The Impact of the CB2 Cannabinoid Receptor in Inflammatory Diseases: An Update

The emergence of inflammatory diseases is a heavy burden on modern societies. Cannabis has been used for several millennia to treat inflammatory disorders such as rheumatism or gout. Since the characterization of cannabinoid receptors, CB1 and CB2, the potential of cannabinoid pharmacotherapy in inflammatory conditions has received great interest. Several studies have identified the importance of these receptors in immune cell migration and in the production of inflammatory mediators. As the presence of the CB2 receptor was documented to be more predominant in immune cells, several pharmacological agonists and antagonists have been designed to treat inflammation. To better define the potential of the CB2 receptor, three online databases, PubMed, Google Scholar and clinicaltrial.gov, were searched without language restriction. The full texts of articles presenting data on the endocannabinoid system, the CB2 receptor and its role in modulating inflammation in vitro, in animal models and in the context of clinical trials were reviewed. Finally, we discuss the clinical potential of the latest cannabinoid-based therapies in inflammatory diseases.


Introduction
Inflammatory diseases are the leading causes of disability and death throughout developing countries [1].Therefore, it is more and more urgent to find a therapeutic target that could improve the conditions of people suffering from inflammatory diseases [2].These disorders evolve from a sustained activation of the immune system that can be localized and/or disseminated throughout the body [3].Chinese traditional healers have used cannabis to treat inflammation since 2000 before Common Era [4].Since the identification of cannabinoid receptors in immune cells, the potential of cannabinoid pharmacotherapy in pain and inflammatory conditions has received much attention [5].Here, we are updating the literature review by Turcotte et al. [6], highlighting the studies of the last decade on the role of the CB 2 receptor in peripheral inflammation and its potential as a pharmaceutical target in inflammatory diseases.

The Type 2 Cannabinoid Receptor (CB 2 )
Our knowledge of the endocannabinoid system started with the elucidation of the structure, metabolism and function of ∆ 9 -tetrahydrocannabinol (∆ 9 -THC) [7][8][9].The first cannabinoid receptor CB 1 was isolated and cloned from a rat cerebral cortex cDNA library.The translated sequence of the identified cDNA was a 473 amino acid protein sequence and a member of the G-protein-coupled family of receptors [10].Two years later, the human cannabinoid receptor CB 2 was cloned from the promyeloid cell line HL-60 [11].The protein encoded by a sequence of 360 amino acids was found to have 44% homology with the CB 1 receptor [12,13].The CB 2 receptor is considered a peripheral receptor due to its predominant expression in various immune cells [14,15].Eosinophils, B cells and natural killer (NK) cells were found to strongly express CB 2 , whereas monocytes, neutrophils, helper T cells and cytotoxic T cells showed mRNA transcript levels ranging from moderate to low [16,17].The membrane expression of CB 2 was also accessed by flow cytometry, and it was shown that circulating B cells, NK cells and monocytes express higher levels of the receptor compared to CD4 + T lymphocytes, CD8 + T lymphocytes and neutrophils [18].Another study confirmed the membrane expression of CB 2 on eosinophils and found that it was increased in eosinophils isolated from the blood of allergic patients compared to those isolated from healthy donors [19].Furthermore, functional CB 2 is detected in perivascular microglia [20] and the brain [21].Therefore, it is suggested that CB 2 is also expressed in the healthy central nervous system albeit at lower transcript levels and not necessarily in neurons [21].
To better understand the expression of CB 2 in the inflammatory context, Turcotte et al. [6] and others comprehensively reviewed the expression of CB 2 in human tissues and animal models and highlighted the role of CB 2 as a regulator of inflammation.For instance, in the gastrointestinal tract, CB 2 is detected in the esophagus, stomach and ileum [22].In humans, CB 2 is expressed by colonic epithelial cells but only under inflammatory conditions [23].The expression of CB 2 by the liver is specifically high during embryonic development [24], and its expression by hepatic myofibroblasts increases under pathological conditions such as liver fibrosis [25] and liver injury [26].In the cardiovascular system, CB 2 is expressed by cardiomyocytes and endothelial cells [27], and its expression increases in an inflammatory context such as atherosclerosis and acute myocardial infarction [28].In the musculoskeletal system, the protein expression of CB 2 by chondrocytes has been detected and shown to be increased in the joint tissues of rats [29,30] and humans with osteoarthritis [31].In the brain, CB 2 expression is still controversial, as it seems to be detected only in the context of neuroinflammation and is attributed to microglia [32].However, other studies have reported that CB 2 is detected in neurons and may mediate brain functions [33].Moreover, the expression of this receptor in innate and adaptive immune cells was recently reviewed by Simard and colleagues [17].In particular, they showed that high CB 2 (Cnr2) expression is found in eosinophils and B cells, while low expression is found in T cells and monocytes [17].Additionally, the expression of CB 2 by different tissues is upregulated in an inflammatory context including neuroinflammation [34][35][36] and acute and chronic inflammation [37,38].Considering the relevance of the CB 2 receptor in the regulation of the immune response, CB 2 is seen as a potential therapeutic target in inflammatory diseases.

CB 2 Signaling Pathway
CB 1 and CB 2 are both G-protein-coupled receptors binding to a trimer of Gα/β/γ proteins and activate in part similar signaling cascades.CB 1 can couple to either Gαs or Gαi/o proteins which respectively activate or inhibit the adenylyl cyclases.Conversely, CB 2 almost exclusively recruits Gαi/o proteins leading to inhibition of adenylyl cyclase activity which reduces AMPc levels and prevents activation of PKA [39].Since PKA activation leads to NF-κB and CREB transactivation [40], Gi/o signaling triggered by CB 2 is believed to prevent the induction of inflammatory genes.Although CB 2 coupling to Gαs has been reported only in rare instances, it was suggested to play an important role in the induction of a pro-inflammatory response dependent on PKA activation [41].CB 2 engagement also leads to the Gβ-and Gγ-dependent activation of p38 and ERK1/2 MAP kinases [41,42].CB 2 -dependent MAPK phosphorylation was shown to be responsible for the activation of pro-inflammatory genes through transcription factors NF-κB and CREB [41,43].Moreover, contrarily to CB 1 , CB 2 cannot trigger potassium channel signaling through Gαs but can initiate Ca 2+ signaling via phospholipase C activation by Gαi/o proteins [44].Finally, the c-terminal tail of CB 1 and CB 2 can also be phosphorylated by G-Protein-Coupled Receptor Kinases (GRK), allowing the recruitment of β-arrestins.CB 2 can bind β-arrestin1 and β-arrestin2 which leads to receptor internalization but also to the phosphorylation of MAPK ERK1/2 [45,46] (Figure 1).MAPK ERK1/2 [45,46] (Figure 1).
A single GPCR can adopt different conformations depending on its interaction with specific agonists, and some conformations can trigger only part of the signaling events attributed to the receptor [47].Evidence of this mechanism, termed biased agonism, has been reported for both CB1 and CB2.In the case of CB2, THC strongly potentiates ERK phosphorylation but is less effective on β-arrestin recruitment and cAMP signaling and does not activate G-protein-coupled inwardly rectifying potassium channels (GIRK).En docannabinoids can activate all signaling pathways, but N-arachidonoylethanolamine (AEA) strongly increases GIRK and ERK activation and is less impactful on cAMP and β arrestin signaling, while 2-AG favors GIRK signaling to other pathways.As for selective CB2 agonists, JWH-133 induces β-arrestin and cAMP signaling more potently than GIRK and ERK activation, while HU910 and HU308 do not show a strong bias toward a specific signaling pathway [43,48].Given that CB2 agonists vary greatly in their selectivity, efficacy and capacity to trigger signaling events, studies using different agonists must be com pared carefully, and agonists that are both selective and not greatly biased toward signal ing pathways should be favored in future studies.A single GPCR can adopt different conformations depending on its interaction with specific agonists, and some conformations can trigger only part of the signaling events attributed to the receptor [47].Evidence of this mechanism, termed biased agonism, has been reported for both CB 1 and CB 2 .In the case of CB 2 , THC strongly potentiates ERK phosphorylation but is less effective on β-arrestin recruitment and cAMP signaling and does not activate G-protein-coupled inwardly rectifying potassium channels (GIRK).Endocannabinoids can activate all signaling pathways, but N-arachidonoylethanolamine (AEA) strongly increases GIRK and ERK activation and is less impactful on cAMP and β-arrestin signaling, while 2-AG favors GIRK signaling to other pathways.As for selective CB 2 agonists, JWH-133 induces β-arrestin and cAMP signaling more potently than GIRK and ERK activation, while HU910 and HU308 do not show a strong bias toward a specific signaling pathway [43,48].Given that CB 2 agonists vary greatly in their selectivity, efficacy and capacity to trigger signaling events, studies using different agonists must be compared carefully, and agonists that are both selective and not greatly biased toward signaling pathways should be favored in future studies.

Cnr2-KO Mice and Inflammation
Cnr2 c deficient animals provide a powerful tool to understand the roles of CB 2 in inflammatory mouse models [49].A majority of studies showed that Cnr2-deficient mice have a pro-inflammatory profile (Table 1), implying that CB 2 is involved in the negative regulation of the inflammatory response [50][51][52][53][54][55].In inflammatory models such as lipopolysaccharide (LPS)-induced inflammation [54,56] or alcoholic liver disease [50], the genetic deletion of Cnr2 was associated with an increased production of pro-inflammatory cytokines.The secretion of chemokines and the expression of their receptors also increased leading to a modification of cell migration and infiltration.In acute inflammatory conditions, such as the murine dorsal air pouch model, the absence of Cnr2 altered the recruitment of neutrophils to the site of inflammation [54].
On the other hand, some studies have shown that Cnr2 deficiency can also dampen inflammation, notably, inflammatory states involving eosinophils, as we summarize in Table 2.For instance, mouse models of allergic bronchitis are characterized by the infiltration of activated eosinophils in the airways [57].This could be explained, in part, by the ability of 2-AG, concomitantly with IL-5, to act as a chemoattractant for eosinophils [58].Eosinophil accumulation often correlates positively with disease severity and immune cell infiltration and is effectively reversed by genetic deletion of Cnr2 [59].The absence of Cnr2 alters innate immune cell function as it reduces the ability of dendritic cells to express costimulatory molecules that affect T cell activation and proliferation [53].Furthermore, CB 2 signaling affects macrophage functions by steering them toward those typical of an M2 phenotype [60].
These contrasting studies demonstrate that the inflammatory effect of Cnr2 deletion is not completely understood.It is also crucial to take into consideration other receptors and endocannabinoid-like compounds belonging to the endocannabinoidome (see below), which can compensate for the absence of CB 2 (Tables 1 and 2) [61][62][63].Among these receptors, we highlight PPARγ (peroxisome proliferator-activator receptor gamma), which belongs to the endocannabinoidome [64].Through its affinity for endocannabinoids, PPARγ can downregulate the activation, proliferation and migration of helper T cells [65][66][67][68].Considering that the endocannabinoid system is involved in key biological processes, effects seen in Cnr2 −/− inflammation mice models cannot be strictly attributed to a direct implication of CB 2 in inflammation.Endocannabinoids can also act through other receptors, namely transient receptor potential vanilloid type 1(TRPV1) [69] and PPARs [70], which can counteract the absence of CB 2 and explain the variable outcomes in response to the genetic deletion of Cnr2 [64].The arrows (↑/↓) indicate whether a process is enhanced or reduced, respectively.

Endogenous Ligands
Endogenous molecules with similar functions as ∆ 9 -THC were discovered and named "endocannabinoids" [84].To date, there are two characterized endocannabinoids binding to the cannabinoid receptors [7,13,85,86].N-arachidonoylethanolamine (AEA), also known as anandamide, was first isolated from the porcine brain [7].The second molecule, 2-arachidonoylglycerol (2-AG), was next found in canine intestinal tissue [86].2-AG has a higher efficacy at CB 2 than AEA [64,85].Moreover, since AEA is a partial agonist at CB 2 as compared to 2-AG, when both ligands are present in a competitive way, AEA may act as a competitive antagonist of CB 2 in the presence of 2-AG [87].
Other candidate endocannabinoids derived from arachidonic acid have also been identified.The ester form of arachidonic acid coupled to glycerol, also known as Noladin or 2-AGE, was isolated from the porcine brain in 2001 by Hanus et al. [88].Based on the selective affinity of Noladin for the CB 1 receptor [89], it took longer to identify that Noladin can also act as a full agonist at the CB 2 receptor [90,91].However, in areas such as the brain, where endocannabinoids and their receptors are known to be abundant, the concentration of 2-AGE was similar to that of AEA.[92].In addition, Oka and colleagues showed that in mammals, the ratio of 2-AGE to 2-AG was very low, ranging from 1/500 to 1/4000 [93].Although 2-AGE can act via the two endocannabinoid receptors, the low presence of Noladin both centrally and peripherally could raise the question of the physiological relevance of Noladin.The endocannabinoid family also includes O-arachidonoyl-ethanolamine, also known as Virodhamine, which is the ester of arachidonic acid and ethanolamine [94].Discovered in 2002 by Porter et al., Virodhamine is known to be a CB 2 agonist [94].In peripheral tissues where CB 2 is expressed, such as the spleen, Virodhamine levels were eight times higher than AEA.Although Virodhamine is known to be a CB 2 agonist, it has been shown that it is also an agonist of the GPR55 receptor and that it exerts its effects through this receptor [95].The third component results from the conjugation of arachidonic acid with dopamine, N-arachidonoyl-dopanime, also known as NADA [96], and has been reported to be present in specific regions of the brain including the striatum and hippocampus [96,97].NADA is an endogenous ligand for CB 1 but also for transient receptor potential vanilloid type 1 (TRPV1) and exerts its role in the inflammatory response through this receptor [98].Given that these endocannabinoid ligands activate other receptors of the endocannabinoid system and the endocannabinoidome, such as CB 1 , GPR55 and TRPV1 [96], in this review, we will focus on endogenous ligands that exert an effect through the CB 2 receptor.
Both AEA and 2-AG are lipid mediators derived from the cleavage of membrane phospholipid precursors [99].However, their biosynthesis depends on different enzymatic pathways.Amongst the four known pathways leading to the production of AEA, the most studied is the hydrolysis of N-arachidonoyl-phosphatidylethanolamine (NAPE) by the N-acyl-phosphatidylethanolamine phospholipase D (NAPE-PLD) [100].AEA is primarily degraded by fatty acid amide hydrolase 1 (FAAH 1 ), a serine hydrolase active at alkaline pH [101].N-acylethanolamine-selective acid amidase (NAAA), which is a cysteine hydrolase working in an acidic environment, is another important catabolic enzyme degrading AEA, although with lower affinity/efficacy than with other N-acyl-ethanolamines, such as N-palmitoyl-ethanolamine.Indeed, with varying degrees of selectivity, these enzymes can hydrolyze N-acylethanolamines to fatty acids and ethanolamine [102].
In an attempt to manipulate the bioavailability of endocannabinoids and enhance their benefits on the immune system response, several inhibitors of FAAH and MAGL were developed [101,106].Namely, selective inhibition of FAAH with URB597 is associated with increased levels of N-acylethanolamines such as AEA and reduced production of pro-inflammatory cytokines and immune cell infiltration in the ovalbumin-induced allergic asthma model [107].A study by Genovese et al., confirmed that inhibition of FAAH activity by URB878 significantly reduced the inflammation associated with acute lung injury [108].The genetic deletion of Mgll, which encodes the MAGL, is linked to increased levels of monoacylglycerols [109] and protects mice from diet-induced obesity and associated inflammatory diseases [110].Furthermore, the absence of Mgll helps to reduce inflammation and liver damage when mice are exposed to carbon tetrachloride [111].In addition, using JZL184 to inhibit the activity of MAGL produces antiarthritic effects and reduces the paw inflammation associated with collagen-induced arthritis [112].These results suggest that endocannabinoids modulate inflammation, although these effects cannot be necessarily attributed to CB 2 activation.
The activation of the CB 2 receptor by endogenous ligands was specifically examined in inflammatory models as detailed in Table 3.Briefly, treatments that target CB 2 either with ∆ 9 -THC [113] or the two endocannabinoids AEA and 2-AG [6,114] revealed a dichotomy regarding inflammatory responses.First, most studies highlighted a reduction in immune cell infiltration and secretion of inflammatory mediators associated with CB 2 activation [115][116][117][118].More importantly, these treatments are accompanied by an improvement in symptoms associated with inflammation.For example, Madig et al. observed reduced edema and improved blood-brain barrier function and neurological recovery after treatment with 2-AG in order to specifically target the CB 2 receptor in Theiler's murine encephalomyelitis virus-induced demyelination [79].Conversely, other authors have shown that a pro-inflammatory response is associated with cannabinoid treatment as well as selective CB 2 receptor agonists.As described by Oka et al. in mice treated with oxazolone to induce dermatitis, the number of infiltrating eosinophils is reduced when CB 2 is blocked with SR144528 [119].However, the pro-inflammatory effects of endocannabinoids in some cases can be attributed to the products of endocannabinoid metabolism [114,120].Eicosanoid metabolites derived from arachidonic acid oxidation contribute to inflammation (see Dennis et al. for details [121]).In addition, arachidonic acid and its metabolites modulate type 2 immune responses, which are important in the allergic response through actions on eosinophils and neutrophils [122].
Though the large body of evidence supporting endocannabinoids as anti-inflammatory mediators is cohesive and well established, these contradictory results reaffirm the need to investigate further the mechanisms by which endocannabinoids can exert pro-inflammatory effects.
Pro-inflammatory effects ↓ Cancer cells proliferation ↓ Dendritic cell activation [130] The arrows (↑/↓) indicate whether a process is enhanced or reduced, respectively.

Synthetic Ligands
Studies on marijuana users prompted the potential therapeutic importance of exogenous, synthetic CB 2 receptor agonists even before the identification of endocannabinoid ligands and receptors and inhibitors of endocannabinoid-inactivating enzymes.
In a multivariate study by Tindall et al., conducted on human immunodeficiency virus (HIV)-positive patients who had not yet developed acquired immunodeficiency syndrome, marijuana users were more likely to have a lower percentage of CD4+ cells and a higher percentage of CD8+ cells [131].In addition, HIV carriers who use marijuana exhibit an increased risk of developing bacterial pneumonia and other opportunistic infections [132].Indeed, it was demonstrated that alveolar macrophages in the lungs of marijuana smokers were less effective in their ability to clear bacteria and tumor cells [133].These studies suggest that natural cannabinoids capable of activating CB 2 receptors affect the immune system, and this effect could be exploited as a treatment.
Synthetic compounds such as CP55,940 and WIN55,212-2 were available when CB 2 was cloned, although these synthetic cannabinoids can also activate CB 1 with similar efficiency [11,134].A highly selective antagonist for CB 2 receptors, SR144528, was the first molecule designed in order to investigate CB 2 -mediated endocannabinoid functions in the immune system [135].Since then, researchers have developed a wide variety of CB 2 -specific antagonists or agonists (Table 4) [136].
In 2017, Soethoudt et al. tested 18 pharmacological agonists and antagonists used in preclinical models and concluded that JWH-133, HU308 and HU910 are the agonists with the best affinity for CB 2 and least psycho-chemical consequence [48].JWH-133 (Ki = 3.4 nM) was characterized in 1999 and found to be a potent CB 2 agonist, 200 times more selective for CB 2 than for CB 1 [13,137].HU308 (Ki = 22.7 nM) is a specific agonist for CB 2 and does not appreciably bind to CB 1.Of these three specific agonists, HU910 (Ki = 6 nM) is one of the most recently developed as it was first used in 2012 and has specific affinity for CB 2 .Other CB 2 selective agonists were synthesized, such as GW405833 [138], which has antiinflammatory properties [139][140][141].However, GW405833 may also act as a non-competitive antagonist for CB 1 as Li et al. noted in their study that the anti-allodynic effect of this compound was mediated through CB 1 [142].
Further in vitro and in vivo studies revealed that targeting the CB 2 receptor has immunomodulatory effects in several ways: the induction of apoptosis and antiinflammatory cytokine production, as well as, the repression of cell proliferation, immune cell migration and pro-inflammatory cytokine production [6].We list in Table 5 recent studies targeting CB 2 in animal models of inflammation.Interestingly, in in vitro and in vivo studies in mouse models of ovalbumin-induced acute asthma, JWH-133 enhanced the mobilization of eosinophils [19].Furthermore, in cecal-ligation-and puncture-induced polymicrobial sepsis in rats, JWH-133 induced an anti-inflammatory response by inhibiting apoptosis and NF-κB signaling in the brain, lung, liver and heart [143].When administered to mice, HU308 reduced blood pressure, slowed defecation and caused anti-inflammatory effects [5].HU910 was found to reduce the markers associated with liver injury in the ischemia/reperfusion injury mouse model and to attenuate the levels of pro-inflammatory cytokines and chemokines as well as immune cell infiltration [117].
Furthermore, in the context of mouse models of rheumatoid arthritis, JWH-133 induced an anti-inflammatory response by decreasing the production of proinflammatory cytokines and increasing the polarization of macrophages toward an M2 phenotype [144,145].In addition, Fukuda et al. showed that, in collagen type-II antibody-induced arthritis, JWH-133 reduced arthritis scores and limited bone destruction [146].Moreover, in the monoiodoacetic acid-induced osteoarthritis model, systemic and chronic administration of JWH-133 was associated with a reduction in pain-related behaviors [147].In addition, experiments carried out in CB 2 -overexpressing mice suffering from monosodium iodoacetate-induced joint pain showed a better control of pain manifestation [148].In the case of rheumatic diseases, CB 2 agonists can improve inflammation and reduce pain concomitantly.
Taken together, these findings confirm that targeting CB 2 with pharmacological agents can often improve inflammation in a wide range of inflammatory diseases, by modulating, directly or indirectly, the responses of various immune cells such as eosinophils [143], macrophages [149], neutrophils [54,56] or lymphocytes [15,150].↓ Blood pressure ↓ Pro-inflammatory cytokines [162] The arrows (↑/↓/↔) indicate whether a process is enhanced, reduced or unchanged respectively.

The Effect of CB 2 Activation in Airway Inflammation
The recruitment of circulating immune cells is an amplification factor for pathologies with an inflammatory component [197].This is the case in chronic allergic airway inflammation where the infiltration of the airways by eosinophils, mast cells and T helper lymphocytes is an important step of the pathogenesis [198].Activation of both recruited and airway resident immune cells, namely innate lymphoid type 2 cells (ILC2), leads to the production of pro-inflammatory Th2 cytokines such as IL-4, IL-5 and IL-13 [198][199][200].IL-4 production by Th2 lymphocytes also leads to the isotypic switch of B cells from IgM to IgE antibodies in the lungs [198].These responses enhance mucus secretion by goblet cells, airway hyperresponsiveness and bronchoconstriction [201,202].Since circulating immune cells express high levels of CB 2 [17], their recruitment during an inflammatory pathology causes an influx of such receptors that have the potential to influence disease development.
Studies on healthy volunteers or asthmatic patients treated with ∆ 9 -THC revealed a beneficial increase in bronchodilatation and specific airway conductance in treated individuals [203,204].In the context of an ovalbumin-induced allergic reaction in mice, treatment with ∆ 9 -THC attenuated the Th2 allergic response.Notably, it decreased the gene expression of IL-2, IL-4, IL-5 and IL-13, reduced serum IgE levels and diminished mucus secretion [205].
In order to understand the specific implication of CB 1 and CB 2 in the beneficial effects elicited by ∆ 9 -THC, Newton and Klein studied the effect of this plant cannabinoid in a Th1-driven inflammatory reaction, i.e., that following Legionella pneumophila infection.They found that ∆ 9 -THC decreased the serum levels of the Th1 cytokines, IL-12 and IFNγ by acting specifically through CB 1 .∆ 9 -THC also induced an increase in GATA3 signaling and IL-4 production by activating CB 2 exclusively [123].Furthermore, these authors demonstrated that, in ovalbumin-sensitized mice, CB 2 activation by a selective agonist reduced IgE levels [206].These results emphasize how, although ∆ 9 -THC can tune down the inflammatory reaction of both Th1-and Th2-driven pathologies, this is the sum of independent effects mediated differently through CB 1 and CB 2 activation.Some studies investigated the role of AEA and 2-AG in asthma.Zoerner et al. found that the bronchoalveolar fluids of allergic asthma patients challenged with an allergen contained much greater levels of AEA than patients given saline.Moreover, patients with higher AEA levels also had more severe symptoms of airway inflammation [207].A study by Larose et al. showed that IL-3, IL-5 and GM-CSF greatly potentiate the CB 2dependant chemoattractant effect of 2-AG on eosinophils isolated from healthy donors [58].In ovalbumin-sensitized guinea pigs, the inhibition of FAAH with URB597 reduced Th2 cytokine production and infiltration of immune cells but had marginal effects on airway hyperreactivity.Conversely, the inhibition of MAGL with JZL184 attenuated cytokine production, immune infiltration and airway hyperreactivity [107].
In a mouse model of ovalbumin-induced asthma, Frei et al. showed that the CB 2 agonist JWH-133 increased the chemotaxis, activation and reactive oxygen species production of eosinophils.These effects were reduced in both Cnr2 −/− and eosinophil-deficient mice confirming that CB 2 activation on eosinophils is a key event in allergic airway inflammation [19].Furthermore, Hurrel et al. showed that CB 2 signaling induced the proliferation and activation of ILC2 isolated from mouse lungs [208].They confirmed, using Cnr2 −/− mice, that these animals presented reduced lung inflammation and airway hyperreactivity, which was a consequence of reduced ILC2 activation [208].Ferrini et al. demonstrated that the absence of Cnr2 in a model of common house dust mite sensitization increased NK cell infiltration and activation that prompted the secretion of Th1 cytokines in the airways.This activation of NK cells led to a drastically reduced number of ILC2 cells, which was accompanied by a reduction in Th2 cytokine levels [59].
The studies cited above, in addition to those we summarize in Table 6, suggest that the putative beneficial effect of CB 2 activation is highly context-dependent.Nevertheless, we can emphasize that in a Th2 inflammatory context, including allergy, antagonism of CB 2 is strongly beneficial since it prevents the inflammatory response possibly by switching it toward a Th1-type response.

The Effect of Cannabinoid-Based Treatments in Inflammatory Diseases
Tissue alterations in endocannabinoid concentrations and the expression of endocannabinoid receptors and metabolic enzymes have been associated with several inflammatory conditions, such as neuroinflammation and chronic inflammatory diseases (Figure 2).Based on accumulated findings, several clinical studies on cannabinoid therapies were instigated that we bring to light in Table 7.   Mentioned for the first time in 1995 [233], neuroinflammation is characterized by an inflammatory response within the brain or spinal cord and reported as an indicator and modulator of neurodegeneration [234].Several studies have highlighted the importance of the CB 2 receptor in diseases associated with neuroinflammation [235,236].The CB 2 receptor is expressed by several cells in the brain such as astrocytes, reactive microglia, perivascular microglia, oligodendrocytes, neuronal progenitor cells and cells involved in blood-brain barrier integrity [237].In 2003, Benito et al. found that the CNR2 gene was highly expressed in post mortem brain microglia from subjects with Alzheimer's disease, whereas the expression of the CNR1 gene remained unchanged [238].In vitro studies demonstrated that CB 2 activation by JWH-015 suppressed the production of IFN-γ, TNF-α and nitrous oxide by microglia after stimulation with the amyloid β peptide [239].In addition, JWH-133 blocked the activation of microglia by amyloid β peptide, and this treatment with the CB 2 agonist reduced the production of pro-inflammatory cytokines [240].Furthermore, when neurons and microglia were co-cultured, treatment with JWH-133 prevented the microglia-mediated toxicity of neurons after amyloid β peptide exposure [241].These in vitro results suggest that the beneficial effects of CB 2 activation are associated with the presence of this cannabinoid receptor on microglia and are related to the suppression of their pro-inflammatory activation.
Interestingly, the levels of AEA were found to be lower in the frontal and temporal cortex tissues of post mortem patients with Alzheimer's disease compared to control subjects [242].Furthermore, higher 2-AG plasma levels correlated with better memory and attention in patients with Alzheimer's disease.This suggests a protective mechanism associated with the modulation of the endocannabinoid system [243].
Since it is possible to modulate endocannabinoid levels by inhibiting their hydrolysis, JZL184, which is an irreversible inhibitor of MAGL, was used in an Alzheimer's disease mouse model.Pihlaja et al. observed a decrease in the production of pro-inflammatory mediators by microglia and astrocytes isolated from adult mice treated with JZL184 [244].The inhibition of MAGL has been associated with several positive effects on Alzheimer's disease in animal models, such as amelioration of inflammation and neuronal disorders [245].The inhibition of FAAH activity with a selective irreversible inhibitor, UBR597, increased the availability of AEA [246].Chiurchiu et al. demonstrated that macrophages from patients with Alzheimer's disease treated with URB597 produced less pro-inflammatory cytokines [247].Moreover, the treatment of aged rats with URB597 reduced the expression of IL-1β and TNF-α and restored aged-related disorders [248].
In a transgenic Tg APP 2576 mouse model of Alzheimer's disease that overexpresses the amyloid precursor protein, the authors observed that chronic treatment with JWH-133 was able to decrease microglia activation and reduce COX-2 activation and TNF-α production.More importantly, CB 2 activation helped to reduce cognitive impairment [241].In the APP/PSI mouse model expressing a chimeric human/mouse amyloid precursor protein directly in neurons, the administration of JWH-133 was found to be effective at reducing Tau-hyperphosphorylation [249].
These results support the idea that targeting CB 2 and endocannabinoid metabolism is a therapeutic option for some neuroinflammatory diseases.

Potential in Chronic Inflammation: Inflammatory Bowel Diseases (IBDs)
Crohn's disease and ulcerative colitis are the two major chronic idiopathic IBDs [250].Though several clinical and pathological features differ between these two IBDs [251], both diseases are characterized by intestinal inflammation and alteration of the epithelial barrier [252].This leads to the translocation of bacteria and microbial products from the gut lumen through the intestinal wall.Consequently, an acute inflammatory reaction occurs, which is driven by immune cell infiltration and cytokine production [253].As the disease progresses, an increasingly uncontrolled chronic inflammation develops that leads to tissue destruction [254,255].
Two noteworthy independent studies on IBD patients compared users versus nonusers of cannabis.Storr et al. showed that patients who used cannabis for more than six months were more susceptible to undergoing surgical treatment associated with Crohn's disease [256].On the other hand, Mbachi et al. compared two groups of patients with Crohn's disease, one group of non-cannabis users and a second group of cannabis users.They concluded that cannabis users developed fewer complications, such as fistulizing disease, colectomy or intra-abdominal abscess [257].Furthermore, mucosal tissue from the inflamed region of the colon of IBD patients incubated with the CB 2 agonist JWH-133 for 6 h showed increased epithelial cell proliferation accompanied by decreased MMP-9 and IL-8 secretion [258].This suggests that, in the context of IBD, the protective effect induced by plant cannabinoids might be mediated by the CB 2 receptor.
Even though CB 2 is highly expressed by infiltrating macrophages and plasma cells [23], the receptor is also detected in the esophagus, stomach, ileum and intestine [259].It is expressed by epithelial cells, goblet cells and Paneth cells and increased during the acute phase of IBDs [260].In addition, the Q63R functional variant of the CB 2 protein has recently been significantly associated with Crohn's disease and ulcerative colitis [261].
It is demonstrated that plasma AEA levels are higher in patients with IBD [262].Regarding 2-AG, its levels are higher in patients with Crohn's disease and are associated with increased expression of the 2-AG synthesizing enzyme DAGL-α [263].
Several IBD mouse models such as 2,4,6-Trinitrobenzenesulfonic acid-or Dinitrobenzene sulfonic acid-treated mice are used to mimic colitis [264].First, in Tnfα overexpressing mice treated with TNBS/DSS, it was demonstrated that expression of Cnr2 was increased in immune cells.Furthermore, GP-1a treatment induced the polarization of T lymphocytes into regulatory phenotype secreting IL-10 [184].In the DNBS-induced IBD model, JTE907, a CB 2 -specific inverse agonist, induced phenotypic differentiation of inflammatory T cells into foxp3 positive regulatory T cells secreting TGF-β and IL-10.In these studies, a CB 2 receptor agonist reduced the severity of the disease [265].
These results suggest that targeting CB 2 and manipulating pharmacologically the availability of endocannabinoids are potential therapeutic avenues to improve the quality of life of patients with IBD in terms of pain and disease symptoms but also to improve inflammation associated with these disorders.

Potential in Metabolic Disease: Non-Alcoholic Fatty Liver Disease (NAFLD)
NAFLD encompasses a number of liver diseases ranging from isolated hepatic steatosis to steatohepatitis and irreversible cirrhosis [270].NAFLD reflects the inability of the adipose tissue to perform its function as fat-storage tissue, leading to increased triglyceride uptake by hepatocytes [271].The prevalence of NAFLD increases dramatically with obesity [272], dyslipidemia and type 2 diabetes [273] and is becoming the most common liver disease in developed countries [274].Steatosis without any signs of fibrosis is considered an early condition, whereas the presence of fibrosis predicts chronic progression to severe liver disease [275].
A study with hepatitis-C-positive patients revealed that cannabis users had decreased prevalence of liver cirrhosis, although this did not improve mortality [276].Furthermore, a study comparing the evolution of liver diseases in obese patients suggested that cannabis reduced the prevalence and progression of steatohepatitis [277].
It was established that Cnr2 mRNA is expressed in the liver of morbidly obese women at different stages of NAFLD.Its expression was correlated with the expression of antiinflammatory and pro-inflammatory mediators, indicating that CB 2 may have a dual role in NAFLD and NAFLD-related complications [278].Indeed, because of the high expression of Cnr2 in the damaged liver, it was proposed that this phenomenon could predict the progression of liver disease from chronic hepatitis to irreversible cirrhosis and hepatocellular carcinoma [279].Cnr2 was expressed by hepatocytes but only in patients with non-alcoholic steatosis [280].In addition, Cnr2 was also expressed by Kupffer cells which are key players in immune control in the liver.Kupffer cells recognize many pathogens through various pattern recognition receptors and respond by producing pro-inflammatory cytokines [281].Furthermore, the depletion of Kupffer cells in rats fed with a high-fat diet to induce steatosis protects against hepatic steatosis and insulin resistance, highlighting the importance of these cell types in liver disease [282].These cells are very plastic in response to their environment as they can switch from a pro-inflammatory M 1 to an anti-inflammatory M 2 phenotype [283].Kupffer cells isolated from Cnr2 knock-out mice are more polarized toward an M1 phenotype compared to Kupffer cells form wild-type mice [284].The absence of Cnr2, or treatment with the CB 2 antagonist, AM630, also confers to mice protection against steatosis induced with a high-fat diet [55].On the other hand, rats treated with the agonist AM1241 displayed significant expression of hepatic progenitor cell markers, which indicates that stimulating CB 2 enhances hepatocyte regeneration [181].Furthermore, chronic activation of CB 2 with JWH-133 in cirrhotic rats decreased the arterial pressure, immune cell infiltration and the number of activated stellate cells but more importantly, decreased fibrosis [285].
Taken together, these findings from various human studies and mouse models reveal contradictions in the effects of AEA and 2-AG in liver injury and cast doubt on the value of CB 2 as a therapeutic target for liver diseases.Further studies are warranted to confirm whether (1) CB 2 has a beneficial role in NAFLD and (2) the anti-inflammatory effect of CB 2 is enough to counterbalance its proposed effect of increasing lipid accumulation.It is perhaps more interesting to explore the possibility that endocannabinoids and CB 2 expression are biomarkers of the evolution of liver disease.

Conclusions and Future Direction
In recent years, we have witnessed an upsurge in research aimed at better understanding the role of CB 2 in different inflammatory contexts.Many animal models demonstrate that CB 2 stimulation by endogenous and exogenous ligands leads to an anti-inflammatory response and improves the symptoms of inflammatory diseases.However, it is important to consider the inflammatory context in which the CB 2 receptor is targeted.This is demonstrated in allergic airway inflammation, a Th2-driven response, for which CB 2 activation is detrimental.
Specific agonists and antagonists were developed and used in several clinical studies to target the endocannabinoid system in inflammatory conditions (Table 6).Although only a fraction of these studies have reached successfully their primary outcome, it establishes CB 2 as a promising therapeutic target.Therefore, future efforts should focus on developing CB 2 ligands activating specific signaling pathways and establishing which ligands are effective in the inflammatory context of each disease.

Figure 1 .
Figure 1.Signaling of the CB2 receptor leading to pro-and anti-inflammatory responses.The arrows (↑/↓) indicate an increase or decrease in molecule concentration, respectively.

Figure 1 .
Figure 1.Signaling of the CB 2 receptor leading to pro-and anti-inflammatory responses.The arrows (↑/↓) indicate an increase or decrease in molecule concentration, respectively.

Figure 2 .
Figure 2. Effects of CB2 activation on inflammation in different mouse tissues and pathological condition.The endocannabinoid system via the CB2 receptor has been implicated in inflammatory processes.Based on previous studies in animal models, the arrows (↑/↓) indicate whether a process associated with each pathology is enhanced or reduced, respectively.

Figure 2 .
Figure 2. Effects of CB 2 activation on inflammation in different mouse tissues and pathological condition.The endocannabinoid system via the CB 2 receptor has been implicated in inflammatory processes.Based on previous studies in animal models, the arrows (↑/↓) indicate whether a process associated with each pathology is enhanced or reduced, respectively.6.1.Potential in Neuroinflammation: Alzheimer's Disease

Table 1 .
Pro-inflammatory effects of Cnr2 genetic deletion in mouse models of inflammation.

Table 2 .
Anti-inflammatory effects of Cnr2 genetic deletion in mouse models of inflammation.

Table 3 .
Effects of endocannabinoid treatment in rodent models of inflammation.

Table 5 .
Anti-inflammatory effects of CB 2 activation by agonists in rodent models of inflammation.

Table 6 .
CB 2 modulates the inflammatory response in vivo and in vitro.

Table 7 .
List of clinical trials based on endocannabinoids targeting inflammation.