Ligands of the CB2 cannabinoid receptors augment activity of the conventional antidepressant drugs in the behavioural tests in mice

Although a lot of information can be found on the specific dual role of the endocannabinoid system in the emotional-related responses, little is known whether stimulation or inhibition of the CB receptors may affect the activity of the frequently prescribed antidepressant drugs. Our interests have been particularly focused on the potential influence of the CB2 receptors, as the ones whose central effects are relatively poorly documented when compared to the central effects of the CB1 receptors. Therefore, we evaluated the potential interaction between the CB2 receptor ligands (i.e., JWH133 - CB2 receptor agonist and AM630 - CB2 receptor inverse agonist) and several common antidepressant drugs that influence the monoaminergic system (i.e., imipramine, escitalopram, reboxetine). In order to assess the antidepressant-like effects we used two widely recognized behavioural tests, the mouse forced swim test (FST) and the tail suspension test (TST). Brain concentrations of the tested antidepressants were evaluated by the HPLC method. Intraperitoneal co-administration of per se ineffective doses of JWH133 (0.25 mg/kg) or AM630 (0.25 mg/kg) with imipramine (15 mg/kg), escitalopram (2 mg/kg), and reboxetine (2.5 mg/kg) significantly shortened the immobility time of mice in the FST and the TST, whereas it did not disturb their spontaneous locomotor activity. Furthermore, the brain levels of antidepressants were not changed. Summarizing, the results of the present study revealed that both activation and inhibition of the CB2 receptor function have a potential to strengthen the antidepressant activity of drugs targeting the monoaminergic system. Most probably, the described interaction has a pharmacodynamic background.


Introduction
The endocannabinoid system has attracted clinicians attention for decades as an endogenous homeostatic system associated with a large number of neurotransmitters and their pathways, and implicated in numerous physiological functions, including an inflammatory process, pain, or emotions. The endocannabinoid system consists of the endogenous arachidonate-based lipids referred to as "endocannabinoids" (such as anandamide and 2-arachidonoylglycerol), enzymes responsible for their synthesis (N-acylphosphatidylethanolamine-phospholipase D and diacylglycerol lipase) and degradation (fatty acid amide hydrolase and monoacylglycerol lipase), and the cannabinoid CB receptors (for review see Refs. [1,2]). Thus far, two types of typical CB receptors, encoded by CNR1 and CNR2 genes, have been described. Both CB 1 and CB 2 receptors bind endogenous and exogenous cannabinoids with high affinity. Due to specific mechanism of action of CB receptors, on the one hand the endocannabinoid system can help to restore the brain balance after exposure to different stressors, but on the other hand, it is involved in the pathogenesis of certain mental disturbances, like anxiety, stress, aggressive behavior, or depression. Until recently, only CB 1 receptors were thought to be implicated in the development of mental diseases as the ones that are expressed both in the periphery and the brain. CB 2 receptors were mainly considered as the peripheral receptors. It was believed that they were distributed almost exclusively within the immune cells and thus, that they have only the immunomodulatory role (i.e., regulation of cytokines or migration of immune cells) (for review see Refs. [1][2][3]). However, several research team have found out that CB 2 receptors are also localized in the rodent [4] and human central nervous system [5][6][7]. At first, their presence in the brain was linked to existence of pathological conditions, like Alzheimer's disease [8], multiple sclerosis, amyotrophic lateral sclerosis [9], or tumours [10], but then, they were also identified under physiological conditions [5]. CB 2 receptors are localized in the spinal nucleus, olfactory nucleus, thalamus, hippocampus, amygdala, cerebral cortex, and cerebellum [11,12]. They are mainly distributed post-synaptically, but they can also be found in the presynaptic areas. Functional effects of CB 2 receptors that prove their presence in the brain have also been described in literature. It has been revealed that CB 2 receptors are involved in the inflammatory responses that accompany the neurodegenerative processes [8], in stress-induced neuroinflammation [13], schizophrenia, alcohol preference in mice, alcoholism in humans [14], drug abuse, motor function, and emotionality [4]. Onaivi et al. [4] reported a reduced expression of CB 2 receptors in the striatum and midbrain of mice that developed alcohol preference.
Involvement of CB 2 receptors in the development of depression has also been suggested. Deletion of CB 2 receptors in mice with C57BL/6 J congenic background induced depression-like behavior in the TST [15], whereas genetically-induced lowered CB 2 receptor function (in the Cnr2 heterozygote knock-out mice) was associated with an increased susceptibility to depression when exposed to immune or emotional stressors (i.e., poly I:C or chronic mild stress, respectively) [16]. Reduction of CB 2 receptors in the hippocampus was observed in mice that had been subjected to the chronic unpredictable mild stress [17]. In turn, García-Gutiérrez et al. [17] found out that the genetically-induced overexpression of CB 2 receptors exerted an antidepressant-like effects in the TST and the novelty-suppressed feeding test in mice with swiss ICR congenic background. Moreover, these animals presented a higher level of the brain-derived neurotrophic factor (BDNF) in the hippocampus and they were resistant to stressful depressogenic-like situations in the chronic unpredictable mild stress model. Both stimulation of CB 2 receptors by their selective agonists (GW405833, JWH133, or β-caryophyllene) [18][19][20] as well as inhibition by their selective inverse agonist (AM630) [17,20] produced an antidepressant-like activity in the FST and/or the TST in naïve albino Swiss [20], C57BL/6 [19], ICR Swiss [17] mice, or in rats that underwent chronic constriction injury [18]. Additionally, treatment with CB 2 receptor ligands inhibited the development of the depressive-like behaviour in mice [17] and rats [21] subjected to the chronic unpredictable stress. Ishiguro et al. [14] suspected association between genetic variants of the CB 2 gene and depression in Japanese people.
Although a lot of information can be found on the specific dual role of the endocannabinoid system in the emotional-related responses, little is known whether stimulation or inhibition of CB receptors may affect the activity of the frequently prescribed antidepressant drugs. This subject seems to be quite important, in the view of the fact that more than 322 million people suffer from depression and that the efficacy of available drugs is not satisfactory. Hence, new (even controversial) treatment options are necessary. Our interests have been particularly focused on the potential influence of the CB 2 receptors, as the ones whose central effects are relatively poorly documented when compared to the central effects of the CB 1 receptors. For this purpose, we evaluated the potential interaction between CB 2 receptor ligands (i.e., JWH133 -CB 2 receptor agonist and AM630 -CB 2 receptor inverse agonist) and three common antidepressant drugs that influence the monoaminergic system (i.e., imipraminea tricyclic antidepressant, escitaloprama selective serotonin reuptake inhibitor, and reboxetine a selective noradrenaline reuptake inhibitor). According to the literature data [22], JWH133 has a very high affinity towards CB 2 receptors (i.e., Ki = 3.4 ± 1 nM) and exhibits 200-fold selectivity for them over CB 1 receptors. AM630 also binds selectively to CB 2 receptors, with a CB 2 /CB 1 affinity ratio of 165 and Ki = 31.2 nM at the CB 2 receptor [23]. Being an inverse agonist of CB 2 receptors, AM630 has both affinity towards CB 2 receptors and negative intrinsic activity. It binds to the constitutively activated receptors and decreases their activity below the basal level [24]. Thus, AM630 acts oppositely to JWH133. In order to assess the antidepressant-like effects we used two widely recognized behavioural tests, the mouse FST and the TST.

Animals
The experiments were carried out on male Albino Swiss mice (25-30 g). Animals were maintained in standard cages (8 mice/cage) placed in environmentally controlled rooms (humidity ca. 45-55%, temperature of 22-23°C, 12 h light/dark cycle). They were given water and food ad libitum. All performed procedures were planned in accordance with binding law related to the experimental studies on animal models, and they were approved by the Local Ethics Committee.

Forced swim test (FST)
The FST was performed according to method that we used before [26]. Each mouse was placed individually into glass cylinders (height 25 cm, diameter 10 cm) containing 10 cm of water at 23-25°C and it was left there for 6 min. The immobility time of animals was recorded between the 2 nd and the 6 th min of the test. A given mouse was judged immobile when it stopped struggling and remained floating motionless in the water. Movements necessary to keep animal's head above the water level were acceptable.

Tail suspension test (TST)
The TST was performed according to the method that we used before [27]. Each mouse was suspended by the tail (2 cm from the end of the tail) using adhesive tape and it was left hanging for 6 min. Though the immobility time in the TST can be measured for the whole duration of the test, we recorded the activity of animals for the last 4 min. Similarly to the situation observed in the FST, animals at the beginning of the TST usually show an increased escape-oriented struggling, which lasts ca. 1-2 min [28]. In order to avoid potential confounding of the results, the first 2 min of the test was treated as a habituation period. A given mouse was judged immobile when it stopped moving its body and limbs. Only movements necessary to breathe were acceptable.

Spontaneous locomotor activity
Spontaneous locomotor activity was measured automatically, as was described before [26]. We used an activity meter Opto-Varimex-4 Auto-Track (Columbus Instruments, USA) that contains plexiglas cages with lids (43 × 43 × 32 cm) equipped with a set of four infrared emitters and four detectors monitoring mice movements. Each animal was placed individually in the cage. A distance travelled by a given animal was recorded between the 2 nd and the 6 th min of the test. This time interval corresponded with the one analyzed in the TST and the FST.

Determination of the antidepressant levels in brain homogenates
New cohorts of animals were used for the biochemical studies. The tested mice were decapitated 60 min after injection of the antidepressant drug (with or without JWH133 or AM630). Their brains were dissected and frozen. Imipramine, desipramine (i.e., an active metabolite of imipramine), escitalopram, and reboxetine brain levels were determined with use of the high-performance liquid chromatography (HPLC) method, as was described before [26]. The assays were reproducible with low intra-and interday variation (coefficient of variation < 10%). The extraction efficiencies of the analyzed compounds and the internal standard ranged from 66% to 95%. Concentrations of antidepressants were expressed for the wet brain tissue in ng/g.

Statistical analysis
t-Test or two-way analysis of variance (ANOVA) followed by Bonferroni's post-hoc test were applied, depending on the experimental design. The outcomes were given as the means ± standard error of the mean (SEM). When p < 0.05, differences between the compared groups were treated as significant.

Effects of a concurrent administration of JWH133 and the tested antidepressants in the FST
A single administration of JWH133 (0.25 mg/kg) or a given antidepressant drug (i.e., imipramine -15 mg/kg, escitalopram -2 mg/kg, or reboxetine -2.5 mg/kg) did not significantly influence the immobility time of animals in the FST. When the tested substances were injected in respective combinations, the treated mice were swimming for a longer time than their control counterparts (Fig. 1A). Two-way ANOVA presented: (1)

Effects of a concurrent administration of AM630 and the tested antidepressants in the FST
Concurrent administration of the sub-effective doses of AM630 (0.25 mg/kg) and the antidepressant drug (i.e., imipramine -15 mg/kg, escitalopram -2 mg/kg, or reboxetine -2.5 mg/kg) increased activity of the tested mice in the FST, which was presented in Fig. 2A

Effects of a concurrent administration of JWH133 and the tested antidepressants in the TST
Mice treated with a single injection of JWH133 (0.25 mg/kg) or a given antidepressant drug (i.e., imipramine -15 mg/kg, escitalopram -2 mg/kg, or reboxetine -2.5 mg/kg) stayed immobile for the same duration of time as the animals from the control group. However, when the tested substances were injected in respective combinations, the antidepressant-like effect was detected in the TST (Fig. 1B). The following significant drug-drug interactions were revealed by two-way ANOVA analysis: (1)

Effects of a concurrent administration of AM630 and the tested antidepressants in the TST
After a single administration of AM630 (0.25 mg/kg) or a given antidepressant drug (i.e., imipramine -15 mg/kg, escitalopram -2 mg/ kg, or reboxetine -2.5 mg/kg), behavior of the treated mice in the TST was similar to the one observed for animals from the control group. By contrast, co-administration of AM630 with a given antidepressant drug significantly reduced the immobility time of animals (Fig. 2B). Twoway ANOVA demonstrated: (1)  3.5. Effects of a concurrent administration of JWH133 or AM630 and the tested antidepressants on the spontaneous locomotor activity of mice As presented in Table 1, neither a single administration of the tested CB 2 receptor ligands or antidepressants nor their respective combinations influenced the spontaneous locomotor activity of animals.

Pharmacokinetic studies
Outcomes of the pharmacokinetic studies demonstrated that none of the tested CB 2 receptor ligands (i.e., JWH133 -0.25 mg/kg or AM630 -0.25 mg/kg) elevated the brain levels of imipramine, desipramine, escitalopram, and reboxetine (Table 2). t-test revealed the following re-

Discussion
In the course of intensive studies over the pathomechanism of depression, the endocannabinoid system has emerged as an important item in the development and/or treatment of this disease. Medical and scientific bases provide discrepant information on the cannabinoids effects in mood disorders. Both the pro-depressive and antidepressant activity have been recorded after cannabis consumption, and these effects use to be attributed exclusively to CB 1 receptors. However, quite recently it has been proved that CB 2 receptors can also be found in different brain areas, including those responsible for emotions, like the prefrontal cortex, hippocampus, or amygdala [4,11,17,29]. Thus, CB 2 receptors seem to be also involved in mood changes in humans, though their impact is still relatively poorly documented. Several authors suggest that CB 2 receptors could be an interesting novel target for the treatment of depressive disorders, including the post-stroke depression [17,21]. The antidepressant potential of JWH133 (0.5-1 mg/kg)the CB 2 receptor agonist applied in our study, was observed by Kruk-Słomka et al. [20] in the FST in Swiss mice. The authors demonstrated that the JWH133 treatment exerted the U-shaped anti-immobility pattern, with ineffectiveness of the lowest (0.25 mg/kg) and the highest (2 mg/kg) tested doses. As for the AM630the CB 2 receptor inverse agonist, results from the behavioural studies are controversial. In the abovementioned experiments carried out by the team of Kruk-Słomka [20], only its lowest tested dose (i.e., 0.5 mg/kg) was effective in the FST, whereas the concentrations of 1-3 mg/kg did not increase the swimming time of animals. Quite contrary were the outcomes reported by García-Gutiérrez et al. [17] who demonstrated that an acute administration of 1-3 mg/kg of AM630 produced an antidepressant-like activity in the FST, but only in the wild type of swiss ICR mice; these doses did not change the behaviour of animals with experimentally-induced overexpression of the CB 2 receptors. In fact, the authors revealed that after an acute and chronic administration of AM630 different effects in the biochemical and behavioural tests measuring the depression level in rodents may be recorded, depending on the applied experimental conditions. A 4-week administration of this CB 2 receptor ligand in the stressed wild-type ICR Swiss mice prevented the development of depression-like behaviour measured by the TST and by the sucrose intake test, whereas such a treatment had no impact on the behaviour of the non-stressed wild-type animals. Furthermore, the AM630 therapy per se did not influence the expression of the CB 2 receptor, BDNF gene, or the BDNF protein level in the hippocampus, but this CB 2 receptor ligand significantly reduced the BDNF loss and reversed the reduction in CB 2 receptor levels induced by the unpredictable chronic mild stress [17]. To our knowledge, the present study is the first one that assessed the influence of CB 2 receptor ligands on the activity of conventional antidepressant drugs that affect the monoaminergic system. We examined the impact of co-administration of CB 2 receptor agonist (JWH133) or CB 2 receptor antagonist (AM630) and imipramine (a tricyclic antidepressant), escitalopram (a selective serotonin reuptake inhibitor), or reboxetine (a selective inhibitor of noradrenalin reuptake) on the depression-related behaviour of mice in the FST and the TST, i.e. two most widely used animal tests evaluating the antidepressant-like activity. We demonstrated that JWH133 and AM630 when administered at per se inactive dose of 0.25 mg/kg significantly intensified the antidepressant effects of the applied drugs (also given at the sub-effective doses: 15 mg/kg of imipramine, 2 mg/kg of escitalopram, and 2.5 mg/kg of reboxetine), suggesting an addition or synergism of action. Mice that received single injections of respective combinations (imipramine-, escitalopram-, or reboxetine-CB 2 receptor ligand) were more mobile in the FST and the TST than animals from the control groups. According to the literature data [30,31] this observation can be treated as an indicator of the antidepressant behaviour. Similarly to our observations, both JWH133 and AM630 when given at a non-effective dose of 2 mg/ kg significantly potentiated the antidepressant activity of an antagonist of the muscarinic acetylcholine receptors, i.e. scopolamine, in the FST in Swiss mice [20]. Though there are reports that CB 2 receptor agonists and antagonists may alter spontaneous locomotor activity of animals in a strain-and gender-dependent manner [4], we did not notice any differences in locomotion of mice from the tested and control groups. Similarly, in experiments by Kruk-Słomka et al. [20] no stimulant/inhibitory action of JWH133 (0.25-2 mg/kg) or AM630 (0.5-3 mg/kg) was noticed. Consequently, it is improbable that the antidepressant-like effects observed in the behavioural tests in the present study were false positive.
Generally, the molecular mechanism of the antidepressant-like activity of CB receptor ligands seems to be very complicated and complex. It has not been fully understood yet. Strangely enough, both CB 1 and CB 2 receptor agonists and antagonists may exert the antidepressant-like activity in rodents as well as they can induce the pro-depressive behaviour in animals. Moreover, it was found out that administration of the inactive dose of both CB 2 receptor antagonist (AM630, 2 mg/kg) and CB 1 antagonist (AM251, 0.25 mg/kg) is able to abolish the antidepressant-like effects of the active doses of CB 2 receptor agonist (JWH133, 0.5 and 1 mg/kg) in the FST in naïve Swiss mice. Similarly, in the same strain of rodents, administration of the non-effective dose of both CB 2 receptor antagonist (AM630, 2 mg/kg) and CB 1 antagonist (AM251, 0.25 mg/kg) is able to reverse the antidepressant-like effects of the active doses of CB 1 receptor agonist (oleamide, 10 and 20 mg/kg) [20]. In fact, Onaivi et al. [4] noted that in different neuronal populations CB 1 and CB 2 receptors work to a certain point independently, but to a certain point also cooperatively in regulation of diverse physiological activities affected by cannabinoids. Viewing the above-mentioned, it should not be surprising that both JWH133 and AM630, i.e., the compounds that should act oppositely, potentiated the activity of the tested antidepressants in our study. Most probably, elevation of the serotonin and/or noradrenaline levels leading to potentiation of the serotonergic and/or noradrenergic neurotransmissions was responsible for the observed effects. Both the tested antidepressant drugs as well as cannabinoids are believed to modulate the monoaminergic system [32]. Table 1 Effect of a combined intraperitoneal administration of (A, B) JWH133 or (C, D) AM630 and antidepressant drugs on the spontaneous locomotor activity of mice. Imipramine (15 mg/kg), escitalopram (2 mg/kg), and reboxetine (2.5 mg/kg) were injected 60 min before testing, whereas JWH133 (0.25 mg/kg) and AM630 (0.25 mg/kg) were given 30 min before the experiment. The values represent mean ± SEM (n = 7-8 mice per group). Two-way ANOVA was used for statistical analysis.

Table 2
Effect of JWH133 and AM630 on the brain levels of antidepressants in mice. Franklin et al. [33] demonstrated that CB 2 receptors are implicated in the cannabinoid-induced upregulation of the serotonergic 5-HT 2A receptors in the brain. This effect is most probably mediated by internalization of CB 2 receptors and stimulation of the ERK1/2 signalling pathway. Disturbances in the 5-HT 2A receptor-dependent signalling are involved in the pathomechanism of different neuropsychiatric disorders, like schizophrenia, anxiety, or depression [34,35]. It is also possible that CB 2 receptors modulate the depression-related responses via alteration in immune system. Their increased expression in the brain was observed after an intra-striatal administration of lipopolysaccharide in a rat model of Parkinson's disease and it corresponded with an activation of the microglia [36]. In experiments by Zoppi et al. [13], activation of CB 2 receptors by experimentally-induced overexpression or administration of an CB 2 receptor agonist in ICR Swiss mice exerted the anti-inflammatory or neuroprotective effects via control of glutamate uptake, whereas CB 2 receptors knock-out potentiated stress-induced neuroinflammatory responses. A specific role of neuroimmune interactions in the development of psychiatric disorders has been examined for decades and a range of diverse studies have demonstrated a clear link between chronic immune responses and occurrence of depression [37]. Stimulation of CB 2 receptors in the brain is also known to reduce expression of inducible nitric oxide synthase, production of nitric oxide, generation of reactive oxygen/nitrogen species, and thus, to attenuate the oxidative stress [38]. Additionally, neither JWH133 nor AM630 are exclusively specific for CB 2 receptors, but they are only CB 2 receptor-selective. They also exhibit partial affinity for CB 1 receptors [39,40]. Therefore, pathways connected with CB 1 receptors may also contribute to the effects detected in our experiments.
Since the pharmacokinetic assays showed that neither JWH133 nor AM630 enhanced the brain levels of the tested antidepressant drugs (i.e., imipramine with its active metabolitedesipramine, escitalopram, or reboxetine), it should be assumed that the potentiation of the antidepressant effect observed in our study took place in the pharmacodynamic phase instead of in the pharmacokinetic one. However, based on the experiments of Smaga et al. [41], it is highly unlikely that an acute administration of the antidepressant drugs applied in our study might have changed the expression of cannabinoid receptors (either CB 1 or CB 2 ) in the mouse brain. Such effects couldn't be ruled out after prolonged (14-day) treatment. Reduced levels of CB 1 and/or CB 2 receptors were detected in the dorsal striatum, nucleus accumbens, and/ or cerebellum of male Wistar rats exposed to imipramine (15 mg/kg), whereas increased levels of CB 1 and/or CB 2 receptors in the hippocampus and/or prefrontal cortex with decreased levels of CB 1 receptors in dorsal striatum were recorded after escitalopram treatment (10 mg/ kg) [41]. In the view of the above, further studies over the molecular mechanism of the interactions demonstrated in our experiments are needed.

Conclusions
The results of the present study revealed that both activation and inhibition of the CB 2 receptor function have a potential to strengthen the antidepressant activity of drugs targeting the monoaminergic system. Most probably, the described interaction has a pharmacodynamic background. Though our findings should be treated as the preliminary ones, we assume that the modulation of the endocannabinoid neurotransmission via CB 2 ligands could be a promising strategy as an adjuvant treatment in patients with depression. Such an approach could have additional benefits, since the CB 2 selective ligands are thought to be devoid of central nervous system side effects that may limit the clinical use of other substances with antidepressant potential [19].

Funding
This study was supported by Funds for Statutory Activity of the Medical University of Lublin, Poland. This research did not receive any specific grant from funding agencies in the public, commercial, or notfor-profit sectors.