It is generally accepted that cannabinoid agonists induce a dose-dependent inhibition of motor activity in both humans and laboratory animals and even cause catalepsy with high doses (Gerdeman et al., 2008). In line with previous studies (Sañudo-Peña et al., 2000; Drews et al., 2005; Rodvelt et al., 2007; Polissidis et al., 2010; 2013), our results showed that treatment with the nonselective CB1R agonist WIN55,212-2 dose-dependently suppressed ambulatory activity after single and multiple administrations.

At the neurobiological level, increases in mesostriatal dopamine neurotransmission have been associated with increased spontaneous locomotor activity after treatment with a variety of substances of abuse (Nestler, 2005; Lüscher and Malenka, 2011). However, most of the CNS actions of cannabinoids are mediated via CB1R (Gardner, 2005; Hashimotodani et al., 2007; Vlachou et al., 2008), and activation of these receptors in the striatum is associated with inhibition of motor behaviors (Giuliani et al., 2000; Darmani, 2001; Järbe et al., 2002; Schramm-Sapyta et al., 2007). Several studies have implicated that the mechanism of this CBR-induced hypomotility involves the interaction between CB1R and D1DR/D2DR dopamine receptors at the cellular level and their coupling to the same affector system (Glass and Felder, 1997; Hermann et al., 2002; Andersson et al., 2005; Kearn et al., 2005; Martín et al., 2008).

Importantly, our results indicated that chronic WIN55,212-2 administration did not induce phenomena of tolerance or sensitization of locomotor activity, although the motor-suppressant effects of the highest dose of WIN55,212-2 tested appeared to level off between days 10 and 20 of testing. Indeed, while the results of the present study demonstrated that chronic WIN55,212-2 can produce acute and sustained motor-suppressant effects, they provide no evidence that repeated administration of WIN55,212-2 leads to a behavioral sensitization similar to that observed with other drugs of abuse. Similar results have been obtained after chronic treatment with low doses of Δ⁹-THC or other CB1R agonists (Arnold et al., 1998; Muschamp and Siviy, 2002; Ellgren et al., 2004; Kolb et al., 2006; Varvel et al., 2007), although considerably higher doses of Δ⁹-THC have been reported to produce behavioral sensitization (Cadoni et al., 2001; Rubino et al., 2001). Although our data did not indicate significant tolerance to the motor-suppressant effects of chronic WIN55,212-2, a study by Sim-Selley and Martin (2002) reported that chronic administration of considerably higher doses of WIN55,212-2 in mice produces tolerance to cannabinoid-mediated hypoactivity. Similar results have also been observed with higher doses of Δ⁹-THC (Rodríguez de Fonseca et al., 1994; Romero et al., 1997; Sim-Selley and Martin, 2002; Whitlow et al., 2003). Thus, it appears that behavioral tolerance or sensitization after chronic WIN55,212-2 may depend on several aspects of the experimental protocol, such as species used, dose, and route of drug administration.

The present study indicated that chronic administration of WIN55,212-2 for 20 consecutive days modulated the dopaminergic and GABAergic systems of adult rat brain. In particular, we observed selective alterations in binding and mRNA levels of the DAT as well as dopamine and GABA_A receptors in somatodendritic and terminal regions of the mesostriatal dopaminergic system (SN and striatum; VTA and NAc). Considering that only high doses of WIN55,212-2 significantly affected motor activity, the extent to which the observed neurochemical alterations are causally related to the locomotor activity findings in response to chronic administration of WIN55,212-2 is not readily apparent. It may be argued that changes in dopaminergic and GABAergic neurotransmission contribute to the sustained suppressant effects measured after high doses of WIN55,212-2.

In the present study, the use of the nonselective CB1R agonist WIN55212-2 could raise the question whether the observed alterations are attributed to CB1R or CB2R activation. CB2 cannabinoid receptors are localized primarily in immune cells both in the periphery (Griffin et al., 2000) and brain microglia (Maresz et al., 2007), indicating that they are activated after brain damage or injury. In addition, they are expressed in neurons, including the striatum and midbrain (Gong et al., 2006), but the extent and level of expression remain controversial (Atwood and Mackie, 2010). However, most of the CNS actions of cannabinoids appear to be mediated via CB1R (Gardner, 2005; Hashimotodani et al., 2007; Vlachou et al., 2008). Systemic administration of CB1R agonists such as ⁹Δ-THC and WIN55,212-2 suppress motor activity, especially at higher doses (Darmani, 2001; Polissidis et al., 2013), and this effect is reversed by CB1R specific antagonists.

It is well known that DAT is localized to plasma membranes of axon terminals as well as dendrites of SNpc dopaminergic neurons (Nirenberg et al., 1996) and plays a role in reuptake of dopamine into dendrites and axon terminals (Cheramy et al., 1981). It has been reported that endogenous or exogenous (eg, WIN55,212-2) cannabinoids inhibit DAT activity in vitro (Chen et al., 2003; Steffens and Feuerstein, 2004), while other studies do not show any effect (Cheer et al., 2004; Köfalvi et al., 2005).

In the present study, we observed reduced DAT mRNA and binding levels in VTA and SNpc at all doses of WIN55,212-2, while DAT binding levels were not altered in striatum but were reduced in both NAc shell and core only at the lowest dose. These results suggest that the expression of DAT decreased at both the protein and mRNA levels in mesolimbic and nigrostriatal dopaminergic neurons and this effect is not dose dependent, with the exception of the NAc. It is interesting to point out that the effect of chronic administration of cannabinoids on DAT binding levels is evident only at the somatodendritic level of dopamine neurons and not at their axonal terminals. These brain regional differences may be a consequence of differential effect of CB1R density and function, as it is known that repeated treatment with cannabinoids (including WIN55,212-2) induces downregulation of the cannabinoid receptor (Oviedo et al., 1993; Breivogel et al., 1999; Sim-Selley, 2003; Tanda and Goldberg, 2003; Sim-Selley et al., 2006; Wu et al., 2008). In support of this notion, Moranta et al. (2009) reported brain regional differences in the synthesis of brain monoamines after chronic treatment with WIN55,212-2.

In vivo electrophysiological studies have shown that cannabinoid agonists increase cell firing of the dopaminergic neurons located in SNpc and VTA (French et al., 1997; Melis et al., 2000; Wu and French, 2000). The increased activity of SNpc neurons is in agreement with in vivo microdialysis experiments showing enhanced dopamine release in the striatum after cannabinoid agonist treatment (Tanda et al., 1997; Solinas et al., 2008; Moranta et al., 2009; Polissidis et al., 2010, 2013). Taking into account the above findings and our results of decreased binding and mRNA levels of DAT in SN and VTA, we could suggest that chronic activation of cannabinoid receptors may lead to decreased dopamine uptake by the dendrites of mesolimbic and nigrostriatal dopaminergic neurons, but not by the nigrostriatal terminals. Furthermore, the increased extracellular levels of dopamine in the striatum could be attributed to saturation of DAT (Oleson and Cheer, 2012; Tye et al., 2013) and/or to decreased inhibitory control over dopamine release due to stimulation of D2S (see below).

D2DRs exist in 2 isoforms (D2S, D2L) generated by alternative splicing of the same gene (Giros et al., 1989). D2Ss are localized presynaptically on both the somatodendritic and terminal regions of midbrain dopaminergic neurons (Khan et al., 1998). Activation of these receptors in mesencephalic dopaminergic neurons elicits hyperpolarization and decreases firing rate (Lacey et al., 1987; Mercuri et al., 1989, 1997; Centonze et al., 2002) while inhibiting dopamine release in their somatodendritic (Cragg and Greenfield, 1997) and terminal mesostriatal regions (Starke et al., 1989; Cragg and Greenfield, 1997; Usiello et al., 2000). Thus, dopamine release is under the inhibitory control of D2DRs (Benoit-Marand et al., 2001). In the present study, the mRNA levels of the short isoform of D2 receptor (D2S), which corresponds to the presynaptic D2DR, were reduced in SNpc and VTA after chronic administration of the CB1R agonist, WIN55,212-2. Using [³Η]raclopride, which labels D2DR binding sites but does not distinguish between presynaptic and postsynaptic localization of D2DR, we were unable to detect any changes of the D2DR at the binding site level in SN and VTA. However, decreased D2DR binding levels were found in the striatum of WIN-treated rats. This change could be correlated to the decreased mRNA expression of D2S observed in SNpc and may reflect a reduction in striatal D2S. However, further immunohistochemical studies using an antibody specific for the D2S are required to verify this change. Our finding is in contrast to the results of Ginovart et al. (2012), who showed increased presynaptic D_2/3 autoreceptor binding levels after chronic Δ⁹-THC administration, but are in agreement with the results of Bossong et al. (2009), who showed reduced [¹⁴C]raclopride binding in human striatum after Δ⁹-THC inhalation.

Overall, we could suggest that the somatodendritic regions of the mesostriatal dopaminergic system, the SN and VTA, seem to be more affected than the projection regions, the striatum and NAc, by chronic administration of WIN55,212-2. Decreased mRNA levels of D2S and DAT may lead to increased activity of SNpc and VTA neurons and enhanced release of dopamine at their somatodendritic and terminal fields. This finding may be attributed to the fact that chronic exposure to WIN55,212-2 can cause neuroadaptive alterations (ie, downregulation) of CB1 receptor (density and function) which is region specific (Sim-Selley and Martin, 2002).

It is generally accepted that regulation of dopamine release in striatum is mainly the consequence of alterations in dopaminergic cell firing in the SNpc and the VTA. However, several studies have revealed local regulation of DA release by other neurotransmitters and modulators, such as glutamate. Recent studies suggest that the glutamatergic regulation of dopamine release is inhibitory (Rice et al., 2011). Taking into consideration the above and that activation of CB1R receptors on corticostriatal terminals would inhibit glutamate release, this activation may lead to increased DA release.

Our results have also shown that D1DR mRNA and binding were not altered in striatum; however, in NAc core, both mRNA and binding levels were increased only at the highest dose. This dose-dependent effect is specific for the mesolimbic pathway, and it may be related to the pronounced motor-suppressant effects of the high dose of WIN22,212-2 that persisted over the course of administration. It has been suggested that simultaneous stimulation of CB1R and D1DR reduces the inhibitory activity of direct striatal projection neurons, resulting in a decreased motor response (Martín et al., 2008).

It is well known that NAc plays a pivotal role in reward and aversive learning and learning flexibility (Graybiel, 2008). In NAc, similar to dorsal striatum, MSNs express either D1DR or D2DR along with other receptors and neuropeptides, and their distinct roles in learning have only recently been explored (Nakanishi et al., 2014). Furthermore, recent investigations suggest a differential involvement of D1-MSNs and D2-MSN cell populations in NAc in drug-related behaviors (Laviolette et al., 2008; Hikida et al., 2010; Smith et al., 2013). In particular, distinct roles of D1DR and D2DR in the core and shell of NAc have been implicated in the modulation of reward by nicotine (Laviolette et al., 2008) and in the acquisition of cocaine-related learning (Espana and Jones, 2013; Smith et al., 2013).

Seif et al. (2011) provided evidence that endocannabinoids mediate the ability of DA receptors to enhance action potential firing in NAc core neurons in vitro, requiring coactivation of D1DR and D2DR. The selective upregulation of D1DRs in NAc core observed in the present study after high doses of chronic cannabinoid administration suggest that the core vs shell and D1DR vs D2DR MSNs of NAc may respond differently to repeated cannabinoid administration, and these cell-type specific alterations in NAc core may contribute to cannabinoid-related behaviors.

It has also been shown that CB1R are localized presynaptically on GABAergic neurons in several brain regions (Matsuda et al., 1993) and their activation inhibits GABA release (French et al., 1997; Pistis et al., 2002; Szabo et al., 2002; Lupica and Riegel, 2005; Szabo and Schlicker, 2005). More precisely, in SN pars reticulata, CB1R are located on GABAergic striatonigral terminals. Furthermore, dendrites of dopaminergic neurons in SNpc extend into the SN pars reticulate, where they form synapses with CB1-containing axon terminals (Fitzgerald et al., 2012). Considering this localization of CB1R, a possible mechanism for the increased dopamine release after cannabinoid treatment may involve an indirect disinhibition of dopamine neurons (French et al., 1997; Szabo et al., 2002; Lupica and Riegel, 2005). It is thus suggested that CB1R, because of their localization, can modulate GABA release and in turn control the activity of the dopaminergic cells in the midbrain (Wu and French, 2000; Laviolette and Grace, 2006; Fernández-Ruiz et al., 2010).

Our results have indicated that GABA_A receptor binding levels were reduced in the dorsal striatum at all doses and SN at the highest dose. Therefore, in addition to the reduction of GABA released from the striatonigral terminals of the direct pathway, the observed reduction of GABA_A receptors in SN at least at the higher dose would increase the activity of the GABAergic cells of SNr, leading to hypomobility. Overall, these results indicate that the effects of WIN55,212-2 on motor activity could be mediated at least partially via GABA_A receptors expressed in the nigrostriatal pathway. In conclusion, our data indicate that chronic administration of the cannabinoid agonist WIN55,212-2 did not induce phenomena of tolerance or sensitization of locomotor activity. Furthermore, repeated cannabinoid administration induced neuroadaptive alterations of the dopaminergic and GABAergic systems in a region-, dose-, and neurotransmitter-dependent manner.