Key words: Antagonists, GABA^A: bicuculline. Antagonists, glycine: strychnine. Neurotransmitters: glutamate. Pain: allodynia. Pharmacology: nitric oxide. Receptors: glutamate.

GLYCINE and gamma-aminobutyric acid (GABA) are both inhibitory neurotransmitters that mediate fast synaptic inhibition in the nervous system. [1] Their actions are to bind specifically to glycine and GABA sub A receptors, respectively. This is followed within milliseconds by the gating or opening of an integral chloride ion channel, which results, in general, in the hyperpolarization of the recipient neuronal cell. [2] Pharmacologically, glycine receptors are defined by the antagonism by the convulsive alkaloid strychnine (strychnine-sensitive glycine receptor antagonist), in contrast to the strychnine-insensitive glycine-binding site that is associated with N-methyl-D-aspartate (NMDA) subclass of glutamate receptors. On the other hand, the convulsant alkaloid bicuculline blocks the hyperpolarizing actions of GABA and nerve stimulation. Molecular models show that GABA is isosteric with a GABA-like moiety in the bicuculline molecule, suggesting a competitive interaction on the GABA^Areceptor.

Previous studies demonstrated that glycine and GABA are important in sensory processing in the spinal dorsal horn as inhibitory neurotransmitters. [3-6] Intrathecal administration of strychnine [7-10] or bicuculline [7,9] to conscious mice was reported to induce allodynia, a state of discomfort and pain evoked by innocuous stimuli; the mice showed squeaking, biting, and escaping in response to low-threshold stimuli. A growing body of evidence suggests that the pharmacology of the system activated in the pathologic state "allodynia" may differ from that activated under normal circumstances by high-threshold thermal, chemical, and mechanical stimuli. [11] In fact, it was previously reported that intrathecal administration of the opioid and alpha²-receptor agonists could produce a definitive inhibition of the spinal response to noxious stimuli but had little effect on the strychnine-induced allodynia. [7] On the other hand, adenosine analogs showed the powerful effect on strychnine-induced allodynia at doses that have only a mild analgesic effect on hyperalgesia. [8] Transmission at neural synapses is mediated by a variety of receptors that specify neurotransmitter interactions and transmit information into target cells in the spinal cord. [11] In recent years, much attention has been directed toward the excitatory transmission mediated through the glutamate receptors in the central nervous system, [12-15] and it has been suggested that there are mechanisms whereby interactions between excitatory and inhibitory neurotransmitter systems can modulate signal transmission in the spinal cord. [16] The glutamate receptors are classified in three groups, NMDA, non-NMDA (AMPA-kainate), and metabotropic receptors. [12] It has been reported that the NMDA receptor is a voltage-gated ion channel that, once activated, allows Calcium²+ to enter the neuron. This increase in intracellular Calcium²+ triggers a cascade of events that include activation of the constitutive form of nitric oxide synthase. [17] Nitric oxide diffuses to its site of action, where it activates soluble guanylate cyclase and increases the intracellular content of cGMP. Although strychnine-induced allodynia was reported to be mediated through the NMDA-type glutamate receptor, [7,10] it is not clear whether the bicuculline-evoked allodynia is mediated through the glutamate receptor system or how different the allodynia induced by strychnine and bicuculline are. The current study was designed to assign the involvement of glutamate receptors and nitric oxide system in strychnine- and bicuculline-evoked allodynia and seek the difference in the mechanisms of action between them by use of antagonists for glutamate receptors and inhibitors of nitric oxide system.

Male ddY mice weighing 20+/-2 g were used in this study. The animals were housed under conditions of a 12-h light-dark cycle and a constant temperature of 22+/-2 degrees C and 60+/- 10% humidity. A 27-G stainless-steel needle (0.35 mm OD) attached to a microsyringe was inserted between the L5 and L6 vertebrae by a slight modification of the method of Hylden and Wilcox. [18] Drugs in vehicle were injected slowly into the subarachnoid space of conscious mice at 22 +/-2 degrees C. It was previously confirmed by use of Coomassie brilliant blue that the injected solution did not extend to the cervical segments. [19] .

Studies on allodynia were carried out according to the method reported previously. [20] Control mice were given physiologic saline (5 micro liter). Drug-treatment groups were injected with 5 micro liter of vehicle containing various doses of test agents. After the intrathecal injection, each mouse was placed in an individual 14 x 10 x 12-cm Plexiglas enclosure with wood chips on the floor and observed. Allodynia was assessed once every 5 min over a 50-min period by light stroking of the flank of the mice with a paintbrush. The allodynic response was ranked as follows: 0 = no response; 1 = mild squeaking with attempts to move away from the stroking probe; and 2 = vigorous squeaking evoked by the stroking probe, biting at the probe, and strong efforts to escape. Each mouse was tested for 50 min following intrathecal injection. To evaluate the effects of various doses of blocking agents on strychnine- and bicuculline-induced allodynia, we assessed the scores at 5 min after intrathecal injection of strychnine for the former and the scores at 10 min after intrathecal injection of bicuculline for the latter.

The animals were used for only one measurement in each experiment. This study was conducted with the approval of the local ethics committee and in concordance with the guidelines of the Ethics Committee of the International Association for the Study of Pain. [21] .

Strychnine (mw 334.4; a strychnine-sensitive glycine receptor antagonist) and N^omega-nitro-L-arginine methyl ester (L-NAME; mw 269.7; a nitric oxide synthase inhibitor) were obtained from Wako Life Science (Osaka, Japan). Bicuculline (mw 367.4; a GABA^Aantagonist), methylene blue (mw 373.9; a guanylate cyclase inhibitor), and N^omega-nitro-D-arginine methyl ester (D-NAME; mw 269.7; a stereoisomer of L-NAME) were obtained from Sigma (St. Louis, MO). D(-)-2-Amino-5-phosphonovaleric acid (D-AP5; mw 197.1; a competitive NMDA receptor antagonist), ketamine hydrochloride (mw 274.2; an NMDA channel blocker), 7-chloro-4-hydroxyquinoline-2-carboxylic acid (7-Cl-KYNA; mw 223.6; a noncompetitive NMDA antagonist acting on the glycine binding site), D-glutamyl-aminomethyl sulphonic acid (GAMS; mw 240.2; a kainate receptor antagonist), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; mw 232.2; an AMPA receptor antagonist), and L(+)-2-amino-3-phosphonopropionic acid (L-AP3; mw 169.1; a putative metabotropic glutamate receptor antagonist) were obtained from Research Biochemicals (Natick, MA). L(+)-2-Amino-4-phosphonobutanoic acid (L-AP4; mw 183.1; a putative metabotropic glutamate receptor antagonist) was obtained from Cambridge Research Biochemicals (Cambridge, UK). CNQX and 7-Cl-KYNA were dissolved in a 1:9 mixture of 7% NaHCO³in sterile saline; all other drugs were dissolved in sterile saline on the day of experiments and kept on ice until used.

The statistical analyses were carried out by analysis of variance. Statistical significance (*0.01 < P < 0.05, **P < 0.01) was further examined with Duncan's test for multiple comparison. IC⁵⁰values were calculated using the computer program of probit test.

Intrathecal administration of strychnine and bicuculline resulted in prominent agitation responses, such as vocalization, biting, and escape from the probe, to tactile stimuli applied to the flank. Brushing of the face or tactile stimulation of the fore paws did not elicit any response, indicating that allodynia appeared limited to the caudal dermatomes of the body.

(Figure 1)(A) presents the time courses of allodynia evoked by strychnine (0.25 micro gram/mouse) and bicuculline (1.25 micro gram/mouse). Strychnine-induced allodynia showed the maximum effect at 5 min after intrathecal injection, gradually decreasing over the 50-min experimental period. On the other hand, the bicuculline-induced allodynia was evoked by the first stimulus at 5 min after intrathecal injection, but the maximum effect was observed at 10 min. The response was long-lasting and did not disappear by 50 min. Both strychnine- and bicuculline-induced allodynia showed the respective patterns of time courses similar to those shown in Figure 1(A), over a wide range of doses from 25 ng to 2.5 micro gram/mouse. When the scores of allodynia obtained for the overall 50 min were cumulated and expressed as a percent of the maximum possible score, both strychnine- and bicuculline-induced allodynia showed a gradually increased pattern (25 ng-2.5 micro gram; Figure 1(B)), and mice displayed convulsions at a dose of 25 micro gram/mouse or more. The intrathecal administration of saline in conscious mice had no effect on allodynia.

The effects of various antagonists for the glutamate receptor family on the allodynia were evaluated by the values obtained 5 min after injection of 0.25 micro gram strychnine or 10 min after injection of 1.25 micro gram bicuculline. The scores of allodynia induced by strychnine at 5 min and bicuculline at 10 min were 83.3% and 75.0% of the maximum possible score, respectively, and were taken as 100%.

We first investigated the involvement of the NMDA receptor in the strychnine- or bicuculline-induced allodynia by using D-AP5, ketamine, and 7-Cl-KYNA. The allodynia evoked by strychnine was dose-dependently blocked by D-AP5, ketamine, and 7-Cl-KYNA with IC⁵⁰values of 389 ng, 147 ng, and 8.03 ng, respectively (Figure 2). On the other hand, the allodynia caused by bicuculline was not blocked by D-AP5, ketamine, or 7-Cl-KYNA (Figure 2).

We investigated the involvement of non-NMDA receptors in allodynia caused by strychnine or bicuculline with GAMS and CNQX. The allodynia caused by strychnine was dose-dependently blocked by GAMS and CNQX with IC sub 50 values of 1.17 micro gram and 8.76 ng, respectively (Figure 3). The allodynia caused by bicuculline was dose-dependently blocked by GAMS with an IC⁵⁰value of 214 ng (Figure 3(A)) but not blocked by CNQX (Figure 3(B)).

We further investigated the effect of L-AP3 and L-AP4 on allodynia caused by strychnine or bicuculline. The allodynia caused by bicuculline was dose-dependently antagonized by L-AP4 with an IC⁵⁰value of 85.6 ng (Figure 4(B)) but was partially blocked by L-AP3 (Figure 4(A)). On the other hand, the allodynia caused by strychnine was not antagonized by L-AP3 or L-AP4 (Figure 4). These results demonstrated that NMDA and non-NMDA receptors in the spinal cord were involved in the strychnine-induced allodynia but that kainate and metabotropic receptors were involved in the bicuculline-induced allodynia.

To examine whether the nitric oxide system is involved in inducing allodynia, we investigated the effects of L-NAME and methylene blue on allodynia caused by strychnine and bicuculline. The allodynia caused by strychnine was dose-dependently blocked by L-NAME and methylene blue with IC⁵⁰values of 68.8 pg and 38.6 ng, respectively (Figure 5). D-NAME, an inactive isomer of L-NAME, did not block the strychnine-induced allodynia (data not shown). On the other hand, the allodynia caused by bicuculline was blocked by methylene blue with an IC⁵⁰value of 120 pg (Figure 5(B)) but not altered by L-NAME (Figure 5(A)). These results demonstrate that the nitric oxide system in the spinal cord is involved in both strychnine- and bicuculline-induced allodynia.

It was previously reported that intrathecal administration of strychnine or bicuculline to conscious mice induced allodynia and that strychnine-induced allodynia was dose-dependently relieved by NMDA antagonists. [7,10] In the current study, we first demonstrated that the glutamate receptor system involves the bicuculline- and the strychnine-induced allodynia. However, whereas the latter was inhibited by NMDA receptor and non-NMDA receptor antagonists, the former was inhibited by the kainate receptor antagonist GAMS and metabotropic receptor antagonists (Figure 2, Figure 3, and Figure 4), suggesting that the interactions of strychnine and bicuculline with the glutamate receptor system are different. This was supported by the difference in the blockade by L-NAME and methylene blue of allodynia evoked by strychnine and bicuculline (Figure 5). One of the mechanisms for touch-evoked allodynia was believed to result from removal of tonic or evoked inhibition from pathways relaying information about innocuous tactile stimuli. Yaksh [7] suggested that the blockade of inhibition by spinal strychnine and bicuculline must either be presynaptic on the large primary afferent or postsynaptic on the second-order neuron and activated only by the large afferent input. Glycine binding is found throughout the spinal gray, with that in the dorsal horn being largely found in laminae II, III, and the lateral aspect of V. [22,23] The glycinergic neurons in the laminae II and III receive a major monosynaptic input from myelinated low-threshold cutaneous primary afferents, and glycine is considered to act as a postsynaptic inhibitory transmitter. [5,11] On the other hand, GABAergic neurons are present in laminae I-III of the rat spinal cord, and many of neurons with somata in laminae I-III are inhibitory interneurons containing GABA. [24] These GABA-containing terminals frequently are present in the presynaptic axons at axoaxonic synapses and in presynaptic dendrites in the dorsal horn. [1,6,25] Presynaptic inhibition depends on depolarization of excitatory axon terminals by a transmitter released from other axon terminals that form axoaxonal synapses with the excitatory terminals of the primary afferent neuron. Although GABA produces postsynaptic inhibition by hyperpolarizing the postsynaptic cell, it can act as a depolarizing transmitter on the presynaptic terminals of certain primary afferent neurons to produce presynaptic inhibition. [1] L-Glutamate is a known neurotransmitter of primary afferents and descending projections from the brain and perhaps neurotransmitters of some intrinsic spinal neurons. [14,16,26] Among the glutamate receptor family, the NMDA receptor and AMPA receptor were reported to be located mainly at postsynaptic site in the spinal cord, but the kainate receptor was likely to be located at presynaptic site in the spinal cord. [14] On the basis of the selective depressant action of L-AP4 in spinal and certain hippocampal pathways, the metabotropic receptor was proposed to be located at a presynaptic site and possibly function as autoreceptors, controlling the release of neurotransmitters. [12,13] Nitric oxide has been suggested to act as a retrograde transmitter. That is, activation of the NMDA receptor results in the production of nitric oxide by nitric oxide synthase in a postsynaptic neuron from which it rapidly diffuses to enter the presynaptic neuron. Thus, nitric oxide may modulate excitability and enhance synaptic connection through activation of guanylate cyclase in presynaptic terminals and postsynaptic neurons. The concept of a retrograde transmitter recently came into favor in studies of long-term potentiation, and it has been suggested that this may be how presynaptic and postsynaptic connections in the central nervous system are strengthened as a consequence of frequent use. This may explain the difference in the effects of L-NAME and methylene blue on allodynia evoked by strychnine and bicuculline. Taken together, these results suggest that the action of strychnine may be mediated by glutamate receptors on postsynaptic site, but that the action of bicuculline may be mediated by glutamate receptors on presynaptic site.

We recently reported that intrathecal administration of prostaglandin (PG)-E²or PGF²alpha to conscious mice induced allodynia. [9,20,27-29] The time courses of allodynia evoked by PGE²and PGF²alpha coincided with those by strychnine and bicuculline, respectively. [9] Whereas the PGE²-induced allodynia was inhibited by NMDA and non-NMDA receptor antagonists similar to the strychnine-induced one, the PGF²alpha -induced allodynia was inhibited by kainate and metabotropic receptor antagonists similar to the bicuculline-induced one. Furthermore, whereas the PGE²-induced allodynia was inhibited by L-NAME and methylene blue similar to the strychnine-induced one, the PGF²alpha -induced allodynia was inhibited by methylene blue, but not by L-NAME, similar to bicuculline-induced one. [29] The modes of inhibition of strychnine- and bicuculline-induced allodynia by glutamate receptor antagonists and nitric oxide synthase inhibitor were the same as those of the agents for PGE²- and PGF²alpha -induced allodynia, respectively. [9] Thus, many neurotransmitters are involved in the modulation of incoming pain information through a number of local receptor systems at different sites, and the disorder of the association may evoke allodynia.

Opioid-insensitive pain evoked by innocuous tactile stimulation is one of the most difficult problems in pain management. The features of strychnine- and bicuculline-induced allodynia apparently resemble those of patients suffering from postherpetic neuralgia or causalgia. The clinical allodynia may involve plastic changes in neural connectivity and synaptic strength in the spinal cord. Hao et al. [30] developed an animal model that produces tonic and chronic states of allodynia in rats lasting several days and 1-3 months, respectively, after spinal cord injury induced photochemically by laser irradiation. They demonstrated that, although the NMDA receptor was involved in the development of allodynia through excitotoxicity, once the allodynia had developed, NMDA receptor antagonists were ineffective in relieving it. Furthermore, they showed that systemic L-NAME induced an analgesic effect on chronic allodynia-like behavior and suggested that production of nitric oxide may be involved in the maintenance of this abnormal pain-related condition in rats with spinal cord injury. [31] Similarly, the established allodynia induced by strychnine was not blocked by the glycine receptor agonist taurine and the NMDA receptor antagonist ketamine.* Furthermore, we reported [32] that the established allodynia induced by intrathecal injection of PGE²was blocked by L-NAME but not by the PGE receptor and NMDA receptor antagonists. These results imply that the allodynia, once developed, does not require the continued agonist occupancy of receptors in our animal model. The alteration of gene expression associated with allodynia remains to be clarified. The current pharmacologic studies demonstrate that allodynia is induced by at least two different mechanisms and that strychnine- and bicuculline-induced allodynia could be two distinct models for pathologic pain and serve as screening of drugs for chronic pain.

*Minami T, Hara N, Onaka M, Sakai M, Mori H, Ito S: Unpublished observations. 1995.