ArticlePDF Available

Abstract

Cannabis has a long history of anecdotal medicinal use and limited licensed medicinal use. Until recently, alleged clinical effects from anecdotal reports and the use of licensed cannabinoid medicines are most likely mediated by tetrahydrocannabinol by virtue of: 1) this cannabinoid being present in the most significant quantities in these preparations; and b) the proportion:potency relationship between tetrahydrocannabinol and other plant cannabinoids derived from cannabis. However, there has recently been considerable interest in the therapeutic potential for the plant cannabinoid, cannabidiol (CBD), in neurological disorders but the current evidence suggests that CBD does not directly interact with the endocannabinoid system except in vitro at supraphysiological concentrations. Thus, as further evidence for CBD's beneficial effects in neurological disease emerges, there remains an urgent need to establish the molecular targets through which it exerts its therapeutic effects. Here, we conducted a systematic search of the extant literature for original articles describing the molecular pharmacology of CBD. We critically appraised the results for the validity of the molecular targets proposed. Thereafter, we considered whether the molecular targets of CBD identified hold therapeutic potential in relevant neurological diseases. The molecular targets identified include numerous classical ion channels, receptors, transporters, and enzymes. Some CBD effects at these targets in in vitro assays only manifest at high concentrations, which may be difficult to achieve in vivo, particularly given CBD's relatively poor bioavailability. Moreover, several targets were asserted through experimental designs that demonstrate only correlation with a given target rather than a causal proof. When the molecular targets of CBD that were physiologically plausible were considered for their potential for exploitation in neurological therapeutics, the results were variable. In some cases, the targets identified had little or no established link to the diseases considered. In others, molecular targets of CBD were entirely consistent with those already actively exploited in relevant, clinically used, neurological treatments. Finally, CBD was found to act upon a number of targets that are linked to neurological therapeutics but that its actions were not consistent withmodulation of such targets that would derive a therapeutically beneficial outcome. Overall, we find that while >65 discrete molecular targets have been reported in the literature for CBD, a relatively limited number represent plausible targets for the drug's action in neurological disorders when judged by the criteria we set. We conclude that CBD is very unlikely to exert effects in neurological diseases through modulation of the endocannabinoid system. Moreover, a number of other molecular targets of CBD reported in the literature are unlikely to be of relevance owing to effects only being observed at supraphysiological concentrations. Of interest and after excluding unlikely and implausible targets, the remaining molecular targets of CBD with plausible evidence for involvement in therapeutic effects in neurological disorders (e.g., voltage-dependent anion channel 1, G protein-coupled receptor 55, CaV3.x, etc.) are associated with either the regulation of, or responses to changes in, intracellular calcium levels. While no causal proof yet exists for CBD's effects at these targets, they represent the most probable for such investigations and should be prioritized in further studies of CBD's therapeutic mechanism of action.
1 23
Neurotherapeutics
The Journal of the American Society for
Experimental NeuroTherapeutics
ISSN 1933-7213
Neurotherapeutics
DOI 10.1007/s13311-015-0377-3
Molecular Targets of Cannabidiol in
Neurological Disorders
Clementino Ibeas Bih, Tong Chen,
Alistair V.W.Nunn, Michaël Bazelot,
Mark Dallas & Benjamin J.Whalley
1 23
Your article is protected by copyright and all
rights are held exclusively by The American
Society for Experimental NeuroTherapeutics,
Inc.. This e-offprint is for personal use only
and shall not be self-archived in electronic
repositories. If you wish to self-archive your
article, please use the accepted manuscript
version for posting on your own website. You
may further deposit the accepted manuscript
version in any repository, provided it is only
made publicly available 12 months after
official publication or later and provided
acknowledgement is given to the original
source of publication and a link is inserted
to the published article on Springer's
website. The link must be accompanied by
the following text: "The final publication is
available at link.springer.com”.
REVIEW
Molecular Targets of Cannabidiol in Neurological Disorders
Clementino Ibeas Bih
1
&Tong Chen
1
&Alistair V. W. Nunn
2
&Michaël Bazelot
1,3
&
Mark Dallas
1
&Benjamin J. Whalley
1
#The American Society for Experimental NeuroTherapeutics, Inc. 2015
Abstract Cannabis has a long history of anecdotal medicinal use
and limited licensed medicinal use. Until recently, alleged clinical
effects from anecdotal reports and the use of licensed cannabinoid
medicines are most likely mediated by tetrahydrocannabinol by
virtue of: 1) this cannabinoid being present in the most significant
quantities in these preparations; and b) the proportion:potency
relationship between tetrahydrocannabinol and other plant canna-
binoids derived from cannabis. However, there has recently been
considerable interest in the therapeutic potential for the plant can-
nabinoid, cannabidiol (CBD), in neurological disorders but the
current evidence suggests that CBD does not directly interact with
the endocannabinoid system except in vitro at supraphysiological
concentrations. Thus, as further evidence for CBDs beneficial
effects in neurological disease emerges, there remains an urgent
need to establish the molecular targets through which it exerts its
therapeutic effects. Here, we conducted a systematic search of the
extant literature for original articles describing the molecular phar-
macology of CBD. We critically appraised the results for the
validity of the molecular targets proposed. Thereafter, we consid-
ered whether the molecular targets of CBD identified hold thera-
peutic potential in relevant neurological diseases. The molecular
targets identified include numerous classical ion channels, recep-
tors, transporters, and enzymes. Some CBD effects at these targets
in in vitro assays only manifest at high concentrations, which may
be difficult to achieve in vivo, particularly given CBDs relatively
poor bioavailability. Moreover, several targets were asserted
through experimental designs that demonstrate only correlation
with a given target rather than a causal proof. When the molecular
targets of CBD that were physiologically plausible were consid-
ered for their potential for exploitation in neurological therapeu-
tics, the results were variable. In some cases, the targets identified
had little or no established link to the diseases considered. In
others, molecular targets of CBD were entirely consistent with
those already actively exploited in relevant, clinically used, neu-
rological treatments. Finally, CBD was found to act upon a num-
ber of targets that are linked to neurological therapeutics but that
its actions were not consistent withmodulation of such targets that
would derive a therapeutically beneficial outcome. Overall, we
find that while >65 discrete molecular targets have been reported
in the literature for CBD, a relatively limited number represent
plausible targets for the drugs action in neurological disorders
when judged by the criteria we set. We conclude that CBD is very
unlikely to exert effects in neurological diseases through modula-
tion of the endocannabinoid system. Moreover, a number of other
molecular targets of CBD reported in the literature are unlikely to
be of relevance owing to effects only being observed at
supraphysiological concentrations. Of interest and after excluding
unlikely and implausible targets, the remaining molecular targets
of CBD with plausible evidence for involvement in therapeutic
effects in neurological disorders (e.g., voltage-dependent anion
channel 1, G protein-coupled receptor 55, CaV3.x, etc.) are asso-
ciated with either the regulation of, or responses to changes in,
intracellular calcium levels. While no causal proof yet exists for
CBDs effects at these targets, they represent the most probable
for such investigations and should be prioritized in further studies
of CBDs therapeutic mechanism of action.
Key Words: Cannabidiol .neurological disorders .
mechanism of action .cannabinoid .pharmacology.
*Benjamin J. Whalley
b.j.whalley@reading.ac.uk
1
School of Chemistry, Food and Nutritional Sciences, and Pharmacy,
University of Reading, Whiteknights, Reading RG6 6AP, UK
2
Broadmind Science Ltd, 47 Bow Field, Hook, Hants RG27 9SA, UK
3
GW Pharmaceuticals Ltd, Sovereign House, Vision Park, Chivers
Way, Histon, Cambridge CB24 9BZ, UK
Neurotherapeutics
DOI 10.1007/s13311-015-0377-3
Author's personal copy
Introduction
For millennia, mankind has associated the use of Cannabis
sativa and its ~100 constituent phytocannabinoids (plant can-
nabinoids) with therapeutic usefulness, including for neuro-
logical disorders such as convulsions and pain [19].
However, such traditional use does not constitute valid evi-
dence for the modern medical use of the plant, its extracts or
components, and licensed clinical use of cannabis-based med-
icines remains limited to a small number of disorders such as
pain in multiple sclerosis (MS), appetite stimulation in HIV/
AIDS, and cancer chemotherapy [1012]. Moreover, the ther-
apeutic effects of currently licensed cannabis-based treatments
rely for the most part upon the pharmacological effects of the
principal psychoactive component derived from the C. sativa
plant, Δ
9
-tetrahydrocannabinol (Δ
9
-THC) [13].
More recently, the therapeutic potential of the typically
second most abundant phytocannabinoid, cannabidiol
(CBD), has been investigated in preclinical animal models
and, together with anecdotal and often ambiguous reports of
crude cannabis extracts containing high proportions of CBD
exerting beneficial effects in treatment-resistant pediatric epi-
lepsies [14], has led to formal human clinical trials of CBD in
a number of epilepsies that will report results in late 2015 [15].
However, despite an extensive preclinical evidence base sug-
gesting therapeutic utility for CBD in several neurological
disorders (reviewed in this issue), a proven lack of cannabi-
noid type 1 receptor (CB
1
R)-mediated psychoactivity [16],
and good tolerability, the specific molecular target(s) through
which CBD exerts its reported therapeutic effects remains
undetermined.
Here, we summarize and assess the current evidence for
CBD exerting plausible pharmacological effects via specific
molecular targets (Part 1) before considering separate evi-
dence of the extent to which these targets may be involved
in mediating therapeutic effects in a variety of neurological
disorders (Part 2) that were selected to complement the re-
views also presented in this issue.
Methods
In order to identify molecular targets of CBD within the
existing literature, a PubMed search, using filters that exclud-
ed review articles, for the term Bcannabidiol^was performed.
Results were thenmanually reviewed to determine whether or
not original results describing CBD effects in molecular target
specific assays were presented. Thus, only peer-reviewed,
original publications that included results from assays specific
to a given molecular target were included in this review.
Part 1 also considers the plausibility of a given molecular
target as having the potential to play a role in CBDs therapeutic
effects by virtue of potency and efficacy information where
available. With regard to potency, as most studies reviewed
relied upon evidence derived from preclinical animal models
of disease, we selected a specific concentration beyond which
effects reported by in vitro studies cannot realistically be
achieved in vivo. This guideline was based upon data derived
from a detailed report of CBDs plasma and brain pharmacoki-
netic profile following administration via a number of routes
(orally and intraperitoneally) in both mouse and rat [17]. Thus,
in vitro effects reported in Part 1 that require CBD concentra-
tions 1020 μM were considered supraphysiological and so
such molecular targets were not considered in Part 2.
Part 1: Molecular Targets of CBD
Receptor Targets
Our review of the current literature revealed 11 investigations
of the effect of CBD upon 10 specific receptor targets, which
account for 15% of the known molecular targets of CBD
(Fig. 1, Table 1).
Cannabinoid Receptors
In contrast to Δ
9
-THC, CBD has a very low affinity and
shows little agonist activity at the G protein-coupled
endocannabinoid system (ECS) receptors, CB
1
RandCB
2
R
[29]. However, despite this micromolar affinity, some of the
literature reports CBD as having an antagonistic profile
against CB
1
R/CB
2
R agonists with a nanomolar K
B
[30].
More recently, a statistical meta-analysis of all extant data
describing direct effects of CBD at CB
1
RandCB
2
Rconclud-
ed that there is no direct CBDCB
1
R interaction that can ac-
count for the reported changes in endocannabinoid signaling
[16]. Indeed, the pharmacology of CBD at cannabinoid recep-
tors is not only complex and highly variable [3133], but also
typically occurs at supraphysiological concentrations in vitro,
so rendering any contribution to behavioral effects unlikely.
These clear discrepancies between in vitro and in vivo studies
are a warning to pharmacologists that predicted pharmacology
from molecular assay systems is not always replicated at a
system level.
Adenosine
CBD not only elicits effects within the central nervous system
(CNS), but also within the cardiovascular system. Adenosine
receptors have been implicated in regulating coronary blood
flow and oxygen consumption by cardiac muscle and are
present in the brain, most notably in the forebrain [34,35].
CBD (50 μg/kg) inhibits the subsequent ventricular tachycar-
dia following coronary artery occlusion in ratsan effect
abolished by 8-cyclopentyl-1,3-dipropylxanthine, an
Bih et al.
Author's personal copy
adenosine A
1
receptor antagonist (Table 1). These results
demonstrate that CBD can exert an antiarrhythmic effect, pos-
sibly mediated by the adenosine A
1
receptor [36]. In addition
to effects on the A
1
receptor, A
2
receptor-mediated effects of
CBD have also been reported and claimed to mediate anti-
inflammatory effects of CBD [10,31,37,38]. The effects of
CBD on inflammatory signaling cascades have also been ex-
amined (Table 1)[37]. Here, retinal tumor necrosis factor-α
secretion following lipopolysaccharide injection was sup-
pressed by CBD (1 mg/kg), and the pharmacology of this
response was further dissected using ZM 241385, an adeno-
sine A
2
antagonist. The authors also suggested that the ob-
served CBD effects could arise though modulation of adeno-
sine transport (see BTransporters^). Using in vitro and in vivo
models of MS demonstrated that an adenosine A
2
receptor-
mediated component of the inflammatory response was sup-
pressed by CBD (in vivo:5mg/kg;in vitro:1μM) [10].
However, it was noted that the adenosine A
2
-mediated re-
sponse was only partially responsible for the anti-
inflammatory observations, again through use of the adeno-
sine A
2
antagonist ZM241385. Studies have suggested that
neuroprotective effects of CBD are mediated via adenosine
A
2
receptor modulation [31]. However, this has also been
contested where species differences and animal developmen-
tal stages in the methodology of these conflicting studies may
underlie the different conclusions drawn [38]. Therefore, a
clear link between the reported neuroprotective effects of
CBD and adenosine A
2
receptors has not yet been shown.
Glycine Receptors
Another molecular target of CBD has been revealed by 2
comprehensive studies investigating glycine-mediated
synaptic transmission [21,22]. Pentameric glycine ionotropic
receptors (GlyR) mediate neuropathic pain and inflammation
through Cl
flux where, at present, 4 αsubunit (α
14
)isoforms
and 1 isoform of the βsubunit have been described [39].
Using in vitro electrophysiological assays, CBD (1300
μM) modulated strychnine-sensitive α1andα1βGlyR. At
concentrations > 100 μM, direct activation of these GlyR
was observed [21], although the physiological pertinence of
an effect at such a high concentration remains unknown.
Subsequently, it was shown that in HEK293 cells expressing
α3GlyRs,1μM CBD increased glycine-induced current am-
plitude almost 5-fold in comparison with glycine alone
(Table 1)[22]. Of interest was the observation that
dehydroxyl-CBD (1μM) was more efficacious than CBD as
it initiated a 9-fold increase in current amplitude. This obser-
vation led to the more recent study highlighting a role for
dehydroxyl-CBD in modulating presynaptic GlyR and sug-
gested a homomeric conformational bias for dehydroxyl-
CBD effects [40].
Opioid Receptors
Opioid receptors (ORs) are G
i/o
protein-coupled receptors that
bind opiates [41]. CBD (0.1100.0 μM) may serve as an al-
losteric modulator at μand δORs (Table 1)[28]. In rat cortical
membranes, disassociation of OR agonist [D-Ala
2
,N-
MePhe
4
,Gly-ol]-encephalinfromμORs was seen following
the application of CBD. A similar effect was observed for δ
ORs. For both receptor isoforms, half maximal inhibition was
produced at approximately 10 μM(Table1). However, the
inhibitory constant varied between subtypes [18.4 μM(δiso-
form) and 31.6 μM(μisoform)] [28].
Fig. 1 Pie chart showing the
proportions of different molecular
targets for cannabidiol described
in the reviewed literature. Chart
shows percentage proportions
from a total of 65 targets. Targets
counted were unique and not
counted per literature report
Molecular Targets of Cannabidiol
Author's personal copy
Serotonin Receptors
CBD may also act via serotonin (5-HT) receptors [26,4245].
The 5-HT
1A
receptor is coupled to G protein G
i/o,
where it is
thought to mediate inhibitory neurotransmission [46]. CBD
binds to the 5-HT
1A
receptor within a range of 832 μM
(Table 1)[26]. In cultured Chinese hamster ovary cells, CBD
displaced radiolabelled 8-OH-DPAT [7-(dipropylamino)-5,6,7,
8-tetrahydronaphthalen-1-ol] from 5-HT
1A
receptors. An in-
crease in guanosine 5O-[gamma-thio]triphosphate binding to
G
i/o
and a fall in cyclic adenosine monophosphate was seen,
which suggests that CBD can act as an agonist at the 5-HT
1A
receptor. In addition to 5-HT
1A
agonism, the same study pre-
sented evidence for CBD binding to the 5-HT
2A
receptor. In
NIH/3T3 cells, CBD (832 μM) served as a partial agonist of
the 5-HT
2A
receptor with weaker efficacy than its action at the
5-HT
1A
receptor [26]. 5-HT receptors are also implicated in
autonomic control [47], and again here CBD has been shown
to modulate 5-HT
1A
-mediated responses in vivo [42]. CBD (1
20 mg/kg i.p.) attenuated the stress response in male rats, as
evidenced by cardiovascular parameters, and this attenuation
was masked by co-administration of WAY-1000635, a 5-
HT
1A
antagonist (although with nonspecific effects at dopa-
mine D
4
receptors) [48]. Further studies examining central con-
trol of cardiovascular function have also shown CBD and 5-
HT
1A
interactions [49,50]. Further work using WAY-100635
has highlighted that CBD activation of 5-HT
1A
receptors, lo-
calized to the dorsal periaqueductal gray, mediates reported
panicolytic effects [43]. It is clear that CBD can modulate func-
tions involving 5-HT receptor function, although to what extent
such effects are direct remains unclear, particularly as several
in vivo studies did not include 5-HT receptor antagonist-only
Table 1 Receptor targets of cannabidiol
Target Concentration
range (μM)
EC
50
/IC
50
(μM)
K
i
(μM) Preparation or tissue Assay type Reference
CB
1
3; NSC ND ND A549/human Viability [18]
3; NSC ND ND H460/human Viability
NSC >10 ND ND ND [19]
NSC >30 ND HEK293 membrane/human GTPγS[20]
CB
2
3; NSC ND ND A549/human Viability [18]
3; NSC ND ND H460/human Viability
NSC >10 ND ND ND [19]
NSC >30 ND HEK293 membrane/human GTPγS[20]
Glycine receptor
α1 subunit
1300; (+) 12.3 ND HEK293/ND Patch clamp/current with glycine [21]
1300; (+) 132.4 ND HEK293/ND Patch clamp/current without glycine
Glycine receptor
α1βsubunit
1300; (+) 18.1 ND HEK293/ND Patch clamp/current with glycine
1300; (+) 144.3 ND HEK293/ND Patch clamp/current without glycine
Glycine receptor
α3 subunit
0.0150.00; (+) 3* ND HEK293/ND Patch clamp/current without glycine [22]
GPR18 10
4
100; (+) 51.1 ND HEK293/ND p44/42
MAPK activation
[23]
GPR55 10
3
1; () 0.45 ND Human osteoclasts Rho and ERK1/2 activation [24]
10
3
10
2
;() ND ND Colon Contraction [25]
() 0.445 ND HEK293 membrane/human GTPγS[20]
5-HT
1A
832; () ND ND CHO membrane/human [3H]-8-OH-DPAT ligand binding [26]
16; () ND ND CHO membrane/human [35S]-GTPγSassay
16; () ND ND CHO/human Forskolin
5-HT
2A
832; (+) ND ND NIH 3T3 membrane/rat [3H]-Ketanserin
nAChR α-7 0.1100.0; () 11.3 ND Xenopus oocyte/human Patch clamp/current/acetylcholine [27]
Opioid (δ)0.1100.0; ()* 10.7* 18.4 Cerebral cortex membrane/rat [3H]-NTI binding assay [28]
Opioid (μ)0.1100.0; (-)* 10* 31.6 Cerebral cortex membrane/rat [3H]-DAMGO binding assay
PPARγ3; (+) ND ND A549/human mRNA RT-PCR/Western blot [18]
3; (+) ND ND H460/human mRNA RT-PCR/Western blot
CB
1
= cannabinoid type 1; CB
2
= cannabinoid type 2; GPR = G protein-coupled receptor; 5-HT = serotonin; nAchR = nicotinic acetylcholine receptor;
PPAR = peroxisome proliferator-activated receptor; NSC = no significant change; ND = not described; HEK = human embryonic kidney; CHO =
Chinese hamster ovary; GTPγS=guanosine5-O-[gamma-thio]triphosphate; MAPK = mitogen-activatedprotein kinase; ERK = extracellular regulated
kinase; [3H]-3-OH-DPAT = 7-(dipropylamino)-5,6,7,8-tetrahydronaphthalen-1-ol; [3H]-NTI = naltrindole; [3H]-DAMGO = D-Ala
2
, N-MePhe
4
,Gly-
ol; RT-PCR = reverse transcription polymerase chain reaction; (+) = stimulation; (-) = inhibition
*Estimated from plots in cited paper
Bih et al.
Author's personal copy
study groups. Therefore, a clear demonstration of in vivo CBD
5-HT receptor interaction is required.
G Protein-coupled Receptor 55 (An Orphan G
Protein-coupled Receptor)
More recently, attention has turned to interactions between
CBD and non-endocannabinoid G protein-coupled receptors
(GPCRs) [23,24,51]. The orphan GPCR 55 (GPR55) shares
structural similarities in transmembrane domains 1, 2, and 3
when compared with the cannabinoid receptors, which may
indicate a binding site for cannabinoids [24,52]. A
radiolabelled synthetic analogue of Δ
9
-THC has been shown
to bind GPR55 with an EC
50
of 5 nM. Further studies with
CBD (0.0011.000 μM) have shown an inhibitory effect on
the agonist activity of CP 55,940 at GPR55, with an IC
50
of
445 nM [24]. Functional consequences of this inhibition have
been demonstrated in rat hippocampal preparations [51].
Here, CBD (1 μM) suppressed physiological activation of
the GPR55 receptors, which restricted excitatory output from
pyramidal cells. While the pharmacology and localization of
GPR55 merits further investigation [20,53], CBD has demon-
strable antagonistic effects within rat brain with clear physio-
logical relevance.
Nicotinic Acetylcholine Receptor
Evidence thatCBD modulates nicotinic acetylcholine receptor
(nAchRs) function comes from a study by Mahgoub et al.
[27]. Electrophysiological recordings from Xenopus oocytes
expressing the α-7-nAChR reveal a concentration-dependent
inhibition in the presence of CBD (0.1100.0 μM; Table 1).
Complementary biochemical evidence highlighted a noncom-
petitive binding to the α-7-nAChR. Furthermore, CBD (130
μM) inhibited acetylcholine-induced ion currents recorded in
rat hippocampal slices (IC
50
=12.7μM) [27].
Peroxisome Proliferator-activated Receptors
The peroxisome proliferator-activated receptor (PPAR) γ,oth-
erwise known as the glitazone receptor, is thought to be re-
sponsible for lipid storage and glucose metabolism [18], and
some anticancer effects of CBD are thought to by mediated
through interaction with PPARγ. In human A549 and A460
cancer cell lines, a time-dependent increase in PPARγmRNA
was observed following the application of CBD (13μM).
CBD has also been shown to inhibit tumor cell viability; cell
death assays revealed IC
50
values of 3.47 μM (A549) and
2.80 μM (A460). A complete loss of tumor cell viability
was observed at 8 μM (A549) and 7 μM (A460) [18].
Interestingly, these effects on tumor cells were not prevented
by pharmacological tools acting at CB
1
R, CB
2
R, or transient
receptor potential (TRP) vanilloid-type 1 (TRPV1). These
results do not confirm a direct effect of CBD upon PPARγ
but the extant data warrants its inclusion for completeness.
Furthermore, CBD was shown to displace a fluorescent ligand
of PPARγ(Fluormone PPAR Green, Life Technologies,
Paisley, UK) at a relatively low IC
50
(5 μM) in a fluorescence
polarization assay, indicating that CBD may bind to this pro-
tein [54]. In addition, CBD may be directly involved in
PPARγactivation as 10 μM CBD stimulates activity of the
luciferase reporter gene in HEK293 cells transiently overex-
pressing retinoid X receptor and PPARγ[54].
Enzyme Targets
Our review of the current literature revealed 19 investigations
of the effect of CBD upon 32 specific enzyme targets, which
accounts for 49% of the known molecular targets of CBD
(Fig. 1).
Enzymes Involved in Xenobiotic Metabolism
Cytochrome P450
As would be expected for a xenobiotic, CBD modulates sev-
eral cytochrome P450 (CYP450) enzymes. Although early
work did not show any statistically significant effect of CBD
upon CYP450 function [55], more recent studies (see below)
have reported several CYP450s as potential molecular targets.
The CYP450 enzyme system includes P450 and its nicotin-
amide adenine dinucleotide (phosphate) oxidase-linked reduc-
tase and catalyzes a monoxygenation reaction. In eukaryotes,
it is mainly found in the endoplasmic reticulum and mitochon-
dria where it maintains homeostatic control of lipophilic en-
dogenous compounds and the detoxification of lipophilic xe-
nobiotic compounds by oxidation, making them more water
soluble, and is thus the main component of phase 1 metabo-
lism. It is highly diverse, with many different isoforms [56].
CBD inhibits human recombinant CYP2C19 using a
mixed inhibition mechanism (Ki = 0.793 μM) [57].
Additionally, CBD is a potent inhibitor of this enzyme using
either high-affinity substrates such as omeprazole (IC
50
=1.55
μM) or 2-O-methylfluorescein (IC
50
=1.79μM), or a low-
affinity substrate, (S)-mephenytoin (IC
50
=2.51μM), to acti-
vate the reaction. Irrespective of substrate, CBD fully inhibits
the reaction at 10 μM, indicating that it is a potent inhibitor. It
is unclear whether this is involved in the therapeutic effect of
CBD in CNS disorders as evidence of CYP2C19 expression
in brain remains conflicted [58,59]. Two other CYPs,
CYP2C9 and CYP2D6, are also potential targets for CBD as
CBD also completely inhibits their function at 10 μM[60,
61]. However, CBD exerts a more potent effect upon
CYP2C9 (IC
50
=2.7μM) than CYP2D6 (IC
50
=6.016.52
μM). An additional study of the CYP1 family revealed that
CBD acts as a more potent inhibitor of CYP1A1 (IC
50
=0.537
Molecular Targets of Cannabidiol
Author's personal copy
μM) than CYP1A2 (IC
50
~3.5 μM) and CYP1B1 (IC
50
~5μM) [62]. With regard to potency, 90% of CYP1A1sac-
tivity was inhibited by 2.5 μM CBD, while 10 μM CBD
inhibited approximately 65% and 75% of CYP1A2 and
CYP1B1 activity, respectively.
Finally, it has been reported that CBD may inhibit members
of the CYP3 family [63]. Although, CYP3A4, CYP3A5, and
CYP3A7 have been identified as molecular targets for CBD
(Table 2), it appears that the compound is a more potent in-
hibitorofCYP3A5(IC
50
= 1.65 μM) compared with
CYP3A4 (IC
50
= 11.7 μM) and CYP3A7 (IC
50
=24.7μM).
More importantly, around 90% of CYP3A5s activity was
abolished by 10 μM CBD, whereas CYP3A4 and CYP3A7
required up to 50μMa physiologically implausible concen-
tration [17]: CBD to achieve ~90 and 72% inhibition, respec-
tively. The mode of inhibition determined in this study was
competitive for CYP3A4 and CYP3A5 but mixed for
CYP3A7.
Other Enzymes Involved in Xenobiotic Metabolism
The involvement of CYP450 is only part of the metabolism of
CBD [77], which has been linked to many other enzymes
involved in the control of redox. Oxidation of xenobiotics
can potentially produce highly reactive intermediates and is
why phase I and II metabolism is integrated with electrophilic
sensors [e.g., the nuclear factor (erythroid-derived 2)-like 2
(NRF2)Kelch-like ECH-associated protein 1 pathway),
which can upregulate cytoprotective genes and pathways
[78]. In this regard, CBD can activate NRF2 [79].
Several enzymes implicated in the regulation of redox have
been identified as potential molecular target of CBD [55].
Here, 100 μM CBD applied to mouse hepatic 105,000 g su-
pernatant reduced superoxide dismutase and catalase activity
by approximately 76% and 24%, respectively, and nicotin-
amide adenine dinucleotide (NAD)- and NAD phosphate-
dependent NAD(P)H quinone reductase by 80% and 81%,
respectively, when compared with controls. Moreover, CBD,
at the same concentration, also significantly stimulated gluta-
thione peroxidase and glutathione reductase by 24% and 40%,
respectively. However, it is notable that the concentration of
CBD used in this study was far in excess of that which can be
achieved physiologically [17].
Enzymes Involve in Cholesterol Metabolism
Several studies have reported the action of CBD upon en-
zymes involved in cholesterol metabolism. Although some
of the efficacious concentrations reported are achievable
in vivo, several are not and so can be excluded from
consideration.
Acyl-cholesterol acyltransferases (ACATs) catalyze the
formation of cholesterylester from long-chain fatty acyl-
coenzme A (CoA) and cholesterol. CBD may antagonize the
overall activity of ACAT (without distinguishing the individ-
ual activity of each subtype) in human fibroblast cells [64].
The potency of CBD on cholesteryl ester formation in the
presence of 0.1 mM [
14
C]-oleate-CoA and 5 μg/ml 25-OH-
cholesterol could be estimated to < 10 μM. Furthermore,
30 μM CBD appears to reduce 90% of the formation of this
product.
The synthesis of mevalonate from 3-hydroxy-3-
methylglutaryl-CoA is catalyzed by 3-hydroxy-3-
methylglutaryl-CoA reductase, which has been shown to be
stimulated by CBD [64]. Although an EC
50
for CBD was not
determined, the in vitro production of mevalonate was in-
creased 4.5-fold in human fibroblasts exposed to 30 μM
CBD when compared with cells exposed to 1 μMCBD.The
exact mechanism by which CBD exerts its stimulatory effect
on the enzyme is still unclear.
CBD has also been reported to inhibit progesterone 17α-
hydroxylase, testosterone 6α-hydroxylase, and testosterone
16β-hydroxylase [75]. However, CBD exhibits only weak
efficacy as 10 mM CBD is required to inhibit ~20%, 49%,
and 67% of progesterone 17α-hydroxylase, testosterone 6α-
hydroxylase, and testosterone 16β-hydroxylase, respectively.
Thus, it is unrealistic to propose that these enzymes are mean-
ingfully involved in the therapeutic effect of CBD in neuro-
logical disorders.
Effects on Enzymes Controlling Ceramide
Ceramide is a secondary messenger with a range of ef-
fects, including induction of differentiation to apoptosis. It
can be viewed as part of a stress signaling pathway and
can be produced either via sphingomyelinase-dependent
hydrolysis or by de novo synthesis. It also plays an im-
portant role in membrane rafts and the control of ion
channels [80,81].
Over 3 decades ago, CBD was shown to activate the
breakdown of sphingomyelin in fibroblasts taken from a
patient with NiemenPick disease [76]. Prior to CBD
treatment, these cells exhibited < 3% normal
sphingomyelinase activity [82]. Although no assay was
performed that showed a direct effect of CBD on
sphingomyelinase, the authors demonstrated that 16 μM
CBD reduced sphingomyelin levels by 77%. In contrast,
when the same in vitro assay was performed on normal
fibroblasts (WI-38 human lung fibroblasts) CBD exerted
no statistically significant effect on sphingomyelin hydro-
lysis. Furthermore, the antitumor effect of CBD has been
shown to be independent on the production of ceramide in
human glioma cells (but was associated with a fall in
mitochondrial potential; see below) [83]andthusitre-
mains unclear as to whether or not CBD acts directly on
the ceramide pathway.
Bih et al.
Author's personal copy
Tab le 2 Enzyme targets of cannabidiol
Target Concentration range
(μM); (effect)
EC
50
/IC
50
(μM)
K
i
/K
m
(μM)
Preparation or tissue Assay type Reference
ACAT 130; () <10* ND Fibroblast/human [64]
Arylalkylamine N-acetyltransferase 0.110.0; () ND ND Pineal gland/rat Radiometric assay [65]
Catalase 10; () ND ND Hepatic/mouse [55]
Complex I 210
2
;()* 8.2 ND Pig cortex High-resolution respirometry in malate
and pyruvate
[66]
Complex II 210
2
()* 19.1 ND Pig cortex High-resolution respirometry in succinate
and rotenone
Complex IV 210
2
;()* 18.8 ND Pig cortex High-resolution respirometry in antimycin,
ascorbate and TMPD
Complex I 1560 mg/kg, single
dose (+)
ND ND Prefrontal/cerebral cortex [67]
Complex II ND ND Prefrontal/cerebral cortex/hippocampus
Complex IIIII ND ND Prefrontal/cerebral cortex/hippocampus
Complex IV ND ND Prefrontal/cerebral cortex/striatum/
hippocampus
Complex I 1560 mg/kg/day for
14 days (+)
ND ND Prefrontal/cortex/striatum/hippocampus
Complex II ND ND Prefrontal/cortex/striatum/hippocampus
Complex IIIII ND ND Prefrontal/cortex
Complex IV ND ND Prefrontal/cerebral cortex/striatum
COX1 1010
3
;() ND ND Recombinant/ovine Polarography [68]
318; (NSC)* ND ND Recombinant/ovine Scintillation [69]
COX2 1010
3
;() ND ND Recombinant/human Polarography [68]
318; (NSC)* ND ND Recombinant/ovine Scintillation [69]
0.5mg/mouse/day
for 23 days; (+)
ND ND Tumor/ CD-1 nude mice Polarography
CYP2C19 10
2
10; () 2.51 0.793/50.800 Recombinant/ human HPLC/4'-OH-mephenytoin [57]
10
2
10; () 1.55 ND/1.26 Recombinant/human HPLC/5-OH-Omeprazole
10
2
10; () 1.79 ND/2.32 Recombinant/human FLUOstar OPTIMA
/ fluorescein, BMG Labtech,
Offenburg, Germany
CYP2D6 140; () 6.52 1.16 Recombinant/human FLUOstar OPTIMA
/O-demethylated
metabolite of AMMC
[60]
140; () 6.01 2.69 Recombinant/human Measuring dextrome-thorphan
O-demethylation
CYP3A4 0.150.0; () 11.7 1 Baculovirus-infected insect cells/human HPLC for N-desmethyl-diltiazem detection [63]
CYP3A5 0.150.0; () 1.65 0.195 Baculovirus-infected insect cells/human HPLC for N-desmethyl-diltiazem detection
CYP3A7 0.150.0; () 24.7 12.3 Baculovirus-infected insect cells/human HPLC for N-desmethyl-diltiazem detection
CYP2C9 0.110; ()* 2.7 2.31 Recombinant/human HPLC/ 4-hydroxy-diclofenac [61]
0.110; ()* 2.67 0.964 Recombinant/human HPLC/ 7-hydroxy-warfarin
CYP1A1 0.12.5; () 0.537 0.155 Recombinant/human FLUOstar OPTIMA for resorufin detection
[62,70]
CYP1A2 110; () ~3.5* 2.69 Recombinant/human FLUOstar OPTIMA for resorufin detection
[62]
CYP1B1 1-10; () ~5* 3.63 Recombinant/human FLUOstar OPTIMA for resorufin detection
DAGL-α110
2
; (NSC) ND ND COS-7 cells membrane/human Scintillation counting [14C]-oleic acid [71]
Molecular Targets of Cannabidiol
Author's personal copy
Tab le 2 (continued)
Target Concentration range
(μM); (effect)
EC
50
/IC
50
(μM)
K
i
/K
m
(μM)
Preparation or tissue Assay type Reference
FAA H 110
2
;()* 27.5 ND N18TG2 cell membrane/mouse Scintillation counting [14C]-ethanolamine [19]
150; () 15.2 ND Brain tissue membrane/rat Scintillation counting [14C]-ethanolamine [71]
0.5 mg/mouse/day
for 23 days; (+)
ND ND Tumor/ CD-1 nude mice [3H]-Arachidonic acid detection/RP-HPLC [68]
1016; (+) ND ND U87 glioma cells [3H]-Arachidonic acid detection/RP-HPLC
Glutathione reductase 10
2
; (+) ND ND Hepatic tissue/mouse [55]
GSH peroxidase 10
2
; (+) ND ND Hepatic tissue/mouse
HMG-CoA Reductase 130; (+) ND ND Fibroblast/human [64]
IDO 0.110.0; () ND ND LPS stimulated-THP-1 cells/human HPLC/kynureninetryptophan [72]
LOX-5 2200; () 73.73 ND LOX inhibitor screening assay kit
(Cayman Chemical)
Hydroperoxides treated with chromogen
spectrophotometry
[73]
1010
3
; (NSC) ND ND Purified from barley Spectropho-tometric detection of
hydroperoxides
[68]
0.5mg/mouse/day
for 23 days; ()
ND ND Tumor/ CD-1 nude mice Detection of arachidonic acid
detection/RP-HPLC
LOX-15 0.210.0; () 2.56 ND LOX inhibitor screening assay kit
(Cayman Chemical)
Hydroperoxides treated with chromogen
spectropho-tometry
[73]
1010
3
; (NSC) ND ND Purified from soybean Spectropho-tometric detection of
hydroperoxi-des
[68]
MAGL 110
2
; (NSC) ND ND COS-7 cells homogenate/human Scintillation counting [3H]-glycerol [19]
N-acylethanolamine acid amide
hydrolase
110
2
; (-) >100 ND HEK293 cells membrane/human Scintillation counting [14C]-ethanolamine
NAD(P)H quinone reductase 10
2
;() ND ND Hepatic tissue/mouse NADH-dependent reaction [55]
10
2
;() ND ND Hepatic tissue/mouse NADPH-dependent reaction
NAPE-PLD 0.5 mg/mouse/day
for 23 days;
(NSC)
ND ND Tumor/ CD-1 nude mice Assaying [3H]-anandamide formation [68]
Phosholipase A2 31.9159.5; (+, ) 6.4/134.0 ND Naja naja venom Spectroscopy [74]
Progesterone 17α-hydroxylase 1010
3
;() ND ND Testis microsome/rat HPLC [75]
SOD 10
2
;() ND ND Hepatic tissue/mouse [55]
Sphingomyelinase 16.0063.60; (+)* ND ND Nieman pick Fibroblast/human HPLC, kratos detector and rainin Microsorb
§
[76]
Testosterone 6α-hydroxylase 1010
3
;() ND ND Hepatic microsome/rat [75]
Testosterone 16β-hydroxylase 1010
3
;() ND ND Hepatic microsome/rat
ACAT = acyl-cholesterol acyltransferase; COX = cyclooxygenase; CYP = cytochrome P; DAGL = diacylglycerol lipase; FAAH = fatty acid amide hydrolase; GSH = Glutathione; HMG-CoA = 3-hydroxy-
3-methylglutaryl-coenzyme A; IDO = indoleamine-2,3-dioxygenase; LOX = lipoxygenase; MAGL = monoacylglycerol lipase; NAPE-PLD = N-acyl-phosphatidyl-ethanolamine-selective phospholipase
D; SOD = superoxide dismutase; NSC = no significant change; ND = not described; LPS = lipopolysaccharide; HEK = human embryonic kidney; TMPD = trimethyl pentanediol; HPLC = high-
performance liquid chromatography; AMMC = 3-[2-(N,N-diethyl-N-methylammonium)-ethyl]-7-methoxy-4-methylcoumarin; RP-HPLC = reverse phase HPLC; NADH = nicotinamide adenine dinu-
cleotide; NADPH = nicotinamide adenine dinucleotide phosphate; (+) = stimulation; () = inhibition
*Estimated from plots in cited paper
Ann Arbor, MI, USA
§
Agilent, Santa Clara, CA, USA
Bih et al.
Author's personal copy
Electron Transport Chain
The mitochondrial electron transport chain (ETC) comprises a
group of 4 enzymatic complexes (known as, complex I, II, III,
and IV) involved in mitochondrial bioenergetics. Recent stud-
ies suggest that CBD may alter mitochondrial bioenergetics
via modulation of ETC activity [67]. Moreover, mitochondrial
bioenergetics may also be modified by CBD via interaction
with mitochondrial voltage-dependent anion channel 1
(VDAC1; see BIon Channels^)[84], which may affect the
production of adenosine triphosphate (ATP). VDAC1 is in-
volved in the transport of ions (e.g., calcium) and multiple
metabolites, including adenosine diphosphate [85], the sub-
strate of ATP.
There are still some discrepancies surrounding the nature of
CBDs action on the complexes comprising the ETC. Acute or
chronic in vivo administration (Table 2) of CBD (60 mg/kg,
i.p.) increased the activity of complex I, II/III, and IV in tissue
obtained from cerebral and the prefrontal cortex in rat [67],
while in vitro application of CBD to isolated mitochondria
from pig brain antagonized the activity of complex I (IC
50
=
8.2 μM), II (IC
50
=19.2μM), and IV (IC
50
=18.8μM) [66];
apart from complex I, the IC
50
values presented are inconsis-
tent with the concentrations likely to be achieved by the
in vivo study to which we have compared [17]. Moreover, this
latter study determined that CBD exerted a greater effect on
complexes I and II than on complex IVas 100 μM CBD fully
inhibited complexes I and II, while complex IVretained ~75%
activity. These results were confirmed by a recent in vitro
study which showed that CBD inhibits directly complexes I,
II/III, and IVonly at higher concentrations (50 μM) [86].
In summary, it would seem that CBD acts differently on
mitochondrial complexes depending on the route of adminis-
tration. While the number of studies in this area remains small,
they appear to suggest that CBD effects on ETC complexes
may involve indirect mechanisms (e.g., increase ETC com-
plex expression or mitochondrial biogenesis) that ultimately
result in the bioenergetic increases reported by Valvassori
et al. [67], although a direct interaction between CBD and
complex I may inhibit this enzyme [66].
Arylalkylamine N-Acetyltransferase
Arylalkylamine N-acetyltransferase (AANAT) plays a role in
melatonin synthesis, which controls circadian rhythm, orches-
trated by the pineal gland, although melatonin performs other
functions,including control of neuroexcitation, immune mod-
ulation, and protection against oxidative stress [8789]. The
effect of CBD on norepinephrine-induced activation of
AANAT in the rat pineal gland and in cell-free lysates has
been studied where 10 μM CBD significantly attenuated nor-
epinephrine induction of melatonin synthesis by 35% com-
pared with controls in isolated pineal glands, while 1 μM
inhibited AANAT activity by 35% [65]. This effect was shown
to be independent of CB
1
RorCB
2
R and due to direct
AANAT inhibition demonstrated using cell-free lysates of
the pineal gland where 10 μM CBD inhibited AANAT by
40% [65].
Effect of CBD on Phospholipases
Phospholipases are responsible for the breakdown of complex
lipids and play a key role in lipid signaling. In particular,
phospholipase A2 (PLA2) may play a key role in brain in-
flammation and thus neurodegeneration [90]. CBD has been
reported to act upon Naja naja venom PLA2, while it has no
effect on N-acylphosphatidyl-ethanolamines-hydrolyzing
phospholipase D, monoacylglycerol lipase, or diacylglycerol
lipase-α(Table 2). Using an in vitro assay, the effect of CBD
upon N. naja venom PLA
2
was reportedly biphasic where low
concentration CBD stimulated activity (EC
50
=6.4μM; max-
imum induction = 262% induction at 39.1 μM), while at
higher concentrations, activity was inhibited (IC
50
=134μM;
maximum inhibition ~60 % inhibition at 159.5 μM) [74].
While the reported EC
50
is physiologically achievable
in vivo, it is unclear whether CBD exerts the same effects upon
mammalian PLA2s as they are likely to be structurally and
functionally different from N. naja venom PLA2 [91].
While studying the effect of CBD on phospholipases in-
volved in regulation of endocannabinoids, repeated adminis-
tration of CBD (0.5 mg/mouse) in female CD-1 nude mice for
23 days was shown not to change the in vivo activity of N-
acylphosphatidyl-ethanolaminephospholipase D. Similarly,
the same authors showed that CBD does not act upon the
human monoacylglycerol lipase and diacylglycerol lipase-α
expressed in COS-7 cells [68].
Effect on Fatty Acid Amide Hydrolase
Fatty acid amide hydrolase (FAAH) is an enzyme involved in
the catabolism of fatty acid amides (e.g.,the endocannabinoid,
anandamide) and reports of CBD effects at this target are
conflicted throughout the literature. Two studies that investi-
gated the effects of CBD on membrane extracted FAAH
in vitro reported different IC
50
values (27.5 μMand15.2
μM, respectively) [19,71]; marginal differences that could
conceivably arise via tissue and/or species differences
(Table 2). These results suggest that higher CBD concentra-
tions may exert inhibitory effects on FAAH. In contrast, an
in vitro study in U87 cells suggested that 16 μM CBD stimu-
lated FAAH by ~2-fold when compared with untreated cells
[68]. Furthermore, the samestudy found that mice treated with
0.5 mg CBD peritumorally for 23 days exhibited increased
FAAH activity via a possible post-translational modification
as the protein content did not differ from untreated mice [68].
Molecular Targets of Cannabidiol
Author's personal copy
Consequently, based on the limited and conflicting evi-
dence, it appears that CBDs effects upon FAAH may depend
upon the physiological environment in which it is studied as
cellular or systemic systems reveal CBD-induced FAAH acti-
vation, while membrane extract assays reveal CBD-induced
inhibition. Such an environment-dependent effect suggests
that more than one mechanism may be involvedshort-term
direct effects and longer-term, post-translational modifications
are probably both at play [68]. It is also notable that reported
effects upon FAAH only occur at relatively high micromolar
concentrations.
Cyclooxygenases and Lipoxygenases
Cyclooxygenases (COXs) and lipoxygenases (LOXs) metab-
olize free arachidonic acid into 2-series prostaglandins and 4-
series leukotrienes, respectively [92]. Here, COX1 and COX2
activity was partially inhibited (50% and 40%, respectively)
by CBD at very high concentrations (500 μM) in an in vitro
assay (Table 2)[68]. Although IC
50
values were not deter-
mined in this study, inhibitory effects of CBD were reported
from 50 μM and 100 μM for COX1 and COX2, respective-
lyconcentrations hard to achieve in vivo [17]. Conversely, a
different study showed that 318 μM CBD had no effect upon
the in vitro activities ofCOX1 and COX2 (Table 2)[69], while
in vivo administration of CBD had no effect upon COX2
activity in tumor tissue [68] (Table 2). However, a further
study suggested that CBD may modulate COX2 activity by
stimulating transcription and/or translation in some subtypes
of cells [18]. While the concentrations described above are
unlikely to be physiologically achievable, a more recent study
suggested that 10 μM CBD increased arterial vasorelaxation
in diabetic Zucker rats, which was associated with COX1 and
COX2 stimulation via an allosteric mechanism [93].
CBD has also been reported to act as an inhibitor of 15-
LOX and 5-LOX. However, CBD exerts a more potent effect
on 15-LOX (IC
50
=2.56μM) than 5-LOX (IC
50
=73.73μM)
in vitro [73]. Furthermore, 200 μM CBD cannot fully abolish
5-LOX activity, while 15-LOX activity can be fully blocked
by 10 μM CBD, indicating that the efficacy of CBD is higher
upon 15-LOX. However, these results are in contradiction
with previous work which showed that CBD exerted no effect
upon 5- and 15-LOX in vitro,despitein vivo administration
inhibiting 5-LOX activity in tumor tissue [68].
Indoleamine-2,3-dioxygenase
Indolamine-2,3-dioxygenase (IDO) is activated by inflamma-
tion, in particular by interferon-γand is thought to be involved
in inflammatory-associated depression (cytokine-induced
sickness behavior). Its overactivation may be important in
neurodegeneration as it can result in oxidative stress. It
catalyzes the degradation of ring-containing compounds, in
particular tryptophan to kynurenine [94].
CBD exerts biphasic effects on the kynuenine/tryptophan
ratio in human monocytic cells stimulated with concavlin A or
phytohemagglutinin (PHA), suggesting modulation of IDO
where 0.030.3 μM CBD stimulated them, whereas the IC
50
was about 8.9 μM. The highest concentration tested (16 μM)
almost completely inhibited IDO mRNA expression induced
by PHA. CBD was also shown to inhibit LPS stimulation of
THP-1 cell IDO activity (IC
50
=0.9μM). Given the bidirec-
tional nature of CBD effects upon IDO, it is most likely that
CBD concentrations typically achieved in in vivo studies in
animal models of disease will inhibit IDO [17].
Ion Channel Targets
The available literature revealed 10 investigationsof the effect
of CBD upon 10 specific ion channel targets, which accounted
for 15% of the known molecular targets of CBD (Fig. 1).
TRP Channels
CBD has been reported to act as an agonist at human TRPV1
channel expressed in HEK293 cells when assessed using a
fluorescence-based, high-throughput assay with an efficacy
of ~70% that of a saturating concentration of the positive
control, ionomycin (4 μM) (Table 3)[19]. TRP channels are
present in the plasma membrane of a variety of cells in many
tissues and act as ligand-gated, nonselective cation channels
permeable to sodium, calcium, and magnesium ions [102].
Here, CBD-induced increases in intracellular calcium
([Ca
2+
]
i
) were reportedly abolished by the TRPV1 antagonist
capsazepine (10 μM), suggesting a TRPV1-specific effect of
CBD. The authors also asserted that CBD did not affect
[Ca
2+
]
i
in non-TRPV1-expressing HEK293 cells but did not
present supporting data. Notably, other molecular targets of
CBD are present in HEK293 cells (e.g., VDAC1; see Tables 1,
2,3and 4) but their expression on intracellular membranes is
likely to prevent CBD-mediated effects from being identified
by an assay such as this. This group used the same approach in
a later study that examined anticancer effects of CBD [99],
where CBD effects at human TRPV1 expressed in HEK293
cells were again assessed and reported in the same way as
above. While specific EC
50
and efficacy results for were not
formally presented, an estimate from the concentrationre-
sponse curves published in suggests that CBD was both less
efficacious (~50% vs 70%) and less potent (~6.3 μMvs 3.5
μM) than previously reported [99]. The same group used the
same approach to assess CBD effects at human TRPV1
expressed in HEK293 cells for a third and final time [71].
Again, some disparity between the results presented and those
previously published was evident as CBDs efficacy was con-
sistent with that reported by Ligresti et al. [99] but not that
Bih et al.
Author's personal copy
reported by Bisogno et al. [19]; however, potency differed
from both preceding papers (EC
50:
1.0 μMvs 6.3μM([71]
estimated from Ligresti et al. [99]vs 3.5 μM[19]). Taken
together, these results suggest that CBD can submaximally
activate TRPV1 receptors, although the potency of CBDs
effects remains less welldefined. Analysis of the above results
yields an approximate mean EC
50
value of 3.5 μMwithan
SEM of 1.5 μM, which suggests that the 95% confidence
interval is greater than that values measured and could be a
consequence of variability in TRPV1 in the expression sys-
tems used (V. DiMarzo, personal communication) [110].
In contrast to the above studies, comparable fluorescence-
based Ca
2+
imaging in HEK293 cells expressing rat TRPV1
receptors revealed that CBD (100 μM) produced a response
that was only 21% of that produced by the positive control,
capsaicin (500 nM), which was insufficient to determine
reliablyanEC
50
value [96]. Furthermore, a recent study used
patch clamp techniques to assess the effect of CBD (330 μM)
on the bidirectional current recorded in HEK293 cells overex-
pressing rat TRPV1 receptors that can be evoked by application
of the TRPV1 agonist, capsaicin (110 μM) [98]. Although a
complete concentrationresponse relationship for CBD was not
determined, detectable currents were evoked by 10 and 30 μM
CBD but not by 3 μM. While species-specific differences in
TRPV1 responsiveness could account for efficacy and potency
differences in the preceding studies, the use of rat TRPV1 here
and by Qin et al. [96] justifies the targets consideration. It is
notable that the highest CBD concentration (30 μM) studied by
Iannotti et al. [98] elicited a peak response of ~30% of that
produced by capsaicin (c.f., 21%). While it remains unknown
whether the response to 30 μM CBD represents a saturating
response, it is notable that CBD concentrations > 20 μMare
Table 3 Ion channel targets of cannabidiol
Target Concentration
range (μM)
EC
50
/IC
50
(μM)
K
i
(μM) Preparation or tissue Assay type Reference
Cav3.1 T-type 10
2
10; () 0.82 ND HEK293 cells/human Patch clamp/ current clamp [95]
Cav3.2 T-type 10
2
10; () 0.78 ND HEK293 cells/Hhuman
Cav3.3 T-type 10
2
30; () 3.7 ND HEK293 cells/human
TRPA1 ND-100; () 81.4 ND HEK 293 cells/rat Calcium mobilization assays [96]
10
3
10; ()* 0.096 ND HEK 293 cells/rat Calcium assay/Fluo4-AM [97]
330; (+) ND ND HEK 293 cells/rat Patch clamp/current [98]
10
3
*25; () 0.11 ND HEK 293 cells/rat Calcium assay/Fluo4-AM [71]
TRPM8 10
3
10; ()* 0.14 ND HEK 293 cells/rat Calcium assay/Fluo4-AM [97]
10
3
25*; () 0.06 ND HEK 293 cells/rat [71]
TRPV1 0.110.0; (+) 3.5 ND HEK 293 cells/human Calcium assay/Fluo-3 [19]
10
3
25; (+)* 1 ND HEK 293 cells/human Calcium assay/Fluo4-AM [71]
ND-100; (NSC) ND ND HEK 293 cells/rat Calcium mobilization assays [96]
330; (+) ND ND HEK 293 cells/rat Patch clamp/current [98]
3; (NSC) ND ND A549/human Viability [18]
3; (NSC) ND ND A549/human DNA fragmentation
10
3
10; (+)* 0.7* ND MDA-MB-231 cells Calcium assay/Fluo4 [99]
TRPV2 0.0510
3
; (+)* 31.7 ND HEK 293 cells/human Calcium mobilization assay [96]
0.1200.0; (+)* 22.2 ND U87MG glioma cells/rat Calcium mobilization assay [100]
10
3
25*; (+) 1.25 ND HEK 293 cells/rat Calcium assay/Fluo4-AM [71]
0.05500.00; (+)* 3.7 ND HEK 293 cells/rat Calcium mobilization assay [96]
330; (+) ND ND HEK293 cells/rat Patch clamp/current [98]
3; (NSC) ND ND H460/human Viability [18]
3; (NSC) ND ND H460/human DNA fragmentation
TRPV3 10
1
10
3
; (+)* 3.7 ND HEK 293 cells/rat Calcium assay/Fluo4-AM [101]
TRPV4 10
2
10
2
; (+)* 0.8 ND HEK 293 cells/rat Calcium assay/Fluo4-AM
VDAC1 20; () ND ND Liver VDAC1 channel in
planar lipid bilayer/sheep
Bilayer Clamp BC-525B
amplifier/current
[84]
0.175.0; (+)* ND Kd; 11.2 Liver VDAC1 channel in
planar lipid bilayer/sheep
Thermophoretic analysis/
CBDVDAC1 interaction
TRP = transient receptor potential; TRPA1 = TRP ankyrin type 1; TRPM8 = TRP subfamily M; TRPV1 = TRP vanilloid-type 1; VDAC = voltage-dependent
anion channel; NSC = no significant change; ND = not described; HEK = human embryonic kidney; CBD = cannabidiol; (+) = stimulation; (): inhibition
*Estimated from plots in cited paper
Molecular Targets of Cannabidiol
Author's personal copy
unlikely to be reached in vivo andsoareunlikelytobephysi-
ologically meaningful [17].
In addition to reports of TRPV1 receptor agonism, CBD
has also been reported to interact with the TRPV2 receptor. In
the first study reporting such effects [96], CBD agonized hu-
man TRPV2 receptors overexpressed in HEK293 cells when
assessed using a fluorescence-based assay, a finding subse-
quently confirmed using patch clamp electrophysiology and
extended to reveal similar effects upon rat TRPV2
overexpressed in the same cell line, although with markedly
different potency [EC
50
=3.7μM(human)vs 31.7 μM(rat)]
(Table 3). CBD effects upon rat TRPV2 receptors were also
separately investigated in the same expression system and
using patch clamp electrophysiology [98]. Here, similar re-
sults were obtained where CBD activated TRPV2 at concen-
trations 10 μM, although a complete concentrationre-
sponse relationship was not. Subsequently, TRPV2-mediated
effects of CBD upon the chemosensitivity of a glioblastoma
multiforme cell line to conventional cytotoxic agents were
reported [100]. Interestingly, in addition to activating
TRPV2, CBD also stimulated TRPV2 expression in this cell
line. Finally, such TRPV2-dependent effects of CBD in ma-
lignant cell lines were extended by a similar demonstration of
CBD-induced increases in chemosensitivity in a multiple my-
eloma cell line [111]. In both cases, TRPV2-mediated effects
of CBD in malignant cells required moderate micromolar con-
centrations (20 μM) to achieve significant effects.
A single report has also described CBD effects at rat
TRPV3 and TRPV4 receptors expressed in HEK293 cells
where it acted as a potent agonist in both cases (Table 3),
although with variable efficacy when compared with positive
controls (TRPV3: 50%; TRPV4: 16%) [71].
Several reports have also described CBD interactions with
the transient receptor potential ankyrin type 1 (TRPA1) recep-
tor. A first report stated that CBD acts as a potent and effica-
cious agonist at rat TRPA1 receptors overexpressed in
HEK293 cells (Table 3), as assessed using a fluorescence-
based assay, but went on to report that high concentrations
(100 μM) of CBD failed to elicit [Ca
2+
]
i
increases in mustard
oil-sensitive dorsal root ganglion neurons [97]. Thereafter,
CBD was reported to activate fully rat TRPA1 expressed in
HEK293 cells when assessed in a similar fluorescence-based
assay, but with much lower potency (EC
50
= 81.4 μM;
Table 3)[96]. Subsequently, the first report was recapitulated
by the same group and reported similar effects of CBD at rat
TRPA1 in HEK293 cells but extended the study to show that
CBD can also act as a potent desensitizer (IC
50
=0.16μM) of
this receptor [97]. Finally, CBD effects at TRPA1 receptors
expressed in HEK293 cells were assessed using patch clamp
electrophysiology where CBD (10 μM) evoked a ruthenium
Table 4 Transporter targets of cannabidiol
Target Concentration
range (μM);
(effect)
EC
50
/IC
50
(μM)
K
i
(μM)
Preparation or tissue Assay type Reference
ABCC1 0.0850.00; () 5.5 ND SF9/ human ATPase/N-ethyl-maleimide-
glutathione
[103]
ABCG2 0.0850.00; () 7.3 ND ATPase/ sulfasalazine [104]
Adenosine uptake 10
3
10
2
;() 3.5 ND Striatal tissue synaptosome/rat Dual-label counting
[3H]-adenosine
[105]
Adenosine uptake 10
2
1; () 0.12 ND EOC-20 microglia/ mouse Scintillation counting/
[3H]-adenosine
[106]
Anandamide
uptake (AMT)
550; ()* 22 ND RBL 2H3 cells/rat Scintillation counting/
[14C]-anandamide
[19]
125; ()25.3ND [71]
Dopamine uptake 0.5100.0; ()* 16.2 ND Striatal tissue synaptosome/ rat Dual-label counting/
[3H]-dopamine
[105]
0.00510.000;(-)* ND ND Hippocampus/corpus striatum/
rat
[3H]-dopamine
Glutamate uptake 110
2
;() 43.8 ND Striatal tissue synaptosome/rat Dual-label counting/
[3H]-glutamate
Mg
2+
-ATPase 110
3
;() ND ND Brain cortical vesicules/ rat Phosphate release [107]
NA uptake 0.00510.000;()* ND ND Hippocampus/corpus striatum/
rat
Method: Coyle and Snyder
(1969)/ [3H]-NA
[108]
Thymidine uptake 10
2
1; () 0.19 ND EOC-20 microglia/ mouse Scintillation counting/
[3H]-thymidine
[106]
Choline uptake 0.24240.00; () 15.9 ND Hippocampus synaptosome/rat Method: Whittaker et al. (1964) [109]
AMT = anandamide membrane transporter; Mg
2+
-ATPase = magnesium-activated adenosine triphosphatase; NA = noradrenaline; ND = not described;
(+) = stimulation; () = inhibition
*Estimated from plots in cited paper
Bih et al.
Author's personal copy
red-sensitive bidirectional conductance, although a full con-
centrationresponse relationship was not presented. The
highest concentration of CBD tested (30 μM) was only able
to induce a TRPA1-mediated current that was ~20% of that
evoked by the positive control TRPA1 agonist, mustard oil,
and was rapidly desensitized by continued exposure to CBD
[98].
Finally, CBD has been reported to act as a potent antagonist
(Table 3) at the rat TRP subfamily M (TRPM8) receptor
expressed in HEK293 cells [71,97].
VDAC1 (Mitochondrial Porin)
CBD has also been reported to act upon sheep liver VDAC1.
VDAC1 is a promiscuous ion channel, also known as mito-
chondrial porin, and is most often located on the outer mem-
brane of mitochondria, although has also been reported to be
present on cell plasma membranes [112]. VDAC1 plays a
complex and important role in a variety of cell functions such
as ATP rationing, Ca
2+
homeostasis, apoptosis, and protection
against oxidative stress. The effect of CBD upon VDAC1
channels reconstituted into an asolectin-based planar lipid bi-
layer has been examined using patch clamp electrophysiology
[84]. Here, CBD (20 μM) inhibited VDAC1-mediated cur-
rents to varying extents at different voltage steps (60 mV to
+60 mV). This functional finding was further confirmed via a
thermophoretic analysis, which demonstrated an interaction
between CBD and VDAC1 with a K
d
of 11.2 μM(Table3).
Voltage-gated Calcium Channels (VGCCs)
CBD has been reported to inhibit human T-type voltage-gated
calcium channels (VGCCs) encoded by the Ca
V
3genefamily
when expressed in HEK293 cells and assessed using patch
clamp electrophysiology [95]. CBD was able to abolish fully
conductances via Ca
V
3.1, 3.2 and 3.3 T-type channels with
comparable potency for Ca
V
3.1, 3.2 but lower potency for
Ca
V
3.3 (Table 3). This finding was extended to investigation
of CBD effects upon native T-type conductances in mouse
trigeminal ganglion neurons where similar inhibitory effects
were observed, although a complete concentration response
relationship was not described.
Voltage-gated Sodium Channels
CBD can block voltage-gated sodium channels (VGSCs) in a
number of in vitro assays but, interestingly, this VGSC block-
ade per se does not confer anticonvulsant effects in whole-
animal models of seizure [113]. Here, the effects of CBD
(110 μM) were examined in rat brain slices, cultured mouse
cortical neurons, and human SH-SY5Y cells using patch
clamp electrophysiology and in Chinese hamster ovary cells
expressing human Na
V
1.1 and Na
V
1.2 VGSC subtypes using
a fluorescence-based assay. In all in vitro tests, CBD inhibited
VGSC conductances at micromolar concentrations, although
the block was sudden with no apparent conventional concen-
trationresponse relationship. This was response is suggestive
of a nonspecific effect (e.g., disruption of the lipid membrane
in which the channels are located) and was also seen for the
structurally similar plant cannabinoid, cannabigerol. When
compared in a model of generalized seizure, CBD was strong-
ly anticonvulsant but cannabigerol exerted no effect, suggest-
ing that the apparent in vitro block of VGSC by CBD and
cannabigerol are artefactual.
Transporter Targets
Our review of the current literature revealed 9 investigations
of the effect of CBD upon 13 specific transporter targets,
which accounts for 20% of the known molecular targets of
CBD (Fig. 1). These are further illustrated in Table 4.
Neurotransmitter Transporters
The uptake of neurotransmitters is vital to homeostasis in the
brain and prevents overstimulation of receptors, which can
result in cell death [114]. A major route for neurotransmitter
uptake is through transporter proteins, which shuttle specific
transmitters. Work in the 1970s utilized rat brain synapto-
somes to determine the effect of CBD on the activity of var-
ious transporters (Table 4)[115]. Here, CBD (1100 μM)
showed a concentration-dependent effect, significant at 50
μM, on 5-HT, noradrenaline (NA), and γ-aminobutyric acid
(GABA) uptake, with greater potency for inhibition of 5-HT
and NA uptake (50 μM CBD; 78% and 81% inhibition, re-
spectively). No significant changes were recorded for 1 μM
CBD, which draws on concerns regarding the physiological
relevance of these data. A more recent study employed rat
hippocampal and striatal synaptosomes to investigate the roles
of CBD on neurotransmitter uptake [116]. By measuring
radiolabelled uptake of NA and dopamine, significant inhibi-
tion of NA and dopamine was observed by 1 μM CBD in both
brain regions investigated.
With regard to dopamine transport, further studies showed
a dose-dependent reduction in dopamine uptake following
application of CBD (0.5100.0 μM; IC
50
=16.2μM) in rat
striatal synaptosomes, consistent with a previous study
(Table 4)[105]. Glutamate uptake was again inhibited in a
dose-dependent manner following application of CBD (1
100 μM) to rat striatal synaptosomes. However, unlike
CBDs effect on dopamine uptake, half maximal inhibition
was achieved at the higher concentration of 43.8 μM[105].
These studies indicate that CBD can inhibit uptake of adeno-
sine, dopamine, and glutamate in vitro, but with variable po-
tency. While examining the anticonvulsant properties of CBD
some investigation of the cholinergic system was undertaken
Molecular Targets of Cannabidiol
Author's personal copy
but did not reveal significant changes in choline uptake across
several brain regions following 60 mg/kg intraperitoneally
administered CBD [117]. This finding was in contrast to
in vitro observations where an IC
50
of 15.9 μM was reported
for CBD upon choline uptake in a rat hippocampal prepara-
tion. This study highlights the need for caution in interpreting
in vitro assays and extrapolating and/or making assumptions
regarding the in vivo effects of CBD.
Microglia act as macrophages of the CNS, and are
thought to serve as the first and main form of active im-
mune defense. In previous studies, CBD has been shown
to exert anti-inflammatory effects, but a distinct mecha-
nism of action for this effect remains unresolved [106],
despite hypotheses that CBD exerts anti-inflammatory ef-
fects by facilitating adenosine transmission [10,106]. To
test this hypothesis, the effect of CBD on the uptake of
[
3
H]-adenosine was investigated (see Table 4). In vitro
studies on EOC 20 rat microglia demonstrated that CBD
(11000 nM) inhibited adenosine uptake in a dose-
dependent manner. A similar effect was observed in
RAW264.7 macrophages where a K
i
of 225nM was re-
ported. Further studies, which examined the affinity of
CBD for the equilibrative nucleoside transporter 1
(ENT1) showed that CBD displaced radiolabelled 6-
S-[(4-Nitrophenyl)methyl]-6-thioinosine ([
3
H]-NMBPR)
with a K
i
value of 237 nM, consistent with the result
obtained from RAW264.7 macrophages. These findings
indicate that CBD can inhibit adenosine uptake by bind-
ing ENT1 [106]. Thus, while it is clear that CBD can
modulate adenosine signaling at both the receptor (see
BReceptors^) and transporter levels, the contribution of
these effects to the in vivo pharmacology of CBD still
requires definitive study.
Anandamide Transporters
CBD has been implicated in modulating the metabolism of the
endocannabinoid, arachidonoylethanolamide (AEA; also
anandamide)in vitro [118,119]. Transporter proteins are
involved in regulating AEA-mediated cellular signaling and
have also been reported to be targets for CBD [19,71].
CBD inhibits the anandamide membrane transporter,
which is thought to be responsible for the reuptake of AEA
into cells prior to degradation [110]. Scintillation counting of
[
3
H]-AEA revealed that CBD inhibits AEA uptake in a
concentration-dependent manner within a range of 5 to
50 μM(IC
50
=22μM) in rat basophilic leukemia cells
(RBL-2H3) [19]. Using the same techniqueand cellline, these
findings were replicated by De Petrocellis et al. [71], where
the IC
50
obtained was 25.3 μM, although the CBD concentra-
tion range tested was 125 μM[71]. Previous studies under-
taken by Di Marzo et al. [120] have shown that AEA exerts
anti-inflammatory effects. Therefore, the anti-inflammatory
effects also exerted by CBD may arise from facilitation of
AEA transmission [19].
Overall, while the concentrations at which effects on AEA
transporters have been reported are relatively high, it remains
possible that incomplete inhibition by CBD could underlie
some effects in vivo.
Multidrug Resistance Transporters
ABCC1 or multidrug resistance-associated protein 1, is a
transporter expressed in various tissues and displays a
wide substrate specificity, including important therapeu-
tics. Investigating intracellular substrate retention, using
Fluo3 and vincristine, CBD (IC
50
=128.3 μM and 30.9
μM, respectively) was reported to be a Bpotent^inhibitor
of ABCC1-meditated transport in vitro [103]. The use of
the word potent in this regard is questionable given the
reported IC
50.
Previous work by this group [104]using
flow cytometric analysis of substrate accumulation and
ATPase activity assays also highlighted the ABCG2 trans-
porter as a molecular target for CBD. Again, inhibition
was demonstrated in the presence of 10 and 50 μM
CBD (Table 4). The ATP-dependent transporter P-
glycoprotein has also been implicated in CBD effects
[121]. ATPase assays revealed a CBD concentration-
dependent inhibition of transport activity mediated by P-
glycoprotein (IC
50
=39.6μM). All these studies highlight
in vitro evidence that adds these substrate transporters to
the list of molecular targets; however, in vivo evidence is
still lacking.
Mg-ATPase Transporter
Magnesium-activated ATP (Mg
2+
-ATPase) is a transporter
that is expressed in erythrocytes, where Mg
2+
flows through
the plasma membrane to drive ATP hydrolysis [122]. In ves-
icles obtained from the rat cortex, CBD was reported to be a
potent inhibitor of Mg
2+
-ATPase within a range of 1
1000 μM[107].
Part 2: The Relevance of CBD Molecular Targets
in Neurological Disorders
Part 1 considered the available pharmacological evidence de-
scribing CBDs effects at specific molecular targets, regardless
of any a priori evidence of their involvementin a given disease
or disorder. Here, in part 2, we consider whether: 1) there is a
role for each of CBDs plausible pharmacological targets in
neurological disease; and whether 2) CBDs modulation of a
given target would be beneficial or detrimental.
Bih et al.
Author's personal copy
Epilepsy
Epilepsy is a highly heterogenous neurological disease char-
acterized by the manifestation of spontaneous recurrent sei-
zures and has a significant unmet clinical need whereby >
30% of people with epilepsy do not gain full control of their
seizures from currently available treatments [123]. Cannabis
has a long anecdotal history of use for the treatment of epilep-
sy, although, confusingly, significant reports of proconvulsant
effects are also present in the literature (for a detailed review
see [124]). CBD has a more consistent record and has been
repeatedly shown to exert anticonvulsant effects in a variety of
preclinical models and small, but flawed, clinical trials but the
mechanism through whichthese effects are exerted remains to
be determined [124]. Here, we consider the potential role that
the pharmacological targets identified in part 1 can play in
epilepsy and its symptoms.
CBD Ion Channel Targets in Epilepsy
While CBDs effects at TRPV channel subtypes have been
investigated with some rigor, the involvement of this subtype
of TRP channel in epilepsy remains unclear. In humans, 2
reports describe increases in TRPV1 mRNA and protein ex-
pression in small numbers of patients with mesial temporal lobe
epilepsy, tuberous sclerosis, and focal cortical dysplasia type
IIb, suggesting that TRPV1 may be involved in epilepsy [125].
However, it remains unclear whether the increased TRPV1
expression described reflects an uninvolved and downstream
consequence of disease or an integral part of the process. The
preclinical literature describing TRPV1 involvement in epilep-
sy is conflicted where studies suggest no involvement [126,
127], a proconvulsant effect, or an anticonvulsant effect of
TRPV1 activation [128134]. Overall, a larger number of stud-
ies suggest that the consequences of TRPV1 activation in epi-
lepsy are detrimental which, given the large number of validat-
ed studies asserting overwhelmingly anticonvulsant effects of
CBD in these animal models, suggests that TRPV1 activation
by CBD is not part of its antiepileptic mechanism of action. To
our knowledge, neither TRPV2 nor TRPV3 have been reported
to have any involvement in epilepsy, while only a single study
to date has suggested the involvement of TRPV4 in epilepsy
[135]. Here, TRPV4 blockade was proposed to reduce febrile
seizures in larval zebrafish [136], and so, again, the ability of
CBD to activate TRPV4 would be in direct contrast to its clear
anticonvulsant effects.
With regard to other TRP family members, only a single
preclinical report associates TRPA1 expression changes with
epilepsy without insight into any causal relationship, and no
reports linking TRPM8 and epilepsy have been made [137].
Finally, the potential conflict between TRP agonist-
mediated increases in [Ca
2+
]
i
and anticonvulsant effects of
CBD should be considered alongside the propensity for many
TRP agonists to elicit Ca
2+
-dependent desensitization, includ-
ing CBD in some cases [98,137]. However, it remains to be
seen whether a hypothetical CBD-induced TRP channel de-
sensitization achieved from steady-state CBD dosing could
contribute to the antiepileptic effects reported. In conclusion,
based on the existing evidence, the TRP channels affected by
CBD are unlikely to account for the antiepileptic effect seen.
Although the literature makes few direct links between
VDAC1 and epilepsy, and such that exist remain limited to
the hypothetical, mitochondrial dysfunction that plays a sig-
nificant role in some epilepsies [138,139]. Therefore, while
VDAC1 remains a plausible molecular target for CBD, it is
too early to assert confidently any involvement of VDAC1 in
the mechanism of action underlying CBDs antiepileptic
mechanism but should be investigated further. Related to
VDAC1 as a mitochondrial target is the effect of CBD via
PPARγ, a nuclear receptor type expressed in brain with a
significant role in regulating metabolism, immune response,
and development. However, while such a physiological role
could have profound implications for epilepsy, no specific link
between PPARγand epilepsy has yet been made, and direct
effects of CBD at (c.f., via) this target have yet to be demon-
strated [140].
While VGCCs are long-established molecular targets for
antiepileptic drugs (AEDs) such as lamotrigine,
eslicarbazepine, felbamate, levetiracetam, gabapentin, and
pregabalin, most target N-, P/Q-, and L-type VGCCs and
not the low voltage-activated T-type VGCCs which CBD
can block [141]. T-type channels are, however, blocked by a
smaller number of existing AEDs such as ethosuximide,
zonisamide, and possibly valproate, where this action is
thought to inhibit synchronized depolarization of neuronal
populations which can lead to generalized seizures [141].
Such drugs have found particular utility in the treatment of
absence seizures, while T-type VGCC mutations have been
associated with the pathogenesis of some absence epilepsies
[142]. Thus, while a definitive role for blockade of T-type
Ca
2+
channels in the antiepileptic mechanism of CBD has
not been established, both CBDs potency at this target and
the targets validity in this disease make it a plausible target for
further investigation. It is also notable that CBDseffectsatT-
type channels produces a hyperpolarizing shift of the activa-
tion potential, while changes in steady-state activation and
activity, even at strongly hyperpolarizing potentials, suggest
the ability to induce a block when the channel is in either the
open or closed state. Interestingly, while ethosuximide in-
duces a similar hyperpolarizing shift, it is only able to block
channels when in the open state [143].
CBD Receptor Targets in Epilepsy
CBD exhibits reasonable affinity at plausible concentration
for 5-HT
1A
and 5-HT
2A
receptors, and 5-HT
2A
receptors act
Molecular Targets of Cannabidiol
Author's personal copy
as a target for fenfluramine, a drug for which some evidence
exists supporting efficacy in drug-resistant epilepsies such as
Dravet syndrome [144]. A very limited number of studies
have reported changes in 5-HT receptor expression and func-
tion in people with epilepsy, although it remains unclear
whether this is a consequence of the disease or a component
of pathogenesis. Thus, while 5-HT involvement in pathogen-
esis remains uncertain, some 5-HT receptor subtypes may
represent a valid therapeutic target in epilepsy through which
CBD could be acting [145,146].
GlyR are predominantly expressed in the CNS in neuronal
cells in the brainstem and spinal cord and, as such, there is
much less evidence of their role in disorders of the cerebrum,
such as epilepsy. However, recent research in rodent species
has shown significant, functional GlyR expression in cortex
and hippocampus at least up to postnatal day 14 where they
serve to modulate neuronal network function [147], and
emerging evidence suggests a role in hyperexcitability disor-
ders [148]. These findings suggests that investigation of GlyR
function in healthy and epileptic, mature human cortex iswar-
ranted in order to lend further credence to GlyR-mediated
antiepileptic effects of CBD.
Links between μand δOR function and a variety of brain
disorders, including epilepsy, have been extensively investi-
gated [149,150]. However, OR agonism and antagonism have
each been confusingly associated with both pro- and anticon-
vulsant effects, and so make any antiepileptic therapeutic strat-
egy involving OR targets uncertain. Moreover, the relatively
high micromolar concentrations at which CBD exerts a block
of μand δOR in highly refined and optimized systems (see
Part 1) further argues against this mechanism underlying the
drugs well established antiepileptic effects [151].
GPR55 is an orphan receptor for which the endogenous
ligand lysophosphatidylinositol has been proposed where
GPR55activationtriggersreleaseofCa
2+
from intracel-
lular stores to facilitate neurotransmission in hippocampus
[51]. However, while this functional role for GPR55 in
modulating synaptic transmission has been demonstrated,
its validity as a target for modulation of hyperexcitability
disorders such as epilepsy remains unknown. CBD shows
good affinity for GPR55 and the antagonistic effects re-
ported would be consistent with an attenuation of synaptic
transmission which could be antiepileptic. However, de-
finitive evidence to support this antiepileptic mechanism
of action is required.
nAChRs have been directly implicated in a rare, familial
form of epilepsy, autosomal dominant nocturnal frontal lobe
epilepsy where mutant forms of the α4β2* subtype are widely
expressed in cortex and, more generally, α4 subunit mutations
are associated with febrile seizures [152]. However, the role of
these mutant nAChRs during brain development to produce
this epilepsy or influence febrile seizure susceptibility remains
unclear [153].Thus, while nAChRs are a plausible target for
CBD, their relevance in ictogenesis and hence the therapeutic
effect of the drug remains unknown.
CBD has also been shown to mediate effects via both A
1
and A
2
adenosine receptors in studies of peripheral tissues.
However, both receptor subtypes are present in brain, and
adenosinergic modulation of neuronal activity serves to pro-
tect against cellular damage resulting from excessive metabol-
ic demand (e.g., during seizures) in addition to many agonists
at these receptors exerting significant anticonvulsant effects in
a wide range of animal models [154]. However, exploitation
of adenosine agonists as epilepsy treatments has been unsuc-
cessful owing to the very narrow therapeutic window that can
result in significant adverse events. In particular, those affect-
ing the cardiovascular system in humans and in animal
models, which suggests that direct agonism of A
1
/A
2
recep-
tors by CBD is unlikely and, as such, indirect elevation of
local adenosine levels is not only more plausible but could
represent a viable mechanism underlying CBDsanticonvul-
sant effects [124,155].
CBD Enzyme Targets in Epilepsy
While CYP enzymes are present in the endoplasmic reticulum
of CNS cells, their expression is typically only 0.5% that seen
in hepatic tissues [156]. Moreover, while CNS CYP450 have
been shown to modulate GABA and cholesterol levels, it is
unclear whether a drug effect on CNS CYP450 would confer a
functional change in CNS function through such pathways
[156].
AANAT has not been implicated in epilepsy and so the
pertinence of this target is unknown. Moreover, while neuro-
inflammation has long been established as a hallmark conse-
quence of epilepsy and PLA2 levels are reportedly increased
in some animal models and human studies of epilepsy, the
relevance of this enzyme as a therapeutic drug target remains
uncertain [157]. Similarly, reported effects upon COX and
IDO enzymes could also be argued to exert effects upon epi-
lepsy via involvement in inflammatory processes. However,
conflicting results and the very high CBD concentrations re-
quired for such effects casts doubt over the relevance of these
targets for CBDs effects in epilepsy.
Conversely, FAAH inhibition has been shown to affect
seizure states in animal models and so represents a putative
antiepileptic target [158,159]. However, as CBD has been
claimed to inhibit and activate FAAH, conclusive elucidation
of CBDs effects at this target are required. Furthermore, as
antiepileptic effects of FAAH result in CB
1
R-mediated anti-
convulsant effects and such effects are prone to tolerance,
investigation of the persistence of FAAH-mediated interven-
tions in epilepsy is also required.
Finally, while not exerting a direct therapeutic effect, it
should be noted that CBD-induced inhibition of enzymes such
as CYP3A4, which is responsible for the hepatic metabolism
Bih et al.
Author's personal copy
of a variety of AEDs and can be induced by others [160],
could exert an indirect effect by altering circulating levels of
concurrent medications. The clinical relevance for such indi-
rect effects of CBD on circulating AED levels has recently
been highlighted in the case of clobazam co-administration
[161].
CBD Transporter Targets in Epilepsy
While CBDs reported inhibition of glutamate reuptake is en-
tirely inconsistent with is well establishedantiepileptic effects,
the inhibition of dopamine and adenosine reuptake could play
a role as increased levels of both transmitters have been asso-
ciated with anticonvulsant effects [6,162]. Nevertheless, the
high micromolar concentrations of CBD required to elicit the-
se effects for some targets suggests that further investigation
of physiologically pertinent concentrations at defined molec-
ular targets is required to clarify their involvement, if any.
CBD reportedly inhibits AEA uptake, although the specific
molecular transporter responsible for this role remains to be
identified. Moreover, in the CNS, AEA is less abundant than
the other major endocannabinoid, 2-arachidonoyl glycerol,
which renders a functional role for AEA uptake inhibition in
the antiepileptic effects of CBD uncertain [163]. Furthermore,
AEA uptake inhibition would lead to increased CB
1
Ractiva-
tion that could conceivably produce psychoactive effects that
have never been reported for CBD. Moreover, as describe
above (BCBD Enzyme Targets in Epilepsy^), CB
1
R-mediated
modulation of seizure states is prone to treatment tolerance,
which does not appear to occur with prolonged CBD treat-
ment in animal models.
Movement Disorders
Movement disorders cover a wide range of diseases that hin-
der the ability to initiate and control movement. CBD has been
shown to alleviate the symptoms of dystonia, dyskinesia, and
catalepsy resulting from antipsychotic treatment or neurode-
generative diseases [164,165].
CBD Ion Channel Targets in Movement Disorders
It is known that CBD acts on multiple subtypes of TRPV
channels, and TRPV1 agonists, such as capsaicin, have been
shown alleviate hyperkinesia in rodent models of
Huntingdons disease (HD) by restoring the neurochemical
deficits [164,165]. CBD also desensitizes the TRPV1 recep-
tor, so it is not certain if TRPV1 serves as a plausible target. At
present, neither TRPV2 nor TRPV3 receptors have been re-
ported to play a role in movement disorders. One study has
shown that TRPV4 R269H mutations are responsible for es-
sential tremors in CharcotMarieTooth disease. However,
TRPV4 does not serve as a suitable target for CBD, as high
micromolar concentrations (100 μM) are required for channel
blockade [71]. Regarding other TRP subtypes, no studies
linking TRPA1 or TRPM8 to movement disorders have been
made. Based on the data so far, it is unlikely that interactions
between CBD and TRP channels account for its
antihyperkinetic effects.
Current evidence proves that CBD changes the conduc-
tance of VDAC1 [84]. VDAC1 plays a significant role in
Parkinsons disease (PD), where it is required for the degra-
dation of mitochondria. This causes the loss of dopaminergic
neurons, leading to the movement disorders seen in PD [166].
Thus, VDAC1 may be involved in the effect of CBD in move-
ment disorders.
Activation of PPARγby pioglitazone has been reported to
reverse oral dyskinesia in rodent models of schizophrenia, but
adverse effects on memory retention have been observed
[167]. CBD has been shown to upregulate and activate
PPARγ[18,54], and therefore it may be able to replicate the
effects of pioglitazone.
CBD acts as a potent inhibitor of T-type VGCCS [95]. T-
type VGCC α1G (CACNA1G) subunit expression is signifi-
cantly increased in Purkinje cells in rodent models of ataxia,
which results in the abnormal regulation of motor function
[168]. Numerous studies on rodent models have also shown
that palatal, Parkinsonian, and essential tremors are all depen-
dent on T-type VGCC conductance [169,170]. Tremors were
abolished upon channel blockade by zonisamide, a selective
T-type VGCC antagonist [171]. T-type VGCCs are blocked
by CBD at low doses, and therefore CBD may be effective in
treating the diseases mentioned.
CBD Receptor Targets in Movement Disorders
CBD activates 5-HT
1A
and 5-HT
2A
receptor isoforms with
reasonable potency [26]. Activation of the 5-HT
1A
subtype
by buspirone has been shown to alleviate catalepsy in rodent
models of PD [172]. Similar results were observed following
CBD treatment, where the effect of CBD was abolished by 5-
HT
1A
receptor blockade, but it is not clear if CBD directly
interacted with 5-HT
1A
to exert its anticataleptic effect [50].
Despite this, the 5-HT
1A
subtype seems to be a suitable ther-
apeutic target for the anticataleptic action of CBD.
GlyRs are expressed in the CNS, where they mediate in-
hibitory neurotransmission. Disorders in glycine transmission
have been implicated in spastic disorders, for example re-
search in the 1980s showed a loss of glycine synthesis in the
spinal cord of spastic animals [173]. High micromolar con-
centrations of CBD are required to activate GlyRs; therefore,
GlyRs are not appropriate targets for CBD.
Present research shows that blockade of δORs and μORs
abolishes the extrapyramidal effects of levodopa treatments in
humans [174]. However, high micromolar concentrations of
CBD are required to antagonize either receptor subtype, so
Molecular Targets of Cannabidiol
Author's personal copy
μORs and δORs are not plausible targets for CBD [28]. There
is no evidence linking the GPR55 receptor with movement
disorders.
Α4β2andα7 nAchRs are potently inhibited by CBD [27].
As rapid desensitization of either subtype reduces dyskinesia
[174], it is possible that the blockade of these receptors by
CBD is responsible for its antidyskinetic effect.
CBD has been reported to modulate adenosinergic activity
in vitro. However, no current studies that link the A
1
receptor
with movement disorders have been undertaken. Adenosine
A
2A
receptors are highly expressed in GABAergic neurons
within the indirect pathway of the basal ganglia, where their
activation exerts an inhibitory effect on movement. Blockade
of the A
2A
receptor in Parkinsonian rats abolished the motor
deficits. The side effect profile was favorable owing to low
A
2A
receptor expression in other regions ofthe brain [111]. As
CBD activates A
2A
receptors, this interaction is unlikely to
have any beneficial effect in PD.
CBD Enzyme Targets in Movement Disorders
CBD is a potent inhibitor of CYP2D6 [60]. Following treat-
ment with various antipsychotic agents, tardive dyskinesia
appeared to be greater in patients with the poor metabolizer
phenotype ofCYP2D6, but the difference was not statistically
significant [175]. It is not clear if the inhibition of CYP2D6 by
CBD will be beneficial in preventing antipsychotic-induced
tardive dyskinesia. No other CYP450 subtypes have been in-
vestigated in movement disorders. AANAT and COX2 have
not been implicated in movement disorders; high CBD con-
centrations are required to an interaction with either enzyme,
thus they are not suitable targets. Studies investigating the use
of FAAH inhibitors totreat restless leg syndrome and periodic
limb movement disorder are ongoing, but no publications
have been released so far.
CBD Transporter Targets in Movement Disorders
CBD inhibits the uptake of glutamate, dopamine, and adeno-
sine, but only at high micromolar concentrations [105].
Impaired glutamate uptake has been reported in transgenic
rodent models of HD, but the effect of this on motor function
has not been determined [176]. Dopamine transporter avail-
ability is significantly lower in patients exhibiting symptoms
of PD [177]. Adenosine uptake has not been implicated in this
disease. Based on the data so far, inhibition of glutamate or
dopamine uptake will not be beneficial in the treatment of
movement disorder.
CBD binds an unknown AEA transporter to inhibit AEA
uptake [19,71]. Blockade of AEA uptake by UCM707
abolished hyperkinesia in rodent models of HD and reduced
spasticity in MS rats [178]. However, the elevation in AEA
could produce psychotropic effects and tolerance upon CB
1
R
activation [179], which has not been seen as a side effect of
CBD [180]. Therefore, the binding of CBD to the AEA trans-
porter will not provide a therapeutic effect in movement
disorders.
Neurodegenerative Diseases
Several preclinical studies have shown that CBD can
exert benefit in mouse models of neurodegenerative dis-
ease [181,182]. A common finding among neurodegen-
erative diseases is chronic neuroinflammation and initi-
ation of positive feedback between failure of calcium
homeostasis, mitochondrial dysfunction, and excessive
cell death, where age and oxidative stress are strong
correlates [183185]. Dementia is caused by many dif-
ferent disorders, resulting in deterioration of memory,
mental functions, and behavior. It was estimated that,
globally, in 2010, 35.6 million people were living with
itand it was predicted that this number would double
every year: Alzheimers disease (AD) is generally con-
sidered to be the most common subtype, resulting in
6080% of cases [186]. As it is associated with the
excessive build-up of β-amyloid (Aβ)and
hyperphosphorylated forms of the microtubule-
associated protein, tau, this led to the proposal of the
Bamyloid cascade hypothesis^as the underlying patho-
genic basis for this disease [187]. However, manifesta-
tion of AD can be highly heterogeneous with clear
geneenvironment interactions [188], and precipitating
are factors highly variable, that is, any insult leading
to an excessive and chronic neuroinflammatory response
(e.g., traumatic brain injury) or energy failure [189],
which is often seen in early-onset neurodegenerative
diseases, such as Leigh syndrome [190]. Genome-wide
analysis has highlighted several risk genes, many of
which alter clearance of misfolded proteins and modu-
late inflammation, while environment also influences ex-
pression via systemic inflammation (e.g., obesity),
which may make targeting risk factors and the immune
system important treatment strategies [191].
CBD Receptor Targets in Neurodegeneration
It has been suggested that the ECS may be hypofunctional,
and/or dysfunctional in neurodegenerative diseases, suggest-
ing circumstance-dependent protective or potentially damag-
ing effects (reviewed in [192,193]) and consistent with its
largely adaptive role in mammalian physiology.
In relation to CBD, McPartland et al. [16] conclude that
CBD does not act directly at CB
1
R but may act indirectly at
potentially supraphysiological concentrations by enhancing
ECS tone. However, as discussed in the transporter section
(see below), CBD may also bind tofatty acid binding proteins
Bih et al.
Author's personal copy
(FABPs), which could partly explain its apparent indirect ef-
fects on the ECS among others [194].
As indicated in Table 1, CBD exerts effects via 5-HT re-
ceptors and the serotonergic system, including the 5-HT
1A
receptor. While implicated in neurodegeneration, serotonergic
involvement is complex as both increases and decreases in
receptor density across various regions of the brain have been
associated with cognitive decline [195]. Although not strictly
a neurodegenerative disorder, CBD has been shown to be
effective in a mouse model of hepatic encephalopathy in-
duced by bile duct ligationa condition that has a large ele-
ment of neuroinflammation. In this study, mice received 5
mg/kg of CBD or vehicle for 5 weeks after ligation. Treated
mice were significantly protected, but the effect was largely
negated by co-administration of a 5-HT
1A
receptor antagonist
[196]. However, as there was no bile duct ligation control
group treated with the antagonist only, it is not possible to
determine if the antagonist was blocking CBDs effects di-
rectly, as it could simply be a behavioral effect. Similarly,
Pazos et al. [197] showed that in a pig model of hypoxic
ischemic brain injury, CBD, at a final concentration in brain
tissue of ~0.2 μM, could be protective when given after the
insult. Critically, the protective effects were blocked when a
5-HT
1A