Increased Release of Immunoreactive Cholecystokinin Octapeptide By

EJP 53006

Increased release of immunoreactive cholecystokinin octapeptide by morphine and potentiation of/z-opioid analgesia by CCK B receptor antagonist L-365,260 in rat spinal cord

Yan Zhou, Yu-Hua Sun, Zhi-Wen Zhang and Ji-Shen Han Neuroscience Research Center, Beijing Medical University, Beijing 100083, People's Republic of China

Received 2 November 1992, revised MS received 13 January 1993, accepted 19 January 1993

This is the first report showing, in an in vivo study, that systemic morphine produced a marked (89%, P < 0.01) increase of the cholecystokinin octapeptide (CCK-8) immunoreactivity in the perfusate of the rat spinal cord, an effect completely reversed by naloxone. Since CCK-8 has been shown to possess potent anti-opioid activity at a spinal level, a blockade of the spinal cholecystokinin effect would be expected to potentiate opiate analgesia. With tail flick latency as a nociceptive index, it was found that intrathecal (i.t.) injection of a novel CCK B antagonist L-365,260 produced a marked potentiation of the analgesic effect induced by the/~-opioid agonists morphine (4 mg/kg s.c.) or ohmefentanyl (32 ng i.t.). Similar effects were obtained with the CCK a antagonist devazepide at a dose 40-50 times higher than that of L-365,260. Both devazepide and L-365,260 showed a bell-shaped dose-response curve. The results confirm the notion that an increased release of CCK-8 may constitute a self-limiting process for opioid effects at the spinal level, and that it is the CCK B receptor which mediates the anti-opioid effect of CCK-8 in the rat spipal cord.

CCK receptor antagonists; CCK-8 immunoreactivity; CCK B receptors; Opiate analgesia; Naloxone

1. Introduction ments designed to test these hypotheses were not al- ways convincing and successful due to a lack of ade- Faris et al. (1983) were the first to show that chole- quate research tools. Thus, proglumide, a weak and cystokinin octapeptide (CCK-8) acts as a specific opiate non-specific CCK receptor antagonist, was shown to antagonist in the rat spinal cord, as evidenced by the potentiate morphine analgesia in one report (Watkins experimental finding that i.t. injection of 3.6 ng of et al., 1985) and to have no effect in another report CCK-8 antagonized opioid analgesia produced by foot- (Lehman et al., 1989). Central administration of chole- shock and morphine. This was confirmed by Han et al. cystokinin antiserum was reported to potentiate anal- (1985), who calculated that the dose of CCK-8 for 50% gesia induced by i.c.v, administered /3-endorphin (Itoh reversal of the effect of morphine (5 mg/kg) analgesia et al., 1985) but not by systemically injected morphine in the rat was 1.7 ng and 3.1 ng by i.c.v, and i.t. routes (Ding et al., 1986), probably due to difficulties with the of administration, respectively. Since CCK-8 has an rapid diffusion of the antibody molecule from the abundant and wide-spread distribution in both brain cerebral ventricle into the brain tissue. and spinal cord, one would naturally anticipate that The development of the potent, specific, non-peptide morphine or other endogenously released opioids CCK receptor antagonists, devazepide and L-365,031 would trigger the release of CCK-8 to counteract the for CCK A receptors (Chang and Lotti, 1986; Evens et opioid effect, and that removal of the effect of chole- al., 1986), and L-365,260 (Lotti and Chang, 1989) and cystokinin would result in an augmentation of opioid PD134308 (Wiesenfeld-Hallin et al., 1990) for CCK B analgesia and/or reversal of opiate tolerance. Experi- receptors, greatly facilitated research in this field. Dourish et al. (1990) have shown recently that morphine analgesia is significantly potentiated by L-365,260 Correspondence to: J.S. Han, Neuroscience Research Center, Bei- and by the CCK n antagonist L-365,031 at a 20 times jing Medical University, Beijing 100083, P.R. China. higher dose, suggesting that the CCK B receptor medi- 148 ates the anti-opioid effect. No information, however, is pril (1 ~M) and bestatin (1/zM) was perfused at a rate available concerning the site of action of these two of 1.0 ml/30 min with a push-pull pump. specific antagonists for their anti-opioid activity. Another basic question is whether morphine accel- 2.3. Nociceptive test erates the release of CCK-8 from the CNS. Besides the in vitro studies performed on periaqueductaI grey The rat was restrained in a plastic holder with its (Rattray and Delleroche, 1987), substantia nigra (Be- hind legs protruding. The nociceptive threshold was noliel et al., 1990) and spinal cord (Benoliel et al., measured by the latency of the tail flick response, 1991), the most commonly cited in vivo study is the elicited by radiant heat applied to the lower 1/3 of the preliminary observation of Tang et al. (1984) that the tail. The mean tail flick latency of three measurements addition of morphine (1 p~M) to the fluid perfusing rat taken at 5-rain intervals at the start of the experiment spinal subarachnoid space increases the cholecys- was taken as the basal threshold. The tail flick latency tokinin content of the perfusate. Since they did not taken at 10-min intervals after drug administration is present a statistical analysis in their preliminary report, expressed as the percentage change from basal tail it seemed worthwhile to re-evaluate this important flick latency, with a cutoff limit of + 150% to avoid issue in order to ascertain the extent and the time unnecessary skin damage. course of this reaction. On the basis of the above mentioned information 2.4. Radioimmunoassay of CCK-8 the present study was performed to clarify the follow- ing issues: (a) whether systemic injection of a conven- Cholecystokinin was extracted as reported by Mar- tional dose of morphine (5 mg/kg s.c.) would indeed ley and Rehfeld (1984). Aliquots of spinal perfusate accelerate the release of CCK-8 immunoreactivity from (1.0 ml) were collected in polyethylene tubes in an the rat spinal cord, and (b)whether the novel CCK A ice-water bath and then heated for 10 min in a water receptor antagonist devazepide and the CCK B recep- bath at 95°C. After cooling, the aliquots were cen- tor antagonist L-365,260 would potentiate opioid anal- trifuged at 10000 ×g for 15 min. Supernatants were gesia after i.t. injection and, if so, what is the dose ratio kept at -20°C and dried in a vacuum dryer. The between the two antagonists. residue was taken up in 0.2 ml H20 for the radioimmunoassay of CCK-8. The cholecystokinin antiserum, which was kindly provided by Dr. J.S. Hong (NIEHS, 2. Materials and methods USA), showed no detectable cross reactivity with [MetS]enkephalin, /3-endorphin, gastrin, CCK-7, CCK- 2.1. 1.t. injection of drugs 4, or unsulfated CCK-8. The radioimmunoassay proce- dure for cholecystokinin in the present study was simi- Male Wistar rats weighing 200-250 g were anes- lar to that suggested by Amersham Inc. On the basis of thetized with chlorohydrate (0.4 g/kg i.p.). The i.t. 10 consecutive assays, the sensitivity of the assay was cannula was implanted as described by Yaksh and 0.4 fmol/ml, with 4.6% intra-assay and 9.7% inter-as- Rudy (1976). A PE-10 polyethylene catheter was im- say variations. planted through the cisterna magna down to the subarachnoid space for 7.5 cm to reach the upper border 2.5. Drugs of the lumbar enlargement of the spinal cord. The outer part of the catheter was plugged and fixed to the 125I-Bolting Hunter-CCK-8 was purchased from skin after the wound was closed. Experiments with i.t. Amersham Inc. (UK). CCK-8 was a gift from Squibb injection of drugs started about 20-26 h after the and Sons, Inc. Morphine HC1 was purchased from surgical operation. Drugs were injected via the catheter Qinhai Drug Company (China). Naloxone HC1 was at a volume of 10 /zl, followed by 10 /zl of normal from Sigma (USA). Ohmefentanyl HCI (N-[1-(/3-hy- saline to flush out the cannula. Injection was com- droxy-fl-phenethy)-3-methyl-4-piperidyl]-N-phenylpro- pleted within 30 s. pionamide) was obtained from the Shanghai Institute of Materia Medica. Devazepide (3S(-)-N-(2,3-dihy- 2.2. Perfusion of the subarachnoid space of the spinal dro-l-methyl-2-oxo-5-phenyl-l-H- 1,4-benzodiazepin-3- cord yl)-H-indole-2-carboxamide) and L-365,260 (3R(+)-N- (2,3-dihydro-1-methyl-2-oxo-5-phenyl-1H-1,4-benzodia- Rats were anesthetized with chlorohydrate, and a zepin-3-yl)N'-(3-methyl-phenyl)urea) were donated by PE-10 catheter was inserted 7.5 cm down the subarach- Dr. R.F. Freidinger of the Merck, Sharp and Dohme noid space through the cisterna magna. Another Research Laboratories (USA). Devazepide and L- polyethylene tubing (PE-50) was inserted 1 cm into the 365,260 were first dissolved in dimethylsulfoxide cisterna. Artificial cerebrospinal fluid containing capto- (DMSO) and then diluted with propylene glycol to a 149

final ratio of 20% DMSO" 80% propylene glycol (v/v). Radioimmunoassay revealed that in the control group Morphine HC1, naloxone HC1 and ohmefentanyl were (fig. 1, left panel) the content of CCK-8 immunoreac- dissolved in normal saline. tivity remained rather stable within the range of 14.4 to 15.2 fmol/ml in the three perfusion periods. Adminis- 2.6. Statistical analysis tration of morphine resulted in an 89% increase in CCK-8 immunoreactivity in the first post-morphine Data were analyzed using a muitifactorial analysis of period (P < 0.01 compared with the pre-morphine pe- variance (ANOVA). Where ANOVA revealed a signi- riod) and a 32% increase in the second post-morphine ficant effect, between group comparisons were made period (fig. 1, middle panel). This effect of morphine using Duncan's test. Student's t-test was used for group was totally reversed by naloxone (20 /zg i.t.) adminis= comparisons where appropriate. A probability level of tered 10 rain prior to morphine (fig. 1, right panel). 0.05 was regarded as significant. 3.2. The effect of deuazepide and L-365,260 on nocicep- tit,e threshold 3. Results To assess the effect of the CCK receptor blockers 3.1. Increase in CCK-8 immunoreacticity in spinal per- on tail flick latency, successive doses of devazepide (10, fusates elicited by systemic morphine and the preL'ention 25, 50, 100, 200 and 400 ng) or L-365,260 (0.625, 1.25, of this effect by i.t. naloxone 2.5, 5.0, 10 and 20 ng) in 10 ~1 volumes were administered i.t. at 20-rain intervals. As can be seen from fig. Three groups of eight rats were used in the spinal 2A and B, no significant changes in tail flick latency perfusion experiments. Each group was perfused for were observed over a period of 140 min. three 30-min perfusion periods at a rate of 1 ml/30 rain. At the end of the first perfusion period, an i.t. 3.3. Potentiation of morphine analgesia by decazepide or injection of normal saline (10 /~i) or naloxone (20 ~zg) L-365,260 administered i.t. was given, followed 10 min later by a s.c. injection of normal saline (0.5 ml/rat) or morphine (5 mg/kg). Two groups of nine rats were injected i.t. with 100 Spinal perfusion was started again immediately after ng of devazepide or 10 #l of vehicle, followed 15 min the s.c. injection, which lasted for 60 rain (2 × 30 rain). later by a sc injection of 4 mg/kg morphine. The Perfusate was collected as 30-min fractions (1 ml). results are shown in fig. 3A. The antinociceptive effect of morphine was significantly higher in the devazepide group than in the control group (P < 0.01, ANOVA for the 60-min observation period). A similar experiment 40 was performed with normal saline instead of morphine. i.t. ~ S.C. Z ~ n=8 11 As can be seen from fig. 3A, no significant increase in tail flick latency was found in the two groups (ANOVA, 30 P > 0.05), indicating that devazepide alone was not I analgesic. 0 The above experiments were repeated with L- -6 '0 365,260 (2.5 rig) instead of devazepide (100 ng), and the results are shown in fig. 3B. L-365,260 produced a L,. 10 marked potentiation of morphine analgesia (P < 0.01, I ANOVA for the 60-min observation period). Injection I ! of L-365,260 alone did not produce significant changes 0 0 J 0 in tail flick latency. The dose-response curves for the effect of the two NS + NS NS÷Morphine NX ÷Morphine CCK receptor blockers on morphine analgesia are Fig. 1. The increase in CCK-8 immunoreactivity in the spinal perfusate elicited by systemic morphine and its blockade by i.t. nalox- shown in fig. 4, where changes in tail flick latency were one. Data are shown as means_+S.E.M, of the levels of CCK-8 recorded 30 min after morphine injection (at the peak immunoreactivity in the spinal perfusate (fmol/ml). Samples (1.0 ml) effect). From the two bell-shaped dose-response curves, were collected every 30 rain. Downward arrows indicate i.t. injection optimal doses of devazepide and L-365,260 for potenti- of NS (10 /~1) or naloxone (20 p,g), upward arrows indicate s.c. ation of morphine analgesia were calculated to be 100 injection of NS or morphine (5 mg/kg). The two injections were separated by 10 min. ## P < 0.01 as compared to the baseline level and 2.5 ng, respectively, showing that the CCK B blocker of the same group of animals; * P < 0.05, ** P < 0.01 as compared was 40 times more effective than the CCK A blocker on to the corresponding value in the control experiment shown in the a weight basis and 37 times more effective on molar left panel. basis. 150

A B

>~ r-0

m ..J llIIIll .¢: o TITTTTT 0 .625 1.25 2.5 5.0 10 20 0 10 2s 50 loo 200 400

Devazepide (i.t., ng) L-365,260 (i.t., ng) p- n~6 n-6

i t t i t t l t i t t I L i t i 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140

Minute Minute Fig. 2. Changes in tail flick latency after successive i.t. injections of devazepide (A) and L-365,260 (B). Mean tail flick latency+ S.E.M. is plotted against time in rain. Devazepide or L-365,260 was injected in 10 Izl at 20-min intervals. The doses are shown under the arrows.

150 A B 125

g 100 T r- 41 t- T--! o 75 ~e 50 .2 h )-- 25

-25 i h i i 0 20 40 60 0 20 40 60

Minute Minute Fig. 3. Time course of potentiation of morphine analgesia by devazepide (A) and L-365,260 (B). Percent change in tail flick latency is plotted against time in min. Morphine or normal saline (NS) was injected s.c. at time 0, and devazepide, L-365,260 or vehicle was injected i.t. 15 min prior to morphine or NS, as indicated by the arrows. For symbols in panel A: (o) devazepide+morphine (9), (•) devazepide+NS (9), (©) vehicle + morphine (9), ( zx ) vehicle + NS (7). For symbols in panel B: (o) L-365260 + morphine (9), ( • ) L-365260 + NS (9), (o) vehicle + morphine (9), (zx) vehicle + NS (6).

150

4) ~- 100 41 t- O

_3 14. 50 I-"

VIS V 10 25 50 100 200 400 VIS V 1.25 2.5 5 10 20 Devazepide (ng), i.t. L-365,260 (ng), i.t. Fig. 4. Dose-response curves for the potentiation of morphine analgesia by the CCK A antagonist devazepide (A) and the CCK B antagonist L-365,260 (B) in the rat tail flick test. Shown are data recorded 30 min after the injection of morphine (4 mg/kg s.c.). Devazepide, L-365,260 or vehicle was injected i.t. 15 min prior to morphine. Between group comparisons were made with an ANOVA, followed by Duncan test. * P < 0,05 as compared to vehicle/morphine (V) treatment. V/S: vehicle/normal saline control. 151

150 A 125

100 e- T e- 0 75

50 .3 h T/.9/1 ?-----o. I-- 25

t i L t -25 L I I I 0 20 40 60 0 20 40 60

Minute Minute Fig. 5. Time course of potentiation of ohmefentanyl-induced analgesia by devazepide (A) and L-365,260 (B). The first arrow indicates i.t. injection of devazepide, L-365,260 or vehicle. The second arrow indicates i.t. injection of ohmefentanyl (32 ng) or normal saline (NS). The two arrows were separated by 10 min. For symbols in panel A: (e) devazepide + ohmefentanyl (9), ( • ) devazepide + NS (9), (©) vehicle + ohmefentanyl (9), ( zx ) vehicle + NS (6). For symbols in panel B: (e) L-365260 + ohmefentanyl (9), ( • ) L-365260 + NS (9), (©) vehicle + ohmefentanyl (9), ( zx ) vehicle + NS (7).

3.4. Potentiation of ohmefentanyl-induced analgesia by ANOVA for the 60-min observation period). A similar devazepide and L-365,260 administered i.t. experiment was performed with normal saline instead of ohmefentanyl. No significant changes in tail flick Unlike morphine, which is an opioid agonist with a latency were observed in the two groups, indicating relative preference for /x-opioid receptors, ohmefen- that devazepide alone was not analgesic. tanyl is one of most selective/z-opioid agonist available The above experiments were repeated with L- today (Xu et al., 1985; Goldstein and Naidu 1990). As 365,260 (1.25 ng) instead of devazepide (66 ng). The the site of interaction between opioids and cholecys- results are shown in fig. 5B. Analgesia induced by tokinin would be made clearer if both agonists were ohmefentanyl was markedly potentiated by L-365,260, administered at the spinal level, we decided to see both administered via the i.t. route (P < 0.01, ANOVA whether the analgesia induced by i.t. ohmefentanyl for the whole course of observation period), and L- would be potentiated by i.t. injected CCK receptor 365,260 per se was not analgesic. antagonist. Dose-response curves were made for the effect of Two groups of nine rats were given an i.t. injection devazepide (fig. 6A) and L-365,260 (fig. 6B) on ohme- of 66 ng of devazepide or 10/zl of vehicle, followed 10 fentanyl-induced analgesia. The ordinates show the min later by an i.t. injection of 32 ng of ohmefentanyl. peak analgesic effect to occur at 30 min for devazepide As shown in fig. 5A, ohmefentanyl-induced analgesia and 40 min for L-365,260. The optimal doses for poten- was significantly potentiated by devazepide (P < 0.05, tiation of ohmefentanyl analgesia were 66 ng for de-

150 A n-9 If B n-,

e- 1oo (u

.J b. 5O I-

V/8 V 10 33 66 100 200 VI8 V .625 1.28 2.5 5.0 10 20 Devazepide (ng), i.t. L-365,260 (ng), i.t. Fig. 6. Dose-response curves for the enhancement of ohmefentanyl-induced analgesia by the CCK A antagonist devazepide (A) and the CCK B antagonist L-365,260 (B). Peak effects recorded 30 min (A) or 40 min (B) after the injection of ohmefentanyl (32 ng i.t.) are shown. Devazepide, L-365,260 or vehicle was injected i.t. 10 min prior to ohmefentanyl. Between group comparisons were made with an ANOVA followed by Duncan test. * P < 0.05 as compared to the vehicle/ohmefentanyl (V) treatment. V/S: vehicle/normal saline (NS) control. 152 vazepide and 1.25 ng for L-365,260, as determined is also in line with the finding that CCK B receptors are from the two bell-shaped dose-response curves. Thus present in the rat spinal cord (Dourish and Hill, 1987). the CCK B receptor antagonist was 53 times more We have reported recently that the effect of CCK-8 effective than the CCK A receptor antagonist on a on opioid binding depends on the type of opiate recep- weight basis and 49 times more effective on a molar tor studied. Thus CCK-8 decreased the B .... of g- basis. opioid sites and increased the K d of K-opioid sites, without affecting 8-opioid sites (Wang and Han, 1990). Since morphine is an opioid agonist that recognizes 4. Discussion both /x- and g-opioid sites, it was of interest to com- pare the effect of cholecystokinin on the response to In our hands, the analgesic effect induced by a morphine with that of a pure /x-opioid agonist. The moderate dose (5 mg/kg s.c.) of morphine, as tested by novel fentanyl derivative ohmefentanyl has been char- the tail flick latency, usually peaks at 30 min and acterized as one of the most selective/z-opioid agonists subsides at 60 min. Measurement of CCK-8 immunore- available today (Goldstein and Naidu, 1990). The reactivity in spinal perfusate revealed that the time course sults shown in fig. 3-6 indicate that the profile of the of the cholecystokinin reaction almost paralleled the cholecystokinin antagonism of ohmefentanyl-induced opiate effect: a 89% increase within 30 rain and then a analgesia was almost identical to that of morphine, drop to a 32% increase in the second 30-min period. with an optimal L-365,260 dose of 1.25 ng as compared Thus the increased release of CCK-8 can be regarded to 2.5 ng in the case of morphine, and a devazepide/ as a fast negative feedback control for the morphine L-365,260 ratio of 49 as compared to 37 in the case of effect. In another study we measured CCK-8 im- morphine. From a theoretical point of view, it would munoreactivity in the cisternal CSF of rats that had be interesting to look at the effect of cholecystokinin received chronic morphine treatment for 6 days. There on the analgesia produced by dynorphin, an endoge- was a 70% increase (P < 0.01) on the first day, a 39% nous K-opioid agonist, because it has been shown that increase (P < 0.05) on the third day, and a return to CCK-8 exerts a modulating effect on the affinity of the control level on the sixth day, when tolerance was K-opioid sites but causes a down-regulation of/2-opioid fully developed (Pu, Zhou and Han, to be published). sites (Wang and Han, 1990). We found recently that Therefore the cholecystokinin reaction seems to be L-365,260 was much more effective in potentiating the phasic rather than static in nature. That the effect of analgesic effect elicited by high-frequency (100 Hz) morphine was mediated by opiate receptors in the electroacupuncture, which is known to be mediated by spinal cord was shown by the observation that it was the •-opioid agonist dynorphin in the spinal cord, than totally abolished by i.t. naloxone. it was in potentiating the analgesic effect elicited by The notion that the spinal cord is a strategic site for low-frequency (2 Hz) electroacupuncture, which is me- cholecystokinin/opiate interactions is supported by the diated by the ~//z agonist [MetS]enkephalin (Shen, finding that (a) the analgesic effect of systemic mor- Zhou and Han, to be published). phine was markedly potentiated by cholecystokinin Since the analgesic effect induced by ohmefentanyl, blockers administered intrathecaily, and (b) the inter- a pure /x-opioid agonist, was potentiated by a CCK action between opiate agonist and cholecystokinin re- receptor antagonist, it could be reasoned that #-opioid ceptor antagonist took place when both drugs were receptor activation causes cholecystokinin release. administered into the spinal subarachnoid space. However, in vitro (Benoliel et al., 1991) and in vivo According to Dourish et al. (1990), the optimal dose (Rodriguez and Sacristan, 1989) studies have shown for L-365,260 to potentiate morphine analgesia is 0.2 that the iz-opioid agonist DAGOL ([D-Ala 2, MePhe 4, mg/kg i.p. In a 250-g rat it would be equivalent to 50 GlyolS]enkephalin) suppresses CCK-8 release in the /xg per rat. Assuming an even distribution of the drug spinal cord, whereas the #-opioid agonist DTLET ([D- in the whole body, the optimal dose for i.t. injection of ThrZ,LeuS]enkephalin-Thr ~') activates cholecystokinin L-365,260 to potentiate morphine analgesia would be release from rat spinal cord slices (Benoliel et al., approximately 100 ng which was 40 times higher than 1991). These authors suggested that the effect of mor- the optimal dose found in the present study (2.5 ng phine to increase cholecystokinin release is mediated i.t.). A blood-brain barrier therefore seems to exist for mainly through g-opioid receptors. Consequently, in- this non-peptide molecule. terpretation of the potentiation by CCK receptor an- The finding that the CCK B antagonist was 40-50 tagonists of the analgesic effect of morphine is far from times more effective than the CCK A receptor antago- conclusive. nist is in agreement with that reported by Dourish et The old puzzle of the bell-shaped dose-response al. (1990), who found a 20-fold difference between the curve for CCK receptor blockers on opiate analgesia CCK B receptor antagonist L-365,260 (0.2 mg/kg i.p.) was again prominent in the present study, even when and the CCK A antagonist L-365,031 (4 mg/kg i.p.). It both drugs (opiate agonist and CCK receptor antago- 153 nist) were administered i.t. No adequate explanation nists, and selective blockade of /x-dependent inhibition by K can be provided at the present time. In this context, it receptor stimulation, Neurosci. Lett. 124, 204. Chang, R.S.L. and V.J. Lotti, 1986, Biochemical and pharmacologi- is interesting to mention that Wiesenfeld-Hallin et al. cal characterization of an extremely potent and selective nonpep- (1990) reported that morphine analgesia (hot-plate test tide cholecystokinin antagonist, Proc. Natl. Acad. Sci. U.S.A. 83, in rats) was potentiated by PDl34308, a selective an- 4923. tagonist of CCK~ receptor developed by Hughes and Ding, X.Z., S.G. Fan, S.P. Zhou and J.S. Han, 1986, Reversal of his group (1990). The potentiation increased steadily tolerance to morphine but no potentiation of morphine-induced analgesia by antiserum against cholecystokinin octapeptide, Neu- over the dose range of 0.1-1.0 mg/kg. However, it is ropharmacology 25, 1155. important to note that PD134308 per se became anal- Dourish, C.T. and D.R. Hill, 1987, Classification and function of gesic at 1.0 mg/kg or higher, and the suppressive effect cholecystokinin receptors, Trends Pharmacol. Sci. 8, 207. of PD134308 (1 mg/kg) on the rat spinal flexor reflex Dourish, C.T., M.F. O'Neill, J. Coughlan, S.J. Kitchener, D. Hawley was shown to be naloxone reversible. It is obvious that and S.D. lversen, 1990, The selective CCK B receptor antagonist L-365,260 enhances morphine analgesia and prevents morphine a bell-shaped curve would not appear if the PD com- tolerance in the rat, Eur. J. Pharmacol. 176, 35. pound caused opioid release at higher doses, although Evens, B.E., M.G. Bock, K.E. Rittle, R.M. DiPardo, W.L. Whitter, the later possibility needs to be verified experimentally. D.F. Veber, P.S. Anderson and R.F. Freidinger, 1986, Design of In conclusion, the spinal cord is an important area potent, orally effective, nonpeptidal antagonists of the peptide for cholecystokinin/opioid interactions, and the CCK hormone cholecystokinin, Proc. Natl. Acad. Sci. U.S.A. 82, 4918. Faris, P.L., B.R. Komisaruk, L.R. Watkins and D.J. Mayer, 1983, receptors involved are of the CCK B type. Under physi- Evidence for the neuropeptide cholecystokinin as an antagonist ological conditions, the amount of CCK-8 released is of opiate analgesia, Science 219, 310. very low, which may explain the fact that CCK receptor Goldstein, A and A. Naidu, 1989, Multiple opioid receptors: ligand blockers did not significantly influence the basal noci- selectivity profiles and binding site signatures, Mol. Pharmacol. ceptive threshold. Administration of an opioid may 36, 265. Han, J.S., X.Z. Ding and S.G. Fan, 1985, Is cholecystokinin octapep- accelerate the release of CCK-8 in the spinal cord, tide (CCK-8) a candidate for endogenous anti-opioid substrates?, thereby forming a negative feedback control on the Neuropeptides 5, 399. opioid effect. Removal of this negative control resulted H611t, V., I. Haarmann, M.J. Millan and A. Herz, 1987, Prodynor- in an augmentation of the opioid effect. It should be phin gene expression is enhanced in the spinal cord of chronic pointed out that while the tonicity of both opioid and arthritic rats, Neurosci. Len. 73, 90. Hughes, J., P. Boden. B. Costall, A. Domeney, E. Kelly, D.C. anti-opioid activities is low under normal physiological Horwell, J.C. Hunter, R.D. Pinnock and G.N. Woodruff, 1990, conditions, it may shift to a high tonicity status in Development of a class of selective cholecystokinin type B recep- situations such as chronic pain (Przewlocki et al., 1986; tor antagonists having potent anxiolytic activity, Proc. Natl. Acad. H611t et al., 1987). Chronic pain might activate the Sci. U.S.A. 87, 6728. cholecystokinin system via an increased release of en- Itoh, S., G. Katsuura, K. Yoshikawa and J.F. Rehfeld, 1985, Potenti- ation of beta-endorphine effects by cholecystokinin antiserum in dogenous opioids (Przewlocki et al., 1986), and CCK rats, Can. J. Physiol. Pharmacol. 63, 81. blockers given alone could be analgesic. Lehmann, K.A., M. Schlusener and P. Arabatsis, 1989, Failure of proglumide, a cholecystokinin antagonist, to potentiate clinical morphine analgesia: A randomized double-blind postoperative Acknowledgements study using patient-controlled analgesia (PCA), Anesth. Analg. 68, 51. Lotti, V.J. and R.S.L. Chang, 1989, A new potent and se(ective We thank Dr. R.F. Freidinger, Merck, Sharp and Dohme Re- nonpeptide gastrin antagonist and brain cholecystokinin receptor search Laboratories, for providing devazepide and L-365,260, and (CCK B) ligand: L-365,260, Eur. J. and Pharmacol. 162, 273. Dr. Z.Q. Chi, Shanghai Institute of Materia Medica, for ohmefen- Marley, P.D. and J.F. Rehfeld, 1984, Extraction techniques for tanyl. We are also grateful to Dr. J.S. Hong for the gift of CCK-8 gastrin and cholecystokinin in the rat central nervous system, antiserum used in the present study. This work was supported by J.Neurochem. 42, 1515. NIDA Grant DA 03983, and a grant from the National Natural Przewlocki, R., W. Lason, J. Silberring, A. Herz and B. Przewlocka, Science Foundation of China. 1986, Release of opioid peptides from the spinal cord of rats subjected to chronic pain, NIDA Res. Monogr. 75,422. Rattray, M. and J. Debelleroche, 1987, Morphine action on cholecystokinin octapeptide release from rat periaqueductal grey slices: References sensitization by naloxone, Neuropeptides 10, 189. Rodriguez, R.E. and M.P. Sacristan, 1989, In vivo release of CCK-8 Benoliel, J.J., A. Mauborgne, S. Bourgoin, M. Hamon and F. Ces- from the dorsal horn of the rat: inhibition by DAGOL, FEBS selin, 1990, Modulations of the in vitro release of CCK-8-1ike Lett. 250, 215. material from the rat substantia nigra by mu, delta and kappa Tang, J., J. Chou, M. Iadarola, H.Y.T. Yang and E. Costa, 1984, opioid receptor agonists, in: The Neuropeptide Cholecystokinin, Proglumide prevents and curtails acute tolerance to morphine in eds. J. Hughes, G. Dockray and G. Woodruff (Ellis Horwood, rats, Neuropharmacology 23, 715. Chichester) p. 96. Wang, X.J. and J.S. Han, 1990, Modification by cholecystokinin Benoliel, J.J., S. Bourgoin, A. Mauborgne, J.C. Legrand, M. Hamon octapeptide of the binding of mu-, delta- and kappa-opioid recep- and F. Cesselin, 1991, Differential inhibitory/stimulatory modu- tors, J. Neurochem. 55, 1379. lation of spinal cholecystokinin release by /~ and ~ opioid ago- Watkins, L.R., I,B. Kinscheck and D.J. Mayer, 1985, Potentiation of 154

morphine analgesia by cholecystokinin receptor antagonist prog- depress spinal nociceptive reflexes, Proc. Natl. Acad. Sci. U.S.A. lumide, Brain Res. 327, 169. 87, 7105. Wiesenfeld-Hallin, Z., X.J. Xu; J. Hughes, D.C. Horwell and T. Xu, H., J. Chen and Z.Q. Chi, 1985, Ohmefentanyl: A new agonist H6kfelt, 1990, PD134308, a selective antagonist of cholecys- for k~-opiate receptor, Sci. Sin. 28, 504. tokinin type B receptor, enhances the analgesic effect of mor- Yaksh, T.L. and T.A. Rudy, 1976, Chronic catheterization of the phine and synergistically interacts with intrathecal galanin to spinal subarachnoid space, Physiol. Behav. 17, 1031.