Effects of the glucocorticoid receptor antagonist PT150 on stress-induced fentanyl seeking in male and female rats

Abstract

Rationale:

Opioid use disorder (OUD) is highly comorbid with stress-related disorders, and stress can serve as a trigger for reinstatement of drug seeking. Glucocorticoid receptor (GR) antagonists such as mifepristone (RU-486) may be effective against stress-induced drug seeking. In the current study, PT150 (formerly ORG-34517), a more selective GR antagonist, was tested using two models of stress-induced drug seeking, namely footshock and yohimbine.

Methods:

Adult male and female Sprague-Dawley rats were trained to self-administer fentanyl (2.5 μg/kg/infusion, i.v.) in a model of escalation. Rats then received 7 days of abstinence, followed by extinction; PT150 (0, 50 or 100 mg/kg in Nutella^®; p.o.) treatment started on the first day of extinction training and continued daily until the end of the study. Following 14 days of extinction, rats were tested for reinstatement following footshock and yohimbine (0, 1 or 2 mg/kg; i.p.), tested in counterbalanced order; PT150 or placebo treatment occurred prior to each extinction and reinstatement session.

Results:

Prior to initiation of PT150 treatment, females self-administered greater levels of fentanyl during 1-hr sessions compared to males; however, when switched to 6-hr sessions, males and females self-administered similar levels of fentanyl and showed a similar escalation of intake over time. PT150 had no effect on extinction of self-administration. While both footshock and yohimbine reinstated fentanyl seeking, only footshock-induced reinstatement was decreased by PT150 (50 and 100 mg/kg). The effect of PT150 on footshock-induced reinstatement was driven primarily by males.

Conclusion:

The glucocorticoid antagonist PT150 reduces shock-induced fentanyl seeking, suggesting it may be effective against stress-induced relapse, although the sex difference in response may need further exploration.

Introduction

While self-reported opioid use has remained relatively steady over the last decade, opioid overdose is rising dramatically due to the availability of high-potency synthetic opioids such as fentanyl (Frank and Pollack 2017), with fentanyl accounting for two-thirds of deaths in the United States in 2018 (CDC 2020). Among those suffering post-traumatic stress disorder (PTSD), such as military combat veterans and trauma victims, there may be higher risk for opioid abuse (Dabbs et al. 2014; Danovitch 2016), and thus it is important to examine pharmacotherapies that are effective at treating opioid use disorder (OUD) in the context of stress. The hypothalamic-pituitary-adrenal (HPA) axis, a key neurohormonal stress system, is altered during opioid withdrawal (Houshyar et al. 2001) and stress hormones may mediate the dysphoria caused by opioid withdrawal (Ingallinesi et al. 2012). For example, the glucocorticoid receptor (GR) antagonist mifepristone (RU-486) ameliorates opioid withdrawal signs in rats following a stress challenge (McNally et al. 2005). However, the therapeutic use of mifepristone to treat OUD is limited due to off-target actions, most notably progesterone receptors (Sun et al. 2014).

An alternative potential therapeutic for treating OUD is PT150 (formerly known as ORG-34517), an orally-available GR antagonist with reduced affinity for progesterone receptors relative to mifepristone (Bachmann et al. 2003; Peeters et al. 2004). PT150 is a competitive GR antagonist, preventing translocation of GRs to the nucleus (Peeters et al. 2008), although recent work shows additional activity at androgen receptors (Tatman et al. 2020). Stress can serve as a trigger for the initiation of drug seeking (Mantsch et al. 2016) and blocking GRs in basolateral amygdala with mifepristone has been shown to reduce stress-induced opioid seeking using place conditioning (Karimi et al. 2014). GR antagonism can also block the effects of morphine withdrawal on signaling pathways involved in drug abuse (Navarro-Zaragoza et al. 2017). However, there is currently no information about the effects of PT150 on OUD, withdrawal, and relapse. Since relapse can be triggered by stressful experiences (Lu et al. 2003), PT150 may lessen the impact of these events and reduce vulnerability to stress-induced relapse.

Here we used a rat model of fentanyl self-administration in which escalation of intake was achieved by using long access (6 hr) sessions to model the loss of control over drug use in humans (Wade et al. 2015). We examined the effects of oral PT150 on reinstatement of fentanyl seeking, elicited by either a physical stressor (footshock) or a pharmacological stressor (yohimbine). These two stressors have each been shown to increase drug seeking (Doncheck et al. 2020), but through different mechanisms (Mantsch et al. 2016). Footshock can lead to activation of GRs because it causes increased corticosterone release (Doncheck et al. 2020), while yohimbine activates the sympathetic arousal system through its action as an α2 noradrenergic autoreceptor antagonist. We hypothesized that PT150 would reduce footshock-induced fentanyl seeking through its actions at GRs.

Methods

Animals

A total of 28 male and 32 female adult Sprague-Dawley rats (Envigo-Harlan, Indianapolis, IN, USA), approximately 50 days of age at arrival, were used in this experiment. Rats were housed individually in temperature- and humidity-controlled rooms that were maintained on a light-dark cycle (lights on at 07:00). After arrival, rats acclimated to the colony room and were handled daily for a minimum of one week before surgery. Food and water were available ad libitum until the abstinence phase, when food was restricted to 15 g per day to ensure prompt and complete ingestion of the palatable drug mixture (see below). All experimental procedures were approved by the Institutional Animal Care and Use Committee at the University of Kentucky and the Animal Care and Use Review Office of US Army Medical Research and Development Command, and they conformed to the National Institutes of Health Guide for the Care and Use of Laboratory Animals (8^th Edition). A total of 22 males and 22 females completed all phases of the study.

Surgical procedures

Rats underwent jugular catheter implantation surgery. Rats were treated with carprofen (5 mg/kg, s.c.) the day before surgery and for 2 days after surgery. Male rats were anesthetized with a combination of ketamine (Butler Schein, Dublin, OH, USA), xylazine (Akorn, Inc., Decatur, IL, USA) and acepromazine (Boehringer Ingelheim, St. Joseph, MO, USA), which were mixed together to yield a single cocktail (75/7.5/0.75 mg/kg; 0.15ml/100g body weight; i.p.). Female rats received ketamine and xylazine, but they did not receive acepromazine based on veterinary consultation. During surgery, a silastic catheter was inserted into the right jugular vein and threaded under the skin to an incision on the scalp. A cannula was connected to the end of the catheter and was secured to the skull with dental acrylic and jeweler’s screws. Following surgery, rats recovered for 7 days before the start of self-administration training.

Apparatus

Training was conducted in operant conditioning chambers (28 × 24 × 25 cm; ENV-001; MED Associates, St. Albans, VT, USA) housed in sound-attenuation chambers containing a fan to dampen extraneous noise (ENV-018M; MED Associates). Each chamber was equipped with two retractable levers with a 28-V white cue light located 6 cm above each lever, a 28-V house light centered on the opposite wall, and a syringe pump for drug delivery (PHM-100; MED Associates). A recessed food tray was present between the levers but was not used in this experiment. The floor of each chamber (CT-ENV-005-RDT; MED Associates) was connected to a computer-controlled shock generating unit (ENV-414 and SG-219G-10; MED Associates). For the self-administration sessions, rats were connected to the syringe pump via tubing passed through a metal leash (C313CS; Plastics One) that was attached to a swivel (375/22PS; Instech) above the chamber.

Procedures

Self-administration training

Rats were first trained to self-administer fentanyl (2.5 μg/kg/infusion) through a 7-day autoshaping procedure similar to that described previously (Hofford et al. 2017). This dose of fentanyl was chosen because it elicits robust responding during 1 and 6 hr sessions (Wade et al. 2015). On each of 7 consecutive days, rats first underwent a 1-hr autoshaping session, returned to their home cage for 2 hr, and then underwent a 1-hr contingent self-administration session. During the autoshaping session, the house light was illuminated and an inactive lever (no programmed consequence for pressing; position counterbalanced across rats) was extended continuously. The active lever extended on a random time 6-min interval for the first 25–35 min of the session. The active lever remained extended for 15 s or until it was pressed, then it retracted, an infusion of fentanyl was delivered (0.1 ml over 3.4 s), and the cue lights above the levers illuminated for 20 s. A total of 5 infusions were delivered in each autoshaping session, and then the active lever remained retracted until the end of the session. The subsequent 1-hr contingent self-administration session was similar to the autoshaping session, except that (1) the active lever was available throughout the session, (2) each infusion required an active lever press, and (3) the 20-s illumination of the cue lights was also a timeout period, where lever presses had no programmed consequence.

After 7 days of autoshaping and 1-hr contingent self-administration sessions, rats next completed 21 days of 6-hr contingent self-administration sessions. The extended length of these sessions was designed to induce escalation of intake (Ahmed and Koob 1998). The 6-hr sessions were identical to the 1-hr sessions, except for their longer duration. Because previous research in our laboratory indicated that some rats will develop self-injurious behavior during the 21 6-hr sessions, we provided all rats with an aspen wood chew block (AC-S03, Lomir Biomedical Inc) on the first day of 6-hr self-administration. Chew blocks accompanied the rats in both their home cage and operant cage. To manage any self-injury, we treated individual rats with a mixture of liquid bandage (New-Skin) and metronidazole before each session. Wounds were cleansed with hibiclens and saline, then treated with antibiotic ointment the end of each session and the chamber walls, levers, and floors were sanitized using Clorox wipes or MB-10 spray after the session.

Abstinence and Nutella pre-exposure

After completing the self-administration training regimen, rats underwent 7 days of forced abstinence in their home cage. On the first day of abstinence, rats were simply weighed, but were otherwise left undisturbed. On the second day of abstinence, rats were weighed and food was restricted to 15 g/day. For abstinence days 3–7, approximately 1 g/kg of Nutella was given in the morning, smeared on the wall of the polycarbonate home cage. Rats were fed 15 g of rat chow approximately 2–3 hr later.

Extinction and reinstatement with PT150 treatment

After the 7 days of forced abstinence in the home cage, rats were randomly assigned to receive either 0, 50, or 100 mg/kg PT150 and concomitant extinction training. On the first day of extinction training, rats were dosed with 1 g/kg of a mixture of Nutella and either PT150 or placebo in the home cage, according to their treatment assignment. After 45 min, rats were removed from their home cage and placed in the operant chamber for the start of a 75-min extinction session. The first 15 min of the extinction program consisted of a blackout period, where the lights were off and the levers were retracted. The next 1 hr was identical to the 1-hr self-administration program, except that rats were not connected to the syringe pump and did not receive any infusions. At the end of these sessions, rats were returned to their home cage and fed 15 g of rat chow. Rats received a total of 14 extinction sessions before undergoing reinstatement testing.

To test for the effect of PT150 on stress-induced reinstatement of fentanyl seeking, each rat was tested with both a physical stressor (footshock) and a pharmacological stressor (the sympathomimetic yohimbine). Rats received both types of stressors in two successive randomized blocks, presented in counterbalanced order, with a minimum of 3 additional extinction sessions (no footshock or injection) intervening between each reinstatement test. There were a total of 5 reinstatement tests: 2 tests for the physical stressor (footshock or no footshock) and 3 tests for the pharmacological stressor (0, 1, or 2 mg/kg yohimbine i.p.; injections were given 30 min after the Nutella mixture). For footshock sessions, rats received mild intermittent footshock (0.7 mA; 0.5 s on, every 40 s) during the 15-min initial blackout period in the operant conditioning chamber; reinstatement tests were otherwise identical to extinction sessions.

Data analysis

Data from each phase of the experiment were analyzed using separate Poisson regression models for active and inactive lever presses. Active lever presses for data analyses were only those that occurred in the absence of the signaled TO and thus active lever presses during acquisition and escalation were equivalent to the number of infusions (see Figure 1). For the acquisition period, responses during the 1-hr contingent self-administration period were analyzed and, for the escalation period, the responses from the 6-hr self-administrations were analyzed. These acquisition and escalation sessions were analyzed separately, with sex and future PT150 assignment as between-subject factors and day as a within-subject factor. Analysis of inactive lever presses during the 6-hr self-administration sessions required adjustment of the model in order to achieve convergence, and therefore interactions between future PT150 assignment and day were removed. Extinction was analyzed with sex and PT150 dose as between-subject factors and day as a within-subject factor. Weight changes during abstinence and extinction were examined as a percentage of bodyweight on the first day of abstinence, using a piecewise-linear mixed regression model with sex and PT-150 assignment as between-subject factors and session as a within-subject factor; an additional term was added to account for changes in slope after the onset of food restriction or PT-150. Reinstatement was analyzed as a function of baseline responses (responses on extinction day 14), sex, PT150 assignment, and the within-subject factor of test. Planned comparisons were conducted to examine the effect of PT150 on performance during each test; in addition, the effects of PT150 were also analyzed separately in males and females. Planned comparisons were adjusted for false discovery rate. Similar to initial acquisition and escalation phases, presses on the active lever during extinction and reinstatement testing did not include presses during illumination of the cue signal associated with an infusion. All analyses were conducted using SAS 9.4 utilizing the GENMOD, MIXED and MULTTEST procedures.

Figure 1:

Number of lever presses (mean±SEM) across initial acquisition (1-hr) sessions (Panel A) and escalation (6-hr) sessions (Panel B) plotted as a function of sex; PT150 was not administered during acquisition or escalation. Broken down by sex, active lever pressing significantly increased more rapidly in females than males across 1-hr sessions (Panel A), but no sex differences were obtained across 6-hr sessions (Panel B). Females also emitted significantly more inactive lever presses than males across the 6-hr sessions. [Note that responses on the active lever that occurred during the 20-sec signaled timeout were not included in these figures, and thus number of active presses is equivalent to the number of infusions. However, the total number of active presses, including presses during the timeout, are also available in the supplemental Figure 1S.]

Drugs

Fentanyl HCl was obtained through the NIDA drug supply program (Bethesda, MD) and prepared at a concentration of 2.5 μg/kg for each 0.1 ml infusion, with dose expressed as salt weight. For PT150 (Palisades Therapeutics, Cliffside Park, NJ), the contents of 150 mg tablets of PT150 or placebo were dissolved in 8 drops of peanut oil and mixed into Nutella^® to create a homogenous suspension containing placebo (75 mg/g) or PT150 (50 or 100 mg/g). Yohimbine HCl (Y3125–1G, Sigma-Aldrich) was dissolved in sterile water, with doses expressed as salt weight.

Results

Acquisition:

During 1-hr self-administration training (Figure 1A), active lever responding increased across sessions for the active lever (X² = 34.56, df = 6, p < 0.001) and there was an interaction between session and sex (X² = 14.57, df = 6, p = 0.024), with females increasing active lever responding more than males after the first session. There was no effect of sex or session on inactive lever pressing during 1-hr sessions and there was no significant difference among groups based on future PT150 treatment.

Escalation:

During 6-hr self-administration sessions (Figure 1B), active lever pressing increased across sessions (X² = 41.7, df = 20, p = 0.003), but there was no effect of sex or future PT150 assignment. Sexes differed on inactive lever responding during 6-hr sessions (X² = 4.80, df = 1, p = 0.028), with females responding on the inactive lever more than males. There was no significant difference among groups based on future PT150 treatment.

Effect of PT150 during Extinction:

During extinction (Figure 2), while active lever responding decreased across sessions across all groups (Fig 2A; X² = 35.91, df = 13, p < 0.001), there was no significant effect of PT150 treatment. In addition, although there was no significant sex difference in active lever responding, inactive lever responding differed between sexes (X² = 5.96, df = 1, p = 0.015), with females (Fig 2C) responding overall more than males on the inactive lever (Fig 2B). There was no effect of PT150 on bodyweight during extinction sessions (results not shown).

Figure 2:

Responses on the active lever (mean±SEM) during extinction, plotted as a function of treatment group collapsed across sex (Panel A) or by treatment within males (Panel B) or females (Panel C). Responding decreased significantly across extinction sessions (Panel A), but there was no effect of PT150 treatment. Behavior was similar in males (Panel B) and females (Panel C), although females pressed the inactive lever significantly more than males overall.

Effect of food restriction and PT-150 on body weight during abstinence, extinction and reinstatement:

During the entire course of PT150 treatment (Figure 3), the slope of body weight across sessions relative to abstinence day 1 was non-zero throughout abstinence, extinction and reinstatement (F1,1039 = 5.26, p = 0.022) and changed direction as rats transitioned from abstinence to extinction (F1,977 = 25.8, p < 0.001). There was no effect of sex or PT150 on body weight, although there was a near-significant interaction between sex and the change in slope during extinction (p = 0.077).

Figure 3:

Body weight during abstinence, extinction and reinstatement, as a percent of bodyweight on the first day of abstinence. There was no significant effect of sex or treatment on body weight across the abstinence, extinction and reinstatement periods.

Effect of PT150 during Reinstatement:

During reinstatement testing (Figure 4), reinstatement of active lever responding differed across the 5 test sessions (Fig. 4A; X² = 29.0, df = 4, p < 0.001). In addition, there was a main effect of sex on reinstatement of active lever pressing (X² = 5.68, df = 1, p = 0.017), with females (Fig. 4C) responding overall more than males (Fig. 4B). The interaction between sex and test condition did not reach significance (p = 0.081). Reinstatement was influenced by the covariate of baseline responses (extinction session 14 in Fig 2; X² = 8.79, df = 1, p = 0.003).

Figure 4:

Reinstatement of active lever pressing (mean±SEM), plotted as a function of treatment group collapsed across sex (Panel A) or by treatment within males (Panel B) or females (Panel C). Footshock (Panel A, left) caused stress-induced reinstatement of active lever pressing that was significantly reduced by treatment with either dose of PT150. In contrast, yohimbine-induced reinstatement (Panel A, right) was not significantly altered by PT150. Sex differences in responding occurred across all test conditions, with males (Panel B) responding significantly less than females (Panel C). Exploratory examination of the effects of PT150 separately within males and females revealed that both doses of PT150 reduced footshock-induced reinstatement in males (Panel B), but not in females (Panel C). FDR-adjusted planned comparisons: ^p < 0.05 vs control condition (no shock or vehicle); *p < 0.05 vs placebo treatment; +p < 0.05 vs 1 mg/kg yohimbine.

Collapsed across sex, footshock increased active lever responding relative to the no footshock control condition for each of the treatment conditions (Fig 4A; placebo: p < 0.001; 50 mg/kg PT150: p = 0.017; 100 mg/kg PT150: p = 0.005). More important, while the full analysis did not reveal a significant interaction between PT150 and the test conditions, FDR-adjusted planned comparisons revealed that both doses of PT150 reduced reinstatement relative to placebo (50 mg/kg PT150: p = 0.038; 100 mg/kg PT150: p = 0.038).

Both doses of yohimbine increased active lever responding relative to placebo for each of the treatment conditions (Fig 4A; p’s < 0.01). However, there was no significant effect of PT150 on the response to yohimbine (p’s > 0.15). For the 50 mg/kg PT150 condition, responding was higher following 1 mg/kg yohimbine than 2 mg/kg yohimbine (p = 0.028).

Examination of the effects of PT150 on active lever pressing separately in males (Fig. 4B) and females (Fig. 4C) revealed that the effect of PT150 on footshock-induced reinstatement was driven primarily by data collected from males. For males, reinstatement of active lever responding differed across the 5 test sessions (X² = 16.1, df = 4, p = 0.002) and was affected by baseline responding (X² = 7.03, df = 1, p = 0.008). In males, FDR-adjusted planned comparisons revealed that both doses of PT150 reduced footshock-induced reinstatement relative to placebo (50 mg/kg PT150: p = 0.009; 100 mg/kg PT150: p = 0.018). Additionally, increased responding following footshock relative to the no footshock condition was observed in males treated with placebo (p < 0.001) and 100 mg/kg PT150 (p = 0.014). In contrast, in females, planned comparisons revealed no effect of PT150 on footshock-induced responding and only females treated with placebo demonstrated increased responding for footshock relative to the no footshock condition (p = 0.001). For females, reinstatement of active lever responding differed across the 5 test sessions (X² = 15.9, df = 4, p = 0.003), but this was not affected by baseline responding.

Discussion

The present study examined the effects of the selective GR antagonist PT150 on stress-induced fentanyl seeking in male and female rats. As expected, fentanyl self-administration using a long-access (6 hr) schedule led to an escalation of intake, which is a hallmark sign of OUD (APA 2013; Wade et al. 2015). Following escalation, initiation of the oral PT150 treatment had no effect on extinction of responding using a forced abstinence procedure. However, both doses of PT150 (50 and 100 mg/kg, daily p.o.) reduced shock-induced reinstatement of fentanyl seeking. Interestingly, the effect of PT150 on shock-induced reinstatement was driven primarily by males. There was no reliable effect of PT150 on yohimbine-induced reinstatement. Taken together, these results provide preclinical evidence that PT150 may serve as an effective pharmacotherapy for reducing opioid seeking following a period of abstinence, particularly among males exposed to stressful events.

There were no reliable sex differences in escalation of fentanyl intake, despite females self-administering more fentanyl during the initial 1-hr acquisition sessions that preceded 6-hr escalation sessions. In one previous study (Wade et al., 2015), only male rats were tested on long access sessions to demonstrate escalation of intake for this synthetic opioid. The current findings extend that work to females and also show that the greater intake of fentanyl observed in females during initial 1-hr acquisition sessions dissipates completely across extended 6-hr sessions, thus yielding no reliable sex differences at the end of escalation. This is consistent with a report of sex differences in heroin self-administration, where females initially acquired self-administration more rapidly, but did not differ from males in intake after acquisition (Lynch and Carroll 1999). Importantly, regardless of sex, there were no differences in fentanyl acquisition or escalation before the initiation of PT150 treatment among the treatment groups.

When treatment with PT150 was initiated on the first day of extinction training, one week after termination of the escalation phase, there was no effect of GR antagonism on the rate of extinction. While we did not assess any signs withdrawal, previous work has shown that mifepristone reduces the dysphoria associated with opioid withdrawal (McNally et al. 2005). Similarly, other studies have shown that both PT150 and mifepristone ameliorate ethanol withdrawal (Reynolds et al. 2015; Sharrett-Field et al. 2013). Because withdrawal was not measured in the current study, additional work is needed to determine if there are effects of PT150 on the dysphoria associated with opioid withdrawal. Most important, however, glucocorticoid antagonism with PT150 reduced footshock-induced reinstatement, particularly in males. This effect is consistent with studies demonstrating that blocking GRs can mitigate the effects of restraint stress on opioid withdrawal (McNally et al. 2005) and reduce stress-induced reinstatement of morphine seeking using conditioned place preference (Karimi et al. 2014). However, it is important to note that those previous studies were conducted only in males. Additional work is needed to clarify how females respond to GR antagonism, especially given that PT150 appeared to have little effect on females in the present study.

It is currently unclear why females and males differed in the effects of PT150 on footshock-induced reinstatement. Although PT150 is more selective than mifepristone for GRs (Peeters et al. 2004), there still may have been off-target effects resulting in progesterone receptor antagonism. Low progesterone can enhance stress-induced drug craving in women (Sinha et al. 2007), while supplemental progesterone may inhibit reinstatement in rats (Anker et al. 2007). Although progesterone antagonism does not affect heroin self-administration (Smith et al. 2021), the role of sex or ovarian hormones in opioid withdrawal and reinstatement is not well understood (Kokane and Perrotti 2020) and a limitation of the current study is that we did not collect vaginal smears from female rats. Thus, it is unclear if females may have differed from males due to off-target effects of PT150 on ovarian hormones. Another possibility is that females differed in sensitivity to the drug or the reinstatement paradigm. A recent study has suggested that females are less sensitive to the aversive effects of footshock and engage in less cocaine seeking after footshock (Doncheck et al. 2020). Although we found no sex difference in footshock-induced reinstatement among rats receiving placebo, we did not measure vocalizations or corticosterone release to verify that footshock served as a stressor of similar magnitude for males and females.

In contrast to the effects of PT150 on footshock-induced reinstatement, there was no effect of PT150 on reinstatement induced by yohimbine. While we chose to include yohimbine as a pharmacological stressor because it has been used extensively in previous studies examining the mechanisms of relapse, evidence suggests that yohimbine may potentiate lever pressing non-specifically, rather than specifically due to activation of the HPA stress axis (Chen et al. 2015; Mantsch et al. 2016; Tabbara et al. 2020). Indeed, there are indications that yohimbine induces a conditioned place preference, rather than an aversion (Chen et al. 2015), which calls to question whether yohimbine should be considered a stressor. However, yohimbine does cause corticosterone release (Simms et al. 2012) and increases drug seeking for a wide array of drugs of abuse (Mantsch et al. 2016), which is why it was used to induce drug seeking in the present study. Interestingly, mifepristone does not affect yohimbine-induced alcohol seeking or corticosterone release (Simms et al. 2012), consistent with the lack of effect of PT150 on yohimbine-induced opioid seeking observed here.

There are limitations to the current study that must be considered when interpreting the results. First, there may have been sex differences in the impact of food restriction during extinction and reinstatement testing, as females may have been more sated after receiving 15 g of food chow after each session due to their lower body weight compared to males. This is not likely an important factor, however, because Nutella given prior to the session was administered based on body weight for both sexes and relative body weight changes across the study did not differ between sexes. Nonetheless, to improve the translational validity of the study, it will be important to replicate the current study with rats that have not been food restricted. Second, males and females received different anesthetic cocktails during surgery; i.e., only males received acepromazine based on veterinarian consultation when we encountered higher than expected anesthetic overdoses in females. Since we cannot rule out the possibility that acepromazine had a long-term consequence on behavior, some caution is needed in directly comparing males and females. Third, the use of Nutella as the vehicle for the drug suspension may have altered responding independent of any drug effect. While use of highly palatable solid foods such as Nutella or peanut butter have been used previously for oral drug delivery, including delivery of PT150 (Rice et al. 2018), it may have impacted motivation during sessions because it was given immediately before each extinction and test session. However, we chose this method in order to model voluntary oral self-administration and to avoid the stress involved with experimenter-delivered oral gavage.

From a clinical perspective, there is an urgent need to identify novel treatments that can protect against relapse among individuals with OUD. Compounds targeting the stress system, such as the GR antagonist tested here, are of particular interest given the high prevalence of comorbid OUD and PTSD (Dabbs et al. 2014; Danovitch 2016). Mifepristone is the most studied GR antagonist (Howland 2013), with a completed clinical trial demonstrating that it reduces alcohol craving and intake (Vendruscolo et al. 2015) and an ongoing clinical trial for alcohol use disorder (Shen 2018). Clinical trials are still in the earliest stages for PT150 and these have also focused exclusively on alcohol use disorder, with recent studies launched to examine the safety and pharmacokinetics of PT150 with concurrent alcohol use (NCT04331288 and NCT03548714). The Pharmacotherapies for Alcohol and Substance Abuse (PASA) Consortium, which is funded through the Congressionally Directed Medical Research Program (CDMRP) and is the sponsor of the current study, includes several projects involving PT150. Thus, a more complete picture of the potential efficacy of PT150 against alcohol and OUD will emerge in the coming years.