ABC transporters P-gp and Bcrp do not limit the brain uptake of the novel antipsychotic and anticonvulsant drug cannabidiol in mice

Introduction

Cannabidiol (CBD), a non-psychoactive constituent of cannabis, displays much potential as a novel therapeutic treatment for various CNS disorders including schizophrenia and epilepsy (Bumb, Enning & Leweke, 2015; Iseger & Bossong, 2015; Longo, Friedman & Devinsky, 2015). CBD has anticonvulsant and antipsychotic effects in animal models of epilepsy and schizophrenia (Arnold, Boucher & Karl, 2012; Jones et al., 2012; Jones et al., 2010; Mao et al., 2015). Anecdotal reports and early clinical findings support CBD’s ability to reduce seizure rates in humans with a good safety profile (Rosenberg et al., 2015). Phase 3 clinical trials are currently investigating CBD in the treatment of epilepsy, most notably in severe and pharmacoresistant childhood epilepsies. There is also promising evidence that CBD is a novel antipsychotic, with a phase 2 clinical trial showing CBD reduces symptoms in schizophrenia patients with comparable efficacy to a conventional antipsychotic drug without producing extrapyramidal side-effects, sedation or weight gain (Bumb, Enning & Leweke, 2015; Iseger & Bossong, 2015; Leweke et al., 2012). Numerous further clinical trials are currently underway examining the efficacy of CBD in treating schizophrenia. The precise pharmacodynamic mechanisms responsible for CBD’s anticonvulsant and antipsychotic efficacy are hotly debated and may involve inhibition of degradation of the endocannabinoid anandamide. Although there are numerous other contenders as CBD is a promiscuous drug that interacts with multiple drug targets including G-protein-coupled receptor 55 (GPR55), transient receptor potential vanilloid type 1 (TRPV1) channels, and adenosine transporters (Leweke et al., 2012; McPartland et al., 2015; Rosenberg et al., 2015).

Resistance to treatment is a major stumbling block in the clinical management of epilepsy and schizophrenia. Approximately 30% of both schizophrenia and epilepsy patients do not respond adequately to drug therapy (Hoosain et al., 2015; Van Os & Kapur, 2009) and ATP-binding cassette (ABC) transporters play a role in treatment-resistance (Bebawy & Chetty, 2008; Brandt et al., 2006). ABC transporters are a large superfamily of proteins that actively transport substrates across biological membranes and thus influence the disposition of substrate drugs (Hee Choi & Yu, 2014; Kathawala et al., 2015). The best characterized ABC transporters are P-glycoprotein (P-gp, Abcb1) and breast cancer resistance protein (Bcrp, Abcg2), which are both efflux pumps localized at various pharmacological barriers in the body including the blood brain barrier (Löscher & Potschka, 2005). Many antipsychotic and anticonvulsant drugs are substrates of P-gp, which strongly limits the brain accumulation of these agents by extruding the drugs from the brain parenchyma back into the blood (Boulton et al., 2002; Doran et al., 2005; Luna-Tortós, Fedrowitz & Löscher, 2008; Zhang et al., 2010). Few studies have examined the substrate profile of CNS drugs for Bcrp, although lamotrigine has recently been demonstrated to be a substrate of human and mouse Bcrp (Römermann, Helmer & Löscher, 2015).

There is evidence that genetic variation in P-gp influences treatment response to antipsychotic and antiepileptic drugs such as olanzapine, risperidone, paliperidone (9-hydroxy risperidone, the active metabolite of risperidone), phenobarbital and phenytoin (French, 2013; Wolking et al., 2015). Furthermore, P-gp and Bcrp are upregulated at the BBB in epilepsy and schizophrenia and contribute to pharmacoresistance by limiting the brain uptake and efficacy of anticonvulsant and antipsychotic drugs (Aronica et al., 2005; Bauer et al., 2014; De Klerk et al., 2010; Lazarowski et al., 2007; Van Vliet et al., 2005). Drugs that are not substrates of ABC transporters will then make better therapeutics, as they will be immune to drug resistance mediated by these proteins. Indeed, it has become common practice in drug development to screen against ABC transporters to help isolate lead compounds that are less likely to fail in clinical trials due to suboptimal pharmacokinetic properties. It is therefore important to establish whether CBD is an ABC transporter substrate. We have shown that the main psychoactive constituent of cannabis, ${\mathrm{\Delta }}^9$-tetrahydrocannabinol (THC) is a substrate of both P-gp and Bcrp (Spiro et al., 2012). CBD, an isomer of THC, inhibits both P-gp and Bcrp transport (Feinshtein et al., 2013b; Holland et al., 2007; Zhu et al., 2006), although results have not been consistent for P-gp (Holland et al., 2006). Inhibitors are often substrates so there is a need to clarify whether CBD is a substrate of P-gp or Bcrp and whether this has implications for the brain uptake of the compound.

The present study assesses whether CBD is a substrate of P-gp and Bcrp by utilising mice devoid of these ABC transporter genes singly or in combination. If CBD accumulates at greater concentrations in the brain of ABC transporter knockout animals than wild-type mice then this provides evidence that CBD is an ABC transporter substrate. The results of CBD will be compared with the positive controls risperidone and 9-hydroxy risperidone, as these antipsychotic drugs are established P-gp substrates (Doran et al., 2005). Our results will be useful in the evaluation of CBD as a therapeutic agent for CNS disorders from a pharmacokinetic perspective. If P-gp or Bcrp do not limit the brain concentrations of CBD, it implies that CBD might be free from drug resistance that is mediated by these ABC transporters.

Materials and Methods

Drug treatment

CBD (THC Pharm, Frankfurt, Germany) was dissolved in a mixture of ethanol, Tween 80, and saline (1:1:18) (Long et al., 2013; Todd & Arnold, 2016) and administered via subcutaneous (s.c) injection at a dose of 10 mg/kg (Doran et al., 2005; Pacchioni et al., 2010). Risperidone (Sequoia Pharmaceuticals, Gaithersburg, MD, USA) was dissolved in a solution of 0.9% saline and 1% acetic acid and injected s.c. at 3 mg/kg. All drugs were freshly prepared before use and made at an injection volume of 10 ml/kg of body weight. At numerous time-points post-injection of CBD (1, 2 and 3 h) and risperidone (1 and 3 h), P-gp knockout, Bcrp knockout, P-gp/Bcrp knockout and WT mice were lightly anesthetised with isoflurane and blood collected via cardiac puncture. Blood samples were stored in ethylenediaminetetraacetic acid (EDTA) coated tubes to avoid coagulation and kept on ice before separation of plasma (Spiro et al., 2012). To separate the plasma from the blood, samples were centrifuged at 3,000 rpm for 10 min at 4 °C and the plasma collected in clean eppendorf tubes (Wang et al., 2004). The brains were immediately extracted and snap frozen in liquid nitrogen. Both the brain and plasma samples were stored at –80 °C before LC-MS/MS analysis.

Results

P-gp or Bcrp knockout did not increase brain or plasma concentrations of CBD

Representative chromatogram traces of CBD and the internal standard CBD-D3 are shown in Fig. 1A. Concentrations of CBD in brain and plasma (Fig. 2) were not altered in P-gp, Bcrp or P-gp/Bcrp knockout mice compared to WT mice as supported by overall two-way ANOVA with no main effect of genotype or a genotype by time interaction. A significant main effect of time was observed in both brain (F(3, 60) = 30.5, P < 0.0001) and plasma (F(2, 63) = 34.8, P < 0.0001) samples, with CBD concentrations decreasing over the three hour time period in all genotypes. Tukey’s post-hoc comparisons confirmed that no significant differences were observed between WT when individually compared to P-gp knockout, Bcrp knockout or P-gp/Bcrp knockout mice at 1, 2 or 3 h time-points. Nor were there any significant differences between P-gp knockout and P-gp/Bcrp knockout mice at any time-point.

P-gp or Bcrp knockout did not influence CBD brain/plasma ratios, whereas P-gp knockout profoundly increased risperidone and 9-hydroxy risperidone brain/plasma ratios

No differences were observed between P-gp, Bcrp or P-gp/Bcrp knockout and WT mice (Fig. 3A) in their brain/plasma CBD concentration ratios (with overall a ratio of approximately 2), as two-way ANOVA showed no main effect of genotype and no genotype by time interaction. This was confirmed by post-hoc comparisons. As CBD did not show any altered disposition in ABC transporter knockout mice, we sought to show in our hands that other known transporter substrates, that is risperidone and 9-OH risperidone, display increased brain uptake in knockout animals. This would prove that our negative finding with CBD was not a false negative due to some procedural error.

Representative chromatogram traces of risperidone and 9-OH risperidone are shown in Fig. 1B. Two-way ANOVA indicated a significant overall main effect of genotype in the risperidone and 9-hydroxy risperidone brain/plasma concentration ratios (F(3, 39) = 34.62, P < 0.0001; F(3, 39) = 32.17, P < 0.0001 respectively) (Figs. 3B and 3C). The brain/plasma ratio for risperidone and 9-hydroxy risperidone also tended to increase over time in all genotypes (main effect of time: F(1, 39) = 5.66, P < 0.05; P(1, 39) = 5.65, P < 0.05 respectively). Tukey’s post-hoc analysis showed P-gp knockout and P-gp/Bcrp knockout had significantly greater brain/plasma risperidone concentration ratios than WT mice at both 1 h and 3 h time points respectively (Ps < 0.05). In the P-gp knockout and P-gp/Bcrp knockout mice the brain/plasma risperidone ratios reached as high as 13 whereas the ratio for WT mice was less than 1. Brain/plasma concentration ratios of 9-hydroxy risperidone in P-gp knockout mice and P-gp/Bcrp knockout mice were significantly higher than the WT mice ratio at the 3 h time point (Ps < 0.05). In the P-gp knockout and P-gp/Bcrp knockout mice the brain/plasma 9-hydroxy risperidone ratio reached as high as 8 whereas the ratio for WT mice was less than 1. P-gp/Bcrp knockout mice achieved a significantly higher brain/plasma concentration ratio for 9-hydroxy risperidone than WT mice at 1 h, however P-gp knockout mice failed to reach significance when compared to WT mice at this timepoint. Although, P-gp knockout and P-gp/Bcrp knockout mice were not statistically different in their brain/plasma concentration ratios for risperidone and 9-hydroxy risperidone at either the 1 or 3 h timepoints implying no cooperation between P-gp and Bcrp in the transport of these drugs. Bcrp knockout mice did not display significantly different brain/plasma concentrations of risperidone and 9-hydroxy risperidone to WT at either timepoint.

Discussion

This study shows that the ABC transporters P-gp and Bcrp do not influence the brain uptake of CBD, a novel antipsychotic and anticonvulsant drug. P-gp, Bcrp and P-gp/Bcrp knockout mice did not accumulate greater brain or plasma concentrations of CBD compared to WT mice, nor were the brain/plasma concentration ratios of CBD influenced by knockout of the ABC transporter genes. By comparison, the known substrates of P-gp risperidone and 9-hydroxy risperidone (Doran et al., 2005; Ejsing, Pedersen & Linnet, 2005; Kirschbaum et al., 2008; Wang et al., 2004), displayed increased brain/plasma concentration ratios in P-gp knockout mice compared to WT mice, reaching as high as approximately 13 and 8 times respectively, whereas WT mice only attained ratios <1. Taken together these results suggest P-gp strongly regulates the brain uptake of the antipsychotic drugs risperidone and 9-hydroxy risperidone but not CBD.

That P-gp and Bcrp did not influence the brain uptake of CBD is somewhat surprising given our prior research showing these transporters regulate the brain concentrations of THC, the main psychoactive constituent of cannabis (Spiro et al., 2012). CBD and THC are isomers with very similar lipophilicity. However, CBD is formed when the central pyran ring of THC is opened and the oxygen in the ring is converted into a free hydroxy group (Compton et al., 1992). This subtle chemical modification yields remarkable differences in the pharmacological activity of these drugs. For instance, THC is a partial agonist at CB₁ cannabinoid receptors and therefore elicits profound psychotropic effects, whilst CBD has poor affinity for the orthosteric site of this receptor and doesn’t have appreciable psychoactivity (Laprairie et al., 2015; Long et al., 2010). CBD has a growing list of distinctive properties to THC. For example, CBD inhibits the anandamide degradative enzyme fatty acid amid hydrolase (FAAH) and activates TRPV1 receptors, actions not shared with THC (McPartland et al., 2015; Pertwee, 2008). Our data here provides yet another example, this time of a differential substrate binding character for ABC transporter proteins.

Our finding that CBD is not a substrate of murine P-gp and Bcrp is consistent with in vitro data with cells expressing human transporters. CBD did not stimulate ATPase activity in insect membranes expressing human P-gp and Bcrp, unlike known P-gp and Bcrp substrates verapamil and sulphasalazine respectively (Holland et al., 2007; Zhu et al., 2006). Therefore it appears that our results here may generalize to human transporters, although future studies assessing CBD transport by human ABC transporters using transwell assays would strengthen this viewpoint. We and others have shown that both CBD and THC inhibit P-gp and Bcrp (Feinshtein et al., 2013a; Feinshtein et al., 2013b; Holland et al., 2007; Zhu et al., 2006). While THC appears to be a competitive substrate, as it is actively transported by P-gp and Bcrp (Bonhomme-Faivre et al., 2008; Spiro et al., 2012), CBD’s ability to inhibit the transporters occurs in the absence of active transport. This phenomenon is not without precedent as paracetamol and haloperidol both inhibit P-gp but are not actively transported as substrates (Feng et al., 2008; Novak et al., 2013) and gefitinib inhibits Bcrp while not being a substrate (Galetti et al., 2015).

Future studies are needed to examine whether CBD’s ability to inhibit ABC transporters alters the pharmacokinetics of co-administered drugs that are ABC transporter substrates like risperidone. Interestingly co-administration of CBD with clobazam in children with refractory epilepsy increased plasma concentrations of clobazam and its active metabolite norclobazam by 60 and 500% respectively (Geffrey et al., 2015). Whether norclozabam is an ABC transporter substrate is unknown; however, clobazam is a duel P-gp/Bcrp substrate, opening the possibility that these transporters may contribute to this drug interaction (Nakanishi et al., 2013). In addition, CBD inhibition of ABC transporters appears relevant to pharmacokinetic interactions between CBD and THC, as THC is a dual P-gp and Bcrp substrate, and CBD potentiates some of the effects of THC via increasing brain THC concentrations (Klein et al., 2011; Spiro et al., 2012; Todd & Arnold, 2016). Extended exposure to CBD has also been shown to influence the expression of P-gp in cancer cells that is mediated by CB2 and TRPV1 receptors (Arnold et al., 2012; Holland et al., 2006). Thus a future study assessing the impact of chronic CBD on P-gp expression in brain microvessels is warranted, as CBD-induced upregulation of P-gp may impact upon the pharmacokinetics of co-administered substrate drugs. Another limitation of the current study is that CBD metabolites were not examined, as no analytical standards for these compounds were commercially available. One metabolite of particular interest is 7-OH CBD which appears to be an active anticonvulsant with greater potency than CBD (Jiang et al., 2011; Stott et al., 2015; Ujváry & Hanuš, 2016). Future studies need to address whether CBD metabolites like 7-OH CBD are ABC transporter substrates.

Our results suggest that risperidone and 9-hydroxy risperidone are not Bcrp substrates and that there is no cooperation between P-gp and Bcrp in the transport of these antipsychotic drugs. If these drugs were Bcrp substrates then the Bcrp knockout mice would have displayed greater brain and plasma concentrations of these agents than WT mice. Further, if these antipsychotic drugs were dual substrates of P-gp and Bcrp we would have expected the double P-gp/Bcrp knockout mice to display greater brain and plasma concentrations than P-gp knockout mice alone. Such observations have been made for other drugs, for example the Bcrp substrates prazosin and mitoxantrone display greater brain or plasma concentrations in Bcrp knockout mice than WT mice (Cisternino et al., 2004) and the dual P-gp and Bcrp substrates sunitinib and dasatinib show greater brain concentrations in P-gp/Bcrp knockout mice than P-gp or Bcrp knockout mice alone (Tang et al., 2013).

Cannabidiol is currently being assessed in randomized controlled trials as a novel antipsychotic and anticonvulsant agent, supported by an array of preclinical and human data (Arnold, Boucher & Karl, 2012; Devinsky et al., 2014; Leweke et al., 2012). ABC transporters may play an important role in pharmacoresistance, which is a major stumbling block in the successful treatment of schizophrenia and epilepsy. Indeed the ABC transporter substrate binding character is routinely assessed in the development of novel CNS therapeutics to ensure adequate brain uptake of the drug e.g., blonanserin (Inoue et al., 2012). As can be seen here, the brain uptake of the commonly used antipsychotic drug risperidone and its active metabolite 9-hydroxy risperidone is profoundly limited by P-gp. Similarly many anticonvulsant drugs such as phenytoin, phenobarbital and clobazam are also ABC transporter substrates and subject to poor brain uptake (Nakanishi et al., 2013; Zhang et al., 2010). Moreover, single nucleotide polymorphisms (SNPs) in MDR1 increase the risk of resistance or greater interindividual response to antiepileptic and antipsychotic drugs (Bozina et al., 2008; Li et al., 2014; Shaheen et al., 2014; Vijayan et al., 2012). Our data support that CBD may be free from the complication of reduced brain uptake or varied interindividual response to drug therapy, at least that is mediated by the ABC transporters P-gp and Bcrp. These findings provide evidence for the favourable pharmacokinetic properties of CBD in the treatment of CNS disorders and help build the case for the development of CBD as a therapeutic agent.

Supplemental Information