Introduction

Top Abstract Introduction Methods Results Discussion References

Despite many years of research, the mechanisms involved in the analgesic action of opioids have yet to be identified. Some previous evidence has suggested that EAAs are involved in nociceptive transmission, and that opioid peptides modulate responses to glutamate and its analogs, especially in the spinal cord. It was found that the spinal NMDA-induced hyperalgesia was abolished by i.t. injected opioids (Aanonsen and Wilcox, 1987), and that NMDA receptor antagonists coadministered with morphine facilitated antinociception at a level of the spinal cord (Chapman and Dickenson, 1992). It was also observed that the release of glutamate and aspartate from rat spinal cord slices was attenuated by opioids (Kangrga and Randi , 1991). These results are supported by immunohistochemical and autoradiographic data showing the existence of glutamate, aspartate (De Biasi and Rustioni, 1988) and opioid peptides (Khachaturian et al., 1993), as well as NMDA (Kus et al., 1995) and opioid receptors (Yaksh, 1993) in the substantia gelatinosa of the rat spinal cord. It appears that many of the NMDA effects are mediated by NO that is synthetized from amino acid L-arginine after stimulation of NO synthase by Ca⁺⁺ that fluxes into the cell subsequently to NMDA receptor activation. The consequences of NO production are activation of soluble guanylate cyclase and elevation of the cGMP level (Garthwaite et al., 1988; Bredt and Snyder, 1992; Bruhwyler et al., 1994; Garthwaite and Boulton, 1995). Accordingly, in the spinal cord, NO synthase was found predominately in superficial layers of the dorsal horn and around the central canal, i.e., regions preferentially involved in the sensory transmission (Dun et al., 1993), and an increased expression of NO synthase and/or NADPH-diaphorase a histochemical marker of NO synthase, in the dorsal horn of the spinal cord was observed after noxious stimulation of the rat hind paw (Solodkin et al., 1992; Fiallos-Estrada et al.,1993; Herdegen et al., 1994, Lam et al., 1996). Moreover, it was found that the NMDA-induced release of NO from the rat spinal cord was blocked by NO synthase inhibitors (Ping et al., 1994), and the spinal NMDA-induced hyperalgesia was reduced by i.t. administered inhibitors of NO synthase and guanylate cyclase (Kitto et al., 1992; Meller et al., 1992; Meller et al., 1996). Recent evidence suggests involvement of NO in opioid-mediated effects. Thus it was found that the inhibition of NO synthase attenuates tolerance to the antinociceptive action of morphine (Kolesnikov et al., 1992; Majeed et al., 1994; Bhargava and Zhao, 1996) and enhances the spinal morphine-induced antinociception in the rat (Przewocki et al., 1993). Summing up, the data presented above suggest a possible interaction between NO and opioid systems.

Opioid drugs exert their antinociceptive effect by interacting with three types of receptors, i.e., mu, delta and kappa. The antinociceptive effect of mu and delta receptor agonists after i.c.v., as well as i.t., administration has been demonstrated (Porreca and Burks, 1993; Yaksh, 1993), but there is some controversy about the kappa-mediated antinociception. In contrast to antinociception in the spinal cord, lack of an antinociceptive action of kappa receptor agonists was found at supraspinal sites (Walker et al., 1982). However, the activation of spinal kappa receptors was reported to be less effective in eliciting antinociception in comparison with the activation of spinal mu and delta receptors (Yaksh, 1993). Moreover, some of the kappa receptor-induced effects appeared to be mediated in a nonopioid manner (Przewocki et al., 1983). Antinociceptive effects of exogenously applied opioid receptor agonists were also observed after prolonged noxious stimulation (Porro and Cavazzuti, 1993), which is associated with long-lasting tissue damage, inflammation or neuropathologies and results from a number of complex changes in nociceptive pathways, being probably modulated in the central nervous system in a different way than the pain elicited by short-lasting stimuli (Dray et al., 1994). There is a growing body of evidence that endogenous opioid peptide systems reveal differential and tissue-specific alternations in their activity on long-term noxious stimulation. An enhanced activity of the proenkephalin system is elicited by noxious stimulation of a limited duration, whereas an increase in the biosynthetic activity of the prodynorphin system evokes a most striking and protracted response to chronic pain, especially in the spinal cord (Przewocka et al., 1992; Millan, 1993).

In view of differential involvement of opioid peptides in nociceptive transmission and influence of NO on opioid-mediated effects, the aim of our study was to find out to which extent NO is involved in antinociceptive effects of multiple opioid receptor agonists at a level of the rat spinal cord in both acute and prolonged pain.

	Methods

Top Abstract Introduction Methods Results Discussion References

Animals

Male Wistar rats (300-400 g) from Laboratory Animals Breeding Center (Rembertow, Poland), were used. The rats were housed in single cages lined with sawdust, on a standard 12 hr/12 hr light-dark cycle (8.00 A.M./8.00 P.M.), with food and water ad libitum.

The rats were chronically implanted with i.t. catheters under hexobarbital anaesthesia. They were placed in the David Kopf stereotaxic table, and an incision was made in the atlanto-occipital membrane. A catheter (PE 10, Clay Adams) was carefully introduced to the subarachnoid space at the rostral level of the spinal cord lumbar enlargement according to Yaksh and Rudy (1976). Only animals with a normal motor function were used. Intrathecal injection studies were carried out 5 to 14 days after the surgery. Drugs were dissolved in distilled water and were injected in a volume of 5 µl (single injections) or 10 µl (coadministration), followed by an injection of 10 µl of distilled water to flush the catheter. In different experiments used in this study, control animals were injected i.t. with distilled water and were tested according to the same time schedule as described below for the experimental groups. After completing the experiment, the animals were killed with an overdose of pentobarbital (i.p.).

Analgesia Testing

Acute pain. Antinociceptive effects were evaluated using TF and PP tests. The TF test was carried out using an Analgesia Meter apparatus (mod 33, IITC Inc., Landing, NJ). The animal was gently restrained by hand, and radiant heat was directed onto the animal's tail. The cut-off time was 8 or 16 sec (the latter in experiments with U50,488H). The PP threshold (Randall-Selitto test), necessary to elicit paw withdrawal, was determined using an automatic gauge (Ugo Basile). The animal was gently restrained and an incremental pressure was applied via a piston onto the dorsal surface of the hind paw. The cut-off pressure was 480 g. TF and PP measurements were taken three times at 15- or 10-sec intervals, respectively, and their mean was used for calculations.

Prolonged pain. The rats were lightly anaesthetized with ether, and 100 µl of a 12% formalin solution was injected s.c. into the dorsal surface of the left hind paw. The rat was then placed in a wire cage for observation of the formalin-injected paw. The pain-related behavior was quantified by counting spontaneous flinches and shakes of the injected paw. The flinches and shakes were counted for each animal at several time points: at 0 to 10 min (first phase) and at 10 to 15, 25 to 30, 35 to 40, 45 to 50, 70 to 75 min (second phase) after formalin administration.

Drugs

DAMGO, DPDPE, U50,488H, L-NAME, 7-NI were purchased from RBI (Natick, MA); morphine was obtained from Polfa (Kutno, Poland).

Data Analysis

The results were statistically assessed by an analysis of variance. Intergroup differences were analyzed by Duncan's multiple-range test.

	Results

Top Abstract Introduction Methods Results Discussion References

Acute Pain

The effect of NO synthase inhibition on the antinociceptive effects of opioids, evaluated by the TF and PP tests. 7-NI (50, 100, 400 µg) used in doses which had no effect on the baseline TF and PP latencies (fig. 1), dose-dependently enhanced the effect of morphine (0.5 µg) in both those tests (fig. 1). Morphine given in the above dose caused mild but significant antinociception at 15 min after injection in the TF test. The morphine-induced effect was significantly potentiated by 7-NI at 30 min (100 µg) and 15 to 30 min (400 µg) after their coadministration. In the PP test, morphine (0.5 µg) did not influence the pain threshold, but its coadministration with 7-NI produced antinociception which occurred after 15 min (50 µg) and 15 to 30 min (100, 400 µg) (fig. 1).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   The effect of 7-nitroindazole (7-NI; 50-400 µg) on the morphine (MOR; 0.5 µg)-induced antinociception after their i.t. coadministration, evaluated by the tail-flick (TF) and paw pressure (PP) tests in the rat. Control animals were injected i.t. with distilled water and were tested according to the same time schedule as experimental group. 7-NI used in the above doses did not influence the baseline TF (the results were as follows: 50 µg, 3.7 ± 0.3, 3.5 ± 0.1, 3.6 ± 0.4, 3.8 ± 0.6; 100 µg, 3.7 ± 0.5, 4.1 ± 0.6, 3.6 ± 0.3, 3.7 ± 0.3; 400 µg, 3.7 ± 0.4, 3.6 ± 0.1, 3.8 ± 0.5, 3.8 ± 0.4 after 0, 15, 30 and 60 min, respectively) and PP values (the results were as follows: 50 µg, 200 ± 3.8, 202 ± 3.2, 196 ± 2.4, 209 ± 5.2; 100 µg, 206 ± 2.6, 209 ± 0.6, 208 ± 4.6, 208 ± 5.2; 400 µg, 208 ± 4.0, 198 ± 2.2, 207 ± 5.8, 202 ± 2.4 after 0, 15, 30 and 60 min, respectively). The data are presented as mean ± S.E.M. +Indicates a significant difference (P < .05) between morphine and control. *Indicates a significant difference (P < .05) between 7-NI + morphine and morphine alone. The number of animals used for each dose was 6 to 11.

View larger version (47K): [in this window] [in a new window]   Fig. 2.   The dose-response relationship evaluated at 30 min after i.t. coadministration of L-NAME (50 µg) with morphine (0.05-0.5 µg) (A), and at 15 min after i.t. coadministration of L-NAME (50 µg) with DAMGO (0.1-2.5 ng) (B), L-NAME (50 µg) with DPDPE (0.02-0.5 µg) (C) and L-NAME (50 µg) with U50,488H (10-100 µg) (D) in the tail-flick (TF) and paw pressure (PP) tests in the rat. Control animals were injected i.t. with distilled water and were tested according to the same time schedule as experimental group. The data are presented as mean ± S.E.M. *Indicates a significant difference (P < .05) between L-NAME coadministered with opioid receptor agonists and L-NAME as well as opioid receptor agonists alone. The number of animals used for each dose was 6 to 9.

Prolonged Pain

The effect of NO synthase inhibition on the antinociceptive effects of opioids, evaluated in the formalin test. L-NAME injected in a dose of 100 µg significantly suppressed the formalin-induced behavior in the first, but not the second phase of the formalin test. L-NAME in a dose of 400 µg did not influence the first phase of the formalin test, but produced significant attenuation of the second phase (fig. 3).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   The effect of i.t. administered L-NAME (100-400 µg) on the formalin-induced nociception in the rat. L-NAME was administered 7 min before peripheral formalin administration. The labels on the x-axis indicate the end of a 10-min period of counting flinches and shakes in the first phase (0-10 min), and the end of 5-min periods at several time points of the second phase (10-75 min) of the formalin test. The data are presented as mean ± S.E.M. *Indicates a significant difference (P < .05) between L-NAME and vehicle-treated control. The number of animals used for each dose was 8 to 25.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   The effects of i.t. administered DAMGO (0.5-500 ng), DPDPE (2-50 µg) and U50,488H (10-100 µg) on the formalin-induced nociception in the rat. Opioid receptor agonists were administered 7 min before peripheral formalin administration. The labels on the x-axis indicate the end of a 10-min period of counting flinches and shakes in the first phase (0-10 min), and the end of 5-min periods at several time points of the second phase (10-75 min) of the formalin test. The data are presented as mean ± S.E.M. *Indicates a significant difference (P < .05) between opioid receptor agonists and vehicle-treated control. The number of animals used for each dose was 8 to 25.

View larger version (27K): [in this window] [in a new window]   Fig. 5.   The time course of i.t. administered L-NAME (100 µg) on the antinociception induced by i.t. administered DAMGO (15 ng), DPDPE (10 µg) and U50,488H (50 µg) in the formalin test in the rat. L-NAME and opioid receptor agonists were coadministered 7 min before peripheral formalin administration. Control animals received injections i.t. with distilled water and were tested according to the same time schedule as experimental group. The labels on the x-axis indicate the end of a 10-min period of counting flinches and shakes in the first phase (0-10 min), and the end of 5-min periods at several time points of the second phase (10-75 min) of the formalin test. The data are presented as mean ± S.E.M. *Indicates a significant difference (P < .05) between L-NAME coadministered with opioid receptor agonists and L-NAME as well as opioid receptor agonists alone. The number of animals used for each dose was 8 to 25.

View larger version (37K): [in this window] [in a new window]   Fig. 6.   The dose-response relationship for i.t. administered L-NAME (100 µg) on the antinociception induced by i.t. administered DAMGO (0.5-15 ng), DPDPE (2-10 µg) and U50,488H (10-100 µg), evaluated at 35 to 40 min after formalin injection to the rat. L-NAME and opioid receptor agonists were coadministered 7 min before peripheral formalin administration. Control animals received injections i.t. with distilled water and were tested according to the same time schedule as experimental group. The data are presented as mean ± S.E.M. *Indicates a significant difference (P < .05) between L-NAME coadministered with opioid receptor agonists and L-NAME as well as opioid receptor agonists alone. The number of animals used for each dose was 8 to 25.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Our study shows that 7-NI, a specific inhibitor of neuronal NO synthase, enhances the antinociceptive effect of morphine after their i.t. coadminstration to the rat in the TF and PP tests. This finding supports our previous results that showed that concomitant i.t. injection of L-NAME and morphine elicits strong and profound antinociception in the rat (Przewocki et al.,1993). Moreover, in our study we have found that i.t. coadministration of L-NAME given in subanalgesic doses with morphine as well as with specific mu and delta opioid receptor agonists DAMGO and DPDPE, respectively, produces dose-dependent, strong antinociception in both the TF and PP tests. A similar, although weaker effect was reported after coadministration of L-NAME with U50,488H, a specific kappa receptor agonist, as evaluated by the PP test. Interestingly, no such effect was observed in the TF test. Furthermore, our previous study showed potentiation of the morphine-induced antinociception by hemoglobin that binds the released NO (Przewocki et al.,1993). Thus all the above results indicate that NO in the spinal cord may be tonically active in inhibiting morphine-activated nociceptive pathways. Our study is in line with the recent results of Xu and Tseng (1995) who reported that N-nitro-L-arginine, hemoglobin and methylene blue, an inhibitor of guanylate cyclase and/or NO synthase (Luo et al., 1995), injected i.t. potentiate the antinociception induced by morphine administered i.c.v. The above findings suggest that NO may exert a pronociceptive action at a level of the spinal cord. In fact, inhibition of NO synthase in the spinal cord leads to antinociception (Meller and Gebhart 1993; Przewocki et al., 1993; Meller et al., 1996) and NO-donating substances administered i.t. produce hyperalgesia in response to acute nociceptive stimuli (Kitto et al., 1992; Shibuta et al., 1995; H. Machelska, R. Przewocki, M. Radomski and B. Przewocka, unpublished data). These behavioral data are supported by the results of biochemical studies that show that in the spinal cord the NO donor enhances the release of substance P and calcitonin gene-related peptide, both neuropeptides being principally involved in the nociceptive transmission (Garry et al., 1994), while the NO synthase inhibitor attenuates the NMDA-induced release of glutamate and citrulline a marker of NO release (Sorkin, 1993). Immunohistochemical studies further confirm the pronociceptive action of NO, because it has been found that the c-fos expression induced by noxious mechanical stimuli is reduced by the NO synthase inhibitor administered i.t. (Lee et al., 1992).

However, NO appears to act as a pronociceptive as well as an antinociceptive agent at supraspinal and peripheral sites. It has been found that L-NAME applied i.c.v. produces dose-related inhibition of the formalin-induced paw licking in mice (Moore et al., 1991) and prolongation of the TF and PP latencies, while NO donors administered i.c.v. shorten these latencies in the rat (Machelska et al., unpublished data). Conversely, NO donors have been also reported to exert an antinociceptive action when they are injected both i.c.v. (Duarte and Ferreira, 1992) and i.pl. (Duarte et al., 1990; Ferreira et al., 1991). There are also some conflicting data on the influence of NO synthase inhibition on the antinociception mediated by morphine at supraspinal and peripheral sites. It was found that the i.c.v. morphine-induced antinociception was not affected by N-nitro-L-arginine, hemoglobin or methylene blue, injected i.c.v. (Xu and Tseng, 1995). In contrast, our experiments demonstrated potentiation of the morphine-induced antinociception by L-NAME after their i.c.v. coadministration (data not shown). Moreover, Duarte and Ferreira (1992) found that the morphine-induced antinociception was prevented by methylene blue, but not by N-iminoethyl-L-ornithine, after their i.c.v. administration as evaluated by the TF test in the rat. It was also found that the peripheral morphine-induced antinociception was potentiated by the NO donor injected i.pl. (Ferreira et al., 1991).

Precise mechanisms underlying the opposing effect of NO on nociception are as yet unknown. However, some studies suggest that NO may induce pronociceptive effects at low concentrations and antinociceptive ones at higher concentrations (Kawabata et al., 1994). Meller et al. (1992) demonstrated that low doses of NMDA, which possibly generate low concentrations of NO, produce hyperalgesia, whereas generation of higher levels of NO may be associated with the antinociception produced by administration of higher doses of NMDA. Additionally, Kawabata et al., (1993) suggested that L-arginine, a substrate of NO, could play a dual role in the nociceptive processing in the brain, depending on the pathway in which it is involved; for example, activation of the kyotorphin-Met-enkephalin pathway leads to antinociception, but activation of the NO-cGMP pathway results in hyperalgesia. Moreover, Xu and Tseng (1993) proposed that different descending nociceptive pathways could be involved in the supraspinal nociceptive transmission. They found that the i.c.v. morphine-induced antinociception was enhanced by i.t. L-NAME, whereas the i.c.v. -endorphin-induced antinociception was potentiated by L-arginine, the latter effect being attenuated by i.c.v. L-NAME. For that reason, those authors proposed that stimulation of opioid receptors by -endorphin applied supraspinally induced release of Met-enkephalin and subsequent stimulation of delta opioid receptors in the spinal cord. However, stimulation of mu opioid receptors by morphine given supraspinally activated the spinopetal serotonergic and noradrenergic systems and subsequently stimulated alpha₂-adrenoreceptors and 5-HT receptors in the spinal cord (Xu and Tseng, 1993). Eventually, yet another factor to consider is, that NO may exist in a reduced or an oxidized form, either of which may produce distinct pharmacological effects (Lei et al., 1992; Lipton et al., 1996).

A pronociceptive action of NO in the spinal cord was also observed after prolonged noxious stimulation. Our study has shown that L-NAME administered i.t. attenuates the formalin-induced nociception in the rat. This observation is in agreement with other behavioral and electrophysiological data. It was previously reported that L-NAME injected i.t. reduced the responses of dorsal horn neurons to locally injected formalin in both the first and second phases of the formalin test (Haley et al., 1992), and attenuated the formalin-induced behavior in the second phase (Malmberg and Yaksh, 1993). Additionally, inhibition of the formalin-induced paw licking following i.c.v., i.p. or oral administration of L-NAME to mice, mainly in the second phase (Moore et al., 1991), as well as inhibition of the second peak of firing of dorsal horn neurons after i.v. injection of L-NAME to the rat (Haley et al., 1992) were observed. An antinociceptive action in the second phase of the formalin test was also reported for 7-NI administered i.p. to mice (Moore et al., 1993). The above results indicate preferential inhibition of the second peak of the formalin-induced nociception after the blockade of NO synthase. This effect corresponds to action of the NMDA receptor antagonists which are effective in inhibiting the repetitive C-fiber stimulation (wind-up), which is related to the second phase of the formalin test (Wheeler-Aceto et al., 1990; Chapman et al., 1994). A biochemical study revealed that peripheral formalin injection resulted in a significant increase in the spinal levels of glutamate and aspartate during the first phase only, and produced significant enhancement of the release of citrulline, during both the first and second phases (Malmberg and Yaksh, 1995). Thus, one can expect that stimulation of NMDA receptors during the first phase of the formalin test leads to an increased level of NO in both phases, while blockade of NMDA receptors and/or inhibition of NO synthase resulted in antinociception principally in the second phase.

Intrathecal injection of specific opioid receptor agonists significantly diminishes both phases of the formalin test. The antinociceptive action of the moderately acting doses of opioid receptor agonists is potentiated by L-NAME (i.t.) in the second phase of the formalin test, this effect depends on the dose used of the opioid receptor agonist. Our study corroborates other behavioral and electrophysiological data that showed that both i.t. (Chapman and Dickenson, 1992) and i.c.v. (Calcagnetti et al., 1988) administration of opioid receptor agonists produced analgesia in the formalin test and, that i.t. opioids inhibited the formalin-induced activity of dorsal horn neurons (Dickenson and Sullivan, 1987). In agreement with other findings (Chapman and Dickenson, 1992; Chapman et al., 1994), our study shows that opioids (especially DAMGO and DPDPE) are more effective in inhibiting the formalin-induced nociception in the first phase than in the second one. This is due to the fact that i.t. opioids effectively inhibit the C-fiber-evoked input responses of dorsal horn neurons, which contribute to the first phase of the formalin test (Chapman et al., 1994). In contrast, the enhanced C-fiber-evoked responses of dorsal horn neurons contribute to the second phase of the formalin test that is mediated by NMDA receptors, and is less sensitive to i.t. opioids (Wheeler-Aceto et al., 1990; Chapman et al., 1994). Moreover, we observed that i.t. injected L-NAME potentiated the opioid-mediated antinociception mainly in the second, but not the first phase (except DPDPE) of the formalin response, which may be due to a higher efficacy of L-NAME to inhibit the formalin-induced nociception in the second phase of that test (as described above).

Our results suggest involvement of NO in the antinociception mediated by different types of opioid receptors in the rat spinal cord. We found that inhibition of NO synthase in the spinal cord strongly potentiatiated the mu and delta receptor-mediated antinociception. We also observed potentiantion of the kappa mediated antinociception; however, that effect was weaker compared to the mu and delta mediated antinociception and occurred only in the PP test and at the last two time points of the formalin test. It is suggested that when the intensity of heat (but not pressure) stimulus is increased the antinociceptive efficacy of kappa agonists is diminished (Millan et al., 1989; Millan, 1990). It seems unlikely that this effect significantly contributes to the lack of potentiation of the U50,488H activity by L-NAME, because in the TF test the heat stimulus was adjusted to the moderate/low intensity (basal latencies approx. 6 sec). Moreover, relative weakness of the spinal action of kappa agonists as compared to their systemic activity in the rat was observed (Millan et al., 1989). This discrepancy can be explained by the finding that injury of the spinal cord due to the cannulation may rise the level of dynorphin in the dorsal horn (Przewocki et al., 1988; Millan et al., 1989) which may underline a cross-tolerance to exogenous kappa agonists and their apparently lower efficacy (Millan, 1990).

Many of the commonly used NO synthase inhibitors are nonselective, and they act on neuronal as well as endothelial NO synthase. The blockade of endothelial NO synthase by L-NAME results in vasoconstriction and a subsequent increase in blood pressure (Moore et al., 1993; Semos and Headley, 1994). However, it seems unlikely that the cardiovascular effects of NO synthase inhibitors contribute significantly to potentiation of the opioid-mediated antinociception in the spinal cord, observed in our study. It was found that L-NAME injected both i.t. and i.c.v. in doses comparable (and even much higher) to those used in the present study did not significantly affect the basal blood pressure of rats and mice (Moore et al., 1991; Haley et al., 1992). L-NAME was reported to influence also the resting cerebral blood flow, but this effect was observed after relatively high doses of L-NAME injected i.v. (Yang, 1996). It was also found that L-NAME may act as a competitive muscarinic receptor antagonist (Buxton et al., 1993). However, the endogenous spinal cholinergic system appears to be antinociceptive (Yaksh et al., 1985; Iwamoto and Marion, 1993, 1994), so if this were the predominant effect of L-NAME, it would be expected to be pro- rather than antinociceptive. Moreover, an antimuscarinic activity of L-NAME was found in guinea pig brain (Buxton et al., 1993) but it did not exhibit any affinity for muscarinic receptors in the rat spinal cord (Buccafusco et al., 1995). Additionally, 7-NI, the most selective inhibitor of neuronal NO synthase that neither increases the mean arterial blood pressure in rodents and cats (when injected i.p. or even i.v. at high doses which exceed antinociceptive ones) (Bland-Ward and Moore, 1995), nor exhibts affinity for muscarinic receptors, also potentiated the morphine-induced antinociception in our studies.

The exact mechanism by which inhibition of NO synthase enhances the opioid-mediated antinociception has not been clarified as yet. The most important aspects of the opioid-mediated antinociception in the spinal cord are presynaptic reduction of transmitter release from C-fibers, and an increase in the membrane K⁺ conductance or a decrease in the Ca⁺⁺ conductance (Satoh and Kuraishi, 1991; North, 1991; Kangrga and Randic, 1991). NO is produced by NO synthase which is activated by the Ca⁺⁺ influx, mainly after NMDA receptor activation (Garthwaite, 1991). In the light of the above data it may be expected that opioids are able to prevent activation of NMDA receptors and the subsequent NO synthesis by reducing the presynaptic release of glutamate. The latter assumption is supported by the findings that morphine produces suppression of the formalin-evoked release of citrulline from the rat spinal cord (Malmberg and Yaksh, 1995). Furthermore, opioids are able to prevent the release of Ca⁺⁺ from intracellular stores (mediated by inositol trisphosphate) by reducing the SP release (Collin et al., 1992). Alternatively, it is also possible that potentiation of the opioid antinociception by NO synthase inhibition may result from a reduced activity of intracellular signaling in both the cGMP and cAMP pathways. Further, the inhibition of NO synthase may diminish the NMDA receptor-mediated synaptic efficacy after the nociceptive input and, in consequence, profoundly enhance the opioid antinociception. A direct interaction between NO and opioid receptors may also contribute to the observed effect. It has been recently reported, that the NO-donor SIN-1 may down-regulate the alpha subunit of Gs protein in C6 glioma cells, by increasing ADP ribosylation (Young et al., 1997). A similar interaction between NO and other G proteins coupled to the opioid receptors, e.g., Gi or Go, may also take place. It cannot be excluded, either, that the NO synthase inhibitor influences the metabolism of opioids or their clearance from intrathecal space. Prolongation of the antinociceptive effect of DPDPE and U50,488H by L-NAME seems to be in line with the above suggestion. In conclusion, our study has demonstrated that inhibition of NO synthase potentiates the mu, delta and to a lesser extent, kappa receptor-mediated spinal antinociception in both acute and prolonged pain.