Antioxidant function of phytocannabinoids: Molecular basis of their stability and cytoprotective properties under UV-irradiation

Article abstract of DOI:10.1016/j.freeradbiomed.2021.01.012

In this contribution, a comprehensive study of the redox transformation, electronic structure, stability and photoprotective properties of phytocannabinoids is presented. The non-psychotropic cannabidiol (CBD), cannabigerol (CBG), cannabinol (CBN), cannabichromene (CBC), and psychotropic tetrahydrocannabinol (THC) isomers and iso-THC were included in the study. The results show that under aqueous ambient conditions at pH 7.4, non-psychotropic cannabinoids are slight or moderate electron-donors and they are relatively stable, in the following order: CBD > CBG ≥ CBN > CBC. In contrast, psychotropic Δ⁹-THC degrades approximately one order of magnitude faster than CBD. The degradation (oxidation) is associated with the transformation of OH groups and changes in the double-bond system of the investigated molecules. The satisfactory stability of cannabinoids is associated with the fact that their OH groups are fully protonated at pH 7.4 (pKa is ≥ 9). The instability of CBN and CBC was accelerated after exposure to UVA radiation, with CBD (or CBG) being stable for up to 24 h. To support their topical applications, an in vitro dermatological comparative study of cytotoxic, phototoxic and UVA or UVB photoprotective effects using normal human dermal fibroblasts (NHDF) and keratinocytes (HaCaT) was done. NHDF are approx. twice as sensitive to the cannabinoids’ toxicity as HaCaT. Specifically, toxicity IC₅₀ values for CBD after 24 h of incubation are 7.1 and 12.8 μM for NHDF and HaCaT, respectively. None of the studied cannabinoids were phototoxic. Extensive testing has shown that CBD is the most effective protectant against UVA radiation of the studied cannabinoids. For UVB radiation, CBN was the most effective. The results acquired could be used for further redox biology studies on phytocannabinoids and evaluations of their mechanism of action at the molecular level. Furthermore, the UVA and UVB photoprotectivity of phytocannabinoids could also be utilized in the development of new cannabinoid-based topical preparations.

Full text of DOI:10.1016/j.freeradbiomed.2021.01.012

Free Radical Biology and Medicine 164 (2021) 258–270
Contents lists available at ScienceDirect
Free Radical Biology and Medicine
journal homepage: www.elsevier.com/locate/freeradbiomed
Original article
Antioxidant function of phytocannabinoids: Molecular basis of their
stability and cytoprotective properties under UV-irradiation
,
Jan Vaceka *, Jitka Vostalovaa, Barbora Papouskovab, Denisa Skarupovaa, Martin Kosc,
,
**
Martin Kabelacd, Jan Storchc
a Department of Medical Chemistry and Biochemistry, Faculty of Medicine and Dentistry, Palacky University, Hnevotinska 3, 775 15, Olomouc, Czech Republic
b Department of Analytical Chemistry, Faculty of Science, Palacky University, 17. Listopadu 12, 771 46, Olomouc, Czech Republic
c Department of Advanced Materials and Organic Synthesis, Institute of Chemical Process Fundamentals of the Czech Academy of Sciences, v. v. i., Rozvojova 135, 165
02, Prague 6, Czech Republic
d Department of Chemistry, Faculty of Science, University of South Bohemia, Branisovska 31, 370 05, Ceske Budejovice, Czech Republic
A R T I C L E I N F O
A B S T R A C T
Keywords:
In this contribution, a comprehensive study of the redox transformation, electronic structure, stability and
photoprotective properties of phytocannabinoids is presented. The non-psychotropic cannabidiol (CBD), can-
nabigerol (CBG), cannabinol (CBN), cannabichromene (CBC), and psychotropic tetrahydrocannabinol (THC)
isomers and iso-THC were included in the study. The results show that under aqueous ambient conditions at pH
7.4, non-psychotropic cannabinoids are slight or moderate electron-donors and they are relatively stable, in the
following order: CBD > CBG ≥ CBN > CBC. In contrast, psychotropic Δ9-THC degrades approximately one order
of magnitude faster than CBD. The degradation (oxidation) is associated with the transformation of OH groups
and changes in the double-bond system of the investigated molecules. The satisfactory stability of cannabinoids is
associated with the fact that their OH groups are fully protonated at pH 7.4 (pKa is ≥ 9). The instability of CBN
and CBC was accelerated after exposure to UVA radiation, with CBD (or CBG) being stable for up to 24 h. To
support their topical applications, an in vitro dermatological comparative study of cytotoxic, phototoxic and UVA
or UVB photoprotective effects using normal human dermal fibroblasts (NHDF) and keratinocytes (HaCaT) was
done. NHDF are approx. twice as sensitive to the cannabinoids’ toxicity as HaCaT. Specifically, toxicity IC50
Phytocannabinoids
(photo)stability
Oxidation
Electronic structure
UVA/UVB photoprotection
Keratinocytes
Dermal fibroblasts
values for CBD after 24 h of incubation are 7.1 and 12.8 μM for NHDF and HaCaT, respectively. None of the
studied cannabinoids were phototoxic. Extensive testing has shown that CBD is the most effective protectant
against UVA radiation of the studied cannabinoids. For UVB radiation, CBN was the most effective. The results
acquired could be used for further redox biology studies on phytocannabinoids and evaluations of their mech-
anism of action at the molecular level. Furthermore, the UVA and UVB photoprotectivity of phytocannabinoids
could also be utilized in the development of new cannabinoid-based topical preparations.
1. Introduction
The main phytocannabinoid is THC, Δ9-tetrahydrocannabinol (Scheme
1), whose application leads not only to modulation of the endocanna-
The endocannabinoid system is one of the major receptor-driven
mechanisms leading to the maintenance of homeostasis. We currently
have relatively extensive knowledge of the structure and function of the
two cannabinoid receptors, CB1 and CB2, which play a key role in the
endocannabinoid system. Clearly the best known exogenous cannabi-
noid receptor ligands are phytocannabinoids, which are lipophilic
phenolic terpenoids that have been isolated from Cannabis sativa [1,2].
binoid system, but also manifests a psychotropic effect [3].
C. sativa contains more than 180 phytocannabinoids, with pharma-
cological effects observed for many of them. In this sense, cannabidiol
(CBD) is probably the best known member of the group of so-called non-
psychotropic cannabinoids [4,5], without adverse effects associated
with abuse, nor of prolonged use producing behavioral change, memory
loss or generally chronic intoxication. Other intensively investigated
* Corresponding author.
** Corresponding author.
E-mail addresses: jan.vacek@upol.cz (J. Vacek), storchj@icpf.cas.cz (J. Storch).
https://doi.org/10.1016/j.freeradbiomed.2021.01.012
Received 29 September 2020; Received in revised form 10 December 2020; Accepted 6 January 2021
Available online 13 January 2021
0891-5849/© 2021 Elsevier Inc. All rights reserved.

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
phytocannabinoids include cannabinol, cannabichromene, cannabigerol
(Scheme 1), or cannabivarin and their derivatives or analogues [3]. For
more details on the phytochemical profile of Cannabis sativa, see
Ref. [6].
primarily in the fields of neurological therapy, chronic pain treatment,
dermatology, complementary treatment for cancer therapy and eating
disorders [11]. Their immunomodulatory and anti-inflammatory activ-
ities are promising properties of phytocannabinoids [10].
From the point of view of the real application of phytocannabinoids,
it is necessary to distinguish between the administration of a complex
extract and the pure (isolated) substance. The application form also
plays an important role. Phytocannabinoids are most commonly
administered topically, orally, sublingually, rectally as suppositories or
by inhalation [7]. When complex mixtures of phytocannabinoids are
applied, synergistic or more often additive effects can be achieved
compared to the application of only one phytocannabinoid. It is
important to note that phytocannabinoids do not only act by binding to
cannabinoid receptors, but also modulate the function of transient re-
ceptor potential (TRP) channels, peroxisome proliferator-activated re-
ceptor gamma (PPARγ), cyclooxygenase-2 (COX-2), etc., and are
involved in calcium homeostasis. The pleiotropic, not exclusively
receptor-driven, actions of phytocannabinoids are frequently discussed
today [8].
A great deal of attention has been paid to the in vitro evaluation of the
effects of phytocannabinoids, often without describing their mode of
action at the molecular level. Besides this, it is important to note that
there is limited number of highly valuable mechanistic, pharmacological
and clinical studies. As for the pleiotropic and biphasic mode of action
and frequently discussed entourage effect of cannabinoids, the
involvement of them in oxidative stress and cannabinoid-based antiox-
idant effects have recently begun to be discussed [12]. It was demon-
strated that CBD protects the liver by inhibiting oxidative and nitrative
stress [13] and attenuating inflammatory response [14]. In contrast,
prooxidant effects of synthetic cannabinoids have been described [15].
An overview of the antioxidant action of CBD was recently reported
[16]. However knowledge on the redox biology of other phytocanna-
binoids is very limited.
In the development of new drugs, the synthetic THC analogue
nabilone (marketed as Cesamet) has been applied for the treatment of
nausea and vomiting associated with cancer chemotherapy. Similarly,
dronabinol (or Marinol) has been introduced into clinical practice as an
anti-emetic agent and later as an appetite stimulant. Probably the best
known cannabinoid-based drug is known as nabiximols (trade name
Sativex), which is a mixture of THC and CBD. Recently, a CBD antiepi-
leptic drug known as Epidiolex has also been approved by the FDA. The
extensive testing of CBD’s safety and its introduction into the category of
drugs opens up a wide field for other preparations containing non-
psychotropic cannabinoids [3,9,10]. In recent decades, attention has
also been paid to the development of synthetic derivatives, CB receptor
agonists/antagonists and modulators [3]. In addition, phytocannabi-
noids are also used in the development of new medical devices, cos-
metics or in the development of dietary supplements and in other
selected applications for prophylaxis and complementary therapy. Based
on our knowledge of natural cannabinoids, these substances can be used
2. Material and methods
2.1. Chemicals and general methods
Cannabidiol (CBD) and cannabigerol (CBG) were purchased from
CBDepot Ltd., both at 99% purity. Other commercially available
reagent-grade materials were used as obtained from Sigma-Aldrich,
Acros Organics, and TCI. All solvents (Lach-Ner) were of reagent
grade and used without any further purification. Thin-layer chroma-
tography (TLC) was performed on silica gel 60 F254-coated aluminum
sheets, and compounds were visualized with UV light (254 nm) or
phosphomolybdic acid. Column chromatography was performed using a
Biotage HPFC system (Isolera One) with prepacked silica gel flash col-
umns. The standard Schlenk technique was used for all reactions. 1H and
13C NMR spectra were recorded using a Bruker Avance at 400 MHz (1H
NMR) and 101 MHz (13C NMR). Chemical shifts (δ) are reported in parts
per million (ppm) relative to TMS, and referenced to residuals of CDCl3
Scheme 1. Chemical structures of tested phytocannabinoids.
259

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
(δ = 7.26 and 77.00 ppm, respectively). The coupling constants (J) are
given in Hertz (Hz) with corresponding multiplicity (s = singlet, d =
doublet, t = triplet, q = quartet, m = multiplet).
μ
m) from Kernet Int. (UK). After polishing, the electrode was rinsed
thoroughly with deionized water. The analyses were performed with
Britton-Robinson buffer (titrated to the desired pH with 0.2 M NaOH) at
room temperature with a μAutolab III analyzer (EcoChemie, NL) in a
2.1.1. Preparation of phytocannabinoids
three-electrode setup with a Ag/AgCl/3 M KCl electrode as the reference
and a platinum wire as the auxiliary electrode. Argon was used to
remove oxygen from the supporting electrolyte. Individual settings for
respective voltammetric analyses are given in the Figure legends.
Cannabichromene (CBC) [17] A 100 mL round-bottom flask
equipped with a Dean-Stark trap was charged with 3.00 g (16.64 mmol)
of olivetol, 2.99 mL (17.48 mmol) of citral, 1.76 mL (16.64 mmol) of
tert-butylamine, and 33 mL of toluene. The reaction mixture was
refluxed for 5 hours. After completion, the solvent was removed under
reduced pressure. Flash chromatography using PE/EtOAc (95:5) pro-
vided 3.90 g (12.40 mmol, 74.5%) of CBC as a yellow oil (99% purity).
NMR spectra were in accordance with published data [18].
2.3. Theoretical calculations
Since the cannabinoids are flexible molecules due to the presence of
aliphatic side-chains, a two-stage Monte Carlo approach in the dihedral
space implemented in the program FROG2 [25] was used as the first step
in retrieving the energetically most stable structures of the given com-
pounds. The twenty most favorable conformers found were further
optimized at the DFT level of theory employing the 6-311++G(d, p)
basis set and wB97XD functional, which provides reasonable geometries
and thermodynamics data [26]. The frontier molecular orbitals and
molecular electrostatic potentials were analyzed using the program
Avogadro [27]. The presence of solvent (water, methanol and n-octanol)
was described implicitly using the PCM model [28].
Cannabinol (CBN) [19] A 100 mL round-bottom flask was charged
with 2.00 g (6.36 mmol) of CBD, 3.23 g (12.72 mmol) of iodine, and 50
mL of toluene. The reaction mixture was refluxed for 7 hours. After
completion, the mixture was washed with a saturated solution of
Na2S2O3 and brine. The organic phase was dried over anhydrous MgSO4.
After evaporation of the solvent, flash chromatography using PE/EtOAc
(94:6) provided 0.69 g (2.22 mmol, 35%) of CBN as a yellow oil (99%
purity). NMR spectra were in accordance with published data [20].
Δ8-Tetrahydrocannabinol (Δ8-THC) [21] A Schlenk flask was
charged with 2.00 g (6.36 mmol) of CBD and 33 mL of DCM. The so-
A harmonic vibrational frequency analysis was performed to confirm
that the structures found are the minima at the potential energy surface.
The vertical and adiabatic ionization potentials were calculated by
subtracting the energies of the compound and its cation at the same level
of theory as the previous ab initio calculations were performed. All
quantum mechanical calculations were treated in the program
Gaussian16 [29].
◦
lution was cooled to 0 C, and 0.32 mL (2.54 mmol) of BF3.Et2O was
added dropwise under an Ar atmosphere. The mixture was allowed to
warm to r.t. After an additional hour of stirring, the mixture was poured
into saturated NaHCO3 and extracted with DCM. The combined organic
phases were dried over anhydrous MgSO4. After evaporation of the
solvent, flash chromatography using pentane/Et2O (95:5) provided
1.12 g (3.56 mmol, 56%) of Δ8-THC as a yellow oil (99% purity). NMR
spectra were in accordance with published data [21].
2.3.1. Dissociation of the cannabinoids
Δ9-Tetrahydrocannabinol (Δ9-THC) and Δ8-iso-tetrahydrocan-
nabinol (iso-THC) [22,23] A Schlenk flask was charged with 2.00 g
(6.36 mmol) of CBD and 33 mL of DCM. The solution was cooled to
ꢀ 10 ◦C and 0.32 mL (2.54 mmol) of BF3.Et2O was added dropwise under
an Ar atmosphere. The mixture was stirred at ꢀ 10 ◦C for 120 min. The
reaction was quenched by the addition of saturated NaHCO3. After
extraction with DCM, the combined organic phases were dried over
anhydrous MgSO4. After evaporation of the solvent, flash chromatog-
raphy using pentane/Et2O (97:3) provided 1.04 g (3.31 mmol, 52%) of
Δ9-THC as a yellow oil (99% purity) and 156 mg (0.496 mmol, 7.8%) of
iso-THC as a yellow oil (98% purity). NMR spectra of Δ9-THC are in
accordance with published data [24]. NMR spectra of iso-THC: 1H NMR
(400 MHz, CDCl3) δ 6.28 (d, J = 1.5 Hz, 1H), 6.12 (d, J = 1.5 Hz, 1H),
4.99 (q, J = 1.5 Hz, 1H), 4.93 (s, 1H), 4.54 (s, 1H), 3.46 (q, J = 3.1 Hz,
1H), 2.51–2.40 (m, 2H), 2.34 (s, 1H), 1.93–1.83 (m, 4H), 1.80–1.53 (m,
7H), 1.42–1.21 (m, 7H), 0.92–0.86 (m, 3H). 13C NMR (101 MHz, CDCl3)
δ 157.39, 152.25, 146.07, 142.62, 111.02, 110.78, 107.91, 105.98,
74.63, 43.03, 35.70, 35.45, 31.58, 30.76, 30.48, 29.41, 27.89, 22.65,
22.55, 21.05, 14.01; for more details, see Supplementary Information.
For in vitro cell experiments, methanol (HiPerSolv CHROMANORM
for HPLC, LC-MS grade) was from VWR International s.r.o. (Czech Re-
public). Dulbecco’s modified Eagle’s medium (DMEM), Ham-F12
nutrient mixture, heat-inactivated fetal calf serum (FCS), stabilised
penicillin-streptomycin solution, amphotericin B, hydrocortisone,
adenine, insulin, epidermal growth factor, 3,3’,5-triiodo-L-thyronine,
trypsin, ampicillin, trypsin-EDTA (0.25%), dimethyl sulfoxide (DMSO),
neutral red (NR), and other chemicals were from Sigma-Aldrich (Czech
Republic).
The calculations of pKa were performed based on the procedure [30],
where a corresponding thermodynamics cycle can be found, with the
modification of Pliego [31]. Based on Eq. (1), the calculation of pKa can
be done through the following equations (2) and (3):
HA + H2O ↔ A- + H3O+
ΔGg = Gg(Aꢀ ) + Gg(H3O+) ꢀ Gg(HA) ꢀ Gg(H2O)
(1)
(2)
ΔG = ΔGg + Gsolv(Aꢀ ) + Gsolv(H3O+) ꢀ Gsolv(HA) ꢀ Gsolv(H2O)
(3)
Where Gg corresponds to the chemical potential of the given compound,
and Gsolv to its solvation free energy.
ΔG
RT
Thus, pKa =
ꢀ log[H2O] with [H2O] = 55.49 m ol/dm 3.
(4)
Since the calculated solvation free energy of H3O+ is the main source
of error in this approach, this value was replaced by an experimental one
equal to ꢀ 110.2 kcal/mol [31]. To minimize errors, the pKa values
obtained for HA via equation (4) were scaled by a factor of 1.0165,
corresponding to the ratio between the experimentally found pKa of
phenol (9.88) and the calculated one (9.72). All calculations were per-
formed using the B3LYP/6-311+G(d,p) level of theory. Solvent effects
were considered through the continuum SMD model of water [32].
Besides this, the program Marvin (Marvin 20.8.0, 2020, ChemAxon, http
://www.chemaxon.com) from the software package ChemAxon, which
uses quantitative structure property relationships (QSPRs) based on
various chemical descriptors and is fragment-based, was used for
comparison.
2.2. Electrochemical measurement
2.4. (Photo)stability measurement
The substances were analyzed using square-wave voltammetry
(SWV) with the working electrode being a glassy carbon electrode (GCE,
1 mm diameter disc, BASi, USA). Before each electrochemical experi-
ment, the GCE was polished using diamond spray (particle size was 3
To evaluate the stability/photostability of the studied compounds,
stock solutions in DMSO were diluted in a mixture of phosphate buffers
(50 mM, pH 7.4) with methanol (2:1, v:v). The final concentration of
compounds was 20 μM and of DMSO 2% (v/v). Immediately after the
260

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
preparation of cannabinoid solution, the spectrum scan was done using
an UV-VIS spectrophotometer (UV–2401PC; Shimadzu, Japan) with a
wavelength range of 250–450 nm in a quartz cuvette with a 1-cm path
length. To determine the effect of oxygen on the (photo)stability of
cannabinoids, solutions of the tested compounds were divided into four
aliquots immediately after their preparation and sealed in plastic flasks
with a plugged cap. Two flasks were bubbled with argon (conditions
without oxygen).
incubation, cells were washed twice with PBS and then PBS supple-
mented with glucose (PBS-G; 1 mg/mL) was applied. Randomly, one
plate was then exposed to a non-cytotoxic dose of UVA radiation (5.0 J/
cm2 for NHDF and 7.5 J/cm2 for HaCaT) using a SOL 500 solar simulator
(Dr. Hoenle Technology, Germany) equipped with an H1 filter trans-
mitting at wavelengths of 320–400 nm. The intensity of UVA radiation
was evaluated before each irradiation with a UVA-meter (Dr. Hoenle
Technology, Germany). During irradiation, the plate was incubated on a
cold panel to limit overheating. The second (non-irradiated) plate was
incubated in the dark for the period of irradiation. After UVA exposure,
PBS-G was discarded and serum-free medium was applied. After 24 h
(37 ◦C, 5% CO2), cell damage was evaluated by NR incorporation into
viable cells. The phototoxic effect was evaluated as the % of viability of
control cells.
Two flasks (with and without oxygen) were exposed to UVA radia-
tion (20 J/cm2). A SOL 500 solar simulator equipped with a H1 filter
transmitting at wavelengths of 320–400 nm was used as the source of
UVA light. The UVA output was measured with an UVA-meter (Dr.
¨
Honle UV Technology, Germany). During the irradiation, the flasks were
placed on ice-cold panels to eliminate heating of the solution and
possible thermal decomposition. In parallel during the irradiation, two
flasks (with and without oxygen) were incubated in the dark. Immedi-
ately and 24 h after UVA irradiation, the absorption spectrum was
scanned using an UV-VIS spectrophotometer in the same way as samples
without UVA irradiation in the presence or absence of oxygen. In par-
allel with the spectral evaluation, electrochemical analyses of samples
(sec. 2.2.) and LC-MS analysis (sec. 2.6.) were performed.
2.5.4. UVA and UVB photoprotection potential of test compounds
Two plates with NHDF or HaCaT were pre-treated with CBD, CBC,
CBG and CBN (0.78–6.25 μmol/L) and with DMSO (0.5%, v/v) in serum-
free DMEM for 1 h. Control cells were treated with serum-free medium
containing DMSO (0.5%, v/v) under the same conditions. After incu-
bation, cells were washed twice with PBS and then PBS supplemented
with glucose (1 mg/ml) was applied. Randomly, one plate was then
exposed to a cytotoxic dose of UVA radiation (7.5 J/cm2 for NHDF and
10.0 J/cm2 for HaCaT) using a SOL 500 solar simulator (Dr. Hoenle
Technology, Germany) equipped with a H1 filter transmitting at wave-
lengths of 320–400 nm. During irradiation, the plate was incubated on a
cold panel to limit overheating. The second (non-irradiated) plate was
incubated in the dark for the period of irradiation. After irradiation,
serum-free DMEM was applied to both cells (irradiated and non-
irradiated), and cells were then incubated for 24 h (37 ◦C, 5% CO2).
To study the photoprotective effect against UVB, the plate was
exposed to a cytotoxic dose of UVB radiation (150 mJ/cm2) using the
solar simulator equipped with a H2 filter transmitting at wavelengths of
295–320 nm. The other manipulations with cells were the same. The
intensity of UVA or UVB radiation was evaluated before each irradiation
with a UVA- or UVB-meter (Dr. Hoenle Technology, Germany). Cell
damage was evaluated by NR incorporation into viable cells according
to the following equation:
2.5. Toxicity, phototoxicity and photoprotection
2.5.1. Cell cultures
Normal human dermal fibroblasts (NHDF) were obtained from the
skin fragments of medically healthy adult donors. The tissue specimens
were obtained from patients undergoing plastic surgery at the Depart-
ment of Plastic and Aesthetic Surgery (University Hospital, Olomouc).
The use of skin tissue complied with the Ethics Committee of the Uni-
versity Hospital in Olomouc and Faculty of Medicine and Dentistry,
Palacky University, Olomouc (date: 6.4.2009, ref. number: 41/09). All
patients had given their written informed consent. The skin fragments
and NHDF were cultured as described earlier [33]. NHDF were used
between the 2nd and 4th passage. For experiments, cells were seeded on
96-well collagen-coated plates at a density of 0.5 × 105 cells per cm2.
Experiments were performed in four independent repetitions with the
use of cells from four donors to minimize the individual sensitivity of
donor cells.
⃒
⃒
⃒
⃒
⃒
⃒
⃒
⃒
As ꢀ Anc
Protection (%)
=
100 ꢀ
⋅100
A spontaneously transformed aneuploid immortalized human kera-
tinocyte cell line (HaCaT) was obtained from CLS (Eppeheim, Germany).
HaCaT was grown in culture medium consisting of Dulbecco’s Modified
Eagle Medium (DMEM) supplemented with fetal calf serum (10%, v/v),
penicillin (100 mg/mL) and streptomycin (100 U/mL). For experiments,
cells (between the 50th and 60th passage) were seeded on 96-well plates
at a density of 1.0 × 105 cells per cm2. Experiments were performed in
four independent repetitions.
Apc ꢀ Anc
As … absorbance of sample (cells pre-incubated with test compounds
in serum-free medium and irradiated)
Anc … absorbance of negative control (cells pre-incubated with
DMSO in serum-free medium and non-irradiated, i.e. incubated in
the dark)
Apc … absorbance of positive control (cells pre-incubated with
DMSO in serum free medium and irradiated).
2.5.2. Cytotoxicity of test compounds
NHDF or HaCaT were treated with CBD, CBC, CBG and CBN
(0.78–100 μmol/L) and with DMSO (0.5%, v/v) in serum-free DMEM for
24 h. Control cells were treated with serum-free medium containing
2.6. LC-MS method
◦
DMSO (0.5%, v/v) under the same conditions. After 24 h (37 C, 5%
CO2), cell damage was evaluated by NR incorporation into viable cells
[34]. Medium was discarded and NR solution (0.03%, w/v, PBS) was
applied. After 60 min, the NR solution was discarded, cells were fixed
with a mixture of formaldehyde (0.5%, v/v) and CaCl2 (1%, w/v) in a 1:1
ratio, and then NR was dissolved in methanol (50%, v/v) with acetic acid
(1%, v/v). After 5 min of intensive shaking, absorbance was measured at
540 nm.
UHPLC/MS analyses were performed in an ACQUITY I-Class UPLC
system (Waters, Milford, MA, USA) equipped with a binary solvent
manager, sample manager and column manager. Kinetex Polar C18
(100 × 2.1 mm, i.d. 2.6 μm; Phenomenex, CA, USA) was chosen as the
analytical column. The system was set for binary gradient elution at a
◦
flow rate of 0.6 mL/min and a temperature of 25 C. Mobile phase A
consisted of 0.1% formic acid in water, mobile phase B was 0.1% formic
acid in acetonitrile. The gradient profile started at: 0–11 min 50–70% B,
11–12.5 min 70–100% B, 12.5–13 min 100-50% B, 13–16 min 50% B.
2.5.3. Phototoxicity of tested compounds
Two plates with NHDF or HaCaT were pre-treated with CBD, CBC,
The injection volume was 2 μL.
CBG and CBN (0.78–100
μ
mol/L) and with DMSO (0.5%, v/v) in serum-
The method development took place in a Xevo TQ-S mass spec-
trometer (Waters), a triple quadrupole mass spectrometer with ESI ion
source operated in positive mode. Multiple reaction monitoring (MRM)
free DMEM for 1 h. Control cells were treated with serum-free medium
containing DMSO (0.5%, v/v) under the same conditions. After
261

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
transitions were adapted from Ref. [35]. The capillary voltage was set to
3.0 kV and the sampling cone 30 V. The source temperature and the
the UPLC system via an electrospray ionization (ESI) interface. The
conditions were adapted from the Xevo TQ-S method. The ion source
operated in positive ionization mode with a capillary voltage of 3.0 kV
and a sampling cone at 30 V. The source temperature and the des-
olvation temperature were set to 120 ◦C and 320 ◦C, respectively. The
cone and desolvation gas flows were 150 L/h and 900 L/h, respectively.
The data acquisition range was from 50 to 1200 Da with a scan time of
◦
◦
desolvation temperature were set to 120 C and 320 C, respectively.
The cone and desolvation gas flows were 150 L/h and 900 L/h,
respectively.
For analysis of oxidation products, a high-resolution Synapt G2-S
Mass Spectrometer (Waters Corp., Manchester, UK) was connected to
Fig. 1. The most stable conformer (CFM) with electrostatic potential (ESP) and HOMO/LUMO distributions of each of the studied cannabinoids.
262

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
0.2 s in MSE mode (scan functions enabling the simultaneous acquisition
of low-collision-energy (2 eV) and high-collision-energy (15–30 eV)
mass spectra in a single experiment).
narrow range of 0.3 eV around ꢀ 8 eV (see Table 1 and Supplementary
Information), in agreement with previously published data [37,38] and
independently of the type of solvent used. Molecular electrostatic po-
tential (ESP) surfaces of cannabinoids can be found in Fig. 1. All oxygen
atoms possess a negative potential (red) both in aromatic phenol as well
as in ether groups. An excess of negative charge can also be found on
carbon atoms participating in isolated double bonds and carbon atoms of
the phenyl ring of CBN. Positive potential (blue) is found on all hydrogen
atoms, especially in the phenol ring.
The instrument was calibrated using adducts of sodium formate in
acetonitrile, and the corrections for accurate mass measurement were
achieved using an external reference Leucine-enkephalin (20 μg/L in
mixture of water:acetonitrile:formic acid (100:100:0.2), flow rate of 5
μ
L/min). The LC-MS system was controlled with MassLynx V4.1 (Wa-
ters), and the post-acquisition processing of the data was performed
using the MetaboLynx XS Application Manager (Waters). Concerning
the accurate mass measurement, MS data outlying the range of ±5 ppm
were not considered.
3.2. Redox behavior – oxidative transformation and acido-basic
properties
3. Results and discussion
Oxidation of the cannabinoids was investigated using SWV at a GCE.
At pH 7.4, all test substances undergo an electrochemical transformation
around the potential Ep +0.5 V vs. Ag/AgCl/3 M KCl (Fig. 2). Based on
previously published results, we can designate the investigated mole-
cules as moderately-effective electron-donors, i.e. substances with a
slight reducing potential. Electrochemical analyses confirmed previ-
ously published results on the antiradical and antioxidant properties of
cannabinoids. An example is a robust study in which the authors
monitored the antiradical capacity of Δ9-THC, CBD and C. sativa extracts
using spectrophotometric as well as electrochemical methods [39]. The
results indicate that it is highly dependent on the experimental condi-
tions and the method chosen. If, for example, CBD is studied in an oil
environment, it may also have a higher antioxidant potential than
Here, we focused on the complex investigation of cannabidiol (CBD),
cannabigerol (CBG), cannabinol (CBN), cannabichromene (CBC), Δ8-
tetrahydrocannabinol (Δ8-THC), Δ9-tetrahydrocannabinol (Δ9-THC),
and iso-Δ8-tetrahydrocannabinol (iso-THC), see Scheme 1. The confor-
mational variability, electronic structures, electron-donor and acido-
basic properties were studied using voltammetric and DFT calculation
approaches. The stability profiles of the phytocannabinoids based on
their ability to be oxidized is described together with the photo-
degradation processes by using voltammetry, UV-VIS spectrophotom-
etry and liquid chromatography-mass spectrometry (LC-MS). Finally,
the cytotoxicity and protective properties of CBD, CBC, CBG and CBN
were tested using UVA- and UVB-irradiated spontaneously immortalized
human keratinocyte cell line (HaCaT) and normal human dermal fi-
broblasts (NHDF).
3.1. Conformational variability and electronic structure
The potential energy surface scan of investigated cannabinoids
shows that, unlike the rigid cyclic part of the molecule, the aliphatic side
chains of these molecules are very flexible and there are a large number
of conformers that only differ in energy by a tiny amount. This means
that besides the structure of the global minimum, several other geom-
etries of the given compound will be non-negligibly populated at room
temperature. The most stable conformers are those where a favorable
C–H ...π interaction appears, which is clearly visible in the structure of
the global minimum of CBG, where a strong interaction between the
terpene side chain and aromatic ring occurs. Whereas in the study [36]
the authors considered the conformers containing aliphatic side chains
in an all-trans position, we found that those are typically less stable by
3–12 kJ/mol than the global minimum. It means that they do not appear
among the geometries of the ten most stable conformers listed in Sup-
plementary Information.
Fig. 2. Square-wave voltammograms of cannabinoids in Britton-Robinson
buffer at pH 7. The analyzed compound (20 μM) was accumulated at open
The DFT computational approach shows that for all the studied
circuit potential for 60 s prior to each analysis. The oxygen was removed from
the supporting electrolyte with an argon stream. The SW voltammetric scan was
performed from ꢀ 0.3 to +1.5 V at a frequency of 200 Hz. SWV records for
alkaline and acidic pH can be found in Supplementary Information.
compounds, the HOMOs and LUMOs correspond to the π and π* orbitals
of the benzene ring (Fig. 1). The delocalization of these orbitals is often
expanded over neighboring parts of the compound containing double
bonds. The HOMO energy values of all compounds can be found in a
Table 1
HOMO-LUMO characteristics and HOMO-LUMO gap (all in eV) of the most stable conformer of the studied cannabinoids in different solvents. All calculations were
performed at the wB97XD/6-311++G(d, p) level of theory.
Solvent
Water
Methanol
HOMO
Octanol
HOMO
Compound
HOMO
LUMO
LUMO-HOMO
LUMO
LUMO-HOMO
LUMO
LUMO-HOMO
CBC
ꢀ 7.768
ꢀ 8.160
ꢀ 8.159
ꢀ 7.858
ꢀ 7.988
ꢀ 8.083
ꢀ 8.045
0.704
1.070
1.056
0.556
1.032
1.045
1.045
8.472
9.230
9.215
8.414
9.020
9.128
9.090
ꢀ 7.842
ꢀ 8.139
ꢀ 8.168
ꢀ 7.844
ꢀ 7.972
ꢀ 8.072
ꢀ 8.055
0.687
1.047
1.072
0.566
1.029
1.070
1.072
8.529
9.186
9.240
8.410
9.001
9.142
9.127
ꢀ 7.711
ꢀ 7.942
ꢀ 8.089
ꢀ 7.686
ꢀ 7.825
ꢀ 7.910
ꢀ 7.925
0.807
1.018
1.012
0.727
1.011
1.048
1.045
8.518
8.960
9.101
8.413
8.836
8.958
8.970
CBD
CBG
CBN
Δ8-THC
Δ9-THC
iso-THC
263

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
α
-tocopherol [40]. From the point of view of electrochemical analysis,
Table 3
pKa values for different cannabinoids calculated with program Marvin and based
on ab initio approach.
the oxidation of highly effective low-molecular weight antioxidants
usually occurs at Ep lower than +0.2 V (vs. Ag/AgCl/3 M KCl). This was
recently demonstrated for selected vitamins, their conjugates and trolox,
under identical experimental conditions as in this study [41]. The OH
group or groups occurring on the benzene skeleton of the investigated
molecules participate in the anodic reaction (Scheme 1). With Δ9-THC,
-1 e-/-1 H+ oxidation to form a quinone product was described [42].
The oxidation of cannabinoids is associated, similarly to other
phenolic compounds, with the formation of radical forms, which leads to
the formation of passivating oligo- or polymeric structures on the elec-
trode surface [43]. The anodic reaction is accompanied by adsorption
phenomena in the aqueous medium. This can be used for the adsorptive
accumulation of the cannabinoids on the electrode surface before the
measurement. Cannabinoids are lipophilic molecules, and this facilitates
their interaction with electrode materials or materials designed to pre-
concentrate them [44,45].
Compound
Marvin
Ab initio
CBC
9.47
9.43
9.46
9.32
9.32
9.34
9.86
10.26
10.32
9.82
CBD
CBG
CBN
9.73
Δ8-THC
Δ9-THC
iso-THC
10.48
10.51
10.84
3.3. Stability and photodegradation processes
Stability data related to the long-term storage of medical cannabis,
the decomposition of cannabis oils and pharmaceuticals in solid form
can be found in the literature, see the stability of dronabinol capsules
[47]. Stability data are also available for the purposes of forensic anal-
ysis [48]. In terms of basic research and biochemical studies, no
comprehensive comparative study on the stability and reactivity of
cannabinoids has been published yet. In this study, we focus on the
In this study, SWV experiments were performed at pH 4, 7.4 and 9
(Fig. 2 and Figs. S1-S2 in Supplementary Information). The electro-
chemical conversion of cannabinoids is a pH-dependent process, where
with increasing pH there is a shift in Ep towards less positive potentials,
see Table 2 and Fig. S3 in Supplementary Information. In an alkaline
environment, an accelerated degradation of cannabinoids as well as
other phenolic derivatives can occur in the presence of oxygen. This is
related to the deprotonation of the investigated molecules. With can-
nabinoids, they are only deprotonated in the pH range of 9–10. At
physiological pH, cannabinoids are practically in a protonated state,
which prevents their oxidative degradation (more about the degradation
of cannabinoids below). The calculated pKa values for all studied can-
nabinoids are given in Table 3. The program Marvin predicts the pKa
values of cannabinoids in very narrow range of 0.1, treating the can-
nabinoids as slightly stronger acids than phenol. The ab initio pKa values
are distributed across a larger interval between 9.7-10.8, determining
the cannabinoids to be either comparable (CBG, CBN) or slightly weaker
acids than phenol. Such values agree well with experimentally found
ones published in Ref. [46].
For a more detailed (mechanistic) insight into the reactivity and
oxidative transformations of cannabinoids, we used DFT computational
approaches (Fig. 1). Electron abstraction during cannabinoid oxidation
corresponds to the distribution of HOMO and its energy levels, which in
the aqueous environment range between ꢀ 7 and ꢀ 8 eV (Table 1) and
not only correspond to the measured oxidation potentials (Table 2), but
also to the ionization potentials (Tab. S1 in Supplementary Information)
of the investigated molecules. The relative small HOMO-LUMO gaps and
the lowest ionization potentials were observed for CBN and CBC. In
addition to the fact that the above-mentioned methods allow us to
quantify the electron-donor capacity of cannabinoids, primarily elec-
trochemical SWV analysis can also be used to evaluate their oxidative
stability. The oxidative stability of cannabinoids could be an important
parameter for the development of new cannabinoid preparations. This
will be discussed in the following section.
Table 2
Oxidation potentials (Ep values) for cannabinoids analyzed by SWV in Britton-
Robinson buffer of different pH levels (vs. Ag/AgCl/3 M KCl). For other de-
tails, see Fig. 2.
Fig. 3. Stability of cannabinoids (20
μM) based on measurement of SWV
Compound
pH 4
pH 7.4
pH 9
oxidation peak height over time. The stability evaluation for non-psychotropic
cannabinoids and THC isomers are shown in panels A and B (n = 3). Both the
incubation step and SWV analysis were performed in 0.1 M phosphate buffer
(pH 7.4) with methanol (2:1, v/v). Prior to each analysis, the analyzed com-
pound was accumulated at open circuit potential for 60 s. The oxygen was
removed from the supporting electrolyte with an argon stream. The SW vol-
tammetric scan direction was as follows: from ꢀ 0.3 to +1.5 V. A frequency of
200 Hz was used.
CBC
0.708
0.701
0.673
0.717
0.721
0.698
0.716
0.524
0.506
0.492
0.511
0.511
0.516
0.547
0.432
0.403
0.390
0.425
0.415
0.421
0.431
CBD
CBG
CBN
Δ8-THC
Δ9-THC
iso-THC
264

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
stability and photostability of cannabinoids at pH 7.4 in an aqueous
environment. The results obtained by SWV measurements are shown in
Fig. 3. In these experiments, a decrease in the oxidation peaks (Fig. 2) of
cannabinoids was observed, the decrease of which generally indicates
oxidative degradation of the investigated substances.
performed under the same conditions as for the long-term stability study
(Fig. 3), only the sample was irradiated with UVA at the beginning of the
stability experiment with a single dose of 20 J/cm2. The sample was
analyzed immediately after irradiation (0 h), and also 24 h after the
irradiation, when it was stored in the dark. SWV analysis showed that
the CBD and CBG samples were stable under the chosen experimental
procedure. Moderate photodegradation was observed in CBN and CBC
(Fig. 4). CBC is known to be photochemically unstable and undergoes
[2+2] photocycloaddition to form cannabicyclol [50]. However, when
interpreting the results, we must assume that the SWV analysis corre-
sponds to oxidative degradation and is not an absolute method, mainly
due to the fact that some photodegradation products may also be elec-
troactive or undergo interfering adsorption processes. For this reason,
we applied other methods such as UV-VIS spectrophotometry and LC-MS
analysis. The UV-VIS spectra of the tested substances are shown in
Fig. 5A. Well-developed spectra can only be obtained for CBC and CBN.
Thus, after irradiation, decreases in the absorbance of both substances at
280 nm, which is related to the presence of double bonds, were inves-
tigated. Concretely, 10–20% decreases in the absorbance of CBC and
CBN were observed in the irradiated samples (Fig. 5B and C) in accor-
dance with the electrochemical analysis (Fig. 4A,D). The LC-MS method
was also used to confirm the photodegradation after 24 h, where the
chromatographic peak of the respective substances was monitored, for
details see Figs. S4 and S5 (Supplementary Information). The results
show increased degradation for CBC and CBN consistent with the elec-
trochemical and UV-VIS spectral analyses (Fig. 5D). In general CBD,
which is most commonly used in cosmetics and other topical pharma-
ceuticals today, exhibits good stability, whether or not it has been
irradiated with UVA radiation. The stability of all substances was
For non-psychotropic cannabinoids, the stability study was per-
formed at pH 7.4 for 216 h, in the dark, at room temperature and in the
presence of molecular (atmospheric) oxygen. Oxidation stability de-
creases in the following order: CBD > CBG ≥ CBN > CBC (Fig. 3A). The
reduced stability for CBN and CBC corresponds well with lower HOMO
energy levels and ionization potentials (Tabs. 1 and S1 in Supplementary
Information), which are closely related to the electron-donor capacity of
the investigated cannabinoids. Furthermore, the results are consistent
with a lower stability of CBC in a slightly acidic environment, where
[4+2] cycloaddition occurs to form cannabicitran [49]. The substances
are stable within 1 h, then there is a gradual decrease in the height of
oxidation SWV peaks depending on the kinetics of decomposition
(oxidation) of the examined samples. These measurements are based on
the assumption that if the OH groups of cannabinoids are oxidized
during the stability study, they can no longer be further oxidized in the
electrochemical experiment. After 24 hours, there was a 20–40%
decrease in the oxidation peaks of non-psychotropic cannabinoids. The
stability of the Δ8-THC and iso-THC was observed under the same
experimental conditions, and was comparable to the stability of CBG and
CBN (Fig. 3B). Unlike other cannabinoids, the Δ9- THC was significantly
less stable, and its degradation after 24 h corresponded to approx. 50%.
After three days of incubation, Δ9-THC could no longer be detected.
In further experiments, we focused on a study of the photo-
degradation of non-psychotropic cannabinoids (Fig. 4). Analyses were
Fig. 4. Changes in SWV oxidation peaks of CBC (A), CBD (B), CBG (C) and CBN (D) after UVA irradiation (n = 3). The cannabinoids (20 μM) were dissolved in 0.1 M
phosphate buffer (pH 7.4) with methanol (2:1, v/v). The samples were analyzed by SWV immediately (0 h) and 24 h after UVA irradiation (20 J/cm2). The samples
were incubated in the dark at room temperature in the presence of atmospheric oxygen. For more details on SWV, see Fig. 3.
265

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
Fig. 5. UV-VIS spectra of cannabinoids (50
μ
M) in 0.1 M phosphate buffer (pH 7.4) with methanol (2:1, v/v) (A). UVA-photostabilty of 20
μ
M CBC (B) and CBN (C)
expressed as % of absorbance changes at 280 nm immediately and 24 hours after UVA irradiation. (D) UVA-photostability of cannabinoids (20
μ
M) measured as total
ion current MS response after 24 h of incubation. The incubation of samples was done in the dark at room temperature and in the presence of atmospheric oxygen (n
= 3, for panels B–D).
investigated under ambient conditions in the presence of atmospheric
oxygen. If oxygen is displaced from the solution, some degradation
processes can be significantly suppressed, which corresponds to analyses
of other hydroxy-substituted aromatic compounds [51].
was expressed as the half-maximal cytotoxic concentration IC50
see Table 4. NHDF were more sensitive than HaCaT. The same con-
centration range (0.78–100 M) was used for evaluating the phototoxic
(μM),
μ
potential of the tested compounds. As shown Table 5, the non-toxic dose
of UVA radiation did not accelerate the toxicity of the parent com-
pounds. These results show that the studied compounds have no
phototoxic potential. The higher values of IC50 (lower toxic effect) found
in the phototoxic experiment is associated with a 1-h treatment with the
studied compounds compared to cytotoxicity testing (Table 4), which
was evaluated after the 24-h treatment.
3.4. Cytoprotective effects under UV irradiation
Based on the spectra of the individual cannabinoids (Fig. 5A), it is
clear that CBN and CBC could have a (direct) protective effect against
irradiation in the UVB region, in contrast to CBG and CBD, which do not
absorb at the corresponding wavelengths. Although CBG and CBD do not
show the typical spectrum, they were also included in the testing,
because both cannabinoids can be used in dermal applications. Firstly,
Based on the cytotoxicity findings, non-toxic concentrations of the
studied compounds were used for the UVA- and UVB-photoprotective
experiments (0.781–6.25 μM). For all the tested compounds, a higher
viability (amount of incorporated NR) of UVA- or UVB-irradiated cells
pre-treated with the studied cannabinoids was observed compared to
non pre-treated ones (Fig. 6). In general, the tested compounds exhibited
higher UVA and UVB protection of NHDF than of HaCaT. Our results
the toxicity and phototoxicity of the cannabinoids (0.78–100 μM) were
assessed on dermal cells on primary human dermal fibroblasts (NHDF)
and a cell line of human keratinocytes (HaCaT). HaCaT were used
instead of primary keratinocytes in the pilot screening for phototoxic
and/or photoprotective properties. The toxicity was dose dependent and
Table 5
UVA phototoxicity (IC50 values;
μM) of non-psychotropic cannabinoids on
Table 4
normal human dermal fibroblasts (NHDF) and human keratinocyte cell line
Cytotoxicity (IC50 values;
μM) of non-psychotropic cannabinoids on normal
(HaCaT), n = 4.
human dermal fibroblasts (NHDF) and human keratinocyte cell line (HaCaT), n
= 4.
Compound
NHDF
HaCaT
Compound
NHDF
HaCaT
non-irradiated
irradiated
non-irradiated
irradiated
CBN
CBC
CBG
CBD
7.15 ± 0.65
4.73 ± 0.54
5.88 ± 0.48
7.15 ± 0.65
12.98 ± 1.35
10.18 ± 1.03
16.94 ± 1.74
12.85 ± 1.17
CBN
CBC
CBG
CBD
14.64 ± 1.24
20.75 ± 1.98
21.23 ± 2.05
12.97 ± 1.01
18.74 ± 1.24
20.61 ± 1.55
24.10 ± 2.25
19.18 ± 1.47
12.07 ± 1.10
20.74 ± 1.65
16.21 ± 1.37
10.74 ± 1.00
16.32 ± 1.23
20.61 ± 1.45
20.14 ± 1.99
16.93 ± 1.36
266

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
Fig. 6. UVA (A) and UVB (B) protectivity of non-psychotropic cannabinoids on normal human dermal fibroblasts (NHDF) and human keratinocytes (HaCaT), n = 4.
For more details, see sec. 2.5.4.
demonstrate the photoprotective effects of other phytocannabinoids,
with exception of CBD, on human skin fibroblasts and keratinocytes
(HaCaT) over the entire solar region of UV radiation (UVA and UVB) for
the first time.
shown to affect binding affinity to CB1 and CB2 receptors and PPARγ
[55]. The selected quinone (VCE-003) was proposed as a candidate for
anti-inflammatory effect testing [56]. The oxidation products of can-
nabinoids also show modulation of the activity of liver biotransforma-
tion enzymes [56]. More information on the stability of cannabinoids in
selected formulations can be found in the following studies [57–59].
As for the antioxidant action of phytocannabinoids, they most likely
interact directly with ROS or other reactive molecules to a limited
extent. There is clear evidence that the endocannabinoid system affects
the redox balance of cells [57–59]. The CB1 and CB2 receptors can be
involved in the elimination but also in the production of free radicals,
depending on intrinsic cell status and external stimuli. Most of the
studies targeting redox balance and the endocannabinoid system are
based on the application of CB1 and/or CB2 receptor agonists or antag-
onists in vitro after the application of toxic compounds (often oxidants)
3.5. Biological relevance of the results and discussion
The obtained physicochemical characteristics of the studied canna-
binoids (sec. 3.1. and 3.2) are in good agreement with their stability and
reactivity, regardless of whether the samples were UV-irradiated or not
(sec. 3.3). The THC:CBD ratio plays a key role in practically used
preparations [52]. However, their different degradation kinetics (Fig. 3)
should be taken into account when interpreting experiments (but also
clinical trials) where both THC and CBD are involved. Cannabinoids are
also used in many topical preparations and their (photo)stability cannot
be neglected, because the effects studied in the experiments may not be
related to the biological effect of the parent cannabinoids, but to their
degradation or biotransformation products. It has been shown that the
biological activity of oxidized CBD may be different from the parent
molecule [53], for example, the oxidation product of CBD inhibited
or induction of inflammation using
a lipopolysaccharide-based
approach [60]. In addition, CB1 and CB2 receptors and also other
membrane receptors (PPARγ, TRP, GPR, serotonin (5HT1A) and adeno-
sine (A2A) receptors) and channels are involved in redox homeostasis
[16]. The mode of action involving the above-mentioned receptors is
based on cannabinoid ligand binding, which results in the increase or
decrease in the level of antioxidant or pro-oxidant enzymes. This
mechanism is also called ‘indirect’ or receptor-driven. In addition to
this, there is also the hypothesis that cannabinoids can act as ‘direct’
lipophilic radical scavenging species. The results presented here support
this indirect mechanism of protection due to a moderate or relatively
low electron-donor ability of phytocannabinoids in comparison to other
(established) low-molecular radical scavengers or antioxidants [41,
61–63].
topoisomerase II
α and β, which was not observed for CBD itself. The
degradation of cannabinoids is most likely based on oxidative trans-
formation, which was confirmed electrochemically (Fig. 3). The reac-
tivity of oxidation products (most probably quinones) should be
investigated in further studies with a focus on other redox processes,
including reductive transformations. The electrochemistry of Δ9-THC
metabolites can be found here [54], data on the other cannabinoids are
not available. As for transformation to quinones, a CBG quinone deriv-
ative was identified in a structure-activity relationship study. For this
derivative, oxidative modification of the resorcinol moiety has been
In the context of dermal cytoprotection (sec. 3.4) CBD induces the
267

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
synthesis of the cytoprotective enzyme heme oxygenase 1 (HMOX1) in
keratinocytes in an Nrf2-independent manner. This was accompanied by
an increase in nuclear export and proteasomal degradation of the Nrf2
transcriptional repressor Bach1, and the expression of cytokeratins 16
and 17 in keratinocytes [64]. The fact that CBD induced the expression
of several Nrf2 target genes should be the main mechanism for its
photoprotective action, and could be extrapolated to the biological ac-
tivities of other studied phytocannabinoids. In addition, it was reported
that CBD reduced the redox imbalance caused by exposure to UVB/hy-
drogen peroxide in keratinocytes, estimated by superoxide anion radical
generation, total antioxidant status and consequently lipid peroxidation
product(s) level [65]. The protective effect of CBD on viability of kera-
tiocytes and melanocytes [66] and skin fibroblasts [67] following UV
irradiation was also recently shown.
CBC is shown to be accelerated after exposure to UVA radiation, with
CBD (or CBG) being highly stable in the 24 h experiment. To gain a more
complete view of phytocannabinoids and their topical applications, we
paid attention to an in vitro study of their cytotoxic, phototoxic and
photoprotective effects. Primary cultures of human skin fibroblasts and
the keratinocyte cell line were used for this purpose. NHDF are
approximately twice as sensitive to cannabinoids as HaCaT. Specifically,
IC50 values for CBD after 24 h of incubation are 7.1 and 12.8 μM for
NHDF and HaCaT, respectively. Phototoxicity of cannabinoids was
excluded in both fibroblasts and keratinocytes. Extensive testing has
shown that CBD is the most effective cytoprotectant after the UVA
irradiation of skin cells. As for UVB photoprotective effects, CBN was the
most effective.
Also, CBD stimulated ROS generation in mitochondria that was
accompanied by apoptosis in monocytes [68]. Antioxidant vs.
pro-oxidant effects and their mutual balance or the localization of their
action is in accordance with the pleiotropic action of cannabinoids.
However, we should always perceive the pleiotropic effects of canna-
binoids with regard to the fact that the primary molecular targets are
CB1 and/or CB2 and other receptors. The interaction with other cellular
components will depend on the cell type and the concentration of
applied cannabinoids, etc. This is in accordance with the biphasic effect
of phytocannabinoids and the phenomena that we refer to in the liter-
ature as entourage effects. Phytocannabinoids are redox-active mole-
cules, and also their interaction with other redox-active ligands (or
transition metals) is one of the possible mechanisms of action that
should be considered [69].
Author statement
J.Va. performed electrochemical experiments, J.Vo. and D.S. per-
formed in vitro cell experiments, M.Ka. designed and performed DFT
calculations, M.Ko. and J.S. purified and synthesized phytocannabi-
noids, and J.Va., M.K. and J.Vo. wrote the paper.
Declaration of competing interest
J.S. has financial interest in CB21 Pharma Ltd., CBDepot Ltd. and
PharmaCan Ltd. J.V. is a president of the scientific board of CB21
Pharma Ltd. All other authors declare that they have no conflicts of
interest with the contents of this article.
In summary, phytocannabinoids are relatively weak or moderate
electron donors, and therefore may be involved to a limited extent in the
‘direct’ scavenging of free radicals and the elimination of oxidizing
agents. However, their action as low-molecular weight antioxidants is
significantly limited by their relatively low bioavailability and strictly
dependent on the environment in which they act, e.g. the cytoplasm vs.
cell membranes. Based on the available data, we conclude that the
antioxidant or (photo)/cytoprotective action of phytocannabinoids will
primarily occur ‘indirectly’ by their interaction with specific receptors
and the activation of stress (redox-sensitive) protection signalling
pathways (the Nrf2 pathway). For a more detailed understanding of the
in vivo antiradical and antioxidant properties of phytocannabinoids, it
will be necessary to conduct more extensive research focused mainly on
the ‘direct’/‘indirect’ antioxidant effect of cannabinoids in skin, but also
in the GIT. In any case, the designation of cannabinoids as ‘effective low-
molecular weight antioxidants’ should not be overused. Instead, we
should think about phytocannabinoids as ‘redox-active modulators of
homeostatic mechanisms of the cell’.
Acknowledgements
The authors gratefully acknowledge the financial support of UPOL,
RVO 61989592 and internal grant IGA_LF_2020_022, from the Ministry
of Education, Youth and Sports. Synthetic work was supported by the
project OP EIC Operational Programme (CZ.01.1.02/0.0/0.0/16_084/
0010374). Computational resources were provided by CESNET
LM2015042 and CERIT Scientific Cloud LM2015085, provided under
the programme ‘Projects of Large Research, Development, and Innova-
tion Infrastructures’. The authors are indebted to Nina Kucharikova MSc
(for in vitro cell experiments), Zita Skolarova DiS (for electrochemical
analyses) and Ben Watson-Jones MEng (for linguistic assistance). We
wish to thank Lumir O. Hanus DSc (Hebrew University of Jerusalem,
Israel) for critical reading.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.freeradbiomed.2021.01.012.
4. Conclusions
References
Despite the fact that phytocannabinoids contain hydroxyl groups
bound to double-bonded carbons, they are compounds which are rela-
tively stable under ambient conditions at pH 7.4 in an aqueous medium.
CBD exhibits the highest stability, CBG, CBN and CBC are less stable. The
degradation of these non-psychotropic phytocannabinoids is very slow
in an aqueous medium, and thus does not distort short-term experiments
on the order of hours. In long-term experiments, it is necessary to take
into account the degradation kinetics of individual substances. As for the
two phytocannabinoids that are the most pharmacologically active and
most used in practice, psychotropic Δ9-THC degrades approximately one
order of magnitude faster than CBD. The high stability of phytocanna-
binoids is associated with the fact that their molecules are fully pro-
tonated at pH 7.4 (pKa for the investigated cannabinoids is ≥ 9).
Phytocannabinoids, especially CBD, are now massively applied in
topical form as cosmetics or selected medical devices. Also for this
reason, we investigated the photostability properties of non-
psychotropic CBD, CBG, CBN and CBC. The degradation of CBN and
[1] Y. Gaoni, R. Mechoulam, Isolation, structure, and partial synthesis of an active
constituent of hashish, J. Am. Chem. Soc. 86 (1964) 1646–1647.
[2] F. Santavy, Notes on the structure of cannabidiol compounds, Acta Univ. Palacki.
Olomuc. Fac. Med. 35 (1964) 5–9.
[3] R. Mechoulam, L.O. Hanus, R. Pertwee, A.C. Howlett, Early phytocannabinoid
chemistry to endocannabinoids and beyond, Nat. Rev. Neurosci. 15 (2014)
757–764.
[4] A.A. Izzo, F. Borrelli, R. Capasso, V. Di Marzo, R. Mechoulam, Non-psychotropic
plant cannabinoids: new therapeutic opportunities from an ancient herb, Trends
Pharmacol. Sci. 30 (2009) 515–527.
[5] S. Pisanti, A.M. Malfitano, E. Ciaglia, A. Lamberti, R. Ranieri, G. Cuomo, M. Abate,
G. Faggiana, M.C. Proto, D. Fiore, C. Laezza, M. Bifulco, Cannabidiol: state of the
art and new challenges for therapeutic applications, Pharmacol. Ther. 175 (2017)
133–150.
[6] C.M. Andre, J.F. Hausman, G. Guerriero, Cannabis sativa: the plant of the thousand
and one molecules, Front. Plant Sci. 7 (2016) 1–17.
[7] M.O. Bonn-Miller, C.V. Pollack Jr., D. Casarett, R. Dart, M. ElSohly, L. Good,
M. Guzman, L.O. Hanus, K.P. Hill, M.A. Huestis, E. Marsh, S. Sisley, N. Skinner,
J. Spahr, R. Vandrey, E. Viscusi, M.A. Ware, D. Abrams, Priority considerations for
medicinal cannabis-related research, Cannab. Cannabinoid Res. 4 (2019) 139–157.
268

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
[8] A. Ligresti, L. De Petrocellis, V. Di Marzo, From phytocannabinoids to cannabinoid
receptors and endocanabinoids: pleiotropic physiological and pathological roles
through complex pharmacology, Physiol. Rev. 96 (2016) 1593–1659.
[9] D. Friedman, J.A. French, M. Maccarrone, Safety, efficacy, and mechanisms of
action of cannabinoids in neurological disorders, Lancet Neurol. 18 (2019)
504–512.
[37] A. Kumer, N. Sarker, S. Paul, A. Zannat, The theoretical prediction of
thermophysical properties, HOMO, LUMO, QSAR and biological indics of
cannabinoids (CBD) and tetrahhdrocannabinol (THC) by computational chemistry,
Adv. J. Chem. A 2 (2019) 190–202.
[38] M.N. Sarker, A. Kumer, M.J. Islam, S. Paul, A computational study of
thermophysical, HOMO, LUMO, vibrational spectrum and UV-visible spectrum of
cannabicyclol (CBL), and cannabigerol (CBG) using DFT, Asian J. Nanosci. Mater. 2
(2019) 439–447.
[10] T.W. Klein, Cannabinoid-based drugs as anti-inflammatory therapeutics, Nat. Rev.
Immunol. 5 (2005) 400–411.
[11] K.P. Hill, Medical marijuana for treatment of chronic pain and other medical and
psychiatric problems A clinical review, JAMA, J. Am. Med. Assoc. 313 (2015)
2474–2483.
[39] A.C.M. Hacke, D. Lima, F. de Costa, K. Deshmukh, N. Li, A.M. Chow, J.A. Marques,
R.P. Pereira, K. Kerman, Probing the antioxidant activity of delta-9-
tetrahydrocannabinol and cannabidiol in Cannabis sativa extracts, Analyst 144
(2019) 4952–4961.
[12] G.W. Booz, Cannabidiol as an emergent therapeutic strategy for lessening the
impact of inflammation on oxidative stress, Free Radic. Biol. Med. 51 (2011)
1054–1061.
[40] Tura, M.; Mandrioli, M.; Toschi, T. G. Preliminary study: comparison of antioxidant
activity of cannabidiol (CBD) and alpha-tocopherol added to refined olive and
sunflower oils. Molecules 24; 2019.
[13] L.L. Yang, R. Rozenfeld, D.F. Wu, L.A. Devi, Z.F. Zhang, A. Cederbaum,
Cannabidiol protects liver from binge alcohol-induced steatosis by mechanisms
including inhibition of oxidative stress and increase in autophagy, Free Radic. Biol.
Med. 68 (2014) 260–267.
[41] E. Vavrikova, V. Kren, L. Jezova-Kalachova, M. Biler, B. Chantemargue,
M. Pyszkova, S. Riva, M. Kuzma, K. Valentova, J. Ulrichova, J. Vrba, P. Trouillas,
J. Vacek, Novel flavonolignan hybrid antioxidants: from enzymatic preparation to
molecular rationalization, Eur. J. Med. Chem. 127 (2017) 263–274.
[42] M.A. Balbino, M.M.T. de Menezes, I.C. Eleoterio, A.A. Saczk, L.L. Okumura, H.
M. Tristao, M.F. de Oliveira, Voltammetric determination of delta 9-THC in glassy
carbon electrode: an important contribution to forensic electroanalysis, Forensic
Sci. Int. 221 (2012) 29–32.
[14] P. Mukhopadhyay, M. Rajesh, B. Horvath, S. Batkai, O. Park, G. Tanchian, R.
Y. Gao, V. Patel, D.A. Wink, L. Liaudet, G. Hasko, R. Mechoulam, P. Pacher,
Cannabidiol protects against hepatic ischemia/reperfusion injury by attenuating
inflammatory signaling and response, oxidative/nitrative stress, and cell death,
Free Radic. Biol. Med. 50 (2011) 1368–1381.
[15] B.M. Fonseca, M. Almada, M. Santos, H. Gaspar, F. Carvalho, N. Teixeira,
G. Correia-da-Silva, Prooxidant effects of synthetic cannabinoids on endometrial
cells, results in a strenous rebound adaptative effect, Free Radic. Biol. Med. 120
(2018). S134-S134.
[43] M. Pyszkova, M. Biler, D. Biedermann, K. Valentova, M. Kuzma, J. Vrba,
J. Ulrichova, R. Sokolova, M. Mojovic, A. Popovic-Bijelic, M. Kubala, P. Trouillas,
V. Kren, J. Vacek, Flavonolignan 2,3-dehydroderivatives: preparation, antiradical
and cytoprotective activity, Free Radic. Biol. Med. 90 (2016) 114–125.
[44] R. Nissim, R.G. Compton, Absorptive stripping voltammetry for cannabis detection,
Chem. Cent. J. 9 (2015).
[16] S. Atalay, I. Jarocka-karpowicz, E. Skrzydlewskas, Antioxidative and anti-
inflammatory properties of cannabidiol, Antioxidants 9 (2020) 21.
[17] M.A. Elsohly, E.G. Boeren, C.E. Turner, Constituents of Cannabis sativa L. An
improved method for the synthesis of dl-cannabichromene, J. Heterocycl. Chem.
15 (1978) 699–700.
[45] I. Novak, M. Mlakar, S. Komorsky-Lovric, Voltammetry of immobilized particles of
cannabinoids, Electroanalysis 25 (2013) 2631–2636.
[46] J. Mazina, A. Spiljova, M. Vaher, M. Kaljurand, M. Kulp, A rapid capillary
electrophoresis method with LED-induced native fluorescence detection for the
analysis of cannabinoids in oral fluid, Anal. Methods 7 (2015) 7741–7747.
[47] M.F. Wempe, A. Oldland, N. Stolpman, T.H. Kiser, Stability of dronabinol capsules
when stored frozen, refrigerated, or at room temperature, Am. J. Health Syst.
Pharm. 73 (2016) 1088–1092.
[18] H.S. Yeom, H. Li, Y. Tang, R.P. Hsung, Total syntheses of cannabicyclol,
clusiacyclol A and B, iso-eriobrucinol A and B, and eriobrucinol, Org. Lett. 15
(2013) 3130–3133.
[19] F. Pollastro, D. Caprioglio, P. Marotta, A.S. Moriello, L. De Petrocellis,
O. Taglialatela-Scafati, G. Appendino, Iodine-Promoted aromatization of p
-Menthane-Type phytocannabinoids, J. Nat. Prod. 81 (2018) 630–633.
[20] Y. Li, Y.J. Ding, J.Y. Wang, Y.M. Su, X.S. Wang, Pd-catalyzed C-H lactonization for
expedient synthesis of biaryl lactones and total synthesis of cannabinol, Org. Lett.
15 (2013) 2574–2577.
[48] R.M. White, Instability and poor recovery of cannabinoids in urine, oral fluid, and
hair, Forensic Sci. Rev. 30 (2018) 33–49.
[49] X. Li, Y.R. Lee, Efficient and novel one-pot synthesis of polycycles bearing cyclols
by FeCl3-promoted [2 + 2] cycloaddition: application to cannabicyclol,
cannabicyclovarin, and ranhuadujuanine A, Org. Biomol. Chem. 12 (2014)
1250–1257.
[21] S. Kang, K. Davidsen, L. Gomez-Castillo, H. Jiang, X. Fu, Z. Li, Y. Liang, M. Jahn,
M. Moussa, F. Dimaio, L. Gu, COMBINES-CID: an efficient method for de Novo
engineering of highly specific chemically induced protein dimerization systems,
J. Am. Chem. Soc. 141 (2019) 10948–10952.
[50] L. Crombie, R. Ponsford, A. Shani, B. Yagnitinsky, R. Mechoulam, Hashish
components. Photochemical production of cannabicyclol from cannabichromene,
Tetrahedron Lett. 9 (1968) 5771–5772.
[22] Y. Gaoni, R. Mechoulam, Hashish-VII, The isomerization of cannabidiol to
tetrahydrocannabinols, Tetrahedron 22 (1966) 1481–1488.
[51] J. Vacek, J. Ulrichova, B. Klejdus, V. Simanek, Analytical methods and strategies in
the study of plant polyphenolics in clinical samples, Anal. Methods 2 (2010)
604–613.
[23] Y. Gaoni, R. Mechoulam, The ISO - tetrahydrocannabinols, Isr. J. Chem. 6 (1968)
679–690.
[24] L.J. Cheng, J.H. Xie, Y. Chen, L.X. Wang, Q.L. Zhou, Enantioselective total
synthesis of (-)-delta-THC and (-)-delta-THC via catalytic asymmetric
hydrogenation and S NAr cyclization, Org. Lett. 15 (2013) 764–767.
[25] M.A. Miteva, F. Guyon, P. Frog2 Tuffery, Efficient 3D conformation ensemble
generator for small compounds, Nucleic Acids Res. 38 (2010) W622–W627.
[26] L. Goerigk, S. Grimme, A thorough benchmark of density functional methods for
general main group thermochemistry, kinetics, and noncovalent interactions, Phys.
Chem. Chem. Phys. 13 (2011) 6670–6688.
[52] N. Jikomes, M. Zoorob, The cannabinoid content of legal cannabis in Washington
state varies systematically across testing facilities and popular consumer products,
Sci. Rep. 8 (2018) 4519.
[53] J.T. Wilson, C.A. Fief, K.D. Jackson, S.L. Mercer, J.E. Deweese, HU-331 and
oxidized cannabidiol act as inhibitors of human topoisomerase IIalpha and beta,
Chem. Res. Toxicol. 31 (2018) 137–144.
[54] M. Renaud-Young, R.M. Mayall, V. Salehi, M. Goledzinowski, F.J.E. Comeau, J.
L. MacCallum, V.I. Birss, Development of an ultra-sensitive electrochemical sensor
for delta-9 -tetrahydrocannabinol (THC) and its metabolites using carbon paper
electrodes, Electrochim. Acta 307 (2019) 351–359.
[27] M.D. Hanwell, D.E. Curtis, D.C. Lonie, T. Vandermeerschd, E. Zurek, G.
R. Hutchison, Avogadro: an advanced semantic chemical editor, visualization, and
analysis platform, J. Cheminf. 4 (2012).
[55] A.G. Granja, F. Carrillo-Salinas, A. Pagani, M. Gomez-Canas, R. Negri, C. Navarrete,
M. Mecha, L. Mestre, B.L. Fiebich, I. Cantarero, M.A. Calzado, M.L. Bellido,
J. Fernandez-Ruiz, G. Appendino, C. Guaza, E. Munoz, A cannabigerol quinone
alleviates neuroinflammation in a chronic model of multiple sclerosis,
J. Neuroimmune Pharmacol. 7 (2012) 1002–1016.
[28] G. Scalmani, M.J. Frisch, Continuous surface charge polarizable continuum models
of solvation. I. General formalism, J. Chem. Phys. 132 (2010) 114110.
[29] M.J. Frisch, G.W. Trucks, H.B. Schlegel, et al., Gaussian 16, Revision B.01,
Gaussian, Inc., Wallingford CT, 2016.
[30] M.D. Liptak, G.C. Shields, Accurate pKa calculations for carboxylic acids using
Complete Basis Set and Gaussian-n models combined with CPCM continuum
solvation methods, J. Am. Chem. Soc. 123 (2001) 7314–7319.
[31] J.R. Pliego, Thermodynamic cycles and the calculation of pKa, Chem. Phys. Lett.
367 (2003) 145–149.
[56] K. Watanabe, N. Usami, I. Yamamoto, H. Yoshimura, Inhibitory effect of
cannabidiol hydroxy-quinone, an oxidative product of cannabidiol, on the hepatic
microsomal drug-metabolizing enzymes of mice, J. Pharmacobio-Dyn 14 (1991)
421–427.
[57] S.C. Chew, K.L. Nyam, Oxidative stability of microencapsulated kenaf seed oil
using Co-extrusion Technology, J. Am. Oil Chem. Soc. 93 (2016) 607–615.
[58] V. Raikos, G. Duthie, V. Ranawana, Denaturation and oxidative stability of hemp
seed (Cannabis sativa L.) protein isolate as affected by heat treatment, Plant Foods
Hum. Nutr. (Dordr.) 70 (2015) 304–309.
[32] A.V. Marenich, C.J. Cramer, D.G. Truhlar, Universal solvation model based on
solute electron density and on a continuum model of the solvent defined by the
bulk dielectric constant and atomic surface tensions, J. Phys. Chem. B 113 (2009)
6378–6396.
[33] A. Rajnochova-Svobodova, B. Zalesak, D. Biedermann, J. Ulrichova, J. Vostalova,
Phototoxic potential of silymarin and its bioactive components, J. Photochem.
Photobiol., B 156 (2016) 61–68.
[59] S. Uluata, N. Ozdemir, Antioxidant activities and oxidative stabilities of some
unconventional oilseeds, J. Am. Oil Chem. Soc. 89 (2012) 551–559.
[60] C. Lipina, H.S. Hundal, Modulation of cellular redox homeostasis by the
endocannabinoid system, Open Biol. 6 (2016) 150276.
[34] G. Repetto, A. del Peso, J.L. Zurita, Neutral red uptake assay for the estimation of
cell viability/cytotoxicity, Nat. Protoc. 3 (2008) 1125–1131.
[61] J. Egea, I. Fabregat, Y.M. Frapart, et al., European contribution to the study of ROS:
a summary of the findings and prospects for the future from the COST action
BM1203 (EU-ROS), Redox Biol. 13 (2017) 94–162.
[35] T. Beres, L. Cernochova, S. Cavar Zeljkovic, S. Benicka, T. Gucky, M. Bercak,
P. Tarkowski, Intralaboratory comparison of analytical methods for quantification
of major phytocannabinoids, Anal. Bioanal. Chem. 411 (2019) 3069–3079.
[36] R.S. Borges, J. Batista, R.B. Viana, A.C. Baetas, E. Orestes, M.A. Andrade, K.
M. Honorio, A.B.F. da Silva, Understanding the molecular aspects of
tetrahydrocannabinol and cannabidiol as antioxidants, Molecules 18 (2013)
12663–12674.
[62] E. Hermankova, M. Zatloukalova, M. Biler, R. Sokolova, M. Bancirova, A.G. Tzakos,
V. Kren, M. Kuzma, P. Trouillas, J. Vacek, Redox properties of individual quercetin
moieties, Free Radic. Biol. Med. 143 (2019) 240–251.
[63] A.M. Chiorcea-Paquim, T.A. Enache, E. De Souza Gil, A.M. Oliveira-Brett, Natural
phenolic antioxidants electrochemistry: towards a new food science methodology,
Compr. Rev. Food Sci. Food Saf. 19 (2020) 1680–1726.
269

J. Vacek et al.
Free Radical Biology and Medicine 164 (2021) 258–270
[64] L. Casares, V. Garcia, M. Garrido-Rodriguez, E. Millan, J.A. Collado, A. Garcia-
Martin, J. Penarando, M.A. Calzado, L. de la Vega, E. Munoz, Cannabidiol induces
antioxidant pathways in keratinocytes by targeting BACH1, Redox Biol. 28 (2020)
101321.
[67] A. Gegotek, S. Atalay, P. Domingues, E. Skrzydlewska, The differences in the
proteome profile of cannabidiol-treated skin fibroblasts following UVA or UVB
irradiation in 2D and 3D cell cultures, Cells 8 (2019) 995.
[68] H.Y. Wu, C.H. Huang, Y.H. Lin, C.C. Wang, T.R. Jan, Cannabidiol induced
apoptosis in human monocytes through mitochondrial permeability transition
pore-mediated ROS production, Free Radic. Biol. Med. 124 (2018) 311–318.
[69] P. Schonhofen, L.M. de Medeiros, I.J. Bristot, F.M. Lopes, M.A. De Bastiani,
F. Kapczinski, J.A.S. Crippa, M.A.A. Castro, R.B. Parsons, F. Klamt, Cannabidiol
exposure during neuronal differentiation sensitizes cells against redox-active
neurotoxins, Mol. Neurobiol. 52 (2015) 26–37.
[65] S. Atalay, I. Dobrzynska, A. Gegotek, E. Skrzydlewska, Cannabidiol protects
keratinocyte cell membranes following exposure to UVB and hydrogen peroxide,
Redox Biol. 36 (2020) 101613.
[66] P. Gohad, J. McCoy, C. Wambier, M. Kovacevic, M. Situm, A. Stanimirovic,
A. Goren, Novel cannabidiol sunscreen protects keratinocytes and melanocytes
against ultraviolet B radiation, J. Cosmet. Dermatol. (2020), https://doi.org/
10.1111/jocd.13693 in press.
270