Gen. Pharmac. Vol. 28, No. 3, pp. 443-448, 1997 Copyright © 1997 Elsevier Science Inc. Printed in the USA.

ELSEVIER

ISSN 0306-3623/97 $17.00 + .00 PII S0306-3623(96)00295-9

All rights reserved

Influence of Opiate Tolerance and Calcium Channel Antagonists on the Antinociceptive

Effects of L-Arginine and N°-Nitro L-Arginine

Enrique Contreras, .1 Aniza Germany, 1 Paulo Gonzalez 3 and Beryl Norris 2 DEPARTMENTS OF IpHARMACOLOGY AND 2pHYSIOLOGY,

FACULTY OF BIOLOGICAL SCIENCES, AND 3DEPARTMENT OF PHARMACY, FACULTY OF PHAmaACV, UNIVERSITY OE CONCEPCI6N, COnCEPCI6N, CHILE [FAx: 56-41-24-59-75]

ABSTRACT. 1. The antinociceptive effects induced by L-arginine (L-Arg 300-600 mg sc) or N C- nitro-L-arginine (NOArg 20-70 mg sc) in mice were assessed by the hot-plate test.

2. The antinociception induced by both agents was antagonized by naloxone. L-Arg significantly re- duced the effects of the largest doses of morphine (3, 5, and 10 mg/kg) or pentazocine (7.5, 15, and 30 mg/kg).

3. Morphine antagonized L-Arg-induced antinociception but did not change the responses to NOArg.

4. Diltiazem (10 mg/kg) or verapamil (10 mg/kg) decreased L-Arg antinociceptive responses, whereas the effects of NOArg were enhanced.

5. The antinociceptive effects of L-Arg and NOArg were also tested in mice rendered tolerant to morphine or pentazocine. Whereas the effect of L-Arg were lower in tolerant animals, the responses to NOArg were unchanged.

6. The results suggest the involvement of opiate mechanisms and NO synthesis in L-ARG-induced antinociception and a lesser influence of opiate mechanisms in the antinociception induced by NOArg. Copyright © 1997 Elsevier Science Inc. GEN P~AmaAC 28;3:443--448, 1997.

KEY WORDS. L-arginine, NG-nitro-L-arginine, antinociception, opiates, calcium channel antagonists

I NTRODUCTION

The involvement of nitric oxide (NO) in the analgesic responses to opiates was suggested by the work of Tseng et al. (1992) who dem- onstrated that the antinociceptive effect of [3-endorphin was in- creased by pretreating of mice with L-arginine (L-Arg), whereas the concomitant administration of NC-nitro-L-arginine (NOArg) at- tenuated L-Arg-induced potentiation.

The mechanisms responsible for the analgesic action of L-Arg are not fully understood. According to Xu and Tseng (1993), the aminoacid exerts a specific action on endorphin-induced analgesia, but is ineffective on the antinociception induced by opioids acting through different receptor mechanisms. As NOArg, a selective in- hibitor of NO synthase, decreases neurotransmitter release in brain synaptosomes induced by the administration of excitatory amino- acids (Hirsch et al., 1993), it has been suggested that NO formation might enhance the release of neurotransmitters, including those re- sponsible for pain perception. Kawabata et al. (1992) have proposed the involvement of L-Arg in the synthesis of kyotorphin, a peptide that enhances the release of Met-enkephalin. Other authors (McDonald et a/., 1994) have suggested that NO formed from L-Arg induces antinoc- iception, a proposal base on the demonstration that N20-induced anti- nociception is enhanced by L-Arg administration. The use of the en- zyme inhibitor NOArg in the analysis of the role of NO in nociception is limited by the antinociceptive effects of the drug.

The aim of the present study is to elucidate the possible involve- ment of common mechanisms in the antinociceptive responses to the NO precursor L-Arg and to the NO synthase inhibitor NOArg. We explored the following issues: (1) The effect of both L-Arg and

* To whom correspondence should be addressed. Received 1 November 1995, accepted 28 June 1996.

its nitro derivative on the antinociception induced by the ~- or K-opiates morphine, U50488H, and pentazocine; (2) the influence of tolerance to either morphine or pentazocine on antinociception induced by L-Arg and NoArg; and (3) the influence of two calcium channel antagonist, diltiazem and verapamil, on the responses to L-Arg and NOArg, as it has been demonstrated that the concomi- tant administration of calcium channel antagonists induces syner- gistic effects on opiate-induce antinociception (Contreras et al. 1988; Dierssen et al. 1990).

MATERIALS AND METHODS General

Male adult albino Swiss Webster mice 12-15 weeks of age, weighing 26-30 g from the animal reproduction laboratories of the Depart- ment of Pharmacology of the University of Concepci6n were used in all experiments. Mice were housed in groups of 7, maintained on a 12:12 light:dark cycle at constant room temperature (22±1°C) and allowed free access to food and water. Determinations of anti- nociceptive responses were carried out in the between 14:30 and 18:00 hr under normal room light and temperature (22+2°C) con- ditions. All groups consisted of 14 mice, and each animal was used for one experimental condition.

Drugs

The drugs used were morphine HCI (May and Baker, Dagenham, England); naloxone, a gift of Endo Labs. (Garden City, NY, USA); pentazocine (Winthrop Prods., New York, USA); U50488H (trans3,4- dichloro-N-methyl-N-[2-pyrrolinidyl) cyclohexenil] benzeneacet- amide methane sulphonate) from Research Biochemicals (Natick, MA, USA); L-arginine, D-arginine, NC-nitro-L-arginine, diltiazem,

444 E. Contreras et al.

and verapamil from Sigma Chemical (St. Louis MO, USA). The drugs were dissolved in saline and administered sc in a volume of 10 mL/kg unless otherwise specified.

Analgesic test

The hot-plate test as described by Eddy and Leimbach (1953) was used for assessing analgesia, (temperature kept at 48±0.5°C). The endpoints considered were jumping off the plate or leg kicking. Each mouse was tested twice before drug administration and the values were averaged to obtain a baseline. Reaction times were determined by 30-min intervals over 90 min. To avoid severe animal burning a cut-off time of 30 sec was used. The total antinociceptive response was obtained as the area under the time-response curve, calculated from the experimental values obtained every 30 min.

To study the effects of the association of L-Arg or NOArg and opiate drugs, the aminoacids were given 5 min before opiates admin- istration. To analyze the influence of the calcium antagonists vera- pamil or diltiazem on the antinociceptive responses to L-Arg or NOArg, the antagonist were given ip 10 min before injecting either amino acid.

Induction of tolerance

To induce opiate tolerance, a single sc dose of morphine (100 mg/ kg) or pentazocine (45 mg/kg) was administered 6 hr prior to the assay of test doses. Control groups were injected sc with saline. Mice were considered tolerant to the antinociceptive effects of the opiates when the effects of the test dose of the analgesic in the primed groups differed from those of the saline controls at a level of proba- bility of 0.05.

Statistical analysis

The results are expressed as means of the area under the time- response curves±SEM. The significance of the differences in the mean responses of the experimental groups was determined by anal- ysis of variance (ANOVA) and the Newman-Keul's test. A level of probability of 0.05 was accepted as statistically significant.

RESULTS Antinociception induced by L-Arg and NOArg

Mice injected with L-Arg, D-Arg, or NOArg at the doses indicated in the present experiments did not exhibit changes in gross animal behavior over the observation period compared with the saline- injected control mice.

3OO

~ +t 100 +1 IO0

0 0 3~o 4~o 6;o 3; s'o ;o

L-Arg (mg/kg) NOArg (mg/kg)

FIGURE 1. The antinociceptive effects of the sc administration of L-Arg and NOArg, followed by saline administration, are ex- pressed as the area under the time-response curve±SEM (squares). Naloxone (2 mg/kg sc 5 min after each drug) signifi- cantly reduced the analgesic effects of L-Arg and NOArg (trian- gles). *P<0.05. Naloxone alone was ineffective (n= 14).

TABLE 1. Effects of NOArg on the antinociceptive response to L-Arg in mice

Antinociceptive response (Area ± SEM)

Dose (mg/kg) L-Arg L-Arg + NOArg

300 88 ± 8 57 ± 5 450 112 ± 10 68 ± 8* 600 224 ± 17 81 ± 10"

NOArg was given sc at a dose of 10 mg/kg 5 min before L-Arg administration. Responses (Area -+ SEM) to saline and NOArg (administered alone) were 12 -+ 6 and 60 -+ 5, respectively.* Statistical differences from the effects of L-Arg administered alone (n = 14 mice per group).

The sc administration L-Arg or NOArg significantly increased the reaction time to the hot plate, indicating a concentration- dependent analgesic effect (Fig. 1). The effects were not observed for more than 90 min after their administration. Naloxone (2 rag/ kg, 5 min after each drug) displace the L-Arg dose-response curve to the right, indicating antagonism of the L-Arg-induced antinoci- ception. That the effect of the aminoacid is stereospecific is indi- cated by the finding that D-Arg up to 600 mg/kg did not induce per se an antinociceptive response. Therefore, the analgesia of L-Arg does not seem to be due to unspecific drug effects. The analgesia in- duced by NOArg also was antagonized by naloxone, although the nature of the antagonism could not be established.

The analgesic responses induced by the sc administration of L-Arg were antagonized by 10 mg/kg NOArg (Table 1).

Effects of calcium channel antagonists on antinociception induced by L,Arg and NOArg

Diltiazem and verapamil reduce L-Arg-induced antinociception. In contrast, both diltiazem and verapamil significantly increased the analgesic response induced by NOArg (Fig. 2). The responses to L-Arg and NOArg in mice previously injected with either saline (control groups) or the calcium channel antagonists are shown in Fig. 2.

Diltiazem or verapamil (10 mg/kg ip) were given 10 min before the administration of L-Arg or NOArg. Neither diltiazem nor verapamil induced significant changes per se in the mice reaction time to the pain stimuli. The administration of diltiazem or vera- pamil either alone or in combination with L-Arg or NO-Arg did not

300 ¢~ 200

200

÷' lOO ~ ~

• ~

0 "" 0 ,^" 3;0 450 6;0 10 2'0 4'0 gO

L-Arg (mg/kg) NOArg (mg/kg)

FIGURE 2. Effects of diltiazem 10 mg/kg (solid triangles) and verapamil 10 mg/kg (open triangles) on the antinociceptive re- sponses to L-Arg and NOArg in mice (solid squares). L-Arg and NOArg were given sc; the calcium channel antagonists were ad- ministered ip. Asterisks indicate significant differences as com- pared with the respective control group responses (P<0.05). The calcium channel antagonists given alone did not change the usual reaction time in the hot plate test (n= 14).

L-Arginine and NC'-Nitroarginine Antinociception 445

300-

,<

,o ,

,<

,~s ; ; ,; o . . . . . . . . ~ . s T . s ~s 3o o ® . , s l . s s

Morphine (n~l/kg) Pentazocine (mo/kg) USO488H (mWkg)

FIGURE 3. Effect of L-Arg on the antinociceptive responses to sc administration of morphine, pentazocine, or U50488H. L-Arg 600 mg/kg sc was administered 5 min before the analgesic. Open triangles, responses to L-Arg in control groups, i , responses to opiates plus saline; -A-, the effects of opiates in association with L-Arg. *Indicates significant decrease (P<O.05) in an experimental group as compared with the reaction times in the opiate saline-injected mice (n= 14).

affect gross animal behavior over 120 min as compared to the saline- injected mice.

Effects of L.Arg on the antinociception induced by morphine, pentazocine, or U50488H

A control group was injected with L-Arg (600 mg/kg sc) followed by sc saline 5 rain afterwards. A similar dose of L-Arg given 5 min before morphine or pentazocine significantly reduced the responses to the largest doses of morphine (10 mg/kg) or pentazocine (30 rag/ kg). The antinociceptive effects of U50488H were unchanged (Fig. 3).

Effect of NOArg on the antinociception induced by morphine, pentazocine, or U50488H

The effect of NOArg at 20 mg/kg followed by a saline injection 5 min later was compared with the effect of a similar dose of NOArg given 5 rain before administration of morphine, pentazocine, or U50488H. Fig. 4 shows that the responses to morphine and unaf- fected. NOArg at 3 mg/kg increased U50488H effects.

Morphine antagonized the antinociception induced by L.Arg but not that induced by NOArg

To obtain further information on the apparent mutual antagonism of opiates and L-Arg, the effect of morphine (5mg/kg) on the anti-

nociceptive responses to L-Arg was investigated. As seen in Fig. 5, the administration of the opiate 5 min after L-Arg induced a shift to the right of the dose-response curve of the NO precursor. However, NOArg effects were not changed by the concomitant administra- tion of morphine.

Antinociception of L-Arg and NOArg in mice tolerant to morphine or pentazocine

Acute tolerance to morphine or pentazocine developed following a sc dose of 100 and 45 mg/kg respectively, administered 6 hr before the injection of L-Arg or NOArg (Fig. 6). Opiate administration in- duced tolerance, as demonstrated by a reduction of the effects of challenge doses of both drugs; the response to a test dose of mor- phine (5 mg/kg sc) was reduced from a mean area of analgesia of 206___36 to 40_+6.2 and the effect of pentazocine (15 mg/kg sc) was reduced from 189_+ 18 to 28_+8. The analgesic effect of L-Arg was reduced in morphine- and pentazocine-tolerant mice, indicating some degree of cross-tolerance. However, the intensity of the appar- ent cross-tolerance was greater in animals tolerant to morphine than in pentazocine-pretreated subjects. Although mice tolerant to pen- tazocine also were tolerant to US0488H (Contreras et al., 1993), the acute treatment with either 30, 60, or 90 mg/kg of U50488H did not change the antinociceptive response to test doses of this agent ad- ministered 6 hr later. In contrast, morphine- or pentazocine- induced tolerance showed no cross-tolerance with the NOArg-

-o t " I S 00f# !00 utJ 200 W ~" 200 4 .

i +1 +l t -~1 lOO lOO lOO-

O" ~ ~ lvO C , 0 7.5 1'5 3'0 Morphine (mg/kg) Pentazocine (mg/kg)

Q

o.~s l 's 3'.0 Ur~ I88H (mg/kg)

FIGURE 4. Effect of NOArg 20 mg/kg sc on the antinociceptive responses to morphine, pentazocine, or U50488H. Symbols as in Fig 3. Responses to NOArg in control groups are shown as open triangles. *Significantly higher (P<O.05) than the mean response of the group injected with the analgesic plus saline (n= 14).

446 E. Contreras et al.

.<

uJ o 4 ;

+~ <[ u.I iv" < [

300-

200 ÷

IO0

0 3;o 4~o do L-Arg (mg/kg)

< tn uu o

+1

<

'°° t 3001

200 t

100 t O:

;0 s~ A NOArg (mg/kg)

FIGURE 5. Dose-response curves showing antinociceptive effect (solid squares) of L-Arg and NOArg sc in saline-injected mice. Open triangles, antinociceptive effects of morphine (5mg/kg sc) administered to control groups of mice. Solid triangles, effects of L-Arg or NOArg in association with morphine (5mg/kg sc) given 5 rain after either aminoacid. *Signifi- cant difference (P<0.05) from the effects of L-Arg (n = 14).

induced antinociception as indicated by the lack of reduction of the NOArg-induced antinociception.

DISCUSSION

The present study confirms the results of Harima et al. (1991) , Ka- wabata et al. (1992) , and Xu and Tseng (1993). It indicates that the administration of L-Arg (but not D-Arg) and NOArg induces anti- nociception in mice. The analgesic response of L-Arg has been caus- ally related to the synthesis of NO in the brain (Haley et al., 1992), which is supposed to decrease pain perception. Several authors have claimed that NO exerts a role in painful perception (Moore et al., 1991; Morris et al., 1992; Bredt and Snyder, 1994). The induction of antinociception by NO has been suggested by a number of experi- mental results. Membrane-permeable analogs of cGMP produce sig- nificant antinociception in analgesic tests after caudal reticular for- mation (Levy et al., 1993) or intrathecal administration (Iwamoto and Marion, 1994). Thus, the exact role of NO in pain perception is not fully understood.

An additional hypothesis (Kawabata et al., 1993) proposes that L-Arg may affect nociception by acting as a kyotorphin (L-tyrosyl- L-arginine) precursor m the brain, which is known to be an endoge- nous Met-enkephalin releaser (Takagi et al., 1979). If the analgesic response to L-Arg is mediated through the synthesis of kyotorphin, it may be expected that naloxone antagonism to Met-enkephalin- induced analgesia should allow the expression of the supposed pro- nociceptive response of NO formed from L-Arg administration. However, no facilitation of the pain response was observed after the administration of L-Arg plus naloxone. This inability of L-Arg to fa- cilitate painful responses has been suggested to be the result of the high Km value of kyotorphin synthase compared to the physiologic concentration of the aminoacid (Kawabata et al., 1993); in contrast, the low Km value of NO synthase may explain the finding that exog- enously given L-Arg activates kyotorphin synthesis, but not NO production.

The main purpose of the present work is to gather further infor- mation on the mechanisms involved in the antinociception induced by L-Arg and NOArg. The concomitant administration of L-Arg and opiates decreased the responses to the highest doses of mor- phine and pentazocine; these results are difficult to interpret, but it seems reasonable to consider the interaction of L-Arg with opiate mechanisms in the induction of antinociception.

Our results do not allow us to elucidate the mechanisms involved in the antinociception induced by L-Arc or NOArg. Although anti- nociception induced by both agents are naloxone sensitive, their ad- ministration in association with opiates gave different results. Whereas L-Arg antagonizes morphine and pentazocine effects a dose of 5 mg/kg of morphine antagonized the response to L-Arc,

NOArg did not affect the response to morphine or pentazocine, and a significant increase of the response to 3 mg/kg U50488H was ob- served. The apparent mutual antagonism L-Arc-morphine also was expressed in the attenuation of tolerance to the opiate. These results suggest that NO resulting from L-Arc administration might induce (1) a pronociceptive effect that reduces opiate-induced antinocicep- tion; or (2) the interaction of both drugs in different stages of the processing of nociceptive responses.

Because the induction of tolerance is one of the most characteris- tic properties of opiates, we thought it of interest to elucidate whether its occurrence might affect the response to the aminoacid and its nitro derivative. The significant decrease of the antinocicep- rive responses to L-Arc administered to morphine- or pentazocine- tolerant mice favors the involvement of common mechanims in the responses to opiates and the aminoacid. The results suggest that the L-Arc response is associated with mechanims related to bL- as well as K-agonist opiates. Considering that pentazocine is an I ~- agonist and K-antagonist, the lower degree of tolerance observed after ad- ministration of this drug than that induced by morphine also may reflect a smaller involvement of K- opioid receptors in the analgesic response to L-Arc. This seems to be confirmed by the absence of an- tagonism to U50488H responses.

Although the analgesic mechanism of NOArg is unclear, the pos- sibility that it acts through nonspecific mechanisms has been dis- missed. In this regard, Moore et al., (1991) showed that the drug does not significantly affect blood pressure at dose levels usually em- ployed to induce antinociception. The drug does not induce signifi- cant changes in mice behavior, which could be thought responsible for its antinociceptive response; our results are consistent with this assertion. When NOArg was given concomitantly with morphine or pentazocine, versus L-Arc and opiates, the nitroderivative did not antagonize the effects of the high doses of morphine or pentazocine.

On the other hand, NO-Arc increased the effect of U50488H, which may reflect a lower dependence of the antinociceptive effect of U50488H on NO concentration in the nervous system. This ap- parent low dependence of U50488H on NO functions also has been reported by McDonald et al., (1994) who demonstrated that NO- Arc does not antagonize its analgesic effect.

As in the case of L-Arc, the influence of tolerance to morphine and pentazocine also was investigated in relation to NOArg- induced antinociception. Unlike L-Arc, the responses to the inhibi- tor of NO synthase were unaffected. Although morphine and pen- tazocine tolerance do not affect the antinociception of NO-Arc, its analgesic response does appear to depend on opiate mechanisms be- cause it is definitely antagonized by naloxone. Besides the present results, other studies have demonstrated that L-Arc and NOArg share some common pharmacologic responses and exhibit impor-

L-Arginine and NG-Nitroarginine Antinociception 447

300 - < (n uJ (.9 200

FIGURE6. Responses to L-Arg o-8 100 and NOArg in mice tolerant to < morphine (left panel) or pentazo- uJ cine (right panel). Solid squares, control responses to L-Arg and to NOArg. Open triangles, results ob- tained in morphine-tolerant mice; solid triangles, responses observed in pentazocine-tolerant animals. Mice were rendered tolerant by a single dose (1OO or 45 mg/kg sc) of mor- 300. phine or pentazocine, respectively, ,< administered 6 hr before the test doses of L-Arg or NOArg. All uJ (.9 200 groups consisted of 14 mice. *Sig- ~ :E nificant decrease of the mean value ~ gi compared with the control group, vJ (P<O.05). 100 O .8

< uJ

/ I I I

300 450 600 L-Arg (mglkg)

P NOArg (mglkg)

300t

200

~q 1°° t

0 -

300- < ~) lU (.9 200 ~ S

u~ ~ . 8 100

L-Arg mg/kg)

NOArg (mglkg)

tant differences. Whereas L-Arg increases the rate of tolerance to morphine (Babey et al., 1994), the NO synthase inhibitor NO-Arg partially antagonizes this process (Elliot et al., 1994). In contrast to these results, Dambisya and Lee (1996) have recently reported that NO synthase inhibitors enhance the development of morphine tol- erance and that L-Arg delays the development of the process.

The discreprancies found by different authors may be partly re- lated to differences in the doses used and the duration of treatments for example, Dambisya and Lee (1996) used increasing doses of mor- phine during 2 days, whereas we used a single dose of the analgesic. Moreover, the doses of L-Arg and NOArg administered in our ex- periments were also higher than those used by these authors.

The association of calcium channel antagonists with L-Arg or NO-Arg also induced some interesting differences with respect to the intensity of the analgesic responses to both drugs. L-Arg effects were antagonized, whereas the responses to NO-krg were potenti- ated. It seems reasonable to suggest that diltiazem- or verapamil- induced decrease of intracellular calcium may affect some critical event in intracellular mechanisms related to the analgesic response to L-Arg. According to Ueda et al., (1987), the process of kyotor- phin synthesis is largely independent of intracellular calcium con- centration. In contrast, the constitutive NO synthase requires cal- cium influx to cell, which, after binding to calmodulin, activates the enzyme (Bredt and Snyder, 1990). If it is accepted that the antinoci- ceptive effects of L-Arg are due to NO synthesis, then the calcium channel antagonists reduction of L-Arg analgesia could be ex- plained as the result of decreased calcium entry into the cell. Never- theless, a possible influence of calcium channel antagonists in an unidentified process of kyotorphin Met-enkephalin interaction can- not be excluded.

In summary, the partial antagonism of the antinociceptive effects of L-Arg by naloxone and the induction of an apparent cross-toler- ance with morphine and pentazocine suggest the existence of an

opioid mechanism in L-Arg-induced analgesia. The decrease of the antinociceptive responses to the aminoacid by calcium channel an- tagonists may be explained by assuming that, at least partly, L-Arg antinociception is induced through NO synthesis. On the other hand, NOArg does not affect the responses of morphine and pentaz- ocine to the same extent as does L-Arg, and opiate tolerance does not affect the NOArg response. However, its synergism with the cal- cium channel antagonists may reflect the existence of a molecular mechanism responsible for analgesia in a manner similar to that of several opiates.

This work was supported by FONDECYT grant 195 1053 and University of Concepci6n grant 93.33.70.1.

References Babey A.-M., Kolesnikov Y., Cheng J., Iturrisi C. E., Trifilletti R. R. and

Pastemak G. W. (1994) Nitric oxide and opioid tolerance Neuropharma- co/ogy 33, 1463-1470.

Bredt D. S. and Snyder S. H. (1990) Isolation of nitric oxide synthetase, a caldmodulin-requiring enzyme. Proc. Natl. Acad. Sci. USA 87,682-685.

Bredt D. S. and Snyder S. H. (1994) Nitric oxide: a physiologic messenger molecule. Ann. Rev. Biochem. 63, 175-195.

Contreras E., Tamayo L. and Amigo M. (1988) Calcium channel antagonists increase morphine induced analgesia and antagonize morphine toler- ance. Eur. J. Pharmacol. 148, 463~-66.

Contreras E., Quijada L., Germany A., Fleckenstein R. and Hern~ndez A. (1993) Calcium channel antagonists and adenosine analogues decrease tolerance to opiates pentazocine and U50488H. Gen. Pharmacol. 24, 1203-1206.

Dambisya Y. M. and Lee T.-L. (1996) Role of nitric oxide in the induction and expression of morphine tolerance and dependence in mice. Br. J. Pharmacol. 117, 914-918.

Dierssen M., Fl6rez, J. and Hurld M. A. (1990) Calcium channel modulation by dihydropyridines modifies sufentanil-induced antinociception in acute and tolerant conditions. Naunyn Schmiedebergs Arch. Pharmacol. 342, 559-565.

4 4 8 E. Contreras et al.

Eddy N. B. and Leimbach D.J. (1953) Synthetic analgesics I1 dithienylbu- tenyl and dithienylbutyl amines. J. Pharmacol. Exp. Ther. 107, 385-393.

Elliot K., Minami N., Kolesnikov Y., Pastemak G. W. and Iturrisi C. E. I. (1994) The NMDA receptor antagonists, LY2764614 and MK-801, and the nitric oxide synthase inhibitor N°-nitro-L-arginine, attenuate anal- gesic tolerance to the mu-opioid morphine but not to kappa-opioids, Pain 56, 69-75.

Haley J. E., Dickenson A. H. and Schachter M. (1992) Electrophysiological evidence for a role of nitric oxide in prolonged chemical nociception in the rat. Neuropharmacology 31, 251-258.

Harima A., Shimizu H., and Takagi H. (1991 ) Analgesic effect of L-arginine in patients with persistent pain. Eur. Neuropsychopharmacol. 1,529-533.

Hirsch D. B., Steiner J. P., Dawson T. M., Mammen A., Hayek E. and Sny- der S. H. (1993) Neurotransmitter release regulated by nitric oxide in PC-12 cells and brain synapsotomes. Current Biol. 3, 749-754.

Iwamoto E. T. and Marion L. (1994) Pharmacologic evidence that spinal muscarinic analgesia is mediated by an L-arginine/nitric oxide/cyclic GMP cascade in rats. J. Pharmacol. Exp. Ther. 271,601-608.

Kawabta A., Fukuzumi Y., Fukushima Y. and Takagi H. (1992) Antinocicep- rive effect of L-arginine on the carrageenin-induced hyperalgesia of the rat: possible involvement of central opiopeptidergic systems. Eur. J. Pharmacol. 218, 153-158.

Kawabata A., Umeda N. and Takagi H. (1983) L-arginine exerts a dual role in nociception processing in the brain: involvement of the kyotorphin- met-enkephalin pathway and NO-cyclic GMP pathway. Br. J. Pharma- col. 109, 73 79.

Levy R. A., Proudfit H. K. and Goldstein B. D. (1983) Antinociception fol- lowing microinjection of dibutyryl cyclic nucleotides into the caudal re- ticular formation and periaqueductal grey of the rat brain. Pharmacol. Biochem. Behav. 19, 79-84.

McDonald C. E., Gagnon M. J., Ellenberger E. A., Hodges B. L., Ream J. K., Tousman S. A. and Quock R. M. (1994) lnhibitors of nitric oxide syn- thesis antagonize nitrous oxide antinociception in mice and rats. J. Phar- macol. Exp. Ther. 269, 601-608.

Moore P.-K, Oluyoni A.O., Babbedge R.C., Wallace P. and Hart S. L. ( 1991 ) L-N°-nitroarginine methyl ester exhibits antinociceptive activity in the mouse. Br. J. Pharmacol. 102, 192-202.

Morris R., Southan E., Braid D. J. and Garthwaite J. (1992) Nitric oxide may act as a messenger between dorsal root ganglia neurones and their satel- lite cells. Neurosci. Lett. 137, 29-32.

Takagi H., Shiomi H, Ueda H. and Amano H. (1979) A novel analgesic di- peptide from bovine brain is a possible Me-enkephalin releaser. Nature 282, 410~rl 2.

Tseng L. F., Xu J. Y. and Pieper G. M. (1992) Increase of nitric oxide produc- tion by L-arginine potentiates i.c.v, administered [3-endorphin-induced antinociception in mice. Eur. J. Pharmacol. 21,301-303.

Ueda H., Yoshihara Y., Fukushima N., Shiomi H., Nakamura A. and Takagi H. (1997) Kyotorphin (tyrosyl-arginine) synthetase in rat brain synapto- somes. J. Biol. Chem. 262, 8165-8173.

Xu J. Y. and Tseng L. F. (1993) Increase of nitric oxide by L-arginine po- tentiates [3-endorphin but not m, d or k-opioid agonist-induced antinoc- iception in the mouse. Eur. J. Pharmacol. 236, 137-142.