QNB activity.

The effects of GTPgS on the mAchRs binding properties, muscarinic (agonists and antagonists) displacement experiments were performed. Table 1 shows that IC50 in the absence of GTPgS, for agonists such as carbamylcholine, metacholine and oxotremorine Furthermore, the pseudo-Hill coefficients (np-H) values for these agonists were lower than 1.0, suggesting the presence of agonist receptor heterogeneity. Nonetheless, GTPgS induced dramatic changes in the np-H, which reached values of 1.0 indicating the disappearance of the agonist dependent receptor heterogeneity (Table 1). Moreover, in the presence of GTPgS, the IC50 values for agonists became similar indicating the existence of "one receptor" agonist population. This agonist-dependent muscarinic receptor heterogeneity was analyzed by applying the "two sites" model as the best fitting for all agonists assayed. Table 1 shows that two agonist-binding sites such as the high affinity (RH) and low affinity (RL) states were found. These two agonist states are in a ratio 1:1 in the presence of 5 mM MgCl2. Interestingly, upon addition of 100 mM GTPgS, all agonist displacement curves were significantly shifted to the right as shown in Figure 1. A more careful analysis (Table 1) showed that, GTPgS induced significant changes in the carbamylcholine, metacholine and oxotremorine binding parameters producing a complete disappearance of the RH states together with an increment in the relative amount of receptors in the RL state to 100%.

Figure 1: Comparison of agonist-[3H]QNB competitive curves for bovine tracheal smooth muscle plasma membranes in the presence of 5 mM MgCl2 () and 5 mM MgCl2 plus 100 mM GTPgS (). Specific [3H]QNB binding expressed as percentage of binding in the absence of agonists, carbamylcholine oxotremorine and metacholine. Concentration of [3H]QNB was 625 pM and 2-3 mg of membrane proteins were assayed at 37o. Other details are given in Materials and methods. Each point represents the mean of four different membrane preparations assayed by triplicate. The S. E. of these values was less than 5% of the mean.

Antagonists binding properties were also studied by using competition assays. Compounds such as 4-DAMP (M3 antagonist), AF-DX 116 and methoctramine (M2 antagonist) and pirenzepine (PZ), (M1 antagonist) in the presence of MgCl2 and GTPgS were evaluated. These results are shown in Figure 2 and Table 2. We have previously shown that in the basal condition (Misle et al., 1994), all muscarinic antagonists displacement curves fit the "one site" model with np-H values close to 1.0 indicating that these drugs seem to bind to a homogeneous receptor-binding site. In the present work, methoctramine, and AF-DX 116 did not change their binding isotherms, either in the presence of Mg2+ or GTPgS showing np-H values close to 1.0 (Figure 2 and Table 2). Surprisingly, the antagonists 4-DAMP and PZ showed competitive displacement curves with pseudo-Hill coefficients (np-H) lower than 1.0, suggesting the presence of antagonist receptor heterogeneity in the mAchR. Analysis of the latter antagonist competitive curves obtained in the presence of MgCl2 by applying the "two sites" model unveiled two binding sites in the mAchRs. These new sites were designated as Rh and Rl. The high antagonist affinity (Rh) sites were unmasked in the presence of Mg2+ showing a pIC50 of 8.1 ± 0.2 for 4-DAMP and 6.5 ± 0.2 for PZ, these novel Rh sites disappeared upon addition of GTPgS as shown in Table 2. 4-DAMP and PZ competitive displacement curves also showed changes induced by GTPgS similar to the agonist curves above described. These latter antagonist curves exhibit np-H values of 1.0 indicating the disappearance of the antagonist dependent receptor heterogeneity.

Figure 2: Comparison of antagonist-[3H]QNB competitive curves for bovine tracheal smooth muscle plasma membranes in the presence of 5 mM MgCl2 () and 5 mM MgCl2 plus 100 mM GTPgS (). Specific [3H]QNB binding is expressed as percentage of binding in the absence of antagonist, 4-DAMP, methoctramine, pirenzepine and AF-DX 116. Concentration of [3H]QNB was 625 pM and 2-3 µg of membrane proteins were assayed at 37o. Other details are given in Materials and methods. Each point represents the mean of four different membranes preparations assayed by triplicate. The S. E. of these values was less than 5% of the mean.

This antagonist-dependent heterogeneity may be related to a mixture of M2 and M3 mAchR subtypes as above mentioned. In order to explore this possibility, we treated the native plasma membrane fractions with a selective M3-alkylating agent (4-DAMP mustard), and protecting the M2 subtype with methoctramine as described in Methods. As expected, the alkylated membranes (Table 3) showed IC50 values, which correspond to a M2 pharmacological pattern (4-DAMP · methoctramine > AF-DX 116 > PZ). Interestingly, all antagonist-binding isotherms in the alkylated membranes showed np-H values of 1.0 and fitted for "one" homogeneous binding site, which was unaffected in the presence of GTPgS (Table 3). However, the competition experiments in the presence of the agonist carbamylcholine still showed high (RH) and low sites (RL), (Table 3). Moreover, The competition experiment with this agonist using alkylated membranes were affected by GTPgS in such a way that the high agonist sites (RH) completely disappeared in the presence of this guanine nucleotide analog leaving only low agonist sites (Table 3).

The GTPgS effects seem to be mediated though G proteins. Thus, additional experiments were performed on native membranes using Pertussis toxin (PTX) to inactivate Gi/o proteins as described in Methods. The PTX effect was only evaluated with the drugs that exhibited the "maximal right-shift" responses in the presence of GTPgS, such as the agonist (oxotremorine) and the antagonist, (PZ). Thus, PTX treatment of native membranes produced a complete decrease of the high agonist affinity binding state (RH) for oxotremorine in the presence of Mg2+ (Table 4). However, GTPgS did not modify the latter response, as shown in Table 4. Similarly, when the antagonist, PZ was assayed on PTX-treated membranes, in the presence of Mg2+, the high affinity antagonist-state (Rh) disappeared. Again, GTPgS did not further affect the PZ displacement curve (Table 4) indicating that PTX abolished the PZ dependent Rh state of these mAchRs.

Our results using classic muscarinic agonists such as carbamylcholine, metacholine and oxotremorine support these previous findings. Muscarinic agonists operate through several affinity states of the receptor (Mitchelson, 1998, Hulme et al., 1990). The high agonist affinity state (RH) is the receptor-coupled G-protein complex (G-R*), while the low agonist affinity state (RL) is the free uncoupled receptor (R) (Hulme et al., 1990). In the present work, mAchR agonist heterogeneity was detected in the presence of Mg2+, and disappeared in the presence of GTPgS. These GTPgS effects are explained by the irreversible activation of G-proteins by this nucleotide producing the dissociation of the G-R*, disappearing the RH and displaying the free RL (Birnbaumer et al., 1982; Ham and Gilchrist, 1996). The GTPgS and the PTX effects on agonists binding here described support the fact that PTX-sensitive Go/i-proteins are involved in these mAchR responses. Whenever, the ADP-ribosylation of G-proteins (Go/i) is catalyzed by PTX, its biological action is perturbed (Birnbaumer et al., 1982). Interestingly, in such isolated systems as plasma membrane fragments, the relative amount of mAchRs subtypes should not be affected by GTPgS or PTX treatments.

In the present work the involvement of mAchR subtypes in these GTPgS effects was investigated using specific muscarinic antagonists. Previous studies have shown that GTP analogs have different opposite effects on receptor affinity for antagonists depending on the system under study. In chick heart, Gpp(NH)p increased receptor affinity (Hosey, 1982), but GTPgS decreased receptor affinity in rat parotid membranes (Horn et al., 1991). Thus, [3H]QNB displacement curves were performed with antagonists related to the M2, M4 family such as methoctramine and AF-DX 116 and the M3, M1, M5 family as 4-DAMP and PZ. M2 antagonists (methoctramine and AF-DX 116) bind to tracheal smooth muscle mAchRs with a similar pharmacological pattern either in the presence or absence of GTPgS. Apparently, these compounds are no able to detect the two-mAchR states (Birdsall et al., 1983) induced by Mg2+ and regulated by GTPgS through G-proteins as suggested by PTX experiments. Therefore, these drugs bind with equal affinity to both "G-R*" and "R" states behaving as "neutral antagonists" as defined elsewhere (Costa et al., 1991).

However; M3 and M1 antagonists such as 4-DAMP and PZ, exhibited heterogeneous binding isotherms with two different binding sites in the presence of Mg2+, being sensitive to GTPgS suggesting that G-proteins may be involved in this biological action. This is an original finding related to 4-DAMP and PZ showing a G-protein dependent differential binding in these mAchRs associated with a plasma membranes fraction isolated from tracheal smooth muscle. The apparent heterogeneity of 4-DAMP and pirenzepine seems to be specific for these drugs because under identical experimental conditions specific M2 antagonists such as methoctramine and AF-DX 116 did not modify the displacement pattern curves. In addition, the competitive curves obtained with the four antagonists here studied using the 4-DAMP mustard alkylated membranes, showed a classic M2 profile.

This antagonist heterogeneity showed by 4-DAMP and PZ in these plasma membranes fractions is assumed as two populations of binding sites for muscarinic antagonists as described in rat heart (Hulme et al., 1981), colonic circular smooth muscle (Zhang et al., 1992) and airway smooth muscle (Luchessi et al., 1990; Mak et al., 1992; Roffel et al., 1988). However, there was not direct experimental evidence to support that assumption. In this work, we found that the 4-DAMP alkylated plasma membranes displayed a pharmacological profile that belongs to a M2 subtype as expected after abrogating the M3 subtype. In this sense, our results with 4-DAMP mustard supported the existence of both M2 and M3 mAchR subtypes in these plasma membranes fractions. Also, specific binding affinities for antagonists of the "remaining M2 subtype" were not affected by GTPgS. Nonetheless, the binding affinity for the agonist, carbamylcholine was affected by GTPgS showing the functionality of the "remaining M2 subtype". The latter finding suggests that, under our experimental conditions, the M3 subtype is the mAchR subtype involved in the GTPgS-dependent heterogeneous binding activity here observed. At this time it is difficult to explain the molecular mechanism that allow the "agonist-like behavior" exhibited by 4-DAMP and PZ.

Our results using binding assays indicated that the M3 mAchR is coupled to PTX-sensitive Gi/o. A similar signal transducing-cascade is involved in the heterologous potentiation of PLC activity and PLC enhanced substrate supply in human embryonic kidney cells (Schmidt et al., 1996). It is possible that this PLC-Gi/o coupled to M3 mAchR cascade exists in these plasma membranes fractions and exerts an important role in the contraction/relaxation of tracheal smooth muscle as suggested elsewhere (Yang et al., 1993). Although, GTPgS does not affect M2 mAchR antagonist binding, the agonist binding is indeed affected, suggesting that this subtype is coupled to G-proteins.

mAchR, muscarinic acetylcholine receptor; GTPgS, guanosine 5’-(3-O-thio) triphosphate; DTT, 1,4-dithiothreitol; [3H]QNB, L-[benzilic-4,4'-3H]-quinuclidinyl benzilate; 4-DAMP, 4-diphenylacetoxy-N-methylpiperidine; PTX, Pertussis toxin; AF-DX 116, 11[[2-[(diethyl-amino)methyl]-1-piperidinyl]acetyl]-5,11-dihydro-6H-pyrido-

[2,3-b][1,4]benzodiazepine-6-one.