Materials and Methods

Animals and Tissue Preparation.

Male guinea pigs, weighing 350 to 450 g, were sacrificed by cervical dislocation. Terminal ileum was removed after the 10 cm nearest to the ileocecal junction had been discarded and strips of myenteric plexus with the longitudinal muscle attached (MP-LM) were prepared according to Paton and Vizi (1969).

Single Twitch Stimulation.

Four MP-LM strips (3–4 cm in length) from each animal were mounted in separate 3-ml organ baths perfused with Tyrode's solution of the following composition: 136.9 mM NaCl, 2.7 mM KCl, 1.2 mM CaCl₂, 1.04 mM MgCl₂, 11. 9 mM NaHCO₃, 0.4 mM NaH₂PO₄, and 10 mM glucose, at 37°C, and continuously oxygenated (95% O₂, 5% CO₂, pH 7.2–7.4). An initial tension of 0.5 g was applied, and the longitudinal muscle activity was recorded through an isometric force transducer. After 120-min perfusion at 1 ml/min, longitudinal muscle repetitive twitch contractions (developing tension of 1.2–2.3 g) were evoked by field stimulation (alternate rectangular pulses, 0.1 Hz, 1-ms duration, 550–600 mA) delivered from two platinum electrodes. The use of submaximal voltage was a precaution against the effects of drugs being masked by supramaximal pulses (see Fosbraey and Johnson, 1980).

The effects of drugs were measured as percent variation of the cholinergic twitch contraction height. To avoid desensitization (Bowery et al., 1981; Ong and Kerr, 1983), agonists were applied at 30-min intervals; when the response to any agonist concentration had reached a maximum, tissues were washed and perfused at 1 ml/min for 20 min. Two complete concentration-response curves were conducted in the same tissue. When the effects of antagonists on the agonist inhibition of twitch contraction were evaluated, after a first concentration-response curve for the agonist was completed in standard medium, a second concentration-response curve was made in standard medium (control) or in the presence of the antagonists (one antagonist concentration only per tissue). Antagonists were added to the perfusion medium 20 min before their effects were evaluated. When the effects of adding CGP 47656 on (−)-baclofen was evaluated, after a first concentration-response curve for (−)-baclofen was completed, a second concentration-response curve was made for (−)-baclofen alone (control) or by adding (−)-baclofen together with CGP 47656 at various concentrations. In each experiment, a preparation was run as control; in the experimental conditions used in control preparations, the agonist EC₅₀ value in the second curve did not significantly differ from the EC₅₀ value in the first curve. Data obtained in first (agonists alone) and second (in the presence of antagonists) curves are given in the figures. In preliminary experiments, the effects of compounds on longitudinal muscle contractile responses to acetylcholine were evaluated in MP-LM preparations.

Data Analysis.

Concentration-response curves for agonist inhibition of the twitch contraction were analyzed by a four-parameter logistic function analysis (SigmaPlot software). pD₂ values of agonists were measured as −log EC₅₀ values (EC₅₀ is the concentration producing 50% of the drug maximum effect). The pA₂ value of CGP 52432 was measured according toArunlakshana and Schild (1959); when the concentration range (equal to or less than 10-fold) of the antagonists precluded construction of a Schild plot to provide true pA₂ values, apparent pA₂ values were determined using the relationship pA₂ = log (EC₅₀ ratio − 1) − log (B) in the presence of at least two different antagonist concentrations (B). The pK_Bvalue for CGP 47656 antagonism of (−)-baclofen was estimated from agonist dose ratios producing half-maximum responses in accordance with the model for partial agonism: pK_B = log (EC₅₀ ratio − 1) − log (B) in the presence of two different CGP 47656 concentrations (B). Mean ± S.E. values of determination in n separate experiments are indicated throughout. The statistical significance of the differences between mean values was assessed by the Student's t test. A probability level of P < .05 was taken as statistically significant.

Drugs.

Acetylcholine hydrochloride, GABA, and (−)-bicuculline methobromide were purchased from Sigma Chemical Co. (St. Louis, MO). Phaclofen was purchased from Tocris Cookson (Bristol, UK). (−)-Baclofen, 3-aminopropyl(diethoxymethyl)phosphinic acid (CGP 35348), 3-aminopropyl-n butyl phosphinic acid (CGP 36742), 3-aminopropyl-(difluoromethyl)phosphinic acid (CGP 47656), [3-[[(3,4-dichlorophenyl) methyl]amino]propyl](diethoxymethyl) phosphinic acid (CGP 52432), and [3-{[1-(R)-(3-carboxyphenyl)ethyl]amino}-2-(S)-hydroxy-propyl]cyclohexyl-methyl-phosphinic acid (CGP 56999) were gifts from Novartis (Basel, Switzerland). All compounds were dissolved in distilled water or Tyrode's solution.

Results

Effects of GABA_B Receptor Agonists on Cholinergic Twitch Contraction.

GABA and (−)-baclofen inhibited the cholinergic twitch contraction by a maximum of about 80%, exhibiting quite close potencies (pD₂ for GABA = 5.70 ± 0.08, n = 9; pD₂ for (−)-baclofen: 5.33 ± 0.05, n = 32; Fig.1). GABA, at the concentration of 30 μM, induced transient, bicuculline-sensitive contraction superimposed on the repetitive twitch (data not shown), before inhibition of twitch contraction.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Effects of GABA (●), (−)-baclofen (▪), or CGP 47656 (♦) on cholinergic twitch contractions in guinea pig ileum MP-LM preparations. Twitch contraction height in the presence of drugs was measured as percentage of the contraction before the addition of the drugs (control). Each point represents the mean ± S.E. of 6 to 32 determinations in separate experiments.

The compound CGP 47656 reduced the cholinergic twitch contraction by a maximum of 35.2 ± 3.7% (mean of n = 24 experiments), exhibiting a pD₂ value of 5.42 ± 0.10 (n = 24; Fig. 1). The inhibition of the contraction produced by 30 μM CGP 47656 (32.8 ± 4.5%;n = 8) was abolished in the presence of various GABA_B receptor antagonists, including phaclofen (10 mM), CGP 35348 (300 μM), CGP 52432 (3 μM), and CGP 36742 (30 μM). Occasionally, CGP 47656 (300 μM) elicited a transient bicuculline (30 μM)-sensitive contraction superimposed on the repetitive twitch before inhibition of twitch contraction. Up to the highest concentrations used, CGP 47656 had no effect on the contractile responses to acetylcholine (0.01–0.1 μM; data not shown).

In the presence of the GABA_A receptor antagonist bicuculline (30 μM), the potencies of GABA, (−)-baclofen, and CGP 47656 in inhibiting twitch contraction remained unaffected (pD₂ for GABA = 5.50 ± 0.10,n = 3; pD₂ for (−)-baclofen = 5.17 ± 0.07, n = 18; pD₂for CGP 47656 = 5.20 ± 0.12, n = 4).

To better understand the effect of CGP 47656, apparently characteristic of a partial agonist, concentration-response curves for (−)-baclofen were constructed in the presence of varying concentrations (3, 30, or 300 μM) of CGP 47656. The results, illustrated in Fig.2, show that the effects of low concentrations of (−)-baclofen were potentiated by CGP 47656, whereas the effects of higher concentrations of (−)-baclofen were depressed by CGP 47656; the estimated CGP 47656 pK_Bvalue was 5.62 ± 0.05 (n = 4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Effects of (−)-baclofen and CGP 47656 on cholinergic twitch contractions in guinea pig ileum MP-LM. Concentration-response relationships for (−)-baclofen alone (▪) or added together with 3 μM (■), 30 μM (▿), or 300 μM (▵) CGP 47656 are shown. Twitch contraction height in the presence of drugs was measured as percentage of the contraction before the addition of the drugs (control). Each point represents the mean ± S.E. of three to six separate determinations.

Blockade of (−)-Baclofen Effect by GABA Receptor Antagonists.

The GABA_B receptor antagonists phaclofen, CGP 35348, CGP 36742, and CGP 52432 shifted to the right the concentration-response curve of (−)-baclofen in a way compatible with competitive antagonism (Fig. 3). In confirmation of previous data (Kerr et al., 1990b; Ong et al., 1994), CGP 35348 was more potent than phaclofen, with apparent pA₂ values for CGP 35348 and phaclofen of 5.02 ± 0.09 (n = 4) and 3.90 ± 0.06 (n = 3), respectively. The compound CGP 36742 was almost equipotent with CGP 35348 (apparent pA₂ = 4.88 ± 0.10, n = 6), whereas CGP 52432 exhibited the highest potency (pA₂ = 7.82 ± 0.08,n = 6).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Competitive antagonism of the (−)-baclofen effect by phaclofen, CGP 35348, CGP 52432, and CGP 36742. Phaclofen: concentration-response curves of (−)-baclofen alone (▪) or in the presence of 100 μM (■) or 1000 μM (▿) phaclofen. CGP 35348: concentration-response curves of (−)-baclofen alone (▪) or in the presence of 3 μM (■), 30 μM (▿), or 300 μM (▵) CGP 35348. CGP 52432: concentration-response curves of (−)-baclofen alone (▪) or in the presence of 0.01 μM (○), 0.03 μM (■), 0.1 μM (▿), 0.3 μM (▵), or 3 μM (⋄) CGP 52432. CGP 36742: concentration-response curves of (−)-baclofen alone (▪) or in the presence of 3 μM (■), 30 μM (▿), or 300 μM (⋄) CGP 36742. Twitch contractions in the presence of (−)-baclofen were evaluated as percentages of the contraction measured before adding the drug (control). Antagonists were added 20 min before the effect of (−)-baclofen was evaluated. Each point represents the mean ± S.E. of 3 to 12 determinations in separate experiments.

The compound CGP 56999 potently inhibited the (−)-baclofen effect; however, the shifts of the (−)-baclofen concentration-response curve produced by this drug are suggestive of a noncompetitive antagonism (Fig. 4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

The compound CGP 56999 is a potent noncompetitive antagonist of (−)-baclofen. Concentration-response curves of (−)-baclofen alone (▪) or in the presence of 0.03 nM (■), 0.3 nM (▿), 1 nM (▵), or 3 nM (⋄) CGP 56999 are shown. Twitch contractions in the presence of (−)-baclofen were evaluated as percentages of the contraction measured before the addition of the drug (control). CGP 56999 was added 20 min before the effect of (−)-baclofen was evaluated. Each point represents the mean ± S.E. of three to six determinations in separate experiments.

The GABA_B receptor antagonists, at the highest concentrations used, slightly increased the twitch contraction height (by about 10%) leaving unaltered the longitudinal muscle tone, with the exception of phaclofen, which, at the concentration of 1 mM, evoked a transient, bicuculline (30 μM)-sensitive contraction superimposed on twitch response. None of the antagonists tested affected the contractile responses to acetylcholine; the acetylcholine EC₅₀ values in the presence of the highest antagonist concentrations used ranged from 53 ± 6.7 to 43 ± 5.9 nM (n = 3), which did not differ from 47 ± 5.2 nM (n = 6), the EC₅₀ value in standard medium.

Discussion

GABA and (−)-baclofen inhibited cholinergic twitch contraction in guinea pig ileum MP-LM preparations, exhibiting potencies similar to those found for the two agonists in previous studies with peripheral tissues (Bowery et al., 1981; Ong and Kerr, 1983; Ong et al., 1994). Moreover, in keeping with published reports (Kerr et al., 1987, 1990b;Ong et al., 1994), the receptor acted on by (−)-baclofen was sensitive to both phaclofen and CGP 35348.

Studies performed in our laboratory with CNS nerve terminals have shown that phaclofen- and CGP 35348-sensitive GABA_Breceptors are, however, heterogeneous (Gemignani et al., 1994; Bonanno et al., 1999). In particular, based on the differential sensitivities to other GABA_B-selective ligands, including CGP 47656, CGP 52432, and CGP 36742 (see Table1), two pharmacologically distinct subtypes within the phaclofen- and CGP 35348-sensitive group of GABA_B receptors were found to exist in the rat cerebral cortex, where they are sited, respectively, on somatostatin (SRIF)- and cholecystokinin (CCK)-releasing axon terminals and mediate inhibition of neuropeptide release.

The results of the present work show that the pharmacological profile of the GABA_B receptor inhibiting cholinergic twitch contraction in guinea pig ileum MP-LM can be clearly distinguished from those of the GABA_B receptors regulating peptide release in the rat cerebrocortex. The intestinal receptor exhibits the highest apparent affinity for CGP 52432 and much lower affinity for CGP 36742, whereas the GABA_Breceptor present on SRIF-releasing terminals in the rat cortex shows the highest affinity for CGP 36742. Moreover, CGP 52432 could block the guinea pig intestinal GABA_B receptor with a potency much higher than CGP 35348, whereas the two antagonists displayed identical potency at the rat cortical receptor involved in the release of SRIF. The pharmacology of the guinea pig ileum receptor does not resemble that of the GABA_B receptor regulating CCK release in rat neocortex either. In fact, CGP 36742 was ineffective at the latter receptor up to 100 μM (Bonanno et al., 1999); in contrast, CGP 36742 was an antagonist as potent as CGP 35348 at the guinea pig ileum receptor. Moreover, although arguments based on agonist effects are of limited value to receptor characterization, CGP 47656 up to 100 μM, which is ineffective as an agonist or antagonist at the GABA_B receptor regulating CCK release (Gemignani et al., 1994), inhibited twitch contractions with potency comparable to that of (−)-baclofen, although with lower efficacy (see Fig. 1). Because modulation of cholinergic nerve activity could be of therapeutic value in regulation of gut motility (see Taniyama et al., 1996), compounds acting as partial agonists/antagonists of the GABA_B receptor controlling intestinal cholinergic motor events could be of interest in treating gut motility disorders.

In human cerebral cortex, SRIF- or CCK-releasing nerve terminals are equipped with GABA_B receptors at which the compound CGP 47656 behaved as a full agonist and the antagonist CGP 52432 exhibited higher potency with respect to CGP 35348 (Bonanno et al., 1996; Raiteri et al., 1996; see Table 1). Interestingly, the compound CGP 36742 was recently found to distinguish between these two receptors in human cerebral cortex, as it does in the rat, being an antagonist at the receptor on SRIF-releasing nerve terminals but inactive (up to 100 μM) at the receptor regulating CCK release (Bonanno et al., 1999; see Table 1). Thus, the pharmacology of the GABA_B receptor inhibiting cholinergic twitch contraction in guinea pig ileum resembles that of the receptor regulating SRIF release in human cerebral cortex; however, further studies with even more selective and additional ligands would be required before a firm conclusion can be established. The lower efficacy of CGP 47656 at the guinea pig ileum receptor could simply depend on a lower efficacy of receptor coupling in the ileum. Similarly, 3-aminopropylphosphonic acid and (R)-(−)-β-phenyl-GABA were full agonists at rat CNS GABA_B receptors but partial agonists at the GABA_B receptors in guinea pig ileum myenteric plexus (Luzzi et al., 1986; Kerr et al., 1990b; see also Kerr et al., 1990a; Ong et al., 1993).

CGP 47656 is a selective GABA_B receptor ligand (Froestl et al., 1995), exhibiting a very peculiar behavior. The compound is a full agonist at the GABA_B receptors regulating SRIF or CCK release in human neocortex (Table 1) but a potent antagonist at the GABA_B autoreceptor regulating GABA release in the same human tissue (Raiteri et al., 1996). In the guinea pig ileum, CGP 47656 appears to behave as a partial agonist (see Fig. 1). The drug was able to modify the concentration-response curve of (−)-baclofen in the MP-LM preparation in a way predicted for an interaction between a full agonist and a partial agonist at the same receptor. That CGP 47656 is a partial agonist at the GABA_B receptor acted on by (−)-baclofen is confirmed by the sensitivity of the CGP 47656 effect to various selective GABA_B receptor antagonists.

The compound CGP 56999 was reported to be a potent antagonist at the GABA_B receptors inhibiting GABA and substance P release in the rat spinal cord, whereas it was inactive at the receptors inhibiting glutamate release (Teoh et al., 1996). Unexpectedly, we here found that CGP 56999 behaves as a potent noncompetitive antagonist in the guinea pig ileum, which is in line with the existence of a rich heterogeneity of GABA_B receptors.

Although abundant evidence supports the existence of multiple pharmacological subtypes of the GABA_B receptor, structural information on this receptor has been lacking for a long time. A protein, termed GABA_BR1, was cloned recently (Kaupmann et al., 1997). The protein binds GABA_B receptor antagonists with high affinity, but it does not exhibit the expected functional properties for a GABA_B receptor because it cannot reach the cell membrane on its own (Couve et al., 1998). Another GABA_B-related protein was then cloned and termed GABA_BR2 (Jones et al., 1998; Kaupmann et al., 1998; White et al., 1998; Kuner et al., 1999). The latter protein cannot be a receptor either, because it cannot bind GABA_B antagonists. However, its coexpression with GABA_BR1 produces a heterodimer that reasonably mimics native GABA_B receptors. Clearly, these molecular biology data are far from an explanation of pharmacological heterogeneity. Moreover, both GABA_BR1 and GABA_BR2 appear to be exclusively expressed in the CNS (Kaupmann et al., 1997, 1998; Kuner et al., 1999). If this were the case, new heterodimers should be found in peripheral systems where GABA_B receptor-mediated functions have been described by several authors (for reviews see Bowery, 1993; Chapman et al., 1993; Taniyama et al., 1996).

The GABA_B receptor present in the guinea pig ileum characterized in the present work is pharmacologically distinct from any of the several subtypes previously identified in the rat CNS (see Bonanno et al., 1999) and displays similarity to the GABA_B receptor regulating SRIF release in human neocortex.