0.01-2.0 Mg/Kg, Iv

Drug and Alcohol Dependence, 15 (1985) 353-360 353 Elsevier Scientific Publishers Ireland Ltd.

AN ANALYSIS OF NALTREXONE AND NALOXONE’S POSSIBLE AGONISTIC ACTIONS IN THE DOG

JOSEPH G. WETTSTEIN* and WILLIAM R. MARTIN** University of Kentucky College of Medicine, Department of Pharmacology, Research Facility No. 2, MR-102, 800 Rose Street, Lexington, KY 40536 (U.S.A.) (Received February 5th, 1985)

SUMMARY

Naltrexone (0.01-2.0 mg/kg, i.v.) produced dose-dependent EEG slowing in the conscious dog as did morphine (0.5-8 mg/kg, i.v.) and fentanyl (5-20 pg/kg, i.v.). However, the dose-response curve for nalrexone was not parallel to the morphine or fentanyl dose-response curves. Morphine and fentanyl but not naltrexone also produced dose-dependent miosis and increased the skin twitch reflex latency. When administered into the fourth cerebral ventricle naltrexone (60 pg), morphine (80 pg) and ethylketazocine (30 Erg) produced EEG slowing. Again, naltrexone did not alter the skin twitch latency whereas morphine lengthened it and ethylketazocine reduced it. The pharmacological profiles obtained from different routes of administration indicate that naltrexone is clearly different from morphine, fentanyl and ethylketazocine. However, naltrexone may act as a partial agonist in the production of EEG slowing at a previously unidentified opioid receptor.

Key words: Naltrexone - Naloxone - Partial agonist - EEG slowing - Opioid agonists

INTRODUCTION

The question has arisen as to whether or not naltrexone and naloxone have both opioid agonistic and antagonistic effects. Narcotic-naive human subjects, for example, reported that naltrexone produced sleepiness and

*Present address: Harvard Medical School, New England Regional Primate Research Center, One Pine Hill Drive, Southborough, MA 01772, U.S.A. **To whom all correspondence should be addressed.

0376-8716/85/$03.30 0 1985 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland 354

fatigue [l] . Similarly, naloxone also promoted lethargy and inactivation in normal and manic patients [2,3]. The apparent hypnotic-like effects of naltrexone and naloxone also can become manifest in the EEG. Both drugs decreased the average alpha frequency of the EEG in humans [ 4,5] as did opioid analgesics (morphine, methadone and heroin) [6,8] . These findings indicate that naloxone and naltrexone are capable of producing certain effects similar to those of opioid agonists. It is possible that naltrexone and naloxone produce these EEG and behavioral responses by acting as opioid agonists at sites mediating a ‘hypnotic-like’ effect resulting in the slowing of the EEG. The present study was undertaken to determine the specificity of naltrexone in producing EEG slowing, prolonging the skin twitch reflex latency and inducing miosis by comparing its effects with those of naloxone, morphine, fentanyl and ethylketazocine.

METHODS

Adult female beagle-type dogs, weighing between 8 and 12 kg, were studied. The dogs were maintained on a diet of Purina dog chow and water. All experiments were conducted in conscious dogs between the hours of 0800 and 1700 h. Surgical procedures were carried out under pentobarbital anesthesia and aseptic conditions. Guide cannulas were stereotaxically implanted in the fourth cerebral ventricle. Two stainless steel screw EEG epidural parietal electrodes were placed bilaterally 1 cm from the midline. The guide cannulas and EEG electrodes were cemented in place with dental acrylic. Dogs were allowed to recover from surgery for at least 2 weeks before experiments were begun. Intravenous infusions were given by means of a hypodermic needle (butter- fly infusion set) inserted into a foreleg radial vein attached to a length of tubing and stopcock. Intraventricular (fourth ventricle) injections were given in volumes ranging from 0.1-1.5 ml at a rate of 0.1 ml/min with a calibrated Hamilton syringe operated by a Harvard infusion pump. Three physiological responses, the EEG, pupil diameter and skin twitch reflex latency were monitored during these studies. The EEG was recorded on a Grass polygraph for visual examination. The half amplitude EEG filter settings were 1 and 35 Hz and the sensitivity was 50 pV/cm. EEG activity was integrated using a Grass model 7PlO Summating Voltage Integrator. The output of this integrator is called electrogenesis [ 91. The values obtained from the integrated EEG represents the cumulative millimeters of ramp output obtained per minute. Ramp output of 232 mm/min was obtained when a 0.1 mV, 10 Hz, 50 ms square wave was integrated. The integrated EEG of a conscious dog generated approx. 196 mm of ramp output per minute. This output increased during sleep. In some experiments the EEG 355

was digitized using an analog-to-digital converter (IQS model 401 ESl-2 Universal Signal Interface) and stored on magnetic disk for subsequent power spectral analysis using an IQS model 4000 Biomedical Acquisition and Analysis System in conjunction with an Apple IIe Computer. Pupil diameter was measured from Polaroid photographs of the eye [lo] . Pretreat- ment pupil diameter was approx. 8.6 mm. The skin twitch reflex was evoked and its latency measured according to the method described by Martin et al. ill1 * Physiological responses were measured at 5min intervals for 20 min prior to drug administration in individual dogs. These four measurements were averaged and were considered predrug control. Postdrug or saline responses were measured at 5-min intervals after injection for up to 60 min. Each postdrug response was expressed as a percent of the averaged predrug control value. Data were analyzed using Y-test for both group and paired replicate analysis and procedures for parallel line assays [ 121. The drugs used were naltrexone hydrochloride (Endo), naloxone hydrochloride (Endo), fentanyl citrate (McNeil), morphine sulfate (Merck, Sharp and Dohme) and ethylketazocine methanesulfonate (Sterling-Winthrop). Doses are expressed as the salts.

RESULTS

Figure 1 compares the effects of fentanyl, morphine, naltrexone and naloxone on EEG electrogenesis, skin twitch reflex latency and pupillary diameter. As can be seen both morphine and fentanyl not only produced dose-related changes in these three physiologic phenomena but that all conditions for a valid assay were fulfilled. Both naloxone and naltrexone produced significant increases in electrogenesis and the slopes of their dose-response lines were statistically significant. The slopes of their dose response lines were less than those of morphine and fentanyl. The dose- response curves for naltrexone and naloxone were not parallel to each other and, hence, a valid estimate of the potency ratio for these drugs could not be made. The naloxone dose-effect curve had the lesser slope. The responses produced by 0.125, 0.5 and 2 mg/kg of naltrexone were not significantly different from each other indicating a ceiling response was attained. There was a significant quadratic component to the naloxone dose-response curve; the highest dose of naloxone tested produced a smaller increase in electrogenesis than did intermediate doses. The EEG changes produced by the smallest dose of naltrexone (0.125 mg/ kg) which produced a maximal increase in electrogenesis and a dose of morphine (0.5 mg/kg) which produced an approximately equal effect were nearly the same (Fig. 2). Power spectral analysis of the EEG revealed that these produced an increase in the delta, theta and alpha frequency ranges. 356

260, EEG ELECTROGENESIS

260 SKIN TWITCH REFLEX LATENCY g :;I y /L.p..5~3,.31/ g 140- 8 _.______IOO-

-? 0 20 J 1 0.001 0.0 I 0.1 I IO

DOSE (mg/kg)

Fig. 1. Changes in electrogenesis, skin twitch reflex latency and pupil diameter produced by i.v morphine, fentanyl, naltrexone and naloxone. The physiological responses are expressed as a percent of predrug (% control) effects and each point represents the mean peak response made in five or six dogs. The mean peak effect of saline f 95% confidence limits is represented by a solid horizontal line and dashed lines, respectively. Relative potencies (RP) and their 95% confidence limits are presented for those comparisons for which a valid bioassay was obtained. Two independent groups of five (morphine and fentanyl) and six (naltrexone and naloxone) dogs each are represented. *indicates that the mean effect of drug at the indicated dose was significantly different from that of saline (P < 0.05, paired t-test). Naltrexone ( v-0); naloxone ( -0); morphine (e----e); fentanyl (m-).

The effects of naloxone and naltrexone on pupillary diameter (Fig. 1) were small, however there was a trend for them to produce miosis. Changes in behavior were also noted and characterized using a 5-point ordinal scale (agitated, restless, awake, drowsy and asleep). A significant negative regres- sion was obtained between electrogenesis on the one hand and pupillary diameter and degree of arousal on the other. Morphine and fentanyl produced a marked, dosedependent prolongation in the latency of the skin twitch reflex (Fig. 1). The effects of naltrexone 357

3- ‘kg)

2- zl % :: l-

o- I I I I

3- morphine (0.5 mg/kg)

saline

2- ; 3 : l-

O-

0.5 10 20 30 Hz

Fig. 2. Mean power spectral analyses of the EEG of three dogs taken for 15 min immedi- ately after i.v. administration of naltrexone and morphine. Power is expressed as relative decibels squared.

and naloxone on the skin twitch latency were not significant and were small compared to those of morphine or fentanyl. Intraventricular injection of naltrexone (60 pg in 1.5 ml) increased electrogenesis and decreased pupil diameter but did not alter the latency of the skin twitch reflex (Table II). The same dose of naltrexone in a smaller volume (0.1 ml) also increased electrogenesis. The increases in electrogenesis in the two experiments were not significantly different from each other (120 f 5% vs. 137 + 8%, for the 1.5 and 0.1 ml vols., respectively t = 1.4, d.f. = 7). As in the i.v. studies, the dogs were less alert or aroused after intraventricular naltrexone when compared to control experiments. Intraventricular injection of ethylketazocine (30 Erg) significantly increased electrogenesis, decreased pupil diameter and shortened the latency of the skin twitch reflex (Table I). Intraventricular injection of morphine (80 pg) also increased electrogenesis but increased rather than decreased 358

TABLE I MEAN EFFECTS OF NALTREXONE, ETHYLKETAZOCINE (EKC) AND MORPHINE INJECTIONS INTO THE FOURTH CEREBRAL VENTRICLE OF THE CONSCIOUS DOG The physiological responses, averaged over a 40-min period following injection, are expressed as a percent of predrug (% control) effects. Volume and rate of injection were: naltrexone, 0.1 or 1.5 ml in 1 or 15 min, respectively; EKC, 1.5 ml in 15 min; morphine, 0.1 ml in 1 min; and control, 0.1 ml in 1 min - saline, 1.5 ml in 1 min - Ringer’s lactate. The number of animals used for each treatment and response is indicated in parenthesis under each value.

Physiological Control Naltrexone EKC Morphine responses (60 wg) (30 fig) (89 pg) (% control)

EEG electrogeneais 102 * 3 (10) 123 t 4* (9) 152 r 14* (5) 122 + 8* (5) Pupil diameter 94 + l(l0) 84 * 2* (6) 75 * 5* (5) Skin twitch latency 106 * 5 (10) 107 + 5 (9) 79? 5*(5) 141 +_10* (5)

*Significantly different from control values (P < 0.05; paired t-tests for all EKC responses and naltrexone-pupil diameter; group t-test for the remainder of comparisons).

the latency of the skin twitch reflex (Table I). The animals, after receiving ethylketazocine or morphine, appeared sedated or asleep during most of opioids (p-endorphin, met- and leu-enkephalin) increased EEG power in the

DISCUSSION Intravenous infusions of naltrexone and naloxone produced dose- dependent and statistically significant increases in EEG electrogenesis in the conscious dog. Naltrexone was more effective than naloxone in this regard. Naltrexone also produced a slow wave EEG after an injection into the fourth cerebral ventricle. The effects of naltrexone and naloxone on the EEG in the dog are generally similar to those observed in man [ 4,5]. In the dog, naltrexone increased delta, theta and alpha EEG activity. These changes were similar to those produced by an equi-effective dose of morphine and were associated with sedation. Naltrexone and naloxone also promoted sleepiness and lethargy in man [l-3] . Furthermore, endogenous opioids (fl-endorphin, met- and leu-enkephalin) increased EEG power in the delta, theta and alpha ranges in the rat [ 131. Naltrexone and naloxone are competitive antagonists at mu and kappa opioid receptors [14,15]. Naltrexone is about three times more potent than naloxone in precipitating abstinence in morphine-dependent dogs [ll] . In the present study, naltrexone was about twelve times more potent than naloxone in increasing electrogenesis which is a greater potency ratio 359

than their relative potency as antagonists in the morphine dependent dog. In addition, the naltrexone and naloxone electrogenesis dose-response curves were not parallel as would have been expected if their EEG effects were due to the antagonism of an endogenous-alerting opioid agonist. Further- more, intraventricular injection of naltrexone, morphine and ethylketazo- tine and intravenous infusions of fentanyl and morphine also increased electrogenesis. In addition; naltrexone and naloxone clearly exhibited ceili* effects on electrogenesis which were significantly different from the responses produced by large doses of morphine and fentanyl. These findings suggest that the EEG effects produced by naltrexone and naloxone are not a consequence of it acting as a competitive antagonism of an endogenous opioid ligand which is producing arousal and EEG desynchronization. Further these studies indicate that naltrexone is acting at a brain stem site in increasing EEG electrogenesis. If naltrexone’s agonistic effects were mediated primarily by its actions at mu and kappa receptors it should have had a pharmacological profile similar to those of morphine, fentanyl or ethylketazocine. Yet, intravenous infusions of morphine and fentanyl prolonged the skin twitch latency whereas naltrexone and naloxone did not. Furthermore, ethylketazocine and morphine produced hyperalgesia and analgesia, respectively, after intraventricular administration and naltrexone had no significant effect at this site. Also, fentanyl and morphine but not naltrexone produced analgesia after microinjection into the periaqueductal gray of the dog [ 161. Thus, naltrexone does not alter the nociceptive threshold at sites where mu and kappa agonists are active. Unlike morphine and fentanyl most intravenous doses of naltrexone and naloxone did not have significant effects on pupil size. However, the highest dose of naltrexone (2 mg/kg) produced miosis and the mean pupillary diameter after all naltrexone doses was less than after the saline control. Correlation analysis indicat%d that decreases in pupil size covaried with elevated electrogenesis. Miosis was also observed after intraventricular administration of naltrexone and ethylketazocine, an opioid which causes sedation [ 171. This suggests that the small degree of miosis produced by naltrexone may be due to a sedative or hypnotic effect as a consequence of it acting on the medulla. Although naltrexone and naloxone had some effects like the opioid agonists, the mechanisms by which these changes occur seem to be different for the two groups of drugs. The pharmacological profiles of the drugs obtained from different routes and sites of administration indicate that naltrexone and naloxone are clearly different from morphine, fentanyl and ethylketazocine. Thus, naltrexone and naloxone do not appear to be acting as agonists at mu or kappa sites. If they are acting as agonists in augmenting electrogenesis and producing a slow wave EEG, it is likely at opioid receptors which are not of the mu or kappa type. 360

ACKNOWLEDGEMENT

This study was supported by the University of Kentucky Tobacco and Health Research Institute and was done in part to fulfill requirements for a Ph.D.

REFERENCES

1 J.H. Mendelson et al., Psychoneuroendocrinology, 3 (1979) 231. 2 L.L. Judd et al., Arch. Gen. Psychiatry, 37 (1980) 583. 3 S. File and T. Silver&one, Psychopharmacologia, 73 (1981) 353. 4 J. Volavka et al., Pharmakopsychiatr., 12 (1979a) 79. 5 J. Volavka Life Sci., 25 (1979b) 1267. 6 H.L. Andrews, Psychosom. Med., 5 (1943) 143. 7 W.R. Martin et al., Arch. Gen. Psychiatry, 28 (1973) 286. 8 J. Volavka et al., Arch. Gen. Psychiatry, 30 (1974) 677. 9 L. Goldstein and C. Munoz, J. Pharmacol. Exp. Ther., 132 (1961) 354-353. 10 W.G. Marquardt, W.R. Martin and D.R. Jasinski, Int. J. Addict., 2 (1967) 301. 11 W.R. Martin et al., J. Pharmacol. Exp. Ther., 189 (1974) 759. 12 D.J. Finney, Statistical method in biological assay, C. Griffin and Co. Ltd., London, 1978. 13 L.H. Miller, Pharmacol. Biochem. Behav., 15 (1981) 845. 14 H. Blumberg and H.B. Dayton, Naloxone, naltrexone, and related noroxymorphones, in: Narcotic Antagonists, M.C. Braude et al., (Eds.), Raven Press, New York, pp. 33- 43,1974. 15 W.R. Martin, et al., J. Pharmacol. Exp. Ther., 197 (1976) 517. 16 J.G. Wettstein, S.G. Kamerling and W.R. Martin, Sot. Neurosci. Abstr., 8 (1982) 229. 17 P.E. Gilbert and W.R. Martin, J. Pharmacol. Exp. Ther., 198 (1976) 66.