Kappa Opioids in Rhesus Monkeys. II. Analysis of the Antagonistic Actions of Quadazocine and fl-funaltrexam ine

Transcription

1 /87/2422-O421$2.O/ THE Jouiw*.. or Piunco.ocv AND EXPERMENTAL TnswsuTtcs CopyrightO 1987 by The American Society for Pharmacology and Experimental Therapeutics Vol. 242, No. 2 Printed in U.S.A. Kappa Opioids in Rhesus Monkeys.. Analysis of the Antagonistic Actions of Quadazocine and fl-funaltrexam ine UNDA A. DYKSTRA,2 DEBRA E. GMEREK,3 GAL WNGER and JAMES H. WOODS Departments of Pharmacology (D.E.G., G.W., J.H.W.) and Psychology (J.H.W.), UnWers!ty of Michigan, Mn Arbor, Michigan and Departments of Psychology and Pharmacology (L.A.D.), University of North Carolina, Chapel Hill, North Carolina Accepted for publication Apill 14, 1987 ABSTRACT Kappa opioid agomsts (e.g., ethylketazocine, bremazocine, tifluadom and U-5,488) have been shown to possess a unique profile of effects in both rodents and monkeys. n the rhesus monkey, this includes increases in urinary output, increases in tail-withdrawal latencies from warm water, muscle relaxation and stupor and unique discriminative stimulus properties, as well as withdrawal signs that are distinct from those of morphine. (See previous and subsequent articles in this journal, Dykstra et at., 1987 and Gmerek et at., 1987, as well as Tang and Collins, 1985; Woods et at., 1979; Woods and Gmerek, 1985.) Although unique behavioral profiles among different opioids suggest that these compounds are acting at different receptor types, additional evidence can be obtained by quantifying the potency of a given antagonist to attenuate the effects of agonists thought to have selective activity at different receptor types. The dose-ratio analysis of Schild (1947) was devel- Received for publication April 1, This work was supported by U.S. Public Health Service Grants DA 154, DA 254 and DA 2749 from the National nstitute on Drug Abuse. 2 Recipient of Research Scientist Development Award DA-33 from the National nstitute on Drug Abuse. 3 Present address: Warner-Lambert Pharmaceutical Research Division, 28 Plymouth Road, Ann Arbor, M n rhesus monkeys, kappa oploid agonists have been shown to 1) increase urinary output, 2) increase tail-withdrawal latencies from warm water and 3) produce distinct discriminative stimulus effects. n order to explore further the relation between these effects and activity at the kappa opioid receptor type, the antagonist activity of quadazocine against several kappa oploid agofists was examined with the tail-withdrawal and dwg-discrimination procedures. Quadazocine dose dependently antagonized the icreases in t-withdrawal latency produced by the kappa agonists bremazoane, ethylketazocine and U-5,488, as well as the discriminative stimulus effects of these drugs. The dose-ratio analysis of Schild revealed apparent pa2 values for quadazocine in combination with bremazocine, ethylketazocine and U-5,488 of 6.1, 6.4 and 6.4, respectively, with the tail-withdrawal procedure and 6.3, 6.4 and 6.1, respectively, with the drug-discnmination procedure. Quadazocine also antagonized the effects of a mu agonist (morphine) in the tail-withdrawal procedure, and the apparent pa2 value for these data was 8.2. The activity of the mu-selective alkylating agent, 9-funaltrexamine (fl-fna), was examined alone and in combination with the kappa agonist ethylketazocine in the urinary-output, tail-withdrawal and drugdisaimination procedures. At about 3 to mm postinjection, fl-fna alone produced ethylketazocine-appropriate responding under the drug-discrimination procedure and increased urine output but did not increase tail-withdrawal latencies. At 24 to 48 hr postinjection, fl-fna did not antagonize effects of ethylketazocine in any ofthe three procedures. Under the same conditions of administration, fl-fna did, however, antagonize the effects of mu agonists in the tail-withdrawal procedure and in the drugdiscrimination procedure. oped originally to quantify interactions such as these in vitro and also has been used successfully to characterize interactions between different groups of opioid agonists and antagonists in vivo (e.g., Takemori et at., 1969; Smits and Takemori, 197; Yaksh and Rudy, 1977; Teal and Holtzman, 198). n the present study, the dose-ratio analysis was used to quantify the activity of the long-acting antagonist quadazocine (WN 44, 441) (Ward et at., 1983; Wood et a!., 1984) against morphine, a mu agonist, and four kappa agonists, bremazocine, ethylketazocine, tifluadom and U-5,488. (See Dykstra et at., 1987 for a description of these.) The antagonist effects of quadazocine were examined in the tail-withdrawal procedure described previously by Dykstra and Woods (1986). Quadazocine also was examined in combination with several kappa agonists in monkeys trained to discriminate ethylketazocine from saline. Additional evidence to infer activity at different receptor types can be obtained with selective alkylating agents such as f-fna (Portoghese et at., 198). As an antagonist, $-FNA has a long duration of action and appears to be selective for mu as compared with kappa agonists. For example, fl-fna has been shown to antagonize the effects of morphine and D-Ala2,Met5- Downloaded from jpet.aspetjournals.org at ASPET Journals on May 1, 216 ABBREVATON: -FNA, -funaltrexamine hydrochlonde. 421

2 422 Dykstra at al. enkephalin, but not of nalorphine, in acetic acid-induced stretching tests in mice (Ward et at., 1982b). n addition, - FNA also precipitates withdrawal in morphine-dependent monkeys. On the other hand, fi-fna does not antagonize the acute behavioral effects of ethylketazocine or Mr 233 in rhesus monkeys (Gmerek and Woods, 1985), or does it precipitate withdrawal in U-5,488-dependent monkeys (Gmerek et at., 1987). Agonist properties of /3-FNA include a naloxone-antagonizable inhibition of the electrically stimulated mouse vas deferens preparation (Takemori et at., 1981; Ward et at., 1982b) and analgesic activity in stretching tests but not in tail-ifick tests in mice. This differential analgesic effect is associated with kappa agonists (Ward et at., 1982b). n the present study, the effects of ethylketazocine were examined after treatment with -FNA. Where appropriate, the effects of a mu agonist (morphine or alfentanil) also were examined after treatment with -FNA. The effects of fi-fna alone were examined with the urine-output, tail-withdrawal and drug-discrimination procedures to assess the kappa agonist profile of f3-fna in rhesus monkeys. Tall Withdrawal Methods Animals. Six individually housed male and female rhesus monkeys weighing between 4 and 8 kg were used. Monkeys were allowed free access to food and water in their home cages and received fresh fruit and Vitamins regularly. Procedure and apparatus. The tail-withdrawal procedure and apparatus were identical with that described in Dykstra et al (1987). Briefly, the lower 1 to 12 cm of each monkey s tail were immersed into a thermos of water maintained at 5C, and the time it took for the animal to withdraw its tail from the water was determined. f a monkey did not remove its tail from the water within 2 see, the thermos was removed by the experimenter and the latency was assigned a value of 2 sec (2 sec cutoff time). Experimental sessions were divided into several trials spaced 3 mm apart and conducted approximately twice weekly, between 9 A.M. and NooN. Dose-effect determinations. Dose-effect curves were obtained by administering cumulative doses of morphine, bremazocine, ethylketazocine or U-5,488 alone or in the presence of quadazocine. n the cumulative dosing procedure, monkeys received an injection after each trial (i.e., once every 3 mm). The amount of compound administered after each trial increased the total cumulative dose by either one-fourth or one-half log unit. When quadazocine was examined in combination with the agonists, it was administered immediately before the first trial of the experimental session. Dose-effect curves were also obtained for 3-FNA alone and for ethylketazocine or morphine after fi-fna treatment. These were obtained with a single administration of fi-fna in the following manner. First, a dose-effect curve was obtained for - FNA alone by administering cumulative doses of $-FNA. Twenty-four hours after -FNA administration, a dose-effect curve was obtained for ethylketazocine and, 48 hr after -FNA administration, a dose-effect curve was obtained for morphine. Data analysis. Tail-withdrawal latencies were determined on each trial after doses of each agonist. The latencies were then divided by the cutoff time (i.e., 2 see) and expressed as the percentage of the cutoff value. Data from at least three monkeys were pooled, and regression lines for dose-response curves were calculated by the method of least squares (Tallarida and Murray, 1981). From these lines, the dose of agonist required to increase tail-withdrawal latencies to 1 sec (i.e., 5% of the cutoff values) was determined (A). Dose ratios were obtained for each agonist in the presence of quadazocine by dividing the (A dose of agonist in the presence of a given dose of quadazocine] by the [A dose of agonist when given alone]. The log of dose ratio - 1 was then expressed as a function of the negative log of the molar VoL 242 dose of quadazocine salt. Regression lines were determined for these points; slope, pa2 values and CL were determined (Tallarida and Murray, 1981). Discriminative Stimulus Characteristics Animals. Three male rhesus monkeys were maintained at 85% of their ad thitum feeding weights according to the procedure described previously (Dykstra et at., 1987). All three monkeys had extensive histories of ethylketazocine discrimination. Apparatus and procedure. The monkeys were placed in chambers with two monkey-response levers and trained to respond differentially On the two levers depending on whether they received an injection of ethylketazocine (.32 ) or saline as described in detail in Dykstra et a!. (1987). Each daily experimental session was divided into several 15-min trials. Each trial started with a b-mm blackout period, followed by the onset of a light in the chamber. n the presence of the stimulus light, 1 consecutive responses on the appropriate lever in 5 mm or less resulted in the delivery of 1 3-mg food pellets. Each trial was followed by as many as five additional 15-mm trials to comprise a daily session. Dose-effect determinations. Dose-effect curves were obtained by administering saline before the first trial and administering increasing doses of the agonists before each subsequent trial. n this cumulative dosing procedure, the administered dose increased by one-fourth or one-half log unit with each injection until the monkey made 8% or more of its responses on the ethylketazocine-appropriate lever, until response rates decreased to less than half the rate that occurred on the first (saline) trial or until six trials had been given. When quadazocine was examined in combination with the kappa agonists, it was administered 3 rain before the initial trial of a test session. Dose-effect curves were obtained using ethylketazocine, bremazocine, U-5,488 and -FNA. Twenty-four hours after evaluation of cumulative doses of $- FNA, a dose-effect determination was made for ethylketazocine; 48 hr after $-FNA administration, a similar determination was made for alfentanil Data analysis. The percentage of total responses made on the ethylketazocine-appropriate lever and the overall rate of responding were determined for each trial. Data from all monkeys were pooled, and regression lines for dose-response curves were calculated by the method of least squares (Tallarida and Murray, 1981). From these lines, the dose of agonist required to produce 5% ethylketazocineappropriate responding was determined (A). Dose ratios and pa2 values were calculated as described above. Urine Output Animals. Five individually housed male rhesus monkeys weighing between 4 and 8 kg were used. Monkeys were maintained according to the protocol described in Dykstra et at. (1987). Apparatus and procedure. Monkeys were placed in modified chairs from which urine output could be measured as described in Dykatra et al (1987). Urine was collected at 45-mm intervals during the daily 3-hr experimental session. Dose-effect determinations. -FNA or ethylketazocine was administeredin a single s. c. dose at the beginning ofselected experimental sessions. Ethylketazocine also was examined after -FNA administration. n this case, a single dose of -FNA (1 ) was administered on day 1 and, on the subsequent 2 days,.3 of ethylketazocine was administered at the beginning of the experimental session. Data analysis. The volume of urine collected from each animal during the 3-hr session in the presence of drug was compared with the mean 3-hr urine volume collected during their respective saline control sessions. Values were compared by analysis of variance, followed by Tukey pairwise comparison. Compounds Morphine sulfate (Mallinckrodt, nc., St. Louis, MO) and alfentanil HC1 (Janssen Pharmaceuticals, Beerse, Belgium) were examined as were several kappa agonists, including bremazocine methanesulfonate Downloaded from jpet.aspetjournals.org at ASPET Journals on May 1, 216

3 1987 Kappa Oplolds 423 (Sandoz Corp., Basel, Switzerland), ethylketazocine methanesulfonate (courtesy of Dr. W. Michne, Sterling-Winthrop Research nstitute, Rensselaer, New York) and U-5,488 trans-3,4-dichloro-n-methyl-n- [2-(1-pyrrolidinyl) cyclohexyl]benzeneacetamide methanesulfonate, hydratel (courtesy of Dr. J. Collins, Upjohn Co., Kalamazoo, M). n addition, dose-effect curves for moat of the above compounds were examined in combination with several doses of the opioid antagonist quadazocine methanesulfonate (WN 44, 441, Sterling-Winthrop Research nstitute). Ethylketazocine, morphine and alfentanil also were examined in combination with -FNA (courtesy of Dr. D. Zimmerman, Eli Lilly Co., ndianapolis, N). Doses of all of the compounds were Calculated as the salt and dissolved in sterile water, to which a small amount of lactic acid was added, if necessary. All injections were delivered s. c. to the back in a volume ofo.1 to 1. mt/kg. Tall Withdrawal 8 MORPHNE / - U- 2 - U-.- -J x 4 too z 8 Latencies U5,488 / Resufts Effects of morphine and several kappa agonists in combination with quadazocine. Figure 1 shows the effects of doses of morphine, bremazocine, ethylketazocine and U- 5,488 alone and in the presence of several doses of quadazocine on tail-withdrawal latencies from 5#{176}C water in three monkeys. All four drugs produced dose-dependent increases in tail-withdrawal latencies. When doses ofquadazocine were administered 3 mm before morphine or the kappa agonists, the dose-effect curves for each agonist were shifted to the right. For morphine, a dose of.1 of quadazocine produced approximately a one-half-log-unit shift in the dose-effect curve, and a dose of.1 of quadazocine shifted the morphine dose-effect curve approximately one log unit to the right. n contrast,.1 of quadazocine produced very small shifts in the bremazocine, ethylketazocine and U-5,488 dose-effect curves, and 1. to 1. of quadazocine were required to shift the too 8 i-.3 -i--.3 BREMAZOCNE ETHYLK ETA ZOCNE dose-effect curve for the kappa agonists by about 1 log unit or more. A comparison between bremazocine and ethylketazocine in combination with 1 of quadazocine reveals approximately a one-and-one-half-log-unit shift in the bremazocine curve and a larger (2-log-unit) shift in the ethylketazocine curve. An incomplete determination of the effects of 1 of quadazocine in U-5,488 discrimination are shown because the limited solubility of U-5,488 prevented administration of higher doses. Data for U-5,488 after administration of 1 mg kg of quadazocine were not used in the pa2 analysis described below. Apparent pa2 value for quadazocine in combination with morphine, bremazocine, ethylketazocine or U- 5,488. Schild plots were constructed using dose-ratio values obtained from regression lines calculated from the data in figure 1 and are shown in figure 2. The apparent pa2 value is the point at which the regression line intercepts the x-axis and the dose ratio equals two. The apparent pa2 for morphine (±95% CL) was 8.2 (±.64), whereas it was 6.1 (±.2), 6.4 (±.16) and 6.4 (±.14) for bremazocine, ethylketazocine and U-5,488, respectively. i9-fna in combination with morphine or ethylketazocine. Figure 3 shows the effects of 1 of -FNA and of doses of ethylketazocine or morphine obtained 24 to 48 hr subsequent to this dose of fi-fna on tail-withdrawal latency. For comparison, a dose-effect curve for ethylketazocine or morphine obtained in the absence of fl-fna also is shown. t can be seen that 1 of fi-fna had very little effect alone. Tail-withdrawal latencies were approximately 1.6 sec or 8% of the 2-sec cutoff value. n the absence of fl-fna, both ethylketazocine and morphine produced dose-dependent increases in tail-withdrawal latencies from 5#{176}Cwater. When the effects -c 1 a c.1 a e-&.3 3. Fig. 1. Tad-wfthdrawal latenoes shown as a percentage of the cutoff value (i.e., 2 sec) after J cumulative doses of morphine, bremazocine, ethylketazocane or 11-5,488 alone(#{149}) or n cornbinatlon wfth doses of quadazocine., S,, #{149}, & A, represent cornbkations of each agonist with.1,.1..1,.3, 1. or 1. of quadazocine, respectively. Each point is the mean of one determination in monkeys 17, 172 and 179. ofethylketazocine were examined 24 hr after fi-fna treatment, there was only a very small rightward shift in the ethylketa- Downloaded from jpet.aspetjournals.org at ASPET Journals on May 1,

4 424 Dykstra at al Vol. 242 MORPHNE 2 BREMAZOCNE zocine dose-effect curve. n contrast, when the effects of morphine were examined after -FNA treatment, the morphine dose-effect curve was shifted at least one and one-haiflog units to the right. U5,488 ETHYLKETAZOCNE 8-J:ll U- U. -i z. 8 2 o 8 2 -FNA.1 MORPHNE $-FNA ---r mg/ kg 6.1 ToT T, b. 1O.O - LOG DOSE OF QUADAZOCNE (moles/kg) Fig. 2. Schlld plot obtained for quadazocsne in combination wfth morphine, brernazocine, ethylketazocine or U-5,488 under the tail-withdrawal proosdure. Ordate: kg of dose ratio - 1. Abscissa: negative log of the molar dose of quadazocine. (Note: The corresponding dose Wi nnigrams per kikgram is indicated by the following symbols:,.1 ; S,.1 ;,.1 ; U,.3 ; L, 1. ; A, 1..) Points were calculated from regression lines constructed from the data in figure 1. Dose ratios were obtained by dmding the [Ado dose of agonist in the presence of a given dose of quadazocine] by the [A dose of the agonist alone]. The slopes of the regression lines shown here are -.6, -1., -1.4 and -1.1 for morphine, bremazocine, U- 5,488 and ethylketazocine, respectively.. Fig. 3. Tail-withdrawal latencies shown as a percentage of the cutoff value(i.e., 2 sec) after cumulative doses of ethylketazocine or morphine alone or after 1 of 1-FNA. The ethylketazocine plus $-FNA curve was obtained 24 hr subsequent to fi-fna administration, and the morphine plus 9-FNA curve was obtained 48 hr subsequent to i9-fna administration. S, effects ofethylketazocine or morphinealone;, effects of ethylketazocine or morphine either 24 or 48 hr subsequent to f3-fna. respectively. Each point is the mean of one determination n at least four monkeys and brackets indicate S.E. The point at $-FNA represents the effects of 1 of 19-FNA alone. 3. Discriminative Stimulus Proprties Effects of several kappa agonists alone and in combination with quadazocine. Figure 4 (top) shows the effects of ethylketazocine, bremazocine and U-5,488 alone and in combination with doses of quadazocine on percentage of ethylketazocine-appropriate responding in three monkeys trained to discriminate between ethylketazocine and saline. Each of the kappa agonists produced dose-dependent increases in ethylketazocine-appropriate responding. When quadazocine was administered before each agonist, the dose-effect curve was shifted to the right in a dose-dependent manner. t can be seen that a dose of.1 of quadazocine produced a very small rightward shift in the bremazocine and ethylketazocine doseeffect curves. t did not shift the U-5,488 dose-effect curve to the right (data not shown). A larger dose of quadazocine (1. ) produced shifts in the ethylketazocine and bremazocine curves of approximately 1 log unit. This dose of quadazocine produced smaller and more variable shifts in the U-5,488 doseeffect curve. Larger doses ofquadazocine could not be examined because they decreased rates of responding when administered alone. Apparent pa2 values for quadazocine. A Schild plot was constructed using dose-ratio values obtained from regression lines Calculated from the data in figure 4. Plots for bremazocine, ethylketazocine and U-5,488 are shown on the bottom of figure 4. The apparent pa2 values (±95% CL) for ethylketazocine, bremazocine or U-5,488, respectively, were 6.4 (±.34), 6.3 (±1.1) and 6.1 (no CL obtained because only two doses of quadazocine produced shifts in the U-5,488 curve). 8-FNA in combination with alfentanil or ethylketazocine. Figure 5 shows the effects of 1 of 9-FNA alone on ethylketazocine-appropriate responding and on rate of responding. The effects of ethylketazocine and the mu agonist alfentanil obtained 24 to 48 hr subsequent to this dose of fi- FNA also are shown. For comparison, dose-effect curves for ethylketazocine or alfentanil obtained in the absence of 9-FNA are presented. All monkeys responded on the ethylketazocine-appropriate lever 3 to mm after administration of 1 of -FNA; however, the lowest dose that produced ethylketazocine-appropriate responding differed among monkeys, ranging between 1. and 1.. Twenty-four to 48 hr later, fi-fna did not produce ethylketazocine-appropriate responding in any monkey. When ethylketazocine was administered alone, it produced dose-dependent increases in ethylketazocine-appropriate responding, whereas alfentanil did not. Administration of -FNA 24 hr before ethylketazocine did not shift the ethylketazocine dose-effect curve to the right. ndeed, it may have produced a slight shift to the left. -FNA given 48 hr earlier attenuated the rate-decreasing effects of alfentanil, shifting this curve onehalf log unit to the right. Urine Output The effects of -FNA and ethylketazocine alone on urine output are shown in table 1. As demonstrated in Dykstra et at. (1987), a dose of.3 of ethylketazocine alone increased urine output significantly over controls. When 1 of - Downloaded from jpet.aspetjournals.org at ASPET Journals on May 1, 216

5 1987 Kappa Oplolds OOr T 2. r!.:n ETHYLKETAZOCNE BREMAZOCNE U5, o _1_ -J E 8,.- 2 z C.,. Q ETHYLKETAZOCNE.+f L j- J..L.L A- 2Fmg 1kg ALFENTAN B-FNA B-FNA 3 1 +lhq+24pw +481v i L S ALONE +B-FNA.3 F LOG DOSE QUADAZOCNE (moles 1kg) Fig. 5. Percentage of ethylketazoclne-approprlate responding (top) or percentage of rate of responding (bottom) after cumulative doses of ethylketazocine or alfentanhl alone or after treatment with 1 of fl-fna. The ethylketazoane plus 3-FNA curve was obtained 24 hr subsequent to 9-FNA administration, and the alfentanil plus 13-FNA curve was obtained 48 hr subsequent to 3-FNA administration. #{149}. the effects of ethftetazocine or alfentanil alone;, the effects of ethlketazocine or alfentanil either 24 or 48 hr subsequent to $-FNA, respectively. The points at 9-FNA represent the effects of 1 of -FNA alone at approximately 1, 24 or 48 hr subsequent to administration. Each point s the mean of one determination in at least three monkeys, and vertical lines indicate S.E. vertical lines. FNA were administered immediately before the experimental session, it also increased urine output; increases in urine output were not observed 24 hr after $-FNA administration (data not shown). Table 1 also shows the effects of.3 of ethylketazocine in combination with fi-fna. Administration of fi- FNA 24 hr before ethylketazocine resulted in increases in urine. - l TABLE 1 Three-hour urine output values for ft-fna, alone with ethylketazocine eeemnt Fig. 4. Top graphs: Percentage of ethylketazodne-appropnate responding after cumulative doses of ethylketazocine, brernazocine or U-5,488 alone (#{149}) or in combination with doses of quadazocine. The ordinate indicates the number of responses omitted on the ethylketazoone-approphate lever, expressed as a percentage of the total session responses., #{149}, L represent combinations of each agonist with.1,.3 or 1. of quadazocine, respectively. Each point is the mean of one determination in three monkeys. Bottom graphs: Schild plot Obtained for quadazodne in combination with ethylketazocine, brernazocine or U-5,488 under the drug-discrimination procedure. Ordinate: log of dose ratio - 1. Abscissa: negative log of the molar dose of quadazocine. Symbols are as indicated in the legend to figure 2. The elopes of the regresslon lines shown hereare -.8, -1. and -1.2 for ethylketazocine, bremazocine and U-5,488, respectively. and hr rng/kgs.c. ml Saline Ethylketazocine (.3) 9-FNA (1) 5-FNA (24) Ethylketazoclne (.3) -FNA (48) Ethylketazocine (.3).P <.5; P <.1 vs. saline control. n combination ±SE.) 24±2 83 ± 26* 85 ± 47* 127 ± 23** 94 ± 7** output that were even greater than those observed with ethyl- ketazocine alone. When -FNA was administered 48 hr before ethylketazocine, it produced no marked changes in the effects of ethylketazocine. Discussion n the present study we examined the extent to which the analgesic and discriminative stimulus effects of several kappa agonists might be mediated by activity at the kappa opioid receptor type. An analysis by pa2 was used to evaluate the interaction between several kappa agonists and the opioid antagonist quadazocine. n the tail-withdrawal procedure, the pa2 values for bremazocine, ethylketazocine and U-5,488 were 6.1, 6.4 and 6.4, respectively, whereas the pa2 for morphine was 8.2. These values indicate that kappa-induced analgesia is far less sensitive to blockade by quadazocine than is that of morphine. Previous investigations in which pa2 values have been compared for mu and kappa agonists have consistently found lower pa2 values for kappa compared with mu agonists. For example, Schmauss and Yaksh (1984) found that the pa2 for naloxone antagonism of morphine on the stretching response was 7.19, whereas it was 5.92, 6.9 and 6.39 for ethylketazocine, bremazocine and U-5,488, respectively. Von Voigtlander et at. (1983) report similar. differences for the effects of bremazocine, ethyl- Downloaded from jpet.aspetjournals.org at ASPET Journals on May 1, 216

6 426 Dyksva at al Vol. 242 ketazocine and U-5,488 as compared with those of morphine in the mouse. Differential sensitivity toward naloxone antagonism has also been reported for the analgesic effects of cyclazocine or ethylketazocine as compared with morphine in rhesus and squirrel monkeys (Yaksh and Rudy, 1977; Dykstra, 1983). The pa2 values obtained for the kappa agonists under the drug-discrimination procedure were similar to those obtained under the tail-withdrawal procedure, i.e., 6.3, 6.4 and 6.1 for bremazocine, ethylketazocine and U-5,488, respectively. t is important to note that these values were obtained within a very narrow range of antagonist doses (i.e.,.1-1. ) and, in the case of U-5,488, with only two doses of the antagonist. The lack of complete dose-effect determinations with U-5,488 may account for some of the variability observed with this compound. Because morphine-like compounds do not produce ethylketazocine-appropriate responding in monkeys trained to discriminate between ethylketazocine and saline, pa2 values for morphine-quadazocine interactions could not be obtained in the present experiment. These data can, however, be compared with data obtained in monkeys trained to discriminate between codeine and saline under an identical procedure. n that study, several mu agonists, including morphine, alfentanil and levorphanol, produced codeine-appropriate responding. When those agonists were combined with doses of quadazocine, the pa2 values were 7.7, 7.8 and 7.9 for levorphanol, morphine and alfentanil, respectively (Bertalmio et at., 1984). Although the pa2 values obtained here with the kappa agonists were significantly different from those obtained with morphine and other mu agonists, it is important to note that determination of a pa2 value is based on several assumptions, some of which are difficult to assess in vivo. First, both the agonist and the antagonist should be tested at time of peak activity. Quadazocine was used in the present studies specifically because of its long duration of action. Thus, it was assumed that its antagonist action would not differ during 2- to 3-hr experimental sessions. t is not clear, however, whether the agonists were always examined at time of peak effect, and this is difficult to determine with cumulative dosing procedures. Preliminary studies in our laboratory indicate that pa2 values obtained under cumulative and acute dosing procedures are similar. For example, in the drug-discrimination procedure, the pa2 value obtained for quadazocine in combination with morphine was 7.8 when morphine was administered cumulatively and 7.6 when it was administered acutely (A. J. Bertalmio, personal communication). Nevertheless, the kappa agonists examined here have short durations of action, and it is possible that the cumulative dosing procedure underestimates their potency, especially using the tail-withdrawal procedure, in which doses were administered 3 mm apart. Second, this analysis requires that the agonists and antagornst8 interact in a competitive manner. Although the slopes of the Schild plots obtained here were not always equal to the theoretical value of -1., which suggests a noncompetitive interaction, quadazocine consistently produced parallel shifts in the agonist dose-effect curves and, in many cases, yielded slopes that were close to unity. Other potential factors, such as not meeting the requirements for equilibrium or failing to examine sufficient doses of the antagonist, may explain those cases in which the slopes were not equal to -1.. Another way to differentiate activity at kappa vs. mu opioid receptor types is with the mu-selective alkylating agent 9-FNA. n all three behavioral paradigms, i.e., the tail-withdrawal procedure, urine-output procedure and drug-discrimination procedure, f3-fna was effective in antagonizing the effects of morphine or alfentanil, whereas it had little or no activity as a kappa antagonist. ndeed, there was a slight suggestion that $- FNA augmented the effects of ethylketazocine, both under the drug-discrimination and urine-output procedures. fi-fna has also been shown to augment the effects of ethylketazocine on muscle relaxation (D. E. Gmerek, unpublished observations), as well as the effects of U-5,488 on rat plasma corticosterone levels (Hayes and Stewart, 1985). Similar specificity has been reported for the overt behavioral effects of morphine or ethylketazocine in drug-naive or drugdependent rhesus monkeys. That is, $-FNA has been shown to antagonize the overt behavioral effects of acute doses of morphine but not those of the kappa agonists, ethylketazocine or Mr 233. Moreover, $-FNA has been shown to precipitate abstinence in morphine-dependent monkeys (Gmerek and Woods, 1985). On the other hand, fl-fna does not precipitate withdrawal in monkeys made dependent on the kappa agonist, U-5,488 (Gmerek et at., 1987). These fmdings are also in keeping with other reports of the selective and long-lasting antagonism by -FNA of mu-mediated effects in other situations (Ward et al, 1982a,b; Ward and Takemori, 1983; Frischknecht et at., 1983). n addition to having mu-selective antagonist action in the rhesus monkey, the present study also demonstrated that fi- FNA has kappa-like agonist effects in the drug-discrimination and urine-output procedures. That is, when 9-FNA was administered 3 mm before testing, it produced ethylketazocineappropriate responding in monkeys trained to discriminate ethylketazocine from saline, an effect that could be antagonized by quadazocine (G. Winger, unpublished observations). Other compounds with kappa agonist properties have also been shown to shift dose-effect curves for the discriminative properties of ethylketazocine to the left (G. Winger, unpublished observations). n other situations, j-fna has been shown to produce quadazocine-reversible muscle relaxation and stupor in rhesus monkeys (Gmerek and Woods, 1985). Moreover, it is active in the acetic acid-induced stretching test in mice. This effect can be antagonized by naloxone and reveals an apparent pa2 value of 6.6, similar to that seen for interactions between kappa agonists and naloxone under the same procedure. On the other hand, -FNA alone did not increase tail-withdrawal latencies at 3 mm postinjection. This lack of effect under the tailwithdrawal procedure is not surprising because kappa agonists generally increase tail-withdrawal latencies at doses at least 1 log unit higher than those that increase urine output or produce ethylketazocine-appropriate responding (Dykstra et a!., 1987). fl-fna also is not active on the mouse tail-flick test, a test that generally does not show activity for kappa agonists (Ward et at., 1982a). Taken together, these fmdings suggest that (-FNA displays time-dependent kappa agonist activity, followed by mu-selective antagonist activity in both mice and rhesus monkeys. Acknowledgments The authors wish to express their appreciation to Mel Dickerson for expert technical assistance, to Rebecca McLaughlin for assistance in preparing the manuscript and to T. Kendall Hardin for comments on earlier drafts of the manuscript. Downloaded from jpet.aspetjournals.org at ASPET Journals on May 1, 216

7 1987 References BERmLM, A. J., WNGER, G. D. AND WOODS, J. H.: Mu and kappa receptormediated opioid discriminative effects in rhesus monkeys: pa2 analysis. Fed. Proc. 43: 966, DYKS rra, L A.: Effects of ketocyclazocine and ethylketocyclazocine on electric shock titration. Eur. J. PhSrmSCOL 94: 19-26, DYK8TRA, L A., GMzitEK, D. E., WNCES. G. AND WooDs, J. H.: Kappa opioids in rhesus monkeys L Diuresis, sedation, analgesia and discriminative stimulus effects. J. PhsZmSCOL Exp. Ther. 242:413-42,1987. DYKBTRA, L A. AND WooDs, J. H.: A tail withdrawal procedure for assessing analgesic activity in rhesus monkeys. J. PharmaCOl. Methods 15: , F1uscHKNEcHT, H.-R., SEGFRED, B., RGG, G. AND Wsan, P. G.: nhibition of morphine-induced analgesia and locomotor activity in strains of mice: A comparison of long-acting opiate antagonists. PharmaCOL Biochem. Behav. 19: , GMsrEK, D. E., DyKsm, L A. AND WOODS, J. H.: Kappe opioids in rhesus monkeys ll. Dependence associated with chronic administration. J. Pharmaeel. Rip. Ther. 242: ,1987. GMnasK, D. E. AND WOODS, J. H.: Effects of -fimaltrexamine in normal and morphine-dependent rhesus monkeys Observational studies. J. PharmacOL Exp. Ther. 235: , HAYES, A. G. AND STnw.RT, B. it: Effect of mu and kappa opioid receptor agonists on rat plasma corticosterone levels. Eur. J. PharmacoL 116: 75-79, PoirToonasa, P. S., LARSON, D. D., SAyza, L. M., Firms, D. S. td TKzuom, A. B.: A novel opioid receptor site directed alkylating agent with irreversible narcotic antagonistic and reversible agonistic activities. J. Med. Chem. 23: , 198. SCHLD, H..: pa, a new scale for the measurement of drug antagonism. Br. J. PharniaCOL Chemother. 2: , SCHMALJSS, C. AND YAK8H, T. L.: n vivo studies on spinal opiate receptor systems mediating antinociception. H. Pharmacological profiles suggesting a differential association ofmu, delta and kappa receptors with visceral chemical and cutanecos thermal stimuli in the rat. J. PharmaCOL Exp. Ther. 228: 1-12, SMn s, S. E. AND TAKEMu, A. E.: Quantitative studies on the antagonism by naloxone of some narcotic and narcotic-antagonist analgesics. Br. J. PharmacoL 39: , 197. TAicaMolu, A. E., Kurrmwnno, H. J. AND MLLER, J. W.: Quantitative studies Kappa Oplolds 427 of the antagonism of morphine by nalorphine and naloxone. J. PharmacoL Exp. Ther. 169: 39-45, TAKssoRz, A. E., LARSON, D. L AND PoRToosasa, P. S.: The irreversible narcotic antagonistic and reversible agonistic properties of the fumaramate methyl ester derivative of naltrexone. Eur. J. Pharmacol. 7: , TALLARDA, R J. AND Muiuux, R B.: Manual of Pharmacologic Calculations with Computer Programs, Springer-Verlag, New York, TANG, A. H. AND COLLNS, R J.: Behavioral effects of a novel opioid analgesic, U-5488, in rats and rhesus monkeys. Psychopharmacology 85: , TEAL, J. J. AND HOLTZMAN, S. G.: Stimulus effects of morphine in the monkey Quantitative analysis of antagonism. PharmacoL Biochem. Behav. 12: , 198. VON VoGmNDEn, P. F., LAHT, R. A. AND DENS, J. H.: U-5,488: A selective and structurally novel non-mu (kappa) opioid agonist. J. Pharmacol. Exp. Ther. 224: 7-12, Wan, S. J., PERSON, A. K. AND MCHNE, W. F.: Multiple opioid receptor profile in vitro and activity in vivo of the potent opioid antagonist WN 44, Life Sci. 33: suppl, 33-36, WARD, S. J., PorrocHsss, P. S. AND TAKaMm, A. E.: Pharmacological profiles offi-funaltrexamine (8-FNA) and fi-chlornaltrexamine (-CNA) on the mouse yes deferens preparation. Eur. J. PharinacoL 8: , 1982a. WAnD, S. J., PORTOOHE8E, P. S. AND TAKEMu, A. E.: Pharmacological characterization in vivo of the novel opiate -funaltrexamine. J. PharmacoL Exp. Ther. 22: , 1982b. W.an, S. J. AND TAKEMOm, A. E.: Relative involvement of receptor subtypes in opioid-induced inhibition of gastrointestinal transit in mice. J. PharmacOL Exp. Ther. 224: , WooD, P. L, PA.1 L, C., THAKUR, M. AND RCHARD, J. W.: WN 44,441: A stereospecthc and long-acting narcotic antagonist. Pharmacol. Res. Commun. 1: 46-48, WooDs, J. H. AND GMnazx, D. E.: Substitution and primary dependence studies in animals. Drug Alcohol Depend. 14: ,1985. WOODS, J. H., SMim, C. B., MEDZHRADSKY, F. AND SWAN, H. H.: Preclinical testing of new analgesic drugs. n Mechanisms of Pain and Analgesic Cornpounds, ed. by R. F. Beers, Jr. and E. G. Bassett, pp , Raven Press, New York, YAKBH, T. L AND RUDY, T. A.: A dose-ratio comparison of the interaction between morphine and cyclazocine with naloxone in rhesus monkeys on the shock titration task. Eur. J. PharnacO. 48: 83-92, Send reprint requests to: Linda A. Dykstra, PhD., Department of Psychology, 13A, University of North Carolina, Chapel Hill, NC Downloaded from jpet.aspetjournals.org at ASPET Journals on May 1, 216