Control Of Spasticity In A Multiple Sclerosis Model Is Mediated By CB1, Not CB2

Truth Seeker

New Member
Background and Purpose:

There is increasing evidence to suggest that cannabis can ameliorate muscle-spasticity in multiple sclerosis, as was objectively shown in experimental autoimmune encephalomyelitis models. The purpose of this study was to investigate further the involvement of CB1 and CB2 cannabinoid receptors in the control of experimental spasticity.

Experimental approach:

Spasticity was induced in wildtype and CB1-deficient mice following the development of relapsing, experimental autoimmune encephalomyelitis. Spastic-hindlimb stiffness was measured by the resistance to flexion against a strain gauge following the administration of CB1 and CB2 agonists.

Key Results:

As previously suggested, some CB2-selective agonists (RWJ400065) could inhibit spasticity. Importantly, however, the anti-spastic activity of RWJ400065 and the therapeutic effect of non-selective CB1/CB2 agonists (R(+)WIN55,212—2 and CP55, 940) was lost in spastic, CB1-deficit mice.

Conclusions and Implications:

The CB1 receptor controls spasticity and cross-reactivity to this receptor appears to account for the therapeutic action of some CB2 agonists. As cannabinoid-induced psychoactivity is also mediated by the CB1 receptor, it will be difficult to truly dissociate the therapeutic effects from the well-known, adverse effects of cannabinoids when using cannabis as a medicine. The lack of knowledge on the true diversity of the cannabinoid system coupled with the lack of total specificity of current cannabinoid reagents makes interpretation of in vivo results difficult, if using a purely pharmacological approach. Gene knockout technology provides an important tool in target validation and indicates that the CB1 receptor is the main cannabinoid target for an anti-spastic effect.

There has been recent interest in the therapeutic potential of cannabis for control for a number of symptoms, notably spasticity that often develops as a consequence of multiple sclerosis (MS. Consroe et al., 1997; Pertwee, 2002). Using cannabinoid agonists and antagonists, we were the first to provide objective, experimental evidence for the tonic control of spasticity by the cannabinoid system in the experimental autoimmune encephalomyelitis (EAE) model of MS (Baker et al., 2000, 2001). This supported patient claims for the use of medicinal cannabis (Consroe et al., 1997) and has been validated by the modest improvements of symptoms in more recent clinical trials of cannabinoids in MS (Zajicek et al., 2003, 2005; Vaney et al., 2004; Brady et al., 2004; Wade et al., 2004; Freeman et al., 2006). Although the exact cause of spasticity is not definitively known, it is clear that this results from alterations in the balance, possibly secondary to selective neuronal loss, between excitatory and inhibitory neural circuits (Brown, 1994; Dutta et al., 2006). This results in loss of control of neurotransmission between the muscles and the central nervous system resulting in uncontrolled spastic movements, which in some instances can be treated using GABA receptor agonists (Brown, 1994; Ivanhoe and Reistetter, 2004). After the initial observations in EAE (Baker et al., 2000), the cannabinoid type 1 (CB1) receptor and endocannabinoid system has been shown to regulate synaptic neurotransmission (Howlett et al., 2002; Wilson and Nicoll, 2002) and this action would be consistent with the cannabinoid control of spasticity. In contrast to CB1, there is limited evidence to indicate that normal nerve tissues express cannabinoid type 2 (CB2) receptors (Van Sickle et al., 2005) and they appear to be restricted to leucocytes (Munro et al., 1993; Galiegue et al., 1995; Howlett et al., 2002), although they are expressed by glial cells and may be upregulated in inflamed brain tissue (Maresz et al., 2005; Wotherspoon et al., 2005) and therefore may not be anticipated to control problems of neurotransmission. Surprisingly, however, a CB2 agonist ameliorated and an antagonist transiently worsened spasticity in EAE (Baker et al., 2000), suggesting that CB2 agonists could provide therapies that avoid the psychoactive effects associated with CB1 agonism (Baker et al., 2000; Howlett et al., 2002; Varvel et al., 2005). In animals, cannabimimetic potential is determined by activity in 'tetrad' (hypomotility, hypothermia, ring catelepsy analgesia) tests, which show no response owing to CB2 agonism (Howlett et al., 2002). However, currently there are no absolutely specific cannabinoid reagents (agonists or antagonists) available, which solely act on either of the CB1 or CB2 receptors, and although they may be selective to one or other of the cannabinoid receptors in vitro, at the doses used in vivo, there is the potential for cannabinoids to crossreact with the other CB receptor (Pertwee, 1999; Howlett et al., 2002). Furthermore, there is increasing evidence for additional receptors that mediate cannabimemetic effects (Hajos et al., 2001; Howlett et al., 2002; Begg et al., 2005; Baker et al., 2006), which further complicates the interpretation of pharmacological data. Therefore, receptor—deletion using transgenic technology (Zimmer et al., 1999, Brooks et al., 2002) provides a level of certainty of the role of the CB receptor subtype that is not provided by the CB receptor antagonism alone. This was used to re-evaluate the CB2-mediated control of spasticity during EAE.


Biozzi ABH and ABH mice lacking the CB1 receptor (Cnr1) gene were generated as described previously (Brooks et al., 2002, Pryce et al., 2003). The congenic ABH.Cnr1−/−used for breeding were produced by intercrossing after seven generations of backcrossing. Genomic screening for the absence of wild-type Cnr1, using polymerase chain reaction (Brooks et al., 2002), was performed on parental animals, and the functional deletion of CB1 protein was confirmed in animals used in these studies by the resistance to sedative doses (20mgkg−1 intraperitoneal (i.p.)) of R(+)WIN-55,212—2. These were from in-house bred stock that was maintained in a 12h-light/dark cycle with controlled humidity and temperature. Animals were fed RM-1E diet and water ad libitum. All animal studies conformed to the United Kingdom Animals (Scientific Procedures) Act 1986.


The full CB1/CB2 agonists R(+)WIN-55,212—2 ((R)-(+)-[2,3-dihydro-5-methyl-3-(-4-morpholinylmethyl)pyrrolo-[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthale-nylmethanone mesylate) (Ki CB1=9.9nM, Ki CB2=16.2nM (Rinaldi-Carmona et al., 1994)) and CP55,940 ((1R,3R,4R)-3-[2-hydroxy-4-(1,1-dimethylheptyl)phenyl]-4-(3-hydroxypropyl)cyclohexan-1-ol) (Ki CB1=1.4nM, Ki CB2=1.4nM (Rinaldi-Carmona et al., 1994)) were purchased from RBI/Sigma (Poole, UK) and Tocris Ltd (Bristol, UK). The CB2-selective agonist JWH056 (1-deoxy-δ-8-tetrahydrocannabinol; receptor affinity. Ki CB1=8770nM, Ki CB2=32nM) was provided by Dr J Huffman, Clemson University, South Carolina, USA (Huffman et al., 1996). The CB2-selective agonist RWJ400065 (binding affinity. Ki CB1=600nM, Ki CB2=10nM), forskolin-stimulated cyclic adenosine monophosphate (cAMP) agonism half-maximal inhibitory concentration (IC50) CB1=6600nM, IC50 CB2=6.6nM (Dr D Argentieri and Dr D Ritchie, unpublished observations) and non-selective CB1/CB2 agonist RWJ352303 (Ki CB1=0.6nM Ki CB2=0.3nM), forskolin-stimulated cAMP agonism in SKN cells IC50 CB1=0.64nM, IC50 CB2=0.14nm (Dr D Argentieri and Dr D Ritchie, unpublished observations) compounds were provided by RW Johnson] (Raritan, NJ, USA). These were suspended in intralipid 30% (Pharmacia, Milton Keynes, UK) before i.p. or intravenous (i.v.) injection in 0.1ml. Unanaesthetized animals were placed in a 87 × 112 × 60mm box (Alpha laboratories, Eastleigh, UK) that had a slot cut to allow exit of the tail from the box. The tail was transiently immersed in warm water (about 45°C) to induce vasodilatation, before injection into the tail vein using 30g needles. The RWJ compounds were initially injected i.v. to avoid effects of first-pass metabolism encountered with some compounds, such as tetrahydrocannabinol (Baker et al., 2000), using dose ranges based on pharmacokinetic data (Dr D Argentieri and Dr D Ritchie, unpublished observations). The i.p. route was selected for CP55,950 and R(+)WIN-55,212—2 as it was already known that these agents were rapidly active following i.p. delivery (Baker et al., 2000).

Induction of spasticity

Young adult mice were injected subcutaneously with 1mg of freeze-dried mouse spinal cord homogenate in Freund's adjuvant on days 0 and 7 to induce EAE as described previously (Baker et al., 1990). Animals developed relapsing-remitting episodes of limb paralysis and spasticity typically developed after 2—3 relapses, about 80—100 days postinduction. This was assessed during remission from active paralytic episodes by the force required to bend the hind limb to full flexion against a strain gauge (Baker et al., 2000).

Assessment of body temperature

Temperature was monitored using a thermocouple placed under the hindlimb as previously described (Brooks et al., 2002).

Data analysis and statistical procedures

Each group contained a minimum of five different animals and the results represent the mean±s.e. resistance to flexion force (N) or individual limbs, which were compared using repeated measures analysis of variance or paired t tests using Sigmastat software (Baker et al., 2001). Changes in temperature were compared using paired t tests using Sigmastat Software (Brooks et al., 2002).

In an attempt to validate our previous studies showing an anti-spastic activity of CB2 agonists (JWH133 receptor affinity. Ki CB1=680nM, CB2=3nM. Baker et al., 2000), additional compounds were investigated. Surprisingly, 10mgkg−1 i.v. JWH056, which is less potent at CB2, but with a lower affinity for CB1 (Ki>8μM) than JWH133, failed to inhibit spasticity at 10—60min after injection i.v. (Figure 1a), whereas RWJ352303, a potent non-selective CB1 agonist, inhibited spasticity (Figure 1a). However, a dose-dependent anti-spastic activity was detectable following injection i.v. of a potent CB2 agonist RWJ400065 (Figure 1b). This compound has similar binding affinities to JWH133 and failed to induce observable sedation (unpublished observations) and hypothermia (Figure 2), indicative of CB1 receptor-mediated effects. In contrast, RWJ352303 had the potential to induce 'tetrad-like' effects (Figure 2), but was still active as an anti-spastic agent (Figure 1a), at doses that did not induce 'tetrad-type' effects, shown here by hypothermic responses (Figure 2). However, when 10mgkg−1 i.v. RWJ400065 was injected into Cnr1−/− mice, there was no apparent anti-spastic activity (Figure 1B). To clarify this further, commonly used high-affinity CB1/CB2 non-selective agonists were examined. However, there was no evidence of inhibition of spasticity in CB1-deficient mice with either CP55,940 or R(+)WIN-55, 212—2 compared to significant (P<0.001) inhibitory activity in wild-type mice (Figure 3). This suggested that CB1 and not the CB2 receptors were actually mediating the inhibitory effects of some CB2 agonists.

Discussion and conclusions
Although this study confirms our previous observation (Baker et al., 2000) that 'tetrad inactive' apparent CB2 agonists can show anti-spastic activity, this does not appear to be owing to the direct activity of CB2 receptors. This most likely occurs because CB2 agonists/antagonists (Baker et al., 2000), or possibly their in vivo metabolites, have some affinity for CB1 receptors that may actually mediate the inhibitory effects. The biology of cannabis and the cannabinoid system now indicates that both tetrahydrocannabinol and CB1 receptors are the major mediators for both therapy in spasticity and also the adverse side effects (Howlett et al., 2002; Wilkinson et al., 2003; Varvel et al., 2005). It will be virtually impossible to truly dissociate these two effects, using cannabis. Clinical studies indicate that there is a substantial variability of individuals to tolerate cannabis and tetrahydrocannabinol (Zajicek et al., 2003; Brady et al., 2004; Wade et al., 2004). The apparent therapeutic window, before psychoactive effects, appears to be very small and is consistent with the modest effects in symptom control observed so far (Zajicek et al., 2003, 2005; Brady et al., 2004; Wade et al., 2004; Freeman et al., 2006), which nevertheless validate our original observations in animal models (Baker et al., 2000, 2001; Wilkinson et al., 2003). This variability of individuals to tolerate cannabinoids means that it will be difficult to dose—titrate adequately with potent CB1 agonists and that weak CB1 agonists, such as at the level found in some CB2 agonists, may be preferable for clinical use.

Currently there are two recognized cannabinoid receptors, but there is pharmacological evidence (Breivogel et al., 2001; Hajos et al., 2001; Howlett et al., 2002; Baker et al., 2006; Oz, 2006), some of which is disputed (Kawamura et al., 2006; Takahashi and Castillo 2006), for additional receptors or pathways that mediate cannabimimetic effects. Although the use of gene knockout technology is not without its own limitations, it provides an important tool in target validation. The loss of anti-spastic activity of R(+) WIN55,212—2 and CP55,940, both full CB1/CB2 agonists, in CB1-deficient mice supports the indication that CB1 and not CB2 is mediating the therapeutic anti-spastic effect. Nevertheless, anti-spastic control is feasible in CB1-deficient animals as shown previously with arvanil (Brooks et al., 2002). Arvanil, a potent transient receptor potential vanilloid type 1 (TRPV1) receptor and weak CB1 agonist, can also inhibit spasticity in the presence of CB1/CB2 antagonists and high doses of the TRPV1 antagonist capsazepine (Brooks et al., 2002). It can also induce cannabimemetic 'tetrad-type' responses, such as hypothermia, and hypomotility, in wild-type and Cnr1−/− mice (Brooks et al., 2002). However, capsazepine is a weak TRPV1 antagonist in mice (Correll et al., 2004) and the hypothermia and marked hypomotility induced by 0.5mgkg−1 i.v. arvanil is lost in Trpv1−/− mice (unpublished observations), further indicating the value of receptor knockout animals in target validation. However, cannabinoid receptors can exist as homodimers, and novel heterodimer formations between CB1 receptors and other G-protein-coupled receptors are assumed or are generated (Kearn et al., 2005; Wager-Miller et al., 2002; Rios et al., 2006). Therefore, CB1/CB2 receptor heterodimers or heterodimers between CB1 and any other molecule to which the CB2 agonists may bind would not exist in Cnr1−/− mice and this may have accounted for the loss of activity of RWJ400065 in CB1-deficient mice. Therefore, similar studies in Cnr2−/− mice are required to exclude definitively a role for CB2 in the control of spasticity.

However, the results from this study suggest that our previous data showing control of spasticity with endocannabinoid degradation inhibitors may need to be more cautiously interpreted (Baker et al., 2001). Many of these inhibitors, often based on the structural modifications of anandamide, have low affinity for CB1 receptors and are inactive in 'tetrad' tests, just as CB2-selective agonists appear to be. Compounds believed to inhibit the anandamide transporter, including AM404, VDM11 (Baker et al., 2001), OMDM-1, OMDM-2 (de Lago et al., 2004), UCM707 (de Lago et al., 2006) 0—2093 and 0—3246 (Ligresti et al., 2006) all exhibit anti-spastic activity. However, many of these agents have activity on additional molecules such as TRPV1 vanilloid receptors and the cannabinoid degrading enzyme: fatty acid amide hydrolase (FAAH), which could account for their biological activity (Ralevic et al., 2001; Fowler et al., 2004). Although a site for membranous diffusion of endocannabinoids has been suggested (Moore et al., 2005), the existence of a specific transporter for anandamide, independent of FAAH, has been questioned (Glaser et al., 2003; Ortega-Gutierrez et al., 2004; Kaczocha et al., 2006). Therefore, until the putative endocannabinoid transporter(s) are identified and cloned, it must be considered possible that the therapeutic, anti-spastic effect of cannabinoid re-uptake inhibitors may be explained by alternative mechanisms. However, using FAAH gene knockout mice (ABH.Faah−/−), we have been able to verify the activity of some FAAH inhibitors (Boger et al., 2000) as anti-spastic agents (unpublished observations). The lack of true understanding of diversity of the cannabinoid system and importantly the lack of absolute specificity of current cannabinoid agonists and antagonists (Pertwee, 1999) means that it may be difficult to correctly interpret results, particularly in vivo, if using a purely pharmacological approach.

Source, Graphs and Figures: Control of Spasticity in a Multiple Sclerosis Model is mediated by CB1, not CB2, Cannabinoid Receptors
Top Bottom