CBD as a novel treatment for context induced relapse to cocaine: Neurobiological attenuation of the cocaine-glutamate response via Ca2+ increase (Research Proposal)

Specific Aim / Research Questions:

A growing body of work since the beginning of the 21st century has targeted the endocannabinoid system and its role in learned behaviors associated with addiction. Research specifically focused on antagonism of the CB receptors, especially CB1, has been the primary focus on the role of the endocannabinoids in addiction studies. A thin, but ample, body of research has begun to look at the non-psychoactive component of cannabis, cannabidiol (CBD), as a therapeutically viable alternative in addiction treatment. Parker et al. (2004) were the first to show that CBD could potentiate the extinction of addiction in cocaine conditioned place preference in rodent models. However few studies have attempted to replicate these results and none have attempted to understand the neurobiological underpinnings which underlie CBD’s ability to attenuate cue induced context reinstatement of cocaine seeking behavior. The following research proposal seeks to replicate the results of Parker et al. (2004) applying a more rigorous model of contextual cocaine seeking following the Diergaarde et al. (2008) model of CB1 antagonism from context induced nicotine seeking studies.

Additionally a neurobiological mechanism for CBD’s ability to attenuate context induced relapse to cocaine will be proposed. Drawing from work by Ren et al. (2009) which showed CBD to have protracted effects (tested at 24h and 2 weeks) on inhibition to cue-induced heroin seeking, CBD is here suggested to be a safer and less invasive alternative to CB1 antagonists and other traditional pharmacological treatments for addiction. It is known that cocaine exerts its effects via inhibition of the dopamine transporter (DAT) intracellular system (Lupica & Riegel, 2005). It is proposed here that CBD will have protracted effects on the disinhibition of DAT via a rise in Ca2+ intracellular fluid and a reduction in the cocaine-glutamate response. The present study will use microdialysis to examine hippocampal changes in Ca2+ and glutamate levels in light of the fact that CBD has been shown to be CB1 and CB2 independent in its modulation of endocannabinoid and glutamate signalling.


Figure 1. Reproduced from Ryan et al. (2007) with blatant disregard for copyright infringement. Model represents a proposed model for a negative feedback of endocannabinoid inhibition of the CBD-Ca2+ response. Accumulation of endocannabinoids resulting from cocaine abuse can block the CBX postsynaptic receptor thus inhibiting Ca2+ rises. Griffin (2017) proposes that CBD’s protracted effects would inhibit the CBX-Gq/11-PLC pathway resulting in increased intracellular Ca2+ levels and indirect modulation of presynaptic CB1 and glutamate receptors. Griffin (2017) proposes that this mechanism is akin to the one found in the Ren et al. (2009) heroin model of protracted CBD effects implicating altered AEA hydrolysis and AMPA GluR1 reactivity as mechanisms underlying behavioral and glutamatergic stabilization.

Figure 2 reproduced from drugabuse.gov. Simplified characterization of the interaction of the endocannabinoid system and cocaine abuse.

Glutamate in addiction models – a role for CBD

To examine general interactions of CBD with glutamatergic pathways this proposal looks at studies by Gobira et al. (2015) and Ren et al. (2009) using rodent models. Gobira et al. (2015) were able to demonstrate that administration of CBD at 15, 30, 60, and 90 mg/kg was able to reduce the duration and latency of seizures caused by cocaine intoxication. Ren et al. (2009) were able to demonstrate a protracted effect of CBD resulting in testing at 24 hours and 2 weeks showing an attenuation of cue induced relapse to heroin. The results of Gobira et al. (2015) showed glutamate levels in synaptosomes of cocaine intoxicated mice at 60nmol/mg protein – an increase from 18nmol/mg protein in vehicle (p<0.001). CBD applied to hippocampal synaptosomes of cocaine intoxicated mice resulted in a glutamate decrease from 60 (cocaine w/ CBD vehicle) to 25nmol/mg protein (p<0.001). Ren et al. (2009) found that rats trained to self-administer heroin had greatly reduced AMPA GluR1 protein expression especially in the NAc core (p<0.001). Additionally ‘post-24hrs administration’ testing showed normalization of AMPA GluR1 expression in the core and medial sections of the NAc, with normalization remaining significant in the core at the 2 week testing measure.

Introduction of AM251 (CB1 antagonist) in the presence of CBD failed to reverse the protective effects of CBD in cocaine intoxication (Gobira et al. 2015). This concurs with Ryan et al. (2007) and the conclusion that CBD can modulate endocannabinoid functioning and glutamate release in a CB1/CB2 independent fashion. Gobira et al. (2015) used the cocaine intoxication level of 75mg/kg based on prior research on intoxication levels. It was found that all doses 15, 30, 60, and 90 mg/kg reduced the duration of seizure symptoms. However only a CBD 30mg/kg dosage was found to increase latency (onset until seizure). Additionally the 30mg/kg was shown to have the largest effect on duration of seizure (18 seconds vs. 42 seconds for vehicle). CBD was administered 30mins prior to cocaine and acute administration was show to be protective. In addiction studies this is critical since overdose is often linked to physiological reactions to either novel environments or sudden increases above usual intake.

Ren et al. (2009) showed that CBD at 5 and 20mg/kg had no effect on heroin intake behavior in the 30min and 24hr groups. Vehicle treated animals showed active lever pressing in cue-induced reinstatement trials (26 presses in a 1hr trial). Ren et al. (2009) found that CBD administration 30mins prior to reinstatement did not attenuate responses at a significant level and thus the data is not shown. However testing of a single CBD injection 24hrs prior to reinstatement testing showed a decrease to 16 presses per trial following a light cue (p<0.01). Even more substantial was the difference between vehicle and CBD when tested 2 weeks later with 5mg/kg administered over the last three days prior to testing (20 vs. 8 presses). This effect was not found to impact priming induced relapse. Postmortem evaluations of the NAc showed an attenuation of elevated CB1 receptors associated with heroin use. Additionally the expression of AMPA GluR1 was stabilized even 2 weeks later specifically in the NAc core.

These results taken together demonstrate the protracted and acute protective factors of CBD and normalization of glutamate functioning. Ren et al. (2009) suggest anandamide (AEA) hydrolysis as one potential mechanism through which CBD exerts this effect. A suggestion that will be revisited when discussing Ryan et al. (2007).

The endocannabinoid system in contextual renewal of drug seeking – methodology

Diergaarde et al. (2008) and Ward et al. (2009) conducted experiments looking at the role of the CB1 receptor in contextual drug seeking for nicotine and cocaine respectfully, and provide essential methodological models for the current proposal. Both studies examined the interactions of CB1 antagonist Rimonabant (SR141716) with the extinction of acquired learning related to drug self administration in rodent operant chamber models. Both studies also provide important insights into the role of the endocannabinoid system’s role in addiction as providing the framework for the behavioral aspects of this proposal.

Diergaarde et al. (2008) provide an important framework for the current proposal. In their study of contextual renewal to nicotine seeking, they developed a crucial operant paradigm. Acquisition training was done over 19 session (12 sessions FR1, 3 session FR2, 4 sessions at FR3) until responding between the active and active hole was significantly discriminated at a stable rate (28 active, 7 inactive nose pokes). Four testing groups were created; control (ABB), novel (AAB), renewal (ABA) and no extinction (AxA). ‘A’ in this case represents the context in which self-administration took place. ‘B’ represents the alternative operant chamber. Of primary concern is the renewal (ABA) group which underwent extinction in a different chamber than acquisition but were tested for reinstatement in the same chamber as acquisition. Results in all groups were measured for active vs. inactive nose pokes with only the renewal (ABA) and no extinction group (AxA) groups showing significant responding on the active hole in nicotine seeking testing. Due to these results the renewal (ABA) group and control group (ABB) were tested for the effects of CB1 antagonist rimonabant on cue induced relapse. Results showed that CB1 antagonism greatly attenuated responses in the ABA group (14 active poke for vehicle, 7 for 1mg/kg, and 2 pokes for 3mg/kg). When taken with data that AAB group did not showed altered nicotine seeking, the results in nicotine seeking and rimonabant trials show that this effect is in response to cues specifically tied to that environment.

Ward et al. (2009) also used an operant chamber instead opting for a single operant chamber with two light cues where an amber light indicated drug delivery. Self-administration results demonstrated that mice would self-administer an average of 54 0.3mg/kg infusions of cocaine over a 2hr trial on a FR1 schedule. Extinction burst responding was measured to be 75 nose-pokes , with extinction meeting ≤33% responding by day 9. Results demonstrated that CB1 antagonism attenuated cue induced cocaine seeking during reinstatement. Additionally mice treated with rimonabant took fewer sessions and fewer discrete trials within sessions to reach extinction criteria. Ward et al. (2009) suggest that the role of the endocannabinoid in extinction learning works by facilitating learning as fear conditioning has been shown to be extinguished by CB1 antagonism.. These results taken with results from Diergaarde et al. (2008) suggest that alterations in endocannabinoid signalling may have alterations in either the motivational or learning specific behaviors of animals.

Neurobiology of CBD interactions

Ryan et al. (2006) were the first to demonstrate that CBD directly increases intracellular Ca2+ production in hippocampal neurons. Their results also implicated that this effect was not mediated by CB1 receptors or by TRPV1 receptors, traditional targets of endocannabinoid antagonist treatments, but was modulated in some way by them. In a follow up study, Ryan et al. (2007) conducted tests on rodent hippocampal cultures to examine the effects of endocannabinoids (endogenous, exogenous, and the hydrolysis of), vanilloid receptors, and G proteins. Their results confirmed their 2006 study of an unidentified CBX receptor on the postsynaptic neuron. Inhibition of the CBX receptor via the Gq/11-/phospholipase C (PLC) pathway was shown to greatly facilitate the CBD-Ca2+ response. Xi et al. (2006) used microdialysis to demonstrate the involvement of glutamate receptors mGluR2/3 in cocaine priming increases of glutamate in the NAc. Work by Loweth et al. (2014) further implicates mGluR1 receptors as critically important in preventing the accumulation of CP-AMPAR in the NAc (implicated in the craving of cocaine).

The results of the aforementioned studies when taken together with figures 1 and 2 demonstrate a model by which CBD can cause tonic alterations rather than direct inhibition of cocaine’s behavioral and neurobiological changes. A flaw in the study by Ryan et al. (2007) as admitted by the authors was that proper endocannabinoid stimulation requires membrane depolarization which can not be done in vitro. Additionally the researchers used rat brain pups (1-3 days old) which would not have been exposed to addiction or glutamatergic alterations. The importance of CBX being located on the postsynaptic neuron and CBD’s inhibition of AEA reuptake and hydrolysis is of importance to the proposed model. It is suggested in this proposal that inhibition of AEA by CBD results in tonic signalling and a clearing of 2-AG from the synaptic cleft (figure 2). Ryan et al. (2007) found that exogenously applied AEA and 2-AG greatly reduced the CBD-Ca2+ response. Suggesting that CBD served as its own negative feedback where the PLC pathway was active in reducing Ca2+ signalling. However it is suggested in this proposal that CBD, exerting protracted effects a la Ren et al. (2009) will show elevated Ca2+ in Renewal-CBD conditions versus Renewal-Vehicle. The hope being that this could be replicated in future studies with live Ca2+ measuring.

Xi et al. (2006) implicate a role for mGluR2/3 receptors in mediating endocannabinoid signalling in cocaine-triggered relapse to drug seeking. This concurs with Loweth et al. (2014) who found that mGlur1 stimulation could enhance long-term depression via reduced CP-AMPAR synaptic transmission. According to Cata et al. (2015) CP-AMPAR are located on glutamatergic synapses and are Ca2+ permeable. It is my hypothesis that CBD may be able to clear CP-AMPAR levels via increased Ca2+ intracellular fluid being able to move CP-AMPAR out of the neuron via increased VGCC (voltage gated calcium channel) activity. This hypothesis was created by combining the CP-AMPAR concepts of Loweth et al. (2014) and CBD-Ca2+-VGCC relationships shown by Ryan et al. (2006). However testing of this idea is beyond the scope of the current study. Ryan et al. (2007) additionally showed that inhibition of the Gq/11-/phospholipase C (PLC) pathway via the application of PLC inhibitor U73122 (100nM) six-fold increase compared to CBD control in fluorescent markers of Ca2+ in neurons and four fold in glia. Application of exogenous AEA and 2-AG showed a reduction in CBD-Ca2+ responding. However since a lack of stimulated cell tissue is studied, it is unreliable to think that this effect would not be different in a neuron exhibiting altered AEA/2-AG hydrolysis as seen in figure 2.

In sum these findings show that CBD effects are CB1 and CB2 receptor independent. Increased responsiveness of glutamate responsiveness should be indicative of changes in Ca2+ functioning as it has been previously demonstrated that glutamate modulation can clear calcium permeable CP-AMPA receptors responsible for cocaine craving in cue induced cocaine relapse. These changes may be observable by looking at postmortem Ca2+ levels of rats treated with CBD vs rats treated with vehicle in the current proposal for context induced relapse to cocaine.


Operant Conditions and Training

Operant testing is adapted from Diergaarde et al. (2008). Two operant chambers will be used that will differ in light color and location, floor designs, receptacle location, and contents of the waste tray. The required rate of responding to meet extinction criteria will follow the model of Ward et al. (2009) which established a ≤33% level of responding on active nose-pokes as meeting extinction criteria.

In the vein of the Diergaarde et al. (2008) model, ten rats will be put into four groups; control (ABB), novel (AAB), renewal (ABA) and no extinction (AxA). This is to establish that rats will acquire cocaine related learning similar to nicotine studies and to control for non-context related cues. 8 rats each will be added to the renewal and control groups so that n=18 in each group (renewal and control) for CBD administration and testing.

Cocaine will be administered via implant, and reinforced on a progressive FR schedule until responding on a FR3 schedule is established (Diergaarde et al. 2008). Extinction patterns will follow the same as outlined above in the background section in regards to the Diergaarde model and reinstatement tested in the above stated conditions. Active nose pokes are associated with lights and tones. Reinstatement is tested by responding in response to the cues associated with the acquisition chamber.

Animals and Drugs

Wistar rats will be used to replicate the Diergaarde et al. (2008) model to ensure consistent levels of learning. Rats will receive a 30mg/kg injection of CBD as this was shown to be the optimal dose in the Gobira et al. (2015) study in reducing the cocaine-glutamate response. Acute administration will be tested by injecting CBD or vehicle 30mins prior to reinstatement testing as 30mins was the acute administration time used by both Ren et al. (2009) and Gobira (2015). Animals will be tested again 24hr following their initial testing without an additional CBD dosing. Cocaine will be administered via surgically implanted intravenous catheters (Xi et al. 2006), with a concentration of 0.3mg/kg per infusion (Ward et al. 2009).

Measurements and Dependent Variables

Total number of rats tested for reinstatement in the CBD experiment will be 36. The two groups (renewal and control) will each be subdivided to give 4 groups for comparison:

  1. Renewal – CBD (n=9)
  2. Renewal – Vehicle (n=9)
  3. Control – CBD (n=9)
  4. Control – Vehicle (n=9)

Rats rate of responding will be measured in active hole nose pokes and assessed after 60mins of testing after initial reinstatement and at 24 hrs post administration. Microdialysis probes will be inserted into the hippocampus 12hrs prior to testing as modeled by Xi et al. (2006). Samples will be collected twice during testing (30min mid test and 60min post test) and frozen immediately. Glutamate responses will be measured by fluorometric detection as exemplified by Xi et al. (2006) and Ryan et al. (2007). Finally, hippocampal synaptosomes will be created following 24hr testing and animal sacrifice. Tests for Ca2+ reactions will be made between the four comparison groups.


Behaviorally, it is expected that CBD will attenuate active nose pokes in the renewal group at a significant level. This would insinuate CBD exhibiting an inhibitory effect upon context induced relapse to cocaine. It is also expected that CBD will reduce the cocaine-glutamate response in all condition, with special attention paid to differences between Renewal-CBD and Renewal-Vehicle groups. It is expected the Renewal-CBD group will show dramatically decreased glutamate levels. It is anticipated that responding will only slightly increase in the 24hr group and glutamate levels only show moderate increases. It is anticipated that Renewal-CBD vs. Renewal-Vehicle groups will show significant differences in Ca2+ reactions replicated from Ryan et al. (2007). With the Renewal-Vehicle group showing more inhibition of the Gq/11-/phospholipase C (PLC) pathway. Conversely, it would be expected that the Renewal-CBD group would show disinhibition of the PLC pathway and subsequent increases in Ca2+.


Since no method for active Ca2+ exists it will remain for future studies to study these differences in live models. It is hoped that by comparing the synaptosomes of CBD vs. Vehicle groups that differential brain structures can at least be studied which was lacking in the in vitro models of Ryan et al. (2007). Since enough factors exist in the current model, I picked one dosage of CBD based on prior studies. It will be up to future studies to see if this 30mg/kg dosing is optimal for addiction models. Additionally only synaptosomes following 24 hrs were collected. Future studies should examine acute applications, post 2-week application, regular application, and tapering amount applications.


Aguiar, D.C., Crippa, J., Gobira, P.H., Goncalves, B.D.C., Moreira, F.A., de Oliveira, A.C., Santos, R.P.M., Vieira, L.B., & Vilela, L.R. (2015). Cannabidiol, a cannabis sativa constituent, inhibits cocaine-induced seizures in mice: Possible role of the mTOR pathway and reduction in glutamate release. NeuroToxicology, 50(1). 116-121.

Burton, P., Mechoulam, R., Parker, L.A., Sorge, R.E., & Yakiwchuk, C. (2004). Effect of low doses of 𝚫9-tetrahydrocannabinol and cannabidiol on the extinction of cocaine-induced and amphetamine-induced conditioned place preference learning in rats. Psychopharmacology, 175(3). 360-366.

Cata, R., Jutras-Aswad, D., & Prud’homme, M. (2015). Cannabidiol as an intervention for addictive behaviors: A systematic review of the evidence. Substance Abuse: Research and Treatment, 9(1). 33-38.

Centonze, D., Laricchiuta, D., & Petrosini, L. (2013). Effects of endocannabinoid and endovanilloid systems on aversive memory extinction. Behavioural Brain Research, 256(1). 101-107.

De Vries, T.J., De Vries, W., Diergaarde, L., Rasso, H., & Schoffelmeer (2008). Contextual renewal of nicotine seeking in rats and its suppression by the cannabinoid-1 receptor antagonist Rimonabant (SR141716A). Neuropharmacology, 55(5). 712-716.

Drysdale, A.J., Pertwee, R.G., Platt, B., & Ryan, D. (2006). Cannabidiol-induced intracellular Ca2+ elevations in hippocampal cells. Neuropharmacology, 50(5). 621-631.

Drysdale, A.J., Pertwee, R.G., Platt, B., & Ryan, D. (2007). Interactions of cannabidiol with endocannabinoid signalling in hippocampal tissue. European Journal of Neuroscience, 25(1). 2093-2102.

Dykstra, L.A., Rosenberg, M., Walker, E.A., & Ward, S.J. (2009). The CB1 antagonist rimonabant (SR141716) blocks cue-induced reinstatement of cocaine seeking and other context and extinction phenomena predictive of relapse. Drug and Alcohol Dependence, 105(3). 248-255.

Flores-Barrera, E., Ford, K.A., LaCrosse, A.L., Le, T., Li, X., Loweth, J.A., Milovanovic, M., Olive, M.F., Scheyer, A.F., Szumlinski, K.K., Tseng, K.Y., Werner, C.T., Wolfe, M.E. (2014). Synaptic depression via mGluR1 positive allosteric modulation suppresses cue-induced cocaine craving. Nature Neuroscience, 17(1). 73-84.

Gardner, E.L., Gilbert, J.G., Li, X., Pak, A.C., Peng, X.O., & Xi, Z.H. (2006). Cannabinoid CB1 receptor antagonist AM251 inhibits cocaine-primed relapse in rats: Role of glutamate in the nucleus accumbens. The Journal of Neuroscience, 26(33). 8531-8536.

Higuera-Matas, A., Hurd, Y.L., Morris, C.V., Ren, Y., & Whittard, J. (2009). Cannabidiol, a nonpsychotropic component of cannabis, inhibits cue-induced heroin seeking and normalizes discrete mesolimbic neuronal disturbances

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s