Materials

All reagents used for these studies were of the highest purity available and were obtained from Sigma-Aldrich (St Louis, MO) unless otherwise stated. Heroin hydrochloride was provided by the Drug Supply Program of the National Institute on Drug Abuse, Rockville, MD) unless otherwise indicated. Oxycodone was obtained from Mallinkrodt Pharmaceuticals (Dublin, Ireland). ITI-333, or (6bR,10aS)-8-[3-(4-fluorophenoxy)propyl]-6b,7,8,9,10,10a-hexahydro-1H-pyrido[3’,4’:4,5]pyrrolo[1,2,3-de]quinoxaline-2(3H)-one (Fig. 1), a novel molecule discovered and synthesized in the Medicinal Chemistry Department at ITCI (New York, NY), was administered in all studies as a free base (MW = 381.44 g/mol). Unless otherwise stated, ITI-333 was prepared for in vivo subcutaneous (s.c.) dosing in a vehicle composed of 45% Trappsol (in water) and administered in volumes of 6.67–10 mL/kg.

Fig. 1

Structure of ITI-333. ITI-333, or (6bR,10aS)-8-[3-(4-fluorophenoxy)propyl]-6b,7,8,9,10,10a-hexahydro-1H-pyrido[3’,4’:4,5]pyrrolo[1,2,3-de]quinoxaline-2(3H)-one (MW = 381.4 g/mol)

Animals

All animals were cared for in accordance with the the Guide for the Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council, and all procedures were performed with the approval of the Institutional Animal Care and Use Committees at the respective institutions and contract research organizations. All rodents were experimentally naïve at the start of experiments and were tested only once, unless otherwise stated. All monkeys had received opioids in previous studies but were drug-free for at least two weeks prior to this study. Animal numbers and treatment group sizes for all experiments described here were based on prior studies in the individual laboratories or contract research organizations conducting the work. The minimal number of animals sufficient to obtain statistically reliable results was used.

Receptor binding assays

The binding of ITI-333 to principal receptors of interest was measured in cell-based assays (Eurofins, Celle l’Evescault, France). A broad selectivity screen of 44 neurotransmitter receptors, enzymes, and channels was first used to test for binding of ITI-333 (0.1 µM) to a diverse panel of potential off-target proteins. Binding affinity was measured and expressed as a % inhibition of control-specific binding; inhibition of greater than 50% was considered to represent significant effects of ITI-333 on a given target. Receptor targets demonstrating > 50% binding with ITI-333 in the broad target screen—including 5-HT2A, MOP, dopamine D1 and D2, and adrenergic α1A receptors—were investigated in further detail in individual assays for binding affinity and for functional activity using recombinant human receptors expressed in Chinese Hamster Ovary (CHO) cells and standard methods. The affinity of ITI-333 was determined at 5-HT2A, MOP, D1, D2, and α1A receptors by measuring inhibition of binding of 0.1 nM [125I]DOI, 0.35 nM [3H]DAMGO, 0.3 nM [3H]SCH23390, 0.3 nM [3H]methylspiperone, and 0.1 nM [3H]prazosin, respectively.

In vitro functional assays 5-HT2A receptors

ITI-333 functional activity was evaluated in a whole cell-based assay system in which 5-HT2A receptor-dependent calcium signaling was measured using a recombinant calcium-dependent bioluminescent protein. ITI-333 was studied in the agonist mode and for its ability to antagonize the activity of the 5-HT2A receptor full agonist α-methylserotonin.

MOP receptors

ITI-333 functional activity was examined in CHO cells expressing human recombinant MOP receptors; agonist activity (using 1 µM DAMGO as a control) and antagonist activity (using reversal of DAMGO-inhibited cAMP production) were assessed at drug concentrations between 0.056 nM–10 µM. Buprenorphine and naloxone (0.0056 nM–1 μM) were used as comparators for agonist and antagonist activity, respectively. The functional activity of ITI-333 at MOP receptors was further investigated in a whole cell-based assay using MOP receptor-dependent suppression of adenylyl cyclase activity to determine intrinsic efficacy compared with DAMGO. The effect of ITI-333 on β-arrestin pathway signaling at MOP receptors was also investigated using the PathHunter® β-Arrestin assay (DiscoverX, Fremont, CA) using met-enkephalin as a control.

Dopamine D1 receptors

ITI-333 functional activity at dopamine D1 receptors was determined at a single concentration of 10 μM. Cellular agonist effects were calculated as the percent of control response to 10 μM dopamine; antagonist effects were calculated as the percent inhibition of 300 nM dopamine response in CHO cells expressing human recombinant D1 receptor.

Adrenergic α1A receptors

ITI-333 functional activity on adrenergic α1A receptor-dependent calcium signaling was assessed using a whole cell-based AequoZen assay system. Adrenergic α1A receptor activation was detected via measurement of light emission after addition of increasing concentrations of ITI-333 or the positive control agonist phenylephrine (≤ 100 µM) in CHO-K1 cells with stable co-expression of aequorin and adrenergic α1A receptors. In antagonist mode, the reversal of agonist-induced (50 nM phenylephrine) calcium response by increasing concentrations of ITI-333 (≤ 10 µM) or a positive control (tamsulosin [≤ 1 µM]) was examined.

In vivo functional assaysBlockade of 5-HT2 agonist-induced head twitch

Functional activity of ITI-333 at 5-HT2A receptors was assessed in vivo using a head twitch paradigm in adult (8–10 weeks) male C57Bl/6 mice (Jackson Labs, Wilmington, MA). Mice received a s.c. injection of ITI-333 (n = 15; 0.01–10 mg/kg) or vehicle (n = 3; 45% Trappsol in water) 30 min prior to an intraperitoneal (i.p.) injection of the 5-HT2 agonist 2,5-Dimethoxy-4-iodoamphetamine (DOI; 2.5 mg/kg), after which they were immediately placed in a novel cage for behavioral observation. During the 5-min observation period, head twitches (rapid side-to-side rotational head movements) were counted manually by a blinded observer. The mean number of head twitches for each treatment group was calculated. Curve fitting was performed with the XL-Fit add-on to Microsoft Excel to estimate an ID50 for inhibition of DOI-induced headtwitch by ITI-333, using a 4-parameter logistic fit.

Morphine-induced hyperactivity

Adult (8–10 weeks) male C57Bl/6 mice (Jackson Labs, Wilmington, MA) were habituated to locomotor chambers (Ohio Instruments, Hiram, OH) without treatment, and their spontaneous horizontal locomotor activity recorded for 30 min. Groups of mice (n = 5 each) then received morphine sulfate (32 mg/kg, s.c.) followed immediately by ITI-333 (0.01–3 mg/kg, s.c.) or vehicle. After 30 min, mice were returned to the locomotor chambers and the cumulative distance traveled over 120 min was measured using video-tracking software (AccuScan, AUT Solutions, Fulshear, TX). The effect of ITI-333 was calculated as the cumulative distance traveled after administration of ITI-333 as a percentage of the distance traveled by vehicle-treated mice.

Spontaneous locomotor activity

Adult (8–12 weeks) male Swiss-Webster mice (Hsd:ND4) were injected (subcutaneous) with either vehicle (45% hydroxypropyl-beta-cyclodextrin) or ITI-333 (0.03, 1, or 3 mg/kg) 30 min before locomotor activity assessment. All mice (n = 8/group) were non-habituated and received an intraperitoneal saline injection immediately before placement into activity test chambers (Digiscan; 40.5 × 40.5 × 30.5 cm). Horizontal activity counts, assessed by interruptions of 16 infrared beams, were measured in 10-min bins. A one-way ANOVA and planned comparisons to the vehicle control group were conducted on horizontal activity counts during the 30-min period following placement in the test chamber. Mean horizontal activity counts were fit to a linear function of dose of the descending portion of the dose–effect curve to determine the dose producing half-maximal depressant activity (ID50), where maximal depression equaled 0 counts per 30 min.

Prevention of naloxone-precipitated withdrawal from oxycodone

Adult (8–10 weeks) male C57Bl/6 mice (Jackson Labs, Bar Harbor, ME) were administered saline or oxycodone twice daily for 8 d at increasing daily doses (9, 17.8, 23.7, and 33 mg/kg, s.c. on days 1–2, 3–4, 5–6, and 7–8, respectively). On day 9, mice were injected with ITI-333 (0.3, 1, or 3 mg/kg, s.c.; n = 32 each) or vehicle (n = 16), followed 30 min later with an injection of naloxone (3 mg/kg, s.c. in saline) to precipitate abrupt opioid withdrawal. Immediately following naloxone injections, mice were observed continuously for 30 min; somatic signs of withdrawal, including jumps, wet-dog shakes, paw tremors, backing, ptosis, and diarrhea, were recorded manually by a blind observer. All behaviors were recorded as new incidences when separated by at least 1 s or interrupted by any other normal behavior. Data were analyzed with analysis of variance (ANOVA) tests followed by Tukey tests for multiple comparisons.

Precipitation of withdrawal from oxycodone by ITI-333

Adult (8–10 weeks) male C57Bl/6 mice (Jackson Labs, Bar Harbor, ME) were administered oxycodone as described above. On the morning of the 9th day, mice were administered oxycodone (33 mg/kg, s.c.) followed 2 h later with an injection of ITI-333 (3, 10, or 17.8 mg/kg, s.c.; n = 8 each) or vehicle (n = 8). A separate group of mice was chronically administered saline instead of oxycodone and was challenged with ITI-333 (17.8 mg/kg, s.c.; n = 8) or vehicle (n = 8) on day 9 to evaluate effects of ITI-333 alone. Thirty minutes following vehicle or ITI-333 injections on day 9, mice were individually placed in Plexiglas cages and observed for somatic signs of withdrawal as described above. Data were analyzed with ANOVA followed by Dunnett’s tests for multiple comparisons.

Cue-induced reinstatement of heroin self-administration

Adult (280–300 g upon arrival) male Long-Evans hooded rats (Envigo, Indianapolis, IN) were trained to self-administer heroin. Training sessions were conducted for 5 d/week for 2 h/d using test chambers equipped with two retractable levers, white cue lights positioned above each lever, a 5W house light, and a tone generator. Rats were outfitted with an indwelling catheter in the right external jugular vein using a protocol described previously (Shelton et al. 2013). Rats were trained using a fixed ratio 1 (ie, 1 infusion per lever press; FR1) reinforcement schedule on the active (right-side) lever to deliver a 0.01 mg/kg heroin infusion (0.14 ml/6 s) through the indwelling catheter. For the duration of the infusion, the tone sounded and the stimulus lights above both levers flashed at 3 Hz. Active lever presses during the infusion and inactive lever presses were without consequences. Active and inactive lever presses and activation of lights, pumps, and tones were recorded. Self-administration training continued until three criteria were met: 1) at least 12 self-administration sessions occurred; 2) at least 15 heroin infusions occurred during each of the last four sessions; and 3) at least 125 lifetime heroin infusions had been obtained.

After successful self-administration training, 12 extinction sessions (2 h/d) were conducted in which the house light was illuminated and the levers were extended but infusions were not administered and other scheduled stimuli (tone or light) did not occur. Conditions during subsequent reinstatement testing were identical to those during self-administration training, except that ITI-333 (1, 3, or 10 mg/kg, n = 12 each) or vehicle (n = 12) was administered 30 min prior to testing and heroin self-administration did not occur. Additionally, cues previously associated with heroin infusion were presented non-contingently for 6 s at the start of the reinstatement testing session (i.e., the tone sounded, the stimulus lights above both levers flashed at 3 Hz for 6 s, and the house light was off).

Data were analyzed using individual ANOVAs. If results were found significant, comparisons between groups were conducted using Tukey’s Multiple Comparison tests. Further, a paired, one-tailed t-test was conducted comparing active-lever presses during the last extinction session with those during the reinstatement test session of the vehicle group to determine if the heroin cue conditions used were capable of reinstating responding. In addition, numbers of inactive level presses (i.e., on the left side lever) occurring during the last test session were compared between groups using ANOVA. All statistical tests were conducted using Prism 7 for Macintosh (GraphPad Software, Inc., San Diego, CA), and all comparisons were considered statistically significant if p < 0.05.

Reinforcement of self-administration in heroin-maintained rats

Adult (200–300 g at the start of the study) male Sprague–Dawley rats (Charles River, Margate, Kent, UK) were trained under mild food restriction to lever press for food rewards on a FR3 schedule of reinforcement. When competent in this task, pharmacologically active doses of ITI-333 were defined in these animals by observing the effects of intravenous (i.v.) bolus injections (0.001, 0.003 and 0.01 mg/kg, i.v.) on general behavior and FR3 operant responding for food rewards (n = 3–10/group). Based on this experiment, 0.0003, 0.001, 0.003 and 0.01 mg/kg/inj (i.v.) doses were selected for the self-administration experiment (n = 8/group). Rats were surgically implanted with an indwelling jugular catheter for the i.v. self-administration experiments and trained to self-administer a low dose of heroin (0.015 mg/kg/inj in 0.9% saline) on a FR5 reinforcement schedule. Saline (0.5 ml/kg/inj, i.v.; n = 21) was the non-reinforcing control substance. The reinforcing effects of ITI-333 were investigated on a FR5 schedule of reinforcement.

The mean number of injections per session during the last 3 sessions for rats responding under an FR5 schedule of i.v. self-administration were calculated and then back-transformed and adjusted for between-animal differences. SEMs were calculated from the residuals of the statistical model. ITI-333 was compared to the first saline extinction session by Williams’ test and to heroin acquisition session by Dunnett’s test. The heroin acquisition session was compared to the first saline extinction session using multiple t-tests.

Reinforcement of self-administration in heroin-maintained rhesus monkeys

The potential reinforcing effects of ITI-333 were studied in 3 male and 2 female adult rhesus monkeys (Macaca mulatta), ranging in age from 7–20 years, using an intravenous self-administration procedure. The monkeys were housed in a room maintained on a 14/10-h light/dark cycle at 21 ± 1 °C and relative humidity of 50 ± 10%. They received primate chow (High Protein Monkey Diet; Harlan Teklad, Madison, Wisconsin), peanuts, and fresh fruit daily in the home cage in amounts adequate to maintain age-appropriate body weights.

Surgery

Monkeys were implanted with a chronic s.c. access port (MIDA-PU-C50; Instech Laboratories, Plymouth Meeting, PA) and an indwelling i.v. catheter (e.g., jugular or femoral vein) under anesthesia with ketamine (10 mg/kg, intramuscular) followed by isoflurane (1.5–3%, inhalation) maintenance. Monkeys were allowed at least 2 d for recovery post-surgery, during which time they were examined daily and received an injectable antibiotic (penicillin B&G, 40,000 IU/kg) and an analgesic (meloxicam, 0.2 mg/kg the first day; 0.1 mg/kg/d for up to 4 additional days).

Apparatus

For all studies, monkeys were seated in chairs (Macaque Restrainer model 1R-1R10; Primate Products, Inc., Immokalee, FL) that provided restraint at the neck and arms. During experimental sessions, chairs were located in ventilated, sound-attenuating chambers. Mounted on one wall of each chamber, within easy access to the seated monkeys, was a custom-made operant panel containing two response levers and associated stimulus lights. Only one of the levers (left or right) was activated for an individual monkey; the particular lever that was activated was chosen based on the behavioral history of the monkeys used in this study. Pellet dispensers were located on the outside of each chamber and, for some monkeys, completion of the response requirement resulted in the delivery of 300 mg raspberry-flavored sucrose pellets (Bio-Serv, Flemington, NJ) to a trough mounted under the lever panel. Infusion pumps (model PHM-100; MED Associates, Inc., Fairfax, VT) were also located outside the chamber and were connected to the implanted i.v. catheter with sterile tubing and a Huber point needle. The response panel and infusion pump were connected to and controlled by an interface and computerized system (MED Associates, Inc.). To maintain patency, catheters and ports were flushed and locked after each session with 2.5 ml of heparinized saline (100 U/ml; Hospira Inc., Lake Forest, IL).

Dose-finding study

Monkeys were previously trained to press a lever, in the presence of a distinctive visual stimulus, on a FR10 schedule to receive a food pellet. Daily sessions comprised eight 15-min cycles (2 h). A cycle began with a 10-min timeout, during which the chamber was dark and lever presses had no consequence. After the 10-min timeout was a response period, signaled by the illumination of a green light, during which monkeys could respond under the FR10 schedule for food. The response period ended and the chamber darkened after the delivery of 10 food pellets or 5 min, whichever occurred first.

ITI-333 (0.0001–0.32 mg/kg) was evaluated in two monkeys for its ability to decrease the rate of lever pressing for food. ITI-333 was administered through the implanted i.v. catheter. Tests were conducted no more often than once every four days and only so long as responding was stable, as demonstrated by the last three sessions before the test session in which drug was not administered. Response rate was averaged across cycles to obtain a mean rate for each session and then averaged across the three sessions; responding was considered stable when the rate for each individual session was > 75% of the mean rate for the three sessions.

Heroin (baseline) self-administration and vehicle substitution (extinction)

Monkeys (n = 4) were trained to respond under an FR 30 schedule for heroin (0.0032 mg/kg/infusion, i.v.) as described previously (Maguire et al. 2019). Thereafter, vehicle replaced heroin (i.e., extinction) for a minimum of 4 sessions and until monkeys received fewer than 8 infusions in each of 3 consecutive sessions in which the average response rate was less than 20% of the average response rate for sessions when heroin was available (for each individual monkey). In separate sessions, ITI-333 (0.01, 0.032, and 0.1 mg/kg/inf, i.v.) was substituted for vehicle. Each dose was studied for a minimum of 5 and a maximum of 10 sessions. A “priming” (noncontingent) infusion of ITI-333 was administered immediately before each session. Following assessment of each dose of ITI-333, vehicle was available for self-administration (i.e., washout) for a minimum of 4 sessions.

Heroin self-administration retest

After completion of sessions with the three doses of ITI-333, monkeys were tested again with heroin (0.0032 mg/kg/inf, i.v.) for a minimum of 5 sessions according to the criteria described above.

Potential of ITI-333 to Induce Pharmacological Tolerance/Physical Dependence on Withdrawal

Adult (200–250 g at the beginning of the study) male Sprague–Dawley rats (Charles River, Margate, Kent, UK) were tested for signs of physical dependence upon withdrawal of ITI-333, morphine, or vehicle (10% Trappsol + 1% Tween 80 in water). During the baseline phase (Day -6 to Day 0), all animals received vehicle by s.c. and p.o. routes at 08:00 h to familiarize them with dosing and handling. On day 0, rats were randomized to treatment with vehicle, ITI-333, or morphine (positive control). During the drug dosing phase (Day 1 to Day 28), ITI-333 was administered once daily and morphine was administered twice daily, as follows:

Group

Treatment 1 (08:00 h)

Treatment 2 (15:00 h)

Vehicle/Vehicle

Vehicle (2 ml/kg s.c., 5 ml/kg p.o.)

Vehicle (5 ml/kg p.o.)

ITI-333/Vehicle

ITI-333 (0.3 mg/kg s.c.); Vehicle (5 ml/kg p.o.)

Vehicle (2 ml/kg s.c.)

ITI-333/Vehicle

ITI-333 (3 mg/kg s.c.); Vehicle (5 ml/kg p.o.)

Vehicle (2 ml/kg s.c.)

Vehicle/Morphine

Vehicle (5 ml/kg s.c.); Morphine (30 mg/kg p.o.)

Morphine (30 mg/kg p.o.)

Rats received their final treatment doses on Day 28; during the subsequent 7-day drug withdrawal phase (Day 29 to Day 35), no treatments were administered. Throughout each phase of the study, all animals were assessed once daily for behavioral, physical, and physiological signs including body weight, food intake, water intake, and temperature. Statistical tests for comparisons against the vehicle/vehicle group were performed using Williams’ test (ITI-333-treated groups) or multiple t-tests (morphine-treated group).

Pulmonary assessment

Adult (9–10 weeks, 270–332 g at the start of the study) male Sprague–Dawley rats (n = 6/group; Charles River) were habituated for 2 d in a plethysmograph chamber prior to the experiment. On the testing day, each rat was placed in the chamber and baseline respiratory parameters were obtained for 5 min. Rats were then removed from the chamber and administered ITI-333 (0.3, 1, 3 mg/kg, s.c.) or vehicle. Rats were returned to the plethysmograph chamber immediately following dosing, and respiratory parameters—including respiratory rate, tidal volume, and minute volume—were measured during 5-min intervals at 30 min (± 3 min), 1, 2, and 4 h (± 5 min). Values for the ITI-333-treated group were compared to vehicle controls and baseline using a two-way repeated measures ANOVA followed by a Bonferroni Test for Multiple Comparisons (SigmaStat v.2.03). P values < 0.05 were considered statistically significant.

Measurement of GI propulsion

GI peristalsis was measured using a motility indicator (transit of a 10% suspension of activated charcoal in 0.25% methylcellulose) administered orally 15 min posttreatment with vehicle, ITI-333, or morphine (positive control). Adult (8 weeks of age, 218–246 g) male Sprague Dawley rats (Charles River) were given a single s.c. dose of vehicle, ITI-333 (0.3, 1, or 3 mg/kg in vehicle), or morphine (5 mg/kg in sterile water; n = 8/group). Rats were euthanized 30 min after charcoal administration and intestines were removed; the length of the intestine (pyloric sphincter to caecum) and the distance traveled by the charcoal as a fraction of that length were measured for each rat. In a second experiment, the ability of ITI-333 to block morphine-induced inhibition of GI motility in rats was studied by administering a single dose of ITI-333 (3 mg/kg, s.c.) either 30 min after or 60 min prior to administration of morphine (5 mg/kg). Values for charcoal motility were expressed as a percent effect (distance traveled divided by the total length of the intestines). Mean values were calculated for each group and percent effect was calculated using the following formula:

$$\%effect=\frac \times 100$$

Group comparisons were performed using ANOVA with Tukey HSD test for multiple comparisons. P values < 0.05 were considered statistically significant.