- Email: [email protected]
Effects of the novel analgesic, cizolirtine, in a rat model of neuropathic pain
Pain 104 (2003) 169–177 www.elsevier.com/locate/pain
Effects of the novel analgesic, cizolirtine, in a rat model of neuropathic pain Vale´rie Kaysera,*, Antonio Farre´b, Michel Hamona, Sylvie Bourgoina a
INSERM U288, NeuroPsychoPharmacologie Mole´culaire, Cellulaire et Fonctionnelle, Faculte´ de Me´decine Pitie´-Salpeˆtrie`re, 91, Boulevard de l’Hoˆpital, 75634 Paris Cedex 13, France b Laboratorios Dr Esteve, S.A., Avenue Mare de De´u de Montserrat, 221, 08041 Barcelona, Spain Received 5 September 2002; accepted 18 December 2002
Abstract Cizolirtine (5-{[(N,N-dimethylaminoethoxy)phenyl]methyl}-1-methyl-1H-pyrazol citrate) is a centrally acting analgesic with a currently unknown mechanism of action, whose efficacy has been demonstrated in various models of acute and inflammatory pain in rodents. Further studies were performed in order to assess its potential antinociceptive action in a well-validated model of neuropathic pain, i.e. that produced by unilateral sciatic nerve constriction in rats. Animals were subjected to relevant behavioural tests based on mechanical (vocalization threshold to paw pressure) and thermal (struggle latency to paw immersion in a cold (108C) water bath) stimuli, 2 weeks after sciatic nerve constriction, when pain-related behaviour was fully developed. Acute pretreatment with 2.5– 10 mg/kg p.o. of cizolirtine reversed both mechanical and thermal allodynia. These effects were antagonized by prior injection of the a2-adrenoceptor antagonist idazoxan (0.5 mg/kg i.v.), but not the opioid receptor antagonist naloxone (0.1 mg/kg i.v.). On the other hand, cizolirtine (10 mg/kg p.o.) produced no motor deficits in animals using the rotarod test. Our study showed that cizolirtine suppressed pain-related behavioural responses to mechanical and cold stimuli in neuropathic rats, probably via an a2-adrenoceptor-dependent mechanism. These results suggest that cizolirtine may be useful for alleviating some neuropathic somatosensory disorders, in particular cold allodynia, with a reduced risk of undesirable side effects. q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. Keywords: Cizolirtine; Idazoxan; Naloxone; Antinociception; Neuropathic pain
1. Introduction Cizolirtine (5-{[(N,N-dimethylaminoethoxy)phenyl]methyl}-1-methyl-1H-pyrazol citrate) is a centrally acting analgesic with a currently unknown mechanism of action (Alvarez et al., 2000; Monck, 2001). This compound is characterized by a complete lack of anti-inflammatory properties and shows no affinity for m, d or k opioid receptors (Alvarez et al., 2000). Nevertheless, the antinociceptive efficacy of cizolirtine has been demonstrated in various animal models of acute and inflammatory pain (Alvarez et al., 2000), thereby suggesting that this drug may be an alternative to opioids for alleviating such pain (Matthew et al., 2000). Interestingly, blockade of a2adrenoceptors by idazoxan significantly reduced cizolirtineinduced analgesia, whereas, in contrast, pretreatment with * Corresponding author. Tel.: þ 33-1-4077-9709; fax: þ33-1-4077-9790. E-mail address: [email protected] (V. Kayser).
desipramine, a selective inhibitor of noradrenaline (NA) reuptake, potentiated the cizolirtine action indicating its mediation through adrenoceptor-dependent mechanisms (Alvarez et al., 2000). However, cizolirtine has no affinity for adrenoceptors and NA transporters, suggesting an indirect activation of NA neurotransmission by the drug. Another notable effect of this novel analgesic drug is its inhibitory influence on the release of substance P and calcitonin gene-related peptide at the spinal level in rats (Ballet et al., 2001). This effect can also be prevented by pretreatment with idazoxan, further emphasizing the implication of a noradrenergic link in the central action of cizolirtine (Ballet et al., 2001). Because previous studies showed that activation of a2adrenoceptors at the spinal level exerts some antinociceptive effects in neuropathic pain (Yaksh, 1999), we addressed the question of whether cizolirtine could also be efficient to alleviate this type of pain. Indeed, a preliminary study in a
0304-3959/03/$30.00 q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. doi:10.1016/S0304-3959(02)00497-9
170
V. Kayser et al. / Pain 104 (2003) 169–177
small number of patients suffering from traumatic neuropathic pain showed that a dose of 115 mg of cizolirtine twice a day for 21 days yielded a marked improvement in a subgroup of five patients with primary allodynia (Shembalkar et al., 2001). In the present study, we evaluated the antinociceptive action of cizolirtine in a well-established model of neuropathic pain, i.e. that which consists of a unilateral chronic constriction injury of the sciatic nerve in rats (Bennett and Xie, 1988). A behavioural test (vocalization threshold to paw pressure), that allowed clear-cut demonstration of the antinociceptive activity of a2-adrenoceptor agonists, antidepressants, sympatholytic and anticonvulsant drugs (Kayser and Christensen, 2000), was used to assess the effect of acute treatment with cizolirtine. In addition, because cold allodynia is a characteristic symptom of neuropathic pain (Frost et al., 1988), we also determined the effect of cizolirtine on the latency of struggle response to paw immersion in a non-noxious cold (108C) water bath (Attal et al., 1990). Finally, we also tested whether a2adrenoceptor blockade by idazoxan could prevent the effects of cizolirtine in this model of neuropathic pain, as previously observed in relevant models of inflammatory pain (Alvarez et al., 2000).
2. Materials and methods 2.1. Animals Experiments were performed on male Sprague –Dawley rats (300 – 350 g; Charles River, St-Aubin-le`s-Elbeuf, France) housed in groups of five per cage. Rats were maintained on a 12/12 h light/dark cycle at an ambient temperature of 21 ^ 18C, and were allowed free access to food and water. Experiments were carried out in accordance with both the European Community Council Directive (86/609/EEC) and the Ministry of Agriculture regulations. In addition, we adhered to the guidelines of the Committee For Research and Ethical Issues of the International Association for the Study of Pain (1983). 2.2. Surgical procedure Sciatic nerve injury was made following a procedure (Attal et al., 1990) adapted from Bennett and Xie (1988). Briefly, rats were anesthetized with pentobarbital (50 mg/ kg, i.p.) and external area of the right thigh was shaved and swabbed with 70% ethanol. A skin incision was made in the cleaned area, and the sciatic nerve was exposed following blunt dissection and gently freed from adhering tissues. Four ligatures (5-0 chromic catgut, applied 1– 2 mm apart) were then placed around the nerve. Care was taken to ensure that the ligatures were not too tight so as to occlude perineural blood flow. Finally, the separated muscle was stitched and the incision was closed with silk thread. In sham-operated
control animals, identical surgical procedures were employed, but the sciatic nerve was left intact. 2.3. Drugs and doses Cizolirtine (Esteve Labs, Barcelona, Spain) was given orally through a feeding cannula in a volume of 10 ml of water. The doses, 2.5, 5 and 10 mg/kg, were chosen on the basis of previous studies in rodents (Alvarez et al., 2000). The a2-adrenoceptor antagonist, idazoxan (Res. Biochem. Int., Natick, MA, USA), was injected in saline (0.9% NaCl, 0.1 ml) into a lateral tail vein 30 min before cizolirtine administration (10 mg/kg per os). The selected dose of idazoxan, 0.5 mg/kg i.v., was previously shown to completely prevent the antinociceptive effect of the prototypical a2-adrenoceptor agonist clonidine, without affecting, on its own, the nociceptive threshold in this model of neuropathic pain (Kayser et al., 1995). Naloxone (Res. Biochem. Int.) was injected in saline 15 min before cizolirtine (10 mg/kg per os) at a dose of 0.1 mg/kg i.v., that was previously shown to completely prevent the antinociceptive effect of a maximally effective dose (1 mg/kg i.v.) of morphine in the same model (Christensen et al., 1998). Paired control rats were treated with water (10 ml p.o. per rat) and/or saline (0.1 ml i.v. per rat) using the same routes of administration. 2.4. Behavioural testing Animals were fasted for 16 h before testing, but water was always provided ad libitum. Testing sessions, beginning at 9:30 AM, were conducted in a quiet room. The experimenter was blind to the drug and the dose tested. Rats were not acclimatized to the test situations beforehand. Each animal received either cizolirtine, idazoxan, naloxone or drug combinations only once and was used in only one experiment. 2.4.1. Mechanical test The vocalization thresholds, expressed as grams, were determined according to a modification of the Randall and Selitto method (Kayser and Christensen, 2000). An increasing pressure was applied through a dome-shaped plastic tip (diameter ¼ 1 mm) onto the dorsal surface of the hindpaw using a Ugo Basile analgesimeter (Comerio, Italy). The tip was positioned between the third and fourth metatarsus (into the sciatic nerve territory) and pressure was applied until the rat squeaked. This response represents a more integrated nociceptive behaviour than paw withdrawal and allows reliable assessment of the respective potencies of various antinociceptive drugs in this pain model (Kayser and Christensen, 2000). Preoperative vocalization threshold (mean of two consecutive stable thresholds) was determined for both hindpaws in each rat. Then, 2 weeks after surgery, when the abnormal pain behaviour was at its maximum (Attal et al., 1990), a control threshold (mean of two consecutive stable
V. Kayser et al. / Pain 104 (2003) 169–177
thresholds) was determined before administration of the drug. Vocalization thresholds were then measured every 10 min, until they returned to baseline. 2.4.2. Cold test The struggle latencies, expressed as seconds, were determined after immersion of the hindpaw into a 108C water bath (Ministat MHUB 11; Bioblock Scientific, Illkirch, France) as described previously (Attal et al., 1990). The rat was gently held by the trunk, leaving the limbs freely exposed. The paw was then lowered into the water bath until the rat exhibited a struggle reaction, or for a maximum of 15 s when the rat did not exhibit such a reaction. Preoperative latency (mean of two consecutive stable trials 20 min apart) was determined for both hindpaws in each rat. Then, 2 weeks after surgery, a control latency (mean of two consecutive stable latencies) was determined before administration of the drug. Struggle latencies were then measured every 20 min, until they returned to baseline. The 20-min interval between successive measurements was necessary because abnormal reactions lasting for more than 15 min after a thermal stimulus have been reported in this model (Attal et al., 1990). 2.4.3. Rotarod test In a separate group of neuropathic rats ðn ¼ 8Þ, locomotor function was tested using the Ugo Basile accelerating rotarod (model 7750 for rats). This apparatus consists of a base platform and a rotating horizontal rod (7 cm in diameter, 50 cm in length) with a non-skid surface. The rod is divided by five disks into four sections of equal length in which four rats can be tested simultaneously. Under each drum section, 26 cm below the rod on the platform, a V-shaped counter-trip plate is positioned. Animals were acclimatized to the revolving drum and habituated to handling in order to avoid stress during testing. The rod was set to accelerate from 4 to 40 rpm in a 5-min period. The integrity of motor coordination was assessed as the performance time on the rod measured from the start of acceleration until the animal fell from the drum onto the counter-trip plate. Rats were acclimatized to acceleration by three training runs. The mean of the performance time determined at the fourth and fifth training runs served as the reference control value (expressed as seconds). The performance time was then measured every 20 min for a total of 80 min after per os administration of cizolirtine or water, and compared with the control value. 2.5. Measurement of skin temperature of the rat paw Because of the possible occurrence of interferences between the pharmacological effects of drugs and adaptive thermoregulation in the thermal test (Le Bars et al., 2001), we measured skin temperature at the level of hindpaws in neuropathic rats ðn ¼ 12Þ before and after cizolirtine
171
(10 mg/kg p.o.) or vehicle (water) administration. Measurements were made using the thermocouple thermometer Digisense (15 mm in diameter; Model N8 8528-10; Cole – Parmer Instrument Co., Chicago, IL) in a quiet room maintained at constant temperature (22 – 238C). The rats were left in their cage and the thermocouple was gently placed on the lateral side of the plantar surface of the hindpaw, in the territory of the sciatic nerve. Stable temperature readings were obtained after 10 s with a precision of 0.18C. For each rat, two consecutive basal measurements were averaged and the temperature was then measured successively on the contralateral and the nerveinjured side every 20 min after drug or vehicle administration for a total of 240 min. 2.6. Statistical analyses All data are expressed as means ^ SEM. A paired Student’s t-test was employed to compare the mechanical thresholds and cold water response latencies before and after the surgery. The overall effects of treatment (areas under the curve, AUCs) were calculated by use of the trapezoidal rule. Statistical significance of the data was analyzed by one-way analysis of variance (ANOVA). The observed significances ðP , 0:05Þ were then confirmed with the Tukey’s test.
3. Results 3.1. Responses to mechanical and thermal stimuli in sciatic nerve-ligatured and sham-operated rats 3.1.1. Mechanical test In agreement with previous observations (Christensen and Kayser, 2000), the vocalization thresholds to paw pressure were markedly decreased 2 weeks after the sciatic nerve ligature (Table 1). At this time, the mean threshold for the nerve-injured paw was 196 ^ 7 g (P , 0:001 vs. the preconstriction value 303 ^ 12 g, n ¼ 72). The vocalization threshold to paw pressure of the contralateral paw was also diminished, but to a minor extent (265 ^ 7 g vs. the preconstriction value 304 ^ 8 g, P , 0:01, n ¼ 72). In contrast, the vocalization thresholds of the sham-operated rats did not differ from those measured in intact animals (302 ^ 13 g for the ipsilateral paw; 303 ^ 7 g for the contralateral paw, n ¼ 72) (Table 1). 3.1.2. Cold test In further agreement with previous studies (Christensen and Kayser, 2000), struggle latencies to cold stimulus applied to the nerve-injured paw were decreased by several seconds, 2 weeks after the nerve ligature (Table 2). At this time, the mean struggle latency at 108C was 7.0 ^ 0.3 s (P , 0:001 vs. the preconstriction value 14.6 ^ 0.8 s, n ¼ 72). In contrast, the response latencies in sham-
172
V. Kayser et al. / Pain 104 (2003) 169–177
Table 1 Maximal vocalization thresholds for the nerve-injured and contralateral paws in sciatic nerve-ligatured rats, and for both hindpaws in sham-operated rats, in the paw pressure test before and after administration (adm.) of vehicle, cizolirtine (2.5–10 mg/kg p.o.) and/or idazoxan (0.5 mg/kg i.v.) or naloxone (0.1 mg/kg i.v.) Treatment adm.
Vehicle (water) Cizolirtine 2.5 Cizolirtine 5 Cizolirtine 10 Idazoxan 0.5 Idazoxan 0.5 þ cizolirtine 10 Naloxone 0.1 Naloxone 0.1 þ cizolirtine 10
Sciatic nerve-ligatured rats
Sham-operated rats (right and left pawsa)
Nerve-injured right paw
Contralateral left paw
Before adm.
After adm.
Before adm.
After adm.
194 ^ 7 186 ^ 6 192 ^ 8 204 ^ 8 200 ^ 9 198 ^ 5 195 ^ 6 201 ^ 8
201 ^ 6 270 ^ 12b 293 ^ 8b 306 ^ 7b 210 ^ 11 206 ^ 7c 201 ^ 8 315 ^ 12b
266 ^ 6 268 ^ 5 274 ^ 6 255 ^ 5 270 ^ 10 265 ^ 7 256 ^ 8 270 ^ 7
270 ^ 6 343 ^ 12b 341 ^ 13b 309 ^ 15b 282 ^ 14 269 ^ 15c 265 ^ 13 340 ^ 15b
Before adm.
After adm.
300 ^ 10 290 ^ 18 294 ^ 18 312 ^ 12 310 ^ 17 300 ^ 12 308 ^ 10 305 ^ 11
310 ^ 10 378 ^ 6b 369 ^ 17b 414 ^ 21b 312 ^ 20 314 ^ 11c 312 ^ 18 395 ^ 6b
Results are expressed as grams. Each value is the mean ^ SEM of nine independent determinations. The mean preoperative thresholds were 303 ^ 8 and 304 ^ 9 g for the right and the left hindpaw, respectively (means ^ SEM, n ¼ 144). a Values for the right and left paws did not differ and were pooled for the sake of clarity. b P , 0:01 vs. before administration (Tukey’s test). c P , 0:01 vs. cizolirtine (10 mg/kg p.o.) alone.
operated rats did not differ from those measured in intact animals (13.5 ^ 1.2 s, n ¼ 10). 3.2. Antinociceptive effects of cizolirtine 3.2.1. Mechanical test Cizolirtine produced significant increases in vocalization thresholds for the nerve-injured paw at all doses tested (Table 1, Fig. 1). Indeed, within 10– 40 min following administration of the two highest doses, vocalization thresholds raised up to values close to those found in intact
animals. Thereafter, the effect of cizolirtine progressively vanished, and vocalization thresholds decreased back to predrug values of 40 – 60 min after administration of 2.5, 5 or 10 mg/kg of the compound, respectively (Fig. 1). In the contralateral paw, cizolirtine also produced an increase in the vocalization thresholds that lasted for , 30 min (Table 1). However, the overall effect (AUC) of
Table 2 Maximal struggle latencies for the nerve-injured paw in sciatic nerveligatured rats, in the cold test before and after administration (adm.) of vehicle, cizolirtine (2.5–10 mg/kg p.o.) and/or idazoxan (0.5 mg/kg i.v.) or naloxone (0.1 mg/kg i.v.) Treatment
Vehicle (water) Cizolirtine 2.5 Cizolirtine 5 Cizolirtine 10 Idazoxan 0.5 Idazoxan 0.5 þ cizolirtine 10 Naloxone 0.1 Naloxone 0.1 þ cizolirtine 10
Nerve-injured paw Before adm.
After adm.
7.3 ^ 0.5 7.4 ^ 0.2 7.1 ^ 0.4 6.7 ^ 0.3 6.7 ^ 0.2 7.0 ^ 0.5 6.8 ^ 0.5 7.0 ^ 0.3
7.6 ^ 0.7 8.0 ^ 0.9 13.1 ^ 0.6a 13.4 ^ 1.0a 7.1 ^ 0.9 7.2 ^ 0.7b 7.6 ^ 0.9 13.8 ^ 0.9a
Results (in seconds) are expressed as means ^ SEM of nine independent determinations.
The mean preoperative latency was 14.6 ^ 0.8 s for the right hindpaw (mean ^ SEM, n ¼ 72). a b
P , 0:01 vs. before administration (Tukey’s test). P , 0:01 vs. cizolirtine (10 mg/kg p.o.) alone.
Fig. 1. Time-course changes in the vocalization threshold in the mechanical pressure test applied to the nerve-injured hindpaw in sciatic nerve-ligatured rats. Effects of cizolirtine. Cizolirtine (2.5, 5 or 10 mg/kg) or vehicle (water) was administered per os at time 0, and vocalization threshold (in grams) was measured every 10 min for up to 60–70 min thereafter. Each value is the mean ^ SEM of nine independent determinations (one determination per rat). Values at time 0 corresponded to vocalization thresholds in untreated sciatic nerve-ligatured rats. The horizontal grey band indicates the range of vocalization threshold values for unoperated rats (303 ^ 12 g, n ¼ 72). *P , 0:05, **P , 0:01 vs. vehicle-treated nerve-injured rats (Tukey’s test).
V. Kayser et al. / Pain 104 (2003) 169–177
cizolirtine was significantly larger on the nerve-injured paw than on the contralateral paw (P , 0:05 for 2.5, P , 0:01 for 5 or 10 mg/kg of the compound, respectively). In sham-operated rats, cizolirtine significantly increased the vocalization thresholds to paw pressure for either the ipsilateral or the contralateral hindpaw (Table 1). At the three doses tested, similar effects were noted for both hindpaws. Comparison of overall effects (AUC) showed that cizolirtine-induced increase in vocalization thresholds in sham-operated rats did not differ ðP ¼ 0:16Þ from that found for the contralateral hindpaw in sciatic nerveligatured rats. In contrast to cizolirtine, vehicle (water) administration did not affect vocalization thresholds in both sciatic nerveligatured and sham-operated rats (Table 1, Fig. 1). 3.2.2. Cold test Vehicle or 2.5 mg/kg of cizolirtine had no effect in sciatic nerve-ligatured rats (Table 2, Fig. 2). By contrast, the other two doses of cizolirtine produced a significant increase in the struggle latency (Table 2, Fig. 2). Thus, a doubling in the latency was noted 20 min after administration of the drug at 5 mg/kg. The effect then declined progressively to reach pre-drug values 120 min after the treatment (Fig. 2). A more prolonged effect was seen with 10 mg/kg of cizolirtine. However, at its maximum (60 – 80 min after administration), the effect caused by 10 mg/kg of cizolirtine was not higher than that observed only after 5 mg/kg of the drug (Table 2, Fig. 2). Nevertheless, significant differences were noted between the two doses because the effect of
173
cizolirtine was clearly dose-dependent with regard to the duration of the effect (r ¼ 0:83, P , 0:001) (Fig. 2). Neither vehicle nor cizolirtine (2.5 – 10 mg/kg) significantly affected struggle latencies in sham-operated rats (not shown). 3.3. Effects of idazoxan and naloxone on the antinociceptive effects of cizolirtine On their own, neither idazoxan (0.5 mg/kg i.v.) nor naloxone (0.1 mg/kg i.v.) affected the vocalization thresholds to mechanical stimulation as well as the struggle latencies in the cold test applied to the injured and the contralateral sides in both sciatic nerve-ligatured and shamoperated rats (Tables 1 and 2). However, pretreatment with idazoxan completely prevented the increases in vocalization thresholds and struggle latencies normally observed after per os administration of cizolirtine at 10 mg/kg. In contrast, naloxone pretreatment did not modify these effects of cizolirtine (Tables 1 and 2). 3.4. Lack of effect of cizolirtine on the rotarod performance time Like its vehicle (water), cizolirtine at the highest dose tested, 10 mg/kg p.o., did not significantly alter the rotarod performance time for at least 80 min after its administration in sciatic nerve-ligatured rats (Table 3). 3.5. Lack of effect of cizolirtine on hindpaw skin temperature Hindpaw skin temperature was measured at various times (20 – 240 min) after per os administration of cizolirtine (2.5 – 10 mg/kg) or water to sciatic nerve-ligatured rats maintained at an ambient temperature of 22– 238C. At all times studied, skin temperature did not significantly differ between rats having received either cizolirtine or water p.o. All along the experiment, the skin temperature of the nerveinjured hindpaw remained stable at 26.7 ^ 1.38C and that of the contralateral hindpaw at 26.5 ^ 1.48C (means ^ SEM, n ¼ 12), in sciatic nerve-ligatured rats treated with cizolirtine or vehicle.
4. Discussion Fig. 2. Time-course changes in the struggle latency in the cold (108C) test applied to the nerve-injured hindpaw in sciatic nerve-ligatured rats. Effects of cizolirtine. Cizolirtine (2.5, 5 or 10 mg/kg) or vehicle (water) was administered per os at time 0, and struggle latency (in seconds) was measured every 20 min for up to 120–220 min thereafter. Each value is the mean ^ SEM of nine independent determinations (one determination per rat). Values at time 0 corresponded to struggle latencies in untreated sciatic nerve-ligatured rats. The horizontal grey band indicates the range of struggle latency values for unoperated rats (14.6 ^ 0.8 s, n ¼ 72). *P , 0:05, **P , 0:01 vs. respective 0 time values (Tukey’s test).
Rats with a unilateral constriction injury of the sciatic nerve clearly exhibited abnormal pain sensitivity, with decreased vocalization thresholds to mechanical stimuli and decreased struggle latencies in response to 108C cold stimuli, as already shown in earlier studies (Kayser and Christensen, 2000). Peroral administration of cizolirtine was found to inhibit pain-related behaviours associated with neuropathic injury after the neuropathy had been established. Interestingly, after p.o. administration of [14C]cizo-
174
V. Kayser et al. / Pain 104 (2003) 169–177
Table 3 Rotarod performance time before and after p.o. administration of vehicle or cizolirtine (10 mg/kg) in sciatic nerve-ligatured rats Treatment
Vehicle (water) Cizolirtine
Before administration
108 ^ 10 116 ^ 15
Post-administration time (min) 20
40
60
80
118 ^ 2 111 ^ 24
113 ^ 10 115 ^ 23
109 ^ 24 119 ^ 24
105 ^ 29 119 ^ 18
Results (in seconds) are expressed as means ^ SEM of four independent determinations. None of the values after cizolirtine treatment were significantly different ðP . 0:05Þ from those in vehicle-treated rats and from pretreatment values.
lirtine within the same dose range as that chosen in our study, peak plasma concentrations of radioactivity are reached within 20 min (Martinez et al., 1999), which is in accordance with the time of the peak effects that we observed in both mechanical and thermal tests.
result is consistent with a recent report showing that cizolirtine inhibits hyperalgesia in rat models of inflammatory pain, without altering motor coordination or exerting sedative effects (Alvarez et al., 2000).
4.1. Effects of cizolirtine on mechanical pain-related behavioural response
4.2. Effects of cizolirtine on thermal pain-related behaviours
Using a centrally integrated test, the measure of the vocalization threshold to paw pressure, we showed that cizolirtine reversed the abnormal reactivity to mechanical stimuli in rats with peripheral mononeuropathy. Indeed, at the doses of 5 and 10 mg/kg p.o., cizolirtine increased the threshold for the nerve injured paw up to a value not different from that found in both sham-operated and intact animals. This effect was close to the one previously found in the same model of neuropathic pain after an i.v. injection of 0.3 mg/kg of morphine (Christensen et al., 1998) or 0.1 mg/ kg of the a2-adrenoceptor agonist clonidine (Kayser et al., 1995). The effect of cizolirtine was observed not only on the nerve-injured side, but also, although to a significantly lower extent, on the contralateral side. At the same doses, 2.5 –10 mg/kg p.o., cizolirtine also produced a significant antinociceptive effect in sham-operated animals, which was roughly comparable in magnitude with that observed for the contralateral hindpaw in sciatic nerve-ligatured rats. Interestingly, the effects of cizolirtine were prevented by i.v. pretreatment with the a2-adrenoceptor antagonist idazoxan, suggesting their mediation through mechanisms related to those triggered by direct and indirect a2-adrenoreceptor agonists such as clonidine (Kayser et al., 1995) and S12813-4 (Kayser et al., 1992), respectively. It has to be emphasized that idazoxan had, on its own, no significant hyperalgesic effect, thereby suggesting that a2-adrenoceptor-mediated inhibitory control of pain signal transmission is probably not tonically active in this neuropathic pain model (see also Kayser et al., 1995). While a2-adrenoceptor agonists are known to exert analgesic effects, they also have accompanying effects such as sedation and hypotension (Puig et al., 2000). Our experiments showed that cizolirtine, in the same dose range as that used to alleviate pain in neuropathic rats, did not decrease the performance time in the rotarod test. This
Another interesting observation in this study relates to the potent effect of cizolirtine on cold allodynia (i.e. abnormal reactions elicited by paw immersion in water at 108C). Cold allodynia is a characteristic symptom of neuropathic pain and is particularly resistant to pharmacological interventions in the present pain model (Kayser et al., 1995; Ida¨ npa¨ a¨ n-Heikkila¨ et al., 1997; Jasmin et al., 1998; Kayser and Christensen, 2000). However, cizolirtine was found to produce a marked, dose-dependent effect in the cold test, which lasted much longer than its effect in the mechanical test. While the lowest dose, 2.5 mg/kg p.o., of cizolirtine was essentially inactive, 5 mg/kg p.o. of the drug almost doubled the struggle latency in response to paw immersion in a 108C water bath. The highest dose tested, 10 mg/kg p.o., exerted the same effect but for a longer time compared with 5 mg/kg of cizolirtine (Fig. 2). This effect of cizolirtine (10 mg/ kg) was prevented by pretreatment with idazoxan, which confirmed its mediation through a2-adrenoceptors, like that already noted about cizolirtine-induced alleviation of pain caused by mechanical stimulation. Because changes in skin temperature may influence the response of rats to thermal stimuli (Hole and Tjølsen, 1993; Le Bars et al., 2001), we measured this parameter from 20 to 240 min after cizolirtine administration. Indeed, previous studies in the hot-plate test showed that an increase in skin temperature could lead to apparent hyperalgesia, whereas a reduction in skin temperature could lead to apparent analgesia (Tjølsen et al., 1991; Hole and Tjølsen, 1993). In the present study, we found that cizolirtine did not cause any significant change in paw skin temperature, which indicated that the effect of cizolirtine in the cold test could not be accounted for by some interfering alteration in peripheral thermoregulatory mechanisms.
V. Kayser et al. / Pain 104 (2003) 169–177
4.3. Possible mechanisms underlying the effects of cizolirtine in neuropathic rats We previously found that morphine and selective opioid receptor agonists injected i.v. are unable to reduce cold allodynia in neuropathic rats (Ida¨ npa¨ a¨ n-Heikkila¨ et al., 1997). In contrast, cizolirtine did reduce allodynia-related response in the cold test indicating that its action does not involve opioid-dependent mechanisms. Accordingly, we demonstrated that pretreatment with the opioid receptor antagonist naloxone (0.1 mg/kg i.v.) did not prevent the increases in vocalization thresholds and struggle latencies normally observed after per os administration of cizolirtine at 10 mg/kg. In line with these data, a recent report showed that this compound is not recognized by m, d or k opioid receptors in binding assays (Alvarez et al., 2000). Even though cizolirtine has no affinity for a2-adrenoceptors (Alvarez et al., 2000), it might activate descending noradrenergic pathways which are known to participate in an inhibitory control of nociception (Fairbanks et al., 2002). Previous studies demonstrated that the blockade of a2adrenoceptors by idazoxan inhibits the effects of cizolirtine (Alvarez et al., 2000; Ballet et al., 2001) and further supports to this observation were obtained in our own studies in sciatic nerve-ligatured rats. Conversely, the potentiation of noradrenergic neurotransmission by desipramine, probably through a facilitation of the agonistic action of endogenous NA at a2-adrenoceptors, results in an increased antinociceptive potency of cizolirtine (Alvarez et al., 2000). Like that caused by cizolirtine, there is evidence that analgesia due to opioids (Suh et al., 1989; Hylden et al., 1991) and nicotine (Rogers and Iwamoto, 1993), is partly mediated through a2adrenoceptor-mediated mechanisms. However, data obtained in the cold test clearly showed that cizolirtine exerted specific effects independent of opioid-related mechanisms. In this context, attention has to be paid to one of the main neurotransmitters of primary afferent fibers, i.e. substance P (Ho¨ kfelt et al., 1977). Substance P has been shown to play a key role in cold thermoreception and thermal allodynia through the stimulation of neurokinin1 (NK1) receptors in the superficial laminae of the dorsal spinal cord (Doyle and Hunt, 1999; McLeod et al., 1999). Furthermore, specific interactions between substance P and NA with regard to thermal hyperalgesia were recently evidenced in mutant mice lacking the gene coding for dopamine b-hydroxylase, the enzyme responsible for synthesis of NA from dopamine (Jasmin et al., 2002). Interestingly, these mice were found to exhibit a substance P-mediated chronic hyperalgesia to thermal, but not mechanical stimuli, which could be counteracted by restoring the synthesis of NA (Jasmin et al., 2002), and consequently, the activation of the pain controlling spinal a2c-adrenoceptors (Fairbanks et al., 2002). Because both an upregulation of NK1 receptors in the dorsal horn (Aanonsen et al., 1992; Malmberg and Basbaum, 1998; Cruce et al., 2001) and an increased production of substance P by primary
175
afferent neurons (Marchand et al., 1994; Noguchi et al., 1994) do occur in rats with neuropathic pain such as those bearing chronic ligatures of the sciatic nerve which were used herein, it can be proposed that thermal allodynia in these animals in fact results from a tonic activation of substance Pergic neurotransmission at the spinal level. Accordingly, cizolirtine would act through an a2c-adrenoceptor-dependent mechanism to reduce excessive substance P-ergic neurotransmission and the associated thermal allodynia. In line with this hypothesis, cizolirtine, like direct or indirect a2adrenoceptor agonists (Bourgoin et al., 1993; Collin et al., 1994), was recently found to exert an inhibitory influence on the spinal release of substance P in rats (Ballet et al., 2001). Furthermore, this effect could be prevented by the a2adrenoceptor antagonist idazoxan at the same dose as that used herein to normalize the behavioural response to paw immersion in water at 108C (B. Aubel, personal communication). Altogether, these data suggest that the mechanism of action of cizolirtine may resemble that of clonidine whose effects are also prevented by a2-adrenoceptor blockade (Tjølsen et al., 1990) and which inhibits the spinal release of substance P (Bourgoin et al., 1993; Collin et al., 1994). Indeed, previous studies also showed that clonidine reduces mechanical and cold allodynia-like behaviours as efficiently as cizolirtine in sciatic nerve-ligatured rats (Kayser et al., 1995). However, in contrast to clonidine, cizolirtine does not bind to a2-adrenoceptors (Alvarez et al., 2000), and further investigations are needed to assess whether cizolirtine, as expected from the aforementioned hypothesis, actually activates descending noradrenergic projections within the dorsal horn of the spinal cord. In addition to a2-adrenoceptors, imidazoline1 (I1) receptors have been proposed to contribute to the antinociceptive – antiallodynic effects of agonists such as clonidine and moxonidine, especially because (i) these drugs act as mixed a2 – I1 agonists (Fairbanks and Wilcox, 1999), and (ii) the a2-adrenoceptor antagonist most often used to prevent their effects, idazoxan, is also an I1 receptor antagonist (Bousquet et al., 2000). In the same line, one can ask whether the inhibitory effect of idazoxan on cizolirtineinduced antinociception might involve I1 receptors. In any case, the effects of cizolirtine cannot be ascribed to a direct stimulation of I1 receptors because this drug, at least up to 10 mM, does not bind to these receptors (J.A. GarciaSevilla, personal communication). However, it cannot be excluded that cizolirtine may induce the release of endogenous ligand(s) acting at I1 receptors, thereby producing at least part of its antinociceptive effects. Further studies with antagonists acting selectively at a2-adrenoceptors or I1 receptors are needed to definitively answer the question of the respective contributions of these receptors in the antinociceptive effects of cizolirtine. To conclude, it has to be emphasized that at the doses used to produce antinociceptive effects, cizolirtine, in contrast to direct a2-adrenoceptor agonists such as clonidine
176
V. Kayser et al. / Pain 104 (2003) 169–177
(Puig et al., 2000), exerted neither sedation nor motor disturbances. Accordingly, cizolirtine appears as a promising compound for the treatment of neuropathic pain, especially to reduce cold allodynia, with a low risk of undesirable side effects.
Acknowledgements This research was supported by grants from INSERM and Laboratorios Dr Esteve S.A.
References Aanonsen L, Kajander KC, Bennett GJ, Seybold VS. Autoradiographic analysis of 125I-substance P binding in rat spinal cord following chronic constriction injury of the sciatic nerve. Brain Res 1992;596:259 –68. Alvarez I, Andreu F, Buxens J, Colombo M, Dordal A, Fort M, Gutierrez B, Farre´ AJ. Pharmacology of cizolirtine: a new analgesic agent. Methods Find Exp Clin Pharmacol 2000;22:211–21. Attal N, Jazat F, Kayser V, Guilbaud G. Further evidence for a “painrelated” behaviour in a model of unilateral peripheral mononeuropathy. Pain 1990;41:235 –51. Ballet S, Aubel B, Mauborgne A, Polie´ nor A, Farre´ A, Cesselin F, Hamon M, Bourgoin S. The novel analgesic, cizolirtine, inhibits the spinal release of substance P and CGRP in rats. Neuropharmacology 2001;40: 578– 89. Bennett GJ, Xie YK. A peripheral mononeuropathy in rat that produces abnormal pain sensation like those seen in man. Pain 1988;33:87–107. Bourgoin S, Pohl M, Mauborgne A, Benoliel JJ, Collin E, Hamon M, Cesselin F. Monoaminergic control of the release of calcitonin generelated peptide- and substance P-like materials from rat spinal cord slices. Neuropharmacology 1993;32:633 –40. Bousquet P, Dontenwill M, Greney H, Feldman J. Imidazoline receptors in cardiovascular and metabolic diseases. J Cardiovasc Pharmacol 2000; 35:S21–5. Christensen D, Ida¨ npa¨ a¨ n-Heikkila¨ JJ, Guilbaud G, Kayser V. The antinociceptive effect of combined systemic administration of morphine and the glycine/NMDA receptor antagonist, (þ)-HA966, in a rat model of peripheral neuropathy. Br J Pharmacol 1998;125:1641–50. Christensen D, Kayser V. The development of pain-related behaviour and opioid tolerance after neuropathy-inducing surgery and sham surgery. Pain 2000;88:231 –8. Collin E, Frechilla D, Pohl M, Bourgoin S, Mauborgne A, Hamon M, Cesselin F. Differential effects of the novel analgesic, S12813-4, on the spinal release of substance P-and calcitonin gene-related peptide-like materials in the rat. Naunyn Schmiedebergs Arch Pharmacol 1994;349: 387– 93. Committee for Research and Ethical Issues of the IASP. Ethical standards for investigations of experimental pain in animals. Pain 1983;16: 109– 10. Cruce WL, Lovell JA, Crisp T, Stuesse SL. Effect of aging on the substance P receptor, NK-1, in the spinal cord of rats with peripheral nerve injury. Somatosens Mot Res 2001;18:66– 75. Doyle CA, Hunt SP. A role for spinal lamina I neurokinin-1-positive neurons in cold thermoreception in the rat. Neuroscience 1999;91: 723– 32. Fairbanks CA, Stone LS, Kitto KF, Nguyen HO, Posthumus IJ, Wilcox GL. a2c-adrenergic receptors mediate spinal analgesia and adrenergic– opioid synergy. J Pharmacol Exp Ther 2002;300:282–90. Fairbanks CA, Wilcox GL. Moxonidine, a selective a2-adrenergic and imidazoline receptor agonist, produces spinal antinociception in mice. J Pharmacol Exp Ther 1999;290:403–12.
Frost SA, Raja SN, Campbell JN, Meyer RA, Khan AA. Does hyperalgesia to cooling stimuli characterize patients with sympathetically maintained pain (reflex sympathetic dystrophy)? In: Dubner R, Gebhart GF, Bond MR, editors. Pain research and clinical management. Amsterdam: Elsevier, 1988. p. 151–6. Ho¨ kfelt T, Ljungdahl A, Terenius L, Elde R, Nilsson G. Immunohistochemical analyses of peptide pathways possibly related to pain and analgesia: enkephalin and substance P. Proc Natl Acad Sci USA 1977; 74:3081– 5. Hole K, Tjølsen A. The tail-flick and formalin tests in rodents : changes in skin temperature as a confounding factor. Pain 1993;53:247 –54. Hylden JL, Thomas DA, Iadarola MJ, Nahin RL, Dubner R. Spinal opioid analgesic effects are enhanced in a model of unilateral inflammation/ hyperalgesia: possible involvement of noradrenergic mechanisms. Eur J Pharmacol 1991;194:135–43. Ida¨ npa¨ a¨ n-Heikkila¨ JJ, Guilbaud G, Kayser V. Prevention of tolerance to the antinociceptive effects of systemic morphine by a selective cholecystokinin-B receptor antagonist in a rat model of peripheral neuropathy. J Pharmacol Exp Ther 1997;282:1366–72. Jasmin L, Kohan L, Franssen M, Janni G, Goff JR. The cold plate as a test of nociceptive behaviors: description and application to the study of chronic neuropathic and inflammatory pain models. Pain 1998;75:367–82. Jasmin L, Tien D, Weinshenker D, Palmiter RD, Green PG, Janni G, Ohara PT. The NK1 receptor mediates both the hyperalgesia and the resistance to morphine in mice lacking noradrenaline. Proc Natl Acad Sci USA 2002;99:1029–34. Kayser V, Desmeules J, Guilbaud G. Effects of noradrenergic agonists on animal chronic pain models: arthritic and mononeuropathic rats. In: Besson JM, Guilbaud G, editors. Towards the use of noradrenergic agonists for the treatment of pain. Amsterdam: Excerpta Medica, 1992. p. 151–67. Kayser V, Desmeules J, Guilbaud G. Systemic clonidine differentially modulates the abnormal reactions to mechanical and thermal stimuli in rats with peripheral mononeuropathy. Pain 1995;60:275–85. Kayser V, Christensen D. Antinociceptive effect of systemic gabapentin in mononeuropathic rats depends on stimulus characteristics and level of test integration. Pain 2000;88:53–60. Le Bars D, Gozariu M, Cadden SW. Animal models of nociception. Pharmacol Rev 2001;53:597– 652. Malmberg AB, Basbaum AI. Partial sciatic nerve injury in the mouse as a model of neuropathic pain: behavioral and neuroanatomical correlates. Pain 1998;76:215–22. Marchand JE, Wurm WH, Kato T, Kream RM. Altered tachykinin expression by dorsal root ganglion neurons in a rat model of neuropathic pain. Pain 1994;58:219–31. Martinez L, Serafini MT, Pretel MJ, Garcia-Soret A, Moragon T, Puig S. Absorption, distribution, metabolism and excretion of cizolirtine, a new analgesic compound, in rat and dog. Xenobiotica 1999;29:859–71. Matthew IR, Ogden GR, Frame JW, Wight AJ. Dose response and safety of cizolirtine citrate (E-4018) in patients with pain following extraction of third molars. Curr Med Res Opin 2000;16:107–14. McLeod AL, Julien JP, Cuello AC, Krause JE, Ribeiro-da-Silva A. Ectopic substance P-immunoreactive boutons are preferentially presynaptic to neurokinin-1 receptor immunoreactive dendrites in the spinal white matter of transgenic mice. Brain Res 1999;836:1–8. Monck N. Cizolirtine. Laboratorios Dr Esteve. Curr Opin Invest Drugs 2001;2:1269–72. Noguchi K, Dubner R, De Leon M, Senba E, Ruda MA. Axotomy induces preprotachykinin gene expression in a subpopulation of dorsal root ganglion neurons. J Neurosci Res 1994;37:596–603. Puig MM, Samso E, Pol O. Analgesic effect of a2-adrenergic agonists. Dolor 2000;15:179–85. Rogers DT, Iwamoto ET. Multiple spinal mediators in parenteral nicotineinduced antinociception. J Pharmacol Exp Ther 1993;267:341–9. Shembalkar P, Taubel J, Abadias M, Arezina R, Hammond K, Anand P. Cizolirtine citrate (E-4018) in the treatment of chronic neuropathic pain. Curr Med Res Opin 2001;17:262–6.
V. Kayser et al. / Pain 104 (2003) 169–177 Suh HH, Fujimoto JM, Tseng LL-F. Differential mechanisms mediating bendorphin-and morphine-induced analgesia in mice. Eur J Pharmacol 1989;168:61 –70. Tjølsen A, Lund A, Hole K. The role of descending noradrenergic systems in regulation of nociception: the effects of intrathecally administered aadrenoceptor antagonists and clonidine. Pain 1990;43:113–20.
177
Tjølsen A, Rosland JH, Berge OG, Hole K. The increasing-temperature hotplate test : an improved test of nociception in mice and rats. J Pharmacol Methods 1991;25:241 –50. Yaksh T. Alpha-2-adrenergic agonists as analgesics. In: Sawynok J, Cowan A, editors. Novel aspects of pain management: opioids and beyond. New York, NY: Wiley-Liss, 1999. p. 179–202.
Effects of the novel analgesic, cizolirtine, in a rat model of neuropathic pain Vale´rie Kaysera,*, Antonio Farre´b, Michel Hamona, Sylvie Bourgoina a
INSERM U288, NeuroPsychoPharmacologie Mole´culaire, Cellulaire et Fonctionnelle, Faculte´ de Me´decine Pitie´-Salpeˆtrie`re, 91, Boulevard de l’Hoˆpital, 75634 Paris Cedex 13, France b Laboratorios Dr Esteve, S.A., Avenue Mare de De´u de Montserrat, 221, 08041 Barcelona, Spain Received 5 September 2002; accepted 18 December 2002
Abstract Cizolirtine (5-{[(N,N-dimethylaminoethoxy)phenyl]methyl}-1-methyl-1H-pyrazol citrate) is a centrally acting analgesic with a currently unknown mechanism of action, whose efficacy has been demonstrated in various models of acute and inflammatory pain in rodents. Further studies were performed in order to assess its potential antinociceptive action in a well-validated model of neuropathic pain, i.e. that produced by unilateral sciatic nerve constriction in rats. Animals were subjected to relevant behavioural tests based on mechanical (vocalization threshold to paw pressure) and thermal (struggle latency to paw immersion in a cold (108C) water bath) stimuli, 2 weeks after sciatic nerve constriction, when pain-related behaviour was fully developed. Acute pretreatment with 2.5– 10 mg/kg p.o. of cizolirtine reversed both mechanical and thermal allodynia. These effects were antagonized by prior injection of the a2-adrenoceptor antagonist idazoxan (0.5 mg/kg i.v.), but not the opioid receptor antagonist naloxone (0.1 mg/kg i.v.). On the other hand, cizolirtine (10 mg/kg p.o.) produced no motor deficits in animals using the rotarod test. Our study showed that cizolirtine suppressed pain-related behavioural responses to mechanical and cold stimuli in neuropathic rats, probably via an a2-adrenoceptor-dependent mechanism. These results suggest that cizolirtine may be useful for alleviating some neuropathic somatosensory disorders, in particular cold allodynia, with a reduced risk of undesirable side effects. q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. Keywords: Cizolirtine; Idazoxan; Naloxone; Antinociception; Neuropathic pain
1. Introduction Cizolirtine (5-{[(N,N-dimethylaminoethoxy)phenyl]methyl}-1-methyl-1H-pyrazol citrate) is a centrally acting analgesic with a currently unknown mechanism of action (Alvarez et al., 2000; Monck, 2001). This compound is characterized by a complete lack of anti-inflammatory properties and shows no affinity for m, d or k opioid receptors (Alvarez et al., 2000). Nevertheless, the antinociceptive efficacy of cizolirtine has been demonstrated in various animal models of acute and inflammatory pain (Alvarez et al., 2000), thereby suggesting that this drug may be an alternative to opioids for alleviating such pain (Matthew et al., 2000). Interestingly, blockade of a2adrenoceptors by idazoxan significantly reduced cizolirtineinduced analgesia, whereas, in contrast, pretreatment with * Corresponding author. Tel.: þ 33-1-4077-9709; fax: þ33-1-4077-9790. E-mail address: [email protected] (V. Kayser).
desipramine, a selective inhibitor of noradrenaline (NA) reuptake, potentiated the cizolirtine action indicating its mediation through adrenoceptor-dependent mechanisms (Alvarez et al., 2000). However, cizolirtine has no affinity for adrenoceptors and NA transporters, suggesting an indirect activation of NA neurotransmission by the drug. Another notable effect of this novel analgesic drug is its inhibitory influence on the release of substance P and calcitonin gene-related peptide at the spinal level in rats (Ballet et al., 2001). This effect can also be prevented by pretreatment with idazoxan, further emphasizing the implication of a noradrenergic link in the central action of cizolirtine (Ballet et al., 2001). Because previous studies showed that activation of a2adrenoceptors at the spinal level exerts some antinociceptive effects in neuropathic pain (Yaksh, 1999), we addressed the question of whether cizolirtine could also be efficient to alleviate this type of pain. Indeed, a preliminary study in a
0304-3959/03/$30.00 q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. doi:10.1016/S0304-3959(02)00497-9
170
V. Kayser et al. / Pain 104 (2003) 169–177
small number of patients suffering from traumatic neuropathic pain showed that a dose of 115 mg of cizolirtine twice a day for 21 days yielded a marked improvement in a subgroup of five patients with primary allodynia (Shembalkar et al., 2001). In the present study, we evaluated the antinociceptive action of cizolirtine in a well-established model of neuropathic pain, i.e. that which consists of a unilateral chronic constriction injury of the sciatic nerve in rats (Bennett and Xie, 1988). A behavioural test (vocalization threshold to paw pressure), that allowed clear-cut demonstration of the antinociceptive activity of a2-adrenoceptor agonists, antidepressants, sympatholytic and anticonvulsant drugs (Kayser and Christensen, 2000), was used to assess the effect of acute treatment with cizolirtine. In addition, because cold allodynia is a characteristic symptom of neuropathic pain (Frost et al., 1988), we also determined the effect of cizolirtine on the latency of struggle response to paw immersion in a non-noxious cold (108C) water bath (Attal et al., 1990). Finally, we also tested whether a2adrenoceptor blockade by idazoxan could prevent the effects of cizolirtine in this model of neuropathic pain, as previously observed in relevant models of inflammatory pain (Alvarez et al., 2000).
2. Materials and methods 2.1. Animals Experiments were performed on male Sprague –Dawley rats (300 – 350 g; Charles River, St-Aubin-le`s-Elbeuf, France) housed in groups of five per cage. Rats were maintained on a 12/12 h light/dark cycle at an ambient temperature of 21 ^ 18C, and were allowed free access to food and water. Experiments were carried out in accordance with both the European Community Council Directive (86/609/EEC) and the Ministry of Agriculture regulations. In addition, we adhered to the guidelines of the Committee For Research and Ethical Issues of the International Association for the Study of Pain (1983). 2.2. Surgical procedure Sciatic nerve injury was made following a procedure (Attal et al., 1990) adapted from Bennett and Xie (1988). Briefly, rats were anesthetized with pentobarbital (50 mg/ kg, i.p.) and external area of the right thigh was shaved and swabbed with 70% ethanol. A skin incision was made in the cleaned area, and the sciatic nerve was exposed following blunt dissection and gently freed from adhering tissues. Four ligatures (5-0 chromic catgut, applied 1– 2 mm apart) were then placed around the nerve. Care was taken to ensure that the ligatures were not too tight so as to occlude perineural blood flow. Finally, the separated muscle was stitched and the incision was closed with silk thread. In sham-operated
control animals, identical surgical procedures were employed, but the sciatic nerve was left intact. 2.3. Drugs and doses Cizolirtine (Esteve Labs, Barcelona, Spain) was given orally through a feeding cannula in a volume of 10 ml of water. The doses, 2.5, 5 and 10 mg/kg, were chosen on the basis of previous studies in rodents (Alvarez et al., 2000). The a2-adrenoceptor antagonist, idazoxan (Res. Biochem. Int., Natick, MA, USA), was injected in saline (0.9% NaCl, 0.1 ml) into a lateral tail vein 30 min before cizolirtine administration (10 mg/kg per os). The selected dose of idazoxan, 0.5 mg/kg i.v., was previously shown to completely prevent the antinociceptive effect of the prototypical a2-adrenoceptor agonist clonidine, without affecting, on its own, the nociceptive threshold in this model of neuropathic pain (Kayser et al., 1995). Naloxone (Res. Biochem. Int.) was injected in saline 15 min before cizolirtine (10 mg/kg per os) at a dose of 0.1 mg/kg i.v., that was previously shown to completely prevent the antinociceptive effect of a maximally effective dose (1 mg/kg i.v.) of morphine in the same model (Christensen et al., 1998). Paired control rats were treated with water (10 ml p.o. per rat) and/or saline (0.1 ml i.v. per rat) using the same routes of administration. 2.4. Behavioural testing Animals were fasted for 16 h before testing, but water was always provided ad libitum. Testing sessions, beginning at 9:30 AM, were conducted in a quiet room. The experimenter was blind to the drug and the dose tested. Rats were not acclimatized to the test situations beforehand. Each animal received either cizolirtine, idazoxan, naloxone or drug combinations only once and was used in only one experiment. 2.4.1. Mechanical test The vocalization thresholds, expressed as grams, were determined according to a modification of the Randall and Selitto method (Kayser and Christensen, 2000). An increasing pressure was applied through a dome-shaped plastic tip (diameter ¼ 1 mm) onto the dorsal surface of the hindpaw using a Ugo Basile analgesimeter (Comerio, Italy). The tip was positioned between the third and fourth metatarsus (into the sciatic nerve territory) and pressure was applied until the rat squeaked. This response represents a more integrated nociceptive behaviour than paw withdrawal and allows reliable assessment of the respective potencies of various antinociceptive drugs in this pain model (Kayser and Christensen, 2000). Preoperative vocalization threshold (mean of two consecutive stable thresholds) was determined for both hindpaws in each rat. Then, 2 weeks after surgery, when the abnormal pain behaviour was at its maximum (Attal et al., 1990), a control threshold (mean of two consecutive stable
V. Kayser et al. / Pain 104 (2003) 169–177
thresholds) was determined before administration of the drug. Vocalization thresholds were then measured every 10 min, until they returned to baseline. 2.4.2. Cold test The struggle latencies, expressed as seconds, were determined after immersion of the hindpaw into a 108C water bath (Ministat MHUB 11; Bioblock Scientific, Illkirch, France) as described previously (Attal et al., 1990). The rat was gently held by the trunk, leaving the limbs freely exposed. The paw was then lowered into the water bath until the rat exhibited a struggle reaction, or for a maximum of 15 s when the rat did not exhibit such a reaction. Preoperative latency (mean of two consecutive stable trials 20 min apart) was determined for both hindpaws in each rat. Then, 2 weeks after surgery, a control latency (mean of two consecutive stable latencies) was determined before administration of the drug. Struggle latencies were then measured every 20 min, until they returned to baseline. The 20-min interval between successive measurements was necessary because abnormal reactions lasting for more than 15 min after a thermal stimulus have been reported in this model (Attal et al., 1990). 2.4.3. Rotarod test In a separate group of neuropathic rats ðn ¼ 8Þ, locomotor function was tested using the Ugo Basile accelerating rotarod (model 7750 for rats). This apparatus consists of a base platform and a rotating horizontal rod (7 cm in diameter, 50 cm in length) with a non-skid surface. The rod is divided by five disks into four sections of equal length in which four rats can be tested simultaneously. Under each drum section, 26 cm below the rod on the platform, a V-shaped counter-trip plate is positioned. Animals were acclimatized to the revolving drum and habituated to handling in order to avoid stress during testing. The rod was set to accelerate from 4 to 40 rpm in a 5-min period. The integrity of motor coordination was assessed as the performance time on the rod measured from the start of acceleration until the animal fell from the drum onto the counter-trip plate. Rats were acclimatized to acceleration by three training runs. The mean of the performance time determined at the fourth and fifth training runs served as the reference control value (expressed as seconds). The performance time was then measured every 20 min for a total of 80 min after per os administration of cizolirtine or water, and compared with the control value. 2.5. Measurement of skin temperature of the rat paw Because of the possible occurrence of interferences between the pharmacological effects of drugs and adaptive thermoregulation in the thermal test (Le Bars et al., 2001), we measured skin temperature at the level of hindpaws in neuropathic rats ðn ¼ 12Þ before and after cizolirtine
171
(10 mg/kg p.o.) or vehicle (water) administration. Measurements were made using the thermocouple thermometer Digisense (15 mm in diameter; Model N8 8528-10; Cole – Parmer Instrument Co., Chicago, IL) in a quiet room maintained at constant temperature (22 – 238C). The rats were left in their cage and the thermocouple was gently placed on the lateral side of the plantar surface of the hindpaw, in the territory of the sciatic nerve. Stable temperature readings were obtained after 10 s with a precision of 0.18C. For each rat, two consecutive basal measurements were averaged and the temperature was then measured successively on the contralateral and the nerveinjured side every 20 min after drug or vehicle administration for a total of 240 min. 2.6. Statistical analyses All data are expressed as means ^ SEM. A paired Student’s t-test was employed to compare the mechanical thresholds and cold water response latencies before and after the surgery. The overall effects of treatment (areas under the curve, AUCs) were calculated by use of the trapezoidal rule. Statistical significance of the data was analyzed by one-way analysis of variance (ANOVA). The observed significances ðP , 0:05Þ were then confirmed with the Tukey’s test.
3. Results 3.1. Responses to mechanical and thermal stimuli in sciatic nerve-ligatured and sham-operated rats 3.1.1. Mechanical test In agreement with previous observations (Christensen and Kayser, 2000), the vocalization thresholds to paw pressure were markedly decreased 2 weeks after the sciatic nerve ligature (Table 1). At this time, the mean threshold for the nerve-injured paw was 196 ^ 7 g (P , 0:001 vs. the preconstriction value 303 ^ 12 g, n ¼ 72). The vocalization threshold to paw pressure of the contralateral paw was also diminished, but to a minor extent (265 ^ 7 g vs. the preconstriction value 304 ^ 8 g, P , 0:01, n ¼ 72). In contrast, the vocalization thresholds of the sham-operated rats did not differ from those measured in intact animals (302 ^ 13 g for the ipsilateral paw; 303 ^ 7 g for the contralateral paw, n ¼ 72) (Table 1). 3.1.2. Cold test In further agreement with previous studies (Christensen and Kayser, 2000), struggle latencies to cold stimulus applied to the nerve-injured paw were decreased by several seconds, 2 weeks after the nerve ligature (Table 2). At this time, the mean struggle latency at 108C was 7.0 ^ 0.3 s (P , 0:001 vs. the preconstriction value 14.6 ^ 0.8 s, n ¼ 72). In contrast, the response latencies in sham-
172
V. Kayser et al. / Pain 104 (2003) 169–177
Table 1 Maximal vocalization thresholds for the nerve-injured and contralateral paws in sciatic nerve-ligatured rats, and for both hindpaws in sham-operated rats, in the paw pressure test before and after administration (adm.) of vehicle, cizolirtine (2.5–10 mg/kg p.o.) and/or idazoxan (0.5 mg/kg i.v.) or naloxone (0.1 mg/kg i.v.) Treatment adm.
Vehicle (water) Cizolirtine 2.5 Cizolirtine 5 Cizolirtine 10 Idazoxan 0.5 Idazoxan 0.5 þ cizolirtine 10 Naloxone 0.1 Naloxone 0.1 þ cizolirtine 10
Sciatic nerve-ligatured rats
Sham-operated rats (right and left pawsa)
Nerve-injured right paw
Contralateral left paw
Before adm.
After adm.
Before adm.
After adm.
194 ^ 7 186 ^ 6 192 ^ 8 204 ^ 8 200 ^ 9 198 ^ 5 195 ^ 6 201 ^ 8
201 ^ 6 270 ^ 12b 293 ^ 8b 306 ^ 7b 210 ^ 11 206 ^ 7c 201 ^ 8 315 ^ 12b
266 ^ 6 268 ^ 5 274 ^ 6 255 ^ 5 270 ^ 10 265 ^ 7 256 ^ 8 270 ^ 7
270 ^ 6 343 ^ 12b 341 ^ 13b 309 ^ 15b 282 ^ 14 269 ^ 15c 265 ^ 13 340 ^ 15b
Before adm.
After adm.
300 ^ 10 290 ^ 18 294 ^ 18 312 ^ 12 310 ^ 17 300 ^ 12 308 ^ 10 305 ^ 11
310 ^ 10 378 ^ 6b 369 ^ 17b 414 ^ 21b 312 ^ 20 314 ^ 11c 312 ^ 18 395 ^ 6b
Results are expressed as grams. Each value is the mean ^ SEM of nine independent determinations. The mean preoperative thresholds were 303 ^ 8 and 304 ^ 9 g for the right and the left hindpaw, respectively (means ^ SEM, n ¼ 144). a Values for the right and left paws did not differ and were pooled for the sake of clarity. b P , 0:01 vs. before administration (Tukey’s test). c P , 0:01 vs. cizolirtine (10 mg/kg p.o.) alone.
operated rats did not differ from those measured in intact animals (13.5 ^ 1.2 s, n ¼ 10). 3.2. Antinociceptive effects of cizolirtine 3.2.1. Mechanical test Cizolirtine produced significant increases in vocalization thresholds for the nerve-injured paw at all doses tested (Table 1, Fig. 1). Indeed, within 10– 40 min following administration of the two highest doses, vocalization thresholds raised up to values close to those found in intact
animals. Thereafter, the effect of cizolirtine progressively vanished, and vocalization thresholds decreased back to predrug values of 40 – 60 min after administration of 2.5, 5 or 10 mg/kg of the compound, respectively (Fig. 1). In the contralateral paw, cizolirtine also produced an increase in the vocalization thresholds that lasted for , 30 min (Table 1). However, the overall effect (AUC) of
Table 2 Maximal struggle latencies for the nerve-injured paw in sciatic nerveligatured rats, in the cold test before and after administration (adm.) of vehicle, cizolirtine (2.5–10 mg/kg p.o.) and/or idazoxan (0.5 mg/kg i.v.) or naloxone (0.1 mg/kg i.v.) Treatment
Vehicle (water) Cizolirtine 2.5 Cizolirtine 5 Cizolirtine 10 Idazoxan 0.5 Idazoxan 0.5 þ cizolirtine 10 Naloxone 0.1 Naloxone 0.1 þ cizolirtine 10
Nerve-injured paw Before adm.
After adm.
7.3 ^ 0.5 7.4 ^ 0.2 7.1 ^ 0.4 6.7 ^ 0.3 6.7 ^ 0.2 7.0 ^ 0.5 6.8 ^ 0.5 7.0 ^ 0.3
7.6 ^ 0.7 8.0 ^ 0.9 13.1 ^ 0.6a 13.4 ^ 1.0a 7.1 ^ 0.9 7.2 ^ 0.7b 7.6 ^ 0.9 13.8 ^ 0.9a
Results (in seconds) are expressed as means ^ SEM of nine independent determinations.
The mean preoperative latency was 14.6 ^ 0.8 s for the right hindpaw (mean ^ SEM, n ¼ 72). a b
P , 0:01 vs. before administration (Tukey’s test). P , 0:01 vs. cizolirtine (10 mg/kg p.o.) alone.
Fig. 1. Time-course changes in the vocalization threshold in the mechanical pressure test applied to the nerve-injured hindpaw in sciatic nerve-ligatured rats. Effects of cizolirtine. Cizolirtine (2.5, 5 or 10 mg/kg) or vehicle (water) was administered per os at time 0, and vocalization threshold (in grams) was measured every 10 min for up to 60–70 min thereafter. Each value is the mean ^ SEM of nine independent determinations (one determination per rat). Values at time 0 corresponded to vocalization thresholds in untreated sciatic nerve-ligatured rats. The horizontal grey band indicates the range of vocalization threshold values for unoperated rats (303 ^ 12 g, n ¼ 72). *P , 0:05, **P , 0:01 vs. vehicle-treated nerve-injured rats (Tukey’s test).
V. Kayser et al. / Pain 104 (2003) 169–177
cizolirtine was significantly larger on the nerve-injured paw than on the contralateral paw (P , 0:05 for 2.5, P , 0:01 for 5 or 10 mg/kg of the compound, respectively). In sham-operated rats, cizolirtine significantly increased the vocalization thresholds to paw pressure for either the ipsilateral or the contralateral hindpaw (Table 1). At the three doses tested, similar effects were noted for both hindpaws. Comparison of overall effects (AUC) showed that cizolirtine-induced increase in vocalization thresholds in sham-operated rats did not differ ðP ¼ 0:16Þ from that found for the contralateral hindpaw in sciatic nerveligatured rats. In contrast to cizolirtine, vehicle (water) administration did not affect vocalization thresholds in both sciatic nerveligatured and sham-operated rats (Table 1, Fig. 1). 3.2.2. Cold test Vehicle or 2.5 mg/kg of cizolirtine had no effect in sciatic nerve-ligatured rats (Table 2, Fig. 2). By contrast, the other two doses of cizolirtine produced a significant increase in the struggle latency (Table 2, Fig. 2). Thus, a doubling in the latency was noted 20 min after administration of the drug at 5 mg/kg. The effect then declined progressively to reach pre-drug values 120 min after the treatment (Fig. 2). A more prolonged effect was seen with 10 mg/kg of cizolirtine. However, at its maximum (60 – 80 min after administration), the effect caused by 10 mg/kg of cizolirtine was not higher than that observed only after 5 mg/kg of the drug (Table 2, Fig. 2). Nevertheless, significant differences were noted between the two doses because the effect of
173
cizolirtine was clearly dose-dependent with regard to the duration of the effect (r ¼ 0:83, P , 0:001) (Fig. 2). Neither vehicle nor cizolirtine (2.5 – 10 mg/kg) significantly affected struggle latencies in sham-operated rats (not shown). 3.3. Effects of idazoxan and naloxone on the antinociceptive effects of cizolirtine On their own, neither idazoxan (0.5 mg/kg i.v.) nor naloxone (0.1 mg/kg i.v.) affected the vocalization thresholds to mechanical stimulation as well as the struggle latencies in the cold test applied to the injured and the contralateral sides in both sciatic nerve-ligatured and shamoperated rats (Tables 1 and 2). However, pretreatment with idazoxan completely prevented the increases in vocalization thresholds and struggle latencies normally observed after per os administration of cizolirtine at 10 mg/kg. In contrast, naloxone pretreatment did not modify these effects of cizolirtine (Tables 1 and 2). 3.4. Lack of effect of cizolirtine on the rotarod performance time Like its vehicle (water), cizolirtine at the highest dose tested, 10 mg/kg p.o., did not significantly alter the rotarod performance time for at least 80 min after its administration in sciatic nerve-ligatured rats (Table 3). 3.5. Lack of effect of cizolirtine on hindpaw skin temperature Hindpaw skin temperature was measured at various times (20 – 240 min) after per os administration of cizolirtine (2.5 – 10 mg/kg) or water to sciatic nerve-ligatured rats maintained at an ambient temperature of 22– 238C. At all times studied, skin temperature did not significantly differ between rats having received either cizolirtine or water p.o. All along the experiment, the skin temperature of the nerveinjured hindpaw remained stable at 26.7 ^ 1.38C and that of the contralateral hindpaw at 26.5 ^ 1.48C (means ^ SEM, n ¼ 12), in sciatic nerve-ligatured rats treated with cizolirtine or vehicle.
4. Discussion Fig. 2. Time-course changes in the struggle latency in the cold (108C) test applied to the nerve-injured hindpaw in sciatic nerve-ligatured rats. Effects of cizolirtine. Cizolirtine (2.5, 5 or 10 mg/kg) or vehicle (water) was administered per os at time 0, and struggle latency (in seconds) was measured every 20 min for up to 120–220 min thereafter. Each value is the mean ^ SEM of nine independent determinations (one determination per rat). Values at time 0 corresponded to struggle latencies in untreated sciatic nerve-ligatured rats. The horizontal grey band indicates the range of struggle latency values for unoperated rats (14.6 ^ 0.8 s, n ¼ 72). *P , 0:05, **P , 0:01 vs. respective 0 time values (Tukey’s test).
Rats with a unilateral constriction injury of the sciatic nerve clearly exhibited abnormal pain sensitivity, with decreased vocalization thresholds to mechanical stimuli and decreased struggle latencies in response to 108C cold stimuli, as already shown in earlier studies (Kayser and Christensen, 2000). Peroral administration of cizolirtine was found to inhibit pain-related behaviours associated with neuropathic injury after the neuropathy had been established. Interestingly, after p.o. administration of [14C]cizo-
174
V. Kayser et al. / Pain 104 (2003) 169–177
Table 3 Rotarod performance time before and after p.o. administration of vehicle or cizolirtine (10 mg/kg) in sciatic nerve-ligatured rats Treatment
Vehicle (water) Cizolirtine
Before administration
108 ^ 10 116 ^ 15
Post-administration time (min) 20
40
60
80
118 ^ 2 111 ^ 24
113 ^ 10 115 ^ 23
109 ^ 24 119 ^ 24
105 ^ 29 119 ^ 18
Results (in seconds) are expressed as means ^ SEM of four independent determinations. None of the values after cizolirtine treatment were significantly different ðP . 0:05Þ from those in vehicle-treated rats and from pretreatment values.
lirtine within the same dose range as that chosen in our study, peak plasma concentrations of radioactivity are reached within 20 min (Martinez et al., 1999), which is in accordance with the time of the peak effects that we observed in both mechanical and thermal tests.
result is consistent with a recent report showing that cizolirtine inhibits hyperalgesia in rat models of inflammatory pain, without altering motor coordination or exerting sedative effects (Alvarez et al., 2000).
4.1. Effects of cizolirtine on mechanical pain-related behavioural response
4.2. Effects of cizolirtine on thermal pain-related behaviours
Using a centrally integrated test, the measure of the vocalization threshold to paw pressure, we showed that cizolirtine reversed the abnormal reactivity to mechanical stimuli in rats with peripheral mononeuropathy. Indeed, at the doses of 5 and 10 mg/kg p.o., cizolirtine increased the threshold for the nerve injured paw up to a value not different from that found in both sham-operated and intact animals. This effect was close to the one previously found in the same model of neuropathic pain after an i.v. injection of 0.3 mg/kg of morphine (Christensen et al., 1998) or 0.1 mg/ kg of the a2-adrenoceptor agonist clonidine (Kayser et al., 1995). The effect of cizolirtine was observed not only on the nerve-injured side, but also, although to a significantly lower extent, on the contralateral side. At the same doses, 2.5 –10 mg/kg p.o., cizolirtine also produced a significant antinociceptive effect in sham-operated animals, which was roughly comparable in magnitude with that observed for the contralateral hindpaw in sciatic nerve-ligatured rats. Interestingly, the effects of cizolirtine were prevented by i.v. pretreatment with the a2-adrenoceptor antagonist idazoxan, suggesting their mediation through mechanisms related to those triggered by direct and indirect a2-adrenoreceptor agonists such as clonidine (Kayser et al., 1995) and S12813-4 (Kayser et al., 1992), respectively. It has to be emphasized that idazoxan had, on its own, no significant hyperalgesic effect, thereby suggesting that a2-adrenoceptor-mediated inhibitory control of pain signal transmission is probably not tonically active in this neuropathic pain model (see also Kayser et al., 1995). While a2-adrenoceptor agonists are known to exert analgesic effects, they also have accompanying effects such as sedation and hypotension (Puig et al., 2000). Our experiments showed that cizolirtine, in the same dose range as that used to alleviate pain in neuropathic rats, did not decrease the performance time in the rotarod test. This
Another interesting observation in this study relates to the potent effect of cizolirtine on cold allodynia (i.e. abnormal reactions elicited by paw immersion in water at 108C). Cold allodynia is a characteristic symptom of neuropathic pain and is particularly resistant to pharmacological interventions in the present pain model (Kayser et al., 1995; Ida¨ npa¨ a¨ n-Heikkila¨ et al., 1997; Jasmin et al., 1998; Kayser and Christensen, 2000). However, cizolirtine was found to produce a marked, dose-dependent effect in the cold test, which lasted much longer than its effect in the mechanical test. While the lowest dose, 2.5 mg/kg p.o., of cizolirtine was essentially inactive, 5 mg/kg p.o. of the drug almost doubled the struggle latency in response to paw immersion in a 108C water bath. The highest dose tested, 10 mg/kg p.o., exerted the same effect but for a longer time compared with 5 mg/kg of cizolirtine (Fig. 2). This effect of cizolirtine (10 mg/ kg) was prevented by pretreatment with idazoxan, which confirmed its mediation through a2-adrenoceptors, like that already noted about cizolirtine-induced alleviation of pain caused by mechanical stimulation. Because changes in skin temperature may influence the response of rats to thermal stimuli (Hole and Tjølsen, 1993; Le Bars et al., 2001), we measured this parameter from 20 to 240 min after cizolirtine administration. Indeed, previous studies in the hot-plate test showed that an increase in skin temperature could lead to apparent hyperalgesia, whereas a reduction in skin temperature could lead to apparent analgesia (Tjølsen et al., 1991; Hole and Tjølsen, 1993). In the present study, we found that cizolirtine did not cause any significant change in paw skin temperature, which indicated that the effect of cizolirtine in the cold test could not be accounted for by some interfering alteration in peripheral thermoregulatory mechanisms.
V. Kayser et al. / Pain 104 (2003) 169–177
4.3. Possible mechanisms underlying the effects of cizolirtine in neuropathic rats We previously found that morphine and selective opioid receptor agonists injected i.v. are unable to reduce cold allodynia in neuropathic rats (Ida¨ npa¨ a¨ n-Heikkila¨ et al., 1997). In contrast, cizolirtine did reduce allodynia-related response in the cold test indicating that its action does not involve opioid-dependent mechanisms. Accordingly, we demonstrated that pretreatment with the opioid receptor antagonist naloxone (0.1 mg/kg i.v.) did not prevent the increases in vocalization thresholds and struggle latencies normally observed after per os administration of cizolirtine at 10 mg/kg. In line with these data, a recent report showed that this compound is not recognized by m, d or k opioid receptors in binding assays (Alvarez et al., 2000). Even though cizolirtine has no affinity for a2-adrenoceptors (Alvarez et al., 2000), it might activate descending noradrenergic pathways which are known to participate in an inhibitory control of nociception (Fairbanks et al., 2002). Previous studies demonstrated that the blockade of a2adrenoceptors by idazoxan inhibits the effects of cizolirtine (Alvarez et al., 2000; Ballet et al., 2001) and further supports to this observation were obtained in our own studies in sciatic nerve-ligatured rats. Conversely, the potentiation of noradrenergic neurotransmission by desipramine, probably through a facilitation of the agonistic action of endogenous NA at a2-adrenoceptors, results in an increased antinociceptive potency of cizolirtine (Alvarez et al., 2000). Like that caused by cizolirtine, there is evidence that analgesia due to opioids (Suh et al., 1989; Hylden et al., 1991) and nicotine (Rogers and Iwamoto, 1993), is partly mediated through a2adrenoceptor-mediated mechanisms. However, data obtained in the cold test clearly showed that cizolirtine exerted specific effects independent of opioid-related mechanisms. In this context, attention has to be paid to one of the main neurotransmitters of primary afferent fibers, i.e. substance P (Ho¨ kfelt et al., 1977). Substance P has been shown to play a key role in cold thermoreception and thermal allodynia through the stimulation of neurokinin1 (NK1) receptors in the superficial laminae of the dorsal spinal cord (Doyle and Hunt, 1999; McLeod et al., 1999). Furthermore, specific interactions between substance P and NA with regard to thermal hyperalgesia were recently evidenced in mutant mice lacking the gene coding for dopamine b-hydroxylase, the enzyme responsible for synthesis of NA from dopamine (Jasmin et al., 2002). Interestingly, these mice were found to exhibit a substance P-mediated chronic hyperalgesia to thermal, but not mechanical stimuli, which could be counteracted by restoring the synthesis of NA (Jasmin et al., 2002), and consequently, the activation of the pain controlling spinal a2c-adrenoceptors (Fairbanks et al., 2002). Because both an upregulation of NK1 receptors in the dorsal horn (Aanonsen et al., 1992; Malmberg and Basbaum, 1998; Cruce et al., 2001) and an increased production of substance P by primary
175
afferent neurons (Marchand et al., 1994; Noguchi et al., 1994) do occur in rats with neuropathic pain such as those bearing chronic ligatures of the sciatic nerve which were used herein, it can be proposed that thermal allodynia in these animals in fact results from a tonic activation of substance Pergic neurotransmission at the spinal level. Accordingly, cizolirtine would act through an a2c-adrenoceptor-dependent mechanism to reduce excessive substance P-ergic neurotransmission and the associated thermal allodynia. In line with this hypothesis, cizolirtine, like direct or indirect a2adrenoceptor agonists (Bourgoin et al., 1993; Collin et al., 1994), was recently found to exert an inhibitory influence on the spinal release of substance P in rats (Ballet et al., 2001). Furthermore, this effect could be prevented by the a2adrenoceptor antagonist idazoxan at the same dose as that used herein to normalize the behavioural response to paw immersion in water at 108C (B. Aubel, personal communication). Altogether, these data suggest that the mechanism of action of cizolirtine may resemble that of clonidine whose effects are also prevented by a2-adrenoceptor blockade (Tjølsen et al., 1990) and which inhibits the spinal release of substance P (Bourgoin et al., 1993; Collin et al., 1994). Indeed, previous studies also showed that clonidine reduces mechanical and cold allodynia-like behaviours as efficiently as cizolirtine in sciatic nerve-ligatured rats (Kayser et al., 1995). However, in contrast to clonidine, cizolirtine does not bind to a2-adrenoceptors (Alvarez et al., 2000), and further investigations are needed to assess whether cizolirtine, as expected from the aforementioned hypothesis, actually activates descending noradrenergic projections within the dorsal horn of the spinal cord. In addition to a2-adrenoceptors, imidazoline1 (I1) receptors have been proposed to contribute to the antinociceptive – antiallodynic effects of agonists such as clonidine and moxonidine, especially because (i) these drugs act as mixed a2 – I1 agonists (Fairbanks and Wilcox, 1999), and (ii) the a2-adrenoceptor antagonist most often used to prevent their effects, idazoxan, is also an I1 receptor antagonist (Bousquet et al., 2000). In the same line, one can ask whether the inhibitory effect of idazoxan on cizolirtineinduced antinociception might involve I1 receptors. In any case, the effects of cizolirtine cannot be ascribed to a direct stimulation of I1 receptors because this drug, at least up to 10 mM, does not bind to these receptors (J.A. GarciaSevilla, personal communication). However, it cannot be excluded that cizolirtine may induce the release of endogenous ligand(s) acting at I1 receptors, thereby producing at least part of its antinociceptive effects. Further studies with antagonists acting selectively at a2-adrenoceptors or I1 receptors are needed to definitively answer the question of the respective contributions of these receptors in the antinociceptive effects of cizolirtine. To conclude, it has to be emphasized that at the doses used to produce antinociceptive effects, cizolirtine, in contrast to direct a2-adrenoceptor agonists such as clonidine
176
V. Kayser et al. / Pain 104 (2003) 169–177
(Puig et al., 2000), exerted neither sedation nor motor disturbances. Accordingly, cizolirtine appears as a promising compound for the treatment of neuropathic pain, especially to reduce cold allodynia, with a low risk of undesirable side effects.
Acknowledgements This research was supported by grants from INSERM and Laboratorios Dr Esteve S.A.
References Aanonsen L, Kajander KC, Bennett GJ, Seybold VS. Autoradiographic analysis of 125I-substance P binding in rat spinal cord following chronic constriction injury of the sciatic nerve. Brain Res 1992;596:259 –68. Alvarez I, Andreu F, Buxens J, Colombo M, Dordal A, Fort M, Gutierrez B, Farre´ AJ. Pharmacology of cizolirtine: a new analgesic agent. Methods Find Exp Clin Pharmacol 2000;22:211–21. Attal N, Jazat F, Kayser V, Guilbaud G. Further evidence for a “painrelated” behaviour in a model of unilateral peripheral mononeuropathy. Pain 1990;41:235 –51. Ballet S, Aubel B, Mauborgne A, Polie´ nor A, Farre´ A, Cesselin F, Hamon M, Bourgoin S. The novel analgesic, cizolirtine, inhibits the spinal release of substance P and CGRP in rats. Neuropharmacology 2001;40: 578– 89. Bennett GJ, Xie YK. A peripheral mononeuropathy in rat that produces abnormal pain sensation like those seen in man. Pain 1988;33:87–107. Bourgoin S, Pohl M, Mauborgne A, Benoliel JJ, Collin E, Hamon M, Cesselin F. Monoaminergic control of the release of calcitonin generelated peptide- and substance P-like materials from rat spinal cord slices. Neuropharmacology 1993;32:633 –40. Bousquet P, Dontenwill M, Greney H, Feldman J. Imidazoline receptors in cardiovascular and metabolic diseases. J Cardiovasc Pharmacol 2000; 35:S21–5. Christensen D, Ida¨ npa¨ a¨ n-Heikkila¨ JJ, Guilbaud G, Kayser V. The antinociceptive effect of combined systemic administration of morphine and the glycine/NMDA receptor antagonist, (þ)-HA966, in a rat model of peripheral neuropathy. Br J Pharmacol 1998;125:1641–50. Christensen D, Kayser V. The development of pain-related behaviour and opioid tolerance after neuropathy-inducing surgery and sham surgery. Pain 2000;88:231 –8. Collin E, Frechilla D, Pohl M, Bourgoin S, Mauborgne A, Hamon M, Cesselin F. Differential effects of the novel analgesic, S12813-4, on the spinal release of substance P-and calcitonin gene-related peptide-like materials in the rat. Naunyn Schmiedebergs Arch Pharmacol 1994;349: 387– 93. Committee for Research and Ethical Issues of the IASP. Ethical standards for investigations of experimental pain in animals. Pain 1983;16: 109– 10. Cruce WL, Lovell JA, Crisp T, Stuesse SL. Effect of aging on the substance P receptor, NK-1, in the spinal cord of rats with peripheral nerve injury. Somatosens Mot Res 2001;18:66– 75. Doyle CA, Hunt SP. A role for spinal lamina I neurokinin-1-positive neurons in cold thermoreception in the rat. Neuroscience 1999;91: 723– 32. Fairbanks CA, Stone LS, Kitto KF, Nguyen HO, Posthumus IJ, Wilcox GL. a2c-adrenergic receptors mediate spinal analgesia and adrenergic– opioid synergy. J Pharmacol Exp Ther 2002;300:282–90. Fairbanks CA, Wilcox GL. Moxonidine, a selective a2-adrenergic and imidazoline receptor agonist, produces spinal antinociception in mice. J Pharmacol Exp Ther 1999;290:403–12.
Frost SA, Raja SN, Campbell JN, Meyer RA, Khan AA. Does hyperalgesia to cooling stimuli characterize patients with sympathetically maintained pain (reflex sympathetic dystrophy)? In: Dubner R, Gebhart GF, Bond MR, editors. Pain research and clinical management. Amsterdam: Elsevier, 1988. p. 151–6. Ho¨ kfelt T, Ljungdahl A, Terenius L, Elde R, Nilsson G. Immunohistochemical analyses of peptide pathways possibly related to pain and analgesia: enkephalin and substance P. Proc Natl Acad Sci USA 1977; 74:3081– 5. Hole K, Tjølsen A. The tail-flick and formalin tests in rodents : changes in skin temperature as a confounding factor. Pain 1993;53:247 –54. Hylden JL, Thomas DA, Iadarola MJ, Nahin RL, Dubner R. Spinal opioid analgesic effects are enhanced in a model of unilateral inflammation/ hyperalgesia: possible involvement of noradrenergic mechanisms. Eur J Pharmacol 1991;194:135–43. Ida¨ npa¨ a¨ n-Heikkila¨ JJ, Guilbaud G, Kayser V. Prevention of tolerance to the antinociceptive effects of systemic morphine by a selective cholecystokinin-B receptor antagonist in a rat model of peripheral neuropathy. J Pharmacol Exp Ther 1997;282:1366–72. Jasmin L, Kohan L, Franssen M, Janni G, Goff JR. The cold plate as a test of nociceptive behaviors: description and application to the study of chronic neuropathic and inflammatory pain models. Pain 1998;75:367–82. Jasmin L, Tien D, Weinshenker D, Palmiter RD, Green PG, Janni G, Ohara PT. The NK1 receptor mediates both the hyperalgesia and the resistance to morphine in mice lacking noradrenaline. Proc Natl Acad Sci USA 2002;99:1029–34. Kayser V, Desmeules J, Guilbaud G. Effects of noradrenergic agonists on animal chronic pain models: arthritic and mononeuropathic rats. In: Besson JM, Guilbaud G, editors. Towards the use of noradrenergic agonists for the treatment of pain. Amsterdam: Excerpta Medica, 1992. p. 151–67. Kayser V, Desmeules J, Guilbaud G. Systemic clonidine differentially modulates the abnormal reactions to mechanical and thermal stimuli in rats with peripheral mononeuropathy. Pain 1995;60:275–85. Kayser V, Christensen D. Antinociceptive effect of systemic gabapentin in mononeuropathic rats depends on stimulus characteristics and level of test integration. Pain 2000;88:53–60. Le Bars D, Gozariu M, Cadden SW. Animal models of nociception. Pharmacol Rev 2001;53:597– 652. Malmberg AB, Basbaum AI. Partial sciatic nerve injury in the mouse as a model of neuropathic pain: behavioral and neuroanatomical correlates. Pain 1998;76:215–22. Marchand JE, Wurm WH, Kato T, Kream RM. Altered tachykinin expression by dorsal root ganglion neurons in a rat model of neuropathic pain. Pain 1994;58:219–31. Martinez L, Serafini MT, Pretel MJ, Garcia-Soret A, Moragon T, Puig S. Absorption, distribution, metabolism and excretion of cizolirtine, a new analgesic compound, in rat and dog. Xenobiotica 1999;29:859–71. Matthew IR, Ogden GR, Frame JW, Wight AJ. Dose response and safety of cizolirtine citrate (E-4018) in patients with pain following extraction of third molars. Curr Med Res Opin 2000;16:107–14. McLeod AL, Julien JP, Cuello AC, Krause JE, Ribeiro-da-Silva A. Ectopic substance P-immunoreactive boutons are preferentially presynaptic to neurokinin-1 receptor immunoreactive dendrites in the spinal white matter of transgenic mice. Brain Res 1999;836:1–8. Monck N. Cizolirtine. Laboratorios Dr Esteve. Curr Opin Invest Drugs 2001;2:1269–72. Noguchi K, Dubner R, De Leon M, Senba E, Ruda MA. Axotomy induces preprotachykinin gene expression in a subpopulation of dorsal root ganglion neurons. J Neurosci Res 1994;37:596–603. Puig MM, Samso E, Pol O. Analgesic effect of a2-adrenergic agonists. Dolor 2000;15:179–85. Rogers DT, Iwamoto ET. Multiple spinal mediators in parenteral nicotineinduced antinociception. J Pharmacol Exp Ther 1993;267:341–9. Shembalkar P, Taubel J, Abadias M, Arezina R, Hammond K, Anand P. Cizolirtine citrate (E-4018) in the treatment of chronic neuropathic pain. Curr Med Res Opin 2001;17:262–6.
V. Kayser et al. / Pain 104 (2003) 169–177 Suh HH, Fujimoto JM, Tseng LL-F. Differential mechanisms mediating bendorphin-and morphine-induced analgesia in mice. Eur J Pharmacol 1989;168:61 –70. Tjølsen A, Lund A, Hole K. The role of descending noradrenergic systems in regulation of nociception: the effects of intrathecally administered aadrenoceptor antagonists and clonidine. Pain 1990;43:113–20.
177
Tjølsen A, Rosland JH, Berge OG, Hole K. The increasing-temperature hotplate test : an improved test of nociception in mice and rats. J Pharmacol Methods 1991;25:241 –50. Yaksh T. Alpha-2-adrenergic agonists as analgesics. In: Sawynok J, Cowan A, editors. Novel aspects of pain management: opioids and beyond. New York, NY: Wiley-Liss, 1999. p. 179–202.