Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

The relatively large number of agonists, antagonists and modulators that are available with high affinity and remarkable selectivity for ionotropic glutamate receptors have significantly contributed to the advancement of our knowledge on the physiology and physiopathology of glutamate-mediated neurotransmission (Collingridge and Lester, 1989; Lodge and Collingridge, 1990; Meldrum et al., 1991; Watkins et al., 1990). Glutamate interacts not only with ionotropic receptors but also with G protein-linked receptors, or mGluRs (Nicoletti et al., 1986; Sladeczek et al., 1985). Unfortunately, the availability of agonists and antagonists with high affinity and selectivity for mGluRs is very limited (Roberts, 1995; Schoepp, 1994).

Eight different subtypes of mGluRs have been cloned so far, and they have been subdivided into three groups on the basis of their sequence homology, effector coupling and pharmacology (see the reviews by Knopfel et al., 1995; Nakanishi, 1992; Pin and Duvoisin, 1995; Schoepp and Conn, 1993). The first group comprises mGluR1 and mGluR5, coupled to the stimulation of PLC; the second group comprises mGluR2 and mGluR3, negatively linked to the adenylate cyclase cascade. The third group (mGluR4, mGluR6, mGluR7 and mGluR8) is also negatively linked to adenylate cyclase but can be distinguished from the second group because it can be selectively stimulated by L-(+)-2-amino-4-phosphonobutyric acid. Splice variants have been found for three mGluRs: mGluR1 (mGluR1a, mGluR1b, mGluR1c and mGluR1e), mGluR4 (mGluR4a and mGluR4b) and mGluR5 (mGluR5a and mGluR5b), mostly arising through alternative splicing of the carboxyl-terminal sequence (for details, see Pin and Duvoisin, 1995). To understand the role of mGluRs in brain function and pathology, gene targeting techniques have recently been introduced, such as the development by two independent groups of transgenic mice lacking mGluR1 (Aiba et al., 1994a, 1994b; Conquet et al., 1994). Despite the potential usefulness of these type of models, conflicting data have been obtained in these laboratories as a possible result of compensatory events that may occur in mGluR knock-out animals during development. Therefore, the availability of selective agonists and antagonists still needs to be viewed as an important strategy, not only for possible therapeutic applications but also to understand the functional role of mGluRs.

(Carboxyphenyl)glycines represent the most interesting class of mGluR antagonists characterized so far, but unfortunately none of them are selective. In particular, (+)-MCPG is able to antagonize both mGluR1 and mGluR2, (S)-4C-PG is an antagonist of mGluR1 but an agonist of mGluR2 and (S)-4-carboxy-3-hydroxyphenylglycine is an antagonist of mGluR1 and an agonist of mGluR2 and mGluR5 (Brabet et al., 1995; Cavanni et al., 1994; Hayashi et al., 1994; Kingston et al., 1995; Roberts, 1995; Thomsen et al., 1994a; Watkins and Collingridge, 1994). In the search for more selective agents, we recently reported that AIDA, a 1-aminoindanedicarboxylate in which the (carboxyphenyl)glycine moiety is inserted in a constrained framework, is an antagonist of mGluRs of the first group (Lombardi et al., 1995; Pellicciari et al., 1995). In the present work, we characterize in further detail the pharmacological profile of AIDA by using BHK cells stably transfected with specific mGluRs (mGluR1a, mGluR2, mGluR4a or mGluR5a) or rat brain slices bearing native mGluRs and then evaluating the effects of the drug on the activities of PLC, adenylate cyclase or PLD. Because we have previously shown that stimulation of mGluR1 increases transmitter release, whereas mGluR2 agonists have opposite effects (Lombardi et al., 1993, 1996), we also investigated whether AIDA was able to modify the effects of mGluR agonists on the depolarization-induced output of preloaded D-[³H]aspartate from rat cortical slices. Finally, because mGluR1 knock-out mice exhibit motor deficits and impaired motor learning (Aiba et al., 1994b; Conquet et al., 1994), we administered AIDA i.c.v. to mice and evaluated their gross behavior in an open field, locomotor activity and pain sensitivity.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. (1S,3R)-ACPD, (S)-4C-PG and (+)-MCPG, were purchased from Tocris Cookson (Bristol, UK). AIDA (originally termed UPF-523) was synthesized for most of the present experiments as previously described (Pellicciari et al., 1995). In some experiments, however, the compound was purchased from Tocris Cookson. D-[³H]Aspartate (10-30 Ci/mmol) and the cAMP radioimmunoassay kit (25 Ci/mmol) were from Amersham (Milan, Italy), myo-[2-³H(N)]inositol (10-20 Ci/mmol) and [1,2,3-³H]glycerol (30-60 Ci/mmol) were from DuPont-NEN (Milan, Italy). Forskolin, 3-isobutyl-1-methylxanthine and Dowex AG-1-X8 anion exchange resin (100-200 mesh) were from Sigma Chimica (Milan, Italy). All other reagents were of analytical grade and obtained from Merck (Darmstadt, Germany).

Transfected cell cultures. BHK cells stably transfected with mGluR1a, mGluR2, mGluR4a or mGluR5a were obtained as previously described (Thomsen et al., 1992, 1993) and cultured in Dulbecco's modified Eagle's medium supplemented with 5% dialyzed fetal calf serum, 2 mM glutamine, 0.05 mg/ml gentamycin and 0.1 mg/ml neomycin in a humidified atmosphere (95% air/5% CO₂) at 37°C. In addition, the incubation medium of transfected cells was supplemented with G-418 and metotrexate.

Preparation of brain slices. The preparation of rat brain slices for both neurochemical and release studies was performed as previously described (Lombardi et al., 1993, 1996). Briefly: rat brain regions were rapidly dissected and placed into ice-cold Krebs-bicarbonate buffer (122 mM NaCl, 3.1 mM KCl, 1.2 mM MgSO₄, 0.4 mM KH₂PO₄, 1.3 mM CaCl₂, 25 mM NaHCO₃ and 10 mM glucose) gassed with 95% O₂/5% CO₂. Slices (350 µm thick) were then prepared by use of a McIlwain tissue chopper and placed in gassed Krebs-bicarbonate solution for 1 hr at 37°C to allow functional recovery.

Measurements of second messengers. The assay conditions for measurements of PLC-catalyzed [³H]IP formation in transfected cells expressing mGluRs (Thomsen et al., 1993, 1994a) or in brain slices (Lombardi et al., 1993, 1994; Pellegrini-Giampietro et al., 1988) were previously reported. Adenylate cyclase activity was determined using a cAMP radioimmunoassay kit in transfected cells or brain slices as described (Lombardi et al., 1993; Thomsen et al., 1992, 1993). The mGluR agonist-induced accumulation of labeled PEt in hippocampal brain slices preloaded with [³H]glycerol was used as a measure of PLD activity. Incubation of mGluR agonists and antagonists in the presence of 170 mM ethanol, organic extraction and thin layer chromatography separation of [³H]PEt were performed as recently described in detail (Pellegrini-Giampietro et al., 1996).

D-[³H]Aspartate release. Slices were incubated for 45 min at 37°C in oxygenated Krebs solution containing D-[³H]aspartate (final concentration, 50 nM) and subsequently rinsed for 30 min, with the incubation medium changed twice. Labeled slices were then transferred (two per chamber) to perfusion chambers (0.3 ml volume) and superfused with oxygenated Krebs' solution at 32°C. After 30 min of perfusion, the slices were challenged with a depolarizing solution containing 30 mM KCl (with isomolar reduction of NaCl); mGluR agonists and antagonists were added 5 min before depolarization. Perfusates were collected every 5 min and analyzed for their content of radioactivity (for details, see Beani et al., 1978; Lombardi et al., 1993, 1996).

Intracerebroventricular administration of AIDA and behavioral tests in mice. Injections of AIDA into the right lateral ventricle of male Swiss albino mice were performed in a volume of 5 µl under light ether anesthesia according to Haley and McCorney (1957). The hot plate test was performed by using a stainless steel container thermostatically set at 52°C (O'Callaghan and Holtzman, 1976). The abdominal constriction test (or writhing test) was performed by injecting 0.6% acetic acid i.p. and counting the number of stretching movements for 10 min (Koster et al., 1959). Spontaneous locomotor activity was investigated 5 min after the i.c.v. injection by placing two mice in a clear plastic box over an LKB Animex activity meter for 20 min (Bacciottini et al., 1987).

Statistical analysis. Statistical evaluation of results was performed using analysis of variance and Student's t tests. pA₂ calculations were performed from concentration-response curves using a computer program as described by Tallarida and Murray (1984).

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Effects of AIDA on phosphoinositide hydrolysis in mammalian cells transfected with mGluR1a or mGluR5a. AIDA reduced in a concentration-dependent manner the stimulation of phosphoinositide hydrolysis induced by L-glutamate in transfected BHK cells expressing either mGluR1a or mGluR5a (fig. 1). The IC₅₀ value for this effect was 214 µM (95% confidence limits, 148-280) in mGluR1a cells and >1 mM in mGluR5a cells. Figure 1 shows that at 1 mM, AIDA antagonized the effects of 10 µM glutamate acting on mGluR1a by 90% but those of 5 µM glutamate on mGluR5a by only 32%. Glutamate was used at a higher concentration in mGluR1a cells because the cells are known to exhibit lower sensitivity to agonists (Brabet et al., 1995, present work). In transfected cells expressing mGluR1a, the effects of AIDA against test concentrations (Brabet et al., 1995; Thomsen et al., 1993) of glutamate (10 µM) or (1S,3R)-ACPD (300 µM) gave a similar concentration-response curve (fig. 2). However, when [³H]IP formation was induced by a relatively large concentration of quisqualate (3 µM) in these cells, AIDA (1 mM) reduced PLC activity by only 50%.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   AIDA reduces in a concentration-dependent manner the formation of [3H]IP induced by l-glutamate in transfected BHK cells expressing mGluR1a or mGluR5a. AIDA was applied 5 min before stimulating phosphoinositide hydrolysis with a submaximal concentration of L-glutamate (10 µM in mGluR1a and 5 µM in mGluR5a cells). The results are expressed as percentage of L-glutamate response minus basal levels of [3H]IP formation. In cells expressing mGluR1a, basal and L-glutamate-stimulated [3H]IP formation were 3800 ± 300 and 22,400 ± 2900 dpm/mg of protein, respectively; in mGluR5a cells, the corresponding values were 4100 ± 400 and 19,200 ± 1200 dpm/mg of protein. Values are mean ± S.E.M. of at least three experiments conducted in triplicate.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   AIDA reduces in a concentration-dependent manner the formation of [3H]IP induced by 10 µM L-glutamate, 300 µM (1S,3R)-ACPD and 3 µM quisqualate in transfected BHK cells expressing mGluR1a. AIDA was applied 5 min before stimulating phosphoinositide hydrolysis with the different agonists. The results are expressed as percentage of agonist response minus basal levels of [3H]IP formation. L-Glutamate raised basal levels of [3H]IP formation from 4000 ± 350 to 23,000 ± 1300, of (1S,3R)-ACPD to 19,000 ± 2100 and of quisqualate to 36,500 ± 3700 dpm/mg of protein. Values are mean ± S.E.M. of at least three experiments conducted in triplicate.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effects of increasing concentrations of AIDA on the (1S,3R)-ACPD concentration-response curve for the formation of [3H]IP in transfected BHK cells expressing mGluR1a and mGluR5a. AIDA (at 100, 300 and 500 µM) shifts to the right the (1S,3R)-ACPD curve in mGluR1a but not in mGluR5a cells. In each experiment, the response to 300 µM (1S,3R)-ACPD was considered as 100% and all other values were calculated accordingly. Larger concentrations of (1S,3R)-ACPD (500 µM) did not increase further the formation of [3H]IP. Values are mean ± S.E.M. of at least five experiments conducted in triplicate.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Preexposure to 300 µM AIDA shifts to the left the (1S,3R)-ACPD concentration-response curve for the formation of [3H]IP in transfected BHK cells expressing mGluR1a. Cells were exposed to AIDA for 48 hr and then thoroughly washed before applying (1S,3R)-ACPD. In each experiment, the response to 300 µM (1S,3R)-ACPD in preexposed cells (which were ~30% larger than those of nonpreexposed cells) was considered as 100%, whereas the formation of [3H]IP in control cells was considered as the basal level. All other values were calculated accordingly. Values are mean ± S.E.M. of at least five experiments conducted in triplicate. *P < .05 vs. same concentration in control.

Effects of AIDA on forskolin-activated cAMP formation in mammalian cells transfected with mGluR2 or mGluR4a. Glutamate and (1S,3R)-ACPD inhibit 10 µM forskolin-stimulated cAMP formation in BHK (fig. 5) and in other types of transfected cells expressing either mGluR2 or mGluR4a (Hayashi et al., 1994; Tanabe et al., 1993; Thomsen et al., 1994a). Similarly, the (carboxyphenyl)glycine (S)-4C-PG (10-500 µM) reduced forskolin-activated adenylate cyclase activity in mGluR2-expressing cells in a manner comparable to (1S,3R)-ACPD (fig. 5) (see also Hayashi et al., 1994; Watkins and Collingridge, 1994). The maximal degree of inhibition was ~60% and was achieved at 300 µM. AIDA (1 mM) did not modify the inhibitory action of a concentration of (1S,3R)-ACPD (300 µM) that gave a quasimaximal response in cells expressing mGluR2 (fig. 5) or in cells expressing mGluR4a. However, a modest agonist activity on mGluR2 was observed when 1 mM AIDA was used (fig 5). Because an exchange between AIDA and intracellular glutamate could account for these results (Thomsen et al., 1994b), experiments were performed to rule out this possibility. The negative results obtained indicate that AIDA (1 mM) does not interact with glutamate carriers in BHK cells (data not shown).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effects of (1S,3R)-ACPD, AIDA and (S)-4C-PG on forskolin-activated cAMP formation in transfected BHK cells expressing mGluR2. Values are expressed as percentage of the formation of cAMP induced by 10 µM forskolin (basal levels of cAMP were 4.1 ± 0.6 pmol/mg of protein; forskolin increased these levels by ~10-fold). Values are mean ± S.E.M. of at least three experiments conducted in triplicate. ***P < .001 vs. forskolin alone.

Effects of AIDA on PLC, adenylate cyclase or PLD activity in rat brain slices. Incubation of hippocampal slices with (1S,3R)-ACPD (30-300 µM) stimulated PLC activity in a concentration-dependent manner; at 100 µM, (1S,3R)-ACPD increased the formation of [³H]IPs by ~6-fold over basal values. Figure 6 shows that the latter effect was antagonized by AIDA (1-1000 µM) in a concentration-dependent manner. The particular shape of the inhibitory curve may be ascribable to the fact that at 1 mM AIDA interacts with at least two mGluR subtypes coupled to phosphoinositide hydrolysis (i.e., mGluR1 and mGluR5; see fig. 1). Other mGluR antagonists such as (S)-4C-PG and (+)-MCPG, which do not appear to interact with mGluR5 (Brabet et al., 1995), at 1 mM reduced the effect of 100 µM (1S,3R)-ACPD on [³H]IP formation by only 20% to 35% (fig. 6).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   AIDA, (S)-4C-PG and (+)-MCPG reduce, in a concentration-dependent manner, the formation of [3H]IP induced by (1S,3R)-ACPD in rat hippocampal slices. Slices prelabeled with [3H]inositol were exposed to 100 µM (1S,3R)-ACPD in the presence of increasing concentrations of the antagonists for 15 min in the presence of 10 mM LiCl. The results are expressed as percentage of (1S,3R)-ACPD response minus basal levels of [3H]IP formation. Basal radioactivity found in the fraction corresponding to inositol phosphates was 15,900 ± 2100 dpm/mg of protein; the exposure to 100 µM (1S,3R)-ACPD increased these levels by ~6-fold. Values are mean ± S.E.M. of at least five experiments conducted in triplicate.

View larger version (63K): [in this window] [in a new window]   Fig. 7.   Effects of (1S,3R)-ACPD, AIDA and (+)-MCPG on the PLD-catalyzed formation of [3H]PEt in rat hippocampal slices. Slices were labeled with [3H]glycerol and then incubated with 170 mM ethanol in the presence of the indicated concentrations of mGluR agents. When assayed together with (1S,3R)-ACPD, AIDA or (+)-MCPG were added 5 min before the agonist. PLD activity is expressed as the percentage of incorporation of label into [3H]PEt under (1S,3R)-ACPD-free (basal) conditions. Each column represents the mean ± S.E.M. of at least five experiments conducted in triplicate. aP < .01 vs. basal. bP < .01 vs. (1S,3R)-ACPD alone.

Effects of AIDA on (1S,3R)-ACPD modulation of D-[³H]aspartate output from rat cortical slices. (1S,3R)-ACPD (10-300 µM) significantly potentiates the depolarization-induced release of neurotransmitter in rat cerebro-cortical synaptosomes (Herrero et al., 1992) and cortical slices (Lombardi et al., 1994, 1996), as well as in the rat parietal cortex in vivo (Cozzi et al., 1996), with a pharmacological profile suggesting the involvement of group 1 mGluRs. In other brain areas such as the striatum, a similar concentration of (1S,3R)-ACPD inhibits the depolarization-induced output of D-[³H]aspartate, probably because the agonist activates mGluRs of the second group in this area (Lombardi et al., 1993, 1994). When AIDA was added at 30 to 100 µM to the buffer solution perfusing cortical slices, the potentiating effects of 100 µM (1S,3R)-ACPD on KCl-induced D-[³H]aspartate output were reduced (fig. 8). A larger concentration of AIDA (300 µM) perfused together with (1S,3R)-ACPD not only antagonized its potentiating effect but also reversed it into an inhibition of KCl-induced D-[³H]aspartate outflow. A similar phenomenon was obtained with (S)-4C-PG, which prevented the potentiation of (1S,3R)-ACPD at 30 µM and caused a reduction of KCl-stimulated D-[³H]aspartate output at 100 to 300 µM. On the contrary, at the higher concentration tested (500 µM), (+)-MCPG was able to completely inhibit the potentiation of release induced by (1S,3R)-ACPD without causing a reduction in the depolarization-induced output of D-[³H]aspartate (fig. 8).

View larger version (17K): [in this window] [in a new window]   Fig. 8.   Effects of AIDA, (S)-4C-PG and (+)-MCPG on (1S,3R)-ACPD modulation of D-[3H]aspartate output from rat cortical slices. (1S,3R)-ACPD alone or in the presence of increasing concentrations of antagonists were added 5 min before and maintained during the depolarizing (30 mM KCl) challenge. Left, 30 mM KCl increased basal D-[3H]aspartate output (5200 ± 500 dpm/mg of protein, evaluated as the average of the two 5 min perfusion samples before depolarization) by 2.8 ± 0.1-fold, whereas KCl plus 100 µM (1S,3R)-ACPD increased basal levels by 4.05 ± 0.2-fold. Right, results are expressed as percentage of the D-[3H]aspartate output response to 30 mM KCl plus 100 µM (1S,3R)-ACPD. (+)-MCPG merely antagonized the stimulatory effect of (1S,3R)-ACPD, whereas AIDA and (S)-4C-PG reversed it into an inhibition of KCl-induced transmitter output. Values are mean ± S.E.M. of at least five experiments conducted in triplicate.

Behavioral effects of i.c.v. injections of AIDA in mice. Mice injected i.c.v. with 5 µl of saline containing AIDA (0.01-10 nmol) and then placed in an open field exhibited some difficulty in the initiation of movement. However, mild stimuli, such as a light pinch on the tail, started an apparently normal ambulatorial behavior that could not be distinguished from that of saline-injected controls. No obvious motor coordination deficit or ataxia was present. This apparent difficulty in initiating normal exploratory behavior lasted ~30 min. In other groups of animals, we then investigated the effects of AIDA (0.01-10 nmol i.c.v.) on the licking latency in the hot plate test and on the number of abdominal constrictions after i.p. injection of 0.6% acetic acid. The results are reported in figure 9 and indicate that i.c.v. administration of AIDA caused mild analgesia, which at its maximum was comparable to that obtained with the systemic administration of 5 mg/kg morphine (O'Callaghan and Holtzman, 1976). Unexpectedly, the analgesic effect was no longer present when larger doses (0.1-10 µmol) of AIDA were used (data not shown).

View larger version (10K): [in this window] [in a new window]   Fig. 9.   Analgesic effects of AIDA in the hot plate (top) and writhing (bottom) tests in mice. AIDA was injected i.c.v. at the indicated doses; 15 min later, animals were placed in a hot plate maintained at 52°C and observed for initial pain reaction (paw licking). For the writhing test, 0.6% acetic acid (10 ml/kg) was injected i.p., and 5 min later, the number of abdominal muscle contractions were scored for 10 min. Each point is the mean value ± S.E.M. of at least eight mice. *P < .05 vs. saline-injected mice. **P < .01 vs. saline-injected mice.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Taken together, our results obtained in mammalian transfected cells and rat brain slices show that AIDA is a selective antagonist of group 1 mGluRs with preferential activity on mGluR1a over mGluR5a. The antagonist activity of AIDA on mGluR1a was competitive in nature and was better appreciated when L-glutamate or (1S,3R)-ACPD, rather than relatively elevated concentrations of quisqualate, was used as stimulating agents in transfected cells expressing mGluR1a. Similarly, in a recent study on the effects of (carboxyphenyl) glycine derivatives in transfected cells expressing mGluR1a or mGluR5a, it was reported that the apparent potency of mGluR antagonists depends on the agonist used to activate these receptors (Brabet et al., 1995). This phenomenon was particularly evident when weak antagonists were tested in the presence of potent agonists and may account for the diverse IC₅₀ values that have been reported for mGluR antagonists. We previously reported (Pellicciari et al., 1995) that the IC₅₀ value of AIDA inhibiting the L-glutamate response in cells expressing mGluR1a was considerably lower (7 µM) than that found in the present experiments (214 µM). However, the former data were obtained in a situation in which AIDA caused only a partial blockade on mGluR1a-mediated phosphoinositide hydrolysis (Pellicciari et al., 1995). In this study, AIDA prevents almost completely the stimulation of [³H]IP formation induced by (1S,3R)-ACPD, although it appears to be less potent.

In rat hippocampal slices, another model in which the antagonism of the formation of [³H]IP induced by (1S,3R)-ACPD was evaluated, the maximal inhibition (~70%) observed with 1 mM AIDA was significantly higher than that (~30%) obtained with the same concentrations of (+)-MCPG or (S)-4C-PG, two (carboxyphenyl)glycines known to interact with PLC-linked mGluRs (Brabet et al., 1995; Hayashi et al., 1994; Kingston et al., 1995). In addition, the concentration-response curve of AIDA had a biphasic shape, suggesting that in this region, relatively low concentrations of AIDA (0.1 mM) selectively block mGluR1, whereas larger concentrations (0.3-1 mM) also reduce the effect of (1S,3R)-ACPD on mGluR5. This hypothesis is in line with the observation that both mGluR1 and mGluR5 are expressed in rat hippocampus (Abe et al., 1992; Martin et al., 1992; Masu et al., 1991; Shigemoto et al., 1993).

When transfected BHK cells expressing mGluR1a were preexposed to AIDA for 48 hr, [³H]IP formation was significantly increased under basal conditions and after stimulatory concentrations of (1S,3R)-ACPD. This observation suggests that in transfected BHK cells, mGluR1a responses may be regulated by plastic changes of the system and that under basal conditions, receptors may be partially desensitized by the presence of significant concentrations of L-glutamate in the incubation medium. This hypothesis is in line with observations obtained using a Syrian hamster cell line transfected with mGluR1 plus a glutamate transporter to keep the concentrations of the excitatory amino acid low at the receptor level (Desai et al., 1995); when the transporter was coexpressed, [³H]IP formation induced by mGluR1 agonists was notably higher than that seen in its absence. Plastic changes of mGluR1 have also been described in primary cultures of cerebellar granular cells (Catania et al., 1991; Favaron et al., 1992), but the lack of appropriate antagonists has not yet allowed a detailed study of their basic mechanisms. Studies are currently in progress in our laboratory to elucidate whether the increased mGluR1a response in cells preexposed to AIDA could be due to an increased number of receptors, to changes in receptor affinity or to modifications in effector coupling.

In BHK cells transfected with mGluR2, AIDA had no effect on forskolin-induced stimulation of adenylate cyclase activity of 300 µM but had a modest inhibitory activity at large concentrations (500-1000 µM). Similarly, (S)-4C-PG (30-300 µM) displayed agonist activity on mGluR2 by reducing forskolin-induced cAMP formation (see also Hayashi et al., 1994; Watkins and Collingridge, 1994). AIDA did not affect the inhibitory action of (1S,3R)-ACPD on forskolin-induced cAMP formation in rat striatal or hippocampal slices, but in line with the observations in transfected cells, it displayed agonist activity when large concentrations were used. In these preparation, (+)-MCPG antagonized the effects of (1S,3R)-ACPD, confirming its activity as an antagonist of mGluRs of the second group. In addition, AIDA was tested for its possible interaction with mGluRs coupled to PLD, which have been recently described in brain slices (Boss and Conn, 1992; Holler et al., 1993), and appear to have a pharmacological profile that is distinct from that of any known mGluR subtype linked to PLC or adenylate cyclase (PellegriniGiampietro et al., 1996). (+)-MCPG acts as an agonist/antagonist on PLD-coupled mGluRs, whereas AIDA (1 mM) neither stimulates PLD activity nor modifies the effect of the agonist (1S,3R)-ACPD.

A functional consequence after activation of PLC-coupled mGluRs is the potentiation of depolarization-induced release of transmitter in cortical preparations (Coffey et al., 1994; Lombardi et al., 1994, 1996). On the contrary, activation of mGluRs negatively linked to adenylate cyclase inhibits release in other brain preparations and, in particular, in striatal slices (Anwyl, 1991; Lombardi et al., 1993, 1994; Lovinger, 1991). (+)-MCPG prevents both the stimulatory and the inhibitory effects of (1S,3R)-ACPD on transmitter release (Lombardi et al., 1993), most likely because it is an antagonist acting on mGluRs of both the first and second group. The results reported in this work for other mGluR antagonists on (1S,3R)-ACPD modulation of D-[³H]aspartate output from rat cortical slices are intriguing. When (S)-4C-PG was studied, we observed that increasing concentrations reverted the antagonism of the potentiating effect of (1S,3R)-ACPD into a potent inhibition of KCl-induced output. It is reasonable to propose that the antagonism is mediated by blockade of mGluR1, whereas the inhibition is due to the simultaneous agonist activity of (S)-4C-PG and (1S,3R)-ACPD on mGluRs of the second group negatively linked to cAMP formation, such as mGluR2 (see fig. 5). AIDA also antagonized the positive effect of (1S,3R)-ACPD at lower doses and reverted it into an inhibition of KCl-induced output, although only at the higher concentration tested (1 mM). In this case, because the compound is inactive on mGluR2, 1 mM AIDA could be blocking mGluR1 and mGluR5, allowing the interaction between (1S,3R)-ACPD and mGluRs of the second group. The use of different antagonists at specific concentrations (e.g., 100 µM in fig. 8) may thus reveal the opposite effects of (1S,3R)-ACPD on transmitter release. These data could perhaps be helpful for the interpretation of otherwise contradictory results regarding the effects of mGluR agonists and antagonists on mechanisms leading to long-term potentiation or other forms of synaptic plasticity (Bashir et al., 1993; Brown et al., 1994; Chinestra et al., 1994; O'Connor et al., 1994).

Finally, we injected AIDA i.c.v. in mice and evaluated their behavior. When treated animals were placed in an open field, we expected to observe the symptomatology described in detail for mice lacking mGluR1: wide base standing position, tremor, ataxia and loss of the righting reflex (Aiba et al., 1994b; Conquet et al., 1994). No dose of AIDA, however, elicited such effects, the righting reflex was always present, and the ambulatorial behavior of treated animals was apparently identical to that of control mice. A careful observation of i.c.v. injected mice, however, revealed that they had a tendency to stay immobile in the center of the open field, possibly because they had difficulties in the initiation of movements. A mild stimulation would interrupt this immobility, and the animals would start to explore the new environment in a manner comparable to those injected with saline. It is possible that AIDA injected i.c.v. might not reach the cerebellar cortex in sufficient concentration; this could explain the discrepancies between our study and those using mGluR1 knock-out animals. However, it is also possible that mice lacking mGluR1 display motor deficit in adulthood because the receptor is required to learn specific motor skills during development. Obviously, animals treated with receptor antagonists in adult life will not exhibit motor impairment because they have mastered motor skills at a time when mGluR1 was not blocked.

It has been shown that antagonists of mGluRs of the first group reduce electrophysiological responses evoked by nociceptive stimuli in the spinal cord and in the thalamus (Salt and Eaton, 1994; Young et al., 1994). We thus tested the effects of AIDA injected i.c.v. in two models widely used to uncover drug actions on the nociceptive reflex: the hot plate test and the writhing test. AIDA delayed the pain reaction (paw licking) of mice placed on a hot plate at 52°C and significantly reduced the number of writhes observed after i.p. administration of diluted acetic acid. The analgesic effect of AIDA was comparable to that of 5 mg/kg s.c. morphine. However, it should be noted that AIDA-mediated antinociception had a bell-shaped dose-response curve because it was no longer present when large doses (0.1-10 µmol) of the compound were administered. This could suggest that mGluR1 and mGluR5 have opposite effects on the antinociceptive reflex, although the present experiments do not rule out alternative hypotheses. More potent and selective mGluR antagonists are required to permit a satisfactory explanation for these observations.

In conclusion, AIDA is an antagonist of mGluRs coupled to PLC activity with potency for mGluR1a superior to that for mGluR5a. When used in transfected cells or in brain slices, it does not affect mGluRs of the second or third group at a concentration of 1 mM. Thus, AIDA appears to be one of the most potent and selective mGluR1 antagonist described so far. In addition, AIDA antagonizes the potentiation of (1S,3R)-ACPD on the depolarization-induced output of transmitter from cortical slices, and when injected i.c.v., it has mild analgesic effects.