New Drugs, Old Targets: Tweaking the Dopamine System to Treat Psychostimulant Use Disorders

Abstract

The abuse of illicit psychostimulants, such as cocaine and methamphetamine, continues to pose significant health and societal challenges. Despite considerable efforts to develop medications to treat psychostimulant use disorders (PUD), none have proven effective, leaving an underserved patient population and unanswered questions as to what mechanism(s) of action should be targeted for developing pharmacotherapies. As both cocaine and methamphetamine rapidly increase dopamine (DA) levels in mesolimbic brain regions, leading to euphoria that in some can lead to addiction, targets in which this increased dopaminergic tone may be mitigated have been explored. Further, understanding and targeting mechanisms underlying relapse is fundamental to the success of discovering medications that either reduce the reinforcing effects of the drug of abuse and/or decrease the negative reinforcement or withdrawal/negative affect that occurs during abstinence. A focus on atypical inhibitors of the DA transporter (DAT) and partial agonists/antagonists at DA D₃ receptors (D₃R) are described as two promising targets for future drug development.

I. Introduction to psychostimulant use disorders: cocaine and methamphetamine

a. The problem

While the national focus has been centered on the opioid epidemic, deaths related to psychostimulants, such as cocaine and especially methamphetamine, have surged dramatically. According to the Center for Disease Control (CDC), the National Institute on Drug Abuse (NIDA) and various news sources, overdose fatalities caused by psychostimulants have more than doubled over the last five years (1–5). One of the likely culprits of this increase is co-administration of the opioid, fentanyl. The Journal of the American Medical Association reported a close to eight-fold increase of unprescribed fentanyl detected in urine samples positive for methamphetamine and 18-fold increase for those tested positive for cocaine in 2019 (4, 6). The contamination of deadly fentanyl in non-regulated psychostimulants has thus contributed to the increase in numbers of overdose related deaths of these drugs (7–9). Another cause of this health and societal tragedy is, ironically, the increased public awareness of the dangers of using and abusing opioids, which has led chronic opioid users to shift to psychostimulants that are viewed as less harmful (10, 11). Alarmingly, the grim picture painted by the official statistics does not account for the total number of drug-related deaths. While the numbers of deaths due to overdose does serve as a good metric for the lethality of these addictive substances, there are many other means by which they can inflict long-term health damage that can shorten the users’ lifespan (12). For example, cocaine causes severe cardio- and lung-toxicity over time, and chronic users of methamphetamine not only show dramatic change in mood and behavior, but also exhibit psychotic symptoms (12–16). Taking into consideration all the factors such as these, the numbers of deaths related to drug use could easily be double of what has been reported (12). These alarming trends resulting from increased illicit psychostimulant abuse and subsequent mortality emphasizes the need for effective treatment strategies, including FDA approved pharmacotherapies, of which none currently exists.

b. Potential pharmacotherapies under development

Clinically, a combination of an FDA approved medication with psychosocial treatment, termed medication-assisted treatment, has been shown to be more effective in treating substance use disorders than employing monotherapies exclusively (17–20). However, unlike opioid use disorders for which options such as buprenorphine, methadone and naloxone exists, Psychostimulant Use Disorders (PUDs) lack FDA approved pharmacotherapies (20, 21). Efforts in overcoming this setback have yielded a variety of potential options. For example, an anti-cocaine vaccine is currently being evaluated in a Phase I clinical trial. The dAd5GNE vaccine is composed of a disrupted serotype adenovirus with a GNE hapten, a cocaine analogue, coupled to its capsid protein (22). The administered vaccine induces the immune system to generate anti-cocaine antibodies that have been shown to sequester cocaine molecules in the blood and subsequently prevent cocaine from reaching the brain as the antibodies cannot cross the blood brain barrier (BBB), thereby eliminating cocaine’s stimulant effects (22). Interestingly, the cocaine-antibody complex also protects other cocaine sensitive organs from damage, hence lowering the toxic effects of cocaine in the periphery as well (23). Through this strategy, cocaine self-administration behavior and cocaine-induced hyperactivity in mouse, rat and nonhuman primate pre-clinical models were effectively blocked (22–24). While this vaccine appears to provide an effective solution to cocaine use disorder, a drawback is the necessity of a monthly vaccine regimen, with no endpoint provided, thus, a potentially lifelong treatment (25).

In terms of pharmacotherapeutics, disulfiram is a medication that shows promise in treating cocaine use disorder, although it is FDA approved only to treat alcoholism. For its label purpose, disulfiram inhibits aldehyde dehydrogenase, which induces an aversive reaction to alcohol and deters further intake (26). Clinically, disulfiram has been observed to be effective in decreasing cocaine consumption for dependent individuals, likely through inhibition of dopamine (DA) β-hydroxylase, when combined with psychotherapy, although outcomes are superior in males (26–29). A recent report indicated that patients with genetically high dopamine transporter (DAT) expression levels due to the presence of two rs28363170 10-repeat allele polymorphisms in SLC6A3 (DAT1) may benefit even more from disulfiram treatment (30).

Topiramate, a glutamate antagonist and GABA agonist was able to prevent cocaine self-administration and reinstatement in rodents, however, clinical trials showed insignificant improvement in patients who were administered either topiramate or placebo (31). Bupropion, a DAT inhibitor, was not significantly efficacious in treating either cocaine or methamphetamine use disorder (32, 33). Mirtazapine, which affects monoamine transporters, was effective in decreasing methamphetamine use (34, 35). More recently, cholinergic medications such as galantamine, an acetylcholinesterase antagonist, have shown success in treating cocaine use disorder in a small and a larger (120 patient) clinical trial; further investigation is needed (31).

Perhaps more controversially, the use of DA agonists, or substitution therapy, has emerged as a highly promising pharmacological strategy for cocaine use disorder treatment (31, 36, 37). Opioid agonists such as methadone and buprenorphine have shown success in treating opioid use disorders, however, using the same strategy for PUDs has thus far been rejected. Long-acting amphetamines have been efficacious in treating cocaine use disorder in clinical trials, however, due to stringent criteria, the retention rate of participants was poor, thus studies with a larger population are needed (31, 37). Atypical DAT inhibitors and DA D₃ receptor antagonist/partial agonists have also emerged as promising solutions to PUDs and will be discussed in detail in this review.

II. Emergence of atypical DAT inhibitors as pharmacotherapies for PUD

Cocaine and methamphetamine occupy a similar binding pocket in the DAT protein, which corresponds to the substrate (DA) binding site. Cocaine and its analogues are DAT inhibitors that typically stabilize an open outward-facing conformation of the DAT and prevent DA from being recycled (38–41). Hence, the rewarding effects imparted by cocaine stem primarily from the elevated levels of DA resulting from DAT blockade. In contrast, although methamphetamine also serves as a DAT inhibitor, unlike cocaine, methamphetamine is also a DAT substrate. As such, it is taken up into the cell where it blocks the Vesicular Monoamine Transporter (VMAT) releasing DA and reversing DAT from inside the DA cell body, resulting in increased levels of mesolimbic DA that produce stimulation and euphoria that can lead to abuse (42–48).

In the mid 1990’s, the Newman lab reported a series of benztropine analogues that like the parent drug had binding affinities at the DAT that were comparable or higher than cocaine, but these compounds did not have a cocaine-like behavioral profile, in rodents (49, 50). The term “atypical DAT inhibitor” was defined as compounds that did not display a cocaine-like behavioral profile, despite sharing the common mechanism of action as a DAT inhibitor. Benztropine was chosen as a lead molecule due to structural features that were shared with cocaine (tropane ring) and another DAT inhibitor, GBR12909 (diphenyl ether; Fig. 1). Benztropine was also a clinically used drug for the treatment of symptoms associated with Parkinson’s disease, and there were no reports of its abuse, suggesting it may be atypical.

Fig. 1.

Chemical structures of cocaine, benztropine, JHW007, rimcazole, GBR12909, modafinil, JJC8-016, 8-088, 8-089, 8-091, RDS3-094 with binding data from the Newman lab; ^adata from reference (78).

a. Benztropine and rimcazole analogues

Modifications of the benztropine structure at the tropane N-, 2-, 3-, and 6,7-bridgehead were made over the following decade to advance structure activity relationships (SAR) across the monoamine transporters, as well as sigma₁ receptors (51–62). In addition, decreasing binding affinities at off target sites such as muscarinic M₁ and histamine H₁ receptors was an important and achieved objective to be sure that these actions were not playing a role or masking the psychostimulant behavioral effects of these novel analogues. Of note, benztropine has ~100-fold higher affinity for muscarinic M₁ receptors than DAT, which may be related to its lack of abuse potential (63). Importantly, several interesting lead molecules were discovered, with the most studied, JHW007 (Fig. 1), showing no detectable psychostimulant-like behaviors and demonstrating antagonism of behaviors produced by cocaine or methamphetamine across numerous animal models and species (64–69).

At the same time, a series of rimcazole analogues (63, 70–73) was discovered that in addition to binding with moderately high affinity to DAT also interacted with sigma₁ receptors. These compounds also exhibited an atypical DAT inhibitor behavioral profile (74–76), suggesting that a dual DAT/sigma₁ binding profile might be beneficial for a pharmacotherapeutic to treat PUD. This concept was further expanded with selected benztropine analogues (77). Of note, GBR12909 was reported to bind to sigma₁ receptors (78) and thus it was posited that this dual mechanism may be shared by other compounds that had been or might be developed to treat PUDs.

In addition to the benztropine and rimcazole analogues, other tropane based atypical DAT inhibitors emerged during this time (79, 80) as well as numerous series of GBR12909 analogues, although extensive behavioral evaluation of these latter compounds was not reported. Moreover, both benztropine and GBR12909 were evaluated clinically as potential treatments for cocaine abuse. Benztropine failed to significantly affect responses to acute cocaine administration (81) and further evaluation was not reported. GBR12909 (vanoxerine) was advanced into Phase I clinical trials before failing, due to rate-dependent QTc elongation in healthy subjects (82–86). However, a slow release formulation has recently been reported to have beneficial effects in a cocaine abusing patient population (87).

From the outset, we noted significantly different SAR at DAT with the benztropine analogues compared to cocaine and the many analogues that were synthesized by other labs during that time period (63, 88). We hypothesized that although these two classes of DAT inhibitors share the tropane ring structure, they were likely binding to the DAT differently. Furthermore, that difference in binding might, at least in part, result in a behavioral profile that was not cocaine-like but indeed, could block cocaine from the DAT and hence mitigate cocaine’s psychostimulant actions. Through extensive molecular pharmacology with DAT mutants that either had no effect on cocaine’s binding affinity (e.g., Y156F) or preferred an inward facing conformation (e.g., Y335A) that significantly reduced cocaine’s binding affinity, we discovered that in fact, the benztropines appeared to prefer a more occluded, inward facing conformation of the DAT, compared to cocaine and these data were supported with computational studies (89). Moreover, a recent crystal structure confirmed that cocaine binds to an outward facing conformation of Drosophila melanogaster DAT (dDAT) (39), as predicted empirically and through computational models. No crystal structure of DAT with a benztropine analogue has been reported.

In addition, through ex vivo autoradiography studies, it was shown that although these benztropine analogues penetrate the BBB, they are slow to occupy DAT (e.g., JHW 007 requires >4 h to fully occupy DAT), as compared to cocaine, which fully occupies DAT in <30 min (64, 90). These binding kinetic differences could also play a critical role in their behavioral profiles as it has been shown through extensive PET imaging that cocaine’s rapid BBB penetration and occupancy of DAT is related to feeling of “high” in human subjects (91). Unfortunately, none of the benztropine analogues were advanced to clinical trials.

More recently, inspired by its pharmacological profile and early promise in clinical studies of patients with cocaine or methamphetamine use disorders (92–95), the Newman lab has focused on (±)-modafinil, its (R)-enantiomer (96) and analogues thereof. We originally noted that although not entirely atypical, as we have defined it (i.e. DAT inhibitor without a cocaine-like behavioral profile), R-modafinil had some striking pharmacological features that made it an interesting template to undertake SAR studies. Importantly, although it is a DA uptake inhibitor, it does not appear to have significant addictive liability in humans. Nevertheless, some people use R-modafinil for cognitive enhancement, instead of its FDA approved use for narcolepsy and other sleep disorders (97, 98). It is a mild stimulant in both rodents and humans, but is not self-administered in rodents and can block some cocaine-induced behaviors (97, 98).

b. Modafinil analogues

Armed with 1) SAR in the benztropine class of molecules and 2) a hypothesis that DA uptake inhibitors that prefer a more occluded inward facing conformation may have an atypical behavioral profile, we began a synthetic campaign to identify modafinil analogues that had higher DAT affinities than the parent molecule, with improved solubility and lower abuse potential. Our goal was to discover an atypical DAT inhibitor that had the appropriate drug-like properties as well as a behavioral profile that might be developed as a therapeutic for treatment of PUD.

In our first series of modafinil analogues, we discovered that reducing the terminal amide to an amine improved DAT binding affinity (99, 100) and the salt form of these analogues would impart improved solubility as compared to the parent drug. The first lead in this series, JJC8-016 (Fig. 1) demonstrated a promising behavioral profile in numerous rodent models of cocaine abuse. For example, pretreatment with JJC8-016 dose-dependently inhibited cocaine-enhanced locomotion, cocaine self-administration, and cocaine-induced reinstatement of drug-seeking behavior (101). However, subsequent analysis of JJC8-016 suggested that it may be bind to the human ether-à-go-go-related potassium channel (hERG; (102)), which might predict cardiotoxicity, as had already been reported for its analogue, GBR12909. hERG is involved in the repolarization of the cardiac action potential (83, 86, 103–106). Inhibition of this channel with small molecule drugs can lead to delays in repolarization, which can lead to QT prolongation and the lethal cardiac arrhythmia torsade de pointes. The hERG channel is highly promiscuous and has high affinity for molecules with aromatic (e.g., phenyl rings) and positively charged groups (e.g., protonatable amines). Many FDA-approved drugs have been withdrawn from the market due to cardiotoxicity related to hERG inhibition and thus high affinity for hERG is a red flag for medication development (107). This also poses a significant challenge for designing drugs that are BBB penetrant and that also bind with high affinity to G-protein coupled receptors or monoamine transporters. Hence additional chemical modification of this template was undertaken to further improve the drug-like properties of this molecule, resulting in several new leads (108).

In the next series of modafinil analogues, a piperazine linker was incorporated and several interesting analogues emerged, including JJC8-088, JJC8-089 and JJC8-091 (Fig. 1). Although initial behavioral testing in a murine locomotor assay suggested that the highest DAT affinity analogue, JJC8-088, might be a lead candidate for development, subsequent evaluation showed that it was not as effective in rats exposed to long access (6h) methamphetamine as JJC8-091 (102). Further exploration revealed that JJC8-088 was more cocaine-like than originally predicted and appeared to prefer a more open conformation of the DAT than JJC8-091. Subsequently, JJC8-091 became the new lead molecule in this series (21) and continues to undergo development (http://encepheal.com/). In rodent studies, JJC8-091 reduced cocaine self-administration and cocaine-primed reinstatement of cocaine seeking (Fig. 2–A, C; (21)).

Fig. 2.

Effects of the novel atypical DAT inhibitor JJC8-091 and the novel D₃R partial agonist (±)-VK4-40 on cocaine self-administration (0.5 mg/kg, FR2) (A, B) and cocaine-primed reinstatement of drug-seeking behavior (C, D) in rats extinguished from previous cocaine self-administration. *p<0.05 compared to the vehicle control group.

More recently, in order to further explore SAR and improve pharmacokinetics and metabolic stability, a new series of 2,6-dimethylpiperazine analogues was reported (109). By introducing the 2,6-dimethyl substitution on the piperazine ring, some improvement in drug-like properties was realized. Nevertheless, the piperazine ring remains a metabolically labile functional group in this series of molecules and hence a new series of analogues in which the piperazine ring has been replaced with a piperidine-amine or amino-piperidine function (Giancola et al, unpublished data) are currently being evaluated in rodent models of PUD as well as for pharmacokinetics, metabolic stability and off target actions. In addition, Lubec and colleagues have recently reported novel heterocyclic-based modafinil analogues that may also be promising new leads for PUD therapeutics (110, 111).

III. Introduction to DA D₃R partial agonists/antagonists as potential medications to treat PUD

a. A brief history of D₃R partial agonists/antagonists

By blocking DA reuptake at DAT, psychostimulants such as cocaine and methamphetamine effectively prolong dopaminergic activity in the synapse at both pre- and post-synaptic DA receptors. There are five major DA receptor subtypes, classified as excitatory “D₁-like” receptors (including D₁R and D₂R) or inhibitory “D₂-like” receptors (D₂R, D₃R, and D₄R). Relative to other DA receptor subtypes, the D₃R is of particular interest as a medicinal target due to its restricted distribution in the mesolimbic reward system, including the nucleus accumbens (e.g., ventral striatum), which play an integral role in psychostimulant reward and abuse (112–115). Shortly after cloning of the D₃R in 1990 (115), preclinical studies revealed that D₃R-preferring ligands, such as 7-hydroxy-N,N-di-n-propyl-2-aminotetralin (7-OH-DPAT), reduce cocaine intravenous self-administration in rats (116). The DA D₃R received increasing interest as a medicinal target for PUDs following reports of upregulated D₃R levels in cocaine overdose patients in key reward-related regions of the brain such as the nucleus accumbens, caudate/putamen, and the basolateral amygdala (117–119). More recent PET studies using a D₃R-preferring radioligand, [¹¹C](+)PHNO, confirmed that cocaine users exhibit upregulated D₃R binding in the substantia nigra, hypothalamus and amygdala (120, 121).

Early studies on putative D₃R medications for PUDs were limited by poor selectivity and specificity for the D₃R over the D₂R. Recent advances in medicinal chemistry have led to significant improvements in selectivity and specificity for the D₃R (Fig. 3). For example, the D₃R partial agonist, BP897, is ~70 fold more selective for the D₃R > D₂R. BP897 reduced conditioned locomotor activity to both amphetamine and cocaine and modestly decreased lever responding to cocaine-paired cues in rats (122–125), but was ineffective when co-administered with amphetamine or when cocaine was available for self-administration (122, 126). Similarly, SB277,011A, a D₃R antagonist with approximately 100-fold selectivity for the D₃R over the D₂R (127), dose-dependently reduced cocaine-enhanced brain stimulation reward, suppressed cocaine conditioned place preference, and attenuated cocaine-primed reinstatement behaviors in rats (128–130). However, like BP897, SB277,011A could not reduce cocaine intake under an “easy” fixed ratio 1 (FR1) reinforcement schedule (130).

Fig. 3.

Chemical structures of D₃R antagonists and partial agonists studied as potential medications for PUD, including highly selective bitopic D₃R antagonists and partial agonists (CJB090, PG01037, VK4-116 and VK4-40) developed in the Newman lab (138–140). Binding data from Newman lab; ^adata from reference (141).

In addition to limited efficacy, SB277,011A had poor translational utility due to low bioavailability (<2%) and a short half-life (<20 min) in non-human primates (131, 132). Subsequently, PG01037 and its parent molecule, NGB2904 (Fig. 3) were studied as potentially more viable D₃R antagonists with >100-fold selectivity for D₃R over D₂R. In rats, NGB 2904 reduced progressive ratio breakpoints for cocaine, inhibited both cocaine- and cue-primed reinstatement of cocaine seeking, and reduced cocaine-enhanced brain stimulation reward (133). However, the reward-attenuating effects of NGB 2904 could be overcome by higher doses of cocaine, and NGB 2904 was found to enhance amphetamine-induced locomotor activity in wild type (but not D₃R knockout) mice, suggested possible abuse liability (133, 134). Similar to SB277,011A, PG01037 reduced progressive ratio breakpoints for methamphetamine, suppressed cue-primed reinstatement of methamphetamine seeking, and attenuated methamphetamine-enhanced brain stimulation reward in rats, but failed to alter methamphetamine self-administration under an FR2 reinforcement schedule in rats (135). In non-human primates, PG01037 reduced cocaine choice over food, but tolerance to these effects developed after 5 days of administration (136). Moreover, PG01037 was not effective in reducing methamphetamine choice or intake in the same self-administering rhesus monkeys (136, 137).

b. Cardiovascular implications: Not a D₃R class effect

Despite limited efficacy in reducing psychostimulant intake when drugs are available under low-cost (FR1, FR2) reinforcement conditions, the D₃R antagonists described above significantly attenuated psychostimulant seeking under high motivation (progressive ratio) or reinstatement conditions, suggesting that D₃R antagonists have value in attenuating relapse behaviors. However, translational efforts towards moving D₃R antagonists to the clinic were essentially halted following reports of cardiovascular side effects by GSK598,809 (142), a D₃R antagonist previously evaluated in clinical trials for smoking cessation (ClinicalTrials.gov Identifier: NCT01188967). Briefly, GSK598,809 was found to increase blood pressure in dogs fitted with telemetry transmitters, particularly when combined with cocaine. Because the D₃R is expressed in the kidney, GSK598,809’s blood pressor effects were predicted to apply to all D₃R ligands.

Subsequent investigations following the GSK598,809 telemetry study confirmed that SB277,011A similarly increased blood pressure and heart rate in rats (143). However, two structurally unique D₃R antagonists, R-VK4-116 and R-VK4-40, which exhibit higher D₃R affinity and selectivity than previous D₃R ligands (1735-fold and 305-fold more selective for D₃R over D₂R, respectively), did not potentiate the adverse cardiovascular effects caused by oxycodone or cocaine (140, 143). Instead, R-VK4-116 reduced cocaine-enhanced heart rate and blood pressure. R-VK4-40 similarly attenuated cocaine-induced increases in heart rate, and further reduced blood pressure and heart rate when administered alone. The reasons for which R-VK4-116 and R-VK4-40 do not exert the same adverse cardiovascular effects as previous generations of D₃R antagonists are unclear. Differences in ligand affinity, selectivity, unknown off-target effects, or biased activation of different intracellular D₃R signaling pathways in the kidney may all play a role (143). Regardless of the reasons why, these R-VK4-116 and R-VK4-40 results renewed interest in the D₃R as a medication target for PUD, and provided fresh impetus for the continued development of D₃R ligands in the treatment of psychostimulant and other drug use disorders, including opioid use disorder (139, 140, 144).

c. Emerging evidence supporting D₃R partial agonist/antagonist treatments for PUD

Although D₃R antagonists effectively attenuate the motivation to earn psychostimulants and relapse-related behaviors in rodent models, these compounds have been relatively ineffective in reducing cocaine intake when response requirements are low and/or the reinforcing value of cocaine is high, such as under low FR schedules. As a result, additional efforts have focused instead on partial D₃R agonists as treatments for PUDs. Partial agonists can functionally block the effects of drugs of abuse (due to occupation of their receptor target) but elicit partial activation of their receptor targets under abstinence conditions, thereby potentially mitigating withdrawal effects. As such, partial agonists have historically been more effective at maintaining abstinence and reducing relapse rates than antagonist therapies, particularly for opioid and nicotine use disorders (145, 146).

Prior generations of partial D₃R agonists have shown promise in attenuating psychostimulant seeking in rodent models. The D₃R preferring partial agonist, RGH-188 (Cariprazine, Vraylar®), reduced cocaine intake under an FR1 schedule and suppressed cue-induced reinstatement of cocaine seeking in rats (147). However, while the D₃R partial agonists BP897 and RGH-237 also suppress cue-driven cocaine seeking, these compounds were not effective when cocaine or methamphetamine were available for self-administration under FR schedules (122, 126). CJB090, another partial D₃R agonist from the Newman lab, decreased methamphetamine intake under both FR and progressive ratio schedules and was more effective at attenuating psychostimulant reward than the D₃R antagonist PG01037 in rats (148) but showed mixed efficacy in non-human primates (149).

As with D₃R antagonists, prior generations of partial D₃R agonists have suffered from poor selectivity (BP897 and CJB090 are ~60-70 fold more selective for D₃R > D₂R (150–152)) or poor pharmacokinetic profiles (RGH-237 does not readily penetrate the BBB and is eliminated within ~5 hours of administration (141)). However, a recently developed series of new D₃R ligands may circumvent these limitations (139, 140). As described above, R-VK4-40, which was found to attenuate blood pressor effects of cocaine as described above (143), functions as a D₃R antagonist. Interestingly, the racemic (±)-VK4-40 compound, and its S-VK4-40 enantiomer, both function as D₃R partial agonists. (±)-VK4-40 is ~300 fold selective for the D₃R>D₂R and remains at significant levels in the brain up to 8 hours following oral administration in the rat (Jordan et al., under review). Preliminary results indicate that (±)-VK4-40 not only attenuates cocaine self-administration and cocaine-primed reinstatement (Fig. 2 – B, D). In optical brain-stimulation reward model, in which adenosine-associated virus-mediated channelrhodopsin 2 (AAV-ChR2-eYFP) expression is driven by the DAT gene promotor selectively in VTA DA neurons of the midbrain (Fig. 4 – A, B), (±)-VK4-40, R-VK4-40, or S-VK4-40 (alone significantly inhibits optical brain-stimulation reward maintained by selective stimulation of VTA DA neurons in transgenic DAT-cre mice (Fig. 4 – C, D, E), and pretreatment with either of them significantly blocks cocaine-induced increase in brain-stimulation reward (Fig. 4 – F, G, H). These early findings support the efficacy of partial D₃R agonists or antagonists as putative treatments for PUDs, and critically extend the translational utility of this medication class. As such, the VK compounds are currently undergoing further development as new investigational drugs for treatment of substance use disorders.

Fig. 4.

Effects of the racemic (±)-VK4-40 (a D₃R partial agonist) and its enantiomers R-VK4-40 (a D₃R antagonist) and S-VK4-40 (a D₃R partial agonist) on lever responding for brain-stimulation reward maintained by optogenetic stimulation of VTA DA neurons in DAT-Cre mice. A: Schematic of the experimental model, illustrating AAV-ChR2-EGFP was microinjected into the VTA and an optical fiber was implanted into the VTA to stimulate DA neurons in DAT-Cre mice; B: Representative images, showing that AAV-ChR2-EGFP is selectively expressed in VTA tyrosine hydroxylase (TH)-positive DA neurons. C, D, E: Systemic administration of (±)-VK4-40 (C), R-VK4-40 (D), or S-VK4-40 (E) alone inhibits optical brain-stimulation reward and dose-dependently shifts the stimulation frequency-active lever responding curve downward. F, G, H: Pretreatment with (±)-VK4-40 (F), R-VK4-40 (G), or S-VK4-40 (H) dose-dependently attenuated cocaine-enhanced brain-stimulation reward as assessed by the upward or leftward shift of the stimulation-response curve after cocaine administration. * p < 0.05, ** p < 0.01, *** p < 0.001, compared to vehicle control group.

IV. Summary

There has never been a time when the development of novel strategies and pharmacotherapeutics to treat substance use disorders was more urgent. The loss of lives and livelihood due to health challenges associated with addiction, incarceration and stigma is unprecedented and the resurgence of cocaine and methamphetamine use and abuse is staggering. There is no doubt that targeting multiple mechanisms for medication development is required (153–155) and that a single medication will never be a “cure all” for everyone who suffers from PUD. Targeted medications, focused clinical trials, recognition of other mental health co-morbidities and a plethora of other challenges have made it extremely difficult to identify and ultimately provide medications to be a part of a successful treatment strategy for those who suffer from PUD. Identifying subpopulations that are positively affected by experimental treatments is critical, instead of only determining that a large clinical trial failed. Undoubtedly there are at least subpopulations of people who would benefit from the medications that have failed to receive FDA approval for PUD (e.g., modafinil), but we can’t stop there.

Accelerating research and engaging pharmaceutical industry in this challenging area is critical to our success. Academic labs can do formidable basic and preclinical research, identifying new leads and moving them forward through animal models of PUD. But, that is not enough. There is an enormous chasm between preclinical research and actually getting a drug into Phase 1 clinical trials, much less beyond, and this is going to require more than NIH funded research. Partnerships and ongoing communication that build bridges between academia, government, and private industry are required. We have made slow but steady strides toward identifying the two sets of lead molecules described in this review: atypical DAT inhibitors and dopamine D₃R partial agonists/antagonists. But frankly, reducing rats’ psychostimulant seeking is not the point. Until we can get our lead molecules into the clinic and until we can determine which of our animal models is translational to humans, we are spinning our wheels. Our data support moving forward with either or both of these mechanistic targets and doing further medicinal chemistry to find drug candidates that not only have the desired pharmacological profile, but also appropriate pharmacokinetics, metabolic stability and lack of off target actions that may preclude further development due to toxicity (e.g., hERG activity). These are challenging tasks, but not impossible. Indeed, the NIH Helping to End Addiction Long-term (HEAL) initiative has been developed and funded to address the Opioid Crisis and this is enormously important to ultimately curtail the unprecedented morbidity related to prescription opioid abuse that has led to increased heroin and fentanyl addiction. Lessons learned and critical research in this area will undoubtedly feed into other substance use disorders. And, although one medication or a single targeted mechanism of action is not going to mitigate all substance use disorders, there will be overlap. Indeed, the D₃R partial agonists/antagonists are an excellent example and are currently being developed toward the prevention and treatment of opioid use disorder. However, as we have discussed herein, preclinical data support their efficacy for treatment of PUD. More work and resources are required. Certainly, people who suffer from PUD and indeed all substance use disorders are depending on us.