Single and combined effects of plant‐derived and synthetic cannabinoids on cognition and cannabinoid‐associated withdrawal signs in mice

Abstract

Background and Purpose

It has been suggested that the non‐euphorogenic phytocannabinoid cannabidiol (CBD) can ameliorate adverse effects of Δ⁹‐tetrahydrocannabinol (THC). We determined whether CBD ameliorates cognitive deficits and withdrawal signs induced by cannabinoid CB₁/CB₂ receptor agonists or produces these pharmacological effects on its own.

Experimental Approach

The effects of THC or the CB₁/CB₂ receptor full agonist WIN55212 alone, CBD alone or their combination were tested across a range of doses. Cognitive effects were assessed in C57BL/6 mice in a conditional discrimination task and in the Barnes maze. Cannabinoid withdrawal signs were assessed following precipitated withdrawal by acute administration of the CB₁ receptor antagonist SR141716, the 5‐HT_1A receptor antagonist WAY100635, the TRPV1 receptor antagonist capsazepine or the adenosine A_2A receptor antagonist SCH58261.

Key Results

THC produced significant motor and cognitive impairment in the Barnes maze task, none of which were attenuated by the addition of CBD. CBD alone did not affect cognitive performance. Precipitation of withdrawal signs by SR141716 occurred in mice chronically treated with THC or WIN55,212. These withdrawal signs were not attenuated by addition of chronic CBD. Chronic treatment with CBD alone did not induce withdrawal signs precipitated by SR141716 or WAY100635. Chronic CBD treatment also produced anxiolysis, which was not altered by attempting to precipitate withdrawal‐induced anxiety with a range of antagonists.

Conclusions and Implications

CBD as a monotherapy may prove to be a safer pharmacological agent, than CB₁ receptor agonists alone or in combination with CBD, for the treatment of several disorders.

Linked Articles

This article is part of a themed section on 8^th European Workshop on Cannabinoid Research. To view the other articles in this section visit http://onlinelibrary.wiley.com/doi/10.1111/bph.v176.10/issuetoc

Abbreviations

CBD

cannabidiol

EPM

elevated plus maze

THC

Δ⁹‐tetrahydrocannabinol

Introduction

The non‐euphorogenic cannabinoid cannabidiol (CBD) is emerging as a promising therapeutic agent for the treatment of a range of psychiatric and inflammatory disorders (see Campos et al., 2016). This is driven not only by its anti‐inflammatory and neuroprotective pharmacological profile but by that fact that it is devoid of many of the psychoactive effects produced by the primary phytocannabinoid Δ⁹‐tetrahydrocannabinol (THC). In spite of other shared pharmacological actions, CBD exhibits much lower affinity than THC for cannabinoid CB₁ receptors (Showalter et al., 1996; Bisogno et al., 2001; Thomas et al., 2004; Pertwee et al., 2005). Currently, over 90 clinical trials are planned, ongoing or completed in which CBD is a primary intervention (https%3A%2F%2Fwww.clinicaltrials.gov%2Fct2%2Fresults%3Fcond%3D%26term%3Dcannabidiol%26cntry1%3D%26state1%3D%26recrs%3D).

Sativex is an oromucosal spray extract containing THC, CBD and specific minor cannabinoids and other non‐cannabinoid components developed by GW Pharma, and a search for ‘Sativex’ on the clinicaltrials.gov website yields over 500 results for a vast array of safety and efficacy trials for this cannabinoid‐based therapeutic. Part of the rationale for Sativex was the idea that the incorporation of additional compounds in the Cannabis plant, especially CBD, would mitigate some of the adverse side effects associated with THC. In addition, in a growing number of states where medical marijuana is approved, more products are being made available which contain moderate to high doses of CBD, and more marijuana strains are being produced which yield high levels of CBD in combination with low to high combinations of THC. Therefore, an overarching strategy has been to harness the therapeutic potential of CBD while lessening the adverse effects associated with THC. This strategy is exemplified by the term ‘entourage effect’, used to describe how phytocannabinoids (and other chemicals classes such as terpenes and flavonoids) may produce more than additive actions. For example, CBD may either synergize with or antagonize THC effects to enhance wanted actions and/or antagonize adverse actions (Russo, 2011). We have recently demonstrated, for example, that a 1:1 ratio of CBD to THC synergistically prevented development of mechanical sensitivity in a mouse model of chemotherapy‐induced neuropathic pain, greatly decreasing the dose of each drug required to produce efficacy (King et al., 2017). A continuation of quantitative research is required to determine the scope of these interactions, including how they affect the adverse effects of THC.

Alterations in learning and memory are of the most common adverse pharmacological effects linked with cannabis use. In humans, acute exposure to cannabis or THC has been shown to impair working and episodic memory (Curran et al., 2002; Bossong et al., 2012; Crane et al., 2013). Long‐term impairments in memory have been reported mainly in frequent, heavy users, but confounding factors make it difficult to establish cause–effect relationships between cannabis use and changes in neurocognitive function (Curran et al., 2016). Animal studies have revealed a high density of cannabinoid receptors in memory‐associated brain regions such as the hippocampus, amygdala, striatum and prefrontal cortex (Herkenham et al., 1991; Matsuda et al., 1993), and THC disrupts hippocampal LTP in animal models as well (Hoffman et al., 2007; Misner and Sullivan, 1999). THC administration is associated with cognitive impairment in a variety of experimental rodent assays, including operant and spatial maze models of memory (Brodkin and Moerschbaecher, 1997; Ferrari et al., 1999; Heyser et al., 1993; Lichtman et al., 1995; Mallet and Beninger, 1998; Nakamura et al., 1991; Varvel et al., 2001). Therefore, one objective of the present study was to determine the effects of CBD alone and in combination with THC or the synthetic CB₁/CB₂ receptor agonist WIN55,212 to define effects of single and combined cannabinoid regimens on instrumental conditioning and spatial learning. Previous reports have described impairing effects of CB₁ receptor agonists on both types of learning (see Lichtman et al., 1995; Sokolic et al., 2011). Cannabinoid agonists and CBD were co‐administered in a 1:1 ratio based on dose to model the content of the major phytocannabinoids in Sativex.

Withdrawal upon abstinence is another pharmacological effect associated with chronic marijuana use in humans (Cooper and Haney, 2008; Jones, 1983). A specific cannabis withdrawal syndrome is well recognized and affects approximately 50% of daily users upon cessation of use (Budney et al., 2004), and common symptoms include craving, poor sleep quality, irritability and dysphoria (Allsop et al., 2011). Withdrawal symptoms associated with chronic exposure to CB₁ receptor agonists have also been observed in rodent models (Cook et al., 1998; Rubino et al., 1998; Castane et al., 2004; Huang et al., 2009; Marusich et al., 2014). In the majority of these experiments, administration of the CB₁ receptor selective antagonist SR141716 is used to precipitate withdrawal in animals chronically treated with cannabinoid agonists such as THC or WIN55212. Characteristic observations of cannabis‐associated withdrawal in rodents include the presence or absence of head shakes, paw flutter or scratching, and hypolocomotion. Therefore, another objective of the present study was to determine whether chronic CBD treatment leads to physical dependence and subsequent withdrawal symptoms mediated by actions on the CB₁ receptor, and whether chronic co‐administration of CBD in combination with the CB receptor agonists THC or WIN55212 altered the observed signs of CB₁ receptor‐associated withdrawal. Cannabinoid agonists and CBD were co‐administered in an approximately 1:1 ratio, based on the phytocannabinoid ratio in Sativex, while also testing relevant doses of cannabinoid agonists known to induce physical dependence.

While evidence supporting the pharmacological effects of CBD as being mediated through CB₁ receptors is limited (but see Bisogno et al., 2001; Casarotto et al., 2010; Sartim et al., 2016), more compelling evidence attributes many of CBD's actions to interactions with other receptors. Three commonly reported receptor mechanisms for the pharmacological effects of CBD are (i) direct agonist action at 5‐HT_1A receptors (Russo et al., 2005; Gomes et al., 2011; Ward et al., 2014), (ii) direct agonist action at TRPV1 receptors (Costa et al., 2004; Pertwee et al., 2005) and (iii) inhibition of the equilibrative nucleoside transporter leading to indirect agonist activity at adenosine A_2A receptors (Carrier et al., 2006; El‐Remessy et al., 2008; Liou et al., 2008; Pandolfo et al., 2011). Therefore, a third objective of the present study was to determine whether chronic CBD administration could lead to traditional signs of physical dependence through a mechanism other than through CB₁ receptors. In these studies, we have focused on a potential 5‐HT_1A receptor mechanism as this is the receptor most indicated in the demonstrated CNS effects of CBD, such as anxiolysis and analgesia. As cannabinoid withdrawal has also been associated with anxiety in humans (Ramesh et al. 2011) and mouse models (Huang et al., 2010), we also tested the potential for 5‐HT_1A, TRPV1 or A_2A receptors antagonism to precipitate anxiogenesis in chronically treated CBD mice.

Methods

Animals

All animal care and experimental procedures were approved by the Temple University Institutional Animal Care and Use Committee. The animal facilities at Temple University are accredited by the Association for Assessment and Accreditation of Laboratory Animal Care, and sentinel animals are tested quarterly to maintain a pathogen‐free environment. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny et al., 2010; McGrath and Lilley, 2015).

A total of 335 male C57Bl6 mice (Taconic, 6–8 weeks upon delivery) were used the experiments described here. Animals were group‐housed, four to a cage except where noted, in a temperature‐controlled animal facility. Cages are polycarbonate with cobb bedding on ventilated racks, and each cage contains enrichment items such as bedding materials and chew sticks. Mice were kept in the animal facility for at least 1 week prior to experiments with a light–dark cycle of 12:12 h and had access to food and water ad libitum except where noted. The number of animals per treatment used in the experiments was based on several years of experience with the experimental protocols by the principal investigators.

Conditional discrimination

Instrumental discrimination learning was assessed in mice using an appetitive‐motivated operant conditional discrimination procedure (Bisen‐Hersh et al., 2013) that used a tone/light compound stimulus and temporal duration. Following the habituation period, mice were separated into individual cages, weighed and food‐restricted for 24 h prior to the start of experimental testing. Water remained available ad libitum. On day 1 of training, each mouse was weighed and placed inside an experimental chamber for 15 min before the session started. During each session, the house light illuminated to indicate the availability of a vanilla‐flavoured liquid nutritional drink, Ensure Plus/water (50:50) solution, which would be presented in a 0.01‐cc dipper. These food deliveries were contingent upon correct responding, as defined separately for each phase described below. Each session lasted for 1 h, following which mice were fed 3 g of food and returned to their cages.

The conditional discrimination procedure involved a series of 8 phases, and specific criteria were completed for each prior to progressing to the next phase. ‘Extensive training’ was defined as completion of Phases 1–5, followed by a test in which a discrimination ratio was measured (Phase 6). ‘Rapid training’ was defined as Phase 7, followed by a test in which a discrimination ratio was measured (Phase 8). Briefly, during Phases 1 and 2, a nose‐poke response into the left or right hole, respectively, was reinforced on a fixed ratio schedule of reinforcement. During Phases 3 and 4, an audible tone was presented along with an illuminated stimulus light above the left or right poke hole for 2 or 8 s, respectively, and a nose‐poke response into the correct hole was reinforced. During Phase 5, the duration of the tone/light cue was randomly alternated between short (2 s) and long (8 s), and a nose‐poke response into the correct hole was reinforced. Mice were trained on this set of contingencies to 70% correct responding over two consecutive sessions before the discrimination ratio was assessed during the Phase 6 test session. During Phase 7, the contingencies previously described were reversed and this new set of contingencies was trained for one session before the discrimination ratio was assessed during the Phase 8 test session.

After separate groups of mice were trained to criteria (n = 8 per group), they were pretreated with THC (0–10 mg·kg⁻¹ i.p.), CBD (0–10 mg·kg⁻¹) or CBD + THC (0–10 mg·kg⁻¹ each drug), or WIN55212 (0–3.0 mg·kg⁻¹), CBD (0–3 mg·kg⁻¹) or CBD + WIN55212 (0–3 mg·kg⁻¹ each drug), 15 min prior to days 6 and 8 to test the acute effect of single and combined cannabinoids on discrimination and reversal learning (Veh and THC only). These doses and pretreatment time were selected based on published data (Lichtman et al., 1995, Lichtman and Martin, 1996, Varvel et al., 2001 and Wise et al., 2009).

Barnes maze

Spatial learning was assessed in mice using an aversively‐motivated standard, automated Barnes maze. A commercially available maze was used (Stoelting Co.). Dark curtains were hung to form a 177 × 177 cm square around the maze to exclude extraneous cues. Four spatial cues consisting of simple geometric figures (plus, square, circle and triangle) were hung on the four curtains. Mice were transported to the testing room and allowed 1–2 h to acclimatize. The first habituation trial lasted for 1 min, before the experimenter guided the mouse to the target hole. If the mouse did not enter on its own volition, a gentle tail‐pull was used to encourage entry into the escape tunnel. The escape tunnel was then covered, and the mouse rested for 1 min before moving the mouse to a holding cage for a 3 min inter‐trial interval during which time the maze was cleaned with 70% isopropanol. Acquisition training began 1 h later, with mice being placed on the maze and permitted to explore for 5 min. If the mouse entered the escape tunnel, the trial was terminated, the target hole covered and the mouse permitted 1 min of rest in the target hole. If the mouse did not exit the maze within the trial, the experimenter guided the mouse to the target hole as before and the mouse was kept there for 1 min. Four acquisition trials were performed per day for 4 days, for a total of 16 acquisition trials. The memory retention trial occurred 24 h later on the fifth day. For this, the escape tunnel was removed from the target hole and replaced with a shallow tray. The ability of the mice in finding the target hole was tested in a 5 min trial. Animals were pretreated with THC (5.0–20 mg·kg⁻¹), CBD (5.0–20 mg·kg⁻¹) or a combination of THC + CBD (5.0–20 mg·kg⁻¹) for 30 min prior to the retention trial. Trials were recorded using a CCD monochrome camera and analysed using ANY‐Maze tracking software. Using nose‐point tracking, latency to entrance of the animal's head into the target hole, total time spent in the target zone, total number of entries into the target zone, average speed around the maze and distance travelled were recorded per trial. These doses and pretreatment times were selected based on published data (Lichtman et al., 1995, Lichtman and Martin, 1996, Varvel et al., 2001 and Wise et al., 2009).

Chronic cannabinoid administration and precipitated withdrawal

Separate groups of mice (n = 8 per group) were injected i.p. with vehicle, CBD (5.0 or 20 mg·kg⁻¹), THC (20 mg·kg⁻¹) or WIN55212 (3.0 mg·kg⁻¹) alone or in combination (20 mg·kg⁻¹ CBD + 20 mg·kg⁻¹ THC, or 5.0 mg·kg⁻¹ CBD + 3.0 mg·kg⁻¹ WIN55212) every 12 h for 4.5 days. Selective antagonists for the CB₁ receptor(10 mg·kg⁻¹ SR141716), 5‐HT1A receptor (0.5 mg·kg⁻¹ WAY 100635), TRPV1 receptor (10 mg·kg⁻¹ capsazepine) or adenosine A_2A receptor(0.05 mg·kg⁻¹ SCH 58261) were administered 2 h following the last cannabinoid injection. Doses of THC, WIN55212 and SR141716 were based on extensive pilot work in our laboratory as well as Cook et al. (1998), Huang et al. (2009, 2010) and Lichtman et al. (2001). The dose of WAY 100635 was determined by previous published work by the principal investigators as well as Fogaca et al. (2014) and Rock et al. (2017). The dose of capsazepine was based on Costa et al. (2004). The dose of SCH 58261 was based on Cunha et al. (2008).

Immediately following injection of the antagonist, mice were placed in open field apparatus with clear Plexiglas walls (Med Associates ENV 510, 10.75″ L × 10.75″ W × 8″ H). They were video‐taped for 30 min, and a blinded observer subsequently rated their behaviours in the following four categories: head shakes (turning or twisting of the head side to side), rearing, head scratches (repetitive movements around the neck with the hind paw) and paw flutters (episodic rapid lateral movements of the paws). These withdrawal behaviours were selected based on reports of the most robust withdrawal behaviours seen following chronic cannabinoid treatment precipitated by SR141716, as cited above.

Elevated plus maze

Thirty minutes following injection of the antagonist, mice were placed on the mouse elevated plus maze [EPM; 50 cm high, consisting of two open and two closed arms (28 × 7 cm) in dim light conditions (7 lux)]. To begin the session, mice were placed in the centre of the maze facing an open arm. Exploration of the mouse was videotaped for 5 min. The measure of anxiety is the per cent time spent on the open arms, whereas the total number of open and closed arm entries is considered a measure of locomotor/exploratory activity.

Data and statistical analysis

The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015). All statistical analyses were run using GraphPad Prism version 6 and as described below for the different sets of data. Significance for all statistical tests was set at P < 0.05.

Conditional discrimination. The discrimination ratio defined as [correct responding / (correct + incorrect responding)]*100 was calculated for each mouse at Phases 6 and 8 to measure response accuracy following extensive training and reversal learning respectively. Only mice that reached the 70% criteria at Phase 5 were included in the following analyses. Separate repeated measures one‐way ANOVAs were used to assess the effect of a range of THC, WIN, CBD or combination doses on correct discrimination. In addition, two‐way ANOVAs with Tukey's multiple comparisons tests were run to determine whether single or combined treatments were significantly different from one another.

Barnes maze. Separate one‐way ANOVAs and Dunnett's multiple comparisons tests were run to assess the effect of a range of THC, CBD or combination doses on each variable measured. In addition, two‐way ANOVAs with Tukey's multiple comparisons tests were run to determine whether single or combined treatments were significantly different from one another.

Withdrawal behaviours. One‐way ANOVA was used to determine the effect of acute administration of SR141716 or WAY 100635 and chronic dosing of CBD, THC, WIN‐55212 or combinations on head shakes, paw flutters and scratching. Post hoc analysis was performed using Dunnett's multiple comparisons test to determine the following: (i) which treatment groups were significantly different from vehicle+vehicle treatment; (ii) which chronic treatment groups were significantly different from vehicle+SR141716 or vehicle+WAY 100635 treatment; and (iii) whether combination CBD groups were significantly different from chronic THC or WIN55212 alone groups. EPM: Un‐paired Student's t‐tests were used to determine the effects of acute antagonist treatments (WAY 100635, capsazepine or SCH 58261) in chronic CBD treated mice as compared with their vehicle‐treated controls.

Materials

Δ⁹‐THC, CBD, WIN55212 and SR141716 were provided by the NIDA drug supply programme (Rockville, MD, USA). All cannabinoid compounds were dissolved in a vehicle of 1:1:18 ethyl alcohol: cremophor:saline (v/v). The 5‐HT_1A receptor antagonist WAY 100635 (dissolved in saline), the TRPV1 receptor antagonist capsazepine (dissolved in the cremophor vehicle described above) and the adenosine A_2A receptor antagonist SCH 58621 (dissolved in saline) were purchased from Tocris Bioscience (Ellisville, MO, USA). All agents were injected i.p. based on body weight.

Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 (Alexander et al., 2017a,b,c).

Results

Conditional discrimination

No effects of THC, CBD or THC + CBD treatment were observed on the test for extensive training (Figure 1A) or reversal learning (data not shown). For the WIN/CBD experiment, WIN55212 alone showed a dose‐dependent decrease in performance (Figure 1B). Two‐way ANOVA showed a significant effect of treatment [F _{(2, 63)} = 3.66] but no post hoc differences between treatments. In addition, separate one‐way ANOVAs showed no significant effect of THC, WIN55212, CBD or combinations, compared to their vehicle control.

Figure 1.

Effects of THC or WIN55212 alone or in combination on retention of a tone discrimination task. (A) Treatment with THC, CBD or their combination did not affect performance in the discrimination task following training. (B) Although a decrease was observed for WIN55212 on discrimination performance and a less potent decrease for co‐administration with CBD to reverse this affect, one‐way ANOVAs showed no significant effects of each treatment. Two‐way ANOVA showed a significant effect of treatment across groups, but post hoc analysis determined no significant differences between two particular treatment groups. N = 8 per group. Underneath overarching bracket, * P < 0.0.5, main effect of dose.

Barnes maze

Latency to first target zone entry

THC alone and in combination showed a dose‐dependent increase in the time taken to enter the target zone for the first time. One‐way ANOVA for each compound showed no significant effect of CBD or THC + CBD treatment, but a significant effect of THC on latency to first target zone entry [F(3,24) = 2.735]. Dunnett's multiple comparisons post test showed a significant effect of 20 mg·kg⁻¹ THC compared with vehicle control. Two‐way ANOVA showed a significant effect of cannabinoid treatment on latency [F(2,54) = 3.965], but no post test significance between specific treatment groups. There were no significant main effects of dose or significant interaction (Figure 2A).

Figure 2.

Effects of THC or CBD + THC administered in 1:1 ratio based on dose performance during retention of spatial memory. Top panel: THC and CBD + THC dose‐dependently decrease latency to first target zone entry, total time in target zone and # of entries into target zone. CBD did not alter retention of spatial memory. Bottom panel: THC did not affect total distance travelled or average speed, while CBD increased both locomotor measurements. N = 7 per group. To the left of overarching bracket, * P < 0.0.5, main effect of dose; # P < 0.0.5, main effect of treatment. Next to individual data points, * P < 0.0.5, significantly different from vehicle‐treated group.

Total time in target zone

THC alone and in combination showed a dose‐dependent decrease in the total time spent in the target zone across the test session. One‐way ANOVA for each compound showed no significant effect of CBD treatment, but a significant effect of THC [F(3,24) = 3.132] and THC + CBD [F(3,24) = 4.683] treatment. Dunnett's multiple comparisons post test showed a significant effect of 20 mg·kg⁻¹ THC compared with vehicle control and significant effects of 10 and 20 mg·kg⁻¹ THC + CBD. Two‐way ANOVA showed a significant effect of cannabinoid treatment on latency [F(2,54) = 3.045], but no post test significance between specific treatment groups. There were no significant main effects of dose or significant interaction (Figure 2B).

Number of entries into target zone

THC alone and in combination showed a dose‐dependent decrease in the total number of times the mice entered the target zone during the test period. One‐way ANOVAs for each compound showed no significant effect of THC, CBD or THC + CBD treatment. Two‐way ANOVA showed no significant main effects or significant interaction (Figure 2C).

Total distance travelled

CBD produced a dose‐dependent increase in total distance travelled, while the combination produced an inverted U‐shaped function. THC did not affect total distance travelled on the maze. One‐way ANOVA for each compound showed a significant effect of CBD treatment on increasing total distance travelled [F(3,24) = 3.999] and no significant effect of THC treatment or THC + CBD. Dunnett's multiple comparisons post test showed a significant effect of 10 and 20 mg·kg⁻¹ CBD compared with vehicle control. Two‐way ANOVA showed a significant effect of cannabinoid treatment on total distance [F(2,54) = 27.27] and a significant interaction [F(4,54) = 2.905] (Figure 2D).

Average speed

CBD, THC and their combination all produced an inverted U‐shaped dose effect on average speed on the maze. One‐way ANOVA for each compound showed a significant effect of CBD treatment on increasing average speed [F(3,24) = 9.089] and no significant effect of THC treatment or THC + CBD. Dunnett's multiple comparisons post test showed a significant effect of 10 and 20 mg·kg⁻¹ CBD compared with vehicle control. Two‐way ANOVA showed a significant effect of cannabinoid treatment on total distance [F(2,54) = 24.07], a significant main effect of dose [F(2,54) = 6.034] and no significant interaction (Figure 2E).

Withdrawal behaviours

CB₁ receptor antagonist precipitated withdrawal – THC and CBD

The ability of an acute injection of the CB₁ receptor antagonist SR141716 (10 mg·kg⁻¹ i.p.) to elicit withdrawal signs following chronic administration of 20 mg·kg⁻¹ THC, 20 mg·kg⁻¹ CBD or the combination was tested. Scratching: One‐way ANOVA and Tukey's multiple comparison test showed a significant overall effect of drug treatment and several post hoc significant differences [F(4,35) = 13.9]. The following treatment groups were significantly different from vehicle‐treated mice: vehicle+SR, CBD + SR showed significant increases in scratching behaviour. The following treatment groups were significantly different from SR‐treated mice, suggesting an effect of precipitated withdrawal: THC + SR, CBD + THC + SR showed an attenuation of scratching behaviour elicited by SR alone. The THC + CBD + SR group showed significantly less scratching than the CBD + SR group but no difference as compared with the THC + SR group. Paw tremor: One‐way ANOVA and Tukey's multiple comparison test showed a significant overall effect of drug treatment and several post hoc significant differences [F(4,35) = 6.91]. The following treatment groups were significantly different from vehicle‐treated mice: THC + SR led to a significant increase in paw tremor behaviour. The following treatment groups were significantly different from SR‐treated mice, suggesting an effect of precipitated withdrawal: THC + SR led to a significant increase in paw tremor behaviour compared with SR alone. The THC + CBD + SR group showed a trend toward less paw tremor behaviour as compared with THC + SR, but there was no post hoc significant effect between these two groups. Head shakes: One‐way ANOVA and Tukey's multiple comparison test showed a significant overall effect of drug treatment and several post hoc significant differences [F(4,35) = 3.59]. The following treatment groups were significantly different from vehicle‐treated mice: vehicle+SR, CBD + SR and THC + CBD + SR showed significant increases in head shake behaviour. No treatment groups were significantly different from SR‐treated mice. Lastly, the THC + CBD + SR group showed no difference in rearing compared with the CBD + SR and THC + SR groups. Rearing: One‐way ANOVA and Tukey's multiple comparison test showed a significant overall effect of drug treatment [F(4,35) = 3.69]. The following treatment groups were significantly different from vehicle‐treated mice: THC+ CBD + SR showed significant decreases in rearing behaviour. No treatment groups were significantly different from SR‐treated mice. Lastly, the THC + CBD group showed no difference in rearing compared with the CBD + SR and THC + SR groups (Figure 3).

Figure 3.

Effects of SR‐141716‐precipitated withdrawal from chronic administration of THC, CBD or their combination. Acute administration of SR141716 significantly increased scratching and head shakes compared to acute vehicle administration. Scratching is attenuated in mice chronically treated with THC or THC + CBD. Chronic THC administration also led to significant paw tremor, compared with vehicle or acute SR administration. N = 8 per group. + P < 0.05, significantly different from Veh + Veh; * P < 0.05, significantly different from Veh + SR; # P < 0.05, significantly different from CBD + THC; one‐way ANOVA.

CB₁ receptor antagonist precipitated withdrawal – WIN 55212 and CBD

The ability of an acute injection of the CB₁ receptor antagonist SR141716 (10 mg·kg⁻¹ i.p.) to elicit withdrawal signs following chronic administration of 3.0 mg·kg⁻¹ WIN 55212, 5.0 mg·kg⁻¹ CBD or the combination was tested. Scratching: One‐way ANOVA and Tukey's multiple comparison test showed a significant overall effect of drug treatment and several post hoc significant differences [F(4,38) = 10.2]. The following treatment groups were significantly different from vehicle‐treated mice: vehicle+SR, CBD + SR showed significant increases in scratching behaviour. The following treatment groups were significantly different from SR‐treated mice, suggesting an effect of precipitated withdrawal: WIN+SR, WIN+CBD + SR showed an attenuation of scratching behaviour elicited by SR alone. The WIN+CBD + SR group showed no difference as compared with the WIN+SR group. Paw tremor: One‐way ANOVA and Tukey's multiple comparison test showed a significant overall effect of drug treatment [F(4,38) = 3.20]. The following treatment groups were significantly different from vehicle‐treated mice: veh + SR led to a significant increase in paw tremor behaviour. No treatment groups were significantly different from SR‐treated mice beside vehicle. Head shakes: One‐way ANOVA and Tukey's multiple comparison test showed a significant overall effect of drug treatment and several post hoc significant differences [F(4,38) = 7.39]. The following treatment groups were significantly different from vehicle‐treated mice: vehicle+SR, CBD + SR showed significant increases in head shake behaviour. The following treatment groups were significantly different from SR‐treated mice, suggesting an effect of precipitated withdrawal: WIN+SR, WIN+CBD + SR showed an attenuation of head shake behaviour elicited by SR alone. Lastly, the WIN+CBD + SR group showed no difference in rearing compared with the CBD + SR and WIN+SR groups. Rearing: One‐way ANOVA and Tukey's multiple comparison test showed a significant overall effect of drug treatment [F(4,38) = 7.53]. The following treatment groups were significantly different from vehicle‐treated mice: WIN+SR and WIN+CBD + SR showed significant decreases in rearing behaviour. No treatment groups were significantly different from SR‐treated mice. Lastly, the WIN+CBD + SR group showed no difference in rearing compared with the CBD + SR and WIN+SR groups (Figure 4).

Figure 4.

Effects of SR141716‐precipitated withdrawal from chronic administration of WIN‐55212, CBD or their combination. Acute administration of SR141716 significantly increase scratching, paw tremor and head shakes compared to acute vehicle administration. Scratching and head shakes are attenuated in mice chronically treated with WIN or WIN+CBD. Chronic WIN and WIN+CBD administration also led to significant decreases in rearing. N = 8 per group. + P < 0.05, significantly different from Veh + Veh; * P < 0.05, significantly different from Veh + SR; # P < 0.05, significantly different from WIN+CBD; one‐way ANOVA.

5‐HT_1A receptor antagonist precipitated withdrawal – THC and CBD

One‐way ANOVA showed no significant effects of chronic phytocannabinoid treatment and acute injection of the 5‐HT_1A receptor antagonist WAY 100635 on scratching, head shakes or paw tremors. There was a significant overall effect of treatment on rearing behaviour [F(4,35) = 3.88], but no post hoc significant differences (Figure 5).

Figure 5.

Effects of WAY100135‐precipitated withdrawal from chronic administration of THC, CBD or their combination. Acute administration of WAY100135 produced no effect on scratching, paw tremors, head shakes or rearing and did not precipitate these behaviours in THC, CBD or combination‐treated mice. N = 8 per group. Underneath overarching bracket, * P < 0.05, significantly different overall; one‐way ANOVA.

5‐HT_1A receptor antagonist precipitated withdrawal – WIN55212 and CBD

One‐way ANOVA showed no significant effects of chronic WIN and CBD cannabinoid treatment and acute injection of the 5‐HT1A receptor antagonist WAY 100635 on scratching, head shakes or paw tremors. There was a significant overall effect of treatment on rearing behaviour [F(4,35) = 10.13, P = 0.0001], and several post hoc significant differences. The following treatment groups were significantly different from vehicle‐treated mice: WIN+WAY and WIN+CBD + WAY showed significant decreases in rearing behaviour. These groups were also significantly different from WAY only‐treated mice (Figure 6).

Figure 6.

Effects of WAY100135‐precipitated withdrawal from chronic administration of WIN, CBD or their combination. Acute administration of WAY100135 produced no effect on scratching, paw tremors, head shakes or rearing and did not precipitate these behaviours in WIN, CBD or combination‐treated mice. N = 8 per group. + P < 0.05, significantly different from Veh + Veh; * P < 0.05, significantly different from Veh + SR; # P < 0.05, significantly different from WIN+CBD; one‐way ANOVA.

Elevated plus maze

% time spent in open arm

The ability of the 5‐HT_1A receptor antagonist WAY100635, the TRPV1 receptor antagonist capsazepine or the adenosine A_2A receptor antagonist SCH 58261 to precipitate withdrawal associated anxiogenic behaviour in chronically treated CBD mice was assessed using the EPM. In chronic vehicle‐treated mice, only acute administration of WAY 100635 produced a significant increase in % time in the open arm. Compared with chronic vehicle treatment, chronic CBD treatment also increased % time spent on the open arms. This significant difference was observed in mice pretreated prior to EPM with vehicle, capsazepine or SCH 58261. No significant difference was observed between vehicle and CBD‐treated mice given WAY 100635 prior to EPM (Figure 7).

Figure 7.

Effect of acute antagonist treatment on anxiety‐like behaviour in chronically CBD‐treated mice. (A) Chronic administration of CBD (shaded bars produced significant increases in time spent in the open arm compared to pretreatment matched controls (open bars) with the exception of the WAY pretreatment mice]. (B) WAY and capsazepine pretreatment produced trends toward alterations, but no significant changes, in total arm entries in chronic CBD versus chronic vehicle treated groups. N = 8 per group. * P < 0.05, significantly different from Veh; one‐way ANOVA.

Total arm entries

None of the acute antagonist treatments altered total arm entries on the EPM in chronic vehicle‐treated mice. Chronic CBD treatment also did not alter total arm entries as compared to chronic vehicle treatment (Figure 7B). In chronic CBD‐treated mice administered acute antagonists, WAY 100635 produced a non‐significant decrease (P=0.06) in total arm entries as compared to chronic CBD/acute vehicle treatment as well as chronic vehicle/acute WAY treatment. Also, acute administration of capsazepine produced a non‐significant increase in total arm entries in chronic CBD treated compared with chronic vehicle treated mice.

Discussion

The present studies were designed to test the hypothesis that chronic CBD treatment does not produce cognitive deficits or precipitated withdrawal signs and does not mitigate the cognitive deficits or precipitated withdrawal signs associated with cessation of CB1 receptor agonist use. Several animal studies have been conducted to determine the learning and memory and dependence‐associated effects of the primary phytocannabinoid THC and other mixed CB₁/CB₂ receptor agonists, but fewer studies have addressed the effects of the second most abundant phytocannabinoid CBD alone and in combination with THC. In the present set of experiments, we observed mild effects of acute THC or WIN‐55212 administration on learning and memory, no cognitive impairing effects of CBD treatment alone and little to no evidence that CBD attenuates the learning and memory deficits associated with acute cannabinoid agonist treatment. Similarly, we observed that chronic administration of THC or WIN‐55212 was associated with behavioural signs indicative of precipitated withdrawal following acute administration of the CB₁ receptor antagonist SR‐141716. Chronic administration of CBD was not associated with any behavioural signs indicative of precipitated withdrawal following acute administration of SR141716 or the 5‐HT_1A receptor selective antagonist WAY100635. Chronic co‐administration of CBD with THC or WIN55212 did not alter tolerance to head scratching behaviour seen with CB agonist alone, but did appear to attenuate chronic CB agonist‐associated paw tremors. Taken together, our data suggest that CBD by and large did not attenuate retention deficits or withdrawal behaviours when combined with THC. Additionally, it did not produce these effects on its own. These results are important because Sativex (1:1 THC to CBD ratio) is already in use in many countries and many researchers and lay people have assumed that CBD mitigates some of the adverse effects caused by THC.

We tested the effects of cannabinoid agonists alone and in combination with CBD in an auditory conditional discrimination task with a reversal component and an appetitively motivated spatial reference memory task. Results from our auditory discrimination task revealed that THC at doses of 2.5–10 mg·kg⁻¹ i.p. did not disrupt auditory discrimination following extensive training, nor did it disrupt reversal learning of the task. CBD alone and in combination with THC were also without effect. Previously, Sokolic et al. (2011) reported that THC administration (0.3–10 mg·kg⁻¹ i.p.) disrupted a well‐learned auditory discrimination in Wistar rats, while Mallet and Beninger (1998) reported that visual/auditory discrimination was not affected by 2.0–4.0 mg·kg⁻¹ i.p. THC in Wistar rats (however, subsequent performance on a delayed non‐match to sample reversal of the task was significantly impacted). One explanation for the lack of THC effect on performance in the present task could be related to dose or strain, as THC appears to be relatively less potent in C57Bl6 mice than other rodent models. For example, Varvel et al. (2001) reported that 100 mg·kg⁻¹ THC i.p. was necessary to impair reference memory in male C57Bl/6 mice. This dose of THC also produced significant hypolocomotion and catalepsy which makes it difficult to separate out these effects. One explanation for this relative tolerance could be CB₁ receptor density, as it was shown in one study that CB₁ receptor densities were 25% lower in C57Bl6 mice compared with that of DBA/2 mice. Significant differences in affinity were also observed between the two lines (Hungund and Basavarajappa, 2000). WIN55212 produced a non‐significant trend to disrupt auditory discrimination following extensive training, and this trend appeared to be attenuated by co‐administration of CBD at 2.0 but not 3.0 mg·kg⁻¹ WIN55212. Previous experiments from our laboratory as well as others ( Fan et al., 1994) have shown that this higher dose of WIN55212 produces significant hypolocomotion, so it is unclear whether the trend observed in the discrimination task was specific to cognition or related to motor impairment.

Significant performance‐decreasing effects of THC were observed in the Barnes maze assessing spatial reference memory. THC administration significantly attenuated latency to first target zone entry, the total time spent in the target zone and the number of entries into the target zone. Co‐administration of CBD did not attenuate these effects; the THC + CBD‐treated group looked identical to the THC alone group on all three measures. In addition to these indicators of spatial learning performance, mean speed and total distance travelled were also measured to control for more general changes in behaviour. THC administration at the doses tested did not lead to significant effects on total distance travelled or average speed, and as previously mentioned, significant hypolocomotion was not observed in C57Bl6 mice until 30 mg·kg⁻¹ i.p. in the Varvel et al. (2001) study. CBD administered alone produced a significant increase in total distance travelled and average speed on the Barnes maze. As CBD did not produce a cognitive‐enhancing effect in the maze, we believe that the most likely explanation for this increased locomotor behaviour could be anxiolysis, a behavioural effect that has been associated with CBD treatment in several rodent (see Guimaraes et al., 1990; Moreira et al., 2006; Fogaca et al., 2014) and human (Zuardi et al., 1982; Bergamaschi et al., 2011; Crippa et al., 2011) studies. At the lower dose combinations, THC + CBD tracked the effects of CBD alone, but at the highest combination, THC + CBD tracked the effects of THC alone. In fact for average speed, 20THC + 20CBD showed a non‐significant trend toward decreasing average speed to below baseline and 20 mg·kg⁻¹ THC alone levels. Interestingly, another recent study in rats showed that CBD alone increased locomotor activity but enhanced THC‐induced hypolocomotion (Britch et al., 2017). One explanation for the potentiating effects of CBD on THC is that CBD decreased THC metabolism (Klein et al., 2011; Britch et al., 2017).

Taking together the cognitive performance measurements with the motor performance measurements in the present study, THC produced dose‐dependent deficits in reference spatial learning in the Barnes maze at doses that did not produce overt motor effects. These effects on hippocampal‐dependent spatial learning are similar to those previously observed with the effects of THC in the radial arm maze (Lichtman et al., 1995; Lichtman and Martin, 1996; Wise et al., 2009) and Morris water maze (Da Silva and Takahashi, 2002; Varvel et al., 2001). The combination of THC + CBD did not attenuate these cognitive effects and showed a trend toward producing more performance deficits at the highest combination tested. Hayakawa et al. (2008) reported that combination ratios higher in CBD than THC exacerbated THC‐induced memory deficits on the eight arm radial maze in ddY mice, while Fadda et al. (2004) reported no effect of CBD‐rich extract of Cannabis on the spatial working memory deficits induced by a THC‐rich extract in Lister rats. Wright Jr. et al. (2013) demonstrated in rhesus monkeys that CBD (5.0 mg·kg⁻¹) attenuated the effect of THC (0.5 mg·kg⁻¹) on a visuospatial learning task but not THC‐induced deficits in other behavioural tasks. Lastly, Englund et al. (2013) reported that in healthy human volunteers, pretreatment with 600 mg CBD significantly attenuated episodic memory deficits associated with i.v. administration of 1.5 mg THC. These data suggest that potentiation or antagonism of THC‐induced memory deficits is dependent on several factors including but not limited to (i) the ratio between THC and CBD, (ii) the cognitive domain being assessed, and (iii) the species or strain tested. As has been shown in these studies and others, CBD alone showed no effect on cognitive performance and may show an anxiolytic profile in the task. An additional possibility in previous studies where CBD enhanced effects of THC is that the CBD extract used may contain a small percentage of THC, which at high CBD concentrations may contribute to combination effects. The CBD used in the present study was obtained through the NIDA Drug Supply Program and is >99% pure CBD.

The withdrawal effects following cessation of chronic THC exposure using animal models have been studied since the 1970's (see Wikler, 1976). However, the long half‐life of cannabinoids such as THC and the consequent delay of withdrawal effects make quantifying these effects a challenge. Not surprisingly, the results of abstinence withdrawal studies in laboratory animals have been mixed. The availability of SR141716 was a major advancement toward this research as it allowed researchers to turn from a spontaneous withdrawal model to one of precipitated withdrawal. Determining whether chronic CBD treatment leads to dependence shares this challenge as it shares the pharmacokinetic profile of THC, and an additional challenge is introduced as well. In the precipitated withdrawal model, an appropriate receptor antagonist is administered to a drug‐dependent animal and the antagonist displaces the agonist from the receptor, immediately eliciting withdrawal effects. In the unique case of CBD, direct and indirect effects upon CB₁ receptors are yet to be clearly defined, while direct or indirect effects upon other receptor targets have been more conclusively demonstrated. Therefore in the present study, we sought to determine whether withdrawal signs would be precipitated in chronically treated CBD mice following acute injection of selective antagonists for the CB₁ receptor, 5‐HT_1A receptor, TRPV1 receptor or adenosine A_2A receptor. A second goal was to determine whether chronic co‐administration of CBD potentiated or attenuated withdrawal signs associated with precipitated withdrawal from THC or WIN55212.

Out of the four withdrawal variables measured, changes in head scratching, paw tremor and head shakes were the most consistent behaviours associated with precipitated withdrawal from chronic cannabinoid exposure. Suppression of rearing behaviour appeared to be more indicative of hypolocomotive effects of chronic cannabinoid administration than of precipitated withdrawal. Head scratching is elicited by the CB₁ receptor antagonist SR141716, possibly through interactions with the 5‐HT system (Ward et al., 2009), and the absence of this effect on chronically CB₁ receptor agonist‐treated rodents is believed to reflect tolerance and therefore insensitivity of the CB₁ receptor to the head‐scratch eliciting effects of SR141716 (Cook et al., 1998). In the present study, acute administration of SR141716 failed to elicit head‐scratching behaviour in mice chronically treated with THC, THC + CBD, WIN‐55212 or WIN+CBD. Chronic treatment with CBD alone had no effect on SR41716A‐elicited head scratching behaviour. Acute administration of SR141716 elicited paw tremors in chronic THC, but not chronic WIN55212‐treated mice. This effect was also observed by Cook et al. (1998) and Lichtman et al. (2001). Co‐administration of THC + CBD attenuated this effect in that the combination‐treated group did not show a significant increase in paw tremors, compared with control groups. Acute administration of SR141716 produced a significant number of head shakes, also potentially through its interactions with the 5‐HT system. Chronic administration of THC, CBD or their combination did not lead to tolerance to this effect (THC effect in agreement with Cook et al., 1998). However, chronic administration of WIN55212 or WIN+CBD did lead to tolerance to the acute effect of SR141716 on head shake behaviour. Significant decreases in rearing were observed in mice chronically treated with THC + CBD, WIN55212 and WIN+CBD. As mentioned, because CBD is known to act as a direct agonist at 5‐HT_1A receptors, we also tested the hypothesis that chronic CBD treatment could lead to precipitated withdrawal following acute injection of the 5‐HT_1A receptor antagonist WAY 100635. Acute administration of WAY 100635 produced no overt behavioural effects in mice chronically treated with CBD, THC, WIN55212 or their combinations. Taken together, to our knowledge, this is the first report of the effects of chronic CBD administration alone or in combination with CB receptor agonists on precipitated withdrawal, and the results suggest that dependence to CBD does not occur following chronic treatment and that co‐administration of equal ratios of CBD and THC does not lead to a substantial attenuation of THC dependence.

Because CBD administration is associated with anxiolysis, a final precipitated withdrawal experiment was performed on the EPM with CBD and a range of antagonists for receptors associated with the pharmacological effects of CBD to determine whether withdrawal from chronic CBD exposure would elicit anxiogenic behaviour. This approach has been taken with THC, and Huang et al. (2010) demonstrated that acute administration of SR141716 precipitated an anxiety‐like phenotype on the maze in mice chronically treated with THC. Chronic CBD administration produced significant anxiolysis in mice treated acutely with vehicle, the TRPV1 antagonist capsazepine or the adenosine A_2A receptor antagonist SCH 58261. Chronic CBD followed by administration of WAY 100635 did not produce open arm activity that was significantly different from chronic vehicle/WAY 100635‐treated mice, suggesting that pretreatment with the 5‐HT_1A receptor antagonist blocked the anxiolytic effect CBD. However, quite surprisingly, WAY 100635 produced an anxiolytic profile in the vehicle‐treated mice, so it is difficult to interpret this result. Overall, these results also demonstrate that chronic CBD administration does not lead to dependence and withdrawal.

The present results are important because they add to a growing literature documenting potential antagonistic or additive interactions between the two most abundant phytocannabinoids in the Cannabis plant. The present results are important because they specifically and extensively test the hypothesis that the 1:1 THC‐CBD combination does not produce fewer adverse cognitive or drug withdrawal effects as compared to THC alone. As mentioned above, Sativex is already in use in many countries and is fast‐tracked for late stage clinical trials in the US, based to a large extent on the assumption that CBD mitigates some of the adverse effects caused by THC. While the present results demonstrate that CBD may not increase the safety profile of THC when given in combination, they do support other studies showing a lack of adverse behavioural effects of CBD alone. Therefore, it is also critical to determine the efficacy of CBD alone compared with THC and the CBD + THC combination, to determine optimal cannabinoid‐based treatment strategies for the future. Recently, we have demonstrated, for example, that while both CBD and THC alone prevent development of mechanical sensitivity in a mouse model of chemotherapy‐induced peripheral neuropathic pain, a 1:1 ratio of CBD to THC acted synergistically, leading to an approximate 10‐fold increase in potency in the model (King et al., 2017). Thus, CBD as a monotherapy may represent a safe and effective treatment strategy for neuropathic pain, while co‐administration of equivalent THC + CBD doses may lead to lower doses of each agent being necessary, thereby attaining the goal of decreasing THC‐associated adverse effects.

Author contributions

A.M.M. performed behavioural testing and was involved in statistical analysis and critical review of the manuscript. P.B.S. performed behavioural testing and was involved in statistical analysis and critical review of the manuscript. J.D.F. assisted in the writing of the manuscript. R.F.T. assisted in the conception and design of the work and in critical review of the manuscript. S.J.W. was responsible for the conception and design of the work. She also was responsible for the majority of the statistical analyses and the drafting of the manuscript.

Conflict of interest

The authors declare no conflicts of interest.

Declaration of transparency and scientific rigour

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organisations engaged with supporting research.

Myers, A. M. , Siegele, P. B. , Foss, J. D. , Tuma, R. F. , and Ward, S. J. (2019) Single and combined effects of plant‐derived and synthetic cannabinoids on cognition and cannabinoid‐associated withdrawal signs in mice. British Journal of Pharmacology, 176: 1552–1567. 10.1111/bph.14147.