We have examined the effects of a novel GABA(B) agonist, CGP35024, in models of chronic neuropathic (partial sciatic ligation) and inflammatory (Freund's complete adjuvant) pain in the rat, and its inhibitory action on spinal transmission in vitro. The effects of CGP35024 were compared with L-baclofen and gabapentin. CGP35024 and L-baclofen reversed neuropathic mechanical hyperalgesia following single subcutaneous or intrathecal administration, but did not affect inflammatory mechanical hyperalgesia. Gabapentin only moderately affected neuropathic hyperalgesia following a single administration by either route, but produced significant reversal following daily administration for 5 days. It was only weakly active against inflammatory hyperalgesia following single or repeated administration. The antihyperalgesic effects of L-baclofen and CGP35024, but not gabapentin, were blocked by the selective GABA(B) receptor antagonist CGP56433A. CGP35024 was seven times more potent against neuropathic hyperalgesia than in the rotarod test for motor co-ordination, whilst L-baclofen was approximately equipotent in the two tests. In the isolated hemisected spinal cord from the rat, CGP35024, L-baclofen and gabapentin all inhibited capsaicin-evoked ventral root potentials (VRPs). CGP35024 and L-baclofen, but not gabapentin, also inhibited the polysynaptic and monosynaptic phases of electrically-evoked VRPs, as well as the 'wind-up' response to repetitive stimulation. These data indicate that CGP35024 and L-baclofen modulate nociceptive transmission in the spinal cord to inhibit neuropathic hyperalgesia, and that CGP35024 has a therapeutic window for antihyperalgesia over spasmolysis.

1 The location of the B, bradykinin receptors involved in inflammatory hyperalgesia was investigated.2 No specific binding of the B1 bradykinin receptor ligand [3H]-des-Arg10-kallidin was detected in primary cultures of rat dorsal root ganglion neurones, even after treatment with interleukin-l1 (100 iu ml-'). 3 In dorsal root ganglion neurones, activation of B2 bradykinin receptors stimulated polyphosphoinositidase C. In contrast, B1 bradykinin receptor agonists (des-Arg9-bradykinin up to 10 pM and des-Arg'0-kallidin up to 1 /M) failed to activate polyphosphoinositidase C, even in neurones that had been treated with interleukin-lp (100 iu ml-), prostaglandin E2 (1 gM) or prostaglandin I2 (1 MM).

1 The modulatory eects of mGlu receptors on NMDA-induced potential changes in spinal motoneurones were studied in vitro. 2 Selective activation of mGlu5 receptors by 10 mM (RS)-2-Chloro-5-hydroxyphenylglycine (CHPG; EC 50 =280+24 mM) did not produce any change in the ventral root potential. However, the same concentration of CHPG (10 min perfusion) signi®cantly attenuated the NMDA-induced ventral root depolarization (VRD). The eect persisted for 10 min after washout. NMDA-induced responses returned to control in 30 min. Brief co-application of CHPG and NMDA did not alter the NMDA-induced response indicating lack of direct receptor interaction. 3 The attenuating eect of CHPG on the NMDA-induced VRD was inhibited by the mGluR5 receptor antagonist, 2-methyl-6-phenyl-ethynylpyridine (MPEP). 4 In the presence of CGP56433A, a GABA B receptor antagonist, the NMDA-induced VRD was unchanged. However, NMDA-induced responses were potentiated after 10 min co-application of CHPG and CGP56433A. 5 (2R,4R)-4-aminopyrrolidine-2,4-dicarboxylate ((2R,4R)-APDC), a group II mGlu receptor agonist did not attenuate the NMDA-induced response. 6 Under normal physiological conditions group I mGlu receptor agonists activate at least two populations of neurones: (1) GABA-ergic cells, which could release GABA and inhibit dorsal horn neurones, and (2) deep dorsal horn neurones/motoneurones which express NMDA receptors. Therefore, activation of mGlu5 receptors located on GABA-ergic interneurones could in¯uence any direct potentiating interaction between mGlu5 and NMDA receptors in spinal cord and result in depression of the VRD. In the presence of a GABA B receptor antagonist, the direct synergistic interaction is unmasked. These data suggest that group I mGlu receptors provide a complex modulation of spinal synaptic processes.

No abstract