The present study demonstrates a different control of striatal CB1 receptors on GABAergic neurotransmission in relation to individual differences in the spontaneous response of approach or avoidance to conflicting stimuli and advances a framework for explaining behaviors of approach and avoidance involving the ECS at striatal level.

Approach system is a motivational system activating reward-seeking behavior associated with impulsivity/exploration, while avoidance system is an attentional system promoting inhibition of appetitive responses [40], [41]. Excessive reactivity of these systems has been related to psychopathological disorders, as attention-deficit/hyperactivity disorders (ADHD) or depression on one hand and anxiety on the other hand [7], [8], [42], [43]. The innovative use of experimental protocols made in the current research allowed analyzing spontaneous motivational components that guide behavioral responses ranging from approaching to avoiding stimuli. In the presence of the same conflicting stimuli, while BA animals reacted with balanced responses between approach and avoidance, AV or AP animals respectively exhibited inhibitory or approach responses towards one of the conflicting inputs.

The behavioral differences observed in AV or AP animals were not linked to different levels of anxiety. Conversely, the behaviors linked to exploration and approach were significantly influenced by the category the mice belonged to. Namely, AP mice in A/A Y-Maze displayed the lowest entry latencies and in the OF traveled the longest distances in the arena, contacted longer the object, demonstrating they were indeed approaching and explorative. Furthermore, in S1 ten trials of A/A Y-Maze test, all animals exhibited similar anxiety levels in the very first trials (the latencies of all animals started with very similar values), AP animals were the only subjects maintaining high reactivity when the trials went by (entry latencies progressively increased only in AV and BA groups), at the end of the task AV animals made evident a behavioral inhibition (in the last two trials their latencies reached values significantly higher).

The relation between reward-seeking behavioral activation and exploration/impulsivity has been also found in previous studies reporting that impulsivity and extraversion [44], [45], as well as risk aversion and low motivation [46], [47] are related to each other.

Intriguingly, the individual differences in temperamental traits were reflected in the differences in the CB1-mediated activity of the dorsal striatal network. In fact, CB1-mediated presynaptic control on GABAergic transmission in the dorsal striatal medium spiny neurons was almost absent in AV animals and conversely increased in AP animals. ECS plays a central role in the balancing control between approach and avoidance in both humans [9], [10] and rodents [5], [11] and modulates GABAergic inhibition controlling fine-tuned behaviors [48]. Notably, reward-associated environmental manipulations sensitize CB1-mediated striatal transmission [25], [26], while chronic psycho-emotional stress causes marked down-regulation of CB1-controlled GABAergic striatal transmission [38]. Recent neuroimaging findings in healthy human subjects indicate striatal and prefrontal functional differences in reward processing related to differences in approach/avoidance personality traits [49], [50]. Individual differences on expectation and receipt of reward have been found also in clinical populations, demonstrating abnormal reward processing in psychopathological disorders, as bipolar mania [51], substance dependence [52], schizophrenia [51], [53], ADHD [54] and depression [55]. The present experimental findings fully fit with these functional studies.

In the AV animals the enhancement of the endogenous tone of anandamide (AEA) with URB597 increased number of white choices and decreased entry latencies in the A/A Y-maze as well as enhanced explorative behavior and contact times in the OF test. These behavioral responses were consistently paralleled by the rescue of CB1 receptor sensitivity to HU210, indicating that indeed striatal CB1 receptors modulate spontaneous reward-related processes. The intriguing observation that URB597 reinstates “sensitivity” to HU210 in AV mice deserves further discussion. AV mice fail to respond to HU210, suggesting silencing of CB1 receptors. How may the increase in endocannabinoid signaling by blocking AEA degradation reinstate CB1 control on GABAergic transmission? Notably, in the striatum the enhancement of AEA tone with URB597 inhibits sEPSC but not sIPSC frequency [39] because CB1 receptors controlling glutamate release are the target of AEA, while the other endocannabinoid 2-AG is the preferential endogenous agonist of CB1 receptors controlling GABAergic transmission. In fact, stimulation of 2-AG synthesis with DHPG [56], [57] or following acetylcholine M1 receptor activation [58] reduces GABAergic but not glutamatergic synaptic events. Thus, in AV animals the AEA increase, caused by URB597, reinstates the control of CB1 receptors on sIPSC frequency to HU210, indicating a complex interaction between the two main endocannabinoids and their receptors.

In AP animals the blockade of CB1 receptors with AM251 reduced contact times and explorative behavior in OF test, although it failed to affect white choice number in A/A Y-maze. Electrophysiological recordings in the same treated animals indicated a fully blocked CB1 receptor activity. Thus, AV or AP animals treated with ECS agonists or antagonists tended to fade away their behavioral features, rendering them less inhibited or less “triggered”, respectively. These findings are fully supported by counterbalancing the pharmacological manipulations in AV or AP animals. In fact, AV mice that had a reduced CB1 control on GABAergic neurotransmission when further inhibited by AM251 did not display any behavioral or electrophysiological modification. In parallel, AP animals that had an enhanced CB1 control on GABAergic neurotransmission when further potentiated by URB597 did not display any behavioral or electrophysiological modification.

ECS functional features and the pharmacological properties of the drugs acting on it provide a possible explanation for the different pharmacological efficacy found in the present research. In fact, endocannabinoids are synthesized and released “on demand” after neuronal depolarization [59], [60]. Indeed, by presynaptically reducing both excitatory and inhibitory neurotransmission, the ultimate effect of endocannabinoids depends on nature and amount of neurotransmitters being controlled [12], [14], [17], [22]–[24], [61]. The inhibition of endocannabinoid degradation by URB597 prolongs the neuronal signaling in active synapses only, preserving the spatio-temporal specificity of endocannabinoid activity [62]. Conversely, systemic CB1 receptor blockade by AM251 suppresses both excitatory and inhibitory ECS effects on multiple neuronal populations, explaining thus the different behavioral responses exhibited in the drug presence. Thus, the treatment with URB597 of the “behaviorally inhibited” AV animals enhanced the endocannabinoid tonic control over striatal GABAergic synapses and unhinged the behavioral inhibition featuring these animals. On the converse, the treatment with AM251 of the “explorative/impulsive” AP animals blocked the endocannabinoid tonic control over striatal synapses and prevented the triggered behavior featuring these animals. These findings are fully consistent with the decrease of anxious behaviors, the reduction of isolation-induced ultrasonic vocalizations in pups and the decrease of stress-induced corticosterone release provoked by URB597 injections [63]–[65]. Furthermore, they fit with the increase of the preference for palatable substances produced by administration of exogenous cannabinoids or endocannabinoids [66]–[68], and with the decrease in palatable food intake produced by treatment with AM251 [5], [23], [69].

The present research demonstrates that in responding to the same conflicting stimuli adolescent inbred mice exhibit variance of spontaneous behavior ranging from avoiding to approaching traits and that this behavioral variance is accompanied by a different CB1-mediated control on striatal neurotransmission. Human and rodent adolescents show a prominent motivation towards reward-responsivity, novelty seeking and impulsivity as well as increased vulnerability to affective illness and addiction [27], [29]–[31], [70]–[73]. Moreover, adolescent rats find repeated cannabinoid exposure less aversive than adult rats but exhibit memory deficits and changes in hippocampal protein expression more lasting [74]. Age-dependent differences in the brain levels of endocannabinoids as well as in CB1-mediated effects on synaptic transmission have been described [75]–[77]. However, the features linked to individual behavioral differences present in adolescence appear to be persistent traits maintained across the life-span [32], [33].

Interestingly, the individual predisposition to approach or avoidance demonstrated by the present data extends recent findings reporting differences in impulsivity associated with differences in striatum and nucleus accumbens monoamines [78] in inbred rodents. Phenotypic differences in susceptibility to stress associated with differences in responses to natural and drug rewards were also reported [32]. Because all same-sex members of inbred strains are genetically identical, when animals belonging to same strain are tested under controlled conditions, individual differences among animals have to reflect allelic and functional differences probably modulated by prenatal and postnatal environmental factors or early dominance hierarchies [79]–[81]. Although environmental influences determining the phenotypic variability in inbred subjects are difficult to control and measure, inbred mice raised in rigorously defined environments may show variability in some traits unrelated to genetic and environmental influences [82]. Future studies are needed to delineate the contribution of genetic, epigenetic and environmental variables that may together develop and modulate individual differences.

The behavioral responses to conflicting stimuli mirrored CB1-mediated control on dorsal striatal neuronal transmission. We are aware that the different CB1-mediated control on GABAergic transmission observed in the three behaviorally-characterized groups of animals could be linked to a different sensitivity of the CB1 receptor as an entity, or to differences in the distribution of CB1 receptors or to differences in the amount of CB1 receptors. Distinguishing among the three alternatives requires aimed studies with specific methodological approaches that are in progress. However, this drawback does not weaken the present data since the current findings do evidence that endocannabinoids acting on dorsal striatal neurons influence the spontaneous response to conflicting stimuli and support the involvement of endocannabinoid transmission in approach and avoidance behaviors that often feature not only full-blown psychiatric disorders but also the individual differences in non-pathological temperamental traits.