In this study we uncover the role of CB2Rs in regulating PE-mediated actions and countering chronic stress behavioural and plasticity-induced defects. Taken together, our data show that inhibiting CB2Rs in combination with PE can restore the reduction in AHN and counteract the emotional and cognitive deficits impinged by chronic stress. Furthermore, we show that adjuvant CB2R inverse agonist treatment greatly boosts the positive effects of PE, resulting in behavioural recovery of emotional and cognitive dimensions (Fig. 5a, a’), increased NSC proliferation, neuronal differentiation and survival (Fig. 5b, b’) as well as decreased glial reactivity, myelination and neuroinflammation (Fig. 5c, c’). These observations suggest a pivotal participation of CB2Rs in PE-elicited recovery from chronic stress. Interestingly, a constant finding herein is that CB2R activation alone or in combination with PE yields no positive actions under chronic stress, in comparison with the beneficial effects of CB2R inhibition. Since in physiological conditions CB2R inhibition has a negative impact, our findings highlight a dual role of CB2Rs under physiological versus pathological conditions.
Although the beneficial effects of PE have been long known [60], the molecular mechanisms underlying this response have remained unclear. As recently proposed [61], understanding the molecular and cellular effects of PE, particularly with regard to effects on AHN and emotional and cognitive functions, is crucial to develop new preventive or combating strategies that mimic or enhance the beneficial effects of PE. From a translational standpoint, the therapeutic potential of physical exercise is enormous for the treatment of a variety of brain disorders (e.g., addiction, anxiety, stroke, epilepsy), as previous studies have highlighted [49, 62]. Current literature suggests that PE-elicited effects are mediated by the ECS [38, 41, 42]. In the context of stress, this interaction likely occurs due to PE mobilizing the ECS to replenish energy stores and participate in PE-induced analgesic and mood-elevating effects, consistent with a role of the ECS in both activating and terminating the hypothalamic-pituitary-adrenal axis response to stress [63, 64].
Recent studies started tackling this interaction in the context of NSC regulation and stress-related disorders, with only a few drawing correlations in clinical settings [40, 42, 43, 65]. In fact, this interplay has been showed to rely on the collaboration between ECS components and BDNF, one of the major neurotrophic factors upregulated by PE. For instance, endogenous cannabinoid signalling, specifically through cannabinoid type 1 receptors (CB1Rs), was shown to be required for the effects of voluntary PE in hippocampal NSCs [66]. In line with this, we have previously reported that CB2Rs are necessary for BDNF-mediated NSC proliferation and neuronal differentiation [44]. More broadly, others have found that cannabinoids prevent depressive-like behaviours, an effect that was accompanied by alterations in BDNF expression, in a rat model of posttraumatic stress disorder [67]. Recently, CB2R inhibition was found to dampen BDNF signalling in stressed animals, suggesting a close interaction between CB2Rs and BDNF in emotional response [45]. Interestingly, evidence suggests that exercise-mediated runner’s high (i.e., relaxing state of euphoria after PE) might occur due to the recruitment of peripheral CB1Rs and CB2Rs but, so far, no study has ever focused on whether CB2Rs participate and/or cooperate with PE in NSC regulation and stress response.
Our data now expands this field of research by showing that CB2Rs actively contribute to stress response elicited by PE. Specifically, we show that CB2R inhibition, but not its activation, is essential for PE-mediated antidepressant, anxiolytic and pro-cognitive actions after chronic stress. We observed that, in physiological conditions, CB2R activation had proneurogenic effects while CB2R inhibition triggered depressive-like behaviour and cognitive impairments, which is in accordance with existing literature [68]. However, and importantly, we observed that CB2R inhibition in combination with PE counteracted chronic stress-induced behavioural impairments, especially in the emotional domain. While chronic stress promoted anxiety-like, anhedonic-like behaviours and impaired coping features, CB2R inhibition alone was able to partially recover deficits in adaptive behaviour (i.e., in the FST test) but only in combination with PE it could fully ameliorate other behavioural dimensions. This is in agreement with previous publications showing that CB2R inhibition blocks the effects of chronic stress in behaviour tests related to emotional processing [26, 45], and we are now adding evidence that PE synergizes with this action. Regarding memory function, we observed that both PE alone or in combination with CB2R inhibition reverted chronic stress-induced cognitive impairments, which goes in line with previous studies showing that PE enhances learning and memory processes [60] and that CB2R is important for long-term memory consolidation [68]. Nevertheless, others have also shown that CB2R activation can exert beneficial effects regarding cognitive performance in a mouse model of Alzheimer's disease [69].
Characterization of the neurogenic process in the hippocampal DG revealed that, in agreement with previous reports [56, 57, 70], chronic stress significantly impacts all stages of AHN – cell proliferation, neurogenesis and adult-born neuron survival. Our data goes in line with existing evidence indicating that the AHN deficits induced by chronic stress can affect both dorsal and ventral divisions of the hippocampal DG [58, 70], but preferentially targets the vHip [71, 72]. Although stress may preferentially decrease neurogenesis in the vHip, mood-improving interventions can stimulate neurogenesis in both hippocampal divisions [13], thus countering the effects of stress. Indeed, in our experimental setting, PE alone partially rescued some stress-evoked deficits in AHN, particularly regarding the number of immature neurons and survival of newborn neurons. This is in agreement with previous literature demonstrating that PE increases the number of immature neurons and adult-born neurons following stress [73, 74] but in conflict with evidence showing that PE can also boost cell proliferation [74, 75]. However, PE was able to fully rescue AHN deficits at all evaluated stages only in combination with CB2R inverse agonist AM630, suggesting that a reduction in CB2R constitutive activity is essential to boost the beneficial influence of PE in countering chronic stress. Growing evidence shows that CB2Rs are key regulators of the neurogenic process [16, 44], particularly in pathological contexts [25, 45], although CB2R knockout mice appear to display stable AHN [76], likely due to compensatory mechanisms. In our hands, CB2R inhibition alone was not able to revert any AHN deficits, further reinforcing that CB2R constitutive activity impinges on the actions of PE to shape adult hippocampal NSC rates of proliferation and differentiation. Further, these results suggest that the combined CB2R inhibition and PE treatment triggers a rearrangement in the DG neurogenic microenvironment, potentiating AHN in both dHip and vHip, which would explain the improvement in both dorsal and ventral hippocampal-dependent functions (i.e., cognitive and emotional dimensions).
Most studies focusing on the impact of chronic stress on AHN often disregard the structural segregation along both the hippocampus longitudinal and transverse axes or differences between the supra- and infra-pyramidal blades of the GCL, although newborn neurons localised there differentially modulate inputs to the DG [56, 72, 77]. This functional differentiation is extremely important because it translates into different behavioural responses related to distinct hippocampal functions. The dHip is preferentially involved in cognitive processes, with dorsal granule neurons participating in contextual memory, whereas the vHip is more implicated in emotional processing, with ventral granule neurons contributing to suppress innate anxiety [9, 13, 78]. The DG can be further subdivided into the suprapyramidal blade, located between CA1 and CA3 regions, and the infrapyramidal blade on the opposite side of the hilus [78], with each blade being further dissected into the SGZ and GCL subregions, likely reflecting different contributions to distinct hippocampal functions, such as the processing of spatial and contextual information [77, 79].
While aimed at disentangling the subregional contribution of each cellular population in AHN, we found chronic stress to disrupt subregional distributions of ongoing proliferating cells and newly-generated neurons, an effect that was more evident in the suprapyramidal blades of both dHip and vHip. Accordingly, when looking at immature neurons, we found that chronic stress induced a numerical shift between immature neurons and progenitor cells, from supra- to more infrapyramidal portions and from SGZ to more GCL regions. Importantly, the CB2R inhibition in combination with PE was able to abrogate these subregional changes and recover proliferating and survival patterns to similar levels as control conditions. These results go in line with previous studies showing that stress preferentially targets proliferation in the SGZ area [80] as well as the survival of adult-born neurons in the suprapyramidal blade [56]. Further studies are required to comprehensively understand the importance of these regional differences. As previously suggested [81], CB2R inhibition is likely boosting the beneficial effects of PE in preventing the stress-induced activation of DG mature granule neurons, which is known to critically shape AHN dynamics, likely through BDNF signalling [82].
Neuroinflammation plays a relevant role in depression, with both MDD patients and preclinical models of depression often exhibiting high levels of neuroinflammatory markers [37]. In our hands, we observe that uCMS exposure induces increased glial reactivity, recapitulating previous findings [83, 84]. We detected significant alterations associated with chronic stress in the expression of microglia and myelin, but not astrocytes, in DG granular regions. Taking into consideration the role of CB2Rs in regulating immunity and inflammation [25] and the role of PE as a strong inhibitor of inflammation [85], we confirmed the immunomodulatory role of PE in combination with CB2R inhibition, but not activation, in fighting neuroinflammation by reverting most changes triggered by chronic stress. These results go in line with previous evidence showing that inhibition of CB2Rs attenuates the inflammatory load triggered by challenging situations [86]. Nevertheless, others have also demonstrated that CB2R activation alone prevents stress-induced neuroinflammatory responses [25]. We posit that, in our experimental conditions, CB2R inhibition may be shielding the system against stress-targeted reactive microglia (which express high levels of CB2Rs [87]), thus facilitating PE actions in limiting the neuroinflammatory damage caused by chronic stress. Conversely, PE may be modulating stress-disrupted cellular immunity and CB2R inhibition further alleviating this immunoinflammatory status [88].
Although our results are associative, we suggest that the combined treatment of CB2R inhibition with PE accelerates the incorporation of newborn neurons into the hippocampal circuitry which, in turn, facilitates adaptive and resilient behaviours [89]. However, it is plausible that PE and CB2R modulation are interfering with the critical period that allows appropriate differentiation and survival of adult-born neurons by supplying adult NSCs with proneurogenic, pro-survival supportive factors (e.g., BDNF) that counteract the effects of stress.
Interestingly, AM630 was suggested to act as protean ligand of CB2Rs, which may be reducing the probability of spontaneous activation of the receptor by favouring an active receptor conformation of lower efficacy [90, 91]. This may reconcile some apparently controversial evidence showing beneficial effects of both CB2R activation and inhibition in stress-related conditions, since a protean ligand may act as agonist in systems that are quiescent (no constitutive activity) or as inverse agonist in constitutively active systems [91]. This possibility highlights that distinct stress levels may affect CB2R constitutive activity. Considering that the therapeutic relevance protean ligands may have in setting receptor activity to a constant level [91], one may speculate that, in our results, stress dysregulates CB2R constitutive activity and that AM630 may contribute to activity adjustments towards normal levels. Nevertheless, our data consistently demonstrates a cumulative effect of CB2R inhibition with PE, which dampens the negative consequences of stress, in most cases nullifying them.
In conclusion, our work shows that a strategy coupling CB2R inhibition in combination with PE may be a useful approach to counteract the effects of chronic stress in terms of behaviour and plasticity-related events in the hippocampus. Overall, our observations highlight a new multimodal approach for the treatment of stress-related pathologies. Future studies should focus on understanding whether the effects of this joint treatment expand to other brain regions and which intracellular mediators are behind the actions of this combined strategy. PE and CB2R effects may, indeed, be converging at multiple signalling pathways or, perhaps, have a common effector element. Altogether, given the lack of effective treatments and the significant public health impact of stress-related psychiatric disorders, in particular MDD, the promise of combining lifestyle interventions such as PE with pharmacological targeting of CB2Rs seems relevant and should be further explored.