The present study demonstrates that the central ghrelin signalling system, involving GHS-R1A, is required for indirect measures of the rewarding/reinforcing effects of the cannabinoid/CB1 agonist/WIN55,212-2 that participate in cannabinoid addiction. To our knowledge, our results, for the first time, document that GHS-R1A antagonism significantly reduced the cannabinoid-induced dopamine efflux in the nucleus accumbens shell (NACSh). A low dose of WIN55,212-2 (2.4 mM/0.5 µL) administered into the posterior ventral tegmental area provoked significant accumbens dopamine release together with extracellular endocannabinoids, anandamide/AEA and 2-AG, increase and transient GABA decrease, and all these observed cannabinoid-induced accumbens changes were significantly reduced by the GHS-R1A antagonist/JMV2959 3 mg/kg i.p. pretreatment 20 min before the cannabinoid. In addition, the WIN55,212-2-induced accumbens dopamine turnover was significantly increased by the JMV2959 pretreatment. Further, the behavioural stimulation observed within 20–40 min after the 0.1 mg/kg WIN55,212-2 intraperitoneal administration was significantly reduced by the JMV2959 3 mg/kg pretreatment. These results indicate substantial participation of ghrelin/GHS-R1A mechanisms in the rewarding/reinforcement effects of the CB1R agonist/synthetic cannabinoid WIN55,212-2, which are involved in cannabinoid addiction.

The cannabinoids/CB1R agonists, including WIN55,212-2, are known for their biphasic/dual effects, with low doses inducing the locomotor hyperactivity and with higher doses the hypoactivity. Behavioural stimulation, frequently observed in drugs of abuse, is considered to be a sign of dopaminergic nigrostriatal pathway activation and constitutes to part of the addiction process associated with drugs’ reinforcing effects, hence, contributing to drug dependence [46,47,48,49]. It was described that the low cannabinoid/WIN55,212-2 dose (0.1 mg/kg i.p.) increased the horizontal and vertical activity in habituated rats [47,48]. Also, in our experiment in the LABORAS cage, we observed a significant increase in locomotion, rearing, distance travelled and the overall average speed of behaviour in comparison to the vehicle treated group, within 20–40 min after WIN55,212-2/vehicle administration, when the initial explorative locomotor activity of rats declined and the advanced stimulatory effects of the cannabinoid were unmasked. The JMV2959 (1 and 3 mg/kg) administration 20 min before the cannabinoid (simultaneously starting the habituation period in the LABORAS cage), significantly reduced the monitored cannabinoid-stimulated behavioural parameters; however, only when the higher dose 3 mg/kg JMV2959 was used (p < 0.001). These results are in accordance with previous studies, when JMV2959 in mice or rats reduced locomotor hyperactivity induced by alcohol [50], nicotine [51], cocaine and amphetamine [52], morphine [53,54] and fentanyl [44]. When JMV2959 was administered alone (in the same doses), no significant influence on rat behaviour was observed within the relevant period in LABORAS, which corresponds well with our former study in activity cage/open field (these results are summarized in Table S3 in the Supplementary Materials) [55].

Cannabinoids including the main cannabis constituent tetrahydrocannabinol (THC) and synthetic cannabinoids (e.g., WIN55,212-2) most likely mediate their pleasurable, anxiolytic and rewarding/reinforcing effects through the CB1Rs located within the central brain reward circuits, particularly the VTA and the NAC [3,10,11]. Cannabinoids activate these reward pathways in a manner that is consistent with other drugs of abuse, including triggering dopamine release in the nucleus accumbens shell (NACSh) [1,2,3,19,23,25,56]; however, particular mechanisms through which this occurs seem to be more specific; it would appear that THC (and probably other exogenous cannabinoids) may have somewhat pathway-dependent effects in this brain reward circuit (further research is needed) [57,58]. The cannabinoid mechanisms of effect and the endocannabinoid system involvement are described in more detail in the Supplementary Materials.

Research studies that used cannabinoid experimental addiction models discovered high dose-, strain/genetic-, drug- and brain substructure-sensitivity/dependence that can crucially influence the obtained results (e.g., see reviews [22,26,56]. The reinforcing effects shift rapidly to aversive effects with increasing dose [2,3]. Systemic cannabinoid/THC/WIN55,212-2-induced CB1R-dependent dopamine release was observed specifically in the NACSh and varied depending on the strain of rats assessed [23,25,59]. The THC was intra-cerebrally self-administered in rats, specifically into the posterior VTA and the NAC shell substructures but not into the anterior VTA or the NAC core [45,56]. Likewise, THC administration into the anterior VTA did not produce any dopamine release in rat NACSh (in vivo microdialysis) [60].

In accordance with the findings mentioned above and in the Supplementary Materials, in our in vivo microdialysis study, local administration of the low 2.4 mM/ 0.5 µL dose of WIN55,212-2 into the posterior VTA induced significant dopamine efflux in the NACSh of Wistar rats [1,2,3,19,45,48]. The intracerebral 2.4 mM dose of WIN55,212-2 induced dopamine release with a maximum of 131% of the baseline mean, which approximately corresponds with published microdialysis experiments when WIN55,212-2 was administered systemically (i.p., i.v.) and the dopamine release was prevented by CB1 antagonist (rimonabant) co-administration [25,48].

The intra-VTA administration of WIN55,212-2 also increased accumbens extracellular concentrations of dopamine degradation metabolites: 3-methoxytyramine/3-MT (a maximum of 114% of the baseline mean), 3,4-dihydroxyphenylacetic acid/DOPAC (133%) and homovanillic acid/HVA (142% of the baseline mean). The maximal microdialysate concentration values (40 and 60 min intervals after WIN55,212-2 administration) were used for calculation of the extracellular dopamine metabolic turnover ratios. The cannabinoid application significantly increased the extracellular dopamine metabolic turnover HVA/DA ratio (p < 0.05) in comparison with the control/vehicle + saline group. The 3-MT/DA ratio significantly decreased (p < 0.001) and the DOPAC/DA ratio increased (p < 0.001). Increases in dopamine metabolism, measured with the DOPAC/DA ratio, have been reported in most but not all rodent studies with cannabinoids [19,61,62]. Some resulting inconsistencies within the early ex-vivo studies could be explained due to the technical limitations in detecting the rapid changes in extracellular dopamine/metabolite concentration, detectable by microdialysis techniques used more recently [19].

Pretreatment with the GHS-R1A antagonist/JMV2959 3 mg/kg i.p. 20 min before WIN55,212-2 significantly reduced the cannabinoid-induced dopamine efflux in the NACSh. The same 3 mg/kg i.p. JMV2959 dose “per se” did not influence accumbens dopamine, as it has already been described [50,53,63]. Also, the lower JMV2959 dose 1 mg/kg alone did not induce significant changes in accumbens dopamine/metabolites (see Figure S1 in the Supplementary Materirals). The observed GHS-R1A antagonist effect is in accordance with previous studies in mice/rats experimental models with alcohol [50], stimulants/cocaine, amphetamine [52], nicotine [51] and opioids/morphine and fentanyl [53,54,63]. Contrary to alcohol and stimulant models, the JMV2959 pretreatment did not completely abolish the cannabinoid-induced accumbens dopamine increase. The dopamine extracellular concentrations remained transiently significantly increased (108% of the baseline mean) in comparison with baseline and the vehicle + saline group, similar to studies with opioids, when the JMV2959 pretreatment effects were weaker [53,54,63]. However, in the present study, the JMV2959 + WIN55,212-2-induced transient dopamine increase (40–80 min intervals) was changed to transient moderate but significant decrease (in the 120 min interval; maximum of 86% of the baseline mean, p < 0.001), which was not observed in previous studies. The observed gradually intensified JMV2959 pretreatment effect could indicate some modulatory mechanisms of the GHS-R1A antagonism specifically in combination with the cannabinoid. However, in contrast to previous studies, when the drugs were administered systemically (s.c. or i.v.), in this case the cannabinoid was injected directly into the VTA, which also may impact the effect proportions. Nevertheless, the GHS-R1A antagonism significantly reduced the cannabinoid/CB1R-induced accumbens dopamine efflux, the trigger of reinforcement effects, which indicates the important involvement of the ghrelin/GHS-R1A system in the cannabinoid reinforcement effects, hence, addiction processing.

The appropriate specific involved mechanisms have to be further researched. The CB1Rs are located on various presynaptic axons/inputs in the VTA and the NAC [3,10,11]. Concerning the GHS-R1As, it was described that intra-VTA administered ghrelin induced accumbens dopamine release and the intra-VTA GHS-R1A antagonist supressed this effect [37,64]. The GHS-R1A is expressed on dopaminergic neurons [29], and it has been suggested that GHS-R1A regulates the activity of VTA dopamine neurons via heterodimerization of the GHS-R1A to the D1R (especially with stimulants), as well as by the strong constitutive activity of the GHS-R1A [31,65]. Further, the midbrain GHS-R1As, co-localized with dopaminergic and cholinergic receptors [35,36], activate the cholinergic–dopaminergic reward link and trigger the VTA dopaminergic signalling, which can be blocked by intra-VTA administration of nicotinic cholinergic and glutamatergic receptor antagonists [37,38,66]. Also, GHS-R1A is known to modulate synaptic plasticity in the brain [67,68]. The GHS-R1As are also expressed on the GABA VTA neurones [67] however, the ghrelin-induced activation of VTA dopamine neurons was not blocked in presence of GABA-A antagonist [39]. The CB1Rs and the GHS-R1As are evenly expressed in both the VTA and the NAC and also in hypothalamus. Substantial functional cooperation between the ghrelin and cannabinoid systems in regulation of food intake has been mentioned above [27,28,29,30,67]. Dimerization between CB1R and GHS-R1A has not been directly examined yet, but it was suggested [34]. In the hypothalamus, the GHS-R1A is present at GABAergic presynaptic terminals, and it attenuates GABA release in the hypothalamic neurons [69]. CB1-dependent long-term depression (LTD) in the VTA GABA neurons induced by cannabinoid/THC has been described [70]. The existing findings propose several possible modes of particular ghrelin/cannabinoid interaction mechanisms (e.g., on GABA axons or possibly on glutamate or cholinergic transmission (what we did not test)); however, further research is necessary.

The accumbens extracellular dopamine metabolic turnover increase, induced by intracerebral WIN55,212-2 during its maximal effect (40 and 60 min intervals) in the present study, was significantly augmented after JMV2959 pretreatment in comparison to the WIN55,212-2 + saline group, measured with metabolic ratios HVA/DA, DOPAC/DA and 3-MT/DA (p < 0.001). The substantial increase in final metabolite HVA formation (HVA/DA ratio) was produced by enhanced DOPAC production (DOPAC/DA ratio). This is in accordance with our previous opioid microdialysis studies when a significantly increased formation of DOPAC and HVA in the NACSh after JMV2959 pretreatment before morphine and fentanyl was observed [54,63]. Thus, a significant impact of JMV2959 pretreatment was observed on cannabinoid- and opioid-provoked accumbens dopamine metabolism due to the monoamine oxidase (MAO) (the significant DOPAC and DOPAC/DA increase). Moreover, the relatively high 6 mg/kg i.p. JMV2959 dose, when administered alone, moderately but significantly increased DOPAC and HVA accumbens levels in rats [54]; we did not find any significant influence of single JMV2959 3 mg/kg on the accumbens dopamine/metabolite concentrations [54,63]. Altogether, it seems that GHS-R1A antagonism might be associated with increased metabolism of dopamine by MAO.

Simultaneously with the dopamine release, administration of WIN55,212-2 into the posterior VTA also induced the increased extracellular concentrations of anandamide and 2-AG in the NACSh. The maximum effects were observed in the same period as with dopamine efflux (2-AG, maximum of 120% of the baseline mean in the 60 min interval) or 20 min later (AEA, maximum of 164% of the baseline mean in the 80 min interval). The rapid extracellular endocannabinoids increase in the NACSh induced by the CB1R agonist-provoked dopamine release (possibly modulated by other factors) was expected in accordance with the in-detail described mechanisms in the Supplementary Material [3,19,58].

The 3 mg/kg i.p. JMV2959 pretreatment moderately but significantly reduced the cannabinoid-induced accumbens AEA increase (by a maximum of 12.7% of the baseline mean) as well as the 2-AG elevation (by a maximum of 5.5% of the baseline mean), but the endocannabinoid concentrations remained significantly higher relatively to the vehicle + saline group from the 40 min interval until the end of microdialysis (140 min interval). As in previous studies, the 3 mg/kg i.p. JMV2959 dose did not influence the accumbens AEA or 2-AG concentrations when it was administered alone [43,44]. The JMV2959 effect on the WIN55,212-2-induced accumbens dopamine release (described above) seems more pronounced than its influence of endocannabinoids. In our previous studies with opioids, we observed a massive effect of the ghrelin antagonist on fentanyl/morphine-induced accumbens AEA increase, which was reversed to decrease due to the JMV2959 pretreatment (administered intraperitoneally as well as into the VTA and the NACh). On the contrary, the opioid-induced 2-AG decrease in the NACSh was intensified by the JMV2959 pretreatment (with all mentioned administration modes) [44,63]. However, cannabinoids and opioids influence diversely the endocannabinoid modulations within the NAC [40,41].

In the present study, the cannabinoid-induced increase in accumbens AEA/2-AG levels were reduced moderately but consistently by the JMV2959 pretreatment and several mechanisms can be considered. The JMV2959 pretreatment significantly reduced the accumbens dopamine efflux, an important trigger of both endocannabinoids accumbens formation and release (from the MSNs or FSI as described in the Supplementary Materials) [3,58]. Further, the GHS-R1As are expressed within the NAC (e.g., on dopaminergic and also GABAergic neurons) [35,39]; thus, their antagonism could affect the relevant neural sensitivity, including the dopamine-independent mechanisms. The strong constitutive activity of the GHS-R1A should be also considered [31], although, the GHS-R1A inverse agonist would play a more important role. The dimerizations of GHS-R1A with D1R and D2R has already been confirmed [34]. Endocannabinoids are presumably produced by both accumbens iMSNs (D2R) as well as dMSNs (D1R), but most important endocannabinoid retrograde modulations were identified in the iMSN synapses [3,58]. We can only consider AEA/2-AG accumbens extracellular concentrations, because the consequent modulations were not investigated in our study. The GHS-R1A/D1R dimerization generally amplifies the D1R signalling [34,65]; thus, the GHS-R1A antagonist would reduce it. One of the possible consequences might be lower dMSN endocannabinoid production. The presence of ghrelin (GHS-R1A/D2R dimerization) appears to switch the D2R-intracellular signalling cascade [71]; therefore, the GHS-R1A antagonist would possibly attenuate this effect, so impact on AEA/2-AG production by iMSNs is possible but unclear. Moreover, the glutamate and acetylcholine involving interactions might participate in the endocannabinoid changes, which were not studied. Consequently, it seems that the GHS-R1A antagonism reduces the endocannabinoids modulatory effects induced by CB1R agonist in the NACSh due to the fact of their decreased availability, but further research is necessary to clarify the antagonists’ particular involvements.

The WIN55,212-2 intra-VTA administration also induced a significant decrease in the γ-aminobutyric acid/GABA extracellular concentrations in the NACSh (maximum of 79% of the baseline in an 80 min interval), which was normalised within the last 140 min interval. The accumbens GABA decrease was presumably a consequence of the endocannabinoid presynaptic modulations (DSI and LTD) within the NAC, where CB1Rs are located on both the MSN collaterals and on intrinsic GABA neuron terminals, just as in the VTA [20,70]. The CB1R agonists (CP55940, THC) were found to decrease the accumbens GABA release [20,72]. It was suggested that the (endo)cannabinoid-induced inhibition of GABA release onto the MSNs in the NAC might increase/cause the disinhibition effect of the GABAergic output to the VTA and thus further participate on the cannabinoid reinforcement effects [20].

The JMV2959 pretreatment reversed the cannabinoid-induced accumbens GABA decrease to transiently significant increase (maximum of 107% of the baseline mean in the 80 min interval), which again was normalised within the last 140 min interval. The single 3 mg/kg JMV2959 dose did not significantly influence the accumbens GABA levels as in previous experiments [44]. In our study with opioids, the fentanyl-induced accumbens GABA enhancement was significantly reduced by the JMV2959 pretreatment [44]; however, accumbens GABA levels are influenced by opioids and cannabinoids in different ways [73,74]. The GHS-R1As are expressed on GABA neurons in the VTA, the NAC and elsewhere in the brain [67]. Important interactions between ghrelin and GABA systems have been implicated within the CeA [75] and the hypothalamus [69]. The GHS-R1As are located on the hypothalamic GABAergic presynaptic terminals; the GHS-R1A constitutive activity itself and administration of the GHS-R1A agonist attenuated GABA release. Presumably, the GHS-R1A antagonist and/or inverse agonist would reduce/abolish this effect. Similar interaction can be presumed within the NAC/VTA [67]. It seems that ghrelin and endocannabinoids show similar synaptic plasticity mechanisms, so they might cooperate on GABAergic neurons, where they are both located on the presynaptic axons; moreover, the CB1/GHS-R1A dimers were indicated [34]. Further investigation is necessary for clarifying the particular involved mechanism.

It should also be mentioned that literary data suggested that an intact ghrelin signalling pathway is necessary for cannabinoid-induced food intake. Administration of CB1R agonist in the GHS-R1A knockout mice did not influence the food intake contrary to the wild/control mice, which showed a trend to increase food consumption with higher doses of HU210 injection [30]. In our studies, the food was removed after the drug administrations during the experiments; thus, we could not observe the single JMV2959 dose effects on food intake. No differences were observed in daily food consumption among the rat groups.

Collectively, our results demonstrate that the GHS-R1 antagonist significantly reduced the CB1R agonist/WIN55,212-2-induced behavioural stimulation, the dopamine and endocannabinoid (AEA, 2-AG) release in the NACSh and reversed the accumbens GABA decrease. Also, it simultaneously increased the extracellular dopamine accumbens metabolism through MAO. In our further recent study (in preparation for this Special Issue), we have found that the same GHS-R1A antagonist (JMV2959) significantly reduced the WIN55,212-2 intravenous self-administration (IVSA) and tendency to relapse, the THC-induced condition place preference (CPP) development as well as expression and reduced the THC-induced behavioural stimulation. Altogether, our results suggest substantial involvement of ghrelin/GHS-R1A central signalling in the cannabinoid rewarding/reinforcement pro-addictive effects, which encourages further investigation of the GHS-R1A antagonism as a potential approach to cannabinoid addiction treatment.