Drug overdose deaths in the United States have reached record highs (Ahmad et al., 2023). The explosive increase in deaths is related to licit and illicit opioids while deaths involving cocaine and other psychostimulants also soared dramatically (Ahmad et al., 2023). This crisis spawned additional health consequences, including an increasing incidence of substance use disorders (SUDs), conditions manifested by escalating physical and psychological impairments (SAMHSA, 2021). Psychosocial interventions in combination with medications for opioid use disorder (OUD) reduce mortality, opioid withdrawal, illicit intake, and opioid-seeking in OUD patients, thus substantially improving the odds of successful recovery (Marsden et al., 2019). However, barriers persist in accessing these opioid agonist medications (buprenorphine, methadone) (Sharma et al., 2017), while there are no FDA-approved medications for cocaine use disorder (CUD). Thus, there is a need to pursue creative solutions that promote CUD and OUD recovery.

Preclinical models demonstrated that laboratory animals self-administer psychoactive drugs misused by humans. These models are also useful to study stimuli associated with drug use which contribute to relapse vulnerability during abstinence (Meil and See, 1996; Epstein et al., 2006). These models accelerated our understanding of the circuitry and neurobiological underpinnings of SUDs, providing intriguing new pharmacotherapeutic targets for OUD and CUD. One such target is ghrelin, known as the “hunger hormone.” Ghrelin binds to the growth hormone secretagogue receptor 1α (GHS1αR) to play crucial roles in appetite and adiposity, as well as modulatory roles in many organ systems, including the brain (for reviews) (Engel and Jerlhag, 2014; Zallar et al., 2017). The GHS1αR is a G protein-coupled receptor with high constitutive activity and complex pharmacology (Ramirez et al., 2019). Small molecules with selectivity as GHS1αR antagonists and inverse agonists have been developed for therapeutic applications, including metabolic diseases and obesity (for reviews) (Cameron et al., 2014; Kumar, 2019; Ramirez et al., 2019), including the peptidergic GHS1αR full antagonist JMV2959 (Demange et al., 2007; Moulin et al., 2007; Ramirez et al., 2019).

Converging literature suggests that central ghrelin actions influence the neurochemical and behavioral effects of alcohol, cocaine, and opioid agonists (Jang et al., 2013; Schuette et al., 2013; Sustkova-Fiserova et al., 2017; Abtahi et al., 2018; Zallar et al., 2018; You et al., 2022a; You et al., 2022b; Richardson et al., 2023). While ghrelin administered centrally or systemically increases alcohol consumption (Jerlhag et al., 2009; Farokhnia et al., 2018a), these outcomes are blocked in rodents treated with the brain-penetrant GHS1αR antagonist JMV2959 or other GHS1αR antagonists (Jerlhag et al., 2009; Landgren et al., 2012; Suchankova et al., 2013; Gomez et al., 2015). In human psychopharmacology studies, ghrelin increases both alcohol self-administration and alcohol-seeking in heavy alcohol drinkers, while the clinically available GHS1αR inverse agonist PF5190457 suppresses these behaviors in a similar population (Leggio et al., 2014; Farokhnia et al., 2018b; Lee et al., 2018). Ghrelin potentiates (Dunn et al., 2019) and JMV2959 suppresses (Jerlhag et al., 2010; Wenthur et al., 2019; Sustkova-Fiserova et al., 2020) cocaine conditioned place preference (CPP), an assay that taps into the rewarding effects of drug-associated contextual cues (Bardo and Bevins, 2000; Huston et al., 2013). Intriguingly, lever presses for cocaine-associated cues positively correlates with ghrelin levels in rats trained in cocaine self-administration (Tessari et al., 2007). Intraventricular ghrelin infusion also increases heroin self-administration (Maric et al., 2012), although infusion of the GHS1αR antagonist [D-Lys-3]-GHRP-6 does not alter heroin intake or reinstatement of heroin-seeking evoked by food deprivation (Maric et al., 2012). Systemic administration of JMV2959 decreases expression of morphine CPP in rodents (Jerlhag et al., 2010; Engel et al., 2015; Jerabek et al., 2017), and blunts fentanyl self-administration in rats (Sustkova-Fiserova et al., 2020). Thus, a GHS1αR antagonist may be of utility in mitigating aspects of SUDs which sustain drug intake or promote relapse vulnerability during abstinence.

The present preclinical studies were designed to investigate the relative effectiveness of JMV2959 to impact cocaine or opioid self-administration and drug-seeking. While premised by previous cocaine and opioid self-administration findings, we analyze JMV2959 in a rat self-administration assay in a different strain (Sprague-Dawley) and under distinct operant conditions than prior publications (Sustkova-Fiserova et al., 2020; You et al., 2022a; You et al., 2022b). We selected oxycodone as a highly misused opioid in humans which readily supports self-administration in rodents (Pravetoni et al., 2014; Wade et al., 2015; Neelakantan et al., 2017), has a fast onset of action likely related to its high cerebral accumulation and active influx transport through the blood-brain barrier and a long duration of action (Olkkola et al., 2013). We analyzed the effectiveness of JMV2959 to suppress cocaine or oxycodone self-administration and drug-seeking within a dose-range of JMV2959 (0.5–2 mg/kg) that lacks impact on general behavioral parameters (Landgren et al., 2012; Suchankova et al., 2013; Sustkova-Fiserova et al., 2016; Jerabek et al., 2017; Rodriguez et al., 2018). Thus, in the present study, we tested the hypothesis that pretreatment with the GHS1αR antagonist JMV2959 (0.5–2 mg/kg) would suppress cocaine and/or oxycodone self-administration and cue-induced drug-seeking in a dose-related manner.

Accumulated findings suggest that JMV2959 is efficacious in suppressing the in vivo neurochemical and behavioral effects induced by cocaine and opioid agonists, premising the studies conducted here. We found that under the current experimental conditions, JMV2959 does not impact self-administration of cocaine nor oxycodone. This is in direct contrast with previous studies that show pharmacological blockade of the GHS1αR via JMV2959 is sufficient to blunt cocaine and oxycodone intake (You et al., 2022a; You et al., 2022b). It is possible that the doses of JMV2959 employed here (0.5–2 mg/kg) were below the effective threshold to supersede the reinforcing properties of either self-administration training drug; JMV2959 was reported to blunt oxycodone intake when administered at doses of 2.5 mg/kg or higher (You et al., 2022b). Several additional experimental factors may also explain the discrepancies between the current and published studies (e.g., rat strain, dose of training drug, session length, schedule of reinforcement). Conversely, JMV2959 robustly decreased responding for the discrete cue complex previously associated with either cocaine or oxycodone delivery in male rats in the present study. In agreement, JMV2959 dose-dependently reduced cocaine- and stress-primed reinstatement of cocaine-seeking in male Long-Evans rats (You et al., 2022a). The dose range of JMV2959 (0.5–2 mg/kg) employed here did not alter operant responses on the inactive lever, but increased latency to the first response in the oxycodone-taking study. This effect is primarily driven by two of the rats assessed (n = 10) (Figure 3C). Doses of JMV2959 exceeding 3 mg/kg have been noted to alter general behavioral measures in previous studies in male Wistar or Long-Evans rats and thus were avoided (Landgren et al., 2012; Sustkova-Fiserova et al., 2014; Sustkova-Fiserova et al., 2016; Jerabek et al., 2017). Importantly, our unpublished data indicate that JMV2959 treatment does not alter locomotor activity under the same conditions employed in the present studies (dose, route of administration, and pretreatment time). Male Sprague-Dawley rats received JMV2959 (0–2 mg/kg; i.p.; 20 min) and were placed into open field chambers for 2 hours. A between-subjects two-way ANOVA with the factors of time and treatment time reveal a main effect of time (F_23,644 = 92.87, p < 0.01), but no main effect of treatment (F_3,28 = 0.63, p = 0.60) or time × treatment interaction (F_69,644 = 0.92, p = 0.69). Taken together, these experiments provide compelling evidence that the GHS1αR-ghrelin axis regulates aspects of relapse-like events associated with both cocaine and oxycodone.

Ghrelin is proposed to cross the blood-brain barrier or affect vagal function to evoke centrally mediated effects ultimately through modulation of the meso-corticolimbic, hypothalamic, and hippocampal pathways (Banks et al., 2002; Zigman et al., 2006; Cabral et al., 2017; Rhea et al., 2018). Of relevance to SUDs, the mesolimbic dopamine (DA) pathway from ventral tegmental area (VTA) DA cell bodies to the nucleus accumbens (NAc) is a chief mediator of the rewarding properties of misused drugs and drug-associated cues (for review) (Volkow et al., 2016). The VTA is a key region in which ghrelin is thought to promote motivation for hedonic rewards (for review) (Abizaid et al., 2006). The GHS1αR transcript is enriched in the VTA (Guan et al., 1997; Kojima et al., 1999) and, although expressed in both VTA DA and γ-aminobutyric acid (GABA) neurons (Abizaid et al., 2006; Zigman et al., 2006), accumulated data suggest that the actions of ghrelin in the VTA are mediated by GHS1αR-expressing DA, but not GABA, neurons (Abizaid et al., 2006). Systemic injection of JMV2959 reduces both cocaine- (Jerlhag et al., 2010) and morphine-induced DA efflux in the NAc (Sustkova-Fiserova et al., 2014; Jerabek et al., 2017), presumably from DA neuronal terminals originating in the VTA. Moreover, acquisition of oxycodone self-administration elevates GHS1αR mRNA expression in VTA neurons containing the dopamine transporter (DAT), but not the vesicular GABA transporter (vGAT), further evidencing involvement of VTA dopaminergic cells in oxycodone-motivated behavior (You et al., 2022b). Based upon these findings, JMV2959 was expected to block the GHS1αR localized to VTA DA neurons, reducing the elevation in resulting NAc DA concentrations, and accordingly impacting both cocaine and oxycodone self-administration. However, JMV2959 failed to reduce the intake of either drug under the current experimental conditions.

The interpretation of these findings requires conjecture within the context of the target actions of cocaine and oxycodone to impact the mesoaccumbens DA pathway. The entirety of their pharmacological actions in vivo are not identical, however, both cocaine (Pettit et al., 1990; Weiss et al., 1992) and oxycodone (Vander Weele et al., 2014) result in overflow of NAc DA efflux. Cocaine has a high affinity for monoamine transporters, with the inhibition of DA reuptake processes playing a prominent role in its reinforcing effects (Ritz et al., 1987). In our hands, JMV2959 pretreatment did not alter self-administration at a cocaine dose (0.25 mg/kg/infusion) which is associated with sustained NAc DA overflow at ∼300% of baseline across the session (Pettit et al., 1990; Weiss et al., 1992). A dose of JMV2959 (6 mg/kg) considerably higher than those doses employed here (0.5–2 mg/kg) suppressed, but did not eliminate, NAc DA overflow evoked by an investigator-delivered dose of cocaine (10 mg/kg) (Jerlhag et al., 2010), suggesting that achieved levels of cocaine-evoked DA in the NAc may be insurmountable by the JMV2959 doses employed in the present study.

The functional role of the GHS1αR in neural nodes is tied to reward processes and drug-seeking behaviors (Jang et al., 2013; Schuette et al., 2013; Sustkova-Fiserova et al., 2017; Abtahi et al., 2018). Ghrelin is predominantly produced in the stomach mucosa (Kojima et al., 1999) and is proposed to cross the blood-brain barrier to evoke centrally mediated effects, although there is also evidence for expression of ghrelin-containing neurons within the brain (Banks et al., 2002; Hou et al., 2006; Rhea et al., 2018). Despite the uncertainty surrounding the source of ghrelin in the brain, converging literature suggests that central GHS1αR action or stimulation underlie the efficacy of GHS1αR ligands to influence the behavioral effects of misused drugs, and supports the premise that this system modulates behaviors with underlying deficits in prefrontal executive control, such as cue-evoked drug-seeking (Wellman et al., 2012; Jang et al., 2013; Schuette et al., 2013; Edwards and Abizaid, 2017; Sustkova-Fiserova et al., 2017; Abtahi et al., 2018; Wenthur et al., 2019). Further, ex vivo biochemical data show that acute treatment with JMV2959 decreases the metabolism of the monoamine serotonin within the amygdala, a region implicated in withdrawal and negative affect in the absence of misused substances (Vestlund et al., 2019). Intra-amygdala infusion of a serotonin 5-HT_2C receptor agonist suppresses cocaine-seeking during a cue-reinforced reinstatement test (Pockros-Burgess et al., 2014), as does systemic administration (Cunningham et al., 2011; Cunningham et al., 2013). It is possible that an underlying mechanism through which GHS1αR blockade suppresses cocaine- and oxycodone-seeking is due to JMV2959-induced increases in amygdala serotonin levels and activation of the prominent expression of the 5-HT_2CR population in this region (Pompeiano et al., 1994; Clemett et al., 2000). As discussed above, additional findings also suggest the GHSR1α and central ghrelin actions have been shown to interact with serotonergic systems (Schellekens et al., 2015) as well as dopaminergic (Jerlhag et al., 2011; Jerlhag et al., 2012; Navarro et al., 2022) and cannabinergic systems (Kola et al., 2008).

Intra-VTA JMV2959 suppresses cue-evoked heroin-seeking in food-restricted, but not sated rats, potentially due in part to the elevated plasma ghrelin levels and the neuroplastic changes associated with chronic food restriction (D'Cunha et al., 2013). The rats in our study were freely fed, yet still demonstrated a robust effect of systemic JMV2959 to suppress cue-evoked drug-seeking. One possibility is that the effects of systemic JMV2959 to suppress drug-seeking behavior may be due to effects localized to the prefrontal cortex (PFC) which exerts top-down control over cue-induced dopaminergic signaling from the VTA to the NAc (Koob and Volkow, 2010). Specifically, PFC glutamatergic outputs to the striatum serve a regulatory role over crucial cue-induced DA release (for review) (Kalivas and Volkow, 2005). The GHS1αR is resident in the PFC, but localized function in this region has yet to be studied; thus, it is possible that GHS1αR control of cue-evoked drug-seeking may be due to its actions in the PFC (Guan et al., 1997). Absolute GHS1αR deletion via CRISPR/Cas9 technology has been shown to decrease alcohol-motivated behavior in rats (Zallar et al., 2019a; Zallar et al., 2019b); studies designed to assess the impact of overall GHS1αR knockout relative to localized GHS1αR gene silencing within the PFC on opioid- and cocaine-motivated behaviors will be valuable in the future.

The present studies were conducted in male rats, and given sex-specific distinctions in the ghrelin-GHS1αR in rats, differences in drug self-administration and cue-reinforced drug-seeking are projected (Lopez-Ferreras et al., 2017; Labarthe et al., 2021). Additionally, male Sprague-Dawley rats are more sensitive to the antinociceptive effects of morphine than female rats, and to a lesser degree to the major metabolite (oxymorphone) of oxycodone, while no sex differences were observed in this regard for fentanyl (Peckham and Traynor, 2006). Future comparative studies are needed to ascertain the extent to which the effectiveness of JMV2959 to suppress drug reward-related behaviors is dependent upon sex and sex hormones (Sell et al., 2000; Zhou et al., 2002; Hinds et al., 2023). In addition, we employed a daily, short-access model of drug self-administration to determine the impact of GHSR1α blockade on drug intake and drug-seeking. Notably, long-access (>6 h) cocaine self-administration paradigms produce escalation of drug intake over time with appreciable differences in drug consumption between male and female rats (Algallal et al., 2020). A recent study also reported significant sex differences in escalation of oxycodone intake (12h/session), however this was only the case for one of four heterogeneous strains tested (Doyle et al., 2023). Escalated drug intake in these preclinical models align with several criteria for SUDs outlined by the Diagnostic and Statistical Manual of Mental Disorders 5 including development of tolerance and greater intake with extended use (Hasin et al., 2013). It is important to determine if the suppressive effects of JMV2959 on drug-motivated behaviors persist in long-access self-administration models.

The mechanisms of action for cocaine and oxycodone are independent yet each drug is a powerful reinforcer with high abuse potential. We identified that the GHS1αR antagonist JMV2959 mitigates distinct aspects of cocaine- and oxycodone-associated behaviors in male rats. Future studies are needed to further clarify specific mechanisms for the regulatory control of the ghrelin-GHS1αR axis over cocaine and opioid intake and drug-seeking.