EPD1504: a novel μ-opioid receptor partial agonist attenuates obsessive-compulsive disorder (OCD)-like behaviors

Abstract

Low doses of μ-opioid receptor (MOR) agonists rapidly ameliorate symptoms in treatment-resistant obsessive-compulsive disorder (OCD) patients (10-50% of OCD patients). However, the utility of MOR agonists is limited by their safety liabilities. We developed a novel MOR partial agonist (EPD1540) that has an improved respiratory safety profile when compared to buprenorphine. Buprenorphine is a MOR partial agonist primarily used in the treatment of opiate-use disorder, which in investigator-led trials, has been shown to rapidly ameliorate symptoms in treatment-resistant OCD patients. In this study, we show that doses of EPD1504 and buprenorphine that occupy small fractions of MORs in the CNS (approximately 20%) are as effective as fluoxetine at ameliorating OCD-like behaviors in two different rat models (an operant probabilistic reversal task and marble burying). Importantly, effective doses of EPD1504 did not impair either locomotor activity, or respiration under normoxic or hypercapnic conditions. Additionally, EPD1504 had effects comparable to buprenorphine in the conditioned place preference assay. These results indicate that EPD1504 may provide a safer alternative to buprenorphine for the treatment of OCD patients.

Keywords: buprenorphine (BN); marble burying (MB); obsessive–compulsive disorder (OCD); opioid; partial agonist; probabilistic reversal task; treatment resistant OCD.

Figure 1

EPD1504 is a brain penetrant MOR partial agonist with buprenorphine-like antinociceptive activity in the hotplate test. (A) Chemical structures of EPD1504 and buprenorphine. (B) Representative dose–response curves for cAMP activation in CHO cells expressing human MORs, with EPD1504, the full agonist: DAMGO, antagonist: naltrexone, and comparable partial agonist: buprenorphine (n = 3 replicates per experiment, mean ± SD). (C) Coplot of dose–response curves for antinociceptive activity in the hotplate test measured as the mean of % maximum possible effect (%MPE) between 30 and 120 min after injection of the test compound on the y-axis vs. %MOR occupancy in the thalamus at 60 min after injection of the test compound on the x-axis (rats per dose: %MOR occupancy, n = 3–4, %MPE antinociceptive activity, n = 6–8). (D) Summary table for dose and percent occupancy of MOR for the data shown in (C). The data used to calculate CNS MOR percent occupancy for buprenorphine have been reported in our previous study (35). (E) Time course and summary data is shown in (F) of antinociceptive activity in the hotplate test for doses of EPD1504 10 mg/kg and buprenorphine that occupy approximately 80% of MOR in the thalamus; note that the antinociceptive activity of both agonists was antagonized by naloxone (1 mg/kg) (which crosses the blood–brain barrier), but not by naloxone methiodide 10 mg/kg which does not cross the blood–brain barrier (n = 6–8 rats per dose, ANOVA, **p < 0.01, ns = not significant).

Figure 2

EPD1504 and buprenorphine reduce error responses in a probabilistic reversal task. (A) Schematic of the probabilistic reversal task. A food restricted rat can press either lever in the operant chamber to obtain a food reward. The probability of a reward after pressing the “correct lever” (green with check mark) is 80%, whereas the probability of a reward after pressing the “error lever” is 20%. After eight correct-lever presses the contingencies reverse, i.e., the correct lever becomes the error lever and vice versa. (B) Histogram of 1,410 reversals during baseline sessions (n = 62 rats). Note that rats had learned the contingencies of the task as more correct responses were made during a greater percentage reversal (approximately 60% of reversals). (C) Baseline rate of responding for 3 days. (D) Rate of responding during subsequent test trials. Note that doses that reduced the response rate were excluded from further analysis. (E) Error responses per reversal normalized to the interquartile range of responses during baseline [data in (B)]; (ANOVA, *p < 0.05, **p < 0.01, ***p <0.001, ns = not significant; n = 3–4 rats per group, n = 2 daily sessions).

Figure 3

EPD1504 and buprenorphine reduce marble burying. (A) Timeline of semi-automated method to quantify marble burying. (B) Representative photographs and processed images of marbles on the bedding (n = 15 marbles, before the rat was placed in the chamber), and after the rat had been in the chamber for 40 min. (C) Histogram showing cutoff used to determine whether a marble is buried or visible for 1,155 marbles before the rat was placed in the chamber (open bars), and after the rat was placed in the chamber (filled bars); note that no marbles were classified as buried before the rat was placed in the chamber. (D) Summary data for vehicle treated rats (n = 15). (E) Mean marbles buried for data in (D), and dose–response curves for marbles buried by rats injected with either vehicle (n = 18), fluoxetine (n = 10 per dose), buprenorphine (n = 10 per dose), EPD1504 (n = 11 per dose), or naloxone (n = 10 per dose). Doses are listed as mg/kg below the x-axis. (F) Doses of fluoxetine (3 mg/kg), EPD1504 (1 mg/kg), and buprenorphine (0.01 mg/kg) that reduced the number of marbles buried (did not affect distance traveled in the open-field test, and did not affect rate of lever pressing in the probabilistic reversal task as shown in Figure 2). (G) Timeline and data for rats preinjected with naloxone (1 mg/kg) followed by either EPD1504 (1 mg/kg) or buprenorphine (0.01 mg/kg) [n = 5 per group, *p<0.05, **p<0.01 ANOVA, not significant vs. vehicle treated rats shown in (E)].

Figure 4

EPD1504 has limited effects on respiration. (A–C) Timeline, and summary data (mean minute volume between 30 and 60 min after injection) for effect of test compounds on minute volume under normoxic conditions (normal room air). The doses are listed in mg/kg below the x-axis; note that in contrast to EDP1504 and buprenorphine, fentanyl suppressed respiration compared to vehicle. (D, E) Timeline and data with inset bars at 30 and 90 min after injection that correspond to observed respiratory depression and recovery, respectively after coinjection of diazepam with test compounds. (F) Summary data for effect of coinjection of diazepam with test compounds on respiration at the two time points (30 and 90 min) illustrated in (E). (G) Volcano plot of differences between mean minute volume observed at 30 and 90 min (recovery – depression); note that although coinjection of diazepam with all opioids tested exacerbated respiratory depression in normoxic condition, that only EPD1504 exhibited significant recovery to baseline in the at the 90-min time point. (H,I) Timeline and summary data for effect of test compounds on minute volume in 10% CO₂ (hypercapnic conditions); note that unlike both buprenorphine and fentanyl, EPD1504 did not suppress respiration under hypercapnic conditions (All groups n = 5–6, *p < 0.05, ***p <0.001, and +++p <0.001, **p < 0.01 as indicated; ANOVA post-hoc test).

Figure 5

EPD1504 has a limited dependence liability. Timelines and summary data: in (A,B) for conditioned place preference observed in opioid–naïve rats in (C,D) for somatic signs of withdrawal precipitated by naltrexone following subchronic exposure to EPD1504 and buprenorphine in subcutaneous mini pumps. Note that as detailed in the main text plasma levels observed for both compounds achieve approximately 70% occupancy of CNS MORs. In (E,F) for conditioned (E,F) place aversion observed after injections of naltrexone (Nltrx) (All groups n = 6–11, ns = not significant, *p < 0.05, **p < 0.01 as indicated; ANOVA post-hoc test).