Psychopharmacology (2002) 163:412–419 DOI 10.1007/s00213-002-1038-x

O R I G I N A L I N V E S T I G AT I O N

S. Stevens Negus · Nancy K. Mello David C. Linsenmayer · R. M. Jones Philip S. Portoghese

Kappa opioid antagonist effects of the novel kappa antagonist 5′-guanidinonaltrindole (GNTI) in an assay of schedule-controlled behavior in rhesus monkeys Received: 30 September 2001 / Accepted: 16 January 2002 / Published online: 13 March 2002 © Springer-Verlag 2002

Abstract Rationale: Opioid receptors are divided into three types: kappa, mu, and delta receptors. Receptor-selective antagonists are useful experimental tools for evaluation of opioid receptor-mediated processes. 5′-Guanidinonaltrindole (GNTI) was recently developed as a novel kappa-selective antagonist. Objectives: To evaluate the potency, time course, and selectivity of GNTI's opioid antagonist effects in rhesus monkeys in an assay of schedule-controlled responding. Methods: Five rhesus monkeys were trained to respond under a fixed ratio 30 schedule of food reinforcement. The rate-decreasing effects of the kappa agonists U50,488 and U69,593, the mu agonist morphine, and the delta agonist SNC80 were examined alone and after pretreatment with GNTI (0.1 and 1.0 mg/kg i.m.; 1 h to 14 days). Results: U50,488, U69,593, morphine, and SNC80 dose-dependently decreased response rates in this procedure. GNTI produced a dose- and time-dependent antagonism of the rate-decreasing effects of U50,488. The kappa antagonist effects of GNTI had a slow onset and a long duration of action, and peak antagonist effects were observed after 24 h. A higher dose of 3.2 mg/kg GNTI eliminated responding in one monkey and was not studied further. The antagonist effects of GNTI were kappa selective, because 1.0 mg/kg GNTI also antagonized the effects of U69,593, but not those of morphine or SNC80. Conclusions: These results suggest that GNTI is a potent and selective kappa antagonist with a slow onset and long duration of action in rhesus monkeys. Relative to the prototype kappa antagonist nor-binaltorphimine, GNTI S.S. Negus (✉) · N.K. Mello · D.C. Linsenmayer Alcohol and Drug Abuse Research Center, Harvard Medical School, McLean Hospital, 115 Mill Street, Belmont, MA 02478, USA e-mail: [email protected] Fax: +1-617-8552519 R.M. Jones · P.S. Portoghese Department of Medicinal Chemistry, College of Pharmacy, University of Minnesota, Minneapolis, MN 55455, USA

may have some advantages as a tool for the study of kappa receptor-mediated processes. Keywords Kappa opioid receptor · Rhesus monkey · 5′-Guanidinonaltrindole · Nor-binaltorphimine

Introduction Opioids act at three major types of opioid receptors, the kappa, mu, and delta opioid receptors (Reisine and Pasternak 1996), and receptor-selective antagonists are useful experimental tools for the evaluation of opioid receptor-mediated effects. Until recently, the only available antagonist selective for kappa receptors was norbinaltorphimine (nor-BNI; Portoghese et al. 1987). Nor-BNI binds with more than 100-fold selectivity to kappa receptors as opposed to mu or delta receptors (Emmerson et al. 1994; Takemori et al. 1988), and numerous studies have demonstrated that, under appropriate conditions, nor-BNI selectively antagonizes in vitro and in vivo effects of kappa opioid agonists (Birch et al. 1987; Broadbear et al. 1994; Butelman et al. 1993; Endoh et al. 1992; Horan et al. 1992; Jewett and Woods 1995; Jones and Holtzman 1992; Takemori et al. 1988). However, nor-BNI has relatively low potency in vivo after systemic administration, and it is unusual among opioid antagonists in that it has a slow onset and an extraordinarily long duration of action. For example, in one study in rhesus monkeys (Butelman et al. 1993), a dose of 3.2 mg/kg nor-BNI selectively antagonized the antinociceptive effects of the kappa agonists U50,488 and U69,593, and produced maximal five- to eightfold rightward shifts in the U50,488 and U69,593 dose-effect curves. However, these maximal kappa antagonist effects were not observed until 24–72 h after nor-BNI administration, and the antagonist effects of nor-BNI persisted for 3–4 weeks. For comparison, the mu-selective opioid antagonist quadazocine and the delta-selective opioid an-

413

tagonist naltrindole produced their peak antagonist effects in less than 1 h, and these antagonist effects lasted less than 1 day (Bertalmio and Woods 1987; Negus et al. 1994). The selectivity of nor-BNI also displays an unusual time course. During the first 1–2 h after its administration in vivo, nor-BNI produces a relatively non-selective antagonism of the effects of kappa and mu agonists (Broadbear et al. 1994; Endoh et al. 1992). During the next few hours, the mu antagonist effects wane while the kappa antagonist effects become more pronounced, and nor-BNI produces reliably selective kappa antagonist effects only after several hours. Overall, the low potency, slow onset, long duration of action, and timedependent selectivity of nor-BNI complicate its use as a pharmacological tool. 5′-Guanidinonaltrindole (GNTI) was recently synthesized as a novel kappa receptor antagonist (Jones and Portoghese 2000; Jones et al. 1998; Stevens et al. 2000). Structure-activity, site-directed mutagenesis, and molecular-modeling studies suggest that the indole moiety of GNTI serves as a scaffold that directs the cationic 5′-guanidium substituent (the “address”) to the non-conserved Glu 297 residue of the kappa opioid receptor to form an ionic bond that results in increased affinity and high selectivity (Portoghese 2001). In receptor binding assays, GNTI had subnanomolar affinity for kappa opioid receptors and more than 200-fold selectivity for kappa receptors as opposed to mu and delta receptors (Jones and Portoghese 2000; Stevens et al. 2000). In smooth muscle preparations, GNTI was approximately 2- to 5-fold more potent than nor-BNI and approximately 10-fold more selective than nor-BNI as a kappa antagonist (Jones and Portoghese 2000; Jones et al. 1998; Stevens et al. 2000). Finally, studies of opioid antagonist effects on feeding behavior in rats suggested the GNTI may have a shorter duration of action than nor-BNI (Jewett et al. 2001). These findings suggest that GNTI may have advantages over nor-BNI as a tool for the evaluation of kappa receptor-mediated effects. The purpose of the present study was to further evaluate the potency, time course, and selectivity of the kappa antagonist effects of GNTI with respect to behavioral studies in rhesus monkeys. Behavioral studies were conducted in an assay of food-maintained, schedule-controlled responding. The use of schedule-controlled responding to generate stable behavioral baselines for the evaluation of drug effects was pioneered by Drs. Dews, Morse, and Kelleher (see, for example, Dews 1956; Kelleher and Morse 1968; McMillan and Morse 1967), and this procedure has been used extensively to examine the pharmacology of agonists and antagonists at kappa, mu, and delta opioid receptors in rhesus monkeys (see, for example, Brandt et al. 2001; Butelman et al. 1996; Downs and Woods 1976; McMillan and Morse 1967; Negus et al. 1993, 1994, 1998).

Materials and methods Subjects Three male (89B057, 89B062, and 90B134) and two female (CH96, 075C) rhesus monkeys (Macaca mulatta) weighing 4.5–12 kg were used as subjects. All monkeys had prior exposure to drugs (primarily dopaminergic and opioid compounds) and operant behavioral procedures. The subjects were individually housed, and water was freely available. Their diet consisted of PMI Feeds Jumbo monkey diet (2–6 biscuits/day) and was supplemented with fresh fruit twice daily. A 12-h light/12-h dark cycle was in effect (lights on from 7 a.m. to 7 p.m.). All housing and procedures were in compliance with NIH guidelines on care and use of animal subjects in research, and were approved by the McLean Hospital Institutional Animal Care and Use Committee. Assay of schedule-controlled behavior Apparatus and procedure Experiments were conducted in each monkey’s home cage (dimensions: 60×65×75 cm). The home cages of all monkeys were modified to include an operant response panel (28×28 cm) mounted on the front wall. Three square translucent response keys (6.4×6.4 cm) were arranged 2.54 cm apart in a horizontal row 3.2 cm from the top of the operant panel. Each key could be illuminated by red, green, or yellow stimulus lights (Superbright LEDs). In addition, three circular translucent panels (1.9 cm in diameter) were located in a vertical column below the center response key and could be illuminated by red, green, or yellow stimulus lights (Superbright LEDs). The operant panel also supported an externally mounted pellet dispenser (Gerbrands; model G5210) that delivered 1 g fruit-flavored food pellets (Noyes, Lancaster, N.H., USA) to a food receptacle mounted on the cage beneath the operant response panel. The panel was controlled by a MED-PC interface and an IBM compatible computer programmed in MEDSTATE Notation (MED Associates, East Fairfield, Vt., USA). Training sessions consisted of five consecutive cycles. Each cycle was 15 min long and consisted of two components: a 10-min pretreatment period followed by a 5-min response period. During the pretreatment period, no stimulus lights were illuminated and responding had no scheduled consequences. During the response period, the center key was illuminated yellow, and the subjects could respond for up to ten food pellets on a fixed ratio 30 (FR30) schedule of reinforcement. If all ten food pellets were earned before 5 min had elapsed, the lights were turned off, and responding had no scheduled consequences for the remainder of that response period. All monkeys were trained until they responded at rates greater than 0.5 responses/s during all five cycles for 10 consecutive days. Sessions were conducted 5 days a week. Test sessions were conducted only after a training session during which the monkeys responded at rates greater than 0.5 responses/s for all five cycles. Test sessions were identical to training sessions except that a dose of the test agonist (U50,488, U69,593, morphine, or SNC80) was administered i.m. at the beginning of each 15-min cycle, and each dose increased the total cumulative dose by 1/4 or 1/2 log units. Each agonist was tested from doses that had no effect on response rates up to a dose that decreased response rates to less than 0.1 responses/s, and test sessions consisted of a maximum of five cycles. To assess the potency and time course of the kappa antagonist effects of GNTI, the effects of the kappa agonist U50,488 (0.032–1.0 mg/kg) were examined alone (baseline) and at various times after pretreatment with GNTI (0.1–1.0 mg/kg i.m.; 1 h to 14 days). To provide a complete assessment of the time course of 1.0 mg/kg GNTI, two separate experiments were conducted. During the first experiment, a dose of 1.0 mg/kg GNTI was administered, and the U50,488 dose-effect curve was determined after 3 h and 1, 4, 7, and 14 days. At least 4 weeks after administration of the first dose of GNTI, a second dose of 1.0 mg/kg GNTI was ad-

414

Fig. 1 Rate-decreasing effects of the kappa agonist U50,488 administered alone or after pretreatment with 5′-guanidinonaltrindole (GNTI) or nor-binaltorphimine (nor-BNI). Left panel U50,488 dose-effect curve determined alone and 1 day after pretreatment with 1.0 mg/kg GNTI. Abscissa Dose of U50,488 in mg/kg (log scale), ordinate percent control response rate. Right panel Time course of effects of 1.0 mg/kg GNTI and 1.0 mg/kg nor-BNI on the dose ratio for U50,488. Abscissa Time after GNTI or nor-BNI administration, ordinate U50,488 dose ratio. All points show mean data from three monkeys, and error bars show the SEM. Filled points in the right panel show times at which the antagonist produced a significant increase in the U50,488 ED50 value (see also Table 2)

ministered, and U50,488 dose-effect curves were determined after 1 h and 2 and 10 days. For comparison, the effects of U50,488 were also examined in two experiments after pretreatment with the prototype kappa antagonist nor-BNI (1.0 mg/kg i.m.; 1 h to 7 days). The first experiment tested U50,488 after 3 h and 1, 4, and 7 days. After at least 4 weeks, a second dose of 1.0 mg/kg nor-BNI was administered, and U50,488 was tested after 1 h and 2 days. These antagonism experiments sometimes involved determination of U50,488 dose-effect curves on consecutive days (i.e., 3 h and 1 day after treatment with 1.0 mg/kg GNTI or nor-BNI), and it was possible that rightward shifts in dose-effect curves on the 2nd day could have reflected acute tolerance rather than antagonism. To assess the degree to which acute tolerance might develop to U50,488 under these conditions, we also determined the U50,488 dose-effect curve on 2 consecutive days without any pretreatment. Our results indicated that acute tolerance did not develop (see Results). To assess the selectivity of GNTI, the rate-decreasing effects of the kappa agonist U69,593 (0.001–0.032 mg/kg), the mu agonist morphine (0.1–10.0 mg/kg), and the delta agonist SNC80 (0.032–0.32 mg/kg) were also determined alone and after i.m. treatment with 1.0 mg/kg GNTI. Because experiments with U50,488 indicated that peak antagonist effects of GNTI were observed after 24 h (see Results), experiments with U69,593, morphine, and SNC80 were conducted using a 24-h pretreatment time with GNTI. In addition, the effects of morphine were also examined 1 h after GNTI pretreatment, because previous studies indicated that nor-BNI produces transient mu antagonist effects for the first 1–2 h after its administration (Broadbear et al. 1994; Endoh et al. 1992). All studies with U50,488, U69,593, and morphine were conducted in groups of three monkeys. Experiments with SNC80 were conducted in only one monkey because binding studies have indicated that the selectivity of GNTI for kappa versus delta receptors is greater than its selectivity for kappa versus mu receptors

(Stevens et al. 2000), and GNTI did not alter the rate-decreasing effects of SNC80 in the one monkey tested (see below). Data analysis Operant response rates from each cycle were converted to percent of control using the average rate from the previous training day as the control value. The ED50 was defined as the log dose of the agonist (U50,488, U69,593, morphine, or SNC80) that produced a 50% decrease in the percent control rate of responding. Individual ED50s were calculated by interpolation when only two data points were available (one below and one above 50% control response rate) or by linear regression when at least three data points were available on the linear portion of the dose-effect curve. Individual ED50 values were averaged to yield a mean ED50 value and 95% confidence limits. Because drug doses were incremented on a logarithmic scale, ED50 values were converted to their log values for calculation of means and confidence limits, and then converted back to linear values for presentation in Tables 2 and 3. The ED50 values were considered to be significantly different if 95% confidence limits did not overlap. Studies with GNTI and nor-BNI were conducted in different groups of monkeys with different baseline ED50 values for U50,488. To facilitate comparison of results with GNTI and nor-BNI, dose ratios were also calculated as (agonist ED50 value following antagonist treatment ÷ agonist alone ED50 value), and these dose ratios were plotted in Fig. 1. Drugs 5′-Guanidinonaltrindole 2HCl (GNTI) and nor-binaltorphimine 2HCl (nor-BNI) were synthesized and supplied by Dr. Portoghese and colleagues. Trans-(±)-3,4-dichloro-N-methyl-N-(2-[1-pyrrolidinyl]cyclohexyl)-benzeneacetamide methanesulfonate (U50,488) and (5α,7α,8β)-(+)-N-methyl-N-(7-[1-pyrrolidinyl]-1-oxaspiro[4,5]DEC-8-yl)-benzeneacetamide (U69,593) were purchased from Sigma Chemicals (St. Louis, Mo., USA). Morphine sulfate was supplied by the National Institute on Drug Abuse (Bethesda, Md., USA). SNC80 free base was provided by Dr. Kenner Rice (National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, Md., USA). GNTI, nor-BNI, U50,488, U69,593, and morphine were dissolved in distilled water. SNC80 was dissolved to a concentration of 50 mg/ml in 3% lactic acid in distilled water, and dilutions were made with distilled water. All compounds were administered i.m. in the thigh, and doses refer to the salt or base forms described above.

415 Table 1 Control response rates (±SEM) in responses per second for each monkey during each of the five cycles. Overall average response rates for each monkey are also shown

Table 2 ED50 values (95% CL) in mg/kg for U50,488 administered alone or at various times after pretreatment with 5′-guanidinonaltrindole (GNTI; 0.1 or 1.0 mg/kg) in one group of three monkeys or with norbinaltorphimine (nor-BNI; 1.0 mg/kg) in a separate group of three monkeys

*Significantly different from U50,488 alone #1 as indicated by non-overlapping confidence limits

Cycle number

1 2 3 4 5 Average

Monkey identification number CH96

075C

89B057

89B062

90B134

0.96 (0.08) 0.95 (0.08) 0.94 (0.07) 0.90 (0.11) 1.02 (0.06) 0.95 (0.04)

0.97 (0.05) 1.28 (0.08) 1.42 (0.10) 1.23 (0.08) 1.44 (0.09) 1.27 (0.04)

2.99 (0.19) 3.15 (0.16) 3.03 (0.19) 2.97 (0.18) 2.89 (0.18) 3.0 (0.08)

2.96 (0.07) 2.86 (0.06) 2.75 (0.11) 2.78 (0.10) 2.71 (0.08) 2.81 (0.04)

1.55 (0.19) 1.61 (0.08) 1.50 (0.33) 1.62 (0.31) 1.42 (0.28) 1.54 (0.10)

Treatment

GNTI group

Nor-BNI group

ED50 in mg/kg (95% CL) U50,488 alone #1 U50,488 alone #2

0.063 (0.054–0.074) 0.076 (0.040–0.14)

0.12 (0.070–0.20) –

U50488+0.1 antagonist 3h 1 day 7 days

0.12 (0.062–0.23) 0.17 (0.11–0.27)* 0.078 (0.033–0.18)

– – –

U50,488+1.0 antagonist 1h 3h 1 day 2 days 4 days 7 days 10 days 14 days

0.12 (0.054–0.25) 0.29 (0.15–0.54)* 0.33 (0.16–0.69)* 0.24 (0.10–0.55)* 0.17 (0.047–0.60) 0.17 (0.056–0.50) 0.13 (0.064–0.25) 0.055 (0.044–0.067)

0.094 (0.051–0.18) 0.19 (0.068–0.54) 0.22 (0.14–0.35) 0.12 (0.058–0.25) 0.16 (0.14–0.20) 0.091 (0.055–0.15) – –

Results Control rates of responding Table 1 shows control rates of responding in each monkey during each of the five response cycles on control days. Although response rates varied across monkeys, response rates were stable within each monkey throughout the control sessions. Control response rates were also stable throughout the course of the experiment (data not shown). Potency and time course of GNTI antagonism of U50,488 Figure 1 (left panel) shows that the kappa agonist U50,488 produced a dose-dependent decrease in rates of schedule-controlled responding. The ED50 value is shown in Table 2. Subsequent antagonism experiments sometimes involved determination of U50,488 dose-effect curves on consecutive days (see below), and it was possible that rightward shifts in dose-effect curves on the 2nd day could have reflected acute tolerance rather than antagonism. To assess the degree to which acute tolerance might develop to U50,488 under these conditions, we also determined the U50,488 dose-effect curve on

2 consecutive days without any pretreatment. The ED50 value for a first determination of the U50,488 dose-effect curve (U50,488 alone #1 in Table 2) was similar to the ED50 value for a second determination of the U50,488 dose-effect curve assessed 1 day later (U50,488 alone #2 in Table 2). GNTI produced a dose- and time-dependent antagonism of the rate-decreasing effects of U50,488, and peak antagonism was observed after 24 h. Table 2 shows ED50 values for U50,488 determined at various times after pretreatment with 0.1 and 1.0 mg/kg GNTI. The lower dose of 0.1 mg/kg GNTI significantly increased the U50,488 ED50 value after 1 day, but not after 3 h or 7 days. A higher dose of 1.0 mg/kg GNTI significantly increased ED50 values for U50,488, after 3 h and 1 and 2 days. Mean ED50 values for U50,488 gradually returned to baseline after 14 days. The left panel of Fig. 1 shows the complete dose-effect curve for U50,488 administered alone or 1 day after pretreatment with 1.0 mg/kg GNTI. The right panel of Fig. 1 shows the dose ratios for U50,488 determined at various times after administration of 1.0 mg/kg GNTI. A higher dose of 3.2 mg/kg GNTI was tested in one monkey (monkey 075C). In this monkey, 3.2 mg/kg GNTI eliminated responding and produced mild sedation for several hours after its administration. On the day after administration of 3.2 mg/kg GNTI, response rates re-

416 Fig. 2 Rate-decreasing effects of the kappa agonist U69,593, the mu agonist morphine, and the delta agonist SNC80 administered alone or 1 day after pretreatment with 1.0 mg/kg GNTI. Abscissae Dose of each agonist in mg/kg (log scale), ordinates percent control response rate. All points for U69,593 and morphine show mean data from three monkeys, and error bars show the SEM. Points for SNC80 show data from one monkey

Table 3 ED50 values (95% CL) in mg/kg for U69,593, morphine, and SNC80 administered alone or after pretreatment with 1.0 mg/kg GNTI. “morphine + 1.0 GNTI (1 h)” and “morphine + 1.0 GNTI (24 h)” were determined in different groups of monkeys, and the ED50 value for “morphine alone” is shown for each group. All data show results from studies with three monkeys except data for SNC80, which show results from studies in one monkey Treatment

ED50 value (±SEM)

U69,593 alone U69,593 + 1.0 GNTI (24 h) Morphine alone Morphine + 1.0 GNTI (1 h) Morphine alone Morphine + 1.0 GNTI (24 h) SNC80 alone SNC80 + 1.0 GNTI (24 h)

0.0030 (0.0017–0.0052) 0.013 (0.0062–0.026)* 1.4 (0.56–3.3) 0.53 (0.19–1.4) 0.72 (0.19–2.6) 0.80 (0.43–1.5) 0.09a 0.12a

*Significantly different from drug alone as indicated by non-overlapping confidence limits a Studies with SNC80 were conducted in only one monkey

covered to baseline levels and the monkey appeared normal. Because the mechanisms underlying these effects of GNTI were unknown, studies with U50,488 were not conducted, and this high dose of GNTI was not studied in other monkeys. For comparison with the effects of GNTI, Table 2 also shows ED50 values for U50,488 determined at various times after treatment with the prototype kappa antagonist nor-BNI (1.0 mg/kg). In addition, the right panel of Fig. 1 shows the dose ratios for U50,488 determined after 1.0 mg/kg nor-BNI treatment. Nor-BNI did not significantly increase the U50,488 ED50 value at any of the times tested, and nor-BNI (1.0 mg/kg) produced smaller increases in the dose ratio for U50,488 than GNTI (1.0 mg/kg). A higher dose of nor-BNI was not tested in the present study, because previous studies have already reported that a higher dose of 3.2 mg/kg nor-BNI produced significant and long-lasting antagonism of the rate-decreasing and antinociceptive effects of U50,488 in

rhesus monkeys (Butelman et al. 1993; Negus et al. 1998). Selectivity of the antagonist effects of GNTI Figure 2 shows the effects of 1 day pretreatment with 1.0 mg/kg GNTI on the dose-effect curves for the kappa agonist U69,593, the mu agonist morphine, and the delta agonist SNC80. ED50 values for each agonist alone and after GNTI pretreatment are shown in Table 3. As with U50,488, 1 day pretreatment with 1.0 mg/kg GNTI produced a rightward shift in the U69,593 dose-effect curve and a significant increase in the U69,593 ED50 value. Moreover, the dose ratios for U50,488 and U69,593 1 day after pretreatment with 1.0 mg/kg GNTI were similar (5.3 and 4.3, respectively). However, GNTI did not significantly alter the effects of morphine or SNC80. The effects of morphine were also not altered by 1 h pretreatment with 1.0 mg/kg GNTI (Table 3).

Discussion Control response rates and rate-decreasing effects of opioid agonists and GNTI Response rates engendered under the FR30 schedule of food-maintained responding varied somewhat across monkeys, but for any one monkey, response rates were stable within each experimental session and across sessions. These results agree with many other studies in finding that appropriately designed assays of schedulecontrolled behavior can engender stable behavioral baselines that can be used to evaluate drug effects (see, for example, Dews 1956; Kelleher and Morse 1968; McMillan and Morse 1967; Negus et al. 1993). The kappa agonists U50,488 and U69,593, the mu agonist morphine, and the delta agonist SNC80 all produced dose-dependent decreases in rates of food-maintained responding

417

under this fixed ratio schedule. Other studies also found that kappa, mu, and delta agonists dose-dependently decreased rates of fixed ratio responding (Bergman and Warren 1989; Downs and Woods 1976; McMillan and Morse 1967; Negus et al. 1993, 1998), and the potencies of the agonists in the present study agree with previous studies in rhesus monkeys (Brandt et al. 2001; Negus et al. 1993, 1998). Doses of 0.1 and 1.0 mg/kg GNTI (i.m.) had little effect on response rates, and the antagonist effects of these GNTI doses were readily evaluated (see below). A higher dose of 3.2 mg/kg GNTI eliminated responding and produced mild sedation in the only monkey tested. The pharmacological mechanisms underlying these effects were not examined; however, the relatively rapid onset and short duration of these effects contrast with the slower onset and longer duration of GNTI's kappa antagonist effects (see below), suggesting that these direct behavioral effects of the high dose of GNTI may not have resulted from kappa receptor blockade. High doses of other opioid antagonists have also been found to decrease rates of food-maintained responding, and non-opioid mechanisms may contribute to these effects (France and Morse 1989; Negus et al. 1993). Antagonist effects of GNTI GNTI (0.1–1.0 mg/kg) produced dose- and time-dependent increases in the ED50 value for U50,488. Peak antagonist effects were observed after 1 day and were no longer apparent after 14 days. GNTI (1.0 mg/kg, 1 day pretreatment) also increased the ED50 value for another kappa agonist, U69,593, and dose ratios for U50,488 and U69,593 were similar (5.3 and 4.3, respectively). These findings suggest that GNTI is a kappa opioid antagonist with a slow onset and a long duration of action in rhesus monkeys. These findings are also consistent with previous reports that the rate-decreasing effects of kappa agonists can be antagonized by the prototype kappa antagonist nor-BNI (Jewett and Woods 1995; Negus et al. 1998) as well as by other opioid antagonists with affinity for kappa receptors, such as naltrexone, MR2266, and quadazocine (Bergman and Warren 1989; Negus et al. 1993). GNTI did not antagonize the effects of the mu agonist morphine or the delta agonist SNC80 when GNTI was administered at a dose (1.0 mg/kg) and pretreatment time (1 day) that produced maximal antagonism of U50,488. One hour pretreatment with 1.0 mg/kg GNTI also failed to antagonize the effects of morphine. These findings suggest that GNTI is selective for kappa as opposed to mu or delta opioid receptors. In vitro receptor binding and functional studies also indicated that GNTI had high selectivity for kappa receptors as opposed to mu or delta opioid receptors (Jones and Portoghese 2000; Jones et al. 1998; Stevens et al. 2000).

Comparison of GNTI to the prototype kappa antagonist nor-BNI The prototype kappa agonist nor-BNI has been widely used as a pharmacological tool to evaluate kappa receptor-mediated processes, and it is therefore of interest to compare the pharmacology of GNTI with that of norBNI. The results of the present study suggest that, following systemic administration in rhesus monkeys, GNTI and nor-BNI differ in potency, time course, and selectivity. First, GNTI was slightly more potent than nor-BNI as an antagonist of the rate-decreasing effects of U50,488. In the present study, both 0.1 and 1.0 mg/kg GNTI produced significant increases in the U50,488 ED50 value after 1 day, whereas 1.0 mg/kg nor-BNI was ineffective. However, we have shown previously that a higher dose of 3.2 mg/kg nor-BNI did significantly increase the U50,488 ED50 value in this assay (Negus et al. 1998). Similarly, a dose of 3.2 mg/kg nor-BNI was required to produce five- to tenfold rightward shifts in the dose-effect curves for U50,588- and U69,593-induced antinociception in rhesus monkeys (Butelman et al. 1993). GNTI has a lower molecular weight than nor-BNI (544.5 and 734.8, respectively), but even when this factor is taken into account, GNTI is at least two to three times more potent than nor-BNI in rhesus monkeys. This finding from in vivo studies agrees with the results of in vitro smooth muscle assays, which indicated that GNTI was two to five times more potent as a kappa antagonist than nor-BNI (Jones and Portoghese 2000; Jones et al. 1998; Stevens et al. 2000). In addition, centrally administered (i.c.v.) GNTI was more potent than nor-BNI as an antagonist of U50,488-induced feeding in rats (Jewett et al. 2001). The higher potency of GNTI may facilitate some types of studies, such as behavioral studies in primates, that require relatively large amounts of drug. GNTI also appears to have a slightly faster onset and shorter duration of action than nor-BNI. In the present study, the maximum rightward shift in the U50,488 doseeffect curve was observed 1 day after administration of 1.0 mg/kg GNTI. The antagonist effects of GNTI were no longer significant after 4 days, although antagonist effects were still apparent in some monkeys for up to 10 days. ED50 values for U50,488 returned to control levels in all monkeys after 14 days. In contrast, a roughly equieffective dose of 3.2 mg/kg nor-BNI produced maximal antagonism of U50,488-induced antinociception after 3 days, and these effects persisted for approximately 3 weeks (Butelman et al. 1993). GNTI was also reported to be shorter acting than nor-BNI as an antagonist of U50,488-induced feeding (Jewett et al. 2001). However, it should be noted that GNTI still has a relatively slow onset and long duration of action in comparison to other opioid antagonists. For example, the mu-selective opioid antagonist quadazocine and the delta-selective opioid antagonist naltrindole produced their peak antagonist effects in less than 1 h, and these antagonist effects lasted less than 1 day (Bertalmio and Woods 1987; Negus et al. 1994). The reason for the relatively

418

long duration of action of GNTI and nor-BNI is not known, but another structurally dissimilar kappa-selective antagonist was also recently developed and shown to have a very long duration of action (Thomas et al. 2001; J.B. Thomas, personal communication). Thus, an unusually long duration of action is a common feature for the three kappa-selective antagonists currently described in the literature. Finally, GNTI may be more selective than nor-BNI. Previous studies have shown that the selectivity of the antagonist effects of nor-BNI varies across time. During the first 1–2 h after its systemic administration in rodents, nor-BNI produced a relatively non-selective antagonism of the effects of both kappa and mu agonists (Broadbear et al. 1994; Endoh et al. 1992). During the next few hours, the mu antagonist effects diminished while the kappa antagonist effects became more pronounced. Thus, nor-BNI acted as a relatively non-selective opioid antagonist for the first 1–2 h after its administration, and it produced reliably selective kappa antagonist effects only after several hours. In comparison, GNTI antagonized only the effects of kappa agonists at the doses and times evaluated in the present study. In particular, 1.0 mg/kg GNTI did not alter the ED50 value for morphine after either 1 h or 1 day. These results suggest that effective kappa antagonist doses of GNTI did not produce either an initial or a delayed mu antagonist effect. Conclusion The present results confirm and extend the characterization of GNTI as a selective kappa opioid antagonist. Relative to the prototype kappa antagonist nor-BNI, GNTI was slightly more potent, it had a slightly faster onset and shorter duration of action, and it did not produce mu antagonist effects during the first hour after its administration. These characteristics may make GNTI more suitable than nor-BNI as a pharmacological tool for the evaluation of kappa receptor-mediated processes for some types of studies. However, GNTI still had a slow onset and long duration of action relative to most other opioid antagonists. The mechanisms that underlie the unusually long duration of currently available kappa antagonists requires further evaluation. Acknowledgements This work was supported by grants RO1DA02519, RO1-DA11460, RO1-DA-01533, U19-DA11007, and K05-DA00101 from the National Institute on Drug Abuse. The authors would like to thank Kate Banks D.V.M. for veterinary assistance.

References Bergman J, Warren PH (1989) Effects of kappa opioids on schedule-controlled behavior of squirrel monkeys. J Pharmacol Exp Ther 248:1102–1108 Bertalmio AJ, Woods JH (1987) Differentiation between mu and kappa receptor-mediated effects in opioid drug discrimination: apparent pA2 analysis. J Pharmacol Exp Ther 243:591–597

Birch PJ, Hayes AG, Sheehan MJ, Tyers MB (1987) Norbinaltorphimine: antagonist profile at k opioid receptors. Eur J Pharmacol 144:405–408 Brandt MR, Furness MS, Rice KC, Fischer BD, Negus SS (2001) Studies of tolerance and dependence with the delta-opioid agonist SNC80 in rhesus monkeys responding under a schedule of food presentation. J Pharmacol Exp Ther 299:629–637 Broadbear JH, Negus SS, Butelman ER, Costa BR de, Woods JH (1994) Differential effects of systemically administered norbinaltorphimine (nor-BNI) on k-opioid agonists in the mouse writhing assay. Psychopharmacology 115:311–319 Butelman ER, Negus SS, Ai Y, Costa BR de, Woods JH (1993) Kappa opioid antagonist effects of systemically administered nor-binaltorphimine (nor-BNI) in a thermal antinociception assay in rhesus monkeys. J Pharmacol Exp Ther 257:1269–1276 Butelman ER, Negus SS, Lewis JW, Woods JH (1996) Clocinnamox antagonism of opioid suppression of schedule-controlled responding in rhesus monkeys. Psychopharmacology 123:320–324 Dews PB (1956) Modification by drugs of performance on simple schedules of positive reinforcement. Ann NY Acad Sci 65: 268–281 Downs DA, Woods JH (1976) Morphine, pentazocine and naloxone effects on responding under a multiple schedule of reinforcement in rhesus monkeys. J Pharmacol Exp Ther 196: 298–306 Emmerson PJ, Liu M-R, Woods JH, Medzihradsky F (1994) Binding affinity and selectivity of opioids at mu, delta and kappa receptors in monkey brain membranes. J Pharmacol Exp Ther 271:1630–1637 Endoh T, Matsuura H, Tanaka C, Nagase H (1992) Nor-binaltorphimine: a potent and selective k-opioid receptor antagonist with long-lasting activity in vivo. Arch Int Pharmacodyn 316: 30–42 France CP, Morse WH (1989) Pharmacological characterization of supersensitivity to naltrexone in squirrel monkeys. J Pharmacol Exp Ther 250:928–936 Horan P, Taylor J, Yamamura HI, Porreca F (1992) Extremely long-lasting antagonistic actions of nor-binaltorphimine (norBNI) in the mouse tail-flick test. J Pharmacol Exp Ther 260:1237–1243 Jewett DC, Woods JH (1995) Nor-binaltorphimine: an ultra-long acting kappa-opioid antagonist in pigeons. Behav Pharmacol 6:815–820 Jewett DC, Grace MK, Jones RM, Billington CJ, Portoghese PS, Levine AS (2001) The kappa-opioid antagonist GNTI reduces U50,488-, DAMGO-, and deprivation-induced feeding, but not butorphanol- or neuropeptide Y-induced feeding, in rats. Brain Res 909:75–80 Jones DNC, Holtzman SG (1992) Long term k-opioid receptor blockade following nor-binaltorphimine. Eur J Pharmacol 215:345–348 Jones RM, Portoghese PS (2000) 5′-Guanidinonaltrindole, a highly selective and potent k-opioid receptor antagonist. Eur J Pharmacol 396:49–52 Jones RM, Hjorth SA, Schwarts TW, Portoghese PS (1998) Mutational evidence for a common k antagonist binding pocket in the wild-type k and mutant µ(K303E) opioid receptors. J Med Chem 41:4911–4914 Kelleher RT, Morse WH (1968) Determinants of the specificity of behavioral effects of drugs. Ergeb Physiol Biol Chem Exp Pharmakol 60:1–56 McMillan DE, Morse WH (1967) Some effects of morphine and morphine antagonists on schedule-controlled behavior. J Pharmacol Exp Ther 157:175–184 Negus SS, Burke TF, Medzihradsky F, Woods JH (1993) Effects of opioid agonists selective for mu, kappa and delta opioid receptors on schedule-controlled responding in rhesus monkeys: antagonism by quadazocine. J Pharmacol Exp Ther 267: 896–903 Negus SS, Butelman ER, Chang KJ, DeCosta BR, Winger G, Wood JH (1994) Behavioral effects of the systemically active

419 delta opioid agonist BW373U86 in rhesus monkeys. J Pharmacol Exp Ther 270:1025–1034 Negus SS, Gatch MB, Mello NK, Zhang X, Rice K (1998) Behavioral effects of the delta-selective opioid agonist SNC80 and related compounds in rhesus monkeys. J Pharmacol Exp Ther 286:362–375 Portoghese PS (2001) From models to molecules: opioid receptor dimers, bivalent ligands, and selective opioid receptor probes. J Med Chem 44:2259–2269 Portoghese PS, Lipkowski AW, Takemori AE (1987) Binaltorphimine and norbinaltorphimine, potent and selective kappa opioid receptor antagonists. Life Sci 40:1287–1292 Reisine T, Pasternak G (1996) Opioid analgesics and antagonists. In: Hardman JG, Limbird LE (eds) The pharmacological basis of therapeutics, 9th edn. McGraw Hill, New York, pp 521–555

Stevens WC, Jones RM, Subramanian G, Metzger TG, Ferguson DM, Portoghese PS (2000) Potent and selective indolomorphinan antagonists of the kappa-opioid receptor. J Med Chem 43:2759–2769 Takemori AE, Ho BY, Naeseth JS, Portoghese PS (1988) Norbinaltorphimine (norBNI), a highly selective kappa-opioid receptor antagonist in analgesia and receptor binding assays. J Pharmacol Exp Ther 246:255–258 Thomas JB, Atkinson RN, Rothman RB, Fix SE, Mascarella SW, Vinson NA, Xu H, Dersch CM, Lu Y, Cantrell BE, Zimmerman DM, Carroll FI (2001) Identification of the first trans-(3R,4R)-dimethyl-4-(3-hydroxyphenyl)piperidine derivative to possess highly potent and selective opioid kappa receptor antagonist activity. J Med Chem 44:2687–2690