Quantification of observable behaviors induced by typical and atypical kappa-opioid receptor agonists in male rhesus monkeys

Abstract

Rationale.

Kappa-opioid receptor (KOR) agonists are antinociceptive but have side effects that limit their therapeutic utility. New KOR agonists have been developed that are fully efficacious at the KOR but may produce fewer or reduced side-effects of typical KOR agonists.

Objectives.

We determined behavioral profiles for typical and atypical KOR agonists purported to differ in intracellular-signaling profiles as well as a mu-opioid receptor (MOR) agonist, oxycodone, using a behavioral scoring system based on Novak and colleagues (1992, 1998) and modified to quantify drug-induced effects (e.g., Duke et al. 2018).

Methods.

Six adult male rhesus monkeys were administered a range of doses of the typical KOR agonists, U50–488H (0.0032–0.1 mg/kg) and salvinorin A (0.00032–0.01 mg/kg), the atypical KOR agonists, nalfurafine (0.0001–0.001 mg/kg) and triazole 1.1 (0.01–0.32 mg/kg), the MOR agonist, oxycodone (0.0032–0.32 mg/kg), and as controls, cocaine (0.032–0.32 mg/kg) and ketamine (0.32–10 mg/kg). For time-course determinations, the largest dose of each KOR agonist or MOR agonist were administered across timepoints (10–320 min). In mixture conditions, oxycodone (0.1 mg/kg) was followed by KOR-agonist administration.

Results.

Typical KOR agonists produced sedative-like and motor-impairing effects. Nalfurafine was similar to typical KOR agonists on most outcomes, and triazole 1.1 produced no effects on its own except for reducing scratch during time-course determinations. In the mixture, all KOR agonists reduced oxycodone-induced scratching, U50–488H and nalfurafine reduced species-typical activity, and U50–488H increased rest/sleep posture.

Conclusions.

Atypical “biased” KOR agonists produce side-effect profiles that are relatively benign (triazole 1.1) or reduced (nalfurafine) compared to typical KOR agonists.

Introduction

Kappa-opioid receptor (KOR) agonists have been investigated as potential therapeutics for a number of symptoms and disorders including pain and substance use. While they lack the broad spectrum of pain reduction produced by mu-opioid receptor (MOR) agonists, KOR agonists produce antinociception against specific pain modalities in humans and nonhumans (Aldrich and McLaughlin 2009; Kivell and Prisinzano 2010) and reduce abuse-related effects of drugs, including MOR agonists (Prisinzano et al. 2005). KOR agonists also reduce pruritus in rodents, nonhuman primates, and humans, including pruritus produced by MOR agonists (Beck et al. 2019; Inan et al. 2019; Kamimura et al. 2017; Ko and Husbands 2009; Phan et al. 2012). The clinical feasibility of KOR agonists has been limited by side effects like psychotomimesis, dysphoria, and sedation (Mores et al. 2019; Pfeiffer et al. 1986).

New KOR agonists have been introduced that are fully efficacious at the KOR but appear to lack certain behavioral effects typical of the drug class. One example, nalfurafine, is approved for clinical use in humans for treatment of intractable pruritus (Remitch®; Inui 2015). Nalfurafine is atypical in that it does not produce psychotomimesis or dysphoria in humans at doses that produce overt sedation, although it does produce KOR-mediated antinociception and reductions in pruritus in humans and nonhumans (Hasebe et al. 2004; Inui 2015; Ko and Husbands 2009). Triazole 1.1 was introduced as a novel KOR agonist that was highly selective for and fully efficacious at the KOR and produced KOR-mediated antinociception and reductions in pruritus in rodents (Brust et al. 2016; Zhou et al. 2013). Unlike the typical KOR agonist, U50–488H, triazole 1.1 did not affect locomotion or brain dopamine levels at antinociceptive and anti-pruritic doses. As such, atypical KOR agonists like nalfurafine and triazole 1.1 appear to have more favorable therapeutic indices than typical KOR agonists.

Efforts to determine the mechanistic underpinnings of the atypical behavioral profiles of KOR agonists like nalfurafine and triazole 1.1 have focused on nuanced, molecule-specific signaling characteristics at the KOR. The KOR is a G-protein coupled receptor (GPCR) that interacts with extracellular ligands to produce an array of intracellular signaling responses (Darcq and Kieffer 2018). Advances in the study of GPCR pharmacology have resulted in the development of molecules that selectively activate certain intracellular pathways over others in a ligand-specific manner, referred to as biased agonism (Mores et al. 2019; Dogra and Yadov, 2015; Wootten et al. 2018). Using biased agonists, G-protein signaling and the recruitment of β-arrestin have emerged as primary effectors of the behavioral effects of KOR agonists. Activation of the G-protein pathway has been interpreted to be the principal mediator of antinociception and anti-pruritus, while recruitment of β-arrestin is hypothesized to be associated with a number of “side effects” like dysphoria and sedation (see Mores et al. 2019; Dogra and Yadov, 2015). The implication of these findings is that G-protein biased KOR agonists will be more therapeutically selective than typical KOR agonists like U50–488H (a benchmark drug often used to represent an “unbiased” KOR agonist). However, links between biased signaling at GPCRs and behavioral phenotypes remain controversial in part because assessments of signaling bias have been shown to vary between approaches (Brust et al. 2016; Dunn et al. 2019; Kaski et al. 2019) and between the species of receptors used within an approach (Schattauer et al. 2017). Accordingly, behavioral studies that systematically compare “biased” to “unbiased” compounds are needed to test the veracity of the claim that G-protein signaling bias at the KOR is associated with atypical behavioral profiles.

We have developed and validated a behavioral scoring system based on Novak et al. (1992, 1998) to quantify drug effects across a range of species-typical and drug-induced behaviors in nonhuman primates (Duke et al. 2018; Platt et al. 2002; Ruedi-Bettschen et al. 2013). Some behaviors are applicable to known MOR or KOR agonist effects (e.g., scratch [itch]; rest/sleep posture [sedation]; retch/vomit and facial rub [nausea/gastrointestinal distress]). Other species-typical behaviors can be used to determine relative sensitivities to direct effects of drugs (e.g., locomotion, self-groom, tactile/oral manipulation). While previous studies have investigated behavioral effects of KOR agonists in nonhuman primates (Butelman et al. 2009; Butelman et al. 2010; Dykstra et al. 1987), no study has investigated atypical KOR agonists in nonhuman primates or in a design that includes a range of behaviors as extensive as those quantified by our approach.

The current study compared behavioral effects of typical and atypical KOR agonists purported to differ in signaling characteristics at the KOR in rhesus monkeys. Using U50,488H as an “unbiased” or typical KOR compound, dose-response and time-course assessments for salvinorin A (unbiased), nalfurafine (purported G-protein bias), and triazole 1.1 (purported G-protein bias) were compared across behaviors. To our knowledge, triazole 1.1 has yet to be tested in nonhuman primates. We predicted that atypical KOR agonists would produce fewer side effects including sedative-like and motor-impairing effects compared with typical KOR agonists. Oxycodone was tested to provide a MOR-agonist comparator and as a mixture with KOR agonists to determine if atypical behavioral profiles affected the ability of KOR agonists to produce characteristic modifications of MOR-mediated effects (e.g., reduction of MOR-induced scratching).

Methods

All procedures were approved by the University of Mississippi Medical Center’s Institutional Animal Care and Use Committee and were conducted in accordance with the National Research Council’s Guide for Care and Use of Laboratory Animals (8^th edition, 2011).

Subjects

Six male adult rhesus monkeys (Macaca mulatta, weighing 9.0–15.0 kg) served as subjects. At the start of the experiment, subjects 0099, 1265, and 0342 were experimentally naïve, 5286, RQ7733, and 5264 had experimental histories of food and drug self-administration (Huskinson et al. 2016; unpublished data). Subjects were individually housed in stainless steel cages (each unit: 0.76 m × 0.76 m × 0.86 m; Carter2 Systems, Inc., Beaverton, OR) that allowed visual and olfactory interaction with other monkeys. Subjects received water ad lib, were fed standard biscuits (Teklad 25% Monkey Diet, Harlan/Teklad, Madison, WI) to maintain healthy body weights, and fruits or vegetables and foraging materials were provided daily. Rooms were maintained on a 12-hr light/12-hr dark schedule (lights on at 0600 hr).

Surgery and Apparatus

Pre-surgery, subjects were injected with atropine sulfate (0.04 mg/kg, i.m.) and ketamine hydrochloride (10–20 mg/kg, i.m.) followed by inhaled isoflurane and preoperative antibiotics (cefazolin; 20–25 mg/kg, i.m.) and analgesics (carprofen, 2–4 mg/kg, s.c. and buprenorphine SR, 0.05 mg/kg, s.c.). Under aseptic conditions, a single lumen silicon catheter (Cole-Parmer; ID: 0.76 mm, OD: 1.65 or 2.46 mm) was implanted into a brachial, jugular, or femoral vein with the tip terminating near the right atrium. The distal end of the catheter was passed subcutaneously to the mid-scapular region, where it exited the subject’s back. After surgery, subjects were fit with a jacket, and the catheter was threaded through a tether and connected to a single-lumen swivel (Lomir Biomedical, Inc., Malone, NY). Catheter material was attached to the swivel from the exterior of the homecage and connected to a 0.22 μm Millipore filter and plastic syringe that was housed in a custom-made box. Postoperative analgesics (carprofen 4 mg/kg, p.o) were given daily for 3 days, and antibiotics (usually Keflex, 22.2 mg/kg, p.o. or i.m.; Eli Lilly & Company, Indianapolis, IN) were given when recommended by veterinary staff. Catheters were flushed daily with heparinized saline (40–100 U/ml). If a catheter became nonfunctional, it was removed, and a new catheter was implanted once health was verified by veterinary staff.

General Behavioral Observation Procedure

Observation sessions were conducted at the same time each day, using the focal animal behavioral scoring system described by Novak and colleagues (1992, 1998) which we modified to include drug-induced behaviors (Duke et al. 2018; Platt et al. 2002). Observers met a 90% inter-observer reliability criterion prior to the experiments and were blind to drug treatments and hypotheses. Behaviors (see Table 1) were scored by recording the presence of each behavior in 15-s intervals during 5-min sessions. Scores were calculated as the number of intervals in which a behavior occurred, with a maximum possible score of 20 for each behavior.

Table 1.

Behavioral categories, abbreviations, and definitions.

Behavior	Brief Description
Passive Visual	Animal is standing or sitting motionless with eyes open
Locomotion	At least two directed steps in the horizontal and/or vertical plane
Self-Groom	Picking, scraping, spreading or licking of an animal’s own hair
Tactile/Oral Exploration	Any tactile or oral manipulation of the cage or environment
Scratch	Vigorous strokes of the hair with fingers or toenails
Stereotypy	Any repetitive, ritualized pattern of behavior that serves no obvious function
Forage	Sweeping and/or picking through wood chip substrate
Vocalization	Species-typical sounds emitted by monkey (not differentiated into different types)
Threat/Aggress	Multifaceted display involving one or more of the following: Open mouth stare with teeth partially exposed, eyebrows lifted, ears flattened or flapping, rigid body posture, piloerection, attack (e.g., biting, slapping) of inanimate object or other monkey
Cage Shake	Any vigorous shaking of the cage that may or may not make noise
Yawn	To open mouth wide and expose teeth
Body Spasm	An involuntary twitch or shudder of the entire body; also “wet dog” shake
Present	Posture involving presentation of rump, belly, flank, and/or neck to observer or other monkey
Drink	Mouth contact to fluid delivery sipper
Facial Rub	Excessive wiping of nose or chin on the home cage or with hand or arm
Fear Grimace	Grin-like facial expression involving the retraction of the lips exposing clenched teeth; may be accompanied by flattened ears, stiff, huddled body posture, screech/chattering vocalizations
Lip Smack	Pursing the lips and moving them together to produce a smacking sound, often accompanied by moaning
Lip Droop	Bottom lip drooping, showing bottom teeth
Vomit/Wretch	Expulsion of food or fluid through mouth or nose or making the sound or movement of vomiting
Tremor/Jerk	A tremor or jerk of a part of the body (e.g., head, two limbs)
Observable Ataxia	Any slip, trip, fall, loss of balance.
Rest/Sleep Posture	Idiosyncratic posture adopted by monkeys during rest or sleep, easily roused; eyes open <3 s after stimulus
Moderate Sedation	Atypical loose-limbed posture (e.g., propped on the cage by the body or a limb), eyes closed, delayed response to external stimuli (> 3 s)
Deep Sedation	Atypical loose-limbed posture, eyes closed, does not respond to external stimuli

For sedation measures (see Figure 1), we used a strategy similar to Duke et al. (2018) and based on standards used for anesthesia of pediatric patients (American Society of Anesthesiologists 2002). When a subject was observed with eyes closed for 3 s, an assessment of the animal’s responsiveness to external stimuli was determined. The observer walked at a normal pace toward the cage, spoke the animal’s name, and tapped the cage with his/her pen. If the subject’s eyes opened within 3 s of any of these stimuli, rest/sleep posture was scored. If the subject’s eyes opened within 3–15 s following stimuli presentation and was observed to be assuming an atypical posture, observers scored moderate sedation. If the subject’s eyes did not open within 15 s following stimuli presentation and was observed to be assuming an atypical posture, observers scored deep sedation. The session was divided into 5, 60-s periods, and the assessment of sedation was initiated during a 60-s period if at any time the animal presented with its eyes closed. If the animal resumed with eyes closed within 3 s of the initial sedation assessment, the result of the initial assessment was recorded for the remaining 15-s intervals of the 60-s period, and if eyes remained closed, the assessment was repeated at the beginning of the next 60-s period. However, if the animal resumed with eyes open after the initial assessment or at any time during the 60-s period for more than 3 s, eyes closing again initiated a new sedation assessment. A flowchart in Figure 1 illustrates these steps for sedation scoring.

Fig. 1. Sedation Assessment Flowchart.

Scoring rules for sedation measures based on standards used for anesthesia of pediatric patients (American Society of Anesthesiologists 2002) and based on prior observation experiments (Duke et al. 2018).

Behavioral profiles for each drug and experimental condition were determined after subjects habituated to the presence of observers. Vehicles and drugs were given intravenously (i.v.) by an experimenter who was not blinded to the drug condition. Injections were delivered manually with a 3-ml syringe containing the vehicle or drug solution followed by an additional 3-ml injection of heparinized saline. Injections were delivered as quickly as possible, taking care to push at a rate that prevented a catheter line from coming disconnected at the swivel or other area along the system (typically within a 15 to 30-s period). Mondays and Thursdays were baseline days, and saline was administered 5 min prior to observation sessions. Tuesdays and Fridays were test days, and details are described below for each condition: dose-response determinations, time-course assessment, and mixtures of oxycodone and KOR agonists.

Dose-response determinations of single drugs.

We determined behavioral profiles following administration of saline, an 80% propylene glycol vehicle, and a range of doses of each drug. Drugs evaluated were the MOR agonist, oxycodone (0.0032–0.32 mg/kg); two typical KOR agonists, U50–488H (0.0032–0.1 mg/kg) and salvinorin A (0.00032–0.01 mg/kg); and two atypical KOR agonists, nalfurafine (0.0001–0.001 mg/kg) and triazole 1.1 (0.01–0.32 mg/kg). Cocaine (0.032–0.32 mg/kg) was evaluated as a centrally active (negative) control not expected to produce sedation and ketamine (0.32–10.0 mg/kg) as a centrally active (positive) control expected to produce sedation. On test days, vehicle or drug was administered i.v., 5 min before observation sessions. Each vehicle was administered 4 times, each oxycodone dose was administered twice, each KOR agonist dose was administered once, and each cocaine or ketamine dose was administered at least once.

Time-course assessment of single drugs.

Time-course assessments were completed for saline, 0.32 mg/kg oxycodone, 0.1 mg/kg U50–488H, 0.01 mg/kg salvinorin A, 0.001 mg/kg nalfurafine, and 0.32 mg/kg triazole 1.1. These were the largest doses evaluated in the dose-response determinations, and with the exception of triazole 1.1, significantly reduced species-typical activity, which comprises behaviors that are sensitive to the direct effects of drugs. Five subjects completed salvinorin A, because the sixth subject (5264) ran out of veins for catheterization prior to assessment of salvinorin A. On test days, saline or drug was administered i.v., 10, 20, 40, 80, 160, or 320 min prior to the observation session, and subjects completed all timepoints once. To blind observers to drug condition and control for any behavioral response to the injection itself, saline was administered 5 min prior to each observation session during time-course assessments.

Mixtures of oxycodone and KOR agonists.

Five subjects completed this condition. On test days, oxycodone (0.1 mg/kg, i.v.) was delivered 5 min prior to an initial observation session. Immediately after the initial session ended, saline, 80% propylene glycol, or a dose of a KOR agonist was delivered, followed 5-min later by a second observation session. The two largest doses from the dose-response of single drugs were used for U50–488H, nalfurafine, and triazole 1.1. Each subject experienced each KOR agonist dose once, and each vehicle twice.

Subjects completed oxycodone and KOR agonist dose-response determinations before time-course and mixture conditions. For dose-response and time-course determinations, subjects experienced each KOR agonist in a counterbalanced order. Generally, all doses or timepoints were completed for each KOR agonist before moving to the next one; however, vehicle, oxycodone, cocaine, and ketamine administrations were interspersed across the entire study in an effort to minimize the number of consecutive KOR agonist tests. In mixture conditions, subjects experienced oxycodone plus vehicle or KOR agonist in an irregular order.

Data Analysis

For dose-response determinations, saline and 80% propylene glycol were not statistically different for any behavior and were combined in the overall analyses. For each drug and each behavior, a separate repeated-measures one-way analysis of variance (ANOVA) was conducted with dose as a within-subjects variable and the results of each ANOVA are reported in Table 2. Geisser-Greenhouse corrections were applied, and Dunnett’s multiple comparisons were used to compare each dose to vehicle. For time-course assessments, a separate two-way ANOVA was conducted for each drug and each behavior with saline vs. drug as one repeated measure and timepoint as another and the results of each ANOVA are reported in Table 3. Bonferroni t-tests were used to compare each timepoint following drug administration to the equivalent saline timepoint. In mixture conditions, saline and 80% propylene glycol were not statistically different and were averaged in the analyses. For each drug and each behavior, a separate repeated-measures one-way ANOVA was conducted with drug condition (initial session oxycodone, second session vehicle, and second session with each dose of each KOR agonist) as the repeated measure. Dunnett’s multiple comparisons were used to compare each second session to oxycodone.

Table 2.

Statistical outcomes from overall analyses. Values in bold-face font met statistical significance at p<0.05

Dose-Response Determinations of Single Drugs: Main Effects of Dose
Drug	Species-Typical	Passive Visual	Rest/Sleep Posture	Lip Droop	Scratch	Facial Rub
U50–488H typical KOR agonist	F(2.7,13.7)=8.0, p<0.05	F(2.2,10.8)=3.4, p=0.07	F(1.4,6.8)=3.8, p=0.09	F(2.3,11.7)=5.0, p<0.05	F(1.6,8.2)=9.4, p<0.05	F(1.3,6.3)=0.9, p=0.40
Salvinorin A typical KOR agonist	F(2.2,12.3)=7.3, p<0.05	F(1.7,8.4)=5.3, p<0.05	F(1,5)=5.3, p=0.07	F(2.0,9.9)=7.4, p<0.05	F(2.6,13.2)=9.7, p<0.05	F(1.4,7.0)=5.0, p=0.05
Nalfurafine atypical KOR agonist	F(2.5,12.5)=3.6, p<0.05	F(1.9,9.7)=1.8, p=0.21	F(1,5)=1.0, p=0.36	F(1.1,5. 7)=1.6, p=0.26	F(2.6,12.8)=5.5, p<0.05	F(2.0,9.9)=1.1, p=0.37
Triazole 1.1 atypical KOR agonist	F(2.1,10.6)=0.8, p=0.49	F(1.7,8.7)=0.7, p=0.51	F(1,5)=1.0, p=0.36	F(1.0,5.0)=1.0, p=0.37	F(1.9,9.7)=1.1, p=0.38	F(1.8,8.8)=1.5, p=0.28
Oxycodone mu-opioid agonist	F(2.4,12.0)=6.4, p<0.05	F(1.9,9.3)=4.9, p<0.05	F(1,5)=1.0, p=0.36	F(1.3,6.7)=2.6, p=0.15	F(2.4,12.0)=3.7, p=0.05	F(1.5,7.6)=7.7, p<0.05
Cocaine centrally active (positive) control	F(2.0,10.1)=7.1, p<0.05	F(2.1,10.6)=8.6, p<0.05	F(1,5)=1.0, p=0.36	F(1.0,5.1)=1.0, p=0.37	F(2.2,11.0)=1.6, p=0.24	F(1.6,7.8)=1.0, p=0.38
Ketamine Centrally active (negative) control	F(1.9,9.7)=11.3, p<0.05	F(2.5,12.5)=5.4, p<0.05	F(1,5)=1.0, p=0.36	F(1.8,8.9)=5.5, p<0.05	F(1.5,7.4)=8.4, p<0.05	F(1.4,7.0)=0.9, p=0.40

Table 3.

Statistical outcomes from two-way repeated-measures ANOVAs. F values in bold-face font met statistical significance at p<0.05

Time-Course Assessment of Single Drugs Main Effects Timepoint (row 1), Saline vs. Drug (row 2), and the Interaction (row 3)
Drug	Species-Typical	Passive Visual	Rest/Sleep Posture	Lip Droop	Scratch	Facial Rub
U50–488H 0.1 mg/kg typical KOR agonist	F(5,25)=2.6	F(5,25)=2.9	F(5,25)=1.8	F(5,25)=1.9	F(5,25)=0.4	F(5,25)=0.7
	F(1,5)=13.6	F(1,5)=2.9	F(1,5)=6.0	F(1,5)=8.9	F(1,5)=31.8	F(1,5)=0.2
	F(5,25)=4.4	F(5,25)=3.5	F(5,25)=1.9	F(5,25)=2.2	F(5,25)=2.5	F(5,25)=1.3
Salvinorin A 0.01 mg/kg typical KOR agonist	F(5,20)=0.8	F(5,20)=0.7	F(5,20)=1.3	F(5,20)=1.9	F(5,20)=1.2	F(5,20)=0.4
	F(1,4)=3.6	F(1,4)=7.6	F(1,4)=2.3	F(1,4)=2.2	F(1,4)=27.9	F(1,4)=5.1
	F(5,20)=0.5	F(5,20)=1.1	F(5,20)=1.9	F(5,20)=1.9	F(5,20)=1.2	F(5,20)=0.4
Nalfurafine 0.001 mg/kg atypical KOR agonist	F(5,25)=1.3	F(5,25)=1.4	F(5,25)=0.8	F(5,25)=0.6	F(5,25)=0.9	F(5,25)=0.7
	F(1,5)=7.2	F(1,5)=4.7	F(1,5)=7.5	F(1,5)=8.6	F(1,5)=45.9	F(1,5)=1.2
	F(5,25)=0.3	F(5,25)=0.3	F(5,25)=1.1	F(5,25)=0.6	F(5,25)=1.0	F(5,25)=1.3
Triazole 1.1 0.32 mg/kg atypical KOR agonist	F(5,25)=0.9	F(5,25)=1.2	F(5,25)=1.0	F(5,25)=1.0	F(5,25)=1.0	F(5,25)=1.1
	F(1,5)=0.3	F(1,5)=0.1	F(1,5)=1.0	F(1,5)=1.0	F(1,5)=18.0	F(1,5)=2.1
	F(5,25)=0.7	F(5,25)=1.1	F(5,25)=1.0	F(5,25)=1.0	F(5,25)=1.3	F(5,25)=0.9
Oxycodone 0.32 mg/kg mu-opioid agonist	F(5,25)=0.2	F(5,25)=1.2	F(5,25)=0.7	F(5,25)=1.4	F(5,25)=2.2	F(5,25)=2.3
	F(1,5)=5.4	F(1,5)=2.4	F(1,5)=0.2	F(1,5)=3.1	F(1,5)=6.8	F(1,5)=7.0
	F(5,25)=2.6	F(5,25)=2.0	F(5,25)=1.4	F(5,25)=0.9	F(5,25)=2.4	F(5,25)=1.8

Drugs

Oxycodone and cocaine hydrochloride were provided by the National Institute on Drug Abuse drug supply program (Rockville, MD). U50–488H, nalfurafine hydrochloride, and salvinorin A were provided by Dr. Thomas Prisinzano (University of Kentucky; Lexington, KY). Triazole 1.1 [2-(4-(furan-2-ylmethyl)-5-((4-methyl- 3-(trifluoromethyl)benzyl)thio)-4H-1,2,4-triazol-3-yl)pyridine] was synthesized and provided by Dr. Bruce Blough at the Research Triangle Institute (Research Triangle Park, NC), and ketamine hydrochloride was purchased from Henry Schein, Inc. Oxycodone, nalfurafine, cocaine, and U50–488H were dissolved in 0.9% sterile saline, ketamine was diluted in 0.9% sterile saline, and salvinorin A and triazole 1.1 were dissolved in propylene glycol and diluted in 0.9% sterile saline to a final concentration of 80% propylene glycol. Each injection (3 ml) was passed through a 0.22 μm Millipore filter prior to administration.

Results

Figure 2 shows scores for dose-response (left panels) and time-course determinations (right panels) for species-typical activity (A), passive visual (B), rest/sleep posture (C), lip droop (D), and scratch (E) for KOR agonists. Shaded areas represent the mean and SEM following vehicle. Species-typical activity is a global measure that includes tactile/oral exploration, locomotion, foraging, and self-groom and was used to capture the different types of activity that different subjects displayed during baseline and subsequent disruption following drug administration. Data for other behaviors from Table 1 were not statistically significant and are not presented.

Fig. 2. Dose-Response (left panels) and Time-Course (right panels) Determinations for KOR Agonists.

Mean (+SEM) score following administration of each KOR agonist for species-typical activity (A panels), passive visual (B panels), rest/sleep posture (C panels), lip droop (D panels), and scratch (E panels). For time-course determinations (right panels), the dose administered across timepoints was 0.1 mg/kg for U50–488H (circles), 0.01 mg/kg of salvinorin A (upward triangles), 0.001 mg/kg of nalfurafine (downward triangles), and 0.32 mg/kg of triazole 1.1 (squares).The shaded area represents the mean (+/−SEM) following vehicle administration. Filled symbols represent doses or timepoints for each drug that were statistically different from vehicle (p’s<0.05).

Both typical KOR agonists, U50488H (0.032 and 0.1 mg/kg) and salvinorin A (0.01 mg/kg), reliably reduced species-typical activity (Fig. 2A, left panel). For U50–488H, this effect persisted until the 40-min timepoint; however, for salvinorin A, this reduction in activity was short lived and not evident at any timepoint beyond 5 min (Fig. 2A, right panel). The reduction in species-typical activity was accompanied by a concomitant increase in passive visual following administration of U50–488H (0.032 mg/kg) and salvinorin A (0.01 mg/kg; Fig. 2B, left panel). During time-course determinations, passive visual only was elevated reliably at the 20-min timepoint for U50–488H (Fig. 2B, right panel). Neither of the typical KOR agonists increased rest/sleep posture during dose-response determinations (Fig. 2C, left panel). Despite not reaching statistical significance in the overall analyses (Table 3), both U50–488H and salvinorin A significantly increased rest/sleep posture at the 10-min timepoint (Fig. 2C, right panel), indicating a somewhat delayed onset of sedation-like effects that were relatively short lived. The largest dose of both typical KOR agonists significantly increased lip droop (Fig. 2D, left panel), a measure purported to indicate muscle relaxation (Weerts et al. 1998). For U50–488H, this effect persisted for 20 min, and for salvinorin A, for 10 min (Fig. 2D, right panel). Finally, both typical KOR agonists reduced baseline levels of scratching at multiple doses (Fig. 2E, left panel) an effect that persisted across all timepoints except 320 min for U50–488H (Fig. 2E, right panel). For salvinorin A, significant reductions in scratch occurred at 10 and 160 min (Fig. 2E, right panel).

The atypical KOR agonist nalfurafine had effects similar to the typical KOR agonists in that the largest dose (0.001 mg/kg) reduced species-typical activity (Fig. 2A, left panel) to a significant level at the 10-, 20-, and 160-min timepoints (Fig. 2A, right panel). This decrease was accompanied by an increase in passive visual at 0.00032 mg/kg (Fig. 2B, left panel) despite the overall analysis for nalfurafine not reaching statistical significance (Table 2). In the time course, passive visual was not reliably present (Fig. 2B, right panel), perhaps because rest/sleep posture began to emerge, significantly so at the 20-min time point (Fig. 2C, right panel) even though rest/sleep posture was not evident in the dose-response determinations (Fig. 2C, left panel). Nalfurafine also did not significantly affect lip droop during dose-response determinations (Fig. 2D, left panel), but significant elevations in lip droop emerged at 20, 40, and 160 min (Fig. 2D, right panel). Finally, like the typical KOR agonists, nalfurafine decreased scratching during dose-response determinations (Fig. 2E, left panel), and like U50–488H, this effect was long lasting, persisting from 10–160 min (Fig. 2E, right panel). In contrast, the atypical KOR agonist, triazole 1.1, had effects unlike both typical KOR agonists and nalfurafine in that, at the doses tested, it did not significantly affect any of the behaviors measured during the dose-response (Fig. 2, left panels) or any of the behaviors measured during time-course determinations with the exception of scratch (Fig. 2, right panels). For scratch (Fig. 2E, right panel), there were significant main effects of saline vs. drug (Table 3) following triazole 1.1 administration, and the post-hoc analyses nearly met statistical significance at the 10- and 40-min timepoints (p’s=0.05).

Figure 3 shows scores for dose-response (left panels) and time-course determinations (right panels) for species-typical activity (A), passive visual (B), scratch (C), and facial rub (D) for oxycodone, cocaine, and ketamine. Oxycodone (0.32 mg/kg), cocaine (0.1–0.18 mg/kg), and ketamine (3.2–10 mg/kg) decreased species-typical activity (Fig. 3A, left panel). For oxycodone, the reduction in species-typical activity occurred at 10 and 20 min (Fig. 3A, right panel), but the overall analyses were not significant (Table 3). As species-typical activity decreased, passive visual tended to increase. This is reflected by main effects of dose (Fig. 3B, left panel; Table 2). For oxycodone, passive visual remained elevated at 10 min, however, the overall analyses were not significant (Table 3). Oxycodone did not increase scratching during dose-response determinations, narrowly missing significance (Table 2). Cocaine also did not alter scratching, and ketamine decreased scratching (3.2–10 mg/kg). During time-course determinations for oxycodone, scratching was significantly elevated at 10 and 20 min (Fig. 3C, right panel). Oxycodone increased facial rubbing (Fig. 3D, left panel), a measure purported to indicate gastrointestinal distress (Weerts et al. 1998), while cocaine and ketamine had no significant effects on this measure. For oxycodone, facial rub was elevated at 10 and 80 min during time-course determinations (Fig. 3D, right panel). Importantly, the largest ketamine dose (10 mg/kg) resulted in deep sedation (data not shown); subjects spent the majority of these sessions in deep sedation (M=17.3, SEM=2.7).

Fig. 3. Dose-Response (left panels) Determinations for Oxycodone and Control Drugs and Time-Course Determinations for Oxycodone (right panels).

Mean (+SEM) score following administration of oxycodone (circles), cocaine (upward triangles), and ketamine (downward triangles) for species-typical activity (A panels), passive visual (B panels), scratch (C panels), and facial rub (D panels). For time-course determinations (right panels), the oxycodone dose administered across timepoints was 0.32 mg/kg. The shaded area represents the mean (+/−SEM) following vehicle administration. Filled symbols represent doses or timepoints for each drug that were statistically different from vehicle (p’s<0.05).

Figure 4 shows average scores for species-typical activity (A), rest/sleep posture (B), and scratch (C) for oxycodone plus vehicle, U50–488H, nalfurafine, and triazole 1.1. Species-typical behavior was reduced by U50–488H and nalfurafine, indicated by significant main effects of drug condition [F(1.2,5.0)=8.1, p<0.05 and F(1.5,6.0)=6.0, p<0.05, respectively] but not by triazole 1.1 [p>0.05]. For U50–488H, the reduction in species-typical activity was accompanied by an increase in rest/sleep posture [F(1.2,4.9)=26.1, p<0.05]. However, rest/sleep posture was not significantly affected by nalfurafine or triazole 1.1 [p’s>0.05]. Finally, oxycodone-induced scratching was significantly reduced by all three KOR agonists [U50–488H, F(1.5,6.2)=17.9, p<0.05; nalfurafine, F(1.6,6.5)=16.3, p<0.05; triazole 1.1, F(1.9,7.6)=8.4, p<0.05].

Fig. 4. Mixtures of Oxycodone and KOR Agonists.

Mean (+SEM) score following administration of 0.1 mg/kg oxycodone prior to the first observation session (left of dashed line) and following administration of vehicle or two doses of each KOR agonist prior to the second observation session (right of dashed line). Data are shown for species-typical activity (panel A), rest/sleep posture (panel B), and scratch (panel C). Asterisks represent doses of each drug that were statistically different from oxycodone from the first session (p’s<0.05)

Discussion

The current study investigated behavioral effects of KOR agonists purported to differ in G-protein signaling bias and that resulted in a distinct behavioral profile compared with the MOR agonist, oxycodone and non-opioid drugs, cocaine and ketamine. Within the KOR-agonist class, U50–488H and salvinorin A produced effects that were qualitatively similar, while the atypical KOR agonist, nalfurafine, exhibited some slight departures across effects. Triazole 1.1 produced the most atypical profile. Time-course assessments for the KOR agonists revealed effect durations that were consistent with previous reports (Brust et al. 2016; Butelman et al. 2009; Endoh et al. 2001). All KOR agonists reduced oxycodone-induced scratching, and U50–488H and nalfurafine reduced species-typical activity, but only the combination of oxycodone and U50–488H resulted in an increase in rest/sleep posture.

In dose-response determinations, the typical KOR agonists, U50–488H and salvinorin A, produced qualitatively comparable effects across all behaviors. Salvinorin A was the more potent of the two drugs, a relation that replicates previous reports (Baker et al. 2009; Butelman et al. 2010). The atypical KOR agonist, nalfurafine, produced effects that were similar to typical KOR agonists. For example, nalfurafine decreased species-typical activity and scratch and increased passive visual. Consistent with their known anti-pruritic effects, U50–488H, salvinorin A, and nalfurafine decreased scratching at doses that did not disrupt species-typical behavior. Decreases in scratch also can result from sedation-like or motor-impairing effects as was observed with ketamine. However, with ketamine, reductions in scratch only appeared at doses that also disrupted species-typical activity. Unlike the typical KOR agonists, nalfurafine produced no immediate effect (i.e., after 5 min) on lip droop, although this behavior and rest/sleep posture were detected at later time points (see below for discussion on time course). To the extent that muscle relaxation (e.g., lip droop) contributes to motor impairment, coupled with the late onset of rest/sleep posture, these latter results are consistent with previous findings that nalfurafine produces sedative-like effects in nonhuman primates at sufficient doses (Endoh et al. 2001).

The atypical KOR agonist, triazole 1.1, produced no significant effects at any dose on any of the scored behaviors in the dose-response determinations. The dose selection for triazole 1.1 was based on its relative potency to U50–488H in behavioral tests with rats and mice (Brust et al. 2016). In that report, the relative potencies for U50–488H and triazole 1.1 depended on the behavioral endpoint being measured. U50–488H was approximately three-fold more potent than triazole 1.1 for thermal antinociception and suppression of scratching induced by chloroquine phosphate. However, unlike U50–488H, triazole 1.1 produced no effects on locomotor behavior and no reductions in brain dopamine concentrations at any of the doses (or concentrations) tested (Brust et al. 2016). In the current report, U50–488H produced lip droop at a three-fold lower dose than the highest dose of triazole 1.1 tested and reduced species-typical activity and increased passive visual at a dose that was ten-fold lower than the highest dose of triazole 1.1 tested. In the time-course assessments, triazole 1.1 reduced baseline levels of scratching and decreased oxycodone-induced scratching at a dose that had no effect on other measures, suggesting a divergence in its potency to reduce scratching relative to its potency to produce reliable effects on other behaviors compared with other KOR agonists tested. It should be noted that higher doses of triazole 1.1 could have revealed more KOR agonist-typical effects. However, challenges with solubility and the fact that this study was the first to test the drug in nonhuman primates precluded our testing of higher doses. Future work may include higher doses in nonhuman primates once dose parameters have been thoroughly explored via the intravenous route in rodents (work that is underway in our laboratory).

Using the highest doses of KOR agonists tested in dose-response analyses, time-course measurements were conducted for the same behaviors. For the drugs that produced significant effects in the dose-response tests (U50–488H, salvinorin A, and nalfurafine), the most systematic time-course effects were seen in lip droop and scratch. As expected, salvinorin A had the shortest duration of action, followed by U50–488H and then nalfurafine. Interestingly, nalfurafine’s peak effect on rest/sleep posture occurred at 20 min and on lip droop occurred at 40 min, suggesting that its time of onset is relatively delayed (for these behaviors) compared to the typical KOR agonists, each of which produced maximum effects at 10 min for rest/sleep posture and 5 min for lip droop. These findings are in agreement with previous reports demonstrating that nalfurafine produces peak antinociception and sedation-like or motor-impairing effects at later time points than U50–488H in nonhuman primates and mice (Endoh et al. 1999, 2001). Interestingly, not all of nalfurafine’s effects were delayed. Reductions in species-typical activity and scratch and increases in passive visual were observed 5 min following administration in dose-response tests.

For a number of reasons, KOR agonists have been investigated as adjunct treatments with MOR agonists. KOR agonists reduce the abuse-related effects of MOR agonists and produce antinociceptive effects in rodents (Kaski et al. 2019; Minervini et al. 2018; Townsend et al. 2017) and in nonhuman primates (Ko and Husbands 2009; Negus et al. 2008; Zamarripa et al. 2020), findings that could lead to dose-sparing strategies for MOR agonists when the two are combined. Moreover, it is well established that KOR agonists reduce pruritus, including itch that is induced by MOR agonists (see Beck et al. 2019). However, there is the possibility that combinations of KOR and MOR agonists could produce undesirable effects. In the current report, oxycodone-induced scratching was reduced by U50,488H and nalfurafine. Notably, U50–488H and nalfurafine also reduced species-typical activity, but only the largest U50,488H dose increased rest/sleep posture when administered after oxycodone, suggesting that doses of U50–488H and nalfurafine that reduced MOR agonist-induced scratching also caused sedation-like and/or motor-impairing effects. Triazole 1.1 reliably reduced oxycodone-induced scratching at the highest dose tested without causing sedation-like or motor-impairing effects, consistent with a previous report in rodents (Brust et al. 2016). Taken together, our results suggest, when combined with MOR agonists, atypical KOR agonists produce a side-effect profile that is relatively benign (i.e., triazole 1.1) or reduced (i.e., nalfurafine) compared with the typical KOR agonist, U50–488H.

A limitation of this study is that the experiments were conducted in male subjects only. Conclusions based on the current data should therefore be made with the understanding that sex differences in responsivity to KOR activation have been widely reported (Becker and Chartoff 2019; Chartoff and Mavrikaki 2015). For example, activation of KORs in mice produced greater anhedonic-like effects in males than in females tested in the ICSS procedure (Russell et al. 2014). However, the KOR agonist U50,488H was more potent at producing conditioned place aversions in female than in male mice (Robles et al. 2014). Moreover, KOR agonists were more potent at increasing prolactin levels in female than in male human and nonhuman primates (Butelman et al. 2007; Kreek et al. 1999) but less potent at producing antinociception in female nonhuman primates (Negus et al. 2002). Whether or not atypical KOR agonists will exhibit comparable departures in behavioral effects in female monkeys relative to the typical KOR agonists studied in the current report remains to be determined.

When administered alone, our results demonstrate that atypical KOR agonists purported to be G-protein biased do not necessarily produce similar behavioral profiles in nonhuman primates. With the exception of time of onset, nalfurafine alone produced effects that were similar to typical KOR agonists while triazole 1.1 had no effects on its own, with the exception of reducing overall levels of scratching during time-course determinations, across the dose range tested. Similar behavioral differences between nalfurafine and triazole 1.1 have been reported in rats. Lazenka et al (2018) reported that anti-pruritic doses of nalfurafine also reduced intra-cranial self-stimulation (ICSS), while doses of triazole 1.1 as much as 8 times higher than an anti-pruritic dose did not affect ICSS responding in rats (Brust et al. 2016). It is possible that structural distinctions in the scaffoldings of nalfurafine and triazole 1.1 are responsible for signaling nuances beyond G-protein signaling and ß-arrestin recruitment that may account for the behavioral differences observed between these atypical KOR agonists. Thus, based on the current and previous data, we conclude that structural modifications to KOR agonists can increase therapeutic selectivity, but the mechanism(s) by which selectivity is increased cannot solely be accounted for by G-protein signaling bias. Other mechanisms yet to be determined may play a role in increasing therapeutic selectivity of KOR agonists.