Cannabidiol inhibits synaptic transmission in rat hippocampal cultures and slices via multiple receptor pathways

Introduction

The Asian hemp plant, Cannabis sativa, has been used therapeutically for centuries but until recently most research has focused on Δ⁹-tetrahydrocannabinol (THC), the most abundant phytocannabinoid in cannabis. THC, acting via the cannabinoid CB₁ or CB₂ receptors (nomenclature follows Alexander et al., 2009), is thought to underlie most of the therapeutic uses of cannabis, which include use as an analgesic, an antiepileptic and an antiemetic (see Ben Amar, 2006; Turcotte et al., 2010). However, there remains controversy over the therapeutic use of cannabis because of the adverse psychotomimetic effects of THC and its potential to cause cognitive impairments (Fletcher and Honey, 2006; D'Souza et al., 2008). Nevertheless, cannabis contains numerous cannabinoids and these different constituents may modulate the effects of THC and have useful effects of their own. This has led to intense investigation into the therapeutic potential of other, non-psychoactive, constituents of cannabis (Russo and Guy, 2006; Pertwee, 2008; Izzo et al., 2009).

In particular cannabidiol (CBD), which lacks negative psychotropic effects when administered to humans even in high doses (Bhattacharyya et al., 2009), has emerged as an interesting compound due to its potential therapeutic application in a number of neurological and neuropsychiatric disorders including addiction, diabetic neuropathy, epilepsy, schizophrenia and stroke and it is currently used at a 1:1 ratio with THC in Sativex® for the treatment of chronic pain in patients suffering from multiple sclerosis and advanced cancer (see Russo and Guy, 2006; Izzo et al., 2009). Despite the fact that CBD is used clinically, the exact mechanism(s) by which it exerts its therapeutic action remains unclear, although several targets have been identified which could underlie its therapeutic actions in CNS disorders (see Russo and Guy, 2006; Pertwee, 2008 and Izzo et al., 2009). In contrast to THC, CBD has a low affinity for CB₁ receptors and acts as a CB₁ receptor antagonist and a CB₂ receptor inverse agonist (Showalter et al., 1996; Bisogno et al., 2001; Thomas et al., 2007). Furthermore, CBD has been proposed as an agonist at 5-HT_1A receptors (Russo et al., 2005) and this has been suggested to mediate the anxiolytic and anti-ischaemic properties of CBD (Campos and Guimarães, 2008; Resstel et al., 2009). CBD also acts either as an agonist or antagonist at transient receptor potential cation channels (TRPs; De Petrocellis et al., 2008; Qin et al., 2008) and has also been shown to inhibit the Ca_v3 subfamily of Ca²⁺ channels (Ross et al., 2008). In addition, CBD can modulate intracellular Ca²⁺ levels when tested on cultured hippocampal neurons, an effect proposed to be mediated via Ca_v1 Ca²⁺ channels, Ca²⁺ release from intracellular stores and the modulation of mitochondrial function (Drysdale et al., 2006; Ryan et al., 2009).

Despite increasing evidence to suggest that CBD has multiple targets (Pertwee, 2008; Izzo et al., 2009), the effect of CBD on basal synaptic transmission has yet to be determined although a recent study has demonstrated clear anti-epileptiform and antiseizure activity (Jones et al., 2010). In the present study, we have examined the effect of CBD upon synaptic activity in primary hippocampal cultures and acute hippocampal slices. Our data show for the first time that CBD is able to modulate basal hippocampal synaptic activity and we suggest a role for CB₁ and 5-HT_1A receptors in these observations. These data provide some mechanistic insights into the effects of CBD and highlight that further investigations into the actions of CBD in the CNS are required in order to fully elucidate the contribution of CBD to the detrimental and therapeutic roles of cannabis.

Methods

Results

Cannabidiol reduces spontaneous action potential frequency in cultured hippocampal neurones

Application of CBD reduced the synaptically driven spontaneous action potential (AP) frequency observed in primary hippocampal cultures in a concentration-dependent manner (0.1 µM, no significant change; 1 µM, 88 ± 8% decrease; 10 µM, 100% decrease; all n= 5, P < 0.001, Figure 1A, B). Confirmation of the synaptically driven nature of the spontaneous APs was shown by their sensitivity to the ionotropic glutamate receptor antagonists, DL-AP5 (100 µM) and NBQX (20 µM) (data not shown). In contrast, CBD was without effect on the resting membrane potential (V_m) at all concentrations tested (0.1 µM, 0.2 ± 2.1 mV from a control V_m of −60.7 ± 2.1 mV; 1 µM, 1.5 ± 1.6 mV from a control V_m of −57.6 ± 1.6 mV; 10 µM, 0.3 ± 0.7 from a control V_m of −62.0 ± 3.7 mV; all n= 5 and P > 0.05). Similar reductions in spontaneous AP frequency were observed following the application of the CB₁ agonist, WIN 55,212-2 (0.1 µM Figure 1B) and the CB₁ partial agonist, THC (10 µM, Figure 1B), as has been shown previously (Shen and Thayer, 1999; Bajo et al., 2009; Roloff and Thayer, 2009). To determine if the effects of CBD were mediated by a Gα_i/_o G-protein coupled receptor (GPCR), experiments were performed on cultures treated with Pertussis toxin (PTX, 200 ng·mL⁻¹, 18 h). The CBD-induced reduction in spontaneous AP frequency was abolished following PTX treatment, compared to CBD alone (1 µM, Figure 1B).

In order to examine whether the CBD-induced effects involved CB₁ receptor activation, we utilized the proposed CB₁ receptor antagonist AM251 and the structurally dissimilar CB₁ receptor inverse agonist LY320135. Application of AM251 (30 nM) resulted in an initial dramatic increase in AP firing rate, presumably due to the inhibition of the endogenous cannabinoid tone, that in the majority of cases led to a significant depolarization of the membrane potential and, as a consequence, to a loss of AP firing (Figure 2A). Thus the effect of AM251 on CBD-induced effects could not be investigated in our culture system. In contrast, application of LY320135 (1 µM) by itself had no significant effect on the AP firing frequency (14.9 ± 4.2% decrease, n= 7) but did significantly inhibit the decrease in AP firing frequency observed following CBD (1 µM) application, compared to CBD alone (Figure 2C). With CBD proposed as a 5-HT_1A receptor agonist, we examined in our culture preparation whether the 5-HT_1A receptor was involved in the CBD-mediated effects. Application of the 5-HT_1A receptor agonist, 8-hydroxy-2(di-N-propylamino)tetralin (8-OH-DPAT, 10 µM) significantly reduced the spontaneous AP frequency (Figure 2B, C) without affecting the V_m (1.2 ± 1.3 mV decrease from a control V_m of −59.4 ± 1.2, n= 5). The ability of 8-OH-DPAT to reduce spontaneous AP frequency was significantly inhibited in the presence of the 5-HT_1A receptor antagonist, WAY100135 (300 nM), compared with 8-OH-DPAT alone (Figure 2C). However, WAY100135 (300 nM) had no effect on CBD-mediated inhibition (1 µM) (Figure 2C), suggesting that in our culture system, the effects of CBD and 8-OH-DPAT were mutually exclusive.

Cannabidiol inhibits synaptic transmission in hippocampal slices

Having established the effect of CBD application in hippocampal cultures, we next investigated whether CBD modulates synaptic transmission in acute hippocampal slices. Application of CBD (10 µM) reversibly inhibited fEPSPs by 39 ± 13% (n= 6, P < 0.05, Figure 3A). The CBD-induced inhibitions were associated with a significant increase in the fEPSP paired-pulse facilitation ratio which increased from the control ratio of 1.65 ± 0.07 to 2.09 ± 0.13 in the presence of CBD (n= 6, P < 0.01, Figure 3B). This finding, aligned with the PTX sensitivity of the CBD effects observed in primary cultures, suggest that a presynaptic GPCR underlies the CBD-mediated inhibition of synaptic transmission. As in experiments in the culture preparation, we utilized the proposed CB₁ receptor antagonist, AM251 and the CB₁ receptor inverse agonist LY320135 in our slice preparation to investigate the involvement of CB₁ receptors. To our surprise, both AM251 (2 µM) and LY320135 (1 µM) abolished the inhibition of synaptic transmission by CBD (Figure 4A, E). Despite the lack of effect of the 5-HT_1A receptor antagonist WAY100135 on CBD-mediated effects in cultures, we also investigated the role of 5-HT_1A receptors in the CBD-mediated effects on synaptic transmission in the slice preparation. Strikingly, WAY100135 (300 nM) abolished the CBD mediated inhibition of synaptic transmission (Figure 4B, E). In addition and similar to its effect in cultures, 8-OH-DPAT (10 µM), like CBD, significantly inhibited basal synaptic transmission (Figure 4C, E). However, in contrast to CBD, it was without effect on the fEPSP paired-pulse facilitation ratio (control 1.71 ± 0.11, 8-OH-DPAT 1.73 ± 0.13). With evidence existing for AM251 acting via non-CB₁ receptor mechanisms (Pertwee, 2005), the specificity of the two putative CB₁ receptor inhibitors was investigated against 8-OH-DPAT-mediated inhibitions. Intriguingly, AM251 (2 µM) significantly inhibited the 8-OH-DPAT (10 µM)-induced depression (Figure 4D, E) whereas LY320135 (1 µM) was without effect (Figure 4E).

Discussion

The physiological actions of CBD, a purported non-psychoactive constituent of cannabis, have recently been an area of intense interest especially as it has been proposed to be beneficial in a number of clinical settings. In the present study, we show for the first time that CBD inhibits basal synaptic activity in both hippocampal cultures and acute slices.

Unlike THC, whose physiological effects are thought to be mediated predominantly through CB receptors, CBD has been shown to modulate a multitude of signalling mechanisms (Russo and Guy, 2006; Pertwee, 2008; Izzo et al., 2009). Utilizing in vitro CNS preparations, CBD has been shown to increase intracellular Ca²⁺ levels when investigated in cultured hippocampal neurons, an effect proposed to be mediated via Ca_v1 Ca²⁺ channels, Ca²⁺ release from intracellular stores and the modulation of mitochondrial function (Drysdale et al., 2006; Ryan et al., 2009). However, these studies were not extended to include the effect of CBD on synaptic activity although CBD did reduce intracellular Ca²⁺ levels under conditions of high neuronal excitability or epileptiform activity (Ryan et al., 2009), a finding supported by a recent study which shows that CBD has anti-epileptiform activity in hippocampal slice models of epilepsy (Jones et al., 2010). Our findings suggest a mechanism by which CBD inhibits synaptic activity in hippocampal culture preparations. We propose that CBD acts via a Gα_i/_o coupled GPCR as its effect was abolished following pre-incubation with the Gα_i/_o G-protein uncoupler, PTX. However, as spontaneous APs observed in our investigation are synaptically driven and with CBD decreasing intracellular Ca²⁺ levels in the neuronal soma under certain conditions (Ryan et al., 2009), a direct action of CBD on presynaptically located Ca²⁺ channels or their inhibition through the activation of presynaptic GPCRs cannot be unequivocally ruled out. Indeed, an inhibition of Ca_v3-type Ca²⁺ channels by CBD has been shown in heterologous expression systems and sensory neurones (Ross et al., 2008). However, as this Ca²⁺ channel subtype is not thought to be involved in neurotransmitter release within the hippocampus (Wheeler et al., 1994; Catterall, 2000), the action of CBD on this channel subtype probably does not underlie its effect in the current study. Moreover, modulation of intracellular stores and mitochondria has also been shown to affect synaptic transmission in CNS preparations (Billups and Forsythe, 2002; Fitzjohn and Collingridge, 2002; Le Magueresse and Cherubini, 2007); however, their potential role in the CBD-mediated effects observed here are beyond the focus of the present study.

As THC and WIN 55,212-2 both decreased synaptic activity in agreement with previous studies (Shen and Thayer, 1999; Ohno-Shosaku et al., 2005; Straiker and Mackie, 2005; Bajo et al., 2009), this indicates that the CBD-mediated effects are not an artefact of our culture preparation. This was further verified by observations that CBD inhibited basal synaptic transmission and increased the paired-pulse facilitation ratio in acute hippocampal slices. An increase in paired-pulse facilitation ratio is indicative that CBD is acting via a presynaptic receptor (Baskys and Malenka, 1991) and this finding, aligned to the sensitivity to PTX of the CBD-mediated actions in hippocampal cultures, suggested that the actions of CBD were mediated, at least in part, via a presynaptic Gα_i/_o coupled GPCR. Hence we investigated the potential role, or lack thereof, of CB₁ receptors in the CBD-mediated effects. To our initial surprise, we observed a decrease in the CBD-mediated effects in slices by the proposed CB₁ receptor antagonist AM251 and the structurally dissimilar CB₁ receptor inverse agonist LY320135. As mentioned earlier, CBD is proposed to have a low affinity for CB₁ receptors with no agonist activity and potentially act as a CB₁ receptor antagonist (Showalter et al., 1996; Bisogno et al., 2001; Thomas et al., 2007; Jones et al., 2010). It is therefore unlikely that the observed CBD effects are mediated through a direct activation of CB₁ receptors. However, an alternative explanation may account for the effects seen with AM251 and LY320135. As CBD has been shown to inhibit the reuptake and metabolism of the endocannabinoid anandamide (Rakhshan et al., 2000; Bisogno et al., 2001), the application of CBD could lead to an increase in the endocannabinoid tone and thus lead to a decrease in synaptic transmission via CB₁ receptor activation, as has been shown previously for exogenously applied endocannabinoids in culture (Straiker and Mackie, 2005; Hashimotodani et al., 2007) and in slices under conditions where degradation was inhibited (Bajo et al., 2009). It should be noted, however, that care needs to be taken when interpreting the results obtained using AM251 as this proposed selective CB₁ receptor antagonist abolished the effect of 8-OH-DPAT in the present study, indicating non-CB₁ receptor effects, as proposed previously (Pertwee, 2005). Despite this, the results obtained with LY320135 do indeed suggest a role for CB₁ receptors in the CBD-mediated effects. Taken together, these data imply the direct activation of 5-HT_1A receptors and the indirect activation of CB₁ receptors in the CBD-mediated effects. However, other signalling mechanisms previously shown to be modulated by CBD may also contribute to the observed effects and cannot be excluded.

An area of current interest regarding the actions of CBD is its proposed action as a putative agonist at the 5-HT receptor subtype, 5-HT_1A (Russo et al., 2005). Other studies also support the involvement of 5-HT_1A receptors in CBD-mediated effects as 5-HT_1A receptor antagonists have been shown to block the anxiolytic effects of CBD observed in animal models of anxiety, depression and stress (Campos and Guimarães, 2008; Resstel et al., 2009; Zanelati et al., 2010). In our experiments, WAY100135 abolished the inhibitory effects of CBD on basal synaptic transmission in acute slices implicating 5-HT_1A receptors in the CBD-mediated effects observed here. In agreement with this, activation of 5-HT_1A receptors with 8-OH-DPAT inhibited basal hippocampal synaptic transmission, as shown in previous studies (Schmitz et al., 1995; 1998; Pugliese et al., 1998). This raises the possibility that CBD could potentially modify hippocampal-dependent behaviours through the activation of 5-HT_1A receptors. However, the effects of 5-HT_1A receptor activation on synaptic transmission in acute slices have been proposed to be mediated via postsynaptic activation of G-protein-coupled inwardly rectifying K⁺ channels (Andrade and Nicoll, 1987; Lüscher et al., 1997) and the lack of effect of 8-OH-DPAT on fEPSP paired-pulse ratio in the current study is consistent with this. It should be noted that 8-OH-DPAT reduced spontaneous AP firing in cultures without affecting the membrane potential. 5-HT receptor-mediated K⁺ channel activation has been shown to run down in cultures in whole cell mode (Yakel et al., 1988) whilst differences between cultures and slices have been reported previously with regard to the K⁺ channel type involved following 5-HT receptor activation (Premkumar and Gage, 1994). As our culture experiments are in whole cell mode and are performed only when a stable firing frequency has been obtained, we suggest this is the explanation for the lack of effect on membrane potential seen with 8-OH-DPAT. However, further investigation into the disparity on the membrane potential effects between culture and slice preparations is beyond the scope of this study.

In conclusion, we have shown that CBD reduces synaptic activity in in vitro hippocampal preparations and propose a role for the direct activation of 5-HT_1A receptors along with an indirect activation of CB₁ receptors in these CBD-mediated effects. These data offer some clues into the mechanisms through which CBD may be reducing regional brain activity (Fusar-Poli et al., 2009; Bhattacharyya et al., 2010) and epileptiform activity (Jones et al., 2010) as reported in recent studies and suggest that further investigations into the actions of CBD in the CNS are required in order to fully elucidate its therapeutic potential.

Conflicts of interest

There are no conflicts of interest regarding the content of this manuscript.