The influence of the CB1 receptor ligands on memory and learning processes has been widely documented by various experiments and clinical studies [8, 62–66]. Nevertheless, the results are still contradictory. In this part of the present elaboration, we will summarize the effects of the CB1 receptor ligands, including CB1 receptor agonist and CB1 antagonists (and inverse agonists) on different memory stages.

Studies have demonstrated that an acute administration of synthetic CB1 receptor agonists: CP55940 and HU-210 attenuated acquisition of memory in various animal models, e.g., the water maze test (WMT), the object recognition task (ORT), and the contextual fear conditioning (CFC) test [67–71]. Similarly, Mazzola et al. [68] and others [71] confirmed these effects of direct activation of natural CB1 receptor agonist—Δ9-tetrahydrokannabinol (Δ9-THC). Δ9-THC (3.0, 5.0, and 6.0 mg/kg) injected intra-peritoneally (i.p.) 30 min before the learning trial significantly impaired memory acquisition using the passive avoidance (PA) task in rats. This deterioration was reversed by pretreatment with 1 mg/kg of rimonabant (SR141716A), a CB1 receptor antagonist. Moreover, indirect stimulation of CB1 receptors impaired acquisition of memory in the recognition memory test [72].

In turn, Pamplona and Takahashi investigated whether CB1/CB2 receptor agonist, WIN 55,212-2, is able to influence the acquisition of fear conditioning using tone and contextual versions [69]. They revealed that this compound (2.5 and 5.0 mg/kg, i.p.), administered before conditioning and before testing, impaired memory processes in the CFC and did not affect the freezing behavior induced by tone presentation; therefore, non-state-dependent effects of WIN 55,212-2 on tested animals were observed. During the course of further studies, selective CB1 antagonists (SR141716A and SR147778) were administered, in order to establish whether impaired contextual conditioning would be prevented. Preadministration of SR141716A (1.0 mg/kg, i.p.) or SR147778 (1.0 mg/kg, i.p.) has in fact prevented the impairment. These findings demonstrate that an acute administration of WIN 55,212-2 dose-dependently impairs the acquisition of memory in the CFC test, which is a hippocampus-dependent learning and memory task and does not affect tone fear conditioning, which is considered independent of the hippocampus [73, 74]. Likewise, chronic administration of WIN 55,212-2 significantly impaired spatial memory in rats evaluated in the WMT [75]. Additionally, Kruk-Slomka et al. [64] revealed that WIN 55,212-2 impaired both acquisition and consolidation of memory in PA test in mice. The evidence gathered from this experiment reaffirms that the effects of the CB receptor agonists are selective for the hippocampus-dependent aversive memories in rats.

Contradictory data concerning the influence of CB1 on memory consolidation has also been reported. It has been demonstrated that post-training administration of CB1 receptor agonist HU-210 as well as a combined CB1/CB2 receptor agonist WIN 55,212-2 attenuated consolidation of memory in the CFC test, the WMT, and the ORT [7, 76–78]. Indeed, Maćkowiak et al. [77] investigated the role of CB1 receptors in hippocampal-dependent memory consolidation using HU-210. The results indicated that HU-210 impaired the consolidation of fear memory in the CFC test. These detrimental effects were abolished by a CB1 receptor antagonist AM251. These findings may suggest the involvement of CB1 receptors in memory and learning processes. The results of the studies indicated also that AM251 blocked the effects of HU-210 on freezing behavior but did not affect memory consolidation in the CFC on its own. Thus, the blockade of CB1 receptors does not affect consolidation of contextual memory [79] and disrupts memory consolidation in a step-down inhibitory avoidance (IA) [80, 81].

Similar effects of WIN 55,212-2 on memory consolidation were observed in spatial memory formation using the WMT [78]. Yim et al. administered WIN 55,212-2 systemically and intra-cranially to assess both methods of drug delivery. They demonstrated that this CB1/CB2 receptor ligand impairs the consolidation of long-term spatial memory. Similar long-term memory impairments were observed in both systemic injections and intra-hippocampal infusions of WIN 55,212-2. As CB2 receptors are not expressed in the hippocampus, therefore, the observed impairments provide an indirect support that this effect was reached by targeting CB1 receptors [78]. Nonetheless, it has been noted that intra-basolateral amygdala (intra-BLA infusion) of WIN 55,212-2 facilitated memory consolidation in rats evaluated in the IA task or had no effect in the mentioned animal model [82, 83]. On the other hand, post-training intra-hippocampal injection of this drug impaired consolidation of memory in several behavioral tasks [84]. WIN 55,212-2 was also evaluated in the experiments conducted by Clarke et al. [76]. They examined the effects of post-training activation of hippocampal CB receptors on the consolidation of object recognition memory. The results of the study were in agreement with the evidence provided by Yim et al. [78]. WIN 55,212-2 impaired the consolidation phase of memory formation. Amnestic effect of this compound was completely reversed by coadministration of CB1 receptor antagonist, AM251, as well as mimicked by CB1 receptor agonist, ACEA, but not by CB2 receptor agonist, JWH-015 [78]. These findings are also in agreement with the results published by Moshfegh et al. [85]. They used a step-down PA task as a model of learning. The results indicated that post-training administration of WIN 55,212-2 produced an amnestic response. All the effects described previously endorsed the hypothesis that the memory impairments were due to activation of hippocampal CB1 receptors.

Systemic administration of CB1 receptor antagonists, e.g., rimonabant (SR141716A) or AM251, has been tested in various learning paradigms alone or coadministered with CB1 agonists [61, 63, 64, 81, 86–89].

SR141716A is a selective and potent CB1 receptor antagonist [90]. It also presents features of an inverse agonist [91]. An acute, pretraining administration of SR141716A facilitated the acquisition of memory in rodents observed in the PA test, the elevated T-maze (ETM) test, and social recognition memory task [61, 89] and impaired the acquisition of memory assessed in the spatial memory test [92]. Additionally, systemic, post-training administration of rimonabant enhanced memory consolidation in the radial arm maze (RAM) [89]. Also, Robinson et al. studied the effects of SR141716A on spatial learning and memory formation using the WMT [92]. Two experiments were performed. In the first one, rimonabant was administered i.p., before or immediately after training. The results indicated that systemic administration before training induced deficits in acquisition of spatial reference memory; however, pretraining before drug treatment eliminated this effect. The experiment revealed that rimonabant-induced memory deficits appeared as a result of anxiogenic effects of the drug. Post-training injections had no effect. In the second experiment, rimonabant was administered intra-hippocampally during the training and the results indicated that this drug enhanced acquisition learning and exerted no effect on consolidation of memory. Subcutaneous injections did not affect memory in any way [92].

The results from these described experiments demonstrated that rimonabant produced various effects dependent upon the route of administration, the timing of infusion, and the dose of the drug. Similarly, Lichtman reported an improvement of memory acquisition induced by administration of SR141716A [93]. Likewise, Wolff and Leander showed the enhancement of the consolidation processes when the animals were tested in the RAM test [89]. Furthermore, Wolff and Leander [89] proved a dose of 1.0 mg/kg of SR141716A to be effective. The higher dose of 3.0 mg/kg did not improve the consolidation, what is consistent with the results obtained by Lichtman [93] in the same task. Pro-cognitive effects of rimonabant were also shown in the experiments performed by Takahashi et al. [88]. Administration of 1.0 mg/kg of rimonabant produced an improvement in memory acquisition and consolidation in the ETM task. Neither lower (0.5 mg/kg) nor higher (2.0 mg/kg) doses were able to improve acquisition. Additionally, facilitation of short-term olfactory memory in the social recognition test was described by Terranova et al. [88]. On the other hand, Marsicano and colleagues failed to prove any effect of rimonabant on the consolidation of aversive memories [94]. This result may be accounted to the different mouse strain. The single dose used by Marsicano et al. [94] was also higher than the maximum dose used in the studies described previously.

AM251 is a member of the same CB group of diarylopyrazoles as SR141716A, presenting the features of CB1 receptor antagonist and inverse agonist [4]. The post-training administration of AM251 interfered consolidation of memory-related processes in the step-through IA task or CFC task [82, 87]. De Oliveira Alvares et al. [81] investigated the effects of intra-hippocampal administration of AM251 in two behavioral tasks: the IA and the open field (OF) habituation task. The results indicated that AM251 exerted a disruptive effect on memory consolidation in the IA test, but not in the OF habituation test. Similarly, Kruk-Slomka and Biala [63] confirmed that an acute injection of AM251 improved the short-term and long-term memory stages (acquisition, consolidation, and/or retrieval) in the IA task. The effect seemed to be purely mnemonic since the drug showed no motor performance effects, which could favor a false positive for the intermediate dose in the IA test session. It needs to be highlighted that the amnestic effect took place with the lower, more selective dose, not with the higher one (that one could bind to the non-specific targets in the hippocampus). The different responses observed in two behavioral tasks require explanation. The findings of the study suggest that hippocampal endocannabinoids are not acting upon the consolidation of the OF habituation task. The fact that the OF was not recruited and the IA was sensitive to AM251 raises the possibility that this system requires some degree of aversiveness or alertness to be recruited. The impairing effect of AM251 on memory consolidation was also confirmed by Bucherelli et al. in 2006 [87]. De Oliveira Alvares and colleagues replicated the previous findings [80]. Afterwards, Bialuk and Winnicka [4] performed a study in an attempt to determine the influence of different doses of AM251 on recognition memory. In order to evaluate the effects of AM251 on acquisition of information, the drug was given 15 min before learning trial, and to establish its influence on consolidation of information, it was injected immediately after the trial in the ORT. The results of the study indicated that AM251 significantly improved both acquisition and consolidation of information; however, these effects were observed only when dose of 1 mg/kg was administered. Higher doses did not exert any influence on it. The memory-improving effect is in an agreement with the results obtained in experiments with SR141716A [4].

Interesting experiments in the context of our paper seem to be the studies of Tan et al. [95]. The authors used CB antagonist, agonist, and reuptake inhibitor, AM251, WIN 55,212-2, and AM404, respectively. The substances were administered bilaterally as an intra-BLA and intra-prelimbic (intra-PLC) microinfusion in rats. The results indicated that pharmacological inhibition of intra-BLA CB1 receptor transmission dose-dependently blocked the acquisition of olfactory fear memory, simultaneously leaving unaffected recall and consolidation of these memories in an olfactory fear conditioning procedure. In addition, activation of CB1 receptor transmission or inhibition of the endocannabinoid reuptake within BLA strongly potentiated the acquisition of fear memory. Moreover, fear memory formation, mediated by CB1 receptor, was blocked when the medial prefrontal cortex (mPFC) was pharmacologically inactivated before intra-BLA activation of CB1. These findings are consistent with the report presented by Campolongo et al., which showed that intra-BLA activation of CB1 receptors can potentiate the encoding of associative memory for IA learning [82]. Previous studies conducted by Roche et al. [96] indicated that intra-BLA blockade of CB1 receptor transmission with rimonabant inhibits the formation of context-dependent fear memory. Overall, the previously mentioned findings suggest that CB1 receptor-dependent transmission within the BLA can influence the magnitude of emotional memory encoding. Additionally, the memory-improving effects of AM251 observed in this study were in agreement with the results obtained by Riedel and Davies [97]. They have reported that while the CB1 receptor agonists impair memory formation, the CB1 receptor antagonists reverse these deficits or act as memory enhancers.

Lin et al. studied the effects of another CB1 receptor antagonist AM281 on the formation of contextual fear memory in adult mice [98]. AM281 (2.5 mg/kg) was injected both intra-peritoneally and intra-hippocampally to assess the influence on memory acquisition. These results indicate that CB1 receptor-mediated signaling within the area of hippocampus negatively regulates the acquisition in contextual fear memory task. Thus, AM281 seems to be a good candidate for memory enhancement; however, further studies in animal models of cognitive dysfunctions are still required. In the context of this subject, it should be noted that Wise et al. decided to determine the effects of a relatively novel and potent CB1 receptor antagonist CE on memory processes. CE is structurally distinct from rimonabant. Wise and colleagues observed that CE significantly enhanced memory consolidation in the RAM procedure [99].

As we described previously, the influence of the CB1 receptor ligands on memory and learning processes has been widely documented by various experiments and clinical studies. Although CB1 receptor ligands are able to improve as well as to impair memory, each of them affects memory in a different way. Such contradictory findings may be connected with differences in behavioral tasks used, handling procedures, e.g., time of drug administration, the kind of drug treatment, or other experimental conditions, as well as doses and CB compounds selected. Therefore, future studies may help to discover and describe the precise role and character of different CB1 receptor compounds (Table 2).