JQ1 attenuates psychostimulant- but not opioid-induced conditioned place preference

Abstract

Epigenetic mechanisms play important roles in the neurobiology of substance use disorder. In particular, bromodomain and extra-terminal domain (BET) proteins, a class of histone acetylation readers, have been found to regulate cocaine conditioned behaviors, but their role in the behavioral response to other drugs of abuse remains unclear. To address this knowledge gap, we examined the effects of the BET inhibitor, JQ1, on nicotine, amphetamine, morphine, and oxycodone conditioned place preference (CPP). Similar to previous cocaine studies, systemic administration of JQ1 caused a dose-dependent reduction in the acquisition of amphetamine and nicotine CPP in male mice. However, in opioid studies, JQ1 did not alter morphine or oxycodone CPP. Investigating the effects of JQ1 on other types of learning and memory, we found that JQ1 did not alter the acquisition of contextual fear conditioning. Together, these results indicate that BET proteins play an important role in the acquisition of psychostimulant-induced CPP but not the acquisition of opioid-induced CPP nor contextual fear conditioning.

Introduction

Aberrant regulation of histone acetylation modifications has been implicated in the pathophysiology of several brain disorders, including substance use disorder (SUD) [1–5]. Acetyl-lysine modifications are added to and removed from histones by epigenetic writer and eraser proteins called, histone acetyltransferases (HATs) and histone deacetylases (HDACs), respectively [6,7]. Epigenetic reader proteins called bromodomains recognize and bind to histone acetylation marks and act as a scaffold for the recruitment of additional chromatin regulators and transcription factors [6]. Both acute and chronic exposure to drugs of abuse elicit changes in histone acetylation levels in reward-related brain regions [8–10]. Furthermore, the functional significance of some proteins that regulate the histone acetylation has been demonstrated in animal models of SUD [11–14]. Thus, these data indicate that targeting histone acetylation mechanisms is a promising therapeutic approach for the treatment of SUD.

Bromodomain and extra-terminal (BET) proteins (BRD2, BRD3, BRD4 and BRDT) are a class of histone acetylation readers that have been shown to play an important role in the neurobehavioral responses to cocaine [15,16]. In previous studies, BRD4 was found to be significantly elevated in the nucleus accumbens (NAc) of rats following cocaine self-administration and conditioned place preference (CPP). In behavioral experiments, systemic and intra-NAc administration of JQ1, a pharmacological inhibitor of BET proteins [41], attenuated cocaine conditioned place preference (CPP) and self-administrations but did not affect locomotor activity or acquisition of lithium chloride conditioned place aversion [15,16]. Although these initial studies have demonstrated the functional importance of BET proteins in cocaine-induced behaviors, a role for BETs in the behavioral response to other drugs of abuse and other types of learning remains unclear. Here, we examined the effects of JQ1 on psychostimulant (nicotine and amphetamine) and opioid (oxycodone and morphine) CPP as well as the effects of BET inhibition on contextual fear conditioning. Our results indicate that JQ1 reduces the acquisition of psychostimulant CPP, but the acquisition of opioid CPP and contextual fear conditioning is unaltered by JQ1.

Materials and Methods

Drug treatments:

(±) JQ1 and the inactive enantiomer (−) JQ1 (Cayman Chemical) were dissolved in 10% DMSO and 10% Tween 80 (v/v) and then diluted to the final concentration using sterile saline, as previously described [15,17]. Vehicle was delivered at the same volume as the JQ1 solution. D-amphetamine, nicotine hydrogen tartrate, morphine sulfate, and oxycodone hydrochloride (Sigma-Aldrich) were dissolved in 0.9% sterile saline and administered by an intraperitoneal (i.p.) injection.

Animals:

Male C57BL/6 mice (8-10 weeks old; Charles River Laboratories) were group-housed under a regular 12 h/12 h light/dark cycle and had ad libitum access to food and water. Animals were housed in a humidity- and temperature controlled in an AAALAC-accredited animal facility at the University of Connecticut or the University of Miami Miller School of Medicine. Experiments were approved by the Institutional Animal Care and Use Committee and conducted according to guidelines established by the U.S. Public Health Service Policy on Humane Care and Use of Laboratory Animals.

Conditioned place preference (CPP):

This CPP procedure used similar methods as previously described [15]. The CPP apparatus was comprised of two adjoined chambers separated by a removable partition. One side of the chamber had black walls and rod grid floors, and the other side had white walls and mesh floor. During a pre-test acclimation session, the divider was removed, and mice were allowed to freely explore both compartments for a duration of 15 min. Time spent in each chamber was recorded using EthoVision tracking software. Mice that spent >65% of the time in one chamber during pretest were removed from the experiment. During the next 3 days of conditioning, mice were injected with saline and restricted to one side of the chamber in the morning (9 AM). In the afternoon (2 PM), mice were injected with the drug of abuse (3 mg/kg of amphetamine, 0.5 mg/kg of nicotine, 5 or 10 mg/kg of morphine, or 3 mg/kg of oxycodone, i.p.) and restricted to the opposite chamber for 30 minutes. Vehicle, JQ1 (25 or 50 mg/kg, i.p.), or (−)JQ1 (50 mg/kg, i.p.) was administered 5 min before each drug conditioning session. The doses used for JQ1, amphetamine, nicotine, morphine, and oxycodone were based on previously published results [15,18–21]. One day after the last conditioning session, mice received a drug-free CPP test where they had free access to both chambers. The preference score was calculated by subtracting the time spent on the drug-paired side during the posttest from the time spent on the same side during the pretest.

Contextual fear conditioning:

Med Associates fear conditioning chambers were positioned inside a sound attenuation box. The test chamber consisted of clear Plexiglas walls and metal grid floors used to deliver a mild electric foot shock. A camera inside the sound attenuation box was used to record movement and freezing behavior. During conditioning, mice were placed in the test chambers for 8 minutes. After 3 minutes of habituation, a white noise (80 dB), which served as the conditioned stimulus (CS), was presented for 30 seconds followed by a mild (2 s, 0.8 mA) foot shock, which served as the unconditioned stimulus (US). The CS–US pairings were presented again at 5 minutes and 7 minutes. Mice were removed from the chambers 30 seconds after the last shock, injected with vehicle or JQ1 (25 or 50 mg/kg, i.p.), and then returned to their home cage. The next day, mice were placed back in the test chambers without a shock for 8 minutes. Freezing behavior was analyzed using Med Associates Video Freeze® software. Freezing was defined as no movement with the exception of those movements associated with breathing. The total freezing time during the test was represented as a percentage.

Data analysis:

Statistical analyses were performed with GraphPad Prism 7.0 software. Mean values from behavioral studies were compared between groups using Analysis of variance (ANOVA). When a significant F-value was observed, comparisons to control were performed using the Newman-Keuls post-hoc test. Data are expressed as means ±SEM and level of significance was set at P < 0.05.

Results

The effects of JQ1 on the acquisition of amphetamine and nicotine CPP

Previous studies have shown that JQ1 is effective at reducing cocaine-induced gene expression and behaviors [15,16] but little is known about a role for BET proteins in the behavioral response to other psychostimulants. Here, we examined the effects of JQ1 on amphetamine and nicotine CPP. In Figure 1A, JQ1 (25 and 50 mg/kg) or the inactive enantiomer (−)JQ1 (50 mg/kg) was administered during the acquisition of amphetamine CPP. Mice treated with 25 and 50 mg/kg of JQ1 showed a significant reduction in amphetamine preference scores compared to vehicle treated mice (One-way ANOVA: F _{(3, 28)} = 4.033, P = 0.0167). However, (−)JQ1 (50 mg/kg) did not alter amphetamine CPP when compared to vehicle treated mice (P > 0.05 via Newmann-Keuls post hoc test) (Figure 1A). In a different group of mice, JQ1 (25 and 50 mg/kg) also reduced the acquisition of nicotine CPP compared to vehicle treated mice (One-way ANOVA: F _{(2, 21}) = 6.441, P = 0.0066) (Figure 1B).

Figure 1. JQ1 attenuates the acquisition of amphetamine and nicotine CPP.

(A) Amphetamine CPP was significantly reduced by JQ1 (25 and 50 mg/kg) compared to vehicle, but the inactive enantiomer (−)JQ1 had no significant effect compared to vehicle treated mice. (B) JQ1 (25 and 50 mg/kg) significantly attenuated the acquisition of nicotine CPP compared to vehicle treated mice. *P < 0.05 and **P < 0.01 indicate a significant difference using Newman-Keuls post hoc test. Data are mean ±SEM.

The effects of JQ1 on the acquisition of morphine and oxycodone CPP

Here, we sought to determine the effects of JQ1 on opioid conditioned place preference. In the first experiment, JQ1 (25 and 50 mg/kg) was administered during the acquisition of morphine (10 mg/kg) CPP. Compared to vehicle treated mice, neither dose of JQ1 was effective at altering morphine CPP (Figure 2A) (One-way ANOVA: F _{(2, 20)} = 1.472, P = 0.2533). Next, to determine if JQ1 had effects using a weaker reinforcer, mice were conditioned with a lower dose of morphine (5 mg/kg). Again, JQ1 had no significant effect on morphine CPP (One-way ANOVA: F _{(2, 21)} = 0.5917, P = 0.5624) (Figure 2B). Finally, mice were treated with JQ1 (25 and 50 mg/kg) during the acquisition of oxycodone (3 mg/kg) CPP. Similar to morphine CPP results, JQ1 had no significant effect on oxycodone CPP (One-way ANOVA: F _{(2, 21)} = 0.453, P = 0.6418) (Figure 2C).

Figure 2. JQ1 does not affect the acquisition of morphine and oxycodone CPP.

Mice were conditioned with 10 mg/kg of morphine (A), 5 mg/kg of morphine (B), or 3 mg/kg of oxycodone (C). Compared to vehicle treated mice, JQ1 (25 and 50 mg/kg) had no effect on the acquisition of morphine (A-B) and (C) oxycodone CPP. Data are mean ±SEM.

The effects of JQ1 on the acquisition of fear conditioning

Previous studies have shown that histone acetylation mechanisms are important for fear conditioning [22–24]. Because BET proteins are readers of histone acetylation, we sought to determine if JQ1 would also alter the acquisition of contextual fear conditioning. Immediately after training, mice were treated with vehicle or JQ1 (25 or 50 mg/kg, i.p.). The next day, freezing behavior was measured. Compared to vehicle treated mice, JQ1 had no effect on the acquisition of contextual fear conditioning (One-way ANOVA: F _{(2, 20}) = 0.1459, P = 0.8651) (Figure 3).

Figure 3. JQ1 does not affect the acquisition of contextual fear conditioning.

Immediately following fear conditioning training, mice were injected with vehicle or JQ1 (25 or 50 mg/kg). The next day, mice were returned to the test chamber and freezing behavior was recorded. Compared to vehicle treated mice, JQ1 had no significant effect on freezing behavior. Data are mean ±SEM.

Discussion

In previous experiments, pharmacological inhibition of BET proteins using JQ1 was shown to attenuate cocaine-seeking behaviors [15,16]. The goal of the current study was to determine if JQ1 also decreases the conditioned response for other drugs of abuse. We revealed that systemic administration of JQ1 significantly reduces the acquisition of amphetamine and nicotine CPP, but not morphine and oxycodone CPP. The findings with amphetamine and nicotine are consistent with previously published data showing a functional role for BET proteins in cocaine CPP and self-administration [15,16]. Although amphetamine- and nicotine-induced changes in BET expression and activity have yet to be fully explored, several BET-target genes that are known to regulate amphetamine- and nicotine-seeking behaviors are reduced by JQ1 or other BET inhibitors. For example, BET inhibition reduces the expression of Arc, Bdnf, Gria1, Gria2 [16, 17,25–27], factors that are elevated in reward-related brain regions by amphetamine and/or nicotine exposure [28–33]. As an indication that BRD4 may play a role in nicotine-induced responses, a study by Saito and colleagues found that repeated nicotine injections caused an upregulation of Brd4 mRNA in mouse mesotelencephalic neurons [27]. Together, these results suggest that BET proteins may have overlapping functions in psychostimulant-induced transcriptional activity and behavioral responses.

Although JQ1 did not alter acquisition of morphine and oxycodone CPP in the current experiments, a previous study showed that histone acetylation modifications, which are read by BET proteins, are elevated in the postmortem brains of chronic opioid abusers, and injections of JQ1 in the dorsal striatum reduced heroin self-administration in rats [34]. Differences in experimental design (CPP vs self-administration, systemic vs. intracranial injections of JQ1, acute vs. chronic exposure to opioids) may explain the divergent effects observed between the two studies. An additional explanation may relate to the role of certain BET-target genes in cocaine versus opioid CPP. For example, JQ1 is known to reduce Bdnf expression, and previous experiments indicate that BDNF suppresses morphine CPP in the ventral tegmental area (VTA), whereas BDNF injections in the NAc enhances cocaine CPP [35,36]. Although we administered JQ1 systemically in the current studies, JQ1-mediated inhibition of Bdnf expression in specific brain regions is a potential mechanism of action that could explain the observed differences between psychostimulants and opioids. However, Bdnf is just one of hundreds of genes altered by BET inhibition (Sullivan et al. 2015; Korb et al. 2015), and future mechanistic studies are needed to explore the dichotomous effects of JQ1 on psychostimulant versus opioid CPP. Furthermore, it is important to note that epigenetic and transcriptional changes that arise during the acquisition of opioid-induced behaviors are likely different compared to the adaptations occurring after chronic opioid use and withdrawal [37–39]. Thus, even though JQ1 did not reduce the acquisition of morphine and oxycodone CPP, prior evidence supports BET inhibitors as a promising therapeutic option in clinically relevant models of opioid use disorder [34].

In addition to testing the effects of JQ1 on psychostimulant and opioid-induce CPP, we also found that JQ1 did not alter the acquisition of contextual fear conditioning. Even though JQ1 reduces the expression of several genes involved in neuroplasticity [15,25–27], previous studies have reported mixed effects of BET inhibition on different types of learning and memory [17,25,27,40]. For example, BET inhibition impaired novel object recognition memory and extinction of remote auditory fear memory but did not affect the acquisition of fear conditioning, lithium chloride conditioned place aversion, Barnes maze, or Y maze [15,17,25,27,40–42]. In other behavioral experiments, JQ1 enhanced learning and memory in contextual fear conditioning and Morris water maze tasks and increased hippocampal long-term potentiation in a mouse model of Alzheimer’s disease [27]. Other behaviors such as locomotor activity, rotarod test, and thigmotaxis, however, were not changed by JQ1 in multiple studies [15,25,27,40–42], an indication that BET inhibition is not producing comprehensive behavioral impairments.

JQ1 has been the primary tool compound used to study the roles of BET proteins in brain disease models [43]. JQ1 and similar pan-BET inhibitors achieve their effect by binding to both bromodomains (BD1 and BD2) within all BET proteins and prevent BET proteins from binding to acetylated histones and proteins [44]. Although relatively few side effects have been identified in preclinical models using JQ1, thrombocytopenia and gastrointestinal toxicity have been reported using similar pan-BET inhibitors in oncology clinical trials [45,46]. In order to decrease the potential toxic liabilities associated with pan-BET inhibitors in humans, more recent research has focused on developing protein- (e.g., BRD4-selective) and domain-selective (e.g., BD1 vs. BD2-selective) BET inhibitors [47–50]. Moreover, if eventually tested in humans with psychostimulant addiction, as compared to cancer patients, it is conceivable that BET inhibitors can be dosed lower and more periodically, mitigating toxicology concerns. Future studies examining these more selective novel compounds and dosing regimens in sophisticated animal models of addiction are needed to advance BET-related therapeutics as a potential safe and effective treatment for substance use disorder.