Discussion

Prior studies have reported that a subset of rats develop unusually high and persistent levels of drug intake when allowed to self-administer MDPV or structurally-related synthetic cathinones (Gannon et al., 2017, 2018b; Doyle et al., 2021b). In addition to being observed across multiple studies and laboratories, because this unusual pattern of drug-taking is enduring and readily transfers to other stimulants (e.g., cocaine, methamphetamine; Gannon et al., 2017), we believe that this high-responder phenotype represents a novel model to investigate the binge-like and/or compulsive patterns of stimulant use often reported by humans. Accordingly, the present study sought to better define the behavioral and pharmacologic determinants of the high-responder phenotype by evaluating MDPV self-administration in rats with histories of reinforcement by drugs with diverse pharmacological mechanisms and nonoverlapping discriminative stimulus effects as well as rats that were provided pharmacologic histories of noncontingent cocaine and MDPV exposure to match those achieved through response-contingent self-administration. In addition to replicating prior findings (Gannon et al., 2017; Doyle et al., 2021b) by demonstrating that a subset of rats responding for MDPV (∼30%), but not cocaine, develop a high-responder phenotype characterized by persistently high levels of drug-taking combined with high levels of responding during postinfusion TO periods, the present studies also showed that: 1) unlike when responding was previously maintained by cocaine, histories of responding maintained by fentanyl, nicotine, or ketamine failed to prevent the development of the MDPV high-responder phenotype; 2) when provided a history of noncontingent cocaine exposure, an intermediate proportion of rats went on to develop the MDPV high-responder phenotype; and 3) when provided a history of noncontingent MDPV exposure, rats failed to develop the high-responder phenotype for cocaine self-administration. Together, these findings suggest that both behavioral and pharmacological histories play a role in the development of the high-responder phenotype, and highlight the importance of response-contingent drug administration in both establishing (MDPV history) and preventing (cocaine history) the aberrant patterns of responding and high levels of drug intake that may be related to the development of substance use disorders in humans.

Previous studies suggested that reinforcement histories with cocaine, but not food, can prevent the development of the MDPV high-responder phenotype (Gannon et al., 2017; Doyle et al., 2021b). The current study replicated and extended this basic finding by determining whether histories of reinforcement by drugs from other pharmacological classes would also interact with the development of the MDPV high-responder phenotype. Although 1 of the 16 (6%) rats with a history of cocaine self-administration developed the high-responder phenotype when MDPV was available for self-administration, rats that were provided histories of fentanyl (50%), nicotine (56%), or ketamine (38%) self-administration went on to develop the MDPV high-responder phenotype at proportions similar to that observed in rats that had only ever self-administered MDPV (31%). These findings suggest a history of cocaine self-administration is unique in its capacity to prevent the development of this high-responder phenotype; however, further studies will be required to determine whether the self-administration of other stimulant drugs (e.g., methamphetamine) is also capable of preventing the high-responder phenotype. In addition, because these studies only assessed the impact of relatively short/limited access (i.e., 90-minute daily sessions) to cocaine reinforcement, it will be important to determine whether cocaine would retain its protective effects if rats were provided histories of cocaine reinforcement under long/extended access or intermittent access conditions, which have been shown to result in escalation of drug-taking, increases in compulsive-like behaviors (Ahmed and Koob, 1998; Vanderschuren and Everitt, 2004), and changes reinforcing effectiveness, cue-induced reinstatement, and dopamine transporter sensitivity (Zimmer et al., 2012; Calipari et al., 2013; Allain et al., 2018; Garcia et al., 2020) that are thought to contribute to substance use disorder-related phenotypes.

To determine whether the contingency between the response and the drug effect was important to the protective effects of cocaine (and the faciliatory effects of MDPV), the current study also incorporated two groups of rats that received noncontingent infusions of cocaine (or MDPV) through a yoked procedure that matched drug exposures (i.e., the timing and number of infusions) to rats that were actively self-administering cocaine (or MDPV). Importantly, when compared with rats that were actively self-administering cocaine (6%) or other drugs (44%; collapsed across MDPV, fentanyl, ketamine, and nicotine), a history of noncontingent cocaine exposure appeared to provide an intermediate protection (19%) against development of the high-responder phenotype. When taken together with the finding that a history of noncontingent MDPV exposure failed to facilitate the development of the high-responder phenotype for cocaine, these findings suggest that the capacity of cocaine to interfere with the development of the high-responder phenotype for MDPV is dependent on both behavioral and pharmacological aspects of these histories. Although the relative importance of the contingencies of reinforcement is likely related to associations that are formed between the response and the drug effects, aspects of the pharmacologic properties of cocaine (and MDPV) that contribute to the protective (and faciliatory) effects are less clear.

Given that MDPV is much more selective than cocaine at inhibiting uptake at DAT relative to SERT (∼750-fold vs. 1.2-fold; Gannon et al., 2018a), it is possible that actions at SERT, or lack thereof, are important for the protective and faciliatory effects of cocaine and MDPV, respectively. Supporting this notion, other DAT-preferring, pyrrolidine-containing cathinones (e.g., 3,4-methylenedioxy-apyrrolidinobutiophenone α-PVP) are also associated with the development of this high-responder phenotype (Gannon et al., 2018b). However, similar phenotypes are not observed in rats that self-administer methylphenidate (Collins et al., 2012), another DAT-selective uptake inhibitor (Markowitz and Patrick, 2008), suggesting that other factors are more important in establishing the high-responder phenotype. Further supporting this, genetic deletion of SERT did not alter the reinforcing effects of cocaine in mice (Thomsen et al., 2009); however, additional studies evaluating MDPV and cocaine self-administration in SERT-knockout mice are needed to rule out a role for SERT in the protective effects of cocaine. Another possibility is that, because the discriminative stimulus effects of cocaine and MDPV overlap (e.g., Gatch et al., 2013; Collins et al., 2016; Seaman et al., 2021), rats are unable to differentiate between conditions in which cocaine or MDPV are available for self-administration. This would be particularly important if potentially aversive effects of cocaine, perhaps related to inhibition of cardiovascular sodium channels (Matthews and Collins, 1983), result in rats first learning to self-administer cocaine in a more regulated and/or cautious manner before gaining access to MDPV. Indeed, bolus doses of MDPV as large as 5.6 mg/kg intravenously in rats (Hambuchen et al., 2017) and 140 mg/kg i.p. in mice (Muskiewicz et al., 2020) have been administered without producing convulsions. Conversely, cocaine-induced convulsions can be observed after bolus doses of ∼9 mg/kg intravenously in rats (Mets et al., 1998) and 56 mg/kg i.p. in mice (Ko et al., 2007). When compared with doses of MDPV and cocaine to maintain half-maximal responding under a PR schedule in rats (0.017 and 0.17 mg/kg/inf, respectively; Gannon et al., 2018a), or cocaine-like discriminative stimulus effects in mice (0.1 and 3 mg/kg, i.p., respectively; Gannon et al., 2016), the potency ratio for MDPV to produce abuse-related and toxic/aversive effects is at least 6- to 70-fold greater than that for cocaine. Thus, the relatively narrow window between the reinforcing/abuse-related effects and the convulsant/aversive effects for cocaine relative to MDPV could be one factor that moderates the level of intake for cocaine, but not MDPV. Although one might hypothesize that the unpredictable nature of noncontingent cocaine infusions would result in their being more aversive than response-contingent cocaine infusions, the contextual associations learned under such scenarios may be insufficient to impact subsequent MDPV self-administration.

Although there are insufficient data to directly compare the onset of action for MDPV and cocaine, a recent study in rats self-administering cocaine (Canchy et al., 2021) suggests that response latencies can be used as a behavioral readout of pharmacokinetic properties of cocaine. Indeed, they reported that upon completing a response requirement, rats withhold responding until sometime after the drug effect is perceived; however, if a cocaine infusion was omitted, rats began responding sooner. The time to peak effect for cocaine (increases in synaptic dopamine levels) is estimated to be tens of seconds (Wise and Kiyatkin, 2011; Canchy et al., 2021), and although little is known about the time to peak effect for MDPV, it is unlikely to be less than 5 seconds, the duration of the TO in the present studies. Nevertheless, if one assumes that the TO responding associated with MDPV self-administration is due to a slower time to peak effect, then one would also expect that substituting MDPV for cocaine would result in more TO responding (i.e., MDPV would be perceived as a delayed/omitted reward), whereas substituting cocaine for MDPV would result in less TO responding (i.e., cocaine would be perceived as a more immediate reward). Given that neither of these outcomes were observed, we do not believe that pharmacokinetics, at least related to the time to peak effect, is a causal factor in facilitating or preventing the high-responder phenotype.

Consistent with patterns of responding maintained by drug reinforcers under simple FR schedules with relatively short postinfusion TOs, very little TO responding was observed when responding was maintained by cocaine, nicotine, or ketamine; however, when fentanyl was available, a subset of rats (3 of 16) made ∼26% of their active lever responses during TOs and earned nearly twice as many infusions than the other rats (∼32 vs. ∼17 infusions per session). Although this suggests that the high-responder phenotype may not be unique to synthetic cathinones, the importance of TO responding to the overall phenotype is not entirely clear. Others have also reported fairly large variability in fentanyl intake in rats whose responding was maintained by intravenous or vapor fentanyl (e.g., Stevenson et al., 2020; Bakhti-Suroosh et al., 2021; McConnell et al., 2021), suggesting that fentanyl may facilitate a high-responder phenotype similar to that seen with MDPV self-administration. Taken together with recent findings suggesting that the high-responder phenotype persists when oxycodone is substituted for MDPV (Gannon et al., 2021), the current findings highlight the need to further explore interactions between this high-responder phenotype and opioid self-administration. Additionally, studies evaluating whether a history of cocaine self-administration can similarly prevent the development of this fentanyl high-responder phenotype could provide insight into the mechanism(s) that underlie the development of these high-responder phenotypes as well as their prevention by cocaine.

In summary, the current study builds on previous work demonstrating that subsets of rats that self-administer MDPV and structurally-related synthetic cathinones develop high levels of drug-taking and responding during periods of drug unavailability (i.e., drug seeking), both of which are key features of substance use disorders in humans. In addition, the present study provided further evidence that a history of cocaine administration can prevent the development of the MDPV high-responder phenotype, and advanced our understanding by demonstrating that these effects (as well as the faciliatory effects of MDPV) are dependent upon the response-contingent administration of these drugs. Although the precise mechanisms that underly the development (and prevention) of this high-responder phenotype remain unclear, that they were not influenced by histories of responding maintained by drugs from other pharmacologic classes suggests that the transition from well regulated and aberrant drug-taking depends on interactions between behavioral/reinforcement and pharmacological histories. Moreover, given that our findings were generated using outbred Sprague-Dawley rats, it is possible that this high-responder phenotype could reflect a genetic predisposition for the development of high levels of drug-taking. Further studies that explore individual differences will be important to better understand interactions among the genetic, behavioral, and pharmacological determinants of vulnerability of individuals to develop a substance use disorder.