Truth Seeker
New Member
Abstract
Dependence on cocaine is still a main unresolved medical and social concern, and in spite of research efforts, no pharmacological therapy against cocaine dependence is yet available. Recent studies have shown that the endocannabinoid system participates in specific stages and aspects of drug dependence in general, and some of this evidence suggests an involvement of the cannabinoid system in cocaine effects. For example, cocaine administration has been shown to alter brain endocannabinoid levels, and the endocannabinoid system has been involved in long-term modifications of brain processes that might play a role in neuro/behavioral effects of psychostimulant drugs like cocaine. Human studies show that marijuana dependence is frequently associated with cocaine dependence, and that the cannabinoid receptor CNR1 gene polymorphism might be related to cocaine addiction. This article will review the main papers in the field showing how a modulation of different components of the cannabinoid system might interact with some of the neurobiological/behavioral effects of cocaine related to its reinforcing effects, evaluated in preclinical models or in clinical settings. The goal of this review will be to provide insights into the complex picture of cocaine abuse and addiction, and to extrapolate from such endocannabinoid-cocaine interactions useful information to test the therapeutic potential of cannabinoid ligands and endocannabinoid-level enhancers against cocaine dependence for future preclinical/clinical trials.
1. INTRODUCTION
A number of studies suggest that the cannabinoid system interacts with cocaine's effects at different levels in the brain. For example, both CB1 agonists and cocaine selectively increase dopamine neurotransmission in the nucleus accumbens shell (1, 2, 3, 4), a brain area believed to play a pivotal role in the addictive effects of drugs (5). It is also interesting to note that cocaine administration results in increased levels of anandamide, one of the endogenous ligands for cannabinoid receptors in the brain (6, 7), and that blockade of CB1 cannabinoid receptors attenuates a dopamine signal triggered by cocaine in the NAc (8).
The cannabinoid system has also been implicated in brain processes related to potentiation or depression of the strength of synaptic signals (9, 10, 11, 12). These so-called processes of long term potentiation and long-term depression activated by cannabinoid neurotransmission might play a significant role in those neurobiological adaptations which are believed to take an important part in the road leading from drug use to drug abuse and addiction (13, 14, 15). Moreover, cannabinoid receptors are located in brain areas that have been assigned a key role in brain functions linked to the development of drug-dependence (16). Plant-derived and endogenous cannabinoid ligands have been suggested to affect neurotransmitter transporter functions (17, 18, 19, 20, 21, 22), which are also pharmacological targets for cocaine (23, 24). Finally, it is interesting to note that cocaine dependence is often diagnosed in the presence of other drug dependencies (25), and marijuana dependence is second only to alcohol dependence as the most frequent comorbid dependence in cocaine addicts (25, 26).
1.1 Overview on cocaine abuse and addiction
Cocaine abuse and addiction are widespread throughout the world, and these phenomena are driven by the ability of this drug to produce reinforcing effects by interacting with specific sites in the central nervous system (23). Cocaine binds to specific transport sites on the cell membranes of neurons that release dopamine, serotonin or noradrenaline. These transport sites have the physiological function of removing neurotransmitters from the extracellular space after their release from nerve terminals (24). Binding of cocaine to the neurotransmitters' transport sites results in blockade of the transport, increased availability of neurotransmitters for binding to their respective receptors, and ultimately increased neurotransmission (23, 27).
Although cocaine also binds and blocks the transporters for serotonin and noradrenaline (24), blockade of the dopamine transporter (DAT) appears to be the pharmacological target of cocaine mainly related to its reinforcing and other behavioral effects (23, 27). Thus, increased availability of dopamine for dopamine receptors is believed to mediate most of the reinforcing actions of cocaine, leading to its abuse liability (1, 3, 23, 27). However, cocaine abuse and addiction are processes that go well beyond the pharmacological effects of cocaine on dopamine neurotransmission, so that many other factors, genetic and environmental for example, have been suggested to play a role in the complex of actions that lead from cocaine use to its abuse and addiction (28, 29, 30, 31).
There is no specific pharmacological therapy available against cocaine's addictive effects, but researchers in the last decades have made progresses in the understanding of the neurobiology underlying cocaine's actions (29, 30). Today, different aspects of cocaine abuse and addiction in humans can be modeled and studied in experimental animals. These animal models have provided valuable results leading to a better understanding of the different factors contributing to cocaine effects (32). Moreover, the models have been invaluable in the study of newly synthesized compounds acting over different neuronal systems that might become candidates for development as cocaine abuse medication (29, 30).
1.2 Overview on the brain endocannabinoid system
The important discoveries in the last two decades in the cannabinoid field have provided useful information about the physiological functions of the endocannabinoid system in the brain and periphery, and the involvement of this system in the etiology and therapy of many human diseases (33, 34, 35). During the last decade researchers have provided evidence for the existence of at least two cannabinoid receptors, CB1 and CB2 (36, 37, 38, 39) and we have now available different selective synthetic agonists and antagonists for both CB1 and CB2 receptors (40, 41), other than classic, plant-derived cannabinoid agonists like delta-9-THC, the active ingredient in marijuana. Although the existence of novel cannabinoid receptors has been suggested in different studies (42, 43, 44), definitive proof of the existence of these cannabinoid receptors has yet to be shown.
Several different endogenous ligands for CB1 and CB2 receptors have been found, such as anandamide, 2AG, noladin-ether (45, 46, 47, 48). At variance with classical neurotransmitters, these endocannabinoids are not stored in vesicles in nerve endings, but are released "on demand" by specific changes occurring in neurons which promote the function of specific enzymes that produce anandamide or 2-AG from membrane-constituent lipids (49, 50, 51, 52). We now have evidence that endogenous cannabinoids released "on demand" can function as retrograde messengers in the synapse (53). Thus, activation of cannabinoid receptors might change the electric equilibrium of the post-synaptic membranes with important modifications for cell-depolarization, neuronal firing, and release of classic neurotransmitters, which ultimately results in a modification of brain neurotransmission (10, 11).
Some of the neurobiological effects of endocannabinoids have been demonstrated to be long-lasting, leading to the so-called processes of long-term potentiation and long-term depression of the synaptic transmission strength (11, 12). It is interesting to note that these long-term effects are in spite of the very brief half life of endogenous cannabinoid ligands in the extracellular space (49, 54, 55).
Anandamide and 2-AG are taken-up into the cell by a not completely defined and still debated mechanism/s: passive diffusion or facilitated transport across the plasma membrane (55, 56, 57, 58, 59). Anandamide is then metabolized into the cell by an intracellular enzyme, the fatty acid amide hydrolase (FAAH) (60, 61), the blockade of which has been shown to prolong the behavioral/neurobiological effects of anandamide (4, 61, 62, 63). In contrast, 2-AG is metabolized through the mono glyceril lipase enzyme (54). In the last few years several pharmacological tools have been discovered that alter the physiological equilibrium of the endocannabinoid system by acting on the metabolic pathways of the endogenous ligands. The result of this modulation is that blockade of endocannabinoid metabolism increases levels of endocannabinoids in specific brain areas in which there is a preexisting cannabinoid tone or where endogenous cannabinoids are released on demand triggered by specific neuronal signals (34, 61). These highly-specific effects contrast with the effects of exogenously administered cannabinoid agonists, which diffuse widely throughout the brain leading to a generalized and contemporaneous activation of multiple brain areas.
It is believed that in humans, cannabinoid receptor agonists, like delta-9-THC, activate cannabinoid CB1 receptors, producing reinforcing effects which are the basis for marijuana abuse (2, 64). CB1 receptors are also involved in other brain mechanisms and functions associated with pathologies in which a cannabinoid intervention might have a therapeutic value (33, 34, 65, 66). Unfortunately, there is no cannabinoid agonist drug available yet that can selectively produce therapeutic effects without having undesired psychotropic, reinforcement-related effects. However, pharmacological tools are now available that can modulate the activity of endogenous cannabinoid ligands in selected brain areas, thereby having therapeutic value (34) without the harmful reinforcing effects produced by delta-9-THC-like drugs with the widespread action. Indeed, interesting actions of such cannabinoid modulators have been already described in preclinical models of anxiety (61, 67) and depression (68), and preclinical tests so far have also shown negative results in procedures involving models of drug abuse (4, 68, 69).
2. Effects of cannabinoid drugs or endocannabinoid modulators in preclinical models of cocaine's behavioral and reinforcing effects
2.1 Cannabinoid interactions with cocaine-induced stimulation of motor activity and behavioral sensitization
2.1.1 Effects after acute administration of cocaine
Psychostimulants, like cocaine, have the ability to influence neuronal activity, which results in stimulation of behavioral activity. Stimulation of locomotor activity induced by cocaine and other psychostimulants has been extensively studied (see, for example citations 70, and 71). Administration of psychostimulants in rodents induces a dose-dependent increase in behavioral activation, which depends on the activation of selected brain striatal areas (72, 73, 74, 75, 76), and that is accompanied by stimulation of DA neurotransmission (3, 77, 78, 79). Although this behavior is not a direct correlate of the reinforcing effects of cocaine, the fact that it can be obtained by injecting cocaine in the NAc (72, 76, 77), and that NAc DA transmission is involved in different aspects of drug abuse and addiction (5, 80, 81), makes monitoring of locomotor activation an easy preclinical target to test drugs that can interfere with the ability of cocaine to stimulate DA neurotransmission and potentially attenuate cocaine-induced reinforcement (82).
Do cannabinoids alter the acute effects of cocaine on stimulation of motor activities?
Pryor and colleagues, 1978 (83), described the interaction between cocaine and delta-9-THC on motor activity, measured as photo-beam interruptions in rats, as one of mutual antagonism dependent on the combination of doses tested. The antagonism was attenuated but still evident after sub-chronic administration of delta-9-THC. In agreement with these data, Ferrari et al., 1999 (84), showed that HU210, a synthetic agonist for CB1 receptors, dose dependently reduced the locomotor activity induced by cocaine, 15 mg/kg i.p. in rats. However, sub-chronic administration of HU210 was not able to modify cocaine-induced locomotion at variance with the previous report. It should be noted that a cannabinoid antagonist was not available in 1978, and that Ferrari et al., 1999 (84), did not test whether a CB1 receptor antagonist could interfere with HU210 actions on cocaine effects. Because of the lack of direct pharmacological involvement of CB1 receptors in cocaine effects, Przegalinski et al., 2005 (85), tried to address this issue in a more recent report. In that study the authors showed that the CB1 receptor agonist WIN 55,212-2, but not its less active enantiomer WIN 55,212-3, dose dependently (3—6 mg/kg i.p.) reduced cocaine-induced hyper-locomotion, without affecting behavior when administered alone. Rimonabant (SR 141716), a selective antagonist of CB1 receptors (86), did not reverse the effects of WIN 55,212-2 on cocaine-stimulated locomotion, suggesting a non-CB1 mediated mechanism of action for WIN 55,212-2. Moreover, the effect was not attributable to unspecific interactions of the highly lipophilic cannabinoid agonist WIN 55,212-2 with hydrophobic sites of proteins or other cell's membrane lipid constituents, since the less active enantiomer did not show any significant effect on cocaine-induced hyper-locomotion (85). Further, the authors showed that the effects of WIN 55,212-2 resembled those of a serotonin 5HT3 antagonist, based on other previously published evidence (87, 88), thus indicating the possibility of an involvement of 5HT3 receptors in the action of cannabinoid agonists on cocaine-induced hyper-locomotion. However, other than a resemblance of effects, there was no direct pharmacological evidence for 5HT3-receptor involvement in the actions of WIN 55,212-2 on cocaine-induce hyper-locomotion (85). It should also be noted that in the report by Przegalinski et al., 2005 (85), the locomotor stimulant effects of cocaine were reduced to a certain extent, though not significantly, by Rimonabant itself, while in a recent report (8) Rimonabant significantly attenuated cocaine-induced hyperactivity, thus suggesting the possibility that both agonist and antagonists of CB1 receptors can negatively modulate the locomotor stimulant effects of cocaine. Also, Rimonabant antagonized cocaine-induced hyper-locomotion in Gerbils habituated to the test cage, but not in non-habituated animals (89). The authors explained their results in terms of a cannabinoid involvement as a permissive element required in the ability of cocaine to override novelty-stimulus satiation in habituated Gerbils. However, lack of effects of Rimonabant on acute cocaine-induced locomotor stimulation has also been reported in habituated mice and rats (90, 91).
Taken together these experiments suggest that an endocannabinoid tone might not be necessary to mediate the acute motor stimulant effects of cocaine in rodents. However, depending on dose, both cannabinoid CB1 receptor agonists and antagonists might modulate this behavioral stimulation induced by cocaine. Moreover, as shown by Cheer et al., 2007 (8), attenuation of cocaine-induced locomotor activity by a CB1 receptor antagonist has a neurochemical correlate. In that study, Rimonabant pretreatment attenuated dopamine neurotransmission, as measured by "fast-scan-cyclic voltammetry", suggesting that a cannabinoid tone might be involved in both neurochemical and behavioral processes triggered by cocaine (8).
2.1.2. Effects in cocaine-sensitized animals
Psychostimulant sensitization is a phenomenon that has been extensively described in rodents. It appears as an increased effectiveness of the psychostimulant drugs to produce behavioral activation in animals previously exposed to single (92) or repeated (93) administration of the same drug. Cross sensitization between different drugs has also been described (94, 95).
It has been suggested that sensitization induced by drugs of abuse may be due to neuronal adaptations that play a role in the development of addictive behaviors. Behavioral sensitization is easily demonstrated in animals, and it has been hypothesized that this adaptation is involved in the transition from drug use to abuse and addiction (see citations 14, and 15, for review). Thus, behavioral sensitization has been the target of different studies searching for drugs able to impair or reduce the effects of this adaptation induced by sub-chronic/chronic administration of cocaine.
Do cannabinoids play a role in the behavioral sensitization induced by cocaine?
There are only a few reports about cannabinoid interactions with development or expression of cocaine-induced behavioral sensitization. Sensitization and cross sensitization have both been reported for cannabinoid agonists and some drugs abused by humans including psychostimulant drugs (94, 95, 96, 97, 98, 99, 100). However, failure to develop cross sensitization in rats chronically treated with a cannabinoid agonist, CP 55,940, and challenged with cocaine has also been reported (101). In the same study, CP 55,940 failed to enhance the sensitivity to cocaine when the drugs were concurrently administered during the development phase. Also, the development of cocaine sensitization did not appear to be under control of CB1 receptors in rodents in another report (90). In this latter study, mice were treated with 20 mg/kg cocaine for 11 days in their home cages. Three days after the last cocaine injection they habituated to an open field, and then tested with cocaine (10 mg/kg). The authors reported a first set of experiments in which Rimonabant 1 mg/kg administered 30 min before the daily cocaine injection significantly reduced the locomotor sensitization induced by cocaine. Such effects, however, were not replicated in a subsequent set of experiments with increasing doses of cocaine (90).
In genetically modified CB1 receptor knock out mice (102), cocaine (10 mg/kg twice daily for 15 days) induced a behavioral sensitization, tested after one week withdrawal in an open field, which was indistinguishable from that induced in wild-type littermate control mice. These findings once more suggest a lack of involvement of the endocannabinoid system in the development of behavioral sensitization induced by cocaine. On the other hand, the expression of cocaine sensitization has been demonstrated to be impaired by cannabinoid CB1 receptor blockade, as shown by administration of the selective CB1 receptor antagonist Rimonabant together with cocaine five days after withdrawal from cocaine (103). Although this interaction needs to be further investigated, results from the cited reports suggest that even if activation of cannabinoid neurotransmission through CB1 receptors might not be required for development of cocaine-induced sensitization, an endogenous cannabinoid tone might be essential for its expression.
Thus, cannabinoid antagonists might decrease the impact of a "sensitized" response to drug stimuli in subjects during abstinence. This prediction matches the results obtained by DeVries et al., 2001 (104), with administration of the cannabinoid receptor antagonist Rimonabant in cocaine abstinent rats during self-administration reinstatement tests (see below the section: "Effects of cannabinoids on animal models of extinction and reinstatement of cocaine-seeking behavior").
2.2 Cannabinoid interactions with cocaine-induced place conditioning
The place conditioning procedure is a model of drug reinforcement and is based on an individual's behavioral choice after a certain number of pairings of drug with one environment and placebo with another. This procedure, though simple, can give an indication of the pleasurable (place preference) or aversive (place avoidance) feelings/perceptions induced by drugs during the conditioning procedure. The place-conditioning allows the study of both development and expression of aversion or preference for a place associated with a certain drug dose (see citations 105, and 106 for review), with the final test performed with animals in a drug-free state. So, the ability of drugs to interfere or impair the effects of the conditioning-drug could be related to a potential therapeutic value, in terms of ability to reduce or block the pleasurable feelings induced by the conditioning drug, or to attenuate the strength of the association between the conditioning drug and the environmental stimuli.
It has been repeatedly shown that cocaine, just like many other drugs abused by humans, supports place preference at certain doses (106). On the other hand, results from place conditioning studies with cannabinoid agonists and antagonists have been inconsistent (see citation 107 for review), and both preference and aversion induced by cannabinoid CB1 receptor agonists and antagonists have been reported under a variety of different experimental conditions and dose ranges (108, 109, 110, 111, 112, 113).
The involvement of the cannabinoid system in processes related to positive reinforcement induced by cocaine in the place conditioning procedure was evaluated by Chaperon et al., 1998 (91), in rats. The authors showed clear evidence that blockade of CB1 receptors by administration of the selective cannabinoid CB1 receptor antagonist Rimonabant impaired the acquisition of cocaine-conditioned place preference, suggesting that the development of such conditioning was under control of a cannabinoid tone in the brain. In the same paper the authors also showed that blockade of CB1 receptors did not impair the expression of cocaine-conditioned place preference when the antagonist of CB1 receptors, Rimonabant, was administered only once, 30 min before the final test, instead of 30 min before each cocaine injection during the conditioning sessions (91). At variance with this report, results from a study by Martin et al., 2000 (102), showed that development and expression of cocaine-conditioned place preference can be established in genetically modified CB1 receptor knock-out mice. These latter results suggest no involvement of the endogenous cannabinoid systems in brain processes related to the motivational effects induced by cocaine in this procedure in genetically modified mice (102). These contradictory findings (91, 102) might be explained on the basis of species differences, or developmental adaptations in the regulatory processes involved in the mediation of cocaine reinforcing effects in mice genetically deficient in CB1 cannabinoid receptors. It should be noted that Rimonabant might have inverse agonist actions on CB1 receptors (114, 115, 116), or might possess actions over pharmacological targets other than on CB1 receptors, for example on the putative CB3 receptor (42, 65). Martin and colleagues (102) did not cite the article by Chaperon et al., 1998 (91), so we do not have their comments on this issue. We also do not have information about the effects of Rimonabant administration on development and expression of cocaine-induced place preference in wild-type mice, an experiment that would allow a direct comparison with data obtained in CB1 KO mice (102).
2.3 Effects of cannabinoids on the discriminative stimulus effects of cocaine
The drug discrimination procedure has been extensively used in animal and human subjects to evaluate the specific perceptions/feelings produced by drug administration (117). It is likely that these subjective effects play an important role in the reinforcing effects of many drugs abused by humans. Indeed, many of these drugs produce euphoric or pleasurable sensations that are believed to be factors contributing to the abuse of substances (117).
The discriminative stimulus properties of cocaine have been evaluated in many differents studies under different conditions and in different species, including human subjects (117, 118). However, there is a very limited set of data available in which the influence of the cannabinoid system has been evaluated in animals trained to discriminate the subjective effects of cocaine from those of placebo. Delta-9-THC did not substitute for cocaine in pigeons trained to discriminate cocaine from placebo, or cocaine from morphine, suggesting that the discriminative stimulus of the cannabinoid agonist and that of cocaine do not overlap (119, 120). In a more recent report (103) the cannabinoid receptor antagonist Rimonabant did not substitute for cocaine, and did not change significantly the discriminative stimulus effects of cocaine in rats.
These limited findings indicate that the endogenous cannabinoid tone mediated by CB1 receptors does not play a role in the expression of the discriminative stimulus of cocaine. However, more detailed studies with a complete range of cannabinoid drugs (agonist, antagonist, and modulators) and doses are needed to further elucidate the role of the cannabinoid system in the subjective effects of cocaine.
2.4 Cannabinoid modulation of the effects of cocaine on thresholds for Intra Cranial Self Stimulation (ICSS)
This procedure is based on studies by Olds and Milner (121), in which rats have the opportunity to learn to repeatedly press a lever in order to obtain an electrical stimulation of the medial forebrain bundle, a component of reward circuits in the brain (122). It has been repeatedly shown that most drugs of abuse are able to lower ICSS thresholds in a dose-dependent fashion (122, 123). This effect has been used as an indirect measure of their ability to produce "reward" through brain processes linked to the medial forebrain bundle. Thus, substances able to reduce or block the ability of drugs of abuse to lower the ICSS threshold may have therapeutic potential as treatments for drug abuse. While the effects of cocaine in this procedure are consistent with its reinforcing effects in animal and human subjects, inconsistent results have been obtained with administration of cannabinoid drugs (124, 125; see also citations 106, and 126, for review). There is only limited information available for combinations of cocaine and cannabinoids studied with the intracranial self stimulation procedure. In a study by Pradhan et al., 1978 (127), co-administration of delta-9-THC in rats resulted in an antagonism of the effects of cocaine, while in experiments in which delta-9-THC was administered 80 min before cocaine, the effects of cocaine were unaffected by the cannabinoid agonist. In agreement with this study, in a more recent report (128) administration of the synthetic cannabinoid agonist WIN 55,212-2 reduced the effects of cocaine on ICSS thresholds. In the same study, administration of the CB1 receptor antagonist, Rimonabant, antagonized the effects of WIN 55-212,2 on cocaine. Also, Rimonabant administration did not affect cocaine actions on this procedure, though only very low doses of the CB1 antagonist were tested (128).
These above cited reports suggest that cannabinoid agonists might have some potential activity against cocaine reinforcing effects in rodents, however, the mechanism/s for these interactions are not clear. Drugs of abuse and intra cranial self stimulation produce reinforcing effects through the same reward circuitry (122), and also increase dopamine neurotransmission in a dopaminergic reward-related area of such circuitry (1, 2, 3, 78, 129, 130) that has been related to their reinforcing effects (5, 79, 80). Thus, it is not apparently clear why cocaine and cannabinoid CB1 receptor agonists, which share the ability to stimulate the brain reward circuitry, may have different and antagonistic effects in this procedure. However, the fact that in rodents cannabinoid agonists do not consistently produce reinforcing effects, but do so only under certain experimental conditions (108, 124, 131, 132, 133, 134; see also 107, 126, and 135), suggests a cautious approach to the outcome of experiments involving reinforcing effects of cannabinoids in these species as compared to the well known consistent reinforcing effects of marijuana in human subjects.
2.5 Effects of cannabinoids on cocaine-maintained self-administration behavior
Although the most direct outcome of this procedure is to predict reinforcing effects and abuse liability in humans for drugs that maintain self-administration behavior in experimental animals, many other aspects of this behavioral procedure can contribute to a better understanding of the behavioral effects of drugs and drug-related cues, their related neurobiology and potential therapeutic interventions (136, 137).
As a drug that is widely abused, cocaine consistently maintains intravenous self-administration behavior across a range of doses in a variety of different species (see, for example, citations 138, and 139). Although the reinforcing effects obtained by administration of plant derived cannabinoids in humans were described in books thousands of years ago, cannabinoid agonists do not consistently maintain self-administration behavior in experimental animals (see citations 107, 126, and 135 for review).
Preclinical studies have shown that pharmacological alterations of the endocannabinoid receptor system might interfere with the ability of cocaine to maintain self-administration behavior. In Long Evans rats, administration of the cannabinoid agonist WIN 55-212,2 reduced the average number of intravenous self-injections of different doses of cocaine under a fixed-ratio 1 schedule, suggesting a potentiation of cocaine effects (140). In the same report, the selective antagonist of CB1 receptors Rimonabant antagonized the effects of WIN 55-212,2 on cocaine self-administration behavior. Rimonabant, however, did not affect the behavior maintained by cocaine when administered alone (140). In agreement with this report, Rimonabant pretreatment did not show any effect on cocaine self-administration under fixed-ratio schedules of behavior in squirrel monkeys (63), an effect that has been recently replicated and confirmed (141, 142). In addition, Rimonabant did not affect the number of cocaine self-injections in Wistar rats studied under a fixed-ratio 5 schedule of drug delivery (103, 104). In agreement with the previous experiments in rats and primates, pre-treating C57Bl/6J mice with Rimonabant before single one-day sessions of cocaine self-administration affected neither the number of nose-pokes maintained nor the total cocaine intake during the 30-min session (90). In contrast the effects of genetic deletion of CB1 receptors have not been entirely consistent, and can differ with the effects of administering the pharmacological antagonist of CB1 receptors, Rimonabant. For example, using the single one-day cocaine self-administration procedure mutant cannabinoid CB1 receptor knock-out mice failed to self-administer morphine, but were still able to self-administer cocaine, suggesting that the endocannabinoid system was a substrate for mediating the reinforcing effects of opioids but not those of cocaine (143). At variance with these findings, a more recent report, using the same mutant cannabinoid CB1 receptor knock-out mice as subjects, showed a reduced number of mice reaching the acquisition criteria, which was also obtained within a lager number of sessions, and a significantly reduced reinforcing efficacy of cocaine (measured as the effort required to obtain a cocaine infusion) as compared to wild-type littermate control mice (144). The authors explain their results, contrasting with those of Cossu et al., 2001 (143), in terms of different experimental conditions; freely moving mice chronically receiving cocaine self-infusions in daily sessions (144) versus mice with restrained mobility receiving cocaine during a single session (143).
Soria and colleagues (144) confirmed their data in mutant CB1 receptor knock-out mice studying the effects of pretreatments with the antagonist of CB1 receptors Rimonabant in wild-type mice on cocaine self-administration under a progressive-ratio schedule. They found that Rimonabant dose-dependently reduced the breaking point achieved by wild-type mice under the progressive-ratio schedule of cocaine self-administration, suggesting that also in normal control mice the endocannabinoid system was a substrate for the reinforcing effects of cocaine (144). It is interesting to note that in this report the authors excluded any involvement of learning/memory processes in the effects of cocaine in mice lacking the CB1 receptors because the same mice successfully acquired an operant response maintained by natural reinforcers (food and water). Also, the reduced efficacy of cocaine as a reinforcer was not the result of a diminished ability of cocaine to stimulate dopamine neurotransmission, measured by intra-cerebral microdialysis in limbic areas of mutant mice as compared to their littermate controls (144). These latter results point toward a more complicated and central involvement of the endocannabinoid system in the addictive effects of cocaine.
As neither deletion nor blockade of CB1 receptors impairs cocaine's ability to stimulate DA neurotransmission, it seems that the effects of modulation of the endocannabinoid system on cocaine self administration may be indirect. In a recent report using Sprague Dawley rats, prior exposure to delta-9-THC did not affect the acquisition and subsequent rates of cocaine self-administration under a fixed-ratio 1 schedule of injection, but did reduce the number of "cocaine-seeking" responses under a progressive-ratio schedule (145). Thus the sensitivity to delta-9-THC exposure may vary with behavioral procedure. Indeed, it has been already reported that changes in reinforcing efficacy of cocaine are hard to detect using fixed-ratio schedules of cocaine self-administration (146). In the report by Panlilio and colleagues (145), delta-9-THC pre-exposure also decreased both the number of entries in the center-zone of an open-field arena, and the time spent in the lighted compartment of the dark-light box (two different predictors of an anxiety response) in cocaine-treated rats. The authors suggested that the reduced effectiveness of cocaine as a reinforcer after delta-9-THC pre-exposure may be because the previous exposure to the cannabinoid agonist enhanced the aversive/anxiogenic effects of cocaine (145).
The contrasting results between species and schedules of drug delivery together suggest that the endocannabinoid system might impair other brain processes related to cocaine self administration which might in turn reduce the ability to detect the reinforcing effects of cocaine. Certainly, a better understanding and clinical prediction of the influence of the endocannabinoid system on the reinforcing effects of cocaine will come from studies of the effects of cannabinoid agonists and antagonists under a variety of procedures (including progressive-ratio schedules) and with a variety of cannabinoid exposure procedures in primates.
2.6 Effects of cannabinoids on animal models of extinction and reinstatement of cocaine-seeking behavior
One of the major obstacles to successfully achieving withdrawal or abstinence from drug use stems from the difficulty of finding effective strategies against relapse to drug use that is thought to result from craving for the abused substance. Other than behavioral therapy, potential pharmacological support might alleviate craving in abstinent individuals, thus reducing the likelihood of relapse. Although the validity of preclinical models of relapse to drug-seeking and drug-taking behaviors is still under discussion (147, 148, 149, 150), these procedures might give some useful information about which drugs and which brain areas might interfere with the processes that underlie relapse.
In the last years, as it has become increasingly clear that the endocannabinoid system plays a role in different aspects related to substance abuse, researchers have explored the involvement of the cannabinoid system in the neurobiology of relapse to drug-seeking (see citation 151 for review). In rats, when cocaine self-administration behavior was extinguished by substituting saline for cocaine injections (152), experimenter administration of increasing doses of delta-9-THC, 3 hours after withdrawal, failed to induce/reinstate cocaine-seeking behavior. However, administration of HU210, a synthetic cannabinoid agonist which has a higher affinity and greater intrinsic activity than delta-9-THC at CB1 receptors, provoked relapse to cocaine-seeking after a prolonged (14 days) withdrawal from cocaine (104). In addition to differences in treatment, the different outcome of these reports could be due to differences in abstinence strength, because in the first and second reports rats were abstinent from cocaine only for three hours (152) and 14 days (104), respectively. These different conditions might suggest time-related adaptations of the brain systems involved in cocaine-seeking behaviors (153), with no influence of the endocannabinoid system during the early stages of withdrawal from cocaine. On the other hand, when the selective antagonists of cannabinoid CB1 receptors Rimonabant and AM251 were administered alone they did not reinstate cocaine-seeking behavior (104, 154), but both drugs reduced or inhibited cocaine-seeking behavior evoked by priming injections of cocaine (104, 154). It is also interesting to note that Rimonabant reduced cocaine seeking behavior induced by cocaine associated cues, but not stress-induced reinstatement (104), suggesting the involvement of the endocannabinoid system only in some specific processes or pathways related to cocaine-seeking behavior. To this end, it is also interesting to note that AM251 did not block the DA-stimulating effects of cocaine in the NAc, but it attenuated the effects of cocaine on glutamate levels in the NAc, which might be a better neurochemical correlate of cocaine-seeking behavior (154).
3. Human studies
Though not overlapping with the purpose of this review, the "gateway drug" hypothesis of marijuana use is an interesting example of a link or association between cocaine dependence and cannabinoid agonists. Following this hypothesis, marijuana consumption could serve as a gateway drug for subsequent use and abuse of other illicit substances in a progression that can lead towards dependence on heroin, cocaine or other psychostimulants (155, 156, 157, 158). There is still an ample debate about the factors studied and taken into consideration, and the data obtained which support this theory (see for example citations 159, and 160). It is interesting to note, however, that independently from the factor/s that link marijuana use to subsequent abuse of other illegal substances, there are also studies that show a very high prevalence of marijuana use and dependence in cocaine dependent individuals (26, 161, 162, 163). Comorbidity of cocaine dependence with other drug dependencies is also very common (25). Thus the occurrence of cannabis dependence in cocaine dependent subjects might suggest that individual vulnerability for a certain drug can be extended to other drugs (160), but also that common risk factors (genetic, environmental, etc.) might jointly influence the individual drug liability and the links between cannabis and other drug dependencies (163).
With regard to genetic risk factors, it is interesting to note that a human genetic variant of the cannabinoid receptor gene CNR1 has been associated with susceptibility of drug dependence in a non-Hispanic, Caucasian population (164). The association was stronger with intravenous drug use, and was greatest for cocaine, amphetamine, and marijuana dependence. Non-significant association data have been instead obtained in replication of the original findings by Comings and colleagues (165, 166). In a recent paper (31) these contrasting data are discussed as the result of the poor definition of the cannabinoid receptor CNR1 gene's structure, regulation and variation. In the same report, indeed, the authors have shown an improved definition of the human cannabinoid receptor CNR1 gene's locus and variants (31), and it has been also shown that haplotypes toward the 5' end of the cannabinoid receptor CNR1 gene's exons and introns differentiate between non-drug users and drug-abusers of European-American, African-American, and Japanese subjects. Also, a recent report by Ballon and colleagues, 2006 (167), supports the association of the cannabinoid receptor CNR1 gene polymorphism with predisposition to cocaine dependence in an African-Caribbean population.
Another interesting association, from a therapeutic perspective, has been evaluated studying the relationship between marijuana use during cocaine abstinence and relapse to cocaine use. In a study by Labigalini et al., 1999 (168), the authors describe a clinical observation based on spontaneous accounts by Brazilian cocaine-crack abusers, undergoing their first psychiatric assessment, reporting use of cannabis in order to get relief from cocaine-withdrawal symptoms. From a sample of 25 subjects, 68 percent ceased to use crack-cocaine during the study, which lasted 9 months. The small sample and the short period of time suggest a cautious approach to the results outcome; further studies are warranted. A more recent study (169) in New York used a much larger sample of cocaine addicts (144 subjects), and examined whether cannabis use after discharge from inpatient treatment could affect relapse to cocaine use. Data from this study showed that about a third of the subjects used cannabis after discharge and that cannabis use substantially increased the risk of relapse to cocaine use, significantly reduced the achievement of sustained remission, and significantly increased relapse to cocaine use after sustained remission (169). In agreement with this latter report is a preclinical investigation (described above, section 2.6) showing cannabinoid-induced reinstatement of cocaine seeking behavior in rodents (104).
4. Summary and concluding remarks
Most of the scientific articles reviewed in the present manuscript have described studies of cannabinoid CB1 receptor agonists and antagonists tested against cocaine effects in preclinical models thought to be predictive of cocaine abuse. These studies have provided interesting results, especially for the ability of cannabinoid antagonists, and Rimonabant in particular, to significantly counteract some of the reinforcing actions of cocaine (104, 143). Collectively, the studies suggest that a cannabinoid tone, impaired by cannabinoid antagonist administration, is indeed involved in many of the reinforcing effects of cocaine which are believed to be responsible for cocaine abuse and addiction. On the other hand, there are no studies available showing interactions of drugs acting as cannabinoid levels modulators/enhancers on cocaine-induced behaviors. These drugs affecting directly the endogenous cannabinoid tone could interfere with cocaine effects in these preclinical procedures and could substantially increase our knowledge about the cannabinoid-neurobiology related to cocaine dependence.
Suggestions about possible genetic predisposition/vulnerability to cocaine dependence from human studies due to variants of the cannabinoid receptor CNR1 gene have given more strength to the link between endocannabinoids and cocaine. Due to the widespread distribution of cannabinoid receptors in the brain, and their abundance in brain areas playing pivotal roles in drug abuse and addiction, the different expression and regulation of cannabinoid receptors induced by genetic differences (31) might be an important factor in the predisposition or vulnerability to drug dependence. For this reason, the potential to directly interact with endocannabinoid tone in selected brain areas, an effect that can be obtained with endocannabinoid uptake inhibitors or metabolism blockers (4, 60, 62)(as shown also in genetically modified mice, 170), as compared to widespread actions of cannabinoid receptors agonists/antagonists, should be one of the next challenges in the research for medications able to counteract the abuse- and dependence-related behavioral/neurobiological effects of cocaine.
Source, Graphs and Figures: Modulation Of The Endo-Cannabinoid System: Therapeutic Potential Against Cocaine Dependence
Dependence on cocaine is still a main unresolved medical and social concern, and in spite of research efforts, no pharmacological therapy against cocaine dependence is yet available. Recent studies have shown that the endocannabinoid system participates in specific stages and aspects of drug dependence in general, and some of this evidence suggests an involvement of the cannabinoid system in cocaine effects. For example, cocaine administration has been shown to alter brain endocannabinoid levels, and the endocannabinoid system has been involved in long-term modifications of brain processes that might play a role in neuro/behavioral effects of psychostimulant drugs like cocaine. Human studies show that marijuana dependence is frequently associated with cocaine dependence, and that the cannabinoid receptor CNR1 gene polymorphism might be related to cocaine addiction. This article will review the main papers in the field showing how a modulation of different components of the cannabinoid system might interact with some of the neurobiological/behavioral effects of cocaine related to its reinforcing effects, evaluated in preclinical models or in clinical settings. The goal of this review will be to provide insights into the complex picture of cocaine abuse and addiction, and to extrapolate from such endocannabinoid-cocaine interactions useful information to test the therapeutic potential of cannabinoid ligands and endocannabinoid-level enhancers against cocaine dependence for future preclinical/clinical trials.
1. INTRODUCTION
A number of studies suggest that the cannabinoid system interacts with cocaine's effects at different levels in the brain. For example, both CB1 agonists and cocaine selectively increase dopamine neurotransmission in the nucleus accumbens shell (1, 2, 3, 4), a brain area believed to play a pivotal role in the addictive effects of drugs (5). It is also interesting to note that cocaine administration results in increased levels of anandamide, one of the endogenous ligands for cannabinoid receptors in the brain (6, 7), and that blockade of CB1 cannabinoid receptors attenuates a dopamine signal triggered by cocaine in the NAc (8).
The cannabinoid system has also been implicated in brain processes related to potentiation or depression of the strength of synaptic signals (9, 10, 11, 12). These so-called processes of long term potentiation and long-term depression activated by cannabinoid neurotransmission might play a significant role in those neurobiological adaptations which are believed to take an important part in the road leading from drug use to drug abuse and addiction (13, 14, 15). Moreover, cannabinoid receptors are located in brain areas that have been assigned a key role in brain functions linked to the development of drug-dependence (16). Plant-derived and endogenous cannabinoid ligands have been suggested to affect neurotransmitter transporter functions (17, 18, 19, 20, 21, 22), which are also pharmacological targets for cocaine (23, 24). Finally, it is interesting to note that cocaine dependence is often diagnosed in the presence of other drug dependencies (25), and marijuana dependence is second only to alcohol dependence as the most frequent comorbid dependence in cocaine addicts (25, 26).
1.1 Overview on cocaine abuse and addiction
Cocaine abuse and addiction are widespread throughout the world, and these phenomena are driven by the ability of this drug to produce reinforcing effects by interacting with specific sites in the central nervous system (23). Cocaine binds to specific transport sites on the cell membranes of neurons that release dopamine, serotonin or noradrenaline. These transport sites have the physiological function of removing neurotransmitters from the extracellular space after their release from nerve terminals (24). Binding of cocaine to the neurotransmitters' transport sites results in blockade of the transport, increased availability of neurotransmitters for binding to their respective receptors, and ultimately increased neurotransmission (23, 27).
Although cocaine also binds and blocks the transporters for serotonin and noradrenaline (24), blockade of the dopamine transporter (DAT) appears to be the pharmacological target of cocaine mainly related to its reinforcing and other behavioral effects (23, 27). Thus, increased availability of dopamine for dopamine receptors is believed to mediate most of the reinforcing actions of cocaine, leading to its abuse liability (1, 3, 23, 27). However, cocaine abuse and addiction are processes that go well beyond the pharmacological effects of cocaine on dopamine neurotransmission, so that many other factors, genetic and environmental for example, have been suggested to play a role in the complex of actions that lead from cocaine use to its abuse and addiction (28, 29, 30, 31).
There is no specific pharmacological therapy available against cocaine's addictive effects, but researchers in the last decades have made progresses in the understanding of the neurobiology underlying cocaine's actions (29, 30). Today, different aspects of cocaine abuse and addiction in humans can be modeled and studied in experimental animals. These animal models have provided valuable results leading to a better understanding of the different factors contributing to cocaine effects (32). Moreover, the models have been invaluable in the study of newly synthesized compounds acting over different neuronal systems that might become candidates for development as cocaine abuse medication (29, 30).
1.2 Overview on the brain endocannabinoid system
The important discoveries in the last two decades in the cannabinoid field have provided useful information about the physiological functions of the endocannabinoid system in the brain and periphery, and the involvement of this system in the etiology and therapy of many human diseases (33, 34, 35). During the last decade researchers have provided evidence for the existence of at least two cannabinoid receptors, CB1 and CB2 (36, 37, 38, 39) and we have now available different selective synthetic agonists and antagonists for both CB1 and CB2 receptors (40, 41), other than classic, plant-derived cannabinoid agonists like delta-9-THC, the active ingredient in marijuana. Although the existence of novel cannabinoid receptors has been suggested in different studies (42, 43, 44), definitive proof of the existence of these cannabinoid receptors has yet to be shown.
Several different endogenous ligands for CB1 and CB2 receptors have been found, such as anandamide, 2AG, noladin-ether (45, 46, 47, 48). At variance with classical neurotransmitters, these endocannabinoids are not stored in vesicles in nerve endings, but are released "on demand" by specific changes occurring in neurons which promote the function of specific enzymes that produce anandamide or 2-AG from membrane-constituent lipids (49, 50, 51, 52). We now have evidence that endogenous cannabinoids released "on demand" can function as retrograde messengers in the synapse (53). Thus, activation of cannabinoid receptors might change the electric equilibrium of the post-synaptic membranes with important modifications for cell-depolarization, neuronal firing, and release of classic neurotransmitters, which ultimately results in a modification of brain neurotransmission (10, 11).
Some of the neurobiological effects of endocannabinoids have been demonstrated to be long-lasting, leading to the so-called processes of long-term potentiation and long-term depression of the synaptic transmission strength (11, 12). It is interesting to note that these long-term effects are in spite of the very brief half life of endogenous cannabinoid ligands in the extracellular space (49, 54, 55).
Anandamide and 2-AG are taken-up into the cell by a not completely defined and still debated mechanism/s: passive diffusion or facilitated transport across the plasma membrane (55, 56, 57, 58, 59). Anandamide is then metabolized into the cell by an intracellular enzyme, the fatty acid amide hydrolase (FAAH) (60, 61), the blockade of which has been shown to prolong the behavioral/neurobiological effects of anandamide (4, 61, 62, 63). In contrast, 2-AG is metabolized through the mono glyceril lipase enzyme (54). In the last few years several pharmacological tools have been discovered that alter the physiological equilibrium of the endocannabinoid system by acting on the metabolic pathways of the endogenous ligands. The result of this modulation is that blockade of endocannabinoid metabolism increases levels of endocannabinoids in specific brain areas in which there is a preexisting cannabinoid tone or where endogenous cannabinoids are released on demand triggered by specific neuronal signals (34, 61). These highly-specific effects contrast with the effects of exogenously administered cannabinoid agonists, which diffuse widely throughout the brain leading to a generalized and contemporaneous activation of multiple brain areas.
It is believed that in humans, cannabinoid receptor agonists, like delta-9-THC, activate cannabinoid CB1 receptors, producing reinforcing effects which are the basis for marijuana abuse (2, 64). CB1 receptors are also involved in other brain mechanisms and functions associated with pathologies in which a cannabinoid intervention might have a therapeutic value (33, 34, 65, 66). Unfortunately, there is no cannabinoid agonist drug available yet that can selectively produce therapeutic effects without having undesired psychotropic, reinforcement-related effects. However, pharmacological tools are now available that can modulate the activity of endogenous cannabinoid ligands in selected brain areas, thereby having therapeutic value (34) without the harmful reinforcing effects produced by delta-9-THC-like drugs with the widespread action. Indeed, interesting actions of such cannabinoid modulators have been already described in preclinical models of anxiety (61, 67) and depression (68), and preclinical tests so far have also shown negative results in procedures involving models of drug abuse (4, 68, 69).
2. Effects of cannabinoid drugs or endocannabinoid modulators in preclinical models of cocaine's behavioral and reinforcing effects
2.1 Cannabinoid interactions with cocaine-induced stimulation of motor activity and behavioral sensitization
2.1.1 Effects after acute administration of cocaine
Psychostimulants, like cocaine, have the ability to influence neuronal activity, which results in stimulation of behavioral activity. Stimulation of locomotor activity induced by cocaine and other psychostimulants has been extensively studied (see, for example citations 70, and 71). Administration of psychostimulants in rodents induces a dose-dependent increase in behavioral activation, which depends on the activation of selected brain striatal areas (72, 73, 74, 75, 76), and that is accompanied by stimulation of DA neurotransmission (3, 77, 78, 79). Although this behavior is not a direct correlate of the reinforcing effects of cocaine, the fact that it can be obtained by injecting cocaine in the NAc (72, 76, 77), and that NAc DA transmission is involved in different aspects of drug abuse and addiction (5, 80, 81), makes monitoring of locomotor activation an easy preclinical target to test drugs that can interfere with the ability of cocaine to stimulate DA neurotransmission and potentially attenuate cocaine-induced reinforcement (82).
Do cannabinoids alter the acute effects of cocaine on stimulation of motor activities?
Pryor and colleagues, 1978 (83), described the interaction between cocaine and delta-9-THC on motor activity, measured as photo-beam interruptions in rats, as one of mutual antagonism dependent on the combination of doses tested. The antagonism was attenuated but still evident after sub-chronic administration of delta-9-THC. In agreement with these data, Ferrari et al., 1999 (84), showed that HU210, a synthetic agonist for CB1 receptors, dose dependently reduced the locomotor activity induced by cocaine, 15 mg/kg i.p. in rats. However, sub-chronic administration of HU210 was not able to modify cocaine-induced locomotion at variance with the previous report. It should be noted that a cannabinoid antagonist was not available in 1978, and that Ferrari et al., 1999 (84), did not test whether a CB1 receptor antagonist could interfere with HU210 actions on cocaine effects. Because of the lack of direct pharmacological involvement of CB1 receptors in cocaine effects, Przegalinski et al., 2005 (85), tried to address this issue in a more recent report. In that study the authors showed that the CB1 receptor agonist WIN 55,212-2, but not its less active enantiomer WIN 55,212-3, dose dependently (3—6 mg/kg i.p.) reduced cocaine-induced hyper-locomotion, without affecting behavior when administered alone. Rimonabant (SR 141716), a selective antagonist of CB1 receptors (86), did not reverse the effects of WIN 55,212-2 on cocaine-stimulated locomotion, suggesting a non-CB1 mediated mechanism of action for WIN 55,212-2. Moreover, the effect was not attributable to unspecific interactions of the highly lipophilic cannabinoid agonist WIN 55,212-2 with hydrophobic sites of proteins or other cell's membrane lipid constituents, since the less active enantiomer did not show any significant effect on cocaine-induced hyper-locomotion (85). Further, the authors showed that the effects of WIN 55,212-2 resembled those of a serotonin 5HT3 antagonist, based on other previously published evidence (87, 88), thus indicating the possibility of an involvement of 5HT3 receptors in the action of cannabinoid agonists on cocaine-induced hyper-locomotion. However, other than a resemblance of effects, there was no direct pharmacological evidence for 5HT3-receptor involvement in the actions of WIN 55,212-2 on cocaine-induce hyper-locomotion (85). It should also be noted that in the report by Przegalinski et al., 2005 (85), the locomotor stimulant effects of cocaine were reduced to a certain extent, though not significantly, by Rimonabant itself, while in a recent report (8) Rimonabant significantly attenuated cocaine-induced hyperactivity, thus suggesting the possibility that both agonist and antagonists of CB1 receptors can negatively modulate the locomotor stimulant effects of cocaine. Also, Rimonabant antagonized cocaine-induced hyper-locomotion in Gerbils habituated to the test cage, but not in non-habituated animals (89). The authors explained their results in terms of a cannabinoid involvement as a permissive element required in the ability of cocaine to override novelty-stimulus satiation in habituated Gerbils. However, lack of effects of Rimonabant on acute cocaine-induced locomotor stimulation has also been reported in habituated mice and rats (90, 91).
Taken together these experiments suggest that an endocannabinoid tone might not be necessary to mediate the acute motor stimulant effects of cocaine in rodents. However, depending on dose, both cannabinoid CB1 receptor agonists and antagonists might modulate this behavioral stimulation induced by cocaine. Moreover, as shown by Cheer et al., 2007 (8), attenuation of cocaine-induced locomotor activity by a CB1 receptor antagonist has a neurochemical correlate. In that study, Rimonabant pretreatment attenuated dopamine neurotransmission, as measured by "fast-scan-cyclic voltammetry", suggesting that a cannabinoid tone might be involved in both neurochemical and behavioral processes triggered by cocaine (8).
2.1.2. Effects in cocaine-sensitized animals
Psychostimulant sensitization is a phenomenon that has been extensively described in rodents. It appears as an increased effectiveness of the psychostimulant drugs to produce behavioral activation in animals previously exposed to single (92) or repeated (93) administration of the same drug. Cross sensitization between different drugs has also been described (94, 95).
It has been suggested that sensitization induced by drugs of abuse may be due to neuronal adaptations that play a role in the development of addictive behaviors. Behavioral sensitization is easily demonstrated in animals, and it has been hypothesized that this adaptation is involved in the transition from drug use to abuse and addiction (see citations 14, and 15, for review). Thus, behavioral sensitization has been the target of different studies searching for drugs able to impair or reduce the effects of this adaptation induced by sub-chronic/chronic administration of cocaine.
Do cannabinoids play a role in the behavioral sensitization induced by cocaine?
There are only a few reports about cannabinoid interactions with development or expression of cocaine-induced behavioral sensitization. Sensitization and cross sensitization have both been reported for cannabinoid agonists and some drugs abused by humans including psychostimulant drugs (94, 95, 96, 97, 98, 99, 100). However, failure to develop cross sensitization in rats chronically treated with a cannabinoid agonist, CP 55,940, and challenged with cocaine has also been reported (101). In the same study, CP 55,940 failed to enhance the sensitivity to cocaine when the drugs were concurrently administered during the development phase. Also, the development of cocaine sensitization did not appear to be under control of CB1 receptors in rodents in another report (90). In this latter study, mice were treated with 20 mg/kg cocaine for 11 days in their home cages. Three days after the last cocaine injection they habituated to an open field, and then tested with cocaine (10 mg/kg). The authors reported a first set of experiments in which Rimonabant 1 mg/kg administered 30 min before the daily cocaine injection significantly reduced the locomotor sensitization induced by cocaine. Such effects, however, were not replicated in a subsequent set of experiments with increasing doses of cocaine (90).
In genetically modified CB1 receptor knock out mice (102), cocaine (10 mg/kg twice daily for 15 days) induced a behavioral sensitization, tested after one week withdrawal in an open field, which was indistinguishable from that induced in wild-type littermate control mice. These findings once more suggest a lack of involvement of the endocannabinoid system in the development of behavioral sensitization induced by cocaine. On the other hand, the expression of cocaine sensitization has been demonstrated to be impaired by cannabinoid CB1 receptor blockade, as shown by administration of the selective CB1 receptor antagonist Rimonabant together with cocaine five days after withdrawal from cocaine (103). Although this interaction needs to be further investigated, results from the cited reports suggest that even if activation of cannabinoid neurotransmission through CB1 receptors might not be required for development of cocaine-induced sensitization, an endogenous cannabinoid tone might be essential for its expression.
Thus, cannabinoid antagonists might decrease the impact of a "sensitized" response to drug stimuli in subjects during abstinence. This prediction matches the results obtained by DeVries et al., 2001 (104), with administration of the cannabinoid receptor antagonist Rimonabant in cocaine abstinent rats during self-administration reinstatement tests (see below the section: "Effects of cannabinoids on animal models of extinction and reinstatement of cocaine-seeking behavior").
2.2 Cannabinoid interactions with cocaine-induced place conditioning
The place conditioning procedure is a model of drug reinforcement and is based on an individual's behavioral choice after a certain number of pairings of drug with one environment and placebo with another. This procedure, though simple, can give an indication of the pleasurable (place preference) or aversive (place avoidance) feelings/perceptions induced by drugs during the conditioning procedure. The place-conditioning allows the study of both development and expression of aversion or preference for a place associated with a certain drug dose (see citations 105, and 106 for review), with the final test performed with animals in a drug-free state. So, the ability of drugs to interfere or impair the effects of the conditioning-drug could be related to a potential therapeutic value, in terms of ability to reduce or block the pleasurable feelings induced by the conditioning drug, or to attenuate the strength of the association between the conditioning drug and the environmental stimuli.
It has been repeatedly shown that cocaine, just like many other drugs abused by humans, supports place preference at certain doses (106). On the other hand, results from place conditioning studies with cannabinoid agonists and antagonists have been inconsistent (see citation 107 for review), and both preference and aversion induced by cannabinoid CB1 receptor agonists and antagonists have been reported under a variety of different experimental conditions and dose ranges (108, 109, 110, 111, 112, 113).
The involvement of the cannabinoid system in processes related to positive reinforcement induced by cocaine in the place conditioning procedure was evaluated by Chaperon et al., 1998 (91), in rats. The authors showed clear evidence that blockade of CB1 receptors by administration of the selective cannabinoid CB1 receptor antagonist Rimonabant impaired the acquisition of cocaine-conditioned place preference, suggesting that the development of such conditioning was under control of a cannabinoid tone in the brain. In the same paper the authors also showed that blockade of CB1 receptors did not impair the expression of cocaine-conditioned place preference when the antagonist of CB1 receptors, Rimonabant, was administered only once, 30 min before the final test, instead of 30 min before each cocaine injection during the conditioning sessions (91). At variance with this report, results from a study by Martin et al., 2000 (102), showed that development and expression of cocaine-conditioned place preference can be established in genetically modified CB1 receptor knock-out mice. These latter results suggest no involvement of the endogenous cannabinoid systems in brain processes related to the motivational effects induced by cocaine in this procedure in genetically modified mice (102). These contradictory findings (91, 102) might be explained on the basis of species differences, or developmental adaptations in the regulatory processes involved in the mediation of cocaine reinforcing effects in mice genetically deficient in CB1 cannabinoid receptors. It should be noted that Rimonabant might have inverse agonist actions on CB1 receptors (114, 115, 116), or might possess actions over pharmacological targets other than on CB1 receptors, for example on the putative CB3 receptor (42, 65). Martin and colleagues (102) did not cite the article by Chaperon et al., 1998 (91), so we do not have their comments on this issue. We also do not have information about the effects of Rimonabant administration on development and expression of cocaine-induced place preference in wild-type mice, an experiment that would allow a direct comparison with data obtained in CB1 KO mice (102).
2.3 Effects of cannabinoids on the discriminative stimulus effects of cocaine
The drug discrimination procedure has been extensively used in animal and human subjects to evaluate the specific perceptions/feelings produced by drug administration (117). It is likely that these subjective effects play an important role in the reinforcing effects of many drugs abused by humans. Indeed, many of these drugs produce euphoric or pleasurable sensations that are believed to be factors contributing to the abuse of substances (117).
The discriminative stimulus properties of cocaine have been evaluated in many differents studies under different conditions and in different species, including human subjects (117, 118). However, there is a very limited set of data available in which the influence of the cannabinoid system has been evaluated in animals trained to discriminate the subjective effects of cocaine from those of placebo. Delta-9-THC did not substitute for cocaine in pigeons trained to discriminate cocaine from placebo, or cocaine from morphine, suggesting that the discriminative stimulus of the cannabinoid agonist and that of cocaine do not overlap (119, 120). In a more recent report (103) the cannabinoid receptor antagonist Rimonabant did not substitute for cocaine, and did not change significantly the discriminative stimulus effects of cocaine in rats.
These limited findings indicate that the endogenous cannabinoid tone mediated by CB1 receptors does not play a role in the expression of the discriminative stimulus of cocaine. However, more detailed studies with a complete range of cannabinoid drugs (agonist, antagonist, and modulators) and doses are needed to further elucidate the role of the cannabinoid system in the subjective effects of cocaine.
2.4 Cannabinoid modulation of the effects of cocaine on thresholds for Intra Cranial Self Stimulation (ICSS)
This procedure is based on studies by Olds and Milner (121), in which rats have the opportunity to learn to repeatedly press a lever in order to obtain an electrical stimulation of the medial forebrain bundle, a component of reward circuits in the brain (122). It has been repeatedly shown that most drugs of abuse are able to lower ICSS thresholds in a dose-dependent fashion (122, 123). This effect has been used as an indirect measure of their ability to produce "reward" through brain processes linked to the medial forebrain bundle. Thus, substances able to reduce or block the ability of drugs of abuse to lower the ICSS threshold may have therapeutic potential as treatments for drug abuse. While the effects of cocaine in this procedure are consistent with its reinforcing effects in animal and human subjects, inconsistent results have been obtained with administration of cannabinoid drugs (124, 125; see also citations 106, and 126, for review). There is only limited information available for combinations of cocaine and cannabinoids studied with the intracranial self stimulation procedure. In a study by Pradhan et al., 1978 (127), co-administration of delta-9-THC in rats resulted in an antagonism of the effects of cocaine, while in experiments in which delta-9-THC was administered 80 min before cocaine, the effects of cocaine were unaffected by the cannabinoid agonist. In agreement with this study, in a more recent report (128) administration of the synthetic cannabinoid agonist WIN 55,212-2 reduced the effects of cocaine on ICSS thresholds. In the same study, administration of the CB1 receptor antagonist, Rimonabant, antagonized the effects of WIN 55-212,2 on cocaine. Also, Rimonabant administration did not affect cocaine actions on this procedure, though only very low doses of the CB1 antagonist were tested (128).
These above cited reports suggest that cannabinoid agonists might have some potential activity against cocaine reinforcing effects in rodents, however, the mechanism/s for these interactions are not clear. Drugs of abuse and intra cranial self stimulation produce reinforcing effects through the same reward circuitry (122), and also increase dopamine neurotransmission in a dopaminergic reward-related area of such circuitry (1, 2, 3, 78, 129, 130) that has been related to their reinforcing effects (5, 79, 80). Thus, it is not apparently clear why cocaine and cannabinoid CB1 receptor agonists, which share the ability to stimulate the brain reward circuitry, may have different and antagonistic effects in this procedure. However, the fact that in rodents cannabinoid agonists do not consistently produce reinforcing effects, but do so only under certain experimental conditions (108, 124, 131, 132, 133, 134; see also 107, 126, and 135), suggests a cautious approach to the outcome of experiments involving reinforcing effects of cannabinoids in these species as compared to the well known consistent reinforcing effects of marijuana in human subjects.
2.5 Effects of cannabinoids on cocaine-maintained self-administration behavior
Although the most direct outcome of this procedure is to predict reinforcing effects and abuse liability in humans for drugs that maintain self-administration behavior in experimental animals, many other aspects of this behavioral procedure can contribute to a better understanding of the behavioral effects of drugs and drug-related cues, their related neurobiology and potential therapeutic interventions (136, 137).
As a drug that is widely abused, cocaine consistently maintains intravenous self-administration behavior across a range of doses in a variety of different species (see, for example, citations 138, and 139). Although the reinforcing effects obtained by administration of plant derived cannabinoids in humans were described in books thousands of years ago, cannabinoid agonists do not consistently maintain self-administration behavior in experimental animals (see citations 107, 126, and 135 for review).
Preclinical studies have shown that pharmacological alterations of the endocannabinoid receptor system might interfere with the ability of cocaine to maintain self-administration behavior. In Long Evans rats, administration of the cannabinoid agonist WIN 55-212,2 reduced the average number of intravenous self-injections of different doses of cocaine under a fixed-ratio 1 schedule, suggesting a potentiation of cocaine effects (140). In the same report, the selective antagonist of CB1 receptors Rimonabant antagonized the effects of WIN 55-212,2 on cocaine self-administration behavior. Rimonabant, however, did not affect the behavior maintained by cocaine when administered alone (140). In agreement with this report, Rimonabant pretreatment did not show any effect on cocaine self-administration under fixed-ratio schedules of behavior in squirrel monkeys (63), an effect that has been recently replicated and confirmed (141, 142). In addition, Rimonabant did not affect the number of cocaine self-injections in Wistar rats studied under a fixed-ratio 5 schedule of drug delivery (103, 104). In agreement with the previous experiments in rats and primates, pre-treating C57Bl/6J mice with Rimonabant before single one-day sessions of cocaine self-administration affected neither the number of nose-pokes maintained nor the total cocaine intake during the 30-min session (90). In contrast the effects of genetic deletion of CB1 receptors have not been entirely consistent, and can differ with the effects of administering the pharmacological antagonist of CB1 receptors, Rimonabant. For example, using the single one-day cocaine self-administration procedure mutant cannabinoid CB1 receptor knock-out mice failed to self-administer morphine, but were still able to self-administer cocaine, suggesting that the endocannabinoid system was a substrate for mediating the reinforcing effects of opioids but not those of cocaine (143). At variance with these findings, a more recent report, using the same mutant cannabinoid CB1 receptor knock-out mice as subjects, showed a reduced number of mice reaching the acquisition criteria, which was also obtained within a lager number of sessions, and a significantly reduced reinforcing efficacy of cocaine (measured as the effort required to obtain a cocaine infusion) as compared to wild-type littermate control mice (144). The authors explain their results, contrasting with those of Cossu et al., 2001 (143), in terms of different experimental conditions; freely moving mice chronically receiving cocaine self-infusions in daily sessions (144) versus mice with restrained mobility receiving cocaine during a single session (143).
Soria and colleagues (144) confirmed their data in mutant CB1 receptor knock-out mice studying the effects of pretreatments with the antagonist of CB1 receptors Rimonabant in wild-type mice on cocaine self-administration under a progressive-ratio schedule. They found that Rimonabant dose-dependently reduced the breaking point achieved by wild-type mice under the progressive-ratio schedule of cocaine self-administration, suggesting that also in normal control mice the endocannabinoid system was a substrate for the reinforcing effects of cocaine (144). It is interesting to note that in this report the authors excluded any involvement of learning/memory processes in the effects of cocaine in mice lacking the CB1 receptors because the same mice successfully acquired an operant response maintained by natural reinforcers (food and water). Also, the reduced efficacy of cocaine as a reinforcer was not the result of a diminished ability of cocaine to stimulate dopamine neurotransmission, measured by intra-cerebral microdialysis in limbic areas of mutant mice as compared to their littermate controls (144). These latter results point toward a more complicated and central involvement of the endocannabinoid system in the addictive effects of cocaine.
As neither deletion nor blockade of CB1 receptors impairs cocaine's ability to stimulate DA neurotransmission, it seems that the effects of modulation of the endocannabinoid system on cocaine self administration may be indirect. In a recent report using Sprague Dawley rats, prior exposure to delta-9-THC did not affect the acquisition and subsequent rates of cocaine self-administration under a fixed-ratio 1 schedule of injection, but did reduce the number of "cocaine-seeking" responses under a progressive-ratio schedule (145). Thus the sensitivity to delta-9-THC exposure may vary with behavioral procedure. Indeed, it has been already reported that changes in reinforcing efficacy of cocaine are hard to detect using fixed-ratio schedules of cocaine self-administration (146). In the report by Panlilio and colleagues (145), delta-9-THC pre-exposure also decreased both the number of entries in the center-zone of an open-field arena, and the time spent in the lighted compartment of the dark-light box (two different predictors of an anxiety response) in cocaine-treated rats. The authors suggested that the reduced effectiveness of cocaine as a reinforcer after delta-9-THC pre-exposure may be because the previous exposure to the cannabinoid agonist enhanced the aversive/anxiogenic effects of cocaine (145).
The contrasting results between species and schedules of drug delivery together suggest that the endocannabinoid system might impair other brain processes related to cocaine self administration which might in turn reduce the ability to detect the reinforcing effects of cocaine. Certainly, a better understanding and clinical prediction of the influence of the endocannabinoid system on the reinforcing effects of cocaine will come from studies of the effects of cannabinoid agonists and antagonists under a variety of procedures (including progressive-ratio schedules) and with a variety of cannabinoid exposure procedures in primates.
2.6 Effects of cannabinoids on animal models of extinction and reinstatement of cocaine-seeking behavior
One of the major obstacles to successfully achieving withdrawal or abstinence from drug use stems from the difficulty of finding effective strategies against relapse to drug use that is thought to result from craving for the abused substance. Other than behavioral therapy, potential pharmacological support might alleviate craving in abstinent individuals, thus reducing the likelihood of relapse. Although the validity of preclinical models of relapse to drug-seeking and drug-taking behaviors is still under discussion (147, 148, 149, 150), these procedures might give some useful information about which drugs and which brain areas might interfere with the processes that underlie relapse.
In the last years, as it has become increasingly clear that the endocannabinoid system plays a role in different aspects related to substance abuse, researchers have explored the involvement of the cannabinoid system in the neurobiology of relapse to drug-seeking (see citation 151 for review). In rats, when cocaine self-administration behavior was extinguished by substituting saline for cocaine injections (152), experimenter administration of increasing doses of delta-9-THC, 3 hours after withdrawal, failed to induce/reinstate cocaine-seeking behavior. However, administration of HU210, a synthetic cannabinoid agonist which has a higher affinity and greater intrinsic activity than delta-9-THC at CB1 receptors, provoked relapse to cocaine-seeking after a prolonged (14 days) withdrawal from cocaine (104). In addition to differences in treatment, the different outcome of these reports could be due to differences in abstinence strength, because in the first and second reports rats were abstinent from cocaine only for three hours (152) and 14 days (104), respectively. These different conditions might suggest time-related adaptations of the brain systems involved in cocaine-seeking behaviors (153), with no influence of the endocannabinoid system during the early stages of withdrawal from cocaine. On the other hand, when the selective antagonists of cannabinoid CB1 receptors Rimonabant and AM251 were administered alone they did not reinstate cocaine-seeking behavior (104, 154), but both drugs reduced or inhibited cocaine-seeking behavior evoked by priming injections of cocaine (104, 154). It is also interesting to note that Rimonabant reduced cocaine seeking behavior induced by cocaine associated cues, but not stress-induced reinstatement (104), suggesting the involvement of the endocannabinoid system only in some specific processes or pathways related to cocaine-seeking behavior. To this end, it is also interesting to note that AM251 did not block the DA-stimulating effects of cocaine in the NAc, but it attenuated the effects of cocaine on glutamate levels in the NAc, which might be a better neurochemical correlate of cocaine-seeking behavior (154).
3. Human studies
Though not overlapping with the purpose of this review, the "gateway drug" hypothesis of marijuana use is an interesting example of a link or association between cocaine dependence and cannabinoid agonists. Following this hypothesis, marijuana consumption could serve as a gateway drug for subsequent use and abuse of other illicit substances in a progression that can lead towards dependence on heroin, cocaine or other psychostimulants (155, 156, 157, 158). There is still an ample debate about the factors studied and taken into consideration, and the data obtained which support this theory (see for example citations 159, and 160). It is interesting to note, however, that independently from the factor/s that link marijuana use to subsequent abuse of other illegal substances, there are also studies that show a very high prevalence of marijuana use and dependence in cocaine dependent individuals (26, 161, 162, 163). Comorbidity of cocaine dependence with other drug dependencies is also very common (25). Thus the occurrence of cannabis dependence in cocaine dependent subjects might suggest that individual vulnerability for a certain drug can be extended to other drugs (160), but also that common risk factors (genetic, environmental, etc.) might jointly influence the individual drug liability and the links between cannabis and other drug dependencies (163).
With regard to genetic risk factors, it is interesting to note that a human genetic variant of the cannabinoid receptor gene CNR1 has been associated with susceptibility of drug dependence in a non-Hispanic, Caucasian population (164). The association was stronger with intravenous drug use, and was greatest for cocaine, amphetamine, and marijuana dependence. Non-significant association data have been instead obtained in replication of the original findings by Comings and colleagues (165, 166). In a recent paper (31) these contrasting data are discussed as the result of the poor definition of the cannabinoid receptor CNR1 gene's structure, regulation and variation. In the same report, indeed, the authors have shown an improved definition of the human cannabinoid receptor CNR1 gene's locus and variants (31), and it has been also shown that haplotypes toward the 5' end of the cannabinoid receptor CNR1 gene's exons and introns differentiate between non-drug users and drug-abusers of European-American, African-American, and Japanese subjects. Also, a recent report by Ballon and colleagues, 2006 (167), supports the association of the cannabinoid receptor CNR1 gene polymorphism with predisposition to cocaine dependence in an African-Caribbean population.
Another interesting association, from a therapeutic perspective, has been evaluated studying the relationship between marijuana use during cocaine abstinence and relapse to cocaine use. In a study by Labigalini et al., 1999 (168), the authors describe a clinical observation based on spontaneous accounts by Brazilian cocaine-crack abusers, undergoing their first psychiatric assessment, reporting use of cannabis in order to get relief from cocaine-withdrawal symptoms. From a sample of 25 subjects, 68 percent ceased to use crack-cocaine during the study, which lasted 9 months. The small sample and the short period of time suggest a cautious approach to the results outcome; further studies are warranted. A more recent study (169) in New York used a much larger sample of cocaine addicts (144 subjects), and examined whether cannabis use after discharge from inpatient treatment could affect relapse to cocaine use. Data from this study showed that about a third of the subjects used cannabis after discharge and that cannabis use substantially increased the risk of relapse to cocaine use, significantly reduced the achievement of sustained remission, and significantly increased relapse to cocaine use after sustained remission (169). In agreement with this latter report is a preclinical investigation (described above, section 2.6) showing cannabinoid-induced reinstatement of cocaine seeking behavior in rodents (104).
4. Summary and concluding remarks
Most of the scientific articles reviewed in the present manuscript have described studies of cannabinoid CB1 receptor agonists and antagonists tested against cocaine effects in preclinical models thought to be predictive of cocaine abuse. These studies have provided interesting results, especially for the ability of cannabinoid antagonists, and Rimonabant in particular, to significantly counteract some of the reinforcing actions of cocaine (104, 143). Collectively, the studies suggest that a cannabinoid tone, impaired by cannabinoid antagonist administration, is indeed involved in many of the reinforcing effects of cocaine which are believed to be responsible for cocaine abuse and addiction. On the other hand, there are no studies available showing interactions of drugs acting as cannabinoid levels modulators/enhancers on cocaine-induced behaviors. These drugs affecting directly the endogenous cannabinoid tone could interfere with cocaine effects in these preclinical procedures and could substantially increase our knowledge about the cannabinoid-neurobiology related to cocaine dependence.
Suggestions about possible genetic predisposition/vulnerability to cocaine dependence from human studies due to variants of the cannabinoid receptor CNR1 gene have given more strength to the link between endocannabinoids and cocaine. Due to the widespread distribution of cannabinoid receptors in the brain, and their abundance in brain areas playing pivotal roles in drug abuse and addiction, the different expression and regulation of cannabinoid receptors induced by genetic differences (31) might be an important factor in the predisposition or vulnerability to drug dependence. For this reason, the potential to directly interact with endocannabinoid tone in selected brain areas, an effect that can be obtained with endocannabinoid uptake inhibitors or metabolism blockers (4, 60, 62)(as shown also in genetically modified mice, 170), as compared to widespread actions of cannabinoid receptors agonists/antagonists, should be one of the next challenges in the research for medications able to counteract the abuse- and dependence-related behavioral/neurobiological effects of cocaine.
Source, Graphs and Figures: Modulation Of The Endo-Cannabinoid System: Therapeutic Potential Against Cocaine Dependence
