The peripheral cannabinoid receptor knockout mice: an update

Abstract

This review gives an overview of the CB₂ receptor (CB₂R) knockout (CB₂R^−/−) mice phenotype and the work that has been carried out using this mutant mouse. Using the CB₂R^−/− mice, investigators have discovered the involvement of CB₂R on immune cell function and development, infection, embryonic development, bone loss, liver disorders, pain, autoimmune inflammation, allergic dermatitis, atherosclerosis, apoptosis and chemotaxis. Using the CB₂R^−/− mice, investigators have also found that this receptor is not involved in cannabinoid-induced hypotension. In addition, the CB₂R^−/− mice have been used to determine specific tissue CB₂R expression. The specificity of synthetic cannabinoid agonists, antagonists and anti-CB₂R antibodies has been screened using tissues from CB₂R^−/− mice. Thus, the use of this mouse model has greatly helped reveal the diverse events involving the CB₂R, and has aided in drug and antibody screening.

Introduction

The peripheral cannabinoid receptor (CB₂R) was the second cannabinoid receptor discovered (Munro et al., 1993), after the central cannabinoid receptor (CB₁R) (Matsuda et al., 1990). Both cannabinoid receptors are G-protein-coupled seven transmembrane (TM) receptors. Human CB₁R and CB₂R share 44% overall homology and 68% at the TM level (Munro et al., 1993). Human CB₁R and mouse CB₁R share 96% homology (Chakrabarti et al., 1995), while human CB₂R and mouse CB₂R share 82% homology (Shire et al., 1996b). Mouse CB₁R and CB₂R share 66% overall homology and 78% at the TM level (Shire et al., 1996a). CB₁R is expressed at high levels in brain tissue and to a lesser extent in peripheral tissues such as the adrenal glands, reproductive organs and on immune cells (Matsuda et al., 1990; Bouaboula et al., 1993; Galiegue et al., 1995). In contrast, CB₂R is mainly expressed in cells of haematopoietic origin. CB₂R expression has been demonstrated in spleen and thymus (Schatz et al., 1997), lymph nodes, Peyer's patches (Lynn and Herkenham, 1994) and immune system-derived cell lines. Human blood cell populations were reported to have different degrees of CB₂R expression with the following rank order: B cells>natural killer cells>monocytes>polymorphonuclear neutrophil cells>CD8⁺ T cells>CD4⁺ T cells (Galiegue et al., 1995; Carayon et al., 1998). The expression level of CB₂R in lymphocytes and macrophages has been shown to vary in relation to cell differentiation and activation state (reviewed in Klein et al., 2003). Thus, Carayon et al. (1998) reported that CB₂R expression varied depending on the stage of B-cell differentiation with virgin and memory B cells, expressing the highest levels of CB₂R mRNA followed by germinal-centre B cells and centroblasts (Carayon et al., 1998). Carlisle et al. (2002) showed that while resident macrophages lack CB₂R expression, thioglycollate-elicited and interferon-γ (IFN-γ)-primed macrophages have high-CB₂R levels (Carlisle et al., 2002). The expression of CB₂R gene in immune tissues has been reported to be 10–100 times that of CB₁R (Galiegue et al., 1995). More recently, CB₂R expression in osteoblasts, osteocytes and osteoclasts was observed (Ofeck et al., 2006). CB₂R expression has also been found in preimplantation embryos (Paria et al., 1995), and more recently in the normal central nervous system (CNS) (Ross et al., 2001; Van Sickle et al., 2005; Ashton et al., 2006; Gong et al., 2006; Onaivi, 2006; Onaivi et al., 2006). The existence of a putative third cannabinoid receptor, GPR55, has been reported. GPR55 is an orphan G-protein-coupled receptor that has low-sequence homology (10–15%), compared to that of CB₁R or CB₂R, and is expressed in the testis at approximately a 15-fold higher level than in the brain (Baker et al., 2006). However, full characterization of this receptor is lacking, and it cannot be concluded that it is a true cannabinoid receptor (Petitet et al., 2006). Recently, Sugiura et al. (2007) were unable to stimulate GPR55 activation using three different cannabinoids (Sugiura et al., 2007). To investigate the role of CB₁R and CB₂R, mutant mice with deletions in these receptors have been developed. Thus far, there are three lines of CB₁R knockout (CB₁R^−/−) mice (reviewed in Valverde et al., 2005), and two CB₂R knockout (CB₂R^−/−) mice, one developed by Buckley et al. (2000) and the one recently developed by Deltagen (San Mateo, CA, USA) and commercially available through Jackson Laboratory (Bar Harbor, ME, USA) (http://jaxmice.jax.org/strain/005786.html#genes). The goal of the present review is to provide an update on the multiple findings investigators have achieved with the use of the CB₂R^−/− mice generated by Buckley et al. (2000).

CB₂R^−/− mice

The CB₂R^−/− mice were generated by homologous recombination. Upon homologous recombination, the last 341 bp of the CB₂R-coding exon in the gene was replaced by the neomycin gene, effectively removing bp 1186 onwards from the coding sequence (see accession U21681 in GenBank). The CB₂R deletion results in a gene lacking the coding region for part of the intracellular loop 3, TM regions 6 and 7 and the C terminus (Buckley et al., 2000). This deletion renders the CB₂R nonfunctional as macrophages derived from CB₂R^−/− mice are unable to respond to THC (Buckley et al., 2000; Chuchawankul et al., 2004). However, these mice are responsive to the psychotropic effects of cannabinoids (Buckley et al., 2000). The CB₂R^−/− mice display no gross morphological differences from their wild-type counterparts, but at the cellular level, the CB₂R^−/− mice are deficient in splenic marginal zone B cells, peritoneal B1a (CD5⁺CD11b⁺CD23^loB220^lo) cells, splenic memory CD4⁺ T cells, and intestinal natural killer cells and natural killer T cells (Ziring et al., 2006). Furthermore, CB₂R^−/− mice have a decreased number of diaphyseal osteoblast precursors, but have an increased osteoclast number and increased activity of their trabecular osteoblasts. This phenotype causes CB₂R^−/− mice to undergo a greater age-related bone loss than their wild-type counterparts, a loss of bone mass that is apparent by 8 weeks of age (Ofeck et al., 2006).

CB₂R^−/− mice and reproduction

The CB₂R^−/− mice are fertile, care for their young and have litter sizes comparable to their wild-type counterparts (Buckley et al., 2000). CB₂R is present in the preimplantation embryo, but not in the oviduct or uterus (Das et al., 1995; Paria et al., 1995, 2001; Wang et al., 2004). CB₂R is expressed from the one-cell through the blastocyts embryonic stage, restricted to blastocyts inner cell mass, but not in the trophectoderm-derived trophoblast stem cells, which are directly involved in implantation (Paria et al., 1995). On the other hand, CB₁R expression is present in the trophectoderm of the preimplantation embryo (Paria et al., 1995; Yang et al., 1996). Congruent with this, was the finding that CB₁R^−/− and CB₁R^−/−/CB₂R^−/−, but not CB₂R^−/− embryos, were trapped in the oviduct (Wang et al., 2006). Furthermore, it has been shown that the number of uterine implantation sites and oviductal embryo transport in CB₂R^−/− pregnant female mice was comparable to that of wild-type pregnant female mice (Wang et al., 2006). Interestingly, embryonic development in CB₂R^−/− mice is retarded (Paria et al., 2001). It was found that on day 3 of pregnancy, only 41.5% of CB₂R^−/− embryos were at the eight-cell stage as compared to the 84% wild-type embryos. There were 24% CB₁R^−/− embryos at the eight-cell stage. On day 4 of pregnancy, 71.3% CB₂R^−/− embryos were blastocysts and 25.7% morulas compared to the 97.5% wild-type blastocysts and 2.5% morulas. There were 61.5% CB₁R^−/− embryos at the blastocyst stage and 37% at the morula stage (Paria et al., 2001). However, CB₂R^−/− mice breeding pairs have normal litter sizes (Buckley et al., 2000), suggesting that the asynchronous CB₂R^−/− embryonic development at days 3 and 4 of pregnancy does not hinder embryonic implantation at day 4 or 5. Taken together, these findings indicate that CB₁R, and not CB₂R, is mostly responsible for successful embryonic implantation, and that both receptors have a role in synchronizing embryonic development. The function of CB₂R in the embryonic stem cells is unknown, but may implicate this receptor in specifying pluripotent inner cell mass cell lineage during blastocyst formation (Paria et al., 2001; Wang et al., 2006).

CB₂R^−/− mice and disease models

CB₂R^−/− mice and bone loss

Cannabinoid receptors have been implicated in bone mass, bone loss and osteoclast activity (Idris et al., 2005; Ofeck et al., 2006). Unregulated osteoclast (bone-resorbing cells) or osteoblast (bone-forming cells) activity can cause bone loss, resulting in diseases such as osteoporosis (Helfrich, 2003). The involvement of CB₁R on bone mass has been reported previously (Idris et al., 2005). CB₂R is expressed in osteoblasts, osteocytes and osteoclasts (Ofeck et al., 2006). Osteoclasts are derived from cells in the myeloid lineage (Boyle et al., 2003; Xing et al., 2005). Osteoblasts, the precursors of osteocytes, are derived from precursor cells in the stromal element of bone marrow (Rickard et al., 1996). CB₂R^−/− mice showed accelerated age-related trabecular bone loss and cortical expansion, although cortical thickness remained the same. CB₂R^−/− osteoclast number and trabecular osteoblast activity were increased. However, there was a significant decrease in the number of diaphyseal osteoblast precursors. In wild-type mice, the CB₂R-specific agonist HU-308 enhanced endocortical osteoblast number and activity and restrained trabecular osteoclastogenesis, apparently by inhibiting proliferation of osteoclast precursors. These findings suggest that CB₂R has a role in bone homeostasis and that this receptor is a potential drug target for the treatment of osteoporosis (Ofeck et al., 2006).

CB₂R^−/− mice and liver disorders

The CB₂R has been implicated to have an antifibrogenic role in the liver (Julien et al., 2005). Liver fibrosis occurs due to chronic liver injury that can eventually lead to cirrhosis and its complications. Julien et al. (2005) found that while normal liver does not express CB₂R, human cirrhotic liver does. In this study, it was found that CB₂R activation had potent antifibrogenic effects, including hepatic myofibroblast growth inhibition and increased apoptosis. Furthermore, when liver fibrosis was induced in wild-type and CB₂R^−/− mice using carbon tetrachloride (CCl₄), they found that CB₂R^−/− mice developed enhanced liver fibrosis compared to their wild-type counterparts. That is, they observed a higher collagen deposition as determined by histological examination and hydroxyproline measurement. In addition, smooth muscle α-actin mRNA (α-SMA) was higher for CCl₄-treated CB₂R^−/− mice compared to CCl₄-treated wild-type or vehicle-treated CB₂R^−/− mice (Julien et al., 2005). Hepatocyte proliferation was also impaired in CB₂R^−/− mice compared to wild-type mice (Deveaux et al., 2007). These findings provide evidence for a protective role of CB₂R in liver injury.

The CB₂R also seems to play a protective role in liver ischaemia/reperfusion (I/R) injury (Batkai et al., 2007). The I/R model of organ ischaemia followed by reperfusion mimics events that may develop in common disease such as myocardial infarction and stroke, coronary bypass surgery and organ transplantation. In the I/R model, liver ischaemia was induced for 60 min by clamping the hepatic artery and portal vein. Reperfusion was then allowed for 90 min or 24 h (Batkai et al., 2007). Liver damage to I/R was greater in CB₂R^−/− mice compared to wild-type mice as determined by increased serum transaminase AST/ALT and myeloperoxidase activities, malondialdehyde formation as an indicator or lipid peroxidation, and an increase in proinflammatory cytokines and chemokines. Histological examination also revealed much more extensive liver injury and neutrophil infiltration in the liver from CB₂R^−/− mice than that from wild-type mice. Interestingly, the hepatic levels of the endocannabinoids anandamide and 2-arachidonoylglycerol (2-AG) were also substantially elevated upon I/R (Batkai et al., 2007).

Taken together, the findings on the involvement of CB₂R in the liver fibrosis model and in the I/R model implicate the CB₂R in the liver repair mechanism.

CB₂R^−/− mice and pain

It is widely recognized that cannabinoids reduce pain in humans and animals. This effect by cannabinoids is thought to be largely mediated by the CB₁R. More recently, however, the implication of CB₂R in pain modulation is being recognized. CB₂R has been shown to modulate acute pain, chronic inflammatory pain, post-surgical pain, cancer pain and pain associated with nerve injury (reviewed in Whiteside et al., 2007). The CB₂R agonist, AM1241, has been shown to exert its antinociceptive effects locally without producing CNS effects (Malan et al., 2001). Furthermore, using three different assays of nociception, it was demonstrated that CB₂R^−/− mice have a lower threshold of pain compared to wild-type mice in the presence of cannabinoids. This was determined using the hot-plate, paw-withdrawal and tail-flick assays. While the hot-plate assay requires supraspinal responses, the paw-withdrawal and tail-flick assays require spinal reflexes. Thus, pain induced by these methods can be inhibited by drugs acting at these peripheral sites. In the absence of cannabinoids, using the Hargreaves' method (Hargreaves et al., 1988), the paw-withdrawal assay showed reduced latency in CB₂R^−/− mice compared their wild-type counterparts. Taken together, these findings provide evidence that CB₂R is involved in acute nociception, and thus, is a likely target to treat acute pain (Ibrahim et al., 2006).

CB₂R^−/− mice and experimental autoimmune encephalomyelitis, a multiple sclerosis model

Cannabinoids have been shown to alleviate spasticity associated with experimental autoimmune encephalomyelitis (EAE) in rodents (Baker et al., 2000; Pryce and Baker, 2007). To determine whether CB₂R is involved in the regulation of autoimmunity, EAE was induced in wild-type and CB₂R^−/− mice on a B10.PL background (Maresz et al., 2007). It was found that CB₂R^−/− mice exhibited a higher incidence of disease, a significantly increased clinical score and a reduced recovery rate than their wild-type counterparts. Furthermore, when EAE was induced in wild-type mice with CB₂R^−/− encephalitogenic T cells by adoptive transfer, there was a more severe clinical disease. The disease was characterized by a higher mortality rate and the presence of increased numbers of infiltrating mononuclear cells. Examination of the lesions revealed that there was an increase in the number of CB₂R^−/− T cells in and around the lesions when EAE was induced with CB₂R^−/− T cells as compared to that induced with wild-type T cells. This increase in T-cell number in the CNS was due to more proliferation and decreased apoptosis of the CB₂R^−/− T cells. Furthermore, the CNS CB₂R^−/− encephalitogenic T cells secreted more proinflammatory cytokines than the wild-type encephalitogenic T cells. This explains the more severe disease in mice receiving the CB₂R^−/− T cells (Maresz et al., 2007). It is known that endocannabinoids are synthesized in the CNS (Salzet et al., 2000) and that endocannabinoids have immune suppressive effects such as induction of apoptosis and inhibition of lymphocyte proliferation (Salzet et al., 2000; Klein et al., 2003), thus these findings provide evidence that the CNS actively suppresses T-cell function through the CB₂R (Maresz et al., 2007).

CB₂R^−/− mice and allergic dermatitis

A model of cutaneous contact dermatitis was used to study the allergic response in mutant mice. It was found that, compared to wild-type mice, CB₁R^−/−, CB₂R^−/− and CB₁R^−/−/CB₂R^−/− mice exhibit enhanced allergic responses to 2,4-dinitrofluorobenzene (DNFB) (Karsak et al., 2007). In this model, mice were sensitized with DNFB on the shaved abdomen. On days 5, 13 and 21, the mice were challenged with DNFB on the right ear. Swelling was then measured on the right ear and compared to the left ear within the same animal. While swelling was increased in the right ear of wild-type mice, swelling was significantly higher in CB₁R^−/−, CB₂R^−/− and CB₁R^−/−/CB₂R^−/− ears. Furthermore, upon injection of the CB₁R and CB₂R antagonists, SR141716A and SR144528, respectively, the swelling was enhanced in wild-type animals. Swelling, however, was decreased by THC, but enhanced by HU-308. Moreover, the levels of the endogenous cannabinoids 2-AG and anandamide were elevated in CB₁R^−/−/CB₂R^−/− mice (Karsak et al., 2007). Interestingly, using a slightly different approach to induce a cutaneous reaction, another group reported a decrease, not an increase in the ear swelling of CB₂R^−/− mice. These investigators passively sensitized the mice by i.v. injection of 1 μg of monoclonal anti-dinitrophenol IgE followed by a topical challenge with DNFB (Ueda et al., 2007). Furthermore, this group reported that oral administration of SR144528 suppressed ear swelling (Ueda et al., 2005, 2007). Although both groups find that the CB₂R is involved in the cutaneous allergic response, their findings, using the CB₂R^−/− mice as well as the CB₂R antagonist, are opposite. While Karsak et al. (2007) implicate the endocannabinoid system in the attenuation of the inflammatory response, Ueda et al. (2005, 2007) propose that CB₂R participate in the induction of the cutaneous reaction. The discrepancy in their findings, and hence conclusions, may be due to the different approaches used to induce the allergic response in wild-type and CB₂R^−/− mice, and in delivering the SR144528 compound.

CB₁R^−/−/CB₂R^−/− and CB₂R^−/− mice in hypotension

Prolonged and profound hypotension is associated with diverse forms of shock, such as haemorrhagic (Wagner et al., 1997), endotoxaemic (Varga et al., 1998) and cardiogenic (Wagner et al., 2001) shock and the hypotension occurring in advanced liver cirrhosis (Batkai et al., 2001). THC, anandamide and 2-AG are known to cause long-lasting hypotension and bradycardia in most animal models (Benowitz and Jones, 1975; Varga et al., 1995; Jarai et al., 2000; Cohen et al., 2002). The cannabinoids anandamide and abnormal cannabidiol are known to induce vasodilation of mesenteric arteries in wild-type and CB₁R^−/−/CB₂R^−/− mice. This vasodilation is inhibited by SR141716A, but not by other CB₁R antagonists (Jarai et al., 1999). A newly discovered endocannabinoid-like molecule, N-arachidonoyl-L-serine (ARA-S) has also been shown to produce endothelium-dependent vasodilation that is not reversed by SR141716 nor by SR144528. ARA-S binds very weakly to CB₁R and does not bind to CB₂R. ARA-S decreases LPS-induced plasma TNF-α levels in wild-type, CB₁R^−/−/CB₂R^−/− and CB₂R^−/− mice (Milman et al., 2006). This LPS-induced hypotension and cardiac contractility were prevented by SR141716. However, LPS-induced hypotension and its inhibition by SR141726 were similar in wild-type, CB₁R^−/− and CB₁R^−/−/CB₂R^−/− mice (Batkai et al., 2004). These findings prove that these cannabinoids cause hypotension via a non-CB₁R/CB₂R mechanism.

CB₂R^−/− mice and atherosclerosis

Atherosclerosis is a chronic inflammatory disease of the large arteries, and is the primary cause of heart disease and stroke in Western countries (Libby, 2002). It has been found that oral administration of THC significantly reduced the progression of atherosclerosis in the apolipoprotein mouse model (ApoE^−/−), an animal model for atherosclerosis. The CB₂R antagonist SR144528 reversed the effect of THC, suggesting the involvement of the CB₂R. Splenocytes derived from ApoE^−/− mice treated with THC showed reduced concanavalin A (Con A)-induced proliferation and IFN-γ secretion. THC was also found to decrease ApoE^−/− and wild-type thioglycollate-induced macrophage migration in response to the monocyte chemoattractant protein-1, but not the migration of CB₂R^−/− macrophages (Steffens et al., 2005). Apoptosis of macrophages is an important event in the pathophysiology of atherosclerosis (Liu et al., 2005). Oxidized low-density lipoproteins (OxLDL) are a major lipid component of atherosclerotic lesions and endocytosis of OxLDL is a potent inducer of apoptosis in cultured macrophages (Reid et al., 1993; Hardwick et al., 1996). Several macrophage processes associated with ongoing atherogenesis are regulated by CB₂R (Zhu et al., 1998; Maccarrone and Finazzi-Agro, 2003; Steffens et al., 2005). It has been found that wild-type peritoneal macrophages treated with OxLDL and its cholesterol derivative 7-ketocholesterol readily undergo apoptosis. OxLDL- and 7-ketocholesterol-induced apoptosis in CB₂R^−/− peritoneal macrophages was significantly decreased compared to wild-type macrophages. Staurosporine-induced apoptosis was similar in wild-type and CB₂R^−/− peritoneal macrophages, indicating that apoptotic mechanism are intact in the CB₂R^−/− macrophages. These findings suggest that CB₂R influences the development and progression of atherosclerotic lesions by mediating the apoptotic response of macrophages to OxLDL (Thewke et al., 2007). Taken together, these observations suggest that drugs targeting the CB₂R may be valuable tools to treat atherosclerosis.

CB₂R and macrophage chemotaxis

A crucial event occurring early in an inflammatory response is the migration of macrophages towards the chemostimulant. It has been found that peritoneal macrophage response to the chemokine RANTES/CCL5 is significantly inhibited by THC, CP55,940 and by the CB₂R-specific agonist O-2137, but not by the CB₁R agonist, ACEA. Moreover, the inhibition by THC was reversed by SR144528 but not by SR141716. THC treatment had a minimal effect on the chemotactic response of CB₂R^−/− peritoneal macrophages (Raborn et al., 2007). These findings implicate the CB₂R in the modulation of macrophage migration in response to chemoattractants.

CB₂R^−/− mice and infectious models

CB₂R^−/− mice and bacterial infections

Δ-9-Tetrahydrocannabinol treatment of mice has been shown to suppress the immune response to Legionella pneumophila, an intracellular bacterium that causes Legionnaires' disease. In this model, THC suppresses Th1 immunity while enhancing Th2 response (Newton et al., 1994; Klein et al., 2000). The Th2-biasing effect of THC involves cannabinoid receptors, suppression of IL-12 and IL-12 receptor and IFN-γ along with an increase in the Th2-biasing transcription factor GATA 3 (Klein et al., 2004). Dendritic cells are potent antigen-presenting cells that upon maturation produce co-stimulatory molecules and cytokines such as IL-12 that bias helper T cells towards Th1 immunity (Kapsenberg, 2003). Since THC treatment significantly suppressed IL-12 serum levels in infected mice, dendritic cells were isolated from wild-type, CB₁R^−/− and CB₂R^−/− mouse bone marrow, they were infected in vitro with L. pneumophila and treated with THC. THC was able to suppress IL12p40 production in dendritic cells derived from all three mice genotypes. When SR141716 was used on CB₂R^−/− dendritic cells and SR144528 on CB₁R^−/− dendritic cells, the antagonists only partially attenuated the THC-induced suppression of IL-12 production by the cells. These findings suggest partial participation of CB₁R and CB₂R in the dendritic response to THC in this infectious model (Lu et al., 2006).

CB₂R^−/− mice and parasitic infections

Infection of wild-type mice with the Plasmodium berghei is a model for cerebral malaria. This infection leads to 100% mortality after 6–7 days of infection. It was found that CB₂R^−/− mice are resistant to cerebral disease after being infected with this parasite. Infection leads to comparable microglial migration in the CNS of wild-type and CB₂R^−/− mice. However, there is an increase in the number of CD11b⁺ cells in spleens of infected CB₂R^−/− mice compared to infected wild-type mice. The reason for the resistance of CB₂R^−/− mice to cerebral malaria is still unknown (Alferink et al., 2007).

CB₂R^−/− mice, lymphocyte proliferation and cytokine production

It is known that cannabinoids and cannabinoid receptors modulate lymphocyte proliferation and cytokine production (reviewed by Klein et al., 2003; Massi et al., 2006). To investigate whether the endogenous cannabinoid 2-AG modulates cytokine production via CB₁R and CB₂R, Ouyang, Kaplan and colleagues used the CB₁R^−/−/CB₂R^−/− double knockout mice. They found that 2-AG suppressed interleukin-2 (IL-2) and IFN-γ production in phorbol myristate 13-acetate/ionomycin (PMA/Io)-treated wild-type mouse splenocytes (Ouyang et al., 1998; Kaplan et al., 2005). In addition, 2-AG suppressed IFN-γ production in splenocytes derived from CB₁R^−/−/CB₂R^−/− mice (Kaplan et al., 2005). 2-AG ether, a nonhydrolysable analogue of 2-AG, also suppressed IL-2 expression in wild-type and CB₁R^−/−/CB₂R^−/− splenocytes (Rockwell et al., 2006). Our findings are consistent with these. We isolated splenocytes and CD4⁺ T cells from wild-type and CB₂R^−/− mice, and treated them in vitro with the T-cell mitogen Con A or with anti-CD3 and anti-CD28 antibodies. The cells were cultured in the presence or absence of 2-AG or WIN 55,212-2. We found that 2-AG and WIN 55,212-2 inhibited the secretion of IL-2 and IFN-γ in wild-type and CB₂R^−/− splenocytes stimulated with Con A (Figures 1 and 2). Furthermore, WIN 55,212-2 inhibited the secretion of IL-2 and IFN-γ in wild-type and CB₂R^−/− and CD4⁺ T cells stimulated with anti-CD3 and anti-CD28 (Table 1). However, WIN 55,212-2 did not alter the secretion of TGF-β in wild-type or CB₂R^−/− splenocytes and CD4⁺ T cells (data not shown). In addition, 2-AG and WIN 55,212-2 did not alter proliferation in wild-type or CB₂R^−/− splenocytes stimulated with Con A. However, WIN 55,212-2 inhibited proliferation in wild-type and CB₂R^−/− CD4⁺ T cells stimulated with anti-CD3 and anti-CD28 antibodies. This suggests that the effect of WIN 55,212-2 on cell proliferation may be responsible for decrease in cytokine production seen with this cannabinoid (Buranapramest, 2006). Taken together, these findings suggest that CB₁R and CB₂R are not involved in the inhibition of splenocyte proliferation and IL-2 and IFN-γ secretion by 2-AG and WIN 55,212-2, respectively.

Figure 1.

Cannabinoids inhibited the secretion of interleukin-2 (IL-2) in CB₂R^+/+ and CB₂R^−/− splenocytes stimulated with concanavalin A (Con A). CB₂R^+/+ or CB₂R^−/− splenocytes (1 × 10⁶ cells per ml per well) were stimulated with Con A (2.5 μg ml⁻¹) and treated with the indicated concentrations of 2-arachidonoylglycerol (2-AG) (a) or WIN 55,212-2 (b). IL-2 secretion levels were determined from 72-h cell culture supernatants by ELISA. Data are expressed as the mean of triplicate samples ±s.d. and are representative of three independent experiments. ^*Significantly different from untreated control, P<0.05.

Figure 2.

Cannabinoid inhibited the secretion of interferon-γ (IFN-γ) in CB₂R^+/+ and CB₂R^−/− splenocytes stimulated with concanavalin A (Con A). CB₂R^+/+ or CB₂R^−/− splenocytes (1 × 10⁶ cells per ml per well) were stimulated with Con A (2.5 μg ml⁻¹) and treated with the indicated concentrations of 2-arachidonoylglycerol (2-AG) (a) or WIN 55,212-2 (b). IFN-γ secretion levels were determined from 72-h cell culture supernatants by ELISA. Data are expressed as the mean of triplicate samples ±s.d. and are representative of three independent experiments. ^*Significantly different from untreated control, P<0.05.

Table 1.

IL-2 and IFN-γ secretion in anti-CD3 and anti-CD28 stimulated CD4⁺ cells treated with WIN 55,212-2

WIN 55,212-2 (nM)	CB2R+/+ cells, IL-2 (pg ml−1±s.d.)	CB2R−/− cells, IL-2 (pg ml−1±s.d.)	CB2R+/+ cells, IFN-γ (pg ml−1±s.d.)	CB2R−/− cells, IFN-γ (pg ml−1±s.d.)
0	49.17±9.42	158.20±52.84	134.50±17.05	139.00±27.02
3	47.64±5.20	115.60±90.35	133.50±32.23	124.40±42.46
10	22.37±15.19	77.04±21.67	87.54±6.35*	80.20±2.86*
31	14.22±2.42*	30.72±9.90*	110.40±23.10	101.50±13.88
100	38.98±27.10	64.98±11.11	122.20±5.86	123.80±24.58
316	48.92±45.70	72.78±53.71	93.67±18.04	96.37±13.11
1000	7.65±1.87*	45.38±0.60*	96.85±12.90	88.89±15.90
3162	13.10±8.50*	18.75±5.36*	91.21±5.21*	82.98±10.50*

Abbreviations: IFN-γ, interferon-γ; IL-2, interleukin-2.

Purified total T cells or CD4⁺ T cells were isolated by negative selection using the Pan T Cell Isolation Kit or CD4⁺ T cell Isolation Kit, respectively (Miltenyi Biotech, Auburn, CA, USA) from spleens. CB₂R^+/+ or CB₂R^−/− CD4⁺ T cells (1 × 10⁵ cells per 0.1 ml per well) were stimulated with anti-CD3 (5 μg ml⁻¹) and anti-CD28 (0.5 μg ml⁻¹) antibodies and treated with the indicated concentrations of WIN 55,212-2. IL-2 and IFN-γ secretion levels were determined from 72-h cell culture supernatants by ELISA. Data are expressed as the mean of triplicate samples ±s.d. and are representative of three independent experiments. ^*Significantly different from untreated control, P<0.05.

CB₂R^−/− mouse cells and intracellular Ca²⁺ levels

The CB₁R^−/−/CB₂R^−/− double knockout mice have also been used to show that the effects of cannabinoids on intracellular calcium increase is not dependent on CB₁R and CB₂R. Rao and Kaminski (2006) found that THC (12.5 μM), cannabinol (20 μM) and HU-210 (20 μM) induced a rise in intracellular calcium in wild-type and CB₁R^−/−/CB₂R^−/− splenocytes. Interestingly, SR141716A and SR144528 (1 μM) attenuated the rise in calcium elicited by the cannabinoids. The partial attenuation of the cannabinoid-induced calcium increase in wild-type and CB₁R^−/−/CB₂R^−/− splenocytes, suggest that the antagonists used may be acting via mechanism distinct from those involving CB₁R and CB₂R.

CB₂R^−/− mice and drug specificity

Recently, the CB₂R^−/− mouse has been used to test drug specificity. The specificity of GW405833 for CB₂R has been tested using this mutant mouse. Inflammatory hyperalgesia was induced in wild-type or CB₂R^−/− mice by an intraplanar injection of Freund's complete adjuvant and 24 h later GW405833 was given intraperitoneally. Tactile allodynia was developed by wild-type and CB₂R^−/− mice in response to the Freund's complete adjuvant injection. While GW405833 was able to reverse this effect in wild-type mice, it did not produce a reduction in tactile allodynia in the CB₂R^−/− mice (Valenzano et al., 2005; Whiteside et al., 2005). Furthermore, it was shown that the central effects of high dose of GW405833 were not mediated by CB₂R as determined by carrying out the hot-plate, tail-flick and rotarod tests in CB₂R^−/− mice and finding similar responses to those found in wild-type mice (Whiteside et al., 2005). These findings show that GW405833 is an effective drug that, at low doses, targets the CB₂R.

The specificity of AM1241 and HU-308 in in vivo LPS-induced TNF-α and IL-10 production has also been tested using the CB₂R^−/− mouse. AM 1241 (50 mg kg⁻¹, i.p.) and HU-308 (30 mg kg⁻¹, i.p.) inhibited LPS-induced TNF-α production. To test if this effect was mediated by the CB₂R, AM1241 and HU-308 were given to wild-type and CB₂R^−/− mice. An hour later, mice were given LPS (0.15 mg kg⁻¹, i.p.). It was found that plasma TNF-α was significantly inhibited and IL-10 significantly elevated in wild-type and CB₂R^−/− mice in response to these agonists. These findings indicate that in vivo administration of AM1241 and HU-308 alter LPS-induced TNF-α and IL-10 production independent of the presence of CB₂R (Huang et al., 2007). However, another study revealed that some LPS-induced physiological events are mediated by the CB₂R. Recently, Duncan et al. (2007) have found that while wild-type animals are able to mount a febrile response to an in vivo LPS (100 μg kg⁻¹, i.p.) challenge, CB₂R^−/− mice do not (Duncan et al., 2007).

Thus, it can be concluded that the CB₂R^−/− mouse model is a very useful tool in drug discovery.

CB₂R^−/− mice and antibody specificity

Tissues from the CB₂R^−/− mice have been used to test the specificity of antibodies raised against the CB₂R. The expression of CB₂R protein in wild-type and CB₂R^−/− mice has been studied by immunohistochemistry in diverse tissues.

Although several laboratories failed to detect CB₂R in the brain (Derocq et al., 1995; Galiegue et al., 1995; Schatz et al., 1997; Carlisle et al., 2002; Ibrahim et al., 2003, 2006), recent studies have revealed the presence of this receptor in diverse brain regions. Using polyclonal antibodies generated against the C terminus of CB₂R from Alpha Diagnostics (San Antonio, TX, USA), Van Sickle et al. (2005) reported the expression of CB₂R protein in neurons of the brainstem from wild-type animals. To test the specificity of their antibody, they carried out immunostaining on the dorsal motor nucleus of the vagus from wild-type and CB₂R^−/− mice. While they found CB₂R expression in the wild-type mice, there was no immunostaining in CB₂R^−/− mice (Van Sickle et al., 2005). Using rabbit anti-human CB₂R polyclonal antibody (Cayman Chemicals, Ann Arbor, MI, USA and Sigma, St Louis, MO, USA), Onaivi, Gong and colleagues have found CB₂R widely expressed in the brain. Unfortunately, the specificity of the Cayman anti-CB₂R antibody was not tested using the CB₂R^−/− brain. Instead, using a different antibody, one raised against the C terminus of CB₂R (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), they showed that while CB₂R expression was found in the interpolar part of spinal 5th nucleus of the wild-type mouse brain, CB₂R expression was not found in the same brain region of CB₂R^−/− mice (Gong et al., 2006; Onaivi, 2006; Onaivi et al., 2006). Coincident with these data is the observation that CB₂R is present in neural progenitors from late embryonic stages to adult brain. Moreover, activation of the CB₂R in vitro promoted neural progenitor cell proliferation, this was not seen in CB₂R^−/− cells. In addition, in vivo treatment with HU-308 increased hippocampal progenitor proliferation in wild-type but not in CB₂R^−/− mice (Palazuelos et al., 2006). On the other hand, Wotherspoon et al. (2005) have not been able to find CB₂R expression in sections of naive mouse dorsal root ganglia or spinal cord using the anti-CB₂R antibody from Cayman. Although CB₂R immunoreactivity was seen following unilateral nerve damage, and was localized to the superficial laminae of the dorsal horn of the spinal cord, ipsilateral to the nerve damage. Antibody specificity was confirmed in CB₂R^−/− spinal cord and spleen sections.

Using the anti-CB₂R antibody from Cayman, expression of CB₂R was also found in mouse atherosclerotic plaques (Steffens et al., 2005) and in osteoblasts, osteocytes and osteoclasts (Ofeck et al., 2006) of wild-type mice. CB₂R antibody specificity was determined by using spleens and bone from CB₂R^−/− mice.

When obtaining data using the currently available polyclonal anti-CB₂R antibodies, one should be cautious. At the moment, there are no CB₂R-specific monoclonal antibodies available. It must therefore be recognized that the fine specificity of such commercial polyclonal antisera may change over time, depending on the bleed and individual animals being immunized during antibody production. Thus, tissues derived from CB₂R^−/− mice will continue to provide a very useful tool to elucidate the specificity of current and future anti-CB₂R antibodies.

Conclusion

The reports reviewed here emphasize the usefulness of the CB₂R^−/− mouse. Using this mutant mouse, investigators have discovered, or confirmed, CB₂R tissue and/or cellular expression under normal or abnormal conditions. Because the CB₂R was originally cloned and reported to be present in immune cells (Munro et al., 1993), it is not surprising that the CB₂R is found in cells of the haematopoietic lineage. The presence of this receptor in preimplantation embryonic stem cells (Paria et al., 1995) is very interesting and may implicate this receptor in the early development of haematopoietic cells and perhaps other cells. We found that in the embryonic rat, CB₂R was expressed in the liver in cells that appeared to be Kuffer cells (Buckley et al., 1998). The CB₂R has also been shown to be expressed in other normal adult cells. These include cells within the CNS (Van Sickle et al., 2005; Gong et al., 2006; Onaivi, 2006; Onaivi et al., 2006) and bone (Ofeck et al., 2006). In other instances, however, CB₂R expression is only evident under certain disease conditions. Thus, CB₂R expression is upregulated in pain models (Wotherspoon et al., 2005), atherosclerosis (Steffens et al., 2005), liver disorders (Julien et al., 2005) and during inflammation (Maresz et al., 2007) (Figure 3). On the basis of the findings using the CB₂R^−/− mouse, investigators have speculated as to the function of this receptor. CB₂R activation is thought to be antinociceptive and anti-inflammatory. It is also thought to be involved in bone homeostasis and in protective mechanisms during atherosclerosis and liver injury. On the other hand, using this mutant mouse, investigators have found that the CB₂R is not involved in hypotension. The implication of CB₂R in so many events lends validity to the notion that CB₂R is biologically functionally relevant. Nevertheless, living organisms are complex and one can reason that CB₂R and its endocannabinoid ligands do not work alone and other important molecular players are equally necessary. Finally, the CB₂R^−/− mouse has been a very useful tool to help determine drug and antibody specificity and promises to continue to be very useful in drug and antibody discovery.

Figure 3.

Using the CB₂R^−/− mice, the CB₂R has been implicated in diverse diseases.

Abbreviations

2-AG

2-arachidonoylglycerol, endogenous cannabinoid

AM1241

synthetic cannabinoid, CB₂R selective agonist

SR141716

N-(piperidin-1-yl)-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide hydrochloride, synthetic cannabinoid, CB₁R selective antagonist

SR144528

N-[(1S)-endo-1,3,3-trimethyl bicycle [2.2.1]heptan-2-yl]-5-(4-chloro-3-methylphenyl)-1-(4-methylbenzyl)-pyrazole-3-carboxamide, synthetic cannabinoid, CB₂R selective antagonist

THC

Δ-9-Tetrahydrocannabinol

WIN 55,212-2

(R)-[2,3-dihydro-5-methyl-3-[(4-morpholinyl)methyl]pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl](1-naphthalenyl)methanone, synthetic cannabinoid agonist

Conflict of interest

The author states no conflict of interest.