Cannabinoid receptor 2 augments eosinophil responsiveness and aggravates allergen‐induced pulmonary inflammation in mice

Abstract Background Accumulation of activated eosinophils in tissue is a hallmark of allergic inflammation. The endocannabinoid 2‐arachidonoylglycerol (2‐AG) has been proposed to elicit eosinophil migration in a CB 2 receptor/Gi/o‐dependent manner. However, it has been claimed recently that this process may also involve other mechanisms such as cytokine priming and the metabolism of 2‐AG into eicosanoids. Here, we explored the direct contribution of specific CB 2 receptor activation to human and mouse eosinophil effector function in vitro and in vivo. Methods In vitro studies including CB 2 expression, adhesion and migratory responsiveness, respiratory burst, degranulation, and calcium mobilization were conducted in human peripheral blood eosinophils and mouse bone marrow‐derived eosinophils. Allergic airway inflammation was assessed in mouse models of acute OVA‐induced asthma and directed eosinophil migration. Results CB 2 expression was significantly higher in eosinophils from symptomatic allergic donors. The selective CB 2 receptor agonist JWH‐133 induced a moderate migratory response in eosinophils. However, short‐term exposure to JWH‐133 potently enhanced chemoattractant‐induced eosinophil shape change, chemotaxis, CD11b surface expression, and adhesion as well as production of reactive oxygen species. Receptor specificity of the observed effects was confirmed in eosinophils from CB 2 knockout mice and by using the selective CB 2 antagonist SR144528. Of note, systemic application of JWH‐133 clearly primed eosinophil‐directed migration in vivo and aggravated both AHR and eosinophil influx into the airways in a CB 2‐specific manner. This effect was completely absent in eosinophil‐deficient ∆dblGATA mice. Conclusion Our data indicate that CB 2 may directly contribute to the pathogenesis of eosinophil‐driven diseases. Moreover, we provide new insights into the molecular mechanisms underlying the CB 2‐mediated priming of eosinophils. Hence, antagonism of CB 2 receptors may represent a novel pharmacological approach for the treatment of allergic inflammation and other eosinophilic disorders.

Hence, antagonism of CB 2 receptors may represent a novel pharmacological approach for the treatment of allergic inflammation and other eosinophilic disorders.
Regarding asthma, levels of eosinophil granule proteins such as major basic protein (MBP) or eosinophil peroxidase (EPO) broadly correlate with disease severity (8). Moreover, it has been shown that patients who receive treatment based on eosinophil counts in sputum have significantly fewer severe asthma exacerbations than patients treated according to standard management therapy (9).
Corticosteroids are currently the most effective treatment to reduce eosinophil numbers in the blood and tissue. However, the pleiotropic effects, especially of orally applied corticosteroids to control severe asthma, can result in potentially harmful sideeffects and thus limit their therapeutic use (10). Therefore, considerable effort has been invested in the development of drugs that can effectively control the trafficking and activation of eosinophils to ameliorate the inflammatory response. Current clinical trials with the eosinophil-targeted mepolizumab revealed an oral glucocorticoid-sparing effect and significantly reduced exacerbation rates in well-selected patients with persistent eosinophilic asthma; however, FEV 1 seemed to remain largely unaffected (11,12). Of note, asthma is a heterogeneous condition with diverse characteristics and disease variants. Thus, stratification of patients by clinical characteristics (phenotypes) and pathogenetic mechanisms (endotypes) should lead to more targeted and personalized approaches to asthma therapy. Endocannabinoids are bioactive lipids released from the cell membrane upon cell activation. The two main endocannabinoids are the arachidonic acid (AA) derivatives 2-arachidonoylglycerol (2-AG) (13) and N-arachidonoylethanolamine (anandamide (AEA)) (14). Most of their actions are mediated by two G-protein-coupled receptors, the cannabinoid receptors 1 and 2 (CB 1 and CB 2 ). In contrast to CB 1 , which is abundantly expressed in the nervous system (15), CB 2 is mainly found in lymphoid organs and cells of the immune system (16). Among others, CB 2 expression has been reported for B cells, monocytes/macrophages, and eosinophils, indicating a crucial immunoregulatory role for CB 2 and its ligands (17).
Up to now, only three studies have examined the direct contribution of the endocannabinoid 2-AG on eosinophil migration in vitro. Oka et al. demonstrated that 2-AG induces the chemotaxis of EoL-1 cells and human peripheral blood eosinophils in a CB 2 -dependent manner, although the order of the pharmacologically effective concentration of 2-AG was significantly higher than that of other chemoattractants (18). Interestingly, another group confirmed this minimal CB 2 -mediated effect of 2-AG and suggested that 2-AG-induced migration is a complex process that may involve other mechanisms, such as cytokine priming, rapid 2-AG metabolism into eicosanoids, and further generation of 15-lipoxygenase metabolites (19).
Moreover, due to the use of different rodent models and varying pharmacological approaches, the in vivo role of CB 2 in eosinophilic disorders is still uncertain. Indeed, a CB 2 inverse agonist was shown to block ovalbumin (OVA)-induced lung eosinophilia in mice (20). Accordingly, the CB 2 antagonist SR144528 mediated beneficial effects in oxazolone-induced contact dermatitis (21). In contrast, Giannini et al. evaluated the effects of the CB 1 /CB 2 receptor agonist CP55,940 on OVAinduced asthma in guinea pigs and concluded that both CB 1 and CB 2 receptors are involved in lung protection (22). Similarly, THC reduced cytokine and IgE level as well as mucus production in OVA-challenged mice (23); however, the use of CB 1 /CB 2 knockout mice revealed that the effects of THC were cannabinoid receptor-independent (24).
Because of the conflicting reports on the pathophysiological role of CB 2 receptors, we set out to explore the direct contribution of specific CB 2 activation on eosinophil effector function. To this end, the effects of the endocannabinoid 2-AG and the potent and selective CB 2 receptor agonist JWH-133 were studied in vitro and in mouse models of acute OVA-induced asthma and directed eosinophil migration. Our results provide clear evidence for a novel CB 2 -induced eosinophil-specific 'priming mechanism' that potentiates eosinophil effector function in vitro and enhances bronchial inflammation in vivo, reflected by both increased airway resistance and eosinophil influx into the airways.

Materials and methods
Detailed materials and methods are provided in the supplement.

Preparation of mouse eosinophils
Bone marrow-derived eosinophils (bmEos) were differentiated ex vivo from unselected bone marrow progenitors using a well-defined cytokine regimen (25).

Flow cytometric staining of CB 2 receptors on eosinophils
Purified human eosinophils were stained with a polyclonal rabbit anti-human CB 2 primary Ab or isotype control, followed by a goat anti-rabbit secondary Ab (AF-647). CB 2 expression was quantified by flow cytometry.

Shape change assay
Isolated eosinophils, PMNL or PBMC were pretreated as indicated and stimulated. Shape change was estimated by flow cytometry as the increase of forward scatter (26,27).

Migration
Chemotaxis and Chemokinesis: Purified human eosinophils or bmEos were pretreated as indicated, placed into the top of a 48-well micro-Boyden chamber (human) or a 96-well chemotaxis plate (bmEos), and were allowed to migrate toward the indicated chemoattractant or vehicle.

Calcium flux
Isolated human eosinophils or bmEos were loaded with Fluo-3-AM in the presence of 0.02% pluronic F-127 (28). Changes in intracellular Ca 2+ were detected as fluorescence increase in the FL1-(530/30 nm) channel.

Adhesion assay under flow
Vena8 TM biochips (Cellix Ltd., Dublin, Ireland) were coated with ICAM-1, and coated channels were superfused with purified eosinophils. Adhesion was monitored using a Hamamatsu ORCA-03G digital camera and CellixVenaFlux software.

CD11b-upregulation
Whole blood samples or PMNL were pretreated as indicated and were incubated with agonists for 30 min at 37°C (29). Samples were stained with anti-CD16-PE-Cy5 and anti-CD11b-PE (ICRF44) Ab. CD11b upregulation was analyzed by flow cytometry.

In vivo chemotaxis
Eight-week-old IL-5Tg mice were treated i.p. with JWH-133 (5 mg/kg/day) or vehicle for three consecutive days. In vivo chemotaxis of eosinophils was induced by intranasal instillation of 4 lg eotaxin-2/CCL24. Bronchoalveolar lavage fluid (BALF) was collected 5 h postinstillation, and migration was evaluated by flow cytometric counting of highly granular (high side scatter) CD11c À /Siglec F + cells.

Mouse model of allergic lung inflammation
Eight-week-old female C57Bl6/N mice were immunized by i.p. injections of 10 lg of OVA adsorbed to Al(OH) 3 on days 0 and 7. Mice were challenged by an aerosol of OVA in saline on days 14 and 16. Additionally, mice received a daily i.p. injection of CB 2 agonist/antagonist (10 mg/kg) or vehicle on day 9 to day 16. On day 17, either airway hyperresponsiveness to methacholine was recorded with the FlexiVent system (Scireq, Montreal, QC, Canada) or BALF was taken and analyzed by flow cytometry.

Results
CB 2 receptor expression is enhanced on eosinophils from allergic donors CB 2 expression has been reported for B cells, monocytes/ macrophages, NK cells, basophils, and eosinophils, and at lower levels, for neutrophils and T cells (16,30,31). In this study, CB 2 surface expression was confirmed by flow cytometric staining in-and off-season in human peripheral blood eosinophils from allergic volunteers with seasonal respiratory symptoms and healthy subjects. Of note, quantification revealed~3.4-fold higher expression of CB 2 on eosinophils from symptomatic allergic donors compared to healthy controls. Interestingly no CB 2 upregulation was observed off-season in eosinophils from asymptomatic allergic donors (Fig. 1).
The CB 2 receptor agonist JWH-133 specifically enhances human and mouse eosinophil function Conflicting data exist regarding the effect of CB 2 activation in eosinophils and neutrophils, and little is known about its impact on basophil function (32,33). Thus, here we explored the effect of the selective CB 2 agonist JWH-133 on human eosinophils, neutrophils, and basophils by means of shape change.
Having encountered a chemotactic factor in vivo, leukocytes immediately begin to rearrange their cytoskeleton and change their shape to facilitate their attachment to microvascular endothelial cells. Such morphological changes can be detected by flow cytometry as changes in the forward scatter properties of the cells (34). Representative plots are provided in Fig. S5. First, purified eosinophils were pretreated with JWH-133 (100 nM) or vehicle, cells were stimulated with serial dilutions of eotaxin-2/CCL24, and shape change was assessed by flow cytometry. As shown in Fig. 2A, JWH-133 led to a significant increase of eosinophil responses; particularly, the sensitivity to eotaxin-2/CCL24 was increased up to threefold. To verify that this effect was mediated through selective CB 2 activation, eosinophils were exposed to the CB 2 antagonist SR144528 (1 lM, 10 min at RT) prior to the assay. As illustrated in Fig. 2B, SR144528 totally prevented the enhancing effect of JWH-133. Similarly, the endocannabinoid 2-AG concentration-dependently increased eotaxin-2/CCL24-induced shape change and elicited a weak, but dose-dependent shape change by itself with a maximum response observed at 500 nM ( Fig. 2C), similar to JWH-133 (Fig. 2D). Emphasizing the cell type-specific impact of selective CB 2 activation, pretreatment with JWH-133 only slightly enhanced basophil shape change as induced by eotaxin-2/CCL24 (Fig. 2E) and did not affect neutrophil responses to IL-8 (Fig. 2F).
The same pattern of CB 2 -mediated priming was observed in adhesion assays under flow conditions. Pretreatment of purified human eosinophils with 100 nM JWH-133 led to a significant twofold increase of firm adherent cells on ICAM-1 coated channels compared to eotaxin-2/CCL24 alone. Again, this effect was completely prevented with the CB 2 antagonist SR144528 (1 lM) (Fig. 3C). Representative pictures are provided in Fig. 3D and E.
To explore the direct contribution of CB 2 activation to eosinophil recruitment, in vitro and in vivo migration experiments were performed. As shown in Fig. 4A, JWH-133 itself elicited a moderate chemotactic response in human eosinophils but not in mouse eosinophils (Fig. S2), similar to the previously noted effect of the endogenous CB 2 agonist 2-AG (19). The maximal chemotactic response (CI~3.9) was observed at~100 nM of JWH-133. At micromolar levels JWH-133 evoked chemokinesis, with a maximal response (CĨ 1.8) observed at 3 lM. Moreover, at low nanomolar concentrations JWH-133 (5 nM) significantly increased the migratory capacity of human eosinophils (Fig. 4B) and bmEos (Fig. 4C) toward eotaxin-2/CCL24 by~80% and 28%, respectively. Emphasizing the physiological importance of these observations, systemic administration of JWH-133 (5 mg/kg/day) for three consecutive days significantly enhanced the eotaxin-2/CCL24-directed accumulation of eosinophils in the airways of IL-5 transgenic mice by~29% (Fig. 4D). To further confirm the direct involvement of CB 2 in the recruitment of mouse eosinophils, bmEos were isolated from WT and CB 2 -KO mice, treated with vehicle or JWH-133 (250 nM), and then allowed to migrate toward serial dilutions of eotaxin-2. Figure 4E shows that the priming effect of the cannabinoid agonist is not present in CB 2 -KO mice, whereas JWH-133 treatment resulted in a significantly increased migratory response of bmEos from WT mice.

CB 2 receptor activation enhances respiratory burst but not degranulation in human eosinophils
At sites of allergic inflammation, activated eosinophils cause tissue damage by the production of reactive oxygen species (ROS) and the release of toxic granule proteins (36). To explore the impact of selective CB 2 activation on eosinophil respiratory burst, purified human eosinophils were applied to 250 nM JWH-133 for 5 min at 37°C and respiratory burst was induced with eotaxin-2/CCL24. As shown in Fig. 5A, JWH-133 itself failed to induce respiratory burst, but significantly increased the eotaxin-2/CCL24 induced ROS production up to~25%. Interestingly, although having significant effects on eosinophil recruitment and ROS production, JWH-133 neither induced nor primed human eosinophil degranulation as assessed by means of CD63 upregulation and EPO release ( Fig. 5B and C).
JWH-133 modulates eosinophil responses in a pertussis toxin (PTX)-insensitive manner CB 2 receptors are known to activate heterotrimeric G i/o type G proteins leading to the inhibition of adenylyl cyclase (AC). Furthermore, CB 2 signaling was shown to involve MAPK (p38 and p42/44) and PI3K activity (37). To explore whether the observed effects depend on Ga i activation, purified human eosinophils were incubated in the presence of pertussis toxin (PTX; 5 lg/ml for 20 min at 37°C) and shape change was induced with PGD 2 (PTX-insensitive) or eotaxin-2/CCL24 (PTX-sensitive). Notably, pretreatment with PTX did not affect the modulatory effects on PGD 2 -induced shape change ( Fig. 6A and B). As expected, PTX pretreatment almost completely blocked the Ga i -dependent CCR3/CCL24 response but not the priming properties of JWH-133 ( Fig. 6C and D).

JWH-133 induces calcium release in human and mouse eosinophils
Having confirmed that CB 2 signaling in human eosinophils is transmitted in a G ai -independent manner, we next aimed to (1 nM), respectively. As expected, Ca 2+ responses to eotaxin-2/CCL24 were almost completely blocked, whereas JWH-133-and 2-AG-induced Ca 2+ responses remained unaffected (Fig. 6H). In contrast, pretreatment with the PLC inhibitor U-73122 (but not its inactive form U73343), as well as the IP 3 receptor antagonist 2-APB abolished Ca 2+ responses induced by JWH-133 and 2-AG indicating that CB 2 interacts with Ga q rather than Ga i . 2-APB is also capable of inhibiting TRP channels. Thus, to exclude the involvement of TRP, control experiments in the presence of 3 mM EGTA were conducted, which showed that Ca 2+ influx from the extracellular space and hence TRP channels are not involved in the Ca 2+ response to CB 2 receptor activation (Fig. 6H).
Having confirmed that JWH-133 augments eosinophil responsiveness to other chemoattractants via MEK1/2 and ROCK signaling in shape change, we investigated the selective contribution of MEK1/2 and ROCK to JWH-133-induced chemotaxis. Similarly, pretreatment with U-0126 and Y-27632 prior to the chemotaxis assay effectively inhibited eosinophil migration toward serial dilutions of JWH-133 ( Fig. S4G and H).

Systemic application of JWH-133 worsens airway hyperreactivity (AHR) in mice
Our in vitro data clearly show that CB 2 ligands significantly contribute to eosinophil activation and responsiveness. To prove the in vivo relevance of these observations we used an acute model of OVA-induced asthma in mice. In brief, mice were immunized to OVA on days 0 and 7, were treated i.p. with JWH-133 (10 mg/kg), SR144528 (10 mg/kg), a combination of both, or vehicle from day 9 to day 16, and were challenged with inhaled OVA aerosol on days 14 and 16. On day 17, either airway hyperresponsiveness to methacholine was recorded, or BALF was taken. As illustrated in Fig. 7, JWH-133 significantly impaired airway resistance and compliance compared to the vehicle group ( Fig. 7A and B), whereas SR144528 pretreat-ment fully prevented these effects of JWH-133 ( Fig. 7C and  D). Of the potent inflammatory lipid mediators comprising the cysteinyl leukotrienes (CysLTs: LTC 4 , LTD 4 , and LTE 4 ), only LTE 4 is stable and abundant in vivo. Besides mast cells, eosinophils are the main source of CysLTs which contribute not only to bronchoconstriction and airway hyperreactivity (39), but as in the case of LTE 4 also to eosinophil recruitment (40). Increased levels of LTE 4 can be detected in urine (41), BALF (42), and exhaled breath condensate (43) of patients after aller- gen challenge. Accordingly, mass spectrometric analysis of BALF from OVA-challenged mice revealed slightly increased LTC4 levels (Fig. 7E) and significantly elevated concentrations of LTD 4 and LTE 4 in JWH-133-treated mice compared to vehicle-treated controls. Again, the effect of JWH-133 could be abolished with SR144528 pretreatment (Fig. 7F and G). We furthermore determined eosinophil counts in the BALF of JWH-133-and vehicle-treated mice. Consistent with our results from the in vitro and in vivo migration assays, flow cytometric analysis revealed threefold higher eosinophil counts in the JWH-133-treated group compared to controls (Fig. 7H). Representative histology images of lung sections are provided in Fig. S6. To confirm the eosinophil-specific effect of the CB 2 agonist to the pathogenesis of allergic asthma, further experi-  in eosinophil-deficient ΔdblGATA mice were conducted. Treatment with JWH-133 worsened airway resistance in WT mice, but mediated beneficial effects on lung parameters in the ΔdblGATA group (Fig. S7A/B). Thus, our results show unequivocally that systemic CB 2 activation directly contributes to the pathophysiology of asthma in mice by enhancing eosinophil migration and effector function.

Discussion
In this study, we propose a novel mechanism of CB 2 -induced priming of eosinophils that may directly contribute to the pathogenesis of eosinophilic diseases. This notion is supported by the observation that pretreatment with the selective and stable CB 2 receptor agonist JWH-133 profoundly increased eosinophil responsiveness toward chemoattractants such as eotaxin-2/CCL24 and PGD 2, with respect to shape change, integrin expression, adhesion, chemotaxis and ROS production . The magnitude of primed responses was dependent on the concentration of JWH-133 and of the chemoattractant. In line with these in vitro data, systemic application of JWH-133 amplified the eotaxin-2/CCL24-directed recruitment of eosinophils into the airways of IL-5Tg mice and exacerbated OVAinduced asthmalike inflammation by increasing eosinophil influx into lungs and worsening of AHR. Oka et al. were the first who demonstrated that EoL-1 cells and human peripheral blood eosinophils express CB 2 but not CB 1 receptors (30). Consistent with these data, we found that purified human eosinophils from healthy subjects express an appreciable amount of CB 2 on mRNA (data not shown) and protein levels. Enhanced gene expression of CB 2 has been shown for lung eosinophils in allergic patients after allergen challenge (44). Accordingly, we could demonstrate that also CB 2 protein expression is significantly increased on the cell surface of peripheral blood eosinophils from symptomatic allergic donors, supporting a role for CB 2 and its ligands in the regulation of allergen-induced eosinophilic inflammation.
Activation with the selective agonist JWH-133 at nanomolar levels only induced a slight chemotactic response in human eosinophils, but amplified eotaxin-2/CCL24 induced eosinophil migration, especially at low concentrations. In contrast, IL-8-induced neutrophil activation remained unaffected by JWH-133, which is in line with previous reports demonstrating that the effects of 2-AG on neutrophils are mainly caused by the degradation of 2-AG to AA and subsequent de novo synthesis of LTB 4 (32).
Preactivation or priming of eosinophils by proinflammatory cytokines in the peripheral blood is a crucial step in the pathogenesis of allergic diseases. Several priming-dependent eosinophil responses, such as migration (45), adhesion (46), and degranulation (47), have been shown to be amplified in allergic patients. Moreover, a 'hyperadhesive' eosinophil phenotype, characterized by increased levels of the adhesion molecule integrin a M b 2 (CD11b/CD18) has been described in allergic patients after allergen challenge (48). Here we could show that exposure to the selective CB 2 receptor agonist JWH-133 prior to stimulation with chemoattractants yields significantly upregulated amounts of CD11b on the surface of both human and mouse eosinophils. Accordingly, JWH-133 also primes eosinophils for an enhanced capacity to adhere to ICAM-1 under physiological flow conditions, an important prerequisite of endothelial transmigration.
To further prove the in vivo relevance of the observed modulating activities of CB 2 , well-established mouse models of directed eosinophil migration and OVA-induced asthma were performed. Of note, we could show that daily treatment with JWH-133 not only aggravated lung parameters, but also led to increased eosinophil counts in the airways of OVA-challenged mice. Interestingly, these effects were completely absent in eosinophil-deficient ΔdblGATA mice, indicating that eosinophils are the major target of JWH-133 in allergic inflammation. Moreover, BALF of JWH-133-treated animals contained significantly higher CysLT levels compared to control animals; again, pointing to a more severe disease state due to systemic CB 2 activation. Noteworthy, Larose et al. demonstrated that 2-AG alone or in combination with platelet-activating factor (PAF) induced CysLT biosynthesis in human eosinophils in vitro. However, 2-AG/PAF-induced synthesis was blocked by MAG lipase inhibitors, indicating that this effect is more related to 2-AG degradation and metabolic transformation into eicosanoids than to selective CB 2 activation (37). In contrast to 2-AG, JWH-133 is not degradable to AA, and therefore cannot be further used for CysLT synthesis indicating a direct correlation exists between the observed enhanced CysLT levels and activation of CB 2. In accordance with our findings, recent work provided further evidence that CB 2 receptor activation is capable of enhancing inflammatory processes. Treatment with the CB 2 agonist JWH-133 potentiated adipose tissue inflammation in mice on high fat diet (49). Notably, 2-AG plasma levels were also found to positively correlate with the body mass index (BMI) in humans (50). Interestingly, the prevalence of asthma, its severity, and reduced responsiveness to standard medication seem to be associated with obesity and high BMI (51). However, the molecular mechanisms leading to these derangements are still poorly understood, but it is tempting to speculate that CB 2 activation by systemically elevated 2-AG levels and facilitation of eosinophil recruitment into the airways might be involved.
In both, human eosinophils and monocytes, 2-AG-induced CB 2 signaling was reported to be Ga i/o -dependent as pretreatment with PTX abrogated the chemotactic responses of these cells (30). Conversely, our data strongly suggest a Ga i/ o /adenylyl cyclase-independent pathway, substantiated by the fact that PTX was unable to prevent eosinophil priming and Ca 2+ flux following CB 2 activation. Furthermore, we observed that the PLC inhibitor U-73122 and the IP 3 receptor antagonist 2-APB were capable of reducing JWH-133-and 2-AG-induced Ca 2+ release. Thus, eosinophil CB 2 receptors seem to interact with Ga q rather than, or in addition to, Ga i proteins . Of note, Shoemaker et al. showed previously that cannabinoid agonists display different rank orders of potencies and receptor occupancies for regulation of intracellular effectors. Endogenous ligands such as 2-AG are more 'efficient' agonists requiring only half the receptor occupancy to elicit same effects as synthetic agonists. Accordingly, we found that 2-AG induced a much stronger Ca 2+ response compared to JWH-133 (52). CB 2 receptors have previously been shown to induce a G bc -dependent MAPK/ERK signaling cascade (53). Here we found that the induction of MEK 1/2 and ROCK activity is likely to be a part of the signaling mechanism accounting for the priming effect of JWH-133. Our observations are in line with previous findings indicating the involvement of MEK1/2-ERK activation in IL-5 and GM-CSF induced priming of human eosinophils (53). Moreover, MEK1/2-ERK and ROCK signaling regulates a variety of proinflammatory cellular processes such as eosinophil migration, degranulation, and respiratory burst (38,55) which afford the progression of eosinophilic inflammation.
In summary, the results of the present study demonstrate for the first time that specific CB 2 activation represents a novel priming process leading to enhanced migratory responsiveness of human and mouse eosinophils in vitro and in vivo. This CB 2mediated amplification of eosinophil migration seems to occur G i/o /adenylyl cyclase-independent, but involves Ga q /MEK/ ROCK signaling. Previous studies in patients with asthma showed a beneficial effect of inhaled or orally taken cannabinoids by dilating bronchial smooth muscles (56,57). In contrast, our data provide evidence for the involvement of the endocannabinoid/CB 2 axis in the progression of allergic inflammatory processes and indicate possible undesirable proinflammatory effects of long-term cannabinoid use. Thus, specific CB 2 receptor antagonism may open a new therapeutic approach for allergic disorders and other eosinophil-driven diseases.

Acknowledgments
We thank Birgit Brodacz, Iris Red, and Kathrin Rohrer for their skilled technical assistance.

Conflicts of interest
The authors declare no conflict of interest.

Supporting Information
Additional Supporting Information may be found in the online version of this article: Data S1 Supplemental Material. Figure S1 The endocannabinoid 2-AG enhances eotaxin-2/ CCL24 induced CD11b upregulation. Figure S2 JWH-133 does not induce chemotaxis in mouse eosinophils. Figure S3 CB 2 desensitization and the effect on eotaxin-2/ CCL24 induced Ca 2+ flux. Figure S4 MEK1/2 and p160 ROCK are involved in the modulating effect of JWH-133. Figure S5 Representative images of eosinophil shape change assessed by flow cytometry. Figure S6 Representative histological pictures of paraffin sections of lungs from OVA-challenged mice. Figure S7 Eosinophils are required for JWH-133 induced aggravation of lung parameters.