Targeting cannabinoid receptor CB₂ in cardiovascular disorders: promises and controversies

Abstract

Cardiovascular disease is the leading cause of death and disability worldwide, which can be largely attributed to atherosclerosis, a chronic inflammation of the arteries characterized by lesions containing immune and smooth muscle cells, lipids and extracellular matrix. In recent years, the lipid endocannabinoid system has emerged as a new therapeutic target in variety of disorders associated with inflammation and tissue injury, including those of the cardiovascular system. The discovery that Δ-9-tetrahydrocannabinol (Δ9-THC), the main active constituent of marijuana, inhibited atherosclerotic plaque progression via a cannabinoid 2 (CB₂) receptor-dependent anti-inflammatory mechanism, and that certain natural and synthetic cannabinoid ligands could modulate the myocardial or cerebral ischaemia–reperfusion-induced tissue damage, have stimulated impetus for a growing number of studies investigating the implication of CB₂ receptors in atherosclerosis, restenosis, stroke, myocardial infarction and heart failure. The aim of this review is to update on recent findings and controversies on the role of CB₂ receptors in cardiovascular disease. Particular emphasis will be placed on novel insights in the potential cellular targets of CB₂ stimulation in cardiovascular system (e.g. endothelial and vascular smooth muscle cells, cardiomyocytes, infiltrating and/or resident monocytes/macrophages and leukocytes, etc.), their interplay and intracellular signalling mechanisms identified, as well as on experimental and clinical studies.

Introduction

The cannabinoid 2 (CB₂) receptor was first cloned from the human leukaemia cell line HL-60 in 1993 (Munro et al., 1993). It is highly expressed in spleen and immune cells and at low levels in various peripheral tissues under normal physiological conditions (Galiegue et al., 1995). In the cardiovascular system, CB₂ receptor expression has been established in the rat myocardium (Lepicier et al., 2007) and at low levels in human cardiomyocytes (Weis et al., 2010; Mukhopadhyay et al., 2010a), coronary endothelial cells and smooth muscle cells (Rajesh et al., 2007a; 2008). Under pathophysiological conditions such as inflammatory stimulation or tissue injury, increased CB₂ receptor expression levels have been reported in the cardiovascular system, which probably reflects a protective response to limit cell or tissue injury (Pacher and Mechoulam, 2011). For example, up-regulation of CB₂ receptor expression has been described in primary human endothelial and smooth muscle cells stimulated by pro-inflammatory triggers and/or mitogens (Rajesh et al., 2007a; 2008; Ramirez et al., 2012), in human and mouse atherosclerotic plaques (Steffens et al., 2005), neointimal lesions following balloon injury (Molica et al., 2012) and in the myocardium of chronic heart failure patients (Weis et al., 2010). Conversely, CB₂ immunostaining was significantly reduced in carotid artery plaques of stroke versus asymptomatic patients (Montecucco et al., 2011). Therefore, one can speculate that CB₂ signalling is part of a protective response against human plaque vulnerability, which is impaired in patients with acute vascular events. Further clinical studies are warranted to validate this hypothesis. Similarly, down-regulation of CB₂ was described in kidney biopsies from patients with advanced kidney nephropathy, suggesting impaired protective CB₂ regulation that counteracts the deleterious effects of signalling through CB₁ receptors (Barutta et al., 2011), consistently with protective effects of CB₂ receptor agonists in models of experimental nephropathy (Mukhopadhyay et al., 2010b; Barutta et al., 2011).

As to the experimental evidence for an implication of CB₂ in cardiovascular disease, there is a mounting number of studies suggesting a protective role for CB₂ receptors in mouse models of atherosclerosis (Steffens et al., 2005; Netherland et al., 2010; Zhao et al., 2010a,b; Hoyer et al., 2011), restenosis (Molica et al., 2012) and myocardial (Defer et al., 2009; Montecucco et al., 2009) and cerebral ischaemia/reperfusion injury (Zhang et al., 2007; 2008; Pacher and Hasko, 2008; Murikinati et al., 2010; Zarruk et al., 2012).

The selectivity of the endocannabinoids and synthetic ligands used in various cardiovascular and other related disorders towards CB_1/2 receptors discussed in the following parts is summarized in Table 1 (reviewed in Pertwee, 2005; Pacher and Mechoulam, 2011).

Table 1.

Ranges of K_i values for certain cannabinoid CB₁ and/or CB₂ receptor agonists or antagonists/inverse agonists for the in vitro displacement of [³H]CP55,940, [³H]HU243 or [³H]BAY38-7271 from CB₁- and CB₂-specific binding sites (reviewed in Pertwee 2005)

Agonis/ligand	CB1Ki value (nM)	CB2Ki value (nM)	Reference
Mixed (without marked CB1/2 selectivity)
Anandamide	61–543	279–1940	Pertwee (2005)
2-AG	58–472	145–1400	Pertwee (2005)
CP55,940	0.5–5	0.7–2.8	Pertwee (2005)
HU-210	0.06–0.1	0.2–0.5	Pertwee (2005)
R-(+)-WIN 55,212-2	1.9–123	0.3–16	Pertwee (2005)
Δ9-THC	5–53	3–75	Pertwee (2005)
CB1 selective agonists
ACEA	1.4–5.3	195 to >2000	Pertwee (2005)
R-(+)-methanandamide	17.9–28.3	815–868	Pertwee (2005)
CB2 selective agonists
JWH015	383	13.8	Pertwee (2005)
JWH133	677	3.4	Pertwee (2005)
O-1966	4071–6039	20.9–25.1	Wiley et al. (2002)
O-3853	1361–1657	3.5–8.5	Zhang et al. (2007)
HU-308	>10 000	22.7	Pertwee (2005)
HU-910	1400	6	Horvath et al. (2012)
Inverse CB2 agonist/antagonist
AM630	5152	31.2	Pertwee (2005)
SR144528	50.3 to >10 000	0.28–5.6	Pertwee (2005)
Inverse CB1 agonist/antagonist
SR141716A	1.8–11.8	515–13200	Pertwee (2005)
AM251	7.5	2290	Pertwee (2005)
AM281	12	4200	Pertwee (2005)
LY320135	141	14900	Pertwee (2005)

Intracellular signalling of CB₂

CB₂ receptors are GPCRs that exert their biological effects through heterotrimeric G_i/0-proteins (Klein et al., 2003; Howlett, 2005). Coupling to G_iα subunits inhibits AC and reduction in cAMP accumulation. This leads to decreased PKA activity and reduced phosphorylation of cAMP response-element binding-protein, thus decreasing gene expression (reviewed in Bosier et al., 2010). In addition, reduction of PKA activity causes decrease in constitutive inhibitory phosphorylation of the MAPK cascade (reviewed in Bosier et al., 2010). Simultaneously, CB₂ activation induces members of the MAPK family through Gβγ, which triggers expression of genes (e.g. involved in cell survival, proliferation or stress response) (reviewed in Klein et al., 2003; Howlett, 2005) (Table 2). A majority of genes induced by CB₂ activation is associated with nuclear translocation of transcription factor NF-κB in promyelocytic cells HL-60 transfected with the CB₂ receptor (Derocq et al., 2000).

Table 2.

Intracellular signalling in vascular and immune cells

Cell type/tissue	Receptor agonist	Intracellular target	Reference
HL-60	2-AG	MEK 1/2 → ERK 1/2, p38, Rho	(Kobayashi et al., 2001; Kishimoto et al., 2003)
Jurkat	JWH015	ERK 1/2	(Ghosh et al., 2006)
Human monocytes	JWH015	Akt, ERK 1/2	(Montecucco et al., 2008)
Human neutrophils	JWH133	ERK 1/2	(Montecucco et al., 2011)
Human coronary artery EC	JWH133, HU-308	NF-κB, Rho	(Rajesh et al., 2007a)
Human sinusoidal EC	JWH133, HU-308, HU-910	NF-κB?	(Batkai et al., 2007; Rajesh et al., 2007b; Horvath et al., 2012)
Human brain EC	JWH133, O-1966	NF-κB?	(Ramirez et al., 2012)
Human coronary artery SMC		Ras, p38 MAPK, ERK 1/2, SAPK/JNK, Akt	(Rajesh et al., 2008)
Mouse heart	JWH133	MEK 1/2 → ERK 1/2, JAK → STAT-3	(Montecucco et al., 2009)

HL-60, human promyelocytic leukaemia cell line; MEK, p44/42 MAPK (kinase activating ERK 1/2); SAPK, stress-activated protein kinase; ‘?’ depicts studies providing indirect evidence on the CB₂-dependent inhibition of NF-κB activation (e.g. inhibition of the expression of various adhesion molecules).

In contrast to CB₁, CB₂ receptor activation does not modulate ion channel function (reviewed in Demuth and Molleman, 2006; Bosier et al., 2010). As a result, CB₂ receptor-mediated Ca²⁺ responses are less pronounced (Felder et al., 1995; Zoratti et al., 2003; Rao et al., 2004; Schuehly et al., 2011) than the potent CB₁ receptor-mediated effects on Ca²⁺ fluxes (reviewed in Bosier et al., 2010). In calf pulmonary artery cells, Zoratti et al. (2003) observed CB₂-dependent increases in cytosolic Ca²⁺ via activation of PLC and subsequent release from endoplasmic reticulum stores.

In addition to G-protein-dependent signalling, GPCRs can also mediate G-protein-independent signalling that involves receptor phosphorylation as well as interaction with regulatory proteins mediating receptor internalization or acting as scaffolds to modulate G-protein-mediated signalling (reviewed in Bockaert et al., 2004; Ritter and Hall, 2009; Smith et al., 2010). So far, available data about CB₂ receptor modification and trafficking are limited and mainly based on stably CB₂ expressing cells. Bouaboula et al. (1999) reported constitutively active, phosphorylated (at serine 352) and internalized CB₂ receptor expression at basal levels in stably transfected CHO cells. The effect was sensitive to the CB₂ antagonist/inverse agonist SR144528, which blocked phosphorylation and up-regulated CB₂ levels at the cell surface. Conversely, CP55,940 agonist treatment resulted in sustained (>8 h) CB₂ receptor phosphorylation. This is in accordance with the data from Derocq et al. (2000) reporting CB₂ phosphorylation in response to CP55,940 activation in CB₂-transfected HL-60 cells, which was maximal after 15 min. In naturally CB₂-expressing microglial cells, receptor expression was shown both internalized and at the cell surface in absence of agonist, while agonist stimulation increased receptor internalisation (Carrier et al., 2004). A recent study also provided evidence for marked functional selectivity of cannabinoid receptor internalization, depending on the cannabinoid ligand used (Atwood et al., 2012).

Chronic exposure of CB₁-expressing cells to agonists results in a loss of response when cells are subsequently challenged after drug washout (Breivogel et al., 1999). This adaptation to prolonged drug exposure is known as receptor desensitization and occurs in part due to uncoupling of receptors from their downstream G-proteins and/or effectors (Smith et al., 2010). In cell lines stably transfected with CB₂, chronic endocannabinoid as well as synthetic agonist treatment resulted in CB₂ receptor desensitization and down-regulation, as determined by the capacity to inhibit AC activity and receptor binding assays (Shoemaker et al., 2005). Following internalization in endocytosis vesicles, it is thought that CB₂ receptors are resensitized by dephosphorylation which allows receptor recycling to the cell surface (Bouaboula et al., 1999; Grimsey et al., 2011).

Whether the above described mechanisms of CB₂ trafficking has physiological relevance in primary vascular and peripheral blood immune cells deserve further investigations.

Intracellular targets of CB₂ activation in vascular and immune cells

The regulation of normal vascular endothelial, smooth muscle and cardiomyocyte cell proliferation, survival, growth and differentiation involves extracellular ligands, growth factors and cytokines that bind to cell surface receptors and activate intracellular signal transduction cascades. In cardiovascular diseases, the role of several MAPK signalling pathways has been recognized, including ERK, JNK and p38 MAPK (Muslin, 2008).

A number of in vitro and in vivo studies have demonstrated the capacity of CB₂ agonists to interact with signalling pathways induced by other cell surface receptors under pathophysiological/inflammatory conditions, suggesting a cross-talk between individual signal transduction pathways (Table 2). For example, CB₂ receptors have been implicated in the modulation of immune cell migration (reviewed in Miller and Stella, 2008). In particular, monocytes treated with the CB₂ agonist JWH015 showed significantly reduced chemokine-induced migration, associated with reduced expression of corresponding chemokine receptors CCR2 and CCR1 as well as IFN-γ-induced adhesion molecule ICAM-1 induction (Montecucco et al., 2008). JWH015 also cross-desensitized human monocytes for chemokine-induced migration by its own chemoattractant properties. The underlying pathways involved PI3K/Akt and ERK 1/2, but not p38 MAPK.

There is also evidence for CB₂-mediated modulation of endothelial cell responses to the pro-inflammatory cytokine TNF-α. The healthy endothelium which separates blood and vessel wall is a highly selective permeable barrier with anti-adhesive properties. Endothelial injury results in increased permeability and subendothelial lipid accumulation, adhesion molecule up-regulation, release of cytokines and growth factors, and adherence of platelets and monocytes (reviewed in Libby et al., 2011). In human primary coronary artery (Rajesh et al., 2007a), liver sinusoidal (Batkai et al., 2007; Rajesh et al., 2007b; Horvath et al., 2012) and brain endothelial cells (Ramirez et al., 2012) CB₂ activation with selective synthetic agonists attenuated the bacterial endotoxin- or TNF-α-induced NF-κB and RhoA activation (Rajesh et al., 2007a), ICAM-1 and VCAM-1 up-regulation (Batkai et al., 2007; Rajesh et al., 2007a,b; Horvath et al., 2012; Ramirez et al., 2012), CCL2 release as well as transendothelial migration and/or adhesion of THP-1 monocytes or neutrophils (Batkai et al., 2007; Rajesh et al., 2007a,b; Horvath et al., 2012).

Vascular smooth muscle cell migration and proliferation are key events during vascular development, in response to injury and atherosclerosis (reviewed in Gerthoffer, 2007). Within the atherosclerotic plaque, smooth muscle cells together with connective tissue encapsulate the necrotic lipid core, thus protecting the plaque from rupturing (reviewed in Lusis, 2000). On the other hand, smooth muscle cells, like macrophages, can express a variety of receptors for lipid uptake and can form foam cells, thereby participating in the cellular accumulation of lipids within plaques. They may also contribute to monocyte recruitment through expression of cytokines and adhesion molecules (reviewed in Doran et al., 2008). In the pathogenesis of restenosis in response to endovascular coronary or peripheral artery interventions, smooth muscle cells play a detrimental role through excessive migration and proliferation, leading to critical reduction in blood flow (reviewed in Ferns and Avades, 2000). In human coronary artery smooth muscle cells, CB₂ activation inhibited the TNF-α-induced proliferation and migration via inhibition of TNF-α-induced Ras, p38 MAPK, ERK 1/2, SAPK/JNK and Akt activation (Rajesh et al., 2008).

In vitro role of CB₂ in oxidized LDL-induced apoptosis

Elevated levels of plasma cholesterol, in particular low-density lipoprotein (LDL), are recognized as a major cardiovascular risk factor and lead to higher concentrations in the subendothelial intimal space. In the intima, LDL is oxidatively modified by reactive oxygen species (ROS) produced in endothelial cells, resident macrophages or smooth muscle cells. Oxidized LDL may injure the endothelium and play a role in the increased leukocyte adherence (Maier et al., 1996; Vora et al., 1997; Wang et al., 1997; Aikawa et al., 2002). Furthermore, oxidized LDL accumulates in macrophages and triggers macrophage apoptosis within atherosclerotic lesions (reviewed in Lusis, 2000). Apoptosis of macrophages might be beneficial for plaque stability if apoptotic bodies are removed. Indeed, it has been demonstrated that impaired macrophage apoptosis triggers lesion formation in mice (Liu et al., 2005). In advanced lesions, however, apoptosis of macrophage-derived foam cells promotes the formation of a pro-thrombotic central lipid pool whose size correlates with plaque instability. Thus, in advanced lesions, macrophage apoptosis could be considered as a pro-atherogenic factor triggering plaque vulnerability and risk of acute plaque rupture (reviewed in Libby et al., 1996).

There is in vitro evidence for a role of CB₂ deficiency in oxidized LDL-induced macrophage apoptosis, which involves modulation of the Akt survival pathway (Freeman-Anderson et al., 2008). The apoptosis rate was significantly reduced in peritoneal macrophages from CB₂ knockout mice as compared with wild-type animals. While oxidized LDL inhibited Akt phosphorylation, the effect was impaired in CB₂-deficient macrophages. In rat peritoneal macrophages, oxidized LDL dose-dependently induced endocannabinoid levels as well as cannabinoid receptor CB₁ and CB₂ expression (Jiang et al., 2009). On the other hand, treatment with a synthetic agonist promoted cholesterol accumulation in these cells in a CB₁-dependent manner.

Effects of plant-derived or synthetic CB₂ agonists in experimental atherosclerosis

The potential implication of CB₂ receptors in atherosclerosis, at least in mice, evolved from the initial discovery that the non-selective agonist Δ9-THC was atheroprotective in a CB₂-dependent matter (Steffens et al., 2005). The inhibitory effect of Δ9-THC on atherosclerotic plaque progression was reversed by the CB₂ antagonist/inverse agonist SR144528. The reduction of plaque size was associated with lower relative plaque macrophage content. In vitro proliferative responses and IFN-γ release were inhibited in splenocytes from Δ9-THC-treated mice, and migration of peritoneal macrophages versus CCL2 was also reduced. CB₂ antagonism reversed the anti-migratory effects, and Δ9-THC did not affect migration of CB₂^−/− macrophages.

In 2010, Zhao et al. provided further evidence for anti-atherosclerotic effects of pharmacological CB₂ activation. In ApoE^−/− mice fed 8 weeks on high-cholesterol diet, daily i.p. injection with the synthetic non-selective agonist WIN55,212-2 during the last 2 weeks before harvest reduced plaque size, macrophage content and expression of adhesion molecules VCAM-1, ICAM-1 and P-selectin (Zhao et al., 2010b). Pharmacologic CB₂ antagonism with AM630 inhibited the atheroprotective effects of WIN55,212-2. In a second study, WIN55,212-2 (administered by daily i.p. injection for 8 weeks before analysis) reduced atherosclerotic plaque formation, lesional macrophage content and mRNA levels of inflammatory markers IL-6, TNF-α and CCL2, as well as NF-κB activation in ApoE^−/− mice fed 16 weeks on high-cholesterol diet (Zhao et al., 2010a).

Lessons from experimental atherosclerosis studies employing genetic CB₂ deficiency

So far, the postulated role for CB₂ in anti-atherosclerotic effects was exclusively based on non-selective agonists combined with pharmacological receptor blockade, which paved the way for further studies investigating the impact of genetic CB₂ receptor deficiency in experimental atherosclerosis. Several groups independently studied atherosclerosis development in LDLR^−/− or ApoE^−/− mouse models, based on double deficiency and/or bone marrow transplantation. Some controversy exists in the reported effects on atherosclerotic plaque size and composition between these studies, which might be only partially explained by the different genetic backgrounds and experimental strategies.

First, Netherland et al. (2010) reported accelerated macrophage and smooth muscle cell infiltration and a less stable plaque phenotype without changes in plaque size in LDLR^−/−CB₂^−/− mice. The CB₂^−/− strain used for generation of double knockout mice was from Nancy Buckley et al. (2000).

Second, Willecke et al. (2011) claimed that neither genetic deficiency nor activation of CB₂ modulated atherogenesis in LDLR^−/− mice. In fact, they reported no changes in plaque size but significantly increased macrophage and lipid plaque content in LDLR^−/−CB₂^−/− mice, which were independently generated with CB₂^−/− mice obtained from Jackson Laboratories (Bar Harbor, ME, USA). Surprisingly, they did not observe any effect on atherosclerotic lesion size, plaque composition or immune cell function in LDLR^−/− mice treated with CB₂ agonist JWH133. Based on the limited pharmacokinetic data provided in this study, it is difficult to assess if the lack of anti-atherosclerotic effect might be due to insufficient frequency and/or dosage of cannabinoid ligand administration. In an attempt to proof the in vivo efficacy of JWH133 administration, the authors performed additional experiments based on thioglycollate-induced peritonitis and found reduced peritoneal macrophage recruitment in JWH133-treated mice. However, no effect was observed on acute TNF-α-induced systemic cytokine release or leukocyte adhesion marker expression. In vitro flow chamber assays also failed to show inhibitory effects on peritoneal macrophage adhesion to endothelial cells. The latter could also be related to the possibility that peritoneal macrophages and/or endothelial cells were already activated during the handling, possible presence of high levels of endocannabinoids in the serum used to culture cells (Marazzi et al., 2011), as well as to the potential rapid internalization of the CB₂ receptors (Atwood et al., 2012) and concomitant activation of CB₁ receptors with opposing consequences (reviewed in Pacher, 2009; Pacher and Mechoulam, 2011), by the extremely high concentration of the JWH133 used. In contrast to the above mentioned study, several recent reports using JWH133 and other CB₂ agonists in vitro, ex vivo or in vivo, confirmed the anti-inflammatory effects of CB₂ activation in endothelium, and its inhibitory effect on monocytes/macrophages and/or leukocyte migration during unrelated pro-inflammatory conditions (Ni et al., 2004; Murikinati et al., 2010; Horvath et al., 2012; Ramirez et al., 2012; Zarruk et al., 2012).

A third study was conducted by Georg Nickenig's group and the CB₂^−/− mice from Andreas Zimmer (Buckley et al., 2000), based on the ApoE^−/− mouse model of atherosclerosis. The authors employed both double deficiency and bone marrow chimeric approach to demonstrate that CB₂ expression in both vascular as well as immune cells influences atherosclerosis development (Hoyer et al., 2011). ApoE^−/−CB₂^−/− mice had significantly increased plaque macrophage infiltration and aortic superoxide production, but a non-significant increase in plaque size. Similarly, lethally irradiated ApoE^−/− mice reconstituted with CB₂^−/− bone marrow had increased macrophage infiltration without significant increase in lesion size. On the other hand, treatment with CB₂ agonist JWH133 significantly reduced atherosclerotic lesion size, macrophage infiltration, superoxide production and improved vascular endothelium-dependent relaxations ex vivo in isolated aortic ring preparations. Interestingly, the authors further reported some changes in aortic levels of endocannabinoids and related lipid mediators (i.e. reduced 2-AG and increased OEA levels) in CB₂^−/− mice on C57BL6 wild-type background. Unfortunately, whether similar endocannabinoid levels are detectable in CB₂-deficient ApoE^−/− mice (on normal chow or high cholesterol diet) remains unclear.

Finally, Delsing et al. studied the effect of immune cell CB₂ deficiency in irradiated LDLR^−/− mice reconstituted with bone marrow from CB₂^−/− mice (from Andreas Zimmer). At the level of the aortic arch but not the aortic sinus, the lesional area was significantly increased in CB₂^−/− bone marrow transplanted mice as compared with mice receiving wild-type bone marrow after 12 weeks of high cholesterol (0.15%) diet (Delsing et al., 2011).

Pharmacological CB₂ activation: a new approach for restenosis prevention?

Since the first percutaneous transluminal coronary angioplasty (PTCA) performed in the 1970s (Gruntzig et al., 1979), this intervention has become one of the most common procedures to reopen occluded vessels in coronary artery disease patients. Despite significant advances in this technique based on inflation of a balloon-tipped catheter and the introduction of stents (including drug-eluting stents), restenosis remains the primary limitation of PTCA (reviewed in Bittl, 1996; Douglas, 2007; Finn et al., 2007).

Restenosis is an inflammatory process in response to arterial injury, leading to secretion of cytokines and growth factors, recruitment of inflammatory cells as well as increased migratory, proliferative and secretory responses of vascular smooth muscle cells (reviewed in Ferns and Avades, 2000; Weber et al., 2004; Gerthoffer, 2007). Efforts to limit this constrictive vascular remodelling process have focused on inhibiting smooth muscle cell proliferation and migration, leading to the development of local stent-based delivery of antiproliferative agents such as sirolimus and paclitaxel (reviewed in Finn et al., 2007). Although these drug-eluting stents reduce rates of restenosis compared with bare metal stents, still a significant percentage of higher-risk patients develop in-stent restenosis (reviewed in Douglas, 2007). Moreover, late stent thrombosis due to the lack of complete endothelial repair has emerged as a major safety concern (reviewed in Joner et al., 2006; Finn et al., 2007; Luscher et al., 2007). Therefore, novel strategies should be targeted on restenosis prevention without impairing the arterial healing process.

Since CB₂ receptor activation inhibited inflammatory proliferation and migration of vascular smooth muscle cells in vitro (Rajesh et al., 2008), we subsequently investigated the effect of CB₂ activation in a mouse model of balloon angioplasty (Molica et al., 2012). As reported in other models of organ injury or inflammation (reviewed in Pacher and Mechoulam, 2011), balloon injury increased vascular CB₂ expression in hypercholesterolaemic ApoE^−/− mice (Molica et al., 2012). Injured vessels of mice treated with CB₂ agonist JWH133 showed reduced intimal and medial thickening, associated with less in situ proliferation, smooth muscle cells and macrophages. Re-endothelialization was not inhibited by treatment with the CB₂ agonist, according to CD31 immunostaining. Conversely, CB₂ deficiency resulted in increased intima formation compared with wild-type mice, whereas JWH133 did not affect intimal formation in CB₂^−/− mice. Apoptosis rates assessed by in situ TUNEL staining were significantly higher in the CB₂ knockouts. In vitro proliferation rates were significantly increased in CB₂^−/− smooth muscle cells compared with wild-type cells. Bone marrow-derived CB₂^−/− macrophages showed enhanced adherence and migration compared with CB₂^+/+ macrophages. The underlying mechanisms involved increased mRNA levels of adhesion molecule ICAM-1, chemokine receptors CCR1 and CCR5, as well as the pro-inflammatory chemokine CCL2.

Implication of CB₂ in myocardial preconditioning based on ex vivo or in vitro models

An implication of the endocannabinoid system in the cardioprotective mechanisms of preconditioning has been initially described in isolated rat heart models (Lagneux and Lamontagne, 2001; Joyeux et al., 2002; Bouchard et al., 2003; Lepicier et al., 2003). However, whether CB₁, CB₂ or both receptors are involved in these models is controversial; furthermore the clinical relevance of these ex vivo model is limited because of the absence of the important inflammatory response (reviewed in Pacher and Hasko, 2008).

Cardioprotective effects of endocannabinoid-mediated CB₂ activation were first reported in LPS-induced preconditioning (Lagneux and Lamontagne, 2001). Perfusion with CB₂ antagonist SR144528 abolished the cardioprotective effect of LPS pretreatment, whereas CB₁ antagonism with rimonabant had no effect. The implication of NO in CB₂-dependent cardioprotection was shown by additional experiments using NOS inhibitor or NO donor respectively. Similarly, blocking of CB₂, but not CB₁ receptors reversed cardioprotection by heat stress-mediated preconditioning (Joyeux et al., 2002). As to endocannabinoid perfusion-mediated cardioprotection, one study reported that only CB₂, but not CB₁ (or only partially) reversed the protective effect (Lepicier et al., 2003). A different study, however, reported inhibition of endocannabinoid-mediated cardioprotection by both CB₁ and CB₂ antagonists (Underdown et al., 2005).

Two studies further investigated the link between endocannabinoid signalling and NO-mediated cardioprotection (Wagner et al., 2006; Lepicier et al., 2007). Both studies suggest the requirement for NO in CB₁, but not CB₂-mediated cardioprotection. By contrast, in isolated neonatal cardiomyocytes preventive effects of the plant-derived cannabinoid Δ9-THC against hypoxia were dependent on NO production and sensitive to CB₂, but not CB₁ antagonism (Shmist et al., 2006).

In vivo role of CB₂ in myocardial ischaemia/reperfusion injury, importance of inflammatory response

A different experimental approach is based on in vivo myocardial ischaemia/reperfusion in anaesthetized rodents, pretreated with cannabinoid receptor agonists or antagonists. The synthetic cannabinoid HU-210 decreased the incidence of ventricular arrhythmias following ischaemia/reperfusion in rats through activation of CB₂ receptors (Krylatov et al., 2001). Treatment with the non-selective agonist WIN55,212-2 before ischaemia significantly reduced the infarct size in mouse hearts (Di Filippo et al., 2004). The CB₂ antagonist AM630, but not the CB₁ antagonist AM251 abolished the effect of WIN55,212-2.

In a clinically more relevant experimental setting, CB₂ activation with JWH133 administered at the end of the ischaemic period significantly reduced the infarct size as compared with vehicle-treated mice (Montecucco et al., 2009). Serum levels of the clinical marker cardiac troponin I, which is released from necrotic cardiomyocytes, were significantly lower in mice treated with JWH133. The infarct size reduction was associated with a decrease in ROS production and neutrophil infiltration into the infarcted myocardium, and activation of cardioprotective signalling pathways (Montecucco et al., 2009). Pre-injection with kinase inhibitors for Akt, ERK 1/2, STAT-3 pathways partially abrogated the JWH133-mediated infarct size reduction. In vitro, JWH133 inhibited TNF-α induced chemotaxis and integrin CD18/CD11b up-regulation on human neutrophils (Montecucco et al., 2009), suggesting a mechanistic explanation that might contribute to the infarct size reduction and potential relevance for human pathology.

In addition, CB₂ receptors have been involved in the cardioprotective effects of remote ischaemic preconditioning, which is a protective phenomenon induced by preceding ischaemia in other organs or vascular beds (Hajrasouliha et al., 2008). Systemic pretreatment with CB₂ antagonist AM630, but not the CB₁ antagonist AM251, abolished the cardioprotective effects of remote preconditioning on infarct size and arrhythmias.

As to a potential clinical relevance for CB₂ signalling in myocardial infarction, a large case-control study enrolling 1968 individuals addressed the involvement of the gene encoding CB₂, CNR2, in the development of myocardial infarction and several cardiovascular risk factors. In particular, a potential association of genetic variations with the development of myocardial infarction and classic cardiovascular risk factors, including arterial hypertension, obesity, hypercholesterolaemia and diabetes mellitus was investigated (Reinhard et al., 2008). However, none of the 13 investigated single nucleotide polymorphisms in the CNR2 gene was associated with myocardial infarction or any of the investigated risk factors.

CB₂ in post-ischaemic repair and heart failure

Acute myocardial infarction leads to necrosis of cardiac myocytes, which induces repair mechanisms that lead to scar formation (Sun, 2009). This process of post-infarction cardiac remodelling involves adaptive changes in shape, size and function of the ventricle, which may ultimately lead to contractile dysfunction and heart failure. Defer et al. (2009) provided substantial evidence for a protective role of CB₂ receptors in ischaemic cardiac myocyte cell death, fibrosis and cardiac dysfunction. Hearts of CB₂^−/− mice had larger infarcts in comparison with wild-type mice and more sustained cell loss 3 days after ischaemia, together with accelerated injury and apoptosis in the non-ischaemic remote myocardium. Furthermore, increased infiltration of macrophages was observed in the infarct-surrounding myocardium of CB₂^−/− mice. Accelerated cardiac remodelling in CB₂^−/− post-ischaemic hearts was documented by higher number of spindle-shaped α-smooth muscle actin-positive myofibroblasts. In vitro, CB₂^−/− cardiomyocytes and fibroblasts were more susceptible to oxidative stress-induced cell death. Long-term effects of cardiac remodelling in CB₂^−/− hearts involved marked fibrosis, accelerated cardiomyocyte hypertrophy, dilative cardiomyopathy and cardiac dysfunction, as reported 4 weeks post infarction. By contrast, wild-type post-ischaemic hearts developed moderate fibrosis and cardiomyocyte hypertrophy, while cardiac function was preserved.

Heart failure is a complication of many diseases that affect the heart, such as coronary artery disease, arterial hypertension, valvular disease or diabetic cardiomyopathy (Jessup and Brozena, 2003; Boudina and Abel, 2007). Recent findings in humans reported an up-regulation of cardiac CB₂ expression and enhanced endocannabinoid AEA and 2-AG levels in patients with chronic heart failure (Weis et al., 2010). Analysis of cannabinoid receptor expression by real-time PCR and immunohistochemistry also revealed a slight but significant down-regulation of cardiac CB₁ receptors in chronic heart failure patients. Concerning potential pathophysiological consequences of CB₂ up-regulation, the authors speculate that it may have a negative inotropic effect due to reduced cAMP levels that may contribute to ventricular weakening. On the other hand, CB₂ receptors might mediate positive inotropic effects through cAMP-independent mechanisms, thus representing a compensatory mechanism to maintain cardiac performance. Moreover, CB₂ up-regulation could represent a protective response to counterbalance chronic heart failure-induced structural changes, as shown in mice (Defer et al., 2009).

In vivo role of CB₂ in cerebral ischaemia/reperfusion injury

Ischaemic stroke is a leading cause of death in developed countries and acquired adult disability. Increased accumulation of CB₂-positive macrophages derived from resident microglia and/or invading monocytes following cerebral ischaemia/reperfusion have previously been reported (Ashton et al., 2007). Numerous recent studies have investigated the role of CB₂ receptors by evaluating the effects of various CB₂ ligands or knockout mice in experimental models of stroke (models of cerebral ischaemia–reperfusion injury) (Ni et al., 2004; Zhang et al., 2007; 2008; Murikinati et al., 2010; Ramirez et al., 2012). Similarly to the effects observed in myocardial ischaemia/reperfusion injury, CB₂ receptor activation limited cerebral infarct size in experimental stroke by attenuating endothelial cell activation, chemokine signalling, inflammatory cell infiltration, glial activation, oxidative/nitrative stress and consequent cell death (Ni et al., 2004; Zhang et al., 2007; 2008; Murikinati et al., 2010; Zarruk et al., 2012). Most of these beneficial effects could be prevented by CB₂ receptor antagonist(s) and/or were absent in CB₂ knockout mice, which often exhibited enhanced injury indicative of the protective role of endocannabinoid system through CB₂ receptors. Principally comparable protection could also be seen in hepatic ischaemia/reperfusion injury models (Batkai et al., 2007; Rajesh et al., 2007b; Horvath et al., 2012), which emphasizes a key role of CB₂ receptors in limiting reperfusion damage in general and also provides a strong rationale for development of this promising approach for clinical use.

Interestingly, recent findings further suggest protective effects of the endocannabinoid-related brain compound N-arachidonoyl-l-serine (araS) in traumatic brain injury (Cohen-Yeshurun et al., 2011). Despite its structural similarity to anandamide, araS exhibits very low affinity at CB₁ and CB₂ receptors (Milman et al., 2006). Nevertheless, treatment with selective CB₂ and transient receptor potential vanilloid 1, but not CB₁ and GPR55 receptor antagonists reversed the protective effects in traumatic brain injury (Cohen-Yeshurun et al., 2011). This raises an intriguing possibility that certain metabolites of araS may exert direct protective effects on these receptors.

Conclusions

Collectively, several lines of evidence discussed above have established that functional CB₂ receptors in cells of the cardiovascular system (e.g. endothelial and vascular smooth muscle cells, cardiomyocytes, fibroblasts and resident immune cells), as well as in infiltrating monocytes/macrophages and leukocytes during various pathological conditions of the cardiovascular system (e.g. atherosclerosis, restenosis, stroke, myocardial infarction and heart failure) may play an important compensatory role in controlling tissue inflammation and injury (Figure 1). In most of the cases, these receptors limit inflammation and associated tissue injury; however, in certain cases or disease states, CB₂ receptor activation may also enhance tissue damage (Pacher and Mechoulam, 2011). On the basis of preclinical results, pharmacological modulation of CB₂ receptors may hold a unique therapeutic potential in stroke, myocardial infarction and atherosclerosis despite some controversies (Figure 1). Nevertheless, the successfully translation of these promising preclinical results to clinical practice requires a better understanding of the underlying pathology, CB₂ pharmacology and signalling, as well as the therapeutic window and long term safety of the use of various CB₂ ligands. Finally, there is an urgent need for better, more selective pharmacological tools that are suitable for clinical application.

Figure 1.

Potential therapeutic targets of CB₂ activation in cardiovascular disorders. EC, endothelial cell; RBC, red blood cell; SMC, smooth muscle cells.

Glossary

2-AG

2-arachidonoylglycerol

Δ9-THC

Δ-9-tetrahydrocannabinol

AEA

anandamide

MCP-1/CCL2

monocyte chemotactic protein-1

OEA

oleoylethanolamine

ROS

reactive oxygen species

Conflict of interest

None.