The first main result of our study is that SI in rats evokes selective alterations in ECS genes expression in the Pfc and Amy, two brain regions with strong connections [60] and which interactions are important in the modulation of affective states and social behaviors [61–64]. Namely, we report significant reductions in mRNA levels for Cnr1 in both brain regions and for Faah gene in the Amy. Several evidence reporting an increased firing at Amy and Pfc might indicate that these events are relevant to evoke the transition from a condition of distress to a possible psychiatric disorder as reviewed by Davidson [65]. Stress exposure firing this specific circuit has been already associated with the downregulation of the ECS and suggested as relevant for the development of an anxious phenotype [66]. No changes were evident in any ECS genes in the other three brain regions here analyzed.
It is well-known that Cb1 receptors mediate most of the effects evoked by cannabinoids in the CNS [67] and we here further support their role in the regulation of stress and emotions, suggested also by their high distribution in the Pfc and Amy, as well as in motivational and cognitive processing [68]. Reduction of Cnr1 expression in Pfc was in fact also reported in an AD mouse model, APP/PS1 mice [69]. These mice did not show at early stages (3 months old) the presence of anxiety, as well as alterations in cognition and social-related behaviors. The regulation of Cnr1 has been thus suggested by the authors as possible biomarker in a presymptomatic stage of cognitive impairment and this might be applied also in our SI rats that could develop cognitive impairment. Moreover, our data on reduced Cnr1 expression in SI fit also with other relevant papers tagging the receptor, for instance: Cb1 blockade decreases social play behavior [70], activation of Cb1 receptors in the Amy can attenuate anxiety responses [71, 72], whereas their antagonism or deletion elicits an anxiogenic response [73]. An earlier study already evaluated the effects of SI on ECS genes expression using immunohistochemistry and, consistent with our findings, Cnr1 resulted down-regulated in the Amy [74]. The authors suggested that lower Cnr1 expression in isolated rats might produce a higher possibility to activate G proteins and thus evoke intracellular second messenger system activation in a more efficient way [74].
As we mentioned in the introduction, we recently reported the central role of Oxtr transcriptional regulation in the same group of rats socially isolated for 5 weeks, and in particular we observed a down-regulation of receptor mRNA levels in the Pfc of ISO rats when compared to controls. This alteration correlated with anxiety-like and stereotyped behaviors [36]. OXTR, which is highly expressed by glutamatergic prefrontal cortical neurons, and Cb1, highly expressed in cortical regions, respectively are known to modulate social recognition [75] and underpin human social–emotional functioning [76].
Interestingly, we here show a significant correlation between the altered expression of OXTR and Cb1 receptors. The Pfc is known to subserve highest-order cognitive abilities and social processing, and it is also the most sensitive brain region to the harmful effects triggered by stress [77]. For this reason, we could hypothesize a possible Endocannabinoid-Oxytocinergic systems interplay in the Pfc of SI rats, partially corroborating an already existing theory of an endocannabinoid signalling mediating oxytocin-driven social reward [37]. Further studies, willing for instance to address oxytocin and endocannabinoid levels correlation in our and other experimental models, are clearly needed to support this hypothesis.
The other relevant finding on ECS genes expression was the reduced levels of Faah mRNA in the Amy of SI rats compared to group-housed animals and this might suggest the enzyme role in processing endocannabinoid signalling on Amy neurons and thus the anxiety processes mediated by this region. It has been in fact already suggested that alterations in Faah activity, evoked by stress or compounds targeting the enzyme [78], would impact Amy functions, including anxiety mechanisms associated with it [42]. It should be considered that ECS genes expression in rat models of SI has been previously investigated by others [74, 79], however beside what already discussed above, other alterations in ECS components’ mRNA have been reported, not consistent with the present findings. Different experimental protocols for SI (8 vs 5 weeks of SI) and methodological approaches (immunohistochemistry protocol vs Real Time PCR) may account for these discrepancies.
Of relevance and consistent with our findings, the down-regulation of Cnr1 [80–83] and reduced levels of FAAH [84, 85] have been also reported in cannabis users, experiencing loneliness [86] and negative effects on social and interpersonal relationships [87].
In order to further analyze the role of epigenetic mechanisms in the downregulation of Faah and Cnr1 genes, we investigated DNA methylation at specific CpG sites and selected histone modifications relevant for activation and inhibition of genes transcription, at the level of both gene promoter regions. We did not report significant differences in DNA methylation levels at any CpG sites under study for both Cnr1 and Faah when comparing SI rats with controls. Instead, we show an increase in H3K27Me3 levels at the Cnr1 promoter region in the Pfc of SI rats consistent with the reduction in gene expression. The same histone modification resulted to be significantly increased at Faah gene promoter in the Amy of SI rats when compared to controls again consistent with the downregulation of enzyme mRNA levels. These data are thus in agreement with the role of trimethylation of Histone-3-Lysine-27 as a modification associated to gene silencing mediating heterochromatin assembly driven by repressive Polycomb complexes [88, 89].
Other studies already reported the dynamic regulation of H3K27 trimethylation evoked by different environmental cues. For instance, chronic stress in mice induces an increase in the epigenetic mark in the hippocampus associated with a reduction in brain-derived neurotrophic factor (Bdnf) [90]. On the other hand, acute stress reduces hippocampal H3K27Me3 [91]. Higher levels of H3K27Me3 were also reported in the orbitofrontal cortex of suicide completers and associated with reduced TrkB gene expression [92].
In the Amy of SI rats we instead did not report any change in H3K27Me3 at Cnr1 gene promoter, whereas we observed a reduction of the activating mark H3K9Ac which role seems not of relevance in the Pfc in our experimental condition. Again, this modulation of the histone modification is consistent and significantly correlated with the reduction in gene expression. In fact, it is known that histone acetylation is associated with gene activation reducing the affinity between DNA and histones [93].
Isolation rearing of rats was already reported to induce a global decrease in H3K9 acetylation [94], whereas in adult mice it was observed an enhancement of the epigenetic mark [95]. Anyhow, SI has been reported to induce alterations in histone acetylation that could contribute to the pathogenesis of major depressive disorder [96].
It should be also taken into account that all the molecular data shown are correlated with the behavioral outcomes we previously reported, and this further support the relevance of these genes regulation in the effects evoked by SI.
Moreover, our data support the existence of an Endocannabinoid-Oxytocinergic systems interplay in the Pfc of ISO rats and suggest a clear role for the transcriptional regulation of selective genes of the ECS (Faah and Cnr1) in the effects evoked by SI in the Pfc and Amy, also pointing out to their epigenetic regulation via specific histone modifications, whereas DNA methylation seems not to be relevant for these genes regulation in our experimental conditions.
In agreement with what recently suggested [97], we provide further evidence on the need for strategies aimed to restore ECS homeostasis in order to counteract the effects of SI and avoid the development of psychiatric conditions, also suggesting the importance of an environmental trigger that could be used to revert those epigenetic mechanisms accounting for the alterations observed. Of particular relevance in the last years, it is the possible role played by nutrition on the prevention of mental disorders [98]. Dietary constituents can in fact impact epigenetic processes to the point that nutritional epigenetics is becoming a new, rapidly expanding research field. [99]. Taking into consideration what we here report, it has been reported that changes in diet methyl donors can evoke alterations in methylation levels of lysine residues of histones 3. For instance, studies showed that H3K27me3 resulted upregulated by a choline supplementation [100] and reduced by a methyl-deficient diet [101]. Thus, it might be helpful to take into account these aspects and assume a proper diet when there might be an environmental trigger like social isolation that might affect mental health.