The complexities of the cardiovascular actions of cannabinoidsMichael D Randall, David A Kendall, and Saoirse O'Sullivan
Br J Pharmacol. 2004 May
The cardiovascular actions of cannbinoids are complex. In general they cause vasorelaxation in isolated blood vessels, while in anaesthetised animals they cause multiphasic responses which involve an early bradycardia and long-lasting hypotension. However, in conscious animals, the picture is one of bradycardia followed by pressor responses. Clearly, the responses to cannabinoids are dependent on the experimental conditions and synthetic cannabinoids and endocannabinoids exhibit different pharmacologies. In terms of mechanisms involved in the vascular responses to cannabinoids, the following have been implicated: the involvement of 'classical' cannabinoid receptors, the involvement of a novel endothelial cannabinoid receptor, the release of nitric oxide, the release of endothelium-derived hyperpolarising factor (EDHF), the activation of vanilloid receptors, metabolism of endocannabinoids to vasoactive molecules, and both peripheral inhibition and central excitation of the sympathetic nervous system.
The cardiovascular effects of both synthetic and endogenous cannabinoids have been extensively examined and reviewed (see Hillard, 2000; Kunos et al., 2000; Ralevic et al., 2002; Randall et al., 2002). What is clear is that the cardiovascular actions of the cannabinoids are complex and appear to be complicated by differences in experimental approach and prevailing conditions. The overwhelming findings from studies on isolated blood vessels are that both endogenous and exogenous cannabinoids are vasodilators. This is mirrored by studies on anaesthetised animals that report hypotensive effects (Varga et al., 1995). However, findings in conscious animals are more complex and do not support the notion that cannabinoids are hypotensive agents (Stein et al., 1996; Lake et al., 1997; Gardiner et al., 2001; Gardiner et al., 2002a; Gardiner et al., 2002b). Similarly, there is no general consensus regarding the molecular target(s) for cannabinoids. Indeed, the field might have become blurred by the assumption that synthetic cannabinoids and endogenous cannabinoids share common pharmacology and that in vitro findings translate to the in vivo situation. The purpose of this review is to summarise the key findings and to attempt to resolve the issues raised by in vitro and in vivo comparisons.comparisons.
Schematic diagram showing possible mechanisms of vasorelaxation to anandamide. Putative mechanisms include (a) endothelium-dependent relaxation coupled to a novel endothelial cannabinoid receptor (CBx) coupled to EDHF (endothelium-derived hyperpolarizing (more ...)
The vascular effects of cannabinoids in isolated arteries (see Figure 1)
In vitro studies have identified that the prototypic anandamide is a potent vasodilator in a number of isolated vascular preparations. A more detailed overview of in vitro effects of anandamide can be found in the following reviews (Kunos et al., 2000; HÃ¶gestatt & Zygmunt, 2002; Randall et al., 2002). Some studies have implicated the endothelium in relaxant responses to anandamide (Pratt et al., 1998; Chaytor et al., 1999; Wagner et al., 1999), with the release of prostanoids (Ellis et al., 1995; Fleming et al., 1999), nitric oxide (Deutsch et al., 1997) or endothelial-derived hyperpolarising factor (EDHF) (Chaytor et al., 1999). Some, but not all, studies have reported that anandamide acts through the stimulation of cannabinoid CB1 receptors, although the intracellular pathway(s) coupling to vasodilatation have not been clearly identified (White & Hiley, 1997). More recently, a novel 'anandamide receptor' has been proposed to exist on the vascular endothelium (JÃ¡rai et al., 1999; Wagner et al., 1999; OffertÃ¡ler et al., 2003) and may be coupled to the release of EDHF.
An important step came in 1999 when Zygmunt and co-workers reported that anandamide stimulates vanilloid receptors on sensory nerves, leading to vasorelaxation via the release of the vasoactive neurotransmitter calcitonin gene-related peptide (CGRP), and this has since been widely confirmed (Ralevic et al., 2000; White et al., 2001; Ho & Hiley, 2003). Subsequent work, however, has suggested that the participation of sensory nerves might depend on the prevailing or experimental conditions and is less important in the absence of a functional nitric oxide system (Harris et al., 2002).
In other studies, anandamide has been shown to inhibit calcium channels (Gebremedhin et al., 1999) and activate various K+ channels (Randall & Kendall, 1997; Randall & Kendall, 1998; White et al., 2001), and these may also account for its vasorelaxant actions.
What is clear from the above is the lack of overall consensus or perhaps diverse mechanisms of action of cannabinoids; the following account will explore potential reasons for mechanistic differences.
The first documentation of the vasorelaxant effects of anandamide was in rabbit cerebral vessels (Ellis et al., 1995), and since then many other blood vessels from different species have been examined.
It is clear that the magnitude of relaxant responses to anandamide differs between preparations. For example, small resistance mesenteric vessels show 100% vasorelaxation to anandamide, while the larger superior mesenteric artery has a maximal relaxation of around 40% (O'Sullivan et al., unpublished observations). Cerebral vessels show a maximum relaxation of 25—50% to anandamide (Ellis et al., 1995; Gebremedhin et al., 1999; Wagner et al., 2001 and, similarly, coronary vessels on an average relax by around 50% (Pratt et al., 1998; White et al., 2001). In rat aortae, the maximum relaxation to anandamide is approximately 20% (O'Sullivan et al., 2004a). In contrast, rat and rabbit carotid arteries (Holland et al., 1999; Fleming et al., 1999) do not relax to anandamide. Such differences between vessels may be due to differences in receptor populations or the prevailing mechanisms. For example, it has been suggested that cannabinoid CB1 receptor expression is associated with mesenteric vessels, but not the thoracic aorta (Darker et al., 1998). There is also evidence that the putative endothelial cannabinoid receptor contributes to vasorelaxation in small mesenteric resistance vessels, but not the main superior mesenteric artery (O'Sullivan et al., unpublished observations).
Although vanilloid receptors on sensory nerves may play an important role in vasorelaxation to anandamide in mesenteric vessels (Zygmunt et al., 1999; Ralevic et al., 2000), this is not the case under all conditions (Harris et al., 2002) and in all vessels. In respect of the latter, it has been shown that sensory nerves do not play a role in coronary vessels from several species (Grainger & Boachie-Ansah, 2001; White et al., 2001; Ford et al., 2002). This indicates that the actions of anandamide may be dependent on vanilloid receptor density and/or density of perivascular nerve in a given blood vessel. This was emphasised in a study by Andersson et al. (2002) who showed that, while anandamide is a full agonist at the vanilloid receptor in mesenteric arteries, it is a weak agonist of this receptor in main bronchi. The authors attributed this to possible differences in receptor reserve and/or cellular uptake between the different tissues. Similarly, Vanheel & Van De Voorde (2001) reported that anandamide produces capsazepine-sensitive hyperpolarisations of daughter branches of the mesenteric artery, but not of the superior mesenteric artery, and suggested that this might relate to the relative density of perivascular sensory nerves or to regional differences in the distribution of vanilloid receptors. Such differences may explain some of the variation in vasorelaxant responses to anandamide between vascular beds, or indeed between different vessels within the same bed.
Some mechanisms of vasorelaxation may be specific to certain tissues, for instance nitric oxide has only been shown to mediate responses to anandamide in renal arteries (Deutsch et al., 1997) but not other vascular beds (Harris et al., 2002).
The vascular responsiveness towards anandamide varies between species. In the rat aorta, anandamide causes approximately a 20% maximal relaxation (O'Sullivan et al., 2004a), while in the rabbit aorta, this has been reported to be 80% (Mukhopadhyay et al., 2002). Similarly, in the rat coronary vessels anandamide causes about 30—40% relaxation (White et al., 2001), 50% relaxation in bovine vessels (Pratt et al., 1998) but 80% relaxation in ovine vessels (Grainger & Boachie-Ansah, 2001). However, anandamide does not cause vasorelaxation in porcine coronary vessels (Fleming et al., 1999). Interestingly, the only work so far in human vessels has shown that anandamide is not a vasorelaxant in myometrial arteries from pregnant women (Kenny et al., 2002).
One important difference to emerge is the involvement of metabolism to arachidonic acid metabolites. Studies are divided into those where anandamide acts directly and those where its actions are dependent on metabolism. For example, there is evidence from bovine and ovine coronary vessels (Pratt et al., 1998; Grainger & Boachie-Ansah, 2001) that vasorelaxation to anandamide is dependent on metabolism via epoxygenase or cyclooxygenase pathways. However, in rat mesenteric vessels it is universally reported that vasorelaxation to anandamide is unaffected by cyclooxygenase inhibition (White & Hiley, 1997). Despite this, in the rabbit mesenteric vessels, Fleming et al. (1999) reported that responses to anandamide were abolished by the cyclooxygenase inhibitor diclofenac, although additional actions of diclofenac cannot be excluded.
The metabolically stable analogue of anandamide, methanandamide, also exhibits vasorelaxant activities (Ralevic et al., 2000) and this would mitigate against metabolism being central to the actions of cannabinoids. Having said that, it is possible that methanandamide (Ralevic et al., 2000) has a greater dependence on sensory nerve activation than anandamide (Harris et al., 2002), presumably due to differences in their relative efficacies at vanilloid and cannabinoid receptors.
The various studies on the vascular actions of ananamide have been carried out under many different conditions, for example in both isolated arterial segments and intact perfused vascular beds, in the absence or presence of cyclooxygenase inhibitors, and against different spasmogens. These different approaches inevitably mean that straightforward comparisons between studies may be difficult.
The archetypal CB1 receptor antagonist SR141716A has been widely used to investigate the involvement of CB1 receptors in vasorelaxation to anandamide but this is confounded by the wide-ranging actions of SR141716A at various concentrations. For example, SR141716A also inhibits myoendothelial gap junctions (Chaytor et al., 1999), which themselves have been implicated in the actions of anandamide via endothelial-derived hyperpolarizing factor (EDHF) (Chaytor et al., 1999; Harris et al., 2002). SR141716A may also antagonise the novel, non-CB1 endothelial cannabinoid receptor proposed by Kunos and co-workers (JÃ¡rai et al., 1999; OffertÃ¡ler et al., 2003).
The pharmacological profile of SR141716A differs depending on the concentration used. For example, White & Hiley (1997) have shown SR141716A at 100 nM to be ineffective against vasorelaxation to anandamide in isolated mesenteric vessels, but 1 μM to be inhibitory. Similarly, Harris et al. (2002) have shown 3 μM, but not 1 μM, to be effective against anandamide-induced vasodilatation in the perfused mesenteric bed. Furthermore, in many studies where more that one CB1 receptor antagonist has been used, SR141716A has been effective at inhibiting anandamide-mediated relaxation, while AM251 or LY320135 have not (Chaytor et al., 1999; White et al., 2001; Ford et al., 2002; Harris et al., 2002).
In addition to anandamide, synthetic cannabinoid compounds and other recently identified endocannabinoids have been reported to have vascular effects in vitro. Many synthetic cannabinoid receptor agonists have been shown to have vasorelaxant effects. Δ9-Tetrahydrocannabinol (THC) causes indomethacin-sensitive relaxation of rabbit cerebral arteries (Ellis et al., 1995) and endothelium-independent vasorelaxation of isolated rabbit mesenteric vessels, which are sensitive to SR141716A (Fleming et al., 1999). Zygmunt et al. (2002) have also recently reported that THC causes release of CGRP from mesenteric vessels, although interestingly, this was not through stimulation of the vanilloid receptor subtype 1, and may involve another novel receptor in the vasorelaxant pathway to cannabinoids. The CB1 receptor agonist HU-210 has also been reported to cause dilatation in the rat-perfused mesenteric bed (Wagner et al., 1999) and SR141716A-sensitive vasorelaxation of coronary and cerebral vasculature (Wagner et al., 2001). In isolated rabbit mesenteric vessels, HU-210 causes endothelium-independent, SR141716A-sensitive vasorelaxation (Fleming et al., 1999). WIN55,212 ((R)-(+)-[2,3-dihydro-5-methyl-3-[(4-morpholino)methyl]pyrrolo-[1,2,3-de]-1,4-benzoxazin-6-yl](1-naphthyl)methanone), another potent CB1 receptor agonist, has been shown to produce vasorelaxtion of feline cerebral vessels and these responses are sensitive to SR141716A (Gebremedhin et al., 1999). However, in rat mesenteric vessels, WIN55,212 causes endothelium-independent vasorelaxation (White & Hiley, 1998), and this does not appear to be through stimulation of either CB1 or CB2 or vanilloid receptors (Ho & Hiley, 2003). Similarly, the cannabinoid analogue, abnormal cannabidiol, elicits an endothelium-dependent non-CB1/CB2/vanilloid relaxation of mesenteric vessels (Ho & Hiley, 2003), and this is thought to be mediated through the novel endothelial cannabinoid receptor (JÃ¡rai et al., 1999; OffertÃ¡ler et al., 2003).
The endocannabinoid, 2-arachidonoylglycerol (2-AG), has been shown to cause endothelium-independent vasorelaxation of rabbit mesenteric vessels, mediated by both CB1 and CB2 receptors (Kagota et al., 2001), but does not have an effect in the perfused rat mesenteric arterial bed (Wagner et al., 1999), possibly due to its instability. We have recently shown that another endocannabinoid, N-arachidonoyl-dopamine (NADA), causes endothelium-dependent vasorelaxation of isolated rat mesenteric vessels that involves the novel endothelial receptor (coupled to EDHF release) and endothelium-independent relaxations via vanilloid receptors (O'Sullivan et al., 2004b).
The cardiovascular effects of cannabinoids in vivo
The in vivo cardiovascular effects of cannabinoids are complex, with both increases and decreases in blood pressure being reported (Stark and Dews, 1980; Dewey, 1986).
Studies in anaesthetised animals
In parallel with studies on cannabinoids in isolated blood vessels, their in vivo cardiovascular actions have also been assessed. In 1995, Varga and co-workers reported that anandamide caused a triphasic response in anaesthetised rats. This included an initial vagally mediated bradycardia with secondary hypotension, a transient pressor effect followed by sustained hypotension, which was sensitive to both the cannabinoid CB1 receptor antagonist, SR141716A and interference with sympathetic control. These early conclusions led to the suggestion that anandamide acted via CB1 receptors to inhibit sympathetic control of blood pressure. A more extensive study in the following year confirmed the triphasic nature of the responses to anandamide, and it was concluded that the sustained depressor effect was due to presynaptic inhibition of sympathetic nerves (Varga et al., 1996). In vitro studies have also confirmed that cannabinoids inhibit sympathetic regulation (see Ralevic, 2003). Cannabinoid-induced sympathoinhibition by synthetic cannabinoids has also been reported in rabbits (Niederhoffer & Szabo, 1999) and rats (Niederhoffer et al., 2003). Significantly, Niederhoffer & Szabo (1999) reported that the synthetic cannabinoid, WIN55212-2, caused depressor effects in pithed rabbits with electrically stimulated, sympathetic tone which were opposed by the CB1 receptor antagonist, SR141716A, but that this was less marked in conscious animals. These findings clearly emphasise the influence of background sympathetic tone on responses to cannabinoids.
Haemodynamic studies in anaesthetised rats have also reported marked hypotension in response to anandamide, which appears to be due to reductions in peripheral resistance (Garcia et al., 2001). These responses were partly sensitive to SR141716A.
Malinowska et al. (2001) have reported that the initial bradycardia and depressor responses to anandamide in anaesthetised rats are due to activation of vanilloid receptors and that the long-lasting hypotensive phase was sensitive to SR141716A and thus presumed to be mediated via CB1 receptors. However, as commented above, SR141716A has a range of actions independent of antagonism of CB1 receptors and its inhibitory effects should be interpreted with caution. Further evidence for the potential participation of sensory nerves in the cardiovascular effects of anandamide comes from studies in anaesthetised rats which showed that intra-arterial injection of anandamide led to hypotension and increased ventilation (Smith & McQueen, 2001). These responses were mimicked by capsaicin, but inhibited by vanilloid receptor antagonists, desensitisation of vanilloid receptors and sectioning of the femoral and sciatic nerves, with the implication that they were due to sensory nerve reflexes evoked by anandamide.
In addition to studies on anandamide, the cardiovascular effects of 2-AG have also been investigated (Mechoulam et al., 1998; JÃ¡rai et al., 2000). In this regard, 2-AG was shown to cause hypotension in anaesthetised rats (Mechoulam et al., 1998) and in anaesthetised mice, there was hypotension and tachycardia which did not appear to be mediated via CB1 receptors but may have involved metabolism to arachidonic acid metabolites (JÃ¡rai et al., 2000). By contrast, similiar cardiovascular effects were observed for a stable analogue of 2-AG, but these appear to have been mediated via CB1 receptors.
In addition to endocannabinoids, a comparative study on the haemodynamic effects of HU210 and anandamide in anaesthetised rats reported that HU 210 caused a profound reduction in cardiac output leading to hypotension that was sensitive to SR141716A, while anandamide did not (Wagner et al., 2001). However, both agents were reported to cause cerebral and coronary vasodilatation, which were sensitive to SR141716A.
Studies in conscious animals
Comparative work in conscious rats has also reported that anandamide caused a profound bradycardia, with a short lived depressor effect but this was followed by a longer lasting pressor effect (Stein et al., 1996). The bradycardic effect was sensitive to cyclooxygenase inhibition and ascribed to the production of arachidonic acid metabolites. Similarly in the conscious rat, Lake et al. (1997) reported that there was vagal activation but the prolonged hypotensive effects reported in anaesthetised animals were absent. Their explanation was that the depressor effect was masked by the pressor effect and since they also reported that the depressor effect was present in conscious hypertensive rats, they speculated that the depressor response was dependent on the level of sympathetic tone. Subsequent studies in conscious rats have underscored the complex nature of the in vivo responses. In this regard, Gardiner et al. (2002a) reported that intravenous anandamide caused a transient pressor effect that was accompanied by regional (hindquarters, mesenteric and renal) vasoconstriction in conscious rats. At higher doses, there was pronounced bradycardia and in some instances there was a depressor effect prior to sustained hypertension, with some degree of hindquarters vasodilatation following constriction. These complex cardiovascular effects were insensitive to the cannabinoid CB1 receptor antagonist AM251. The bradycardia was opposed by atropine and the hindquarter vasodilatation appeared to be mediated via β2-adrenoceptors, possibly due to adrenaline release. Furthermore, when the early bradycardia was blocked with atropine the initial hypotension was absent. Parallel studies on synthetic cannabinoids (WIN-55212-2 and HU 210) in conscious rats (Gardiner et al., 2002b) have reported that these agents caused pressor effects, accompanied by renal and mesenteric vasoconstriction but hindquarters vasodilatation. In contrast to the actions of anandamide, these cardiovascular effects were sensitive to the cannabinoid CB1 receptor antagonist, AM251, but the hindquarters vasodilatation was also inhibited by a β2-adrenoceptor antagonist. Studies on conscious normotensive and hypertensive rats have also demonstrated pressor effects with WIN 55,212-2, which were sensitive to ganglion blockade (Gardiner et al., 2001). From these findings it was concluded that the effects of synthetic cannabinoids were mediated via CB1-receptors linked to increases in sympathetic activity. However, it should be noted that in in vitro studies synthetic cannabinoids reduce sympathetic activity (Ralevic, 2003). A further point to emerge from that study was that AM 251 alone did not affect blood pressure or regional haemodynamics, with the implication that endogenous cannabinoids do not influence cardiovascular control under resting conditions. A similar conclusion was also drawn from the CB1 knockout mice (Ledent et al., 1999). Although the role, if any, of endocannabinoids in cardiovascular regulation remains to be established, Rademacher et al. (2003) reported that SR141716A injected into the nucleus tractus solitarius delayed baroreflex recovery in anaesthetised dogs, with the implication that endocannabinoids might play a role in this regulatory system. Furthermore, in the rat, injection of anandamide into the nucleus tractus solitarius increased baroreflex sensitivity in an SR147116A-sensitive manner, possibly via modulation of GABAergic or glutamergic neurotransmission (Seagard et al., 2004). It was also observed that pharmacologically induced increases in blood pressure were accompanied by increases in endogenous anandamide in the nucleus tractus solitarius, pointing to a possible modulatory role.
In conscious mice, Ledent et al. (1999) identified a biphasic response to anandamide with an initial pronounced depressor effect followed by more sustained hypotension and these changes were accompanied by bradycardia. More importantly, it was reported that the cardiovascular responses to anandamide were absent in CB1-receptor knockout mice and, thus were assumed to be CB1 receptor-mediated.
In vivo studies will inevitability involve both central and peripheral effects and in this respect intracisternal administration of various cannabinoid agonists in conscious rabbits caused both sympathoexcitation and increased vagal output, with bradycardia and at high doses pressor effects were observed (Niederhoffer & Szabo, 1999). In anaesthetised rats, administration of synthetic cannabinoids into the rostral ventrolateral medulla oblongata leads to increased sympathetic activity and hypertension (Padley et al., 2003). Hence, central effects of cannabinoids may oppose their peripheral effects.
Studies in man
Administration of cannabis-derived cannabinoids (including via smoking) in man is associated with pronounced tachycardia (as opposed to bradycardia reported in animals above) (see Dewey, 1986: Jones, 2002). This is accompanied by an increase in circulating noradenaline release but demonstrates rapid tolerance on repeated administration (see Jones, 2002). The tachycardia is also sensitive to SR141716A, implicating the involvement of cannabinoid CB1 receptors (Huestis et al., 2001). According to Jones, the reasons for this difference in man compared to animal studies is not immediately clear but could be related to the high doses in animal studies and also differences in arousal between human volunteers and animals in the conscious and anaesthetised state.
Is there any consensus as to the in vivo actions of cannabionids?
The above studies have highlighted clear differences in the actions of cannabinoids depending on the experimental conditions. Studies in anaesthetised animals report a multiphasic response with a clear initial bradycardia and a final long-lasting hypotensive phase, probably mediated via sympathoinhibition. The bradycardia is also seen in conscious animals but the long-lasting hypotensive phase is not. The lack of a hypotensive phase under 'physiological conditions' could reflect differences in sympathetic activity between the conscious and anaesthetised state. Another possibility is that anaesthetic agents directly influence the responses. In this regard, anandamide has been shown to inhibit the TASK-1 potassium channel that is anaesthetic sensitive (Maingret et al., 2001) and this might have bearing on the differences between the anaesthetised and conscious state. It is also possible that the central effects of cannabinoids might be more susceptible to inhibition by general anaesthetics.
Is there a correlation between in vivo and in vitro effects?
The clear vasorelaxant effects reported from in vitro studies are largely postjunctional and will not be significantly influenced by in vivo control systems such as the autonomic nervous system, and this would certainly contribute towards differences between the two situations. Furthermore, central effects following in vivo administration may also complicate the peripheral effects.
Another consideration is the route of administration. In vitro studies are based on local application and this will lead to local effects, while in vivo studies usually involve systemic administration, with the potential for widespread effects and metabolism.
There is good evidence from in vitro studies that cannabinoids may exert dual effects on vascular control, for example action at vanilloid receptors may lead to sensory nerve-mediated vasodilatation but presynaptic cannabinoid receptors may oppose this (see Ralevic, 2003). Once again the predominant effect may be dependent on the prevailing conditions.
Some of the in vitro actions are uncovered when other systems are inhibited, for example, actions of anandamide via EDHF release are probably accentuated by removal of nitric oxide. However, in the in vivo situation, the physiological significance of EDHF has yet to be established. Once again, the actions of endocannabinoids may be dependent on the experimental conditions. Similarly, the balance between the endocannabinoid and endovanilloid actions of anandamide will also influence the overall effect.
The physiological significance of the cardiovascular effects of endocannabinoids are unclear and it may be that they are of more pathophysiological importance. In this regard, Wagner et al. (1997) demonstrated, in a rat model of haemorrhagic shock, that activated macrophages release anandamide which may contribute towards the hypotension. Similarly in endotoxic shock, the synthesis of 2-AG in platelets and anandamide in macrophages are increased (Varga et al., 1998). The release of anandamide by central neurones under hypoxic conditions, leading to improved blood flow and protection against ischaemia has also been advanced as a pathophysiological role for anandamide (Gebremedhin et al., 1999). In the context of cardiac ischaemia, Lagneux and Lamontagne (2001) reported that cardioprotection of the rat heart against ischaemia by pretreatment with lipopolysaccharide involved endocannabinoids. Subsequent work by that group also reported that palmitoylethanolamide and 2-arachidonoyl glycerol both caused cardioprotection via CB2 receptor activation (Lepicier et al., 2003).
The cardiovascular effects of cannabinoids and in particular the endocannabinoids are complex; their precise molecular targets are diverse and their relative contributions are uncertain. Furthermore, actions in isolated tissues do not necessarily translate to the whole animal situation. In vivo, the responses reported appear to be dependent on the experimental conditions, not least the use of general anaesthetics. However, much is to be gained by identifying the key targets; for example, can the vascular actions be best defined by considering novel cannabinoid receptors? Considering the cardiovascular actions of endocannabinoids, to what extent are they cannabinoids or vanilloids? These questions remain to be answered.
Source with Links and Visuals: The complexities of the cardiovascular actions of cannabinoids