Endocannabinoids and the Brain Immune System: New Neurones at the Horizon?

Due to ongoing cell renewal and turnover, we generate an almost completely new body every second month (1). However, the majority of neurones appears to be excluded from this process. Even though the plasticity of the brain is a prerequisite for learning, memory formation and other cognitive functions, in most of the brain compartments, neuronal cell renewal is substituted by synaptic plasticity and the potentiation of existing pathways and connections. The hippocampus and the olfactory bulb are the only two brain regions where neurogenesis occurs throughout adulthood (2, 3).

Neurogenesis in the hippocampus has been of considerable scientific interest ever since this brain region was shown to be important for memory consolidation (4, 5). Thus, neuronal progenitor cells in the dentate gyrus of the hippocampus are thought to be the gatekeepers of memory (6, 7). However, conclusive evidence proving a strong causal connection between memory performance and the generation of new cells has yet to be shown. Understanding how a new neurone is made allows us to understand stem cell biology in general. Research on adult neurogenesis may develop specific tools for stem cell transplantation and brain regions not capable of intrinsic neurogenesis could be reprogrammed to produce new neurones in the event of disease or injury. Thus, the search for molecules that modulate neurogenesis is ongoing.

Inflammation in the central nervous system (CNS): maintaining the neurogenic balance

Neural progenitor cell proliferation and differentiation depends on their intrinsic properties and local environment and is down-regulated in brain disorders associated with inflammatory reactions. Since pioneering studies demonstrated that lipopolysaccharide-induced inflammation strongly impairs basal hippocampal neurogenesis in rats (8) and that anti-inflammatory treatment restored neurogenesis after endotoxin-induced inflammation and augments neurogenesis after cranial irradiation (9) and seizure (10), much evidence has demonstrated that neuroinflammation and neurogenesis are closely related.

This relationship, however, is complex and difficult, and the neural progenitor cell apparently has to travel like Ulysses between Scylla and Charybdis within the stormy sea of inflammation in order to achieve the final fate of becoming a functional neurone. Whereas evidence suggests that activated microglia have a detrimental effect on the survival of new hippocampal neurones, reversal of inflammation has been shown to restore hippocampal neurogenesis (e.g. after cranial irradiation) (8). In stroke models, treatment with anti-inflammatory agents, such as indomethacin or minocycline, suppressed the activation and proliferation of microglia (11, 12) and enhanced the accumulation of newborn neurones (11, 13), or at least did not interfere negatively with neurogenesis (12). By contrast, inflammation decreased neurogenesis, which appeared to be correlated with up-regulation of microglial production of tumour necrosis factor (TNF)-α (14).

However, newly-formed neurones can survive despite chronic inflammation (10) and, moreover, specifically arise within an inflammatory environment. During specific inflammatory conditions, neurogenesis of adult neural progenitor cells in mice is induced by interleukin (IL)-4- and was mediated, at least in part, by insulin-like growth factor-I (14).

The role of insulin-like growth factor 1 has been supported by a study using an animal model of Alzheimer’s disease: the double-transgenic (APP/PS1) mice. Within this model, the induction of neurogenesis leads to a phenotypic switch from microglial type cells to dendritic-like (CD11c) cells that produce insulin-like growth factor 1 (15). A role for TGF-β in the induction of neurogenesis has also been suggested, but the data are conflicting. Whereas TGF-β has been shown to promote neurogenesis in primary neuronal stem cell cultures and blockade of TGF-β activity diminished the percentage of newborn neurones in vivo (16), TGF-β has also been shown to induce a long-lasting inhibition of neural stem and progenitor cell proliferation and a reduction in neurogenesis in adult neural stem and progenitor cell cultures by induction of G0/1 phase cell cycle arrest (17). Moreover, there is much evidence supporting a role for T cells in neurogenesis. Enhanced neurogenesis has been associated with an increase in CD4-positive T cells and the Th1 cytokine interferon (IFN)-γ (14). Furthermore, hippocampal neurogenesis induced by an enriched environment was associated with the recruitment of T cells and the activation of microglia, whereas, in immune-deficient mice, hippocampal neurogenesis was markedly impaired but could be restored and boosted by CNS-specific T cells (18).

It thus appears that a specific microglial phenotype in association with T cells act to support neurogenesis (14, 15). In this context, both constitutive and postlesion levels of microglial activation differ between neurogenic and non-neurogenic regions (19).

Endocannabinoids balance the CNS immune system and neurogenesis

To avoid inflammatory escalation, the CNS harbours an impressive arsenal of cellular and molecular mechanisms enabling strict control of immune reactions: the so-called ‘immune privilege’. The immune system acts to keep the CNS under surveillance, but in a strictly controlled, limited and well-regulated manner. In recent years, several in vitro, in vivo and clinical studies have suggested that the endocannabinoid system participates crucially in CNS immune control and neuroprotection, therefore playing an important role in the cellular network of communication in, and between, the nervous and immune systems during neuroinflammation and neuronal damage.

In immune system cells, CB₂-receptor-stimulation generates inhibitory signals (20) that are primarily mediated by inhibition of both the cAMP/protein kinase A pathway and NF-AT- and AP-1-dependent transcription. Endocannabinoids also inhibit CB₂-mediated release of TNF-α, IL-1 (21), IL-6 and IL-8 (22, 23) by monocytes/macrophages, and apparently stimulate nitric oxide release (24). In T cells, cannabinoids induce a shift from a Th1- to Th2-type cytokine profile, with changes such as suppression of IFN-α and IL-12, and induction of IL-4 release. It has been demonstrated that cannabinoids can also act at the nuclear level. For example, cannabinoids can: (i) induce κ-α phosphorylation that leads to an enhanced transcription of several apoptotic genes regulated by NF-κ (25); (ii) mediate peroxisome proliferator-activated receptor gamma-dependent inhibition of NF-AT (26); and (iii) promote cell cycle interference by activation of p21waf-1/cip-1 and induction of G1/S phase arrest (27). Endocannabinoids are also chemoattractants involved with the initial stage of macrophage and microglial cell recruitment to the site of tissue damage (28). Conversely, endocannabinoids may also help to control and limit the local immune response, by preventing harmful overactivation. Thus, endocannabinoids may exhibit both pro- and anti-inflammatory functions, dependent on their concentration and the presence of other stimuli or mediators.

The endocannabinoid system obviously controls immune responses via multiple cellular and molecular targets such as the reduced secretion of IL-1β and TNF-α (21), enhanced transcription of several apoptotic genes regulated by NF-κ (25) and blocked G1/S phase transition in stimulated peripheral blood mononuclear cells (27). Generally, cannabinoids may cause dysregulation in cytokine production by immune cells (29), which may disturb the well-regulated immune resonponse. Additionally, there is evidence that endocannabinoids negatively regulate the adaptive immune response of T cells (26, 27, 30). In this context, there are several reports that the endocannabinoid system may play a critical role in the regulation of dendritic cell growth and maturation (31). Importantly, dendritic cells appear to be highly sensitive to cannabinoid-induced apoptosis compared with other immune cells, which has been suggested as the cellular background for the immunosuppressive and anti-inflammatory properties of cannabinoids (25). Due to the fact that the endocannabinoid system influences multiple aspects of immune function, it is likely that endocannabinoids, at least partially, control inflammation-balanced and -regulated neurogenesis.

Many genes involved in neuronal development are also regulated during adult neurogenesis (32). Begbie et al. (33) found that the CB1 gene exhibits a dynamic expression pattern that spatially and temporally follows neuronal differentiation in the early embryo, confirming the impact of the endogenous cannabinoid system in embryonal neuronal differentiation.

Cannabinoids have been shown to control cell fate via different mitogen- and stress-activated protein kinase cascades, proteion kinase B and ceramide in neuronal and immune cells (34, 35). Studies have demonstrated that cannabinoids, via cell cycle interactions, can induce apoptosis (36) in cancer cells both in vivo and in vitro, thus raising the possibility that cancer could be treated with cannabinoids. One other exciting aspect of cannabinoid research is the neuroprotective action of cannabinoids during glutamatergic excitotoxicity, ischaemia and oxidative damage (37). The evidence presented suggests that the endogenous cannabinoid system may influence adult neurogenesis; a process that involves cell proliferation, fate decision and cell survival.

Although considerable evidence shows that the endocannabinoid system influences long-term potentiation and depression and therefore memory formation, there are no studies that link endocannabinoid action on cognitive function to the direct effect on neurogenesis. To investigate the impact of cannabinoids on neurogenesis and potential neurogenesis dependent pathologies, the endocannabinoid would need to be knocked out exclusively in the cells directly involved in this process (e.g.the progenitor cells in the subgranular zone or subventricular zone itself). However, at present, no models exist that significantly impair neurogenesis without influencing other systems. There are attempts to generate progenitor specific CB1 knockout mice, which will enable us to distinguish between direct and indirect effects of cannabinoids on neural progentor cell (NPC) fate and function.

Rueda et al. (38) have shown that the endocannabinoid N-arachidonoyl-ethanolamine (anandamide, AEA) inhibits progenitor cell differentiation through attenuation of the ERK pathway in vitro and that adult neurogenesis in the dentate gyrus (DG) was significantly decreased by the AEA analogue methanandamide and increased by the CB1 antagonist SR141716. Jin et al. (39) also reported increased neurogenesis in the DG using both SR141716 and another antagonist, AM251, claiming that this action was mediated by CB1 receptors and the vanniloid receptor 1. CB1R knockout animals showed a 50% decrease of BrdU labelled cells in the DG whereas application of SR141716 had no effect on progenitor cell proliferation in the DG, but increased BrdU incorporation in the subventricular zone (SVC).

It is important to note that the two available CB1R knockout strains that are used in the cannabinoid research field differ due to the different background strains. The mice from Ledent et al. (40) are backcrossed on the CD1 outbred strain, whereas the mice from Zimmer et al. (41) (homologous recombination) and Marsicano et al. (lox/CRE system) are backcrossed to the inbred C57Bl/6 strain. Whereas application of SR141716 had no effect on BrdU incorporation into the DG of knockout mice with a CD1 background, increased BrdU labelling was detected in the DG of knockout mice with a C57Bl/6 background mice (39). Unfortunately, in this study, the BrdU cell phenotype was not ascertained. Molina-Holgado et al. (42) reported that cannabinoids promote oligodendrocyte progenitor survival. The fate of a newly-generated cell is not described by BrdU incorporation. Markers for neuronal lineage maturation, such as doublecortin, nestin, neuN or calretinin, need to be used to investigate neurogenesis versus plain progenitor cell dynamics (43). Interestingly, the two background strains used to generate CB1 knockout mice differ in the baseline rate of progenitor cell proliferation and neurogenesis. The C57Bl/6 strain have a higher proliferation rate, but a lower survival rate than the CD1 strain (44).

Another crucial factor that differs between the various studies is the variability with which neurogenesis is measured or even defined by different laboratories. Basic studies performed in the 1990s investigated the dynamics of cell proliferation, neurogenesis and gliogenesis in the hippocampus and the olfactory bulb in great detail (45, 46). It has been established that, in order to gain a good measure of neurogenesis, two specific time points are important: 1 day and 4 weeks after BrdU application. The 24-h time point after BrdU provides a measure of progenitor cell proliferation, whereas cells positive for BrdU that are found in the subgranular zone 4 weeks after BrdU application provide a measure of cell survival (6, 7, 47). Some of the studies that investigate certain modulators of adult neurogenesis, such as the cannabinoid system, focus on an intermediate time point of 1 week after BrdU application (39). It is well known that the rate of proliferation and neurogenesis can both increase or decrease depending on how the compound or treatment affects the progenitor cells (6, 7, 46). There are compounds that increase proliferation, such as kainate acid, but decrease cell survival. Voluntary exercise has a greater effect on cell proliferation than cell survival whereas cognitive stimulation does not influence proliferation but promotes cell survival to a greater extent than wheel running (47, 48). Thus, a clear distinction needs to be made as to whether one wishes to measure an effect on proliferation or survival rate. This distinction is less distinct when intermediate time points are measured.

Taken together, it is extremely difficult to compare studies involving the cannabinoid system because variables such as strain, rodent model and methods are so heterogeneous. Thus, although a large amount of important data has been presented so far, inherent differences in study design suggest that a clear picture of how the cannabinoid system regulates adult neurogenesis is not yet apparent.

Studies by Aguado, Galve-Ropher, Guzman and others have described the existence of parts of the endogenous cannabinoid system such as the CB1 receptor, AEA production and FAAH production by progenitor cells while also measuring proliferation and survival rate as well as phenotyping the BrdU positive cells in the hippocampus (49–51). These studies show that CB1 receptor activation promotes cell proliferation and neurosphere generation, an action that is abrogated in CB1-deficient NPs (on a CD1 background) (38, 50). Accordingly, proliferation of hippocampal neuronal progenitors (NPs) is increased in FAAH-deficient mice. A recent study by the same group showed that CB1 receptor ablation prevents the excitotoxicity induced hippocampal basic fibroblast growth factor, brain-derived neurotrophic factor and endothelial growth factor production. The authors link this regulation with CB1 dependent modulation of injury induced NP proliferation and neurogenesis that points towards the already known neuroprotective properties of cannabinoids (50, 51). Jiang et al. (52) showed that the CB1 receptor agonist HU210 promoted proliferation but not differentiation of progenitor cell cultures and that chronic treatment with HU210 produces anxiolytic- and antidepressant-like effects. Therefore, it is possible that there is a link between the functional outcome and the generation of new neurones. As discussed elsewhere, neurogenesis and depression might be non-identical twins. They occur at the same time and space in the same individual, but are completely independent from each other (52). Nevertheless decreased neurogenesis could be used as a biomarker for pathophsychological disease, at least post mortem.

Derivatives of the Cannabis sativa plant are known for their beneficial effects and are already used as medication for multiple sclerosis (53), especially cannabidol, which has great therapeutic potential. The routes of action are not very well understood. The modulation of neurogenesis by the endocannabinoid system might provide a link between the observed therapeutic effects and cellular mechanisms. On the other hand Δ⁹-tetrahydrocannabinol (THC) is known for its psychogenic effects, which can lead to cognitive impairments when used chronically. Indeed, it has been shown that chronic exposure to Δ⁹-THC decreases neurogenesis in the hippocampus of mice (S. A. Wolf, unpublished data). Although acting via the same receptor, Δ⁹-THC and HU210, as well as AEA, have opposite effects on neurogenesis (39, 52). The effects of cannabidol are not thought to be strongly CB1 dependent (54). Cannabidol promotes survival of the NPCs increasing net neurogenesis (S. A. Wolf, unpublished data). This might be one of the mechanisms via which cannabidol is beneficial in certain pathological brain conditions. Again, also in this case neurogenesis might only be a parallel effect of cannabidol action with no causal link to the functional outcome. During a pathological event, enhanced survival of newly-generated neurones promoted by cannabidol might be of medical interest.

Whether neurogenesis proves to only be useful as a post-mortem biomarker, becomes a therapeutic target or gives us insight into stem cell biology, the modulators of this process need to be well understood. In a recent breakthrough, a metabolic biomarker for the detection and quantification of neural progenitor cells in human brain has been described, which enables the detection of low concentrations of progenitor cells by magnetic resonance spectroscopy in humans (55).

Over the past 5 years, it has become evident that cannabinoids have an impact on neuronal progenitor cell proliferation and survival, an effect partially mediated via the CB receptors by influencing NPC cell cycle, differentiation and lineage choice. The route of action and kinetics differ between the endogenous cannabinoids, synthetic compounds, and the plant derived cannabinoids. This is a relatively new area of scientific research and, as a result, comparisons between studies should be made very carefully. The base is just set for more detailed and sophisticated investigations. In further studies that link the two research fields, a close look back to the roots of neurogenesis should be much anticipated.

Conflicts of interest

The authors have declared no conflicts of interest.