Activation of Trace Amine-Associated Receptor 1 Selectively Attenuates the Reinforcing Effects of Morphine

Abstract

Background and Purpose:

Trace amine-associated receptor 1 (TAAR1) plays a critical role in regulating dopamine transmission. Previous studies showed that pharmacologically or genetically manipulating the activity of TAAR1 modulates addiction-like behaviors associated with psychostimulants. However, little is known about whether TAAR1 modulation would regulate the behavioral effects of opioids.

Experimental Approach:

Here, we systematically assessed the effects of the selective TAAR1 partial agonist RO5263397 on the addiction-related and antinociceptive effects of morphine in male rats and mice.

Key Results:

We found that RO5263397 attenuated the expression of morphine-induced behavioral sensitization in wildtype but not TAAR1 knockout mice. RO5263397 shifted the dose-effect curve of morphine self-administration downward and reduced the breakpoint in a progressive ratio schedule of reinforcement, but did not affect food self-administration in rats. RO5263397 decreased the cue- and drug-induced reinstatement of morphine-seeking behavior in rats. RO5263397 alone did not trigger the reinstatement of morphine-seeking behavior or change locomotor activity in rats that had a history of morphine self-administration. However, RO5263397 did not affect the expression of morphine-induced conditioned place preference in mice or rats. RO5263397 did not affect naltrexone-precipitated jumping behavior or naltrexone-induced conditioned place aversion in morphine-dependent mice. Furthermore, RO5263397 did not affect the analgesic effects of morphine in an acute nociception model in mice and a chronic pain model in rats.

Conclusion and Implications:

These results indicated that TAAR1 activation selectively attenuated the reinforcing but not withdrawal or antinociceptive effects of morphine, suggesting that selective TAAR1 agonists might be useful to combat opioid addiction while sparing the strong analgesic effects.

INTRODUCTION

Opioids are widely used in the clinic for pain management due to their highly effective analgesic effects (Busse et al., 2018). However, prolonged use of opioids could lead to opioid addiction, which is characterized by compulsive opioid taking and seeking behaviors as well as a high rate of relapse (Volkow et al., 2019). The opioid epidemic has brought heavy health and economic burdens to individuals and society (Volkow et al., 2019). One of the strategies to combat the opioid epidemic is to develop novel pharmacological treatments that prevent opioid addiction without influencing the analgesic effects of opioids.

It is suggested that attenuating the activity of the mesolimbic dopaminergic system would be a promising strategy to control opioid addiction (Nutt et al., 2015). The trace amine-associated receptor 1 (TAAR1), one of the receptors of endogenous trace amines, has been shown to negatively modulate dopamine transmission (Revel et al., 2011). Previous studies showed that selective TAAR1 agonists could reduce the dopamine release in the nucleus accumbens (NAc) induced by psychostimulants (Liu et al., 2018b; Pei et al., 2017; Pei et al., 2014). Behavioral analyses showed that TAAR1 agonists attenuated while knockout of taar1 augmented addiction-related behaviors associated with psychostimulants (Achat-Mendes et al., 2012; Liu et al., 2018a; Liu et al., 2018b; Pei et al., 2014; Thorn et al., 2014). It is clear that opioids are pharmacologically and mechanistically different from psychostimulants (Badiani et al., 2011; Graziane et al., 2016). However, so far, little is known about the role of TAAR1 in opioid addiction (Liu et al., 2018a). It was shown that knockout of TAAR1 did not affect morphine-induced conditioned place preference (CPP) in mice (Achat-Mendes et al., 2012). However, more systematic studies are needed to elucidate the role of TAAR1 in regulating morphine-induced maladaptive plasticity in the brain and other morphine-associated addiction-related behaviors (Alcantara et al., 2011; Venniro et al., 2016). Given that TAAR1 negatively regulates dopamine transmission that is critical for opioid addiction, we hypothesized that TAAR1 might be involved in other opioid-induced behaviors such as behavioral sensitization and opioid self-administration (Venniro et al., 2016; Vigano et al., 2003).

Our recent study showed that activation of TAAR1 attenuated the pharmacological stressor yohimbine-induced reinstatement of cocaine-seeking behavior, suggesting that TAAR1 might regulate stress responses to contribute to drug addiction (Liu et al., 2020a). Evidence shows that the brain stress systems regulate negative reinforcement associated with opioid withdrawal (Koob, 2020). It is of interest to see whether pharmacological activation of TAAR1 would modulate the negative reinforcement process associated with opioid withdrawal. Besides, it is shown that the TAAR1 agonist 3-Iodothyronamine could reduce pain threshold to hot stimuli, suggesting that TAAR1 might regulate pain sensitivity (Manni et al., 2013). Many behavioral pain models have been developed to assess different types of pain in animals (Laurino et al., 2017). For example, our previous studies used a hot water tail immersion test as an acute thermal pain model and complete Freund’s adjuvant (CFA)-induced inflammation pain as a chronic pain model (Li et al., 2013; Siemian et al., 2018). Certain drug targets may participate in acute and chronic pain differently due to their mechanistic differences (Pogatzki-Zahn et al., 2018). Given the essential role of opioids in pain management, it is important to examine whether TAAR1 agonists would impact the analgesic effects of opioids.

Here, we systemically investigated the effects of the selective TAAR1 partial agonist RO5263397 on behaviors associated with the rewarding and reinforcing, withdrawal, and antinociceptive properties of morphine.

MATERIALS AND METHODS

Materials and methods are briefly described below. Detailed materials and methods are provided in the Supplementary Materials and Methods.

Animals

A total of 124 adult male Sprague–Dawley rats (RRID: RGD_1566457) and 178 male C57BL6J mice (RRID: IMSR_JAX:000664) were used in the study. Dr. Gregory M. Miller at Northeastern University gifted breeding pairs of TAAR1 knockout (TAAR1-KO) mice of C57BL6/J background (Achat-Mendes et al., 2012). All experimental procedures were approved by the Institutional Animal Care and Use Committee, University at Buffalo, the State University of New York, and with the 2011 Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources on Life Sciences, National Research Council, National Academy of Sciences, Washington DC).

Drugs

Dr. Yanan Zhang at Research Triangle Institute provided RO5263397 (purity > 98%). All doses of RO5263397 (3.2 and 5.6 mg/kg, i.p.) for rats were selected based on our previous studies (Liu et al., 2018b). We chose 5.6 mg/kg as the highest dose of RO5263397 for rats in the present study since our previous study found that 10 mg/kg RO5266397 showed a slight inhibitory effect on rates of responding on a food-associated lever (Liu et al., 2018b). Doses of RO5263397 for mice were based on previous study (Revel et al., 2013) and a locomotor activity test in the present study (Figure 1B). RO5263397 was dissolved in the vehicle, which contains 1 part absolute ethanol, 1 part Emulphor-620 (Rhodia), and 18 parts physiologic saline (Liu et al., 2018b). Morphine sulfate was provided by Research Technology Branch, National Institute and Drug Abuse, National Institutes of Health (Rockville, MD, USA), and was dissolved in 0.9% saline. Naltrexone hydrochloride (0.1 and 1 mg/kg, i.p.) was purchased from Sigma-Aldrich and dissolved in saline. CFA was purchased from Difco (Detroit, MI, USA) and dissolved in paraffin oil (Siemian et al., 2018).

Figure 1. Selective TAAR1 partial agonist RO5263397 attenuates morphine-induced behavioral sensitization in wildtype but not TAAR1 knockout mice.

(A) Experimental timelines for locomotion test (I) and behavioral sensitization (II). (B) RO5263397 at the dose of 1 mg/kg (i.p.) only slightly reduced locomotor activity in mice; therefore, we used 1 mg/kg (i.p.) as the maximum dose of RO5263397 in the following mice experiments (n = 7 for the vehicle group; n = 6/group for other groups). (C) RO5263397 at doses of 0.32 and 1.0 mg/kg (i.p.) decreased the expression of morphine-induced sensitization in mice (n = 7 for the WT-0.32 RO526 group, n = 8/group for other groups). (D) RO5263397 at the dose of 1.0 mg/kg did not affect the expression of morphine-induced sensitization in TAAR1 knockout (TAAR1-KO) mice (n= 7/group). Data are expressed as mean ± SEM; *p < 0.05.

Locomotor activity and morphine-induced behavioral sensitization in mice

For the locomotor activity test, RO5263397 (0.1, 0.32, 1.0, and 10 mg/kg, i.p.) or vehicle was administered 10 min before the test session of locomotor activity started. The procedure for morphine-induced behavioral sensitization in mice required a total of 15 days. On day 1, a 105-min session was conducted during which a morphine dose-effect curve was determined by using a cumulative dosing procedure (cumulative doses of 1, 3.2, and 10 mg/kg, i.p.). From day 2 to day 7, mice received daily injections of a single dose of morphine (10 mg/kg, i.p.) and were placed in the locomotion chambers (1 h exploration) immediately after morphine administration. From day 8 to day 14, mice were kept in their home cages with no drug treatment. On day15, a final dose-effect curve of morphine was re-determined, which was identical to the test on day 1 except that RO5263397 (1, 3.2, and 10 mg/kg, i.p.) and vehicle rather than saline were administered 10 min before the test session.

Catheterization Surgery and morphine self-administration in rats

The rats (weighing 280–300 g at the beginning of the experiments) were anesthetized with ketamine and xylazine (75 and 5 mg/kg, respectively, i.p.). Rats were implanted with chronic indwelling jugular catheters as previously described (Liu et al, 2018). The rats were allowed to recover for at least 1 week after surgery. Catheters were flushed daily with 0.2 ml solution of enrofloxacin (4 mg/ml) mixed in a heparinized saline solution (50 IU/ml in 0.9% sterile saline) for 1 week after surgery to preserve catheter patency and prevent infection.

Morphine self-administration procedures are described in the Supplementary Materials and Methods. Briefly, one week after surgery, rats began morphine self-administration (0.3 mg/kg/infusion; 2 hours/session/day). The response requirement was gradually increased from FR 1 to FR 5 in 3 weeks (training) and then maintained under FR 5 for 10 days before the morphine-intake procedure, progressive ratio schedule of reinforcement, and extinction tests as described below. Separate groups of rats were used for the morphine intake test, the progressive ratio schedule of reinforcement, and the reinstatement test.

In this study, three rats were excluded because of catheter failure. Four rats were excluded because they did not achieve stable responses after training (the variance of the total number of injections was <20% for two consecutive days). Separate groups of rats were used for the morphine intake test, the progressive ratio schedule of reinforcement, and the reinstatement test.

Morphine intake test in rats

The morphine intake test was conducted starting from one day after the maintenance of morphine self-administration. Different doses of morphine (0.3, 0.1, and 0.03 mg/kg/infusion; 2 hours/session/day) were substituted for the maintenance dose of morphine among animals. Vehicle or RO5263397 (5.6 mg/kg, i.p.) was administered 10 min before the tests.

Progressive ratio schedule of reinforcement in rats

One day after the maintenance of morphine self-administration, rats underwent a progressive ratio schedule of reinforcement. Rats were allowed to self-administer morphine during a 3-h session. The first response of the session resulted in a drug infusion, after which the response requirements escalated by following the series 1, 2, 4, 6, 9, 12, 15, 20, 25, 32, …, derived from the formula: Response ratio (rounded to nearest integer) = [5e^{(injection numberX0.2)}] – 5 (Liu et al., 2018b). Vehicle or RO5263397 (5.6 mg/kg, i.p.) was administered 10 min before the tests.

Cue- and drug-induced reinstatement of morphine-seeking in rats

One day after the maintenance of morphine self-administration, rats underwent 6 daily extinction sessions (2-h session), during which lever presses had no consequence (no drug or cue). At the end of extinction, all rats in the study reached the extinction criteria (less than 10 active lever responses per 2 h over two consecutive sessions). One day after the last extinction, rats were tested cue-induced reinstatement (FR5; 2 h) during which responding in the active lever (FR5) resulted in the presentation of lights with no drug delivery. A 2-h re-extinction session was conducted 1 day after cue-induced reinstatement. Drug-induced reinstatement (2 h) was performed 24 hour after re-extinction with a counterbalanced design for the two morphine groups of rats. Rats receiving vehicle or RO5263397 (5.6 mg/kg, i.p.) 10 min before cue- and drug-induced reinstatement of morphine-seeking.

To test the non-specific effects of RO5263397, a separate cohort of rats (n = 10) were trained to self-administer morphine and received extinction. One day after extinction, all rats received an injection of RO5263397 (5.6 mg/kg, i.p.) 10 min before an additional extinction session (2 h; no cue or drug delivery) to test whether RO5263397 could reinstate morphine-seeking. One day later, a cue-induced reinstatement of morphine-seeking test was conducted in all rats (no pretreatment of RO5263397). Rats were then randomly assigned into two groups (n = 5/group) for the locomotor activity test. Vehicle or RO5263397 (5.6 mg/kg, i.p.) was administered 10 min before the locomotor activity test.

Morphine-induced conditioned place preference (CPP) in rodents

The procedure for morphine-induced CPP was the same as in our previous study (see detail in Supplementary Materials and Methods) (Liu et al., 2016). For mice, vehicle, 0.32, and 1.0 mg/kg RO5263397 were administered 10 min before the CPP test (Post-test). For rats, 3.2 mg/kg RO5263397 or vehicle (i.p.) was administered 10 min before the CPP test (Post-test 1). One day later, a second test was conducted with a counterbalanced design (Post-test 2). In other words, rats that received RO526 treatment in the post-test 1 were administered vehicle in the post-test 2, and vice versa. 5.6 mg/kg RO5263397 or vehicle (i.p.) was administered 10 min before Post-test 2.

Morphine withdrawal and conditioned place aversion (CPA)

Jumping was counted during a 30-min period immediately after naltrexone administration. Vehicle or 1.0 mg/kg RO5263397 was administered 10 min before naltrexone administration (post-test). For the CPA, the conditioning of CPA was conducted on days 5 and 6 (see detail in Supplementary Materials and Methods). Two post-tests were performed on days 7 and 8. Mice were counterbalanced between post-tests 1 and 2. On day 7, the two groups of mice were administered vehicle or 0.32 mg/kg RO5263397 (post-test 1). On day 8, mice were administered vehicle or 1.0 mg/kg RO5263397 (post-test 2).

Hot water tail immersion test and CFA-induced inflammatory test

In the hot water tail immersion test in mice, vehicle, 0.32, or 1.0 mg/kg RO5263397 was administered 10 min before the test. In the CFA-induced inflammatory test in rats, vehicle, 3.2 and 5.6 mg/kg RO5263397, and 0.1 mg/kg naltrexone were administered 10 min before the test (Figure 5C).

Figure 5. RO5263397 does not affect the antinociceptive effects of morphine.

(A) The experimental timeline for the hot water tail immersion test in mice. (B) RO5263397 at doses of 0.32 and 1.0 mg/kg did not affect the analgesic effects of morphine in the hot water tail immersion test in mice (n = 8/group). (C) RO5263397 did not show an analgesic property in the hot water tail immersion test in mice. (D) The experimental timeline for the CFA-induced inflammatory test in mice. (E) Naltrexone but not RO5263397 prevented analgesic effects of morphine in the CFA-induced inflammatory test in rats (n = 6/group). (F) RO5263397 did not show an analgesic property in the CFA-induced inflammatory test in rats. Data are expressed as mean ± SEM; *p < 0.05.

Group assignment

Animals in the morphine-intake, progressive ratio schedule of reinforcement, and reinstatement of morphine-seeking experiments were assigned into different groups according to the responses in the maintenance of morphine self-administration so that all groups in the same experiment had similar levels of morphine self-administration. Each group of rats then received vehicle or a particular dose of RO5263397 with a random assignment before the subsequent behavioral testing. Two-way repeated-measures ANOVA was conducted to analyze the morphine self-administration data to ensure the groups were not statistically significant. Animals in other experiments were randomly assigned into different groups before the start of experiments. All experiments in this study were designed to generate groups of equal size, using randomisation and blinded analysis. Unequal group size in the final analysis was due to exclusion according to predetermined criteria. Power analysis was used to ensure the group size was sufficient to achieve statistical power.

Data analyses

All results were presented as mean ± SEM and analyzed by the Graphpad Prism 8 software (GraphPad Software, San Diego, CA) (RRID: SCR_002798). Two-tailed unpaired t-tests were conducted for the right panels of Figure 4B and 4D as well as the progressive raito schedule of reinforcement (Figure 2F). One-way ANOVA was conducted for saline rats in the morphine intake experiment (V at the X-axis of Figure 2B) and Figure 3E. All other behavioral data were analyzed by two-way repeated-measures ANOVA with assumed sphericity (within-subjects and between-subjects factors were specifically stated in the Results section) followed by post hoc Bonferroni’s multiple comparisons test. Post hoc tests were conducted only if F in ANOVA achieved the necessary level of statistical significance (P < 0.05) and there was no significant variance in homogeneity (which precludes use of parametric statistics). Group size was selected based on our previous studies. The declared group size is the number of independent values, and that statistical analysis was done using these independent values (i.e., not treating technical replicates as independent values). Statistical analysis was undertaken only for studies where each group size was at least n = 5. Exact P values were reported except for those that were less than 0.05. P < 0.05 was considered statistically significant. The manuscript complies with BJP’s recommendations and requirements on experimental design and analysis (Atlan et al., 2018).

Figure 4. RO5263397 does not affect morphine-induced CPP or morphine withdrawal.

(A) RO5263397 at doses of 0.32 and 1.0 mg/kg did not affect the expression of morphine-induced CPP in mice (vehicle, n = 8; 0.32 RO5263397, n =7; 1.0 RO5263397, n = 8). RO5263397 at the dose of 1.0 mg/kg did not change place preference in mice conditioned with saline (vehicle, n = 7; 1.0 RO5263397, n =8). (B) RO5263397 at dose of 3.2 mg/kg and 5.6 mg/kg did not affect morphine-induced CPP in rats (n = 10/group). (C) RO5263397 at doses of 1.0 mg/kg did not affect the expression of naltrexone-precipitated jumping behavior in morphine-dependent mice (n = 7/group). (D) RO5263397 at dose of 0.32 mg/kg and 1.0 mg/kg did not affect the expression of naltrexone-induced CPA in morphine-dependent mice (left panel: vehicle n = 11; 0.32 RO5263397, n = 13; right panel: vehicle n = 13; 1.0 RO5263397 n = 13). Data are expressed as mean ± SEM; *p < 0.05.

Figure 2. RO5263397 reduces morphine intake and motivation to self-administer morphine in rats.

(A) The experimental timeline for morphine intake. (B) RO5263397 at doses of 3.2 and 5.6 mg/kg (i.p.) reduced morphine intake (n = 8/group). (C) RO5263397 did not affect food self-administration. (D) The experimental timeline for the progressive ratio schedule of reinforcement. No treatment was given during morphine self-administration. Vehicle or RO5263397 was administered before the progressive ratio schedule of reinforcement (vehicle, n = 9; RO5263397, n = 8). (E) The two groups of morphine rats showed no difference between each other. (F) RO5263397 reduced breakpoint, earned infusions of morphine, and total active lever responses in the progressive ratio schedule of reinforcement. Data are expressed as mean ± SEM; *p < 0.05.

Figure 3. RO5263397 decreases cue- and drug-induced reinstatement of morphine-seeking behavior in rats.

(A) The experimental timeline for the cue- and drug-induced reinstatement of morphine-seeking. (B) Rats maintained self-administration followed by extinction. (C) RO5263397 (5.6 mg/kg, i.p.) reduced the cue-induced reinstatement of morphine-seeking (vehicle, n = 8; RO5263396, n = 9). (C) RO5263397 (5.6 mg/kg, i.p.) reduced the cue-induced reinstatement of morphine-seeking (vehicle, n = 9; RO5263396, n = 8). (E) RO5263397 did not trigger reinstatement of morphine-seeking behavior in rats that self-administered morphine (n = 10). Data are expressed as mean ± SEM; *p < 0.05.

Nomenclature of Targets and Ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to Pharmacology (Harding et al., 2018), and are permanently archived in the Concise Guide to Pharmacology 2019/20 (Alexander et al., 2019).

RESULTS

1. Selective TAAR1 partial agonist RO5263397 attenuated morphine-induced behavioral sensitization in wildtype but not TAAR1 knockout mice.

We first tested the effects of the selective TAAR1 partial agonist RO5263397 on locomotor activity in mice (Figure 1A, panel I). RO5263397 slightly decreased locomotion at the dose of 1 mg/kg and dramatically reduced locomotion at the dose of 10 mg/kg in mice (two-way repeated-measures ANOVA for time course; F_{20, 130} = 6.07, P < 0.05; one-way ANOVA for total distance; F_{4, 26} = 16.31, P < 0.05; Figure 1B). Therefore, we used 1 mg/kg RO5263397 as the maximum dose in the following mice experiments. We then examined the effects of RO5263397 on morphine-induced behavioral sensitization (Figure 1A, panel II). Data of morphine-induced behavioral sensitization were analyzed by two-way repeated-measured ANOVA with morphine as within-subjects factor and RO5263397 dose as between-subjects factor. In the wildtype mice, acute administration of morphine dose-dependently increased locomotor activity on day 1 (two-way repeated-measures ANOVA; significant effect of morphine: F_{2, 54} = 14.06, P < 0.05, but no effect of RO5263397 dose: F_{3, 27} = 1.88, P = 0.16 or morphine × RO5263397 dose: F_{6, 54} = 0.65, P = 0.69; Figure 1C). For day 15, two-way repeated-measures ANOVA revealed significant main effects of morphine (F_{2, 54} = 20.31, P < 0.05), group (F_{3, 27} = 6.06, P < 0.05), and an interaction of morphine × RO5263397 dose (F_{6, 54} = 2.52, P < 0.05). Post hoc analysis showed that RO5263397 at doses of 0.32 and 1 mg/kg significantly reduced the expression of morphine-induced behavioral sensitization (day 15; at 10 mg/kg morphine; 0.32 RO5263397 vs vehicle: t₁₃ = 3.51, P < 0.05; 1.0 RO5263397 vs vehicle: t₁₄ = 4.70, P < 0.05; Figure 1C). In the TAAR1-KO mice, morphine dose-dependently induced higher levels of locomotor activity in TAAR1-KO mice in both tests (main effect of morphine; day 1: F_{2, 24} = 22.70, P < 0.05; day 15: F_{2, 24} = 25.84, P < 0.05; Figure 1D). However, RO5263397 (1.0 mg/kg) had no effect on the expression of morphine-induced sensitization on day 15 in TAAR1 knockout mice (main effect of RO5263397, F_{1, 12} = 0.34, P = 0.57; no interaction of RO5263397 × morphine, F_{2, 24} = 0.58, P = 0.57; Figure 1D), indicating that the inhibitory effect of RO5263397 on morphine-induced sensitization was not due to an off-target effect. Taken together, RO5263397 reduced the expression of morphine-induced sensitization in wildtype but not TAAR1 knockout mice.

2. RO5263397 reduced total morphine intake and the motivation to self-administer morphine in rats.

We then asked whether RO5263397 could affect morphine intake in the morphine self-administration model. Vehicle or RO5263397 (3.2 and 5.6 mg/kg, i.p.) was administered 10 min before morphine self-administration (Figure 2A). Two-way repeated-measures ANOVA (morphine as within-subjects factor and RO5263397 dose as between-subjects factor) showed significant main effects of morphine (F_{2, 40} = 9.94, P < 0.05) and RO5263397 dose (F_{2, 20} = 30.59, P < 0.05), but no interaction of morphine × RO5263397 dose (F_{2, 40} = 9.94, P = 0.43). Post hoc analysis showed that both doses of RO5263397 reduced morphine intake at all doses of morphine (all P < 0.05; Figure 2B). RO5263397 did not affect saline administration in the control group that self-administered saline rather than morphine (V on the x-axis of Figure 2B; one-way repeated-measures ANOVA: F_{2, 14} = 2.12, P = 0.17). To examine whether RO5263397 would produce a non-specific inhibition, we tested the effects of RO5263397 on normal food self-administration in rats. One-way ANOVA analysis revealed no significant effect (F_{2, 15} = 0.06, P = 0.94; Figure 2C), suggesting that RO5263397 did not affect food self-administration.

To examine the effects of RO5263397 on motivation to self-administer morphine, we tested the effects of RO5263397 on the breakpoint of morphine self-administration in a progressive ratio schedule of reinforcement (Figure 2D). Rats that self-administered morphine were assigned into two groups according to the performance in the self-administration. For the maintenance of morphine self-administration, two-way repeated-measures ANOVA analysis with day as within-subjects factor and RO5263397 dose as between-subjects factor did not show any significant effect (main effect of day, F_{9, 150} = 0.77, P = 0.64; main effect of RO5263397 dose, F_{1, 150} = 22.70, P = 0.48; interaction of RO5263397 dose × day, F_{4, 36} = 0.32, P = 0.97; Figure 2E). The progressive ratio schedule of reinforcement was conducted one day after morphine self-administration. Two-tailed unpaired t-test showed that 5.6 mg/kg RO5263397 significantly decreased breakpoint (t₁₅ = 6.42, P < 0.05), earned infusion of morphine (t₁₅ = 7.03, P < 0.05), and total active lever presses (t₁₅ = 7.11, P < 0.05; Figure 2F). These results showed that RO5263397 reduced morphine intake and motivation to self-administer morphine in rats.

3. RO5263397 decreased cue- and drug-induced reinstatement of morphine-seeking behavior in rats.

We then asked whether RO5263397 would inhibit relapse to morphine self-administration by using the cue- and drug-induced reinstatement of morphine-seeking behavior (Figure 3A). Rats that self-administered morphine were assigned into vehicle and RO5263397 groups based on the response level during the maintenance of morphine self-administration. Two-way repeated-measures ANOVA with day as within-subjects factor and RO5263397 dose as between-subjects factor was conducted to anlyaze active lever presses for training (main effect of RO5263397 dose, F_{1, 15} = 0.14, P = 0.71; main effect of day, F_{9, 135} = 2.25, P < 0.05; interaction of RO5263397 dose × day, F_{9, 135} = 0.20, P = 0.99). The two groups of rats showed no difference duing extinction (main effect of RO5263397 dose, F_{1, 15} = 0.98, P = 0.34; main effect of day, F_{7, 105} = 11.53, P < 0.05; interaction of RO5263397 dose × day, F_{7, 105} = 0.59, P = 0.76; Figure 3B). RO5263397 (5.6 mg/kg, i.p.) was administered 10 min before the cue- and drug-induced reinstatement of morphine-seeking. For the cue-induced reinstatement, two-way repeated-measures ANOVA with cue as within-subjects factor and RO5263397 dose as between-subjects factor revealed significant main effects of cue (F_{1, 15} = 37.29, P < 0.05) and RO5263397 dose (F_{1, 15} = 4.75, P < 0.05), and an interaction of cue × RO5263397 dose (active lever presses: F_{1, 15} = 10.19, P < 0.05; Figure 3C). Post hoc analysis indicated that morphine-associated discrete cue induced reinstatement of morphine seeking (active lever presses; vehicle vs RO5263397: t₁₅ = 6.77, P < 0.05), which was reduced by RO5263397 (active lever presses; vehicle vs RO5263397: t₁₅ = 3.72, P < 0.05; Figure 3C). For the drug-induced reinstatement, two-way repeated-measures ANOVA with drug as within-subjects factor and RO5263397 dose as between-subjects factor revealed significant main effects of drug (F_{1, 15} = 14.90, P < 0.05) and RO5263397 dose (F_{1, 15} = 5.33, P < 0.05), and an interaction of drug × RO5263397 dose (active lever presses: F_{1, 15} = 7.84, P < 0.05; Figure 3D). Post hoc analysis showed that RO5263397 also attenuated drug-induced reinstatement of morphine-seeking (active lever presses; vehicle vs RO5263397: t₁₅ = 3.59, P < 0.05; Figure 3D). RO5263397 did not affect inactive lever presses in the cue- or drug-induced reinstatement tests (cue: F_{1, 15} = 0.62, P = 0.44; drug: F_{1, 15} < 0.01, P = 0.96; Figure 3C, D).

We then tested whether RO5263397 alone (5.6 mg/kg, i.p.) could reinstate morphine-seeking behavior. A separate cohort of rats were trained to self-administer morphine and received extinction, as described above. We found that discrete cue rather than RO5263397 reinstated morphine-seeking in these rats (active lever presses; one-way repeated-measures ANOVA, F_{2, 18} = 34.45, P < 0.05; Post hoc analysis, RO5263397 vs last extinction: t₁₈ = 0.15, P > 0.99; cue vs last extinction: t₁₈ = 7.26, P < 0.05; Figure 3E). To further determine the potential non-specific effects of RO5263397, a locomotor activity test was conducted with the same rats one day after the cue-induced reinstatement morphine-seeking test. We found that RO5263397 (5.6 mg/kg, i.p.) did not affect the locomotor activity in these rats (total traveled distance in 30 min; Veh vs. RO526: 35.73 ± 3.17m vs. 34.34 ± 1.27 m; t₈ = 0.40, P = 0.70). Collectively, these results indicated that RO5263397 decreased cue- and drug-induced reinstatement of morphine-seeking behavior in rats.

4. RO5263397 did not affect morphine-induced CPP or morphine withdrawal

A previous study showed that knockout of TAAR1 did not affect morphine-induced CPP in mice, suggesting that the endogenous TAAR1 might not be important for morphine reward memory (Achat-Mendes et al., 2012). To further clarify the role of TAAR1 in the rewarding effect of morphine, we examined the effects of RO5263397 on the expression of morphine-induced CPP in mice. Two-way repeated-measures ANOVA with test as within-subjects factor and RO5263397 dose as between-subject factor revealed a significant effect of test (F_{1, 40} = 18.54, P < 0.05) but no main effect of RO5263397 dose (F_{2, 40} = 0.17, P = 0.85) or interaction of test × RO5263397 dose (F_{2, 40} = 0.04, P = 0.96; Figure 4A, left panel). RO5263397 (1.0 mg/kg) also did not affect the place preference in mice conditioned with saline (main effect of RO5263397 dose: F_{1, 13} < 0.01, P = 0.97; Figure 4A, right panel). Given that the species-dependent stereoselectivity of TAAR1 would affect the effects of TAAR1 agonists in different species (Reese et al., 2007, 2014), we further assessed the effects of RO5263397 (3.2 mg/kg) on morphine-induced CPP in rats. Two-way repeated-measures ANOVA of post-test 1 with test as within-subjects factor and RO5263397 dose as between-subjects factor indicated a significant effect of test (F_{1, 36} = 12.88, P < 0.05), but no main effect of RO5263397 dose (F_{1, 36} = 0.01, P = 0.96) or interaction of test × RO5263397 dose (F_{1, 36} = 0.22, P = 0.64; Figure 4B, left panel). Moreover, 5.6 mg/kg RO5263397 also did not affect the expression of morphine-induced CPP in rats (post-test 2; two-tailed unpaired t-test; t₁₈ = 0.20, P = 0.85; Figure 4B, right panel).

We then tested the effects of RO5263397 on morphine withdrawal by using the naltrexone-induced jumping behavior and naltrexone-induced CPA in morphine-dependent mice. For the jumping behavior, two-way repeated-measures ANOVA analysis with test as within-subjects factor and RO5263397 dose as between-subjects factor indicated a significant effect of test (F_{1, 24} = 59.47, P < 0.05), but no main effect of RO5263397 dose (F_{1, 24} = 0.01, P = 0.93) or interaction of test × RO5263397 dose (F_{1, 24} < 0.01, P = 0.96), indicating that RO5263397 did not affect naltrexone-induced jumping (Figure 4C). For the post-test 1 of naltrexone-induced CPA (0.32 mg/kg RO5263397), two-way repeated-measures ANOVA with test as within-subjects factor and RO5263397 dose as between-subjects factor revealed a significant effect of test (F_{1, 22} = 54.78, P < 0.05), but no main effect of RO5263397 dose (F_{1, 22} = 0.43, P = 0.52) or interaction of test × RO5263397 dose (F_{1, 22} = 0.57, P = 0.46; Figure 4D, left panel). RO5263397 at the dose of 1.0 mg/kg also did not affect the expression of naltrexone-induced CPA in the post-test 2 (two-tailed unpaired t-test; t₂₂ = 0.66, P = 0.52; Figure 4D, right panel). These results indicated that RO5263397 did not affect morphine reward memory or morphine withdrawal.

5. RO5263397 did not affect the antinociceptive effects of morphine.

We then asked whether TAAR1 activation would modulate the analgesic effect of morphine by using the hot water tail immersion test in mice. Two-way repeated-measures ANOVA analysis with test as within-subjects factor and RO5263397 dose as between-subjects factor revealed a significant main effect of test (F_{4, 84} = 101.3, P < 0.05) but no main effect of RO5263397 dose (F_{2, 21} = 1.54, P = 0.24) or interaction of test × RO5263397 dose (F_8,84 = 0.51, P = 0.85), indicating that RO5263397 did not affect the analgesic effect of morphine on acute thermal pain (Figure 5B). Further, we showed that RO5263397 alone did not affect tail withdrawal within 60 min after administration (main effect of RO5263397 dose: F_{2, 20} = 0.09, P = 0.91; Figure 5C). TAAR1 might participate in different pain processes (for example, acute vs. chronic pain). We then assessed the effects of RO5263397 (3.2 and 5.6 mg/kg) on chronic inflammatory pain in the CFA-treated rats. It should be noted here that we used the von Frey filament test, which measures mechanical withdrawal threshold, to assess pain in the CFA-treated rats. Two-way repeated-measures ANOVA analysis with test as within-subjects factor and treatment as between-subjects factor revealed significant main effects of test (F_{4, 80} = 138.10, P < 0.05) and treatment (F_{3, 20} = 15.00, P < 0.05) and an interaction of test × treatment (F_12,80 = 11.05, P < 0.05). Post hoc analysis showed that the analgesic effect of morphine was prevented by naltrexone (naltrexone vs. vehicle: for 3.2 mg/kg morphine, t₁₀ = 3.71, P = 0.49; for 10 mg/kg morphine, t₁₀ = 14.20, P < 0.05) but not 3.2 mg/kg RO5263397 (3.2 mg/kg RO5263397 vs. vehicle; for 3.2 mg/kg morphine, t₁₀ < 0.00, P > 0.99; for 10 mg/kg morphine, t₁₀ = 1.77, P = 0.59) or 5.6 mg/kg RO5263397 (5.6 mg/kg RO5263397 vs. vehicle; for 3.2 mg/kg morphine, t₁₀ = 1.75, P = 0.73; for 10 mg/kg morphine, t₁₀ < 0.00, P > 0.99), indicating that RO5263397 did not affect the analgesic effect of morphine on chronic inflammatory pain (Figure 5E). RO5263397 alone did not affect inflammatory pain within 60 min after administration (main effect of group: F_{2, 15} = 0.49, P = 0.62; Figure 5F). Taken together, these results indicated that RO5263397 did not affect the antinociceptive effects of morphine in hot water tail immersion test or CFA-induced inflammatory test.

DISCUSSION

Consistent with our previous studies suggesting that activation of TAAR1 could reduce cue- and drug-induced reinstatement of cocaine-, methamphetamine-, nicotine-seeking behaviors in the rat self-administration model (Jing et al., 2014; Liu et al., 2018b; Thorn et al., 2014), the present study showed that the selective TAAR1 partial agonist RO5263397 significantly attenuated morphine-induced sensitization, morphine intake, motivation to use morphine, and morphine-seeking behavior induced by morphine-associated cue and morphine, suggesting that RO5263397 could be useful to prevent opioid relapse.

The inhibitory effect of RO5263397 might not be due to an off-target effect or non-specific behavioral inhibition. First, RO5263397 was highly selective to rat TAAR1 as it showed high affinity at TAAR1 expressed in HEK293 cells but not many other receptors and enzymes (Revel et al., 2013). Second, the present study showed that RO5263397 did not change locomotor activity or induce reinstatement of morphine-seeking behavior in rats self-administered morphine. Third, our previous studies showed that the same doses of RO5263397 used in the present study did not affect locomotion or cue-induced reinstatement of sucrose-seeking in rats (Jing et al., 2014). Furthermore, we also showed that RO5263397 attenuated morphine intake but did not influence the consumption of standard food pellets in the self-administration models. A previous study showed that RO5263397 dose-dependently increased responding maintained by food in a progressive ratio schedule of reinforcement (Pei et al., 2017). In contrast, the TAAR1 full agonist RO5256390 reduced palatable food reward but not the consumption of a standard diet (Ferragud et al., 2017). Collectively, the effects of TAAR1 agonists on food reward and consumption might depend on various factors, e.g., type of the particular TAAR1 agonist used (i.e., partial or full agonist), type of diet used in the experiment (i.e., standard diet or palatable food), and behavioral paradigm (Moore et al., 2018). Despite these findings, we could not completely rule out the non-specific inhibition of RO5263397 in rats. Future studies would use the TAAR1 knockout rats to clear this issue. We found that knockout of TAAR1 in mice produced a higher locomotor response induced by an acute administration of morphine (vehicle groups in Figure 1C, D), indicating that endogenous TAAR1 plays a role in the acute morphine-induced hyperactivity in mice. The levels of expression of morphine-induced behavioral sensitization were similar between wildtype and TAAR1 knockout mice, which might be due to a ceiling effect. Alternatively, it might be dependent on the dose of morphine administered in the induction phase of morphine-induced sensitization.

Evidence shows that drug reward memory plays a crucial role in drug addiction, and targeting the drug reward memory might be a promising strategy to disrupt drug addiction and prevent relapse selectively (Hyman et al., 2006; Steidl et al., 2017). Our previous study indicated that activation of TAAR1 decreased the retrieval of cocaine-induced CPP (Liu et al., 2016). However, the present study showed that RO5263397 did not affect morphine-induced CPP in both rats and mice. These results are consistent with a previous study, which showed that TAAR1 knockout mice developed a higher level of methamphetamine- but not morphine-induced CPP (Achat-Mendes et al., 2012). It is not paradoxical that activation of TAAR1 attenuated other morphine addiction-related behaviors but did not affect morphine-induced CPP. Although different morphine-associated behaviors share some common mechanisms, particular molecules may be involved distinctly in different addiction-related behaviors, e.g., morphine CPP vs. morphine self-administration. For example, genetic deletion of the δ opioid receptor disrupted morphine-induced CPP but not morphine self-administration (Le Merrer et al., 2011). The δ/μ opioid agonist MMP-2200 did not induce CPP but maintained self-administration in rats (Stevenson et al., 2015). Notably, our results showed that RO5263397 reduced cue- and drug-induced reinstatement of morphine-seeking in the morphine self-administration model. It is commonly accepted that drug-induced CPP is an associative memory model, whereas drug self-administration is an instrumental memory model. Therefore, RO5263397 might differentially affect morphine-related associative and instrumental memories. Given that we examined different memory stages in these two models, i.e., the expression of morphine-induced CPP but the reinstatement after extinction of morphine self-administration, the memory stage could be another factor that might influence the effects of RO2563397. Further studies are required to investigate whether the effects of RO5263397 on morphine-related memories would be dependent on memory type and stage.

The present study suggested that activation of TAAR1 might only inhibit the positive rather than negative reinforcing effects of morphine (Bozarth et al., 1983). It is unlikely that the dose of RO5263397 used in the mice morphine withdrawal experiments was too low. First, the same dose of RO5263397 attenuated morphine-induced behavioral sensitization in mice, indicating that this dose is biologically effective. We also found that even a larger dose of RO5263397 (10 mg/kg) that significantly reduced locomotion did not affect naltrexone-precipitated jumping behavior in mice (data not shown). Second, RO5263397 did not affect the expression of naltrexone-induced CPA when naltrexone was not present. However, the lack of involvement of TAAR1 in the negative reinforcement of opioids may not generalize to other drugs. For example, our preliminary data showed that TAAR1 might regulate nicotine withdrawal-induced anxiety-like behaviors and physical symptoms (data not shown).

A previous study showed that 3-Iodothyronamine (T1AM), an endogenous high-affinity ligand of TAAR1, reduced pain threshold to hot stimuli in mice (Laurino et al., 2017). However, T1AM is a multitarget ligand rather than a selective TAAR1 ligand. T1AM showed inverse agonist action at the human trace amine-associated receptor 5 and induced activation at the α−2A-adrenergic receptor (Dinter et al., 2015a; Dinter et al., 2015b). Furthermore, T1AM induced hypothermia in both wildtype and TAAR1-KO mice (Panas et al., 2010). Therefore, the effects of T1AM on pain response may not suggest a role of TAAR1 in pain processing. A previous study showed that RO5263397 (3.2 mg/kg, s.c.) reduced core body temperature in mice (Black et al., 2017), which may presumably influence the behavioral responses to a hot stimulus. However, in the present study, we did not observe any effect of RO5263397 on the tail withdrawal in the hot water immersion test. It is possible that the dose of RO5263397 (1 mg/kg, i.p.) used in the present study was not high enough to induce hypothermia that could influence responses to a hot stimulus. However, a higher dose of RO5263397 (e.g., 3.2 mg/kg, i.p.) reduced locomotor activity and thus would not be appropriate for the hot water tail immersion test. Although our results showed that RO5263397 did not affect pain response or the analgesic properties of morphine in both acute and chronic pain models, a more careful and systemic examination should be performed before concluding that TAAR1 does not regulate pain. Furthermore, we did not examine the effects of RO5263397 on pain under the condition of chronic morphine use, which could be different from the current findings.

A limitation of the present study is that we did not investigate the potential neural mechanisms of TAAR1 in regulating opioid addiction. However, the present study aimed to examine the role of TAAR1 in opioid-related behaviors systemically. Investigating the neural mechanisms in such a broad spectrum of behaviors is beyond the scope of this study. So far, to the best of our knowledge, there has been no study investigated the neural mechanism underlying TAAR1 in regulating the effects of opioids or opioid addiction. Nevertheless, previous studies have suggested some potential mechanisms underlying TAAR1 signaling transduction (Liu et al., 2018a). For example, TAAR1 may interact with pre-synaptic dopamine D2 autoreceptors to negatively regulate dopamine transmission, which has been suggested to commonly mediate the addictive-like behaviors (Leo et al., 2014). TAAR1 could also interact with post-synaptic D2 receptors to regulate downstream GSK3β signaling pathways (Espinoza et al., 2015; Harmeier et al., 2015). However, it should be noted that these cellular mechanisms in normal conditions may not reflect the function of TAAR1 in opioid addiction that engages particular mechanisms at behavioral and cellular levels (Badiani et al., 2011). Besides, it should be noted that RO5263397 is a TAAR1 partial agonist rather than a full agonist (Revel et al., 2013). Although our previous studies showed that many outcomes of RO5263397 and the full agonist RO5166107 were similar in psychostimulants-related behaviors, we cannot rule out the possibility that full and partial TAAR1 agonists may exert different effects on morphine-related behaviors and be mediated by other mechanisms (Liu et al., 2020b). Therefore, future studies are required to investigate the specific neural mechanisms of TAAR1 and its agonists in opioid addiction. Besides, we used male rats and mice in different behaviors in the study. Given that TAAR1 agonists showed species-dependent stereoselectivity, their behavioral effects in one species might not generalize to other species. Another limitation is that we only used male animals in the present study. As sex is an essential biological variable that modulates opioid analgesia and addiction (Lee et al., 2013), TAAR1 agonists may exert distinct actions on opioid-associated behaviors in males and females. Thus, further studies are required to evaluate the sex differences of TAAR1 in opioid-related behaviors.

CONCLUSION

In summary, the present study showed a critical role of TAAR1 in opioid addiction and suggested that the selective TAAR1 partial agonist RO5263397 may be promising to treat opioid addiction and prevent relapse. Since RO5263397 did not affect the antinociceptive effects of morphine, we suggest that a combination of RO5263397 and opioid prescription might be a strategy to reduce the risk of opioid addiction in patients without affecting the analgesic effects of opioids to manage pain.

Bullet point summary.

What is already known:

• Trace amine-associated receptor 1 (TAAR1) negatively regulates dopamine transmission.

• Knockout of TAAR1 did not affect morphine-induced conditioned place preference.

What this study adds:

• The TAAR1 partial agonist RO5263397 reduced morphine-induced behavioral sensitization, morphine self-administration, motivation to use morphine, and the cue- and drug-induced reinstatement of morphine-seeking.

• RO5263397 did not affect morphine-induced conditioned place preference, morphine withdrawal, or the analgesic properties of morphine.

Clinical significance:

• These results support the potential utility of selective TAAR1 agonists to treat opioid addiction.

Abbreviations:

ANOVA

analysis of variance

cAMP

cyclic adenosine monophosphate

CFA

complete Freund’s adjuvant

CPA

conditioned place aversion

CPP

conditioned place preference

FR

fixed ratio

NAc

nucleus accumbens

RO5263397

(4S)-4-(3-Fluoro-2-methylphenyl)-4,5-dihydro-1,3-oxazol-2-amine

TAAR1

trace amine-associated receptor 1

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request. Some data may not be made available because of privacy or ethical restrictions.