Hippocampal Cb 2 receptors: an untold story

: The field of cannabinoid research has been receiving ever-growing interest. Ongoing debates worldwide about the legislation of medical cannabis further motivates research into cannabinoid function within the central nervous system (CNS). To date, two well-characterized cannabinoid receptors exist. While most research has investigated Cb 1 receptors (Cb 1 Rs), Cb 2 receptors (Cb 2 Rs) in the brain have started to attract considerable interest in recent years. With indisputable evidence showing the wide-distribution of Cb 2 Rs in the brain of different species, they are no longer considered just peripheral receptors. However, in contrast to Cb 1 Rs, the functionality of central Cb 2 Rs remains largely unexplored. Here we review recent studies on hippocampal Cb 2 Rs. While con ﬂ icting results about their function have been reported, we have made signi ﬁ cant progress in understanding the involvement of Cb 2 Rs in modulating cellular properties and network excitability. Moreover, Cb 2 Rs have been shown to be expressed in different subregions of the hippocampus, challenging our prior understanding of the endocannabinoid system. Although more insight into their functional roles is necessary, we propose that targeting hippocampal Cb 2 Rs may offer novel therapies for diseases related to memory and adult neurogenesis de ﬁ cits.


Introduction
Peripheral Cb 2 receptorsa half-truth Our understanding of the endocannabinoid system (ECS) is continuously being shaped by novel discoveries of complex interactions and intricate processes. Since the initial description of the cannabinoid (Cb) receptors, namely the Cb 1 receptors (Cb 1 Rs) (Matsuda et al. 1990) and the Cb 2 receptors (Cb 2 Rs) (Munro et al. 1993), to the discoveries of the "blissful" substances anandamide (Devane et al. 1992) and 2-arachidonylglycerol (Mechoulam et al. 1995;Stella et al. 1997), emerging roles of the ECS in several brain processes have been recognized. Interestingly, the ECS encompasses both the central nervous system (CNS) and peripheral tissues (Mechoulam and Parker 2013). The Cb 1 R is widely distributed in the CNS, with abundant expression in the basal ganglia, cortex, cerebellum, and hippocampus, as shown by radiography (Herkenham et al. 1991), in situ hybridization (Mailleux and Vanderhaeghen 1992;Matsuda et al. 1993) and immunohistochemistry (Egertová and Elphick 2000;Tsou et al. 1998). These receptors are typically expressed on axon terminals and mediate retrograde signaling by endocannabinoids (Castillo et al. 2012). Much less is known about Cb 2 R-mediated signaling. In fact, for several years central Cb 2 Rs were largely overlooked and more attention was given to the peripheral Cb 2 Rs (Atwood and Mackie 2010;Kopach et al. 2012). Most cannabinoid research in the CNS to date has thus been focused on Cb 1 R. Recently, however, Cb 2 Rs have attracted considerable interest as potential modulators in drug-seeking behavior, pain, depression, anxiety, memory, neuroinflammation, and neurodegenerative diseases (Chen et al. 2017). During the last two decades, numerous studies have helped disperse the myth of Cb 2 Rs being exclusively expressed in the periphery. In situ hybridization and quantitative realtime PCR detected Cb 2 R mRNA in the hippocampus, cortex, cerebellum, brainstem, and midbrain of both rodents and nonhuman primates Li and Kim 2015;Liu et al. 2009;Navarrete et al. 2012;Zhang et al. 2014). Moreover, Cb 2 R protein expression has been demonstrated for various brain regions (Ashton et al. 2006;Baek et al. 2008;Brusco et al. 2008;Gong et al. 2006;Van Sickle et al. 2005). Cannabinoid receptor expression has also been investigated in postmortem human brain tissue: Cb 2 R mRNA was found in the human prefrontal cortex while the expression of both Cb 1 R and Cb 2 R proteins was demonstrated in the cerebellum (Rodríguez-Cueto et al. 2014a,b).
Brain Cb 2 Rs are attractive therapeutic targets. Despite the low physiological expression, they are highly inducible under some pathological conditions, and their expression is quickly enhanced in the brain. The upregulation of Cb 2 Rs has been described in several disorders, including neurodegenerative diseases, brain injuries, and neuroinflammation. Among them, Parkinson's disease (Gómez-Gálvez et al. 2016;Grünblatt et al. 2007;Navarrete et al. 2018), Alzheimer's disease (AD) (López et al. 2018), post-traumatic stress disorder (PTSD) (Lisboa et al. 2019;Morena et al. 2018), traumatic brain injury (Tchantchou and Zhang 2013), vascular dementia (Luo et al. 2018), stroke (Zarruk et al. 2012), and neuroinflammation (Torres et al. 2011). The inducibility of Cb 2 R expression in pathological conditions was found mainly in the hippocampus, but there is also evidence for its occurrence across the CNS, including the cerebral cortex and the cerebellum ( Figure 1A). The Cb 2 Rs are also induced in the spinal cord, in in vitro preparations for studying multiple sclerosis and in regions remotely connected to a primary site of focal brain damage during remote cell death ( Figure 1B) (Askari and Shafiee-Nick 2019;Viscomi et al. 2009;Wen et al. 2015). Pathology-induced Cb 2 Rs were found in microglia, astrocytes, oligodendrocytes, and neurons. The activation of Cb 2 Rs in these cells modulates the release of several cytokines, which regulate immune function and inflammatory responses ( Figure 1C).
Regional increases in Cb receptor expression have been shown to modulate the potency and efficacy of exogenous agonists at disease sites, theoretically allowing for targeted activation at the local of injury (Miller and Devi 2011). However, there has only been limited success in transitioning Cb 2 R agonists from preclinical studies to clinical trials (Dhopeshwarkar and Mackie 2014). But it should be noted that even for the case of Cb 1 Rs, which have been extensively studied through decades, there is a paucity of well-established clinical applications. This reflects the complex role of the ECS in modulating brain function and could indicate the need for improvement in preclinical models . Notwithstanding, Cb 2 Rs have remarkable advantages over Cb 1 Rs as potential therapeutic targets (Chen et al. 2017). In contrast to Cb 1 Rs, Cb 2 Rs are less prone to produce psychotropic effects, which have been a key point in the ongoing worldwide debates regarding the legislation of medical marijuana (Dhopeshwarkar and Mackie 2014). Moreover, Cb 2 Rs have lower expression levels and a more specific distribution than Cb 1 Rs in the brain. Thus, Cb 2 R ligands could offer therapeutic treatments without the adverse effects often seen with Cb 1 R-ligands (Onaivi et al. 2011). Furthermore, Cb 2 Rs can balance Cb 1 R activation effects since the former are expressed in neuronal somatodendritic areas, and the latter are predominantly expressed on neuronal terminals . Considering these characteristics, studies aiming to unveil the Cb 2 R role in the neurophysiology of specific brain regions may foster the development of clinically effective Cb 2 R-modulators, which will likely offer novel strategies for treating neuropsychiatric and neurological diseases.
In this review, we discuss recent developments in hippocampal Cb 2 R research. We selected the hippocampus as a major area of interest since it plays a crucial role in cognition, learning, and memory, which are important functions disrupted in neurodegenerative and neuroinflammatory diseases (Leuner and Gould 2010). Moreover, the layers, subregional differences, and cell populations within this structure are well-known, facilitating inferences about potential Cb 2 R roles in hippocampal activity. Although the function of hippocampal Cb 2 R is still largely unexplored, recent reports show that they modulate cellular and network excitability, suggesting a meaningful role in regulating hippocampal output. As will be argued below, further exploring the function of hippocampal Cb 2 R is an important step to understand the cognitive effects of both exogenous and endogenous cannabinoids.

Different cell populations express Cb 2 receptors in the hippocampus
In a series of experiments, Onaivi and colleagues explored the distribution of Cb 2 Rs in the hippocampus, and found them to be broadly distributed among its different subregions (Brusco et al. 2008;Gong et al. 2006;Onaivi 2006;Onaivi et al. 2006) (Figure 2). Together, these experiments also provided the first evidence for the expression of hippocampal Cb 2 Rs in microglia, principal neurons and interneurons, which encouraged further research. Below we discuss the findings for each cell type individually.

Cb 2 receptors in microglia
The function of peripheral Cb 2 Rs as potent immune modulators has been clearly demonstrated (Basu and Dittel 2011;Cabral and Griffin-Thomas 2009;Racz et al. 2008;Turcotte et al. 2016). Accordingly, the fact that the CNS microglial cells express Cb 2 Rs is not surprising since they are the resident macrophages (Perry and Teeling 2013). Microglial Cb 2 R expression in mouse and rat hippocampi was first demonstrated using different Cb 2 R-specific antibodies (Brusco et al. 2008;Gong et al. 2006). More recent studies challenge some of the results previously shown by Onaivi and colleagues. For instance, when combining RNAscope, an ultrasensitive in situ hybridization technique, with immunostaining against the microglial marker Iba1, no overlap could be detected in the CA1 region of healthy rat hippocampi . Some publications also challenge the specificity of Cb 2 R antibodies, and overall results indicate that currently available antibodies may lack specificity and may lead to conflicting outcomes . Considering these results, more validating research needs to be conducted. It is worth noting that, usually, the experiments that failed in finding Cb 2 R

Cerebellum
Parkinson's disease Alzheimer's disease Stress/PTSD Traumatic brain injury Vascular dementia model

A C B
Multiple Sclerosis (EAE model) expression in hippocampal cells (microglia, principal neurons, and interneurons) were conducted using healthy tissue, while inflammatory responses in the brain have been suggested to increase Cb 2 R and its mRNA expression (Guida et al. 2017;Luongo et al. 2010;Palazuelos et al. 2009;Walter et al. 2003). Hence, the absence of microglial Cb 2 R expression in healthy hippocampal tissue would not exclude a potential functional relevance of this receptor during pathological conditions.

Cb 2 receptors in principal neurons
Most Cb 2 Rs in the CA1 region of the hippocampus are expressed in the principal neurons, which are the excitatory pyramidal cells Onaivi 2006). The localization of Cb 2 Rs in these cells is primarily postsynaptic. Still, they can also be observed in the rough endoplasmic reticulum, Golgi apparatus, neuronal cytoplasm, and in dendrites near the plasma membrane (Brusco et al. 2008). These findings support that Cb 2 Rs are synthesized in the soma and subsequently transported to target dendrites. Interestingly, no expression occurs in axon terminals (Brusco et al. 2008), and since hippocampal Cb 1 Rs are mainly expressed presynaptically (Castillo et al. 2012;Kano et al. 2009;Katona et al. 1999;Monory et al. 2015), this indicates a functional difference between Cb 1 and Cb 2 receptors. Moreover, this also suggests that the notion of the ECS as a retrograde signaling system might not be complete (Castillo et al. 2012). The postsynaptic localization of Cb 2 Rs in principal cells thus reveals a more complex role for the ECS in the hippocampus than previously thought.

Cb 2 receptors in interneurons
Hippocampal interneurons consist of a morphologically diverse group of cell types. It has been suggested that there are at least 21 different classes of interneurons in the CA1 region alone (Klausberger and Somogyi 2008). In this region, the Cb 1 Rs are found primarily in GABAergic interneurons (Tsou et al. 1999). In contrast, Cb 2 R mRNA is found in about 20% of both glutamatergic and nonglutamatergic cells in the CA1 . Whether there exists an overlap of Cb 1 R and Cb 2 R expression within specific interneuron types is still open for exploration. Further understanding the localization of Cb 2 R expression might add a new perspective on the action of cannabinoids in the CA1 region, as some of the functions previously thought to be mediated by Cb 1 R could potentially be due to overlooked Cb 2 intra-and/or interneuronal signaling cascades.

Moving from expression to functionality
Correlation does not imply causation, a mantra for science. Similarly, expression does not denote function. Therefore, albeit evident that both Cb 2 R mRNA and protein can be found in the hippocampus, whether this has functional relevance is debatable, as we revisit below.

Synaptic function of hippocampal Cb 2 receptors
Although the levels of Cb 2 R mRNA are significantly lower in the CNS than in the periphery (Onaivi 2006;Van Sickle et al. 2005), there is enough evidence to support a functional importance of Cb 2 R for hippocampal activity.
Traditionally, the functional effects of the ECS in the hippocampus have been credited to Cb 1 R, which modulate presynaptic neurotransmitter release (Monory et al. 2015). Depending on the neuronal cell type and brain region, the activation of Cb 1 R can have opposite effects, either increasing or decreasing excitability (Chevaleyre and Castillo 2003;Miraucourt et al. 2016;Winters et al. 2012). Whether this is also the case for Cb 2 R is an open question. In layers II and V of the medial entorhinal cortex, activation of Cb 2 R decreases the amplitude of spontaneous inhibitory postsynaptic currents through suppression of GABAergic transmission (Morgan et al. 2009). On the other hand, a recent study reported that inhibitory synaptic transmission is not affected by acute activation of Cb 2 Rs in CA1 but rather that chronic activation of Cb 2 Rs results in increased excitatory transmission . The regional or cellspecific factors underlying the different actions of Cb 2 Rs in the hippocampus and medial entorhinal cortex have yet to be identified.
As stated above, whether the low levels of Cb 2 Rs in CA1 are relevant to synaptic function under physiological conditions is still an open question Li and Kim 2016a;Stempel et al. 2016). To further complicate matters, the postsynaptic localization of Cb 2 Rs has also been challenged. Namely, Morgan et al. (2009) observed no changes in the kinetics of miniature inhibitory postsynaptic currents in the presence of a selective Cb 2 R agonist, which would be expected if Cb 2 R were postsynaptically located (Morgan et al. 2009). Therefore, there are important unanswered questions regarding the synaptic function and localization of Cb 2 R, but still enough pieces of evidence suggesting a functional Cb 2 R role in the hippocampus, which could lead to new interpretations of the effects of exogenous and endogenous cannabinoids in this region.

Cb 2 receptors and memory consolidation
As discussed above, there is evidence of upregulation of Cb 2 R in microglia as a response to neuroinflammation, but the functional consequences are not yet clear. In addition, Cb 2 R mRNA and protein expression are upregulated in AD, for which one of the hallmark features is hippocampal-dependent memory impairment. However, how do Cb 2 R relate to memory consolidation? Köfalvi et al. (2016) observed in hippocampal slices of both young and old healthy mice that Cb 2 R activation increases glucose transporters (GLUT) in hippocampal astrocytes and neurons. In contrast, these authors reported that the glucose uptake induced by Cb 2 R activation is impaired in a mouse model of AD (TgAPP mice). TgAPP mice present β-amyloid-burden and object recognition memory impairment. Interestingly, prolonged oral administration of JWH-133, a selective agonist of Cb 2 Rs, rescued hippocampal glucose uptake, diminished β-amyloid levels, and prevented the memory deficit in TgAPP mice (Köfalvi et al. 2016;Martín-Moreno et al. 2012). Dagon et al. (2007) induced hepatic encephalopathy in wild-type and Cb 2 R knockout mice. Hepatic encephalopathy is a neuropsychiatric syndrome caused by liver dysfunction and characterized by impaired glucose oxidative pathways in the brain, amnesia, and confusion. The authors found that treatment with Δ9-tetrahydrocannabinol (THC) increased AMP-activated protein kinase, which in turn stimulated GLUT expression and transport efficiency in the hippocampus. Interestingly, THC also prevented spatial working memory deficit assessed by the eight-arm maze in wild-type mice but not in Cb 2 R knockout animals (Dagon et al. 2007). Given that brain glucose availability controls cognition and memory in humans (Messier 2004) and that central metabolic boosting alleviates the cognitive symptoms of brain disorders (Branconnier 1983), the studies mentioned above support a role of hippocampal Cb 2 Rs in counteracting cognitive impairment via regulation of glucose uptake.
In Cb 2 R knockout mice, hippocampal-dependent long-term contextual fear memory is impaired while hippocampal-independent cued fear memory is not affected (Li and Kim 2016a). A follow-up study showed that knocking out the Cb 2 R gene decreases hippocampal excitatory synaptic transmission, long-term potentiation, and dendritic spine density, indicating that the endogenous activity of Cb 2 R contributes to the maintenance of synaptic function and regulates cognitive functions such as long-term memory (Li and Kim 2016b). These results suggest that the loss of Cb 2 R may lead to hippocam pal-dependent memory deficits, though they should be interpreted with caution since compensatory mechanisms may occur in developmental knockout mice. Furthermore, these results were obtained using a general Cb 2 R knockout mouse line and it is therefore not possible to infer whether memory impairment was specifically due to the loss of Cb 2 Rs in the hippocampus. Subsequently, Li and Kim (2017) used either Cre-dependent overexpression of Cb 2 Rs or CRISPR-Cas9 genome-editing techniques to delete Cb 2 R gene in combination with the injection of adeno-associated viruses into the dorsal hippocampus of transgenic mouse lines. With this approach, they were able to investigate the role of Cb 2 Rs in specific cell R. Visvanathar et al.: Hippocampal Cb 2 receptors: an untold story populations (i.e., pyramidal cells, interneurons, and microglia) and found that increasing or decreasing the expression of Cb 2 Rs in microglia respectively enhances or impairs contextual fear memory (Li and Kim 2017). They also showed that disruption of Cb 2 R expression in CA1 pyramidal neurons enhances spatial working memory, while overexpression reduces anxiety levels as tested by the open field test (Li and Kim 2017). Noteworthy, in studies that genetically modulate Cb 2 R expression it is impossible to disentangle if the memory impairments are due to hindered consolidation, acquisition or retrieval. In this regard, despite the limitations in the current available Cb 2 R agonists/antagonists, pharmacological studies are more advantageous to investigate the functions of CB 2 R in specific memory phases. Nasehi et al. (2017Nasehi et al. ( , 2018 reported that microinjection of a Cb 2 R agonist (Gp1a) into CA1 impairs aversive memory consolidation in rats and mice. Moreover, aversive memory consolidation was further impaired when simultaneously injecting muscimol (an ionotropic GABA receptor agonist), suggesting an interactive effect between Cb 2 and GABA A signaling (Nasehi et al. 2017(Nasehi et al. , 2018. This notion is also supported by Garcia-Gutiérrez and Manzanares (2011), who reported an upregulation of GABA A protein expression after chronic activation of Cb 2 Rs in the cortex. A more recent study showed that Gp1a infusion into CA3 also impairs aversive memory consolidation. This effect was increased by coinfusion of scopolamine (a nonselective antagonist of muscarinic acetylcholine receptors), suggesting that Cb 2 Rs can also interact with cholinergic signaling (Nasehi et al. 2020). The suggestion of an interaction between cannabinoid and cholinergic signaling during memory processing was made previously by Robinson et al. (2010), although without a direct mention of Cb 2 Rs. They reported that intraperitoneal administration of WIN55,212-2 (a nonselective cannabinoid receptor agonist) before spatial memory acquisition caused memory impairment through a mechanism that was independent of Cb 1 R, and that this impairment was reversed by coinfusion of a cholinesterase inhibitor (Robinson et al. 2010). Similarly, studies in rats showed that cannabidiol disrupts consolidation of (specific and generalized) fear memories via Cb 2 R localized in the dorsal hippocampus (Raymundi et al. 2020;Stern et al. 2017).
It should be noted, however, that in addition to the studies showing that activating Cb 2 R signaling has disruptive effects on memory, there is also evidence for heightened Cb 2 R activity improving spatial and fear memory. Chronic treatment with Cb 2 R agonists and Cb 2 R upregulation could rescue spatial memory deficits in mouse models of vascular dementia or AD (Çakır et al. 2019;Lou et al. 2017;Wu et al. 2017). Moreover, Ratano et al. (2018) showed that the endocannabinoid 2-arachidonoilglycerol (2-AG) enhanced memory consolidation in a inhibitory avoidance task through a Cb 2 R-dependent modulation of mTOR signaling. In a previous study of the same group, they also showed that blocking Cb 2 R signaling tended to impair fear memory retention (Ratano et al. 2017). Consistently, chronic treatment with systemic Cb 2 R antagonist aggravated fear memory loss caused by orthopedic surgery ). Nevertheless, downregulation of Cb 2 R expression in hippocampal cells has been associated with impaired spatial memory, object recognition, and fear conditioning acquisition and retention (Tang et al. 2017).
To summarize, it is clear that hippocampal Cb 2 Rs do influence memory, but contradictory results have been reported regarding the exact role of these receptors on memory consolidation and disruption. While Cb 2 R gene overexpression in the dorsal hippocampus enhanced fear conditioning, acute Cb 2 R agonism in the same region impaired fear conditioning consolidation (Li and Kim 2017;Raymundi et al. 2020). Nevertheless, increasing Cb 2 R activity in CA1 and CA3 impaired memory consolidation in inhibitory avoidance, but systemic Cb 2 R agonism enhanced memory retention in this same task (Nasehi et al. 2017(Nasehi et al. , 2018(Nasehi et al. , 2020Ratano et al. 2018). In the Morris water maze, one study reported impaired spatial memory after treatment with a potent cannabinoid receptor agonist (Robinson et al. 2010), but several others showed improvement of spatial learning and prevention of memory deficits with selective Cb 2 R agonism (Çakır et al. 2019;Lou et al. 2017;Sun et al. 2017;Wu et al. 2017) (for an overview, see Table 1). These results suggest that Cb 2 R activation differently modulates aversive and neutral memories. Consistently, Cb 2 R mRNA transcripts in the DG and CA1 regions of the dorsal hippocampus were shown to be increased in stressed and anxious mice, further supporting the inducibility of Cb 2 R expression and indicating a role in coping mechanisms (Robertson et al. 2017). On the other hand, most studies using Cb 2 R knockout or antagonism showed impaired neutral, aversive, and spatial memory (see Table 2). But there is also evidence for contrasting promnesic and amnesic effects of Cb 2 R knockout on the Y-maze and fear conditioning, respectively (Li and Kim 2016b). Therefore, Cb 2 Rs modulate memory and behavior, as well as anxiety and stress, but less is known about their specific role and the exact mechanisms underlying their function (Figure 3). With increasing research interests and emerging techniques such as optogenetics and DREADDs, we hope that these questions will be addressed soon.

Cb 2 receptors and hippocampal adult neurogenesis
Previously we discussed the expression and some of the functions of Cb 2 R present in fully differentiated hippocampal cells. However, what about the hippocampal neural progenitor cells? For decades, scientists believed that the adult brain did not generate new neurons. This belief persisted until Altman and Das (1965) first reported neurogenesis in the DG of adult rodents, a finding that was later confirmed by several studies (Gonçalves et al. 2016;Jorgensen 2018). Currently, it is widely accepted that hippocampal adult neurogenesis happens in the subgranular zone of the DG of humans and several other vertebrates (Gonçalves et al. 2016;Jorgensen 2018). This process has been associated with memory and learning (Cameron and Glover 2015;Deng et al. 2010), mood disorders (Jorgensen 2018;Snyder et al. 2011), and neurological diseases (Horgusluoglu et al. 2017).
In the last decades, several studies have shown an important role of Cb 2 R in hippocampal adult neurogenesis. First, Palazuelos et al. (2006)   (2) ablation of Cb 2 R signaling through knockout impairs adult neurogenesis; and (3) Cb 2 R activation increases hippocampal adult neurogenesis (Palazuelos et al. 2006). Then, in a follow-up study, Palazuelos et al. (2012) reproduced their previous results and extended them by showing that Cb 2 R promotes adult neurogenesis through the activation of the PI3K/Akt/mTORC1 pathway. This results in the inhibition of the cyclin-dependent kinase inhibitor p27Kip1, a protein that inhibits the G 1 -S phase transition in neural progenitor cells (Palazuelos et al. 2012). Later, Avraham et al. (2014) showed that Cb 2 R activation could reverse the deficits in hippocampal adult neurogenesis caused by the human immunodeficiency virus (HIV) glycoprotein 120 (Gp120) and could thus be a potential mechanism to treat HIV-associated neurocognitive disorders (Avraham et al. 2014). On the other hand, a recent study by Rodrigues et al. (2017) partially contradicted the previously mentioned results by showing that activation of Cb 2 Rs alone is not enough to induce the proliferation of DG neural precursor cells. Instead, activation of both Cb 2 Rs and Cb 1 Rs was necessary to induce this proliferation. Furthermore, they showed that, although Cb 2 R agonism is enough to induce differentiation of DG neural precursor cells, Cb 1 R signaling is still needed since its blockade prevents the effect of Cb 2 R agonism. Finally, they suggest the formation of Cb 2 R-Cb 1 R heteromers in these cells, which could be controlling neuronal differentiation (Rodrigues et al. 2017). Moreover, another recent study showed that Cb 2 R-deficient mice have normal hippocampal adult neurogenesis, suggesting that Cb 2 R signaling might not be necessary for basal hippocampal adult neurogenesis (Mensching et al. 2019). However, since the animals used in this study were constitutive knockouts, compensatory mechanisms might have influenced the results.

used both in vitro and in vivo
In sum, these studies strongly suggest that Cb 2 R signaling is important to control/modulate hippocampal adult neurogenesis. However, it is still not fully clear if it acts alone or in conjunction with Cb 1 R signaling or the specific mechanisms involved. Thus, further experiments are necessary to address these questions and whether Cb 2 R signaling is necessary for basal hippocampal adult neurogenesis or is only recruited in specific situations.

Concluding remarks and future perspectives
There is now enough data to support not only the existence but also the functional relevance of hippocampal Cb 2 Rs, which challenges our prior understanding of ECS action in the CNS and warrants further exploration. Hopefully, modern techniques will offer more robust approaches to answering some of the outstanding questions and shed light on the contradictory results in the literature. Among them, we should address the suggested postsynaptic localization of hippocampal Cb 2 Rs, which would crucially differentiate them from presynaptic Cb 1 Rs. Both light-and electron microscopy may help in this regard, and even more advanced techniques such as super-resolution microscopy could produce robust results (Cristino et al. 2017). Moreover, it is worth noting that the studies reviewed by us Figure 3: Cellular mechanisms of Cb 2 R function in a variety of expression systems and animal species. Activation of Cb 2 Rs has a variety of downstream effects via different pathways. These effects are mostly mediated by the activation of the coupled G-protein (represented by its α, β, and γ subunits). Cb 2 R activation can: (1) mediate plasticity in hippocampal CA3 principal neurons via a G-protein and Na + -dependent modulation of the sodium/bicarbonate co-transporter (NBC) (Stempel et al. 2016); (2) lead to a G-protein gated inward rectifying potassium channel (GIRK) mediated cell-autonomous hyperpolarization (Stumpf et al. 2018); (3) through phospholipase C (PLC) production, lead to Ca 2+ release via IP 3 which in turn leads to the opening of Ca 2+ -activated Cl − channels (CACCs) in layer II/III pyramidal neurons of the rat medial prefrontal cortex (den Boon et al. 2012); (4) negatively affect the production of adenylyl cyclase, which, through decreased cyclic adenosine monophosphate (cAMP) production results in reduced activation of protein kinase A (PKA) (Dhopeshwarkar and Mackie 2014); (5) mediate suppression of voltage-gated calcium channels (VGCCs) via reduced activation of PKA (Qian et al. 2017); (6) cause increased phosphorylation of p38 MAPK, JNK1 and JNK2 (Neves et al. 2018); (7) mediate upregulation of the PI3K/Akt pathway, increasing the activity of mTORC1 (composed by mTOR, Raptor, Deptor, mLST8, PRAS40, Ttti1, and Tel2), which in turn inhibits the cyclin-dependent kinase inhibitor p27Kip1 (Cao et al. 2018;Palazuelos et al. 2012); (8) through activation of sphingomyelin phosphodiesterase (SMase), can result in increased ceramide production (Askari and Shafiee-Nick 2019). For full reviews of Cb 2 R function mechanisms and therapeutic potentials, see Aghazadeh Tabrizi et al. (2016) and Cassano et al. (2017). Cb 2 R crystal structure taken from RCSB Protein Data Bank (https://www.rcsb.org/structure/5ZTY) (Li et al. 2019 reported the discovery and active presence of Cb 2 Rs in the mammalian brain. Still, more quantitative approaches are needed to provide information to support differences in Cb 2 R expression along the hippocampal subregions and between their ventral and dorsal portions. This would improve the discussion of the functions of hippocampal Cb 2 Rs since the dorsal and ventral parts differently contribute to memory, anxiety, neurogenesis, and related pathologies (Fanselow and Dong 2010;Nadel et al. 2013). Until now, the most common approach for studying Cb receptor activity has been through pharmacology, with a large degree of uncertainty regarding the specificity of the compounds used (Console-Bram et al. 2012). Novel Cb 2 Rspecific compounds and emerging transgenic tools now offer more targeted methods (Bickle 2016;Nevalainen 2014), and further combining these tools with current knowledge of regional and neuronal diversity should generate significant new insights.
The Cb 2 Rs may play a crucial role in regulating hippocampal-dependent memory formation and adult hippocampal neurogenesis, particularly during neuroinflammatory conditions where Cb 2 Rs are upregulated. The properties of low expression but high-inducibility during pathological conditions should incite more research on therapeutic strategies with selective Cb 2 Rmodulators, especially considering their lower psychoactive effects than those of Cb 1 R-ligands (Chen et al. 2017). Further exploring hippocampal Cb 2 Rs should not only increase our understanding of the ECS but also contribute to the debate on the legislation of medical cannabis that currently concerns several countries worldwide. The emerging field of hippocampal Cb 2 Rs provides avenues for exploration and discovery; insightful times lie ahead.

Outstanding questions
(1) Is there a functional relationship between central Cb 1 and Cb 2 receptors? Can they form heteromers? If so, are they present in adult cells and what would be the functional implications? (2) Are Cb 2 Rs mostly expressed postsynaptically? If so, should we replace the retrograde signaling view of the ECS in the CNS by a more complex signaling dynamics? (3) How are Cb 2 Rs expressed in different hippocampal subregions and neuronal cell types? Would they mark a specific subset of interneurons? And what is the balance of expression between immune and nonimmune cells? (4) What are the mechanisms of action of Cb 2 Rs in the hippocampus? Would the activity of Cb 2 Rs in microglia also influence network states?
(5) Is there a therapeutic role for Cb 2 R ligands, through orthosteric or allosteric binding, in neurodegenerative or neuroinflammatory diseases?