One-pot heterogeneous synthesis of Δ3-tetrahydrocannabinol analogues and xanthenes showing differential binding to CB1 and CB2 receptors

Article abstract of DOI:10.1016/j.ejmech.2014.07.062

Δ⁹-tetrahydrocannabinol (Δ⁹-THC) is the major psychoactive cannabinoid in hemp (Cannabis sativa L.) and responsible for many of the pharmacological effects mediated via cannabinoid receptors. Despite being the major cannabinoid scaffold in nature, Δ⁹-THC double bond isomers remain poorly studied. The chemical scaffold of tetrahydrocannabinol can be assembled from the condensation of distinctly substituted phenols and monoterpenes. Here we explored a microwave-assisted one pot heterogeneous synthesis of Δ³-THC from orcinol (1a) and pulegone (2). Four Δ³-THC analogues and corresponding Δ^4a-tetrahydroxanthenes (Δ^4a-THXs) were synthesized regioselectively and showed differential binding affinities for CB₁ and CB₂ cannabinoid receptors. Here we report for the first time the CB₁ receptor binding of Δ³-THC, revealing a more potent receptor binding affinity for the (S)-(-) isomer (hCB₁K_i = 5 nM) compared to the (R)-(+) isomer (hCB ₁K_i = 29 nM). Like Δ⁹-THC, also Δ³-THC analogues are partial agonists at CB receptors as indicated by [³⁵S]GTPγS binding assays. Interestingly, the THC structural isomers Δ^4a-THXs showed selective binding and partial agonism at CB₂ receptors, revealing a simple non-natural natural product-derived scaffold for novel CB₂ ligands.

Full text of DOI:10.1016/j.ejmech.2014.07.062

European Journal of Medicinal Chemistry 85 (2014) 77e86
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
One-pot heterogeneous synthesis of
D
3-tetrahydrocannabinol
analogues and xanthenes showing differential binding to CB1
and CB2 receptors
,
Ornelio Rosati a *, Federica Messina a, Azzurra Pelosi a, Massimo Curini a,
,
*
Vanessa Petrucci b, Jürg Gertsch b, Andrea Chicca b
a Department of Pharmaceutical Sciences, University of Perugia, Via del Liceo, 1, 06128 Perugia, Italy
b Institute of Biochemistry and Molecular Medicine, University of Bern, Bühlstrasse, 28, CH-3012 Bern, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
D
9-tetrahydrocannabinol (
and responsible for many of the pharmacological effects mediated via cannabinoid receptors. Despite
being the major cannabinoid scaffold in nature,
9-THC double bond isomers remain poorly studied. The
D
9-THC) is the major psychoactive cannabinoid in hemp (Cannabis sativa L.)
Received 18 February 2014
Received in revised form
18 July 2014
Accepted 19 July 2014
Available online 19 July 2014
D
chemical scaffold of tetrahydrocannabinol can be assembled from the condensation of distinctly
substituted phenols and monoterpenes. Here we explored a microwave-assisted one pot heterogeneous
synthesis of
tetrahydroxanthenes (
finities for CB1 and CB2 cannabinoid receptors. Here we report for the first time the CB1 receptor binding
of
D D -
3-THC from orcinol (1a) and pulegone (2). Four 3-THC analogues and corresponding D4a
D
4a-THXs) were synthesized regioselectively and showed differential binding af-
Keywords:
Cannabis sativa
D
3-THC, revealing a more potent receptor binding affinity for the (S)-(ꢀ) isomer (hCB1 Ki ¼ 5 nM)
Ytterbium triflate
Heterogeneous assisted catalysis
compared to the (R)-(þ) isomer (hCB1 Ki ¼ 29 nM). Like
D D
9-THC, also 3-THC analogues are partial ag-
onists at CB receptors as indicated by [35S]GTP
4a-THXs showed selective binding and partial agonism at CB2 receptors, revealing a simple non-natural
natural product-derived scaffold for novel CB2 ligands.
gS binding assays. Interestingly, the THC structural isomers
D
D
3-Tetrahydrocannabinol
9-Tetrahydrocannabinol
D
Tetrahydroxanthene
Cannabinoid receptor
© 2014 Elsevier Masson SAS. All rights reserved.
1. Introduction
elucidation in 1964 [3], the interest in the total synthesis of THC-
type cannabinoids [4,5] culminated with discovery of the canna-
binoid receptors CB1 and CB2 [6,7] and the numerous findings about
therapeutic and adverse effects mediated in particular via CB1 re-
ceptors [8e14]. While numerous pharmacological studies have
Tethrahydrocannabinol (THC) is a naturally occurring tricyclic
terpenophenolic natural product and the major psychoactive con-
stituent of psychoactive hemp (Cannabis sativa L.). In nature it oc-
curs in two isomeric forms,
D
D
9-tetrahydrocannabinol (
D
9-THC) and
been performed on
D
9-THC [1,2], little is known about the canna-
D
8-tetrahydrocannabinol (
8-THC) [1], both of which are psycho-
binoid receptor interaction of D
9-THC double bond isomers and
active [2]. Overall, there are 7 double bond isomers and their 30
stereoisomers of the tetrahydrocannabinol scaffold. In the cannabis
plant, THC occurs mainly as tetrahydrocannabinolic acid. In an
enzymatically catalyzed reaction, geranyl pyrophosphate and oli-
vetolic acid produce cannabigerolic acid which is cyclized by the
stereoisomers or the structurally similar non-natural analogues.
Among the various approaches to the synthesis of THC-like
cannabinoids, the condensation between differently substituted
phenols and monoterpenes was one of the first to be explored [15].
Here we employed this synthetic strategy to generate
D
3-THC an-
enzyme THC acid synthase into
moiety in the aromatic ring stems from the olivetolic acid. Over
time, or when heated, this natural product is decarboxylated to D9
D
9-THC acid. The carboxylate
alogues regioselectively using the natural phenolic secondary
metabolite orcinol (1a) and the widespread monoterpene pulegone
(2) as starting materials, exploring yield and cannabinoid receptor
binding affinity of the products.
-
THC. Since the first isolation of
D
9-THC and its seminal structure
The condensation between olivetol and 2 under acid catalysis
for the preparation of D
3-THC, in its racemic form was investigated
in the early 1940s [16]. Moreover, all the procedures indicated from
the outset several drawbacks, such as the use of benzene, poor
* Corresponding authors.
E-mail addresses: ornelio.rosati@unipg.it (O. Rosati), andrea.chicca@ibmm.
unibe.ch (A. Chicca).
http://dx.doi.org/10.1016/j.ejmech.2014.07.062
0223-5234/© 2014 Elsevier Masson SAS. All rights reserved.

78
O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
Table 1
yields, long reaction times, harsh reaction conditions and homo-
geneous catalysis [15e19]. In the past, several attempts were made
to obtain the enantiomers in a pure form. For instance, the syn-
Reaction of resorcinol 1b and pulegone 2 in the presence of optimized amount of
catalyst under conventional heating.
Solventa
T
ꢁCb
a
-ZrP-SO3H
H2SO4/SiO2
(10 mol %)a
Yb(OTf)3
thesis and isolation of (R)-(þ)
D
3-THC and (S)-(ꢀ)
D
3-THC was
(12 mol %)a
(16 mol %)a
achieved in 1984 [20] but, even though the process of base-
D
9-THC as start-
Time
(day)
Yieldc
Time
(day)
Yieldc
Time
(day)
Yieldc
catalyzed isomerization gave good yields, it used
ing material which is legally restricted and difficult to obtain. The
classic reaction of 1a and 2 mediated by phosphorus oxychloride
reported by Leaf et al. In 1942 [16] was revisited in 1968 by Chazan
and Ourisson [18] who achieved the complete isolation and iden-
tification of all reaction products. The main products obtained were
THC analogue (3a) and the tetrahydroxanthenes 4a and 5a
(Suppl. Scheme 1), with isolated yields of 25%, 15% and 10%
respectively, along with several side products.
3b
4b
3b
4b
3b
4b
DCM
DCE
D
40 R
84 R
2
2
7
2
25
35
e
3
5
2
2
35
32
e
e
2
2
45
51
5
6
101 R
80
Neat
5
a
DCM: CH2Cl2, DCE: (CH2Cl)2, D: dioxane.
R: reflux.
Isolated yields (%).
b
c
The low regioselectivity and poor yields (up to 25%) showed by
this reaction likely discouraged further exploration. Nevertheless,
to date this approach represents the simplest way to synthesize
THC analogues in a one-pot reaction. In addition, the availability of
several substituted resorcinols (i.e. olivetol and orcinol) and (R)- or
(S)-pulegone should make this approach more practical and easy.
allowed to react under conventional heating using either different
organic solvents or in neat condition (Scheme 1).
All experiments required long reaction times (up to 7 days).
Surprisingly, under conventional heating the reaction of 1b and 2
gave the corresponding
as the main product (Table 1). The desired
D D
4a-tetrahydro-1H-xanthene ( 4a-THX) 3b
From the catalytic point of view, to our knowledge,
theses by acidic heterogeneous catalysts have never been tested.
With the aim of evaluating the biological activities of
3-THC
D
3-THC syn-
D
3-THC analogue 4b was
isolated in very poor yield (5%). Since halogenated solvents gave the
best results in this preliminary test 1,2-dichloroethane was chosen
for the reaction.
When the temperature plays a crucial role in a chemical process,
more efficient heating system, such as microwave irradiation, can
be exploited to increase yields and to reduce reaction time [25,26].
The role of the temperature and a possible improvement in the
synthetic yield, the previous reaction setup was repeated under
D
and corresponding analogues, we explored the possibility to
improve the yield and handleability of their syntheses by
microwave-assisted heterogeneous catalysis.
2. Results and discussion
microwave irradiation in
(Scheme 2, Table 2).
a sealed vial in 1,2-dichloroethane
2.1. Chemistry
A general optimization of parameters was made to establish the
best reaction conditions. The hermetically closed reaction envi-
ronment allowed an evaluation of the temperature effect over the
normal solvent boiling point. It was possible to explore a range of
temperatures from 70 to 150 ꢁC, reaching a maximum inner pres-
sure of 13 atm. The reaction mixture was observed to turn into a
very dark red coloration at the end of the reaction time. No internal
cooling system was used in this instance. The optimal reaction time
resulted 120 min. When the reactions were carried out for more
than 120 h, an increase of by-products formation was observed
along with a degradation of the desired products and a poor cata-
lyst recovery. The results reported in Table 2 show an improvement
Two systems of number notation are commonly used for can-
nabinols, IUPAC dibenzopyran based numbering and monoterpene
based numbering [21] (Suppl. Fig. 1). The two systems are often
used simultaneously. For instance,
cannabinol in the dibenzopyran numeration, are also called D1
tetrahydrocannabinol and
3-THC in monoterpene numeration.
Here we refer to
9-THC and 4a-THX using the dibenzopyran
3-THC using the monoterpenoid
D D
9-THC and 6a-tetrahydro-
-
D
D
D
(IUPAC) numeration and to
numeration (Scheme 1).
D
The presence of an acid catalyst is essential to obtain a suc-
cessful condensation reaction of 2 and resorcinols for the prepa-
ration of THCelike compounds. Relying on our previous knowledge
in the yield of D
3-THC 4b as a result of the increased temperature.
Moreover, in this series, Yb(OTf)3 showed a tendency to be the best
performing catalyst and was readily available in our laboratory. In
order to verify the effect of a substituent on the reactivity and
regioselectivity of the reaction, the best reaction conditions were
applied to different resorcinol derivatives (Scheme 2).
on organic synthesis mediated by heterogeneous acid catalysts, a-
zirconium sulphophenyl phosphonate/methanphosphonate [22],
ytterbium triflate (Yb(OTf)3) [23] and sulfuric acid supported on
silica gel [24] were selected as possible solid acid catalysts for the
synthesis of
educts.
D
3-THC analogues using resorcinol (1b) and 2 as
As shown in Table 3, the reaction yield of D
3-THC analogues was
increased when C5-substituted resorcinols 1a and olivetol (1d)
were used as starting materials. Noteworthy, bioactive naturally
In order to verify and compare the catalytic efficiency of the
aforementioned solid acids, equimolar amounts of 1b and 2 were
7
2
3
6
1
8
5
1
4
8a
9a
4a
2'
5'
7
2
9
a)
5
1'
6'
3'
+
+
4
O
HO
OH
8
6
10a
3
HO
O
10
9
4'
HO
O
10
1b
2
3b
4b
Scheme 1. Reagents and conditions: a) Catalyst, solvent, reflux (see Table 1 for detailed information).

O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
79
R
R
R
a)
+
+
O
HO
OH
O
HO
O
3a-e
HO
1a R = 5-Me
1b R = H
1c R = 4-Et
1d R = 5-Pent
1e R = 2-Me
2
4a-e
Scheme 2. a) Catalyst, 1,2-dichloroehtane, Microwave Irradiation.
occurring THC analogues have an alkyl side chain at this position
[27e31].
Table 2
Reaction of resorcinol 1b and pulegone 2 in the presence of optimized amount of
Intriguingly, 5-pentylresorcinol (olivetol, 1d) was completely
unreactive under conventional heating conditions, keeping con-
stant all the other parameters, while under microwave irradiation it
catalyst under microwave irradiation in 1,2-dichloroethane.a
T
ꢁC
a
-ZrP-SO3H
H2SO4/SiO2
(10 mol %)
Yb(OTf)3
(16 mol %)
D
3-THC
(12 mol %)
Yield (%)b
3b
became the best starting material for the synthesis of the
analogues. Although the use of 1a and 1d resulted in a more
regioselective approach to the synthesis of
3-THCs, the reaction
yields (up to 23%) were not satisfactory. Under the above
mentioned reaction conditions the
4a-THX derivatives remained
Yield (%)b
Yield (%)b
D
4b
3b
4b
3b
4b
70
90
130
150
8
21
23
22
3
11
13
14
23
24
23
25
7
14
14
13
10
18
24
27
5
12
15
18
D
the main products in the case of using 5-unsubstituted resorcinols
(Table 3, entries 2 and 4).
In view of the obtained results the attention was focused on the
use of Yb(OTf)3 with the aim to improve the yields of the
analogues. Among several studies performed, the most promising
a
Reaction time: 120 min.
Isolated yields. Cooling OFF.
D
3-THC
b
Table 3
Synthesis of tetrahydro-1H-xanthene and
1,2-dichloroethane.
D
3-THC analogues from resorcinols and pulegone 2 in the presence of optimized amount of catalyst under microwave irradiation in
#
Main products
a
-ZrP-SO3H (12 mol %)
H2SO4/SiO2 (10 mol %)
Yb(OTf)3 (16 mol %)
(3) (%)a
(4) (%)a
(3) (%)a
(4) (%)a
(3) (%)a
(4) (%)a
OH
OH
1
18
10
20
12
12
17
CH
O
CH
O
3
3
3a
4a
2
3
4
15
18
15
9
19
3
18
11
19
7
11
5
27
15
19
8
23
6
HO
O
HO
O
3c
4c
OH
OH
O
(R)-(+)-4d
O
(R)-(+)-3d
HO
O
HO
O
3e
4e
Reaction temperature: 150 ꢁC. Reaction time: 120 min.
a
Isolated yields. Cooling OFF.

80
O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
was the possibility to increase the catalytic capacity of Yb(OTf)3 by
the use of a co-catalyst. The combination of two components as
catalytic system is known in organic synthesis and can produce
several advantages [32,33].
improvement in the catalytic process, leading to a relevant in-
crease of the reaction yield from 23% to 49% (see Table 4 for
compound 4d).
Moreover, the use of ascorbic acid as co-catalyst allowed to
reduce the amount of Yb(OTf)3 from 16 mol % to 2 mol %, with a
great improvement in terms of the process economy. In addition,
1 mol % of the catalyst was still sufficient to catalyze the reaction
when doubling the reaction time.
Unfortunately, despite the improvement of the reaction yields,
YTACA did not allow a quantitative process. This is not addressable
to the behavior of the catalyst but to the formation of unavoidable
side products that lead to the consumption of the starting mate-
rials. In the presence of ascorbic acid alone the reactions between
resorcinol derivatives and 2 did not take place. Also the addition of
ascorbic acid to ZrPeSO3H and H2SO4eSiO2 did not lead to any
improvement in the reaction yield.
The association of two acidic species as catalytic system has
been studied and reported in the literature [34]. The concept of
combined acids can be classified into Brønsted acid-assisted
Lewis acid (BLA), Lewis acid-assisted Lewis acid (LLA), Lewis
acid-assisted Brønsted acid (LBA), and Brønsted acid-assisted
Brønsted acid (BBA) [35]. Under these premises, different
organic and inorganic acids were tested in association with
Yb(OTf)3 and ascorbic acid was found to be the best performing.
With reference to the above mentioned combined acids theory,
Yb(OTf)3 and ascorbic acid (named YTACA for short) can be
defined as Lewis acid-assisted Brønsted acid (LBA) [36]. As
shown in Table 4, the use of YTACA produced an overall
Table 4
Synthesis of tetrahydro-1H-xanthene and tetrahydrocannabinol analogues from resorcinols and pulegone 2 in the presence of Yb(OTf)3 (2 mol %) and ascorbic acid (20 mol %)
under microwave irradiation in 1,2-dichloroethane.
#
Main products
Yb(OTf)3 (2 mol %)/ascorbic acid 1:10
4a-THX (3) (%)a 3-THC (4) (%)a
D
D
OH
OH
1
13 (12)b
34 (27)b
50 (27)b
43 (17)b
12 (18)b
11 (8)b
CH
O
CH
O
3
3
3a
4a
2
HO
O
3b
HO
O
4b
3
4
HO
O
HO
O
3c
4c
OH
OH
i
22 (15)b
49 (23)b
O
O
(R)-(+)-3d
(R)-(+)-4d
OH
OH
ii
21
47
O
O
(S)-(-)-3d
(S)-(-)-4d
5
38 (19)b
8 (6)b
HO
O
HO
O
3e
4e
Reaction temperature: 150 ꢁC. Reaction time: 120 min.
a
Isolated yields. Cooling OFF.
b
In brackets are reported the yields of the reactions carried out in the presence of Yb(OTf)3 without the addition of ascorbic acid (See Table 3).

O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
81
Table 5
Ki values for the binding of D
4a-THXs to CB receptors.
Side chain (R)
CB1 (Ki mean SE)
CB2 (Ki mean SE)
>10
[ gS EC50 on CB2 (Ki mean SE)
35S]GTP
3a
3b
3c
Me
H
Et
Pent
Pent
Me
>10
>10
>10
mM
mM
mM
m
M
n.d.
n.d.
n.d.
4.2 1.1 mM
0.51 0.21
0.18 0.12
0.47 0.15
m
m
m
M
M
M
3d (S)-(¡)
3d (R)-(þ)
3e
2.1 0.4
1.1 0.3
m
m
M
M
0.82 0.11
0.93 0.10
n.d.
m
m
M
M
>10
mM
1.0 0.6 mM
WIN55,212
CP55,940
2.5 1.3 nM
0.52 0.08 nM
0.45 0.9 nM
0.89 0.04 nM
n.d.
9.3 8.1 nM
n.d. not determined.
The use of 5-substituted resorcinols (1a and 1d) and 2 in the
receptors. Despite the limited series of compounds, it is noteworthy
that there is a 10-fold difference for CB1 receptor affinity and 60-
fold for CB2 between the most active (3d isomers) and least
active compounds in the series.
presence of YTACA and under microwave irradiation resulted in
higher yields and regioselectivity for the
ing almost unchanged the yield of the
4a-THX analogues (Table 4).
YTACA also improved the reactivity of 5-unsustituted resor-
D
3-THC analogues, keep-
D
The major structural differences among those compounds occur
in the positioning and the hydrophobicity of the substituents pre-
sent on the tetrahydroxanthene scaffold. In terms of CB1 binding,
most of the molecules were inactive and only the derivatives
bearing a pentyl chain at C-3 exhibited a certain affinity to the re-
cinols (Table 4, entries 2, 3, 5) yielding
main products.
D
4a-THX derivatives as the
This approach was successfully used also with (S)-(ꢀ) 2 and 1d
to generate the corresponding
D
3-THC analogue (S)-(ꢀ) 4d.
In microwave (MW) reactors equipped with an internal air-
cooling system, the reaction can be carried out with the cooling
mode ON at the lower temperature of 120 ꢁC. The reaction afforded
the final products with comparable isolated yields as with the
cooling mode OFF at 150 ꢁC. In MW-assisted syntheses there are
two primary hypotheses about thermal and non-thermal MW ef-
fects ([37] and see Suppl. Methods for more details). We speculate
that the two sets of condition may be equivalent for the final result.
Given the comparability of the resulting yields, the lower temper-
ature is more suitable and compliant with the presence of ascorbic
acid in the catalytic association and resulted to be particularly
important for the recyclability of the catalytic system. The recov-
ered catalyst could be reused with a minimal refresh of ascorbic
acid (20% of the initial amount of ascorbic acid), obtaining com-
parable reaction yields.
ceptor. Interestingly, most of the D
4a-THXs showed a better inter-
action with CB2 receptors, with the (S)-(ꢀ) 3d and (R)-(þ) 3d
stereoisomers being the best ligands with Ki values in the sub-
micromolar range (0.47 0.15
m
M and 0.18 0.12
m
M for (R)-(þ) 3d
and (S)-(ꢀ) 3d, respectively). Interestingly, the replacement of the
hydrophobic pentyl chain with the hydrophilic hydroxyl group at C-
3 did not impair the CB2 receptor binding, and the compounds
bearing a short hydrophobic substituent (methyl or ethyl) in the
neighbor positions showed a similar Ki value as the pentyl de-
rivatives (0.51 0.21 mM and 1.0 0.6 mM for 3c and 3e, respec-
tively) (Fig. 1A and B). (R)-(þ) and (S)-(ꢀ) 3d isomers were further
investigated for their functional activity at CB2 receptors by using
[
35S]GTP
g
S binding assays. As shown in Fig. 1C the two isomers
behaved as agonists, by inducing a concentration-dependent in-
crease of [35S]GTP
S binding. The EC50 values for (S)-(ꢀ) 3d and (R)-
g
(þ) 3d do not significantly differ between each other and are
slightly (2e4 times) higher than the calculated Ki values for the
receptor binding (Table 5). When compared with the full agonist
CP55,940, the 3d isomers show partial agonism at CB2 receptors
(Emax value (% of vehicle) of 152 11, 167 14 and 230 22 for (R)-
(þ) 3d, (S)-(ꢀ) 3d and CP55,940, respectively) (Fig. 1C and Table 5).
2.2. Characterization of
THX on cannabinoid receptor binding and function
D
3-THC analogues and corresponding D4a
-
Next we characterized the properties of the purified
D
3-THC
analogues (ꢂ95% purity) and corresponding
binoid receptor binding and function. As shown in Table 5, D4a
THXs showed moderate binding interactions at CB1 and CB2
D
4a-THXs on canna-
We also characterized the effects of
CB1 and CB2 receptors. As shown in Table 6, compound 4b, 4c and
D
3-THC and analogues on
-
Fig. 1.
D
4a-THXs binding and functional curves at CB2 receptor. Concentration-dependent binding curve at CB2 receptor for (A) the more potent (3c, (S)-(ꢀ) and (R)-(þ) 3d) and (B)
less potent (3a, 3b and 3e)
D
4a-THXs.
D
9-THC was used as positive control. (C) [35S]GTP
g
S binding induced by (S)-(ꢀ) 3d, (R)-(þ) 3d and CP55,940 upon CB2 activation. For A) and B)
data were collected from 5 to 8 independent experiments, for C) data were collected from 3 independent experiments. All the experiments were carried out in triplicate and
presented as mean S.E.

82
O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
Table 6
Ki values for the binding of D
3-THCs to CB receptors.
Side chain (R)
CB1 (Ki mean SE)
1.0 0.2
[
35S]GTP
g
S on CB1 (EC50 mean SE)
CB2 (Ki mean SE)
0.66 0.23
[ gS on CB2 (EC50 mean SE)
35S]GTP
4a
4b
4c
Me
H
Et
Pent
Pent
Me
Pent
m
M
n.d.
n.d.
n.d.
m
M
n.d.
n.d.
n.d.
>10
>10
5
m
m
M
M
>10
>10
m
m
M
M
4d (S)-(¡)
4d (R)-(þ)
4e
2 nM
87.5 28.1 nM
126.5 31.8 nM
n.d.
77.5 30.4 nM
n.d.
17 9 nM
18 12 nM
10.9 7.5 nM
17.4 11.1 nM
n.d.
12.3 6.8 nM
n.d.
29 12 nM
>10
mM
>10
mM
D
9-THC
22 13 nM
2.5 1.3 nM
0.52 0.08 nM
46 10 nM
0.45 0.9 nM
0.89 0.04 nM
WIN55,212
CP55,940
3.9 6.7 nM
9.3 8.1 nM
n.d. not determined.
4e which lack the pentyl chain at C-3 and the phenolic hydroxyl
group at C-1 did not show any significant binding, neither at CB1
nor at CB2.
The 3-methyl-1-hydroxy
tivity, at submicromolar concentrations, to both CB1 and CB2 re-
Although we have investigated a limited set of compounds, we
could observe an interesting preliminary structureeactivity rela-
tionship in regard to CB1 and CB2 receptor binding. Compounds (R)-
(þ) 4d and (S)-(ꢀ) 4d resulted the best ligands for both receptors
D
3-THC (4a) had a weak binding ac-
with an affinity that is comparable to D
9-THC (Fig. 2).
ceptors. Interestingly, compound (R)-(þ) 4d and (S)-(ꢀ) 4d which
The two stereoisomers were also tested on [35S]GTP
gS binding
bear the same substituents and in the same positions as
D
9-THC,
assays to check their functional behavior at CB1 and CB2 receptors.
The results showed that (R)-(þ) and (S)-(ꢀ) 4d activate CB1 re-
ceptors with similar potencies and with a tendency of the (S)-(ꢀ)
isomer to be more potent and efficient than the (R)-(þ) isomer
(Fig. 3A and Table 6). When compared with the full agonist
showed a potent binding to CB1 and CB2 receptors with the (S)-(ꢀ)
isomer being more potent at CB1 compared to the (R)-(þ) isomer (Ki
values: 5
2 nM and 17
6 nM for (S)-(ꢀ) 4d and (R)-(þ) 4d,
respectively). Both isomers also showed potent binding activities at
CB2 receptors, without significant difference between each other.
CP55,940, both 4d isomers behaved as partial agonists like D
9-THC
Fig. 2. (R)-(þ) and (S)-(ꢀ) 4d binding curve at CB1 and CB2 receptor. Concentration-dependent binding curve at (A) CB1 and (B) CB2 receptor for (R)-(þ) 4d, (S)-(ꢀ) 4d and
D
9-THC.
Data were collected from 5 to 6 independent experiments carried out in triplicate and presented as mean S.E.
Fig. 3. [35S]GTP
g
S binding induced by (R)-(þ) 4d, (S)-(ꢀ) 4d and
D
9-THC. Concentration-dependent [35S]GTP
g
S binding curves at (A) CB1 and (B) CB2 receptor for (R)-(þ) 4d, (S)-(ꢀ)
4d and
D
9-THC. Data were collected from 4 to 5 independent experiments carried out in triplicate and presented as mean S.E.

O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
83
(Emax value of 209 16, 240 10, 236 12 and 348 39 for (R)-(þ)
scaffold. The most potent compounds (S)-(ꢀ) 3d and (R)-(þ) 3d
bear the pentyl chain at C-3 and the hydroxyl group at C-1 in
agreement with the SAR described for classical cannabinoids.
Interestingly, the presence of a methyl or an ethyl group in the orto-
position of the hydroxyl group in C-3 increases the affinity at CB2
(3c and 3e), while the type of alkyl chain C-3 is not a strict
requirement for the binding at CB2 receptors.
4d, (S)-(ꢀ) 4d,
D
9-THC and CP55,940, respectively) (Fig. 3A). (R)-(þ)
4d, (S)-(ꢀ) 4d and
D
9-THC also exerted partial agonistic effect at CB2
receptors (Emax value of 133 6, 146 9, 139 13 and 222 9 for
(R)-(þ) 4d, (S)-(ꢀ) 4d,
D
9-THC and CP55,940, respectively) (Fig. 3B
and Table 6).
Despite the 4d isomers have already been described more than
twenty years ago, no biological and pharmacological investigation
has been carried out. Interestingly, the only report on the functional
activity of those molecules was performed in human volunteers
Despite the landscape of CB2 ligands has exploded over the last
years, leading to the identification of numerous potent and selec-
tive receptor agonists, antagonists and inverse agonists [see [46,49]
for review], our findings with the THX analogues open the possi-
bility to investigate an alternative natural product-like scaffold as
selective CB2 ligands. We also describe that (R)-(þ) 3d and (S)-(ꢀ)
3d behave as partial agonists, by increasing the CB2-mediated
[38]. In that study, both 1S and 1R D
3-THC were injected i.v. in six
individuals and the “cannabis-like” effects became evident at doses
of 4, 8 and 16 mg for the 1S isomer, while the 1R isomer elicited
such effects only at 16 mg [38]. As in 1987, the cannabinoid re-
ceptors were not identified yet, this at least suggested that
behave similarly to
9-THC, which was also included in the test, by
interacting with a specific common cellular target. Unlike other
THCs (i.e.
8-THC [39,40]) no further investigation of 3-THC ac-
tivity at CB receptors was performed. We now describe for the first
time the potent CB1 binding of both (R)-(þ) and (S)-(ꢀ)
3-THC
(4d), with the (S)-(ꢀ) isomer showing a 6-fold lower Ki value.
Although in the [35S]GTP
S assay we could not detect any signifi-
cance difference, (S)-(ꢀ) 4d and
9-THC showed a tendency to be
more potent and efficacious than (R)-(þ) 4d. In addition, other
signal transduction pathways (e.g.
-arrestin and Ca2þ) could be
D
3-THCs
activation of G-protein. A deeper investigation of the SAR for D4a
-
THXs binding at cannabinoid receptors needs to be performed
together with the characterization of the functional behavior for
D
D
D
the main CB2-mediated transduction pathways (i.e. cAMP, b-
arrestin and Ca2þ). In the past decade it has become clear that the
concept of agonist at a certain receptor does not exhaustively
describe the pharmacological properties of a ligand. Different ag-
onists at the same metabotropic receptor can transduce the signal
by recruiting different pathways leading to a diverse pharmacology
D
g
D
and/or toxicology [50]. For example, D
9-THC has been recently re-
b
ported to trigger some of its CB1-mediated behavioral effects by
increasing COX-2 expression through the preferential activation of
the G-protein bg-subunit signaling, while 2-AG does the opposite
(i.e. reducing the COX-2 expression) by recruiting the Gai-mediated
signaling pathway [51]. Thus, new scaffolds with particular CB2
receptor binding and functional properties may lead to the iden-
tification of molecules which exhibit new pharmacological
features.
differentially involved in the CB1-mediated biological effects of 4d
isomers. This may account for the different potency in triggering
the “cannabis-like” effects in human subjects [38]. In terms of the
SAR for cannabinoid receptor binding, several studies have inves-
tigated the main structural requirements that make classical can-
nabinoids good or bad ligands for both CB1 and CB2 receptors.
Among the main chemical features for CB1 binding, the presence of
a phenolic hydroxyl group at C-1 is usually essential as well as a
suitable substituent and geometry at C-9 [41]. In general, an 11- or
11-nor-9-hydroxyl group enhances the affinity at CB1 receptors
[42e44]. The alkyl chain present at C-3 clearly plays a relevant role
in the interaction with CB1 receptors by eliciting an optimal binding
when is composed by five to eight carbon atoms [45]. SAR data for
classical cannabinoids on CB2 receptors have been reported to
minor degree, but some important features which make the com-
pounds more selective for CB2 have been suggested, based upon the
3. Conclusions
Ytterbium triflate-ascorbic acid (named YTACA) was used as
association catalyst in the MW-assisted one pot synthesis of
and analogues. The use of YTACA led to an improvement in the D3
THC yield (up to 49%) and in the
3-THC/ 4a-THX ratio (2:1)
D
3-THC
-
D
D
compared with previously reported data reported [16,18]. The
heterogonous catalysis described in our article allowed a fast and
regioselective preparation of the desired products in sufficient
yields for full biological characterizations.
observation that some ligands (i.e. JWH-051, JWH-057, 1-deoxy-D8
-
THC analogues and L59633, see Suppl. Fig. 2 for chemical struc-
tures) are more selective for this cannabinoid receptor subtype
[46]. In particular, the phenolic hydroxyl group at C-1 hinders CB2
selectivity which is increased when it is replaced by hydrogen.
Unlike for CB1 receptors, the length of the alkyl chain at C-3 plays a
We show for the first time that
receptors with a similar potency and exert partial agonism at both
receptor subtypes like
9-THC. In addition, we report that the (S)-
D
3-THC binds to CB1 and CB2
D
(ꢀ) isomer shows a 6-fold higher affinity at CB1 than the (R)-(þ)
isomer in agreement with previous functional results obtained in
human volunteers. We also show that any changes in the type and
position of the substituents on the THC scaffold leads to a loss of
receptor binding. Importantly, we report for the first time the
marginal role for CB2 binding [40,46]. The limited
D
3-THCs SAR
described in our study strictly follows these structural re-
quirements. In fact, compound 4b, 4c and 4e, which bear a hydroxyl
group instead of an alkyl chain at C-3 and lack the hydroxyl group at
C-1 resulted totally inactive at CB1 and CB2 receptors while only (R)-
(þ) 4d and (S)-(ꢀ) 4d, which show the same chemical features as
moderate but selective CB2 receptor binding properties for D4a
-
THXs, which are structural isomers of D
3-THCs.
D
9-THC possess a high affinity binding at both CB receptors. In
In future studies, a more detailed investigation of the SAR for
keeping with this similarity in the binding potency, the two
D
3-THC
D
4a-THXs binding at cannabinoid receptors needs to be performed
stereoisomers also behave as partial agonists at CB1 and CB2 like D9
-
as well as an in vivo pharmacological characterization of the two
(S)-(ꢀ) and (R)-(þ)
3-THC isomers.
THC ([47,48] and our data).
D
D
4a-THXs and classical cannabinoids (THCs) are structural iso-
mers and show a certain degree of similarity in the chemical
structure. Based on this we assume that similar structural re-
quirements described for classical cannabinoids would also be
4. Experimental section
4.1. Materials and methods
essential for
D D
4a-THXs binding at CB receptors. All 4a-THXs
resulted inactive or poorly active at CB1. On the other side, we
noticed an increasing affinity to CB2 receptor depending on the type
and position of the substituents on the tetrahydroxanthene
All chemicals were purchased from the major chemical sup-
pliers as highest purity grade and used without any further puri-
fication. Solvents were dried over standard drying agent and freshly

84
O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
distilled prior to use. Column chromatography was performed by
Merck silica gel 60 (70e230 mesh ASTM) with Hexane/Et2O 98:2 as
eluent, and monitored by TLC on silica gel 60 F254 with detection
by charring with sulfuric acid/p-anisaldehyde 0.5%/0.5% in EtOH
followed by heating at 110 ꢁC. 1H and 13C NMR spectra were
recorded in CDCl3 with a Bruker Avance DPX 400 spectrometer at a
frequency of 400 and 100 MHz, respectively. GCeMS analysis were
obtained by HP-6890 gas chromatograph (dimethyl silicone col-
umn, 12.5 m) equipped with an HP-5973 mass-selective detector at
an ionizing voltage of 70 eV.
4.2.1. (9R)-3,9,9-trimethyl-2,3,4,9-tetrahydro-1H-xanthen-6-ol
(3b)
22.2
Yield 34%. Yellow oil, [
a]
D
þ 64.0ꢁ (c 0.17 in CHCl3) 1H NMR
¼ 1.05 (d, J ¼ 6.0 Hz, 3H, CH3), 1.29
(400 MHz, CDCl3, 25 ꢁC, TMS)
d
(m,1H),1.34 (s, 3H, CH3), 1.36 (s, 3H, CH3),1.81 (m, 3H), 2.18 (m, 3H),
5.08 (bs, 1H, OH), 6.37 (d, J ¼ 2.6 Hz, 1H, ArHeX), 6.53 (dd, J ¼ 8.5,
2.6 Hz, 1H, ArHeY), 7.16 (d, J ¼ 8.5 Hz, 1H, ArHeZ). 13C NMR
(100 MHz, CDCl3, 25 ꢁC, TMS)
d
¼ 21.7, 22.6, 29.2, 29.4, 30.2, 31.6,
33.3, 35.5, 102.6, 110.4, 111.7, 123.3, 127.6, 142.1, 151.4, 154.5. GCeMS
m/z: 242, 217, 172, 145, 118, 102, 78. C16H20O2 HRESI-MS: [MꢀH]ꢀ,
m/z 243.13853 calctd ¼ 244.14633, purity ¼ 98%.
GCeMS analysis were obtained with HP-6890 gas chromato-
graph (dimethyl silicone column, 12.5 m) equipped with an HP-
5973 mass-selective detector at an ionizing voltage of 70 eV. Op-
tical rotations ([
polarimeter.
a
]D) were measured with a JASCO DIP-1000 digital
4.2.2. (3R)-7ethyl-3,9,9-trimethyl-2,3,4,9-tetrahydro-1H-xanthen-
6-ol (3c)
20.6
For HRESI-MS all samples were diluted with methanol in order
to obtain stock solutions (5 ng/mL), to be analyzed by a Thermo
Scientific LTQ-Orbitrap XL mass spectrometer equipped with an
electrospray ion source (Thermo Scientific, Germany) and operated
under Xcalibur 2.1 version software. Negative ionization mode for
the MS analysis was used with data-dependent automatic switch-
ing between MS and MS/MS acquisition modes. The instrument
was calibrated using the manufacturer's calibration standards. The
scan was collected in the Orbitrap at a resolution of 30,000 in a m/z
range of 150e800 amu. The source voltage was 4.5 kV and capillary
voltage ꢀ35 kV was, the tube lens offset 126 V and the capillary
temperature was set at 280 ꢁC, auxiliary gas was set at 5 and no
sheath gas was used.
Yield 50%. Dark Yellow oil, [
a]
þ 43.6ꢁ (c 0.16 in CHCl3), 1H
D
NMR (400 MHz, CDCl3, 25 ꢁC, TMS)
d
¼ 1.04 (d, J ¼ 6.0 Hz, 3H, CH3),
1.24 (t, J ¼ 7.5 Hz, 3H, CH3) 1.28 (m, 1H), 1.35 (s, 3H, CH3), 1.37 (s, 3H,
CH3), 1.81 (m, 3H), 2.21 (m, 3H), 2.61 (q, J ¼ 7.5 Hz, 2H, CH2), 4.89
(bs, 1H, OH), 6.31 (s, 1H, ArH), 7.02 (s, 1H, ArH). 13C NMR (100 MHz,
CDCl3, 25 ꢁC, TMS)
d
¼ 14.7, 21.7, 22.6, 23.1, 29.2, 29.4, 30.2, 31.6,
33.3, 35.5,102.3,110.5,122.8,124.7,126.9,142.1,149.4,152.1. GCeMS
m/z: 272, 257, 241, 227, 215, 200, 128, 115. C18H24O2 HRESI-MS:
[MꢀH]ꢀ, m/z 271.16923 calctd ¼ 272.17763, purity ¼ 98%.
4.2.2.1. (3R)-6-penthyl-3,9,9-trimethyl-2,3,4,9-tetrahydro-1H-
xanthen-8-ol ((R)-(þ)3d). Yield 22%. Pale-brown oil, [
a
]
24.0 þ 36.0ꢁ
D
d
HPLC analysis was performed using a HP 1100 system equipped
with a UV/Vis detector with chiral column Lux Amilose-2 (chlori-
nated amylose phenylcarbamate) (250 ꢃ 4.6 mm ID). Eluent Hex-
ane/Isopropanol 98:2. Flow 1 mL/min. Detection at wavelength
254 nm. All chromatograms were run at 25 ꢁC.
(c 0.19 in CHCl3), 1H NMR (400 MHz, CDCl3, 25 ꢁC, TMS)
¼ 0.91 (t,
J ¼ 7 Hz, 3H, CH3), 1.04 (d, J ¼ 6.2 Hz, 3H, CH3), 1.28 (m 7H, 3ꢃ CH2,
1ꢃ CH) 1.52 (s, 3H, CH3), 1.55 (s, 3H, CH3), 1.61 (m, 2H, CH2), 1.84 (m,
2H, CH2), 2.22 (m, 2H, CH2), 2.46 (t, J ¼ 7.6 Hz, 2H, CH2), 4.77 (bs, 1H,
OH), 6.19 (d, J ¼ 1.7 Hz, 1H, ArH) 6.31 (d, J ¼ 1.7 Hz, 1H, ArH). 13
C
Microwave-assisted reactions were carried out using a single-
mode cavity dedicated reactor (Biotage Initiator™). Reactions
were performed either with temperature or power-controlled
programs in glass vials (0.5e2 mL, 2e5 mL or 5e25 mL, depend-
ing on the scale) sealed with a Teflon septum. Power-controlled
experiments were performed with simultaneous cooling of the
vials by means of pressurized air (maximum 15 bar). All tempera-
tures were measured externally by an IR sensor. The reaction time
was counted when the reaction mixture reached the stated tem-
perature for temperature-controlled experiments. In the case of
power controlled experiments the specified reaction time corre-
sponds to the total irradiation time. Pressure was measured by a
non-invasive sensor integrated into the cavity lid.
NMR (100 MHz, CDCl3, 25 ꢁC, TMS)
d
¼ 14.4, 21.9, 22.2, 23.0, 26.6,
27.0, 29.4, 31.0, 31.90, 31.92, 33.6, 35.5, 35.6, 109.0, 111.1, 112.4, 113.9,
140.3, 142.7, 151.7, 154.6. GCeMS m/z: 314, 299, 285, 271, 257, 243,
228, 218, 205, 189, 125, 96. C21H30O2 HRESI-MS: [MꢀH]ꢀ, m/z
313.21632 calctd ¼ 314.22445, purity ¼ 98%.
4.2.2.2. (3S)-6-penthyl-3,9,9-trimethyl-2,3,4,9-tetrahydro-1H-
xanthen-8-ol ((S)-(ꢀ)3d). Yield 22%. Pale-brown oil, [
a]
d
24.0e35.6ꢁ
¼ 0.91 (t,
D
(c 0.78 in CHCl3), 1H NMR (400 MHz, CDCl3, 25 ꢁC, TMS)
J ¼ 7 Hz, 3H, CH3), 1.04 (d, J ¼ 6.2 Hz, 3H, CH3), 1.28 (m 7H, 3ꢃ CH2,
1ꢃ CH) 1.52 (s, 3H, CH3), 1.55 (s, 3H, CH3), 1.61 (m, 2H, CH2), 1.84 (m,
2H, CH2), 2.22 (m, 2H, CH2), 2.46 (t, J ¼ 7.6 Hz, 2H, CH2), 4.77 (bs, 1H,
OH), 6.19 (d, J ¼ 1.7 Hz, 1H, ArH) 6.31 (d, J ¼ 1.7 Hz, 1H, ArH). 13
C
NMR (100 MHz, CDCl3, 25 ꢁC, TMS)
d
¼ 14.4, 21.9, 22.2, 23.0, 26.6,
4.2. General procedure for THXs and
D
3-THCs synthesis
27.0, 29.4, 31.0, 31.90, 31.92, 33.6, 35.5, 35.6, 109.0, 111.1, 112.4, 113.9,
140.3, 142.7, 151.7, 154.6. GCeMS m/z: 314, 299, 285, 271, 257, 243,
228, 218, 205, 189, 125, 96. C21H30O2 HRESI-MS: [MꢀH]ꢀ, m/z
313.21553 calctd ¼ 314.22458, purity ¼ 98%.
In a microwave hermetically sealable vial an equimolar quantity
of R-(þ)-pulegone or S-(ꢀ)-pulegone and resorcinol derivatives
1aee are placed. 1,2-dichloroethane (1e4 mL) was used as solvent.
The catalyst is added (either
a-zyrconium sulphenylphosphonate
[(a-Zr(O3PCH3)1.2(O3PC6H4SO3H)0.8], sulfuric acid supported on
4.2.3. (3R)-3,5,9,9-tetramethyl-2,3,4,9-tetrahydro-1H-xanthen-6-
silica gel [H2SO4eSiO2], ytterbium triflate [Yb(OTf)3], ytterbium
triflate-ascorbic acid [YTACA 1:10]. The vial was sealed with its cap
ol (3e)
23.1
Yield 38%. Yellow oil, [
a]
þ 53.6ꢁ (c 0.35 in CHCl3), 1H NMR
¼ 0.99 (d, J ¼ 6.0 Hz, 3H, CH3), 1.23
D
and
a
microwave irradiation (20e240 min, cooling OFF/ON,
(400 MHz, CDCl3, 25 ꢁC, TMS)
d
90 ꢁCe150 ꢁC, 150e350 W mean irradiation depending on reaction
temperatures) was applied. At the end of the programmed reaction
time, the resulting mixture was filtered on a Buchner funnel in
order to recover the catalyst and evaporated under vacuum. Effi-
cient and complete separation of the major reaction product is
achieved by column chromatography on silica gel, eluting with
Hexane/Et2O (98:2).
(m, 1H), 1.28 (s, 3H, CH3), 1.30 (s, 3H, CH3), 1.78 (m, 3H), 2.10 (s, 3H,
CH3), 2.15 (m, 3H), 4.72 (bs, 1H, OH), 6.47 (d, 1H, J ¼ 8.5 Hz, ArH),
6.97 (d, 1H, J ¼ 8.5, ArH). 13C NMR (100 MHz, CDCl3, 25 ꢁC, TMS)
d
¼ 8.1, 21.4, 22.3, 28.9, 29.0, 29.9, 31.4, 33.3, 35.2, 109.3, 110.1, 111.2,
122.6, 123.5, 142.0, 149.1, 152.1. GCeMS m/z: 258, 243, 227, 216, 201,
189, 175, 115.
calctd ¼ 258.16198, purity ¼ 98%.
C
17H22O2 HRESI-MS: [MꢀH]ꢀ, m/z 257.15363

O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
85
4.2.4. (9R)-6,6,9-trimethyl-7,8,9,10-tetrahydro-6H-benzo[c]
5 mM MgCl2, 0.5% fatty acid free BSA, pH 7.4). Filters were added to
45 of MicroScint20 scintillation liquid and radioactivity
chromen-3-ol (4b)
mL
22.2
Yield 12%. Pale yellow oil, [
a]
þ 98.6ꢁ (c 0.18 in CHCl3), 1H
¼ 1.07 (d, J ¼ 6.5 Hz, 3H, CH3),
measured with the 1450 MicroBeta Trilux top counter (PerkinElmer
Life Sciences). Data were collected from at least three independent
experiments performed in triplicate and the non-specific binding
was subtracted. Results were expressed as [3H]-CP55940 bound in
% of binding to vehicle-treated samples. Ki values were calculated
applying the ChengePrusoff equation. IC50 values were derived
from the concentration-dependent curves.
D
NMR (400 MHz, CDCl3, 25 ꢁC, TMS)
d
1.28 (m, 2H, CH2), 1.32 (s, 3H, CH3), 1.43 (s, 3H, CH3), 1.79 (m, 2H,
CH2), 1.93, (m, 1H, CH), 2.11 (m, 1H, CH), 2.43 (m, 1H, CH), 4.97 (bs,
1H, OH), 6.35 (d, J ¼ 2.8 HZ, 1H, ArH), 6.38 (dd, J ¼ 2.8, 8.3 Hz, 1H,
ArH), 6.99 (d, J ¼ 8.3 Hz, 1H, ArH). 13C NMR (100 MHz, CDCl3, 25 ꢁC,
TMS)
d
¼ 21.9, 24.5, 25.5, 25.7, 28.3, 30.9, 32.9, 78.6, 103.6, 107.5,
117.9, 122.7, 122.8, 131.3, 153.2, 155.5. GCeMS m/z: 244, 229, 214,
202, 174, 136, 118, 102, 78. C16H20O2 HRESI-MS: [MꢀH]ꢀ, m/z
243.13829 calctd ¼ 224.14633, purity ¼ 98%.
4.4. [35S]GTP
gS binding assay
The assay was carried out using 5
pared in-house from CHO-hCB2 and CHO-hCB1 cells as previously
reported [52]. Membranes were diluted in silanized plastic tubes
m
g of clean membrane pre-
4.2.5. (9R)-2-ethyl-6,6,9-trimethyl-7,8,9,10-tetrahydro-6H-benzo
[c]chromen-3-ol (4c)
22.7
Yield 11%. Yellow oil, [
a]
þ 19.9ꢁ (c 0.17 in CHCl3), 1H NMR
¼ 1.09 (d, J ¼ 6.5 Hz, 3H, CH3), 1.23
D
with 200
mL of GTPyS binding buffer (50 mM TriseHCl, 3 mM
(400 MHz, CDCl3, 25 ꢁC, TMS)
d
MgCl2, 0.2 mM EGTA and 100 mM NaCl at pH 7.4 supplemented
(t, J ¼ 2.5 Hz, 3H, CH3), 1.27 (m, 2H, CH2), 1.31 (s, 3H, CH3), 1.42 (s,
3H, CH3), 1.80 (m, 2H, CH2), 1.96, (m, 1H, CH), 2.13 (m, 1H, CH), 2.45
(dd, J ¼ 4.4, 16.3 Hz, 1H, CH), 2.59 (q, J ¼ 7.5, 2H, CH2), 4.72 (bs, 1H,
OH), 6.35 (d, J ¼ 2.8 HZ, 1H, ArH), 6.29 (s, 1H, ArH), 6.88 (s, 1H, ArH).
with 0.5% of BSA fatty acid free) in the presence of 10 mM of GDP and
0.1 nM of [35S]-GTPyS 1250 Ci/mmol). The mixture was kept on ice
until the binding reaction was started by adding the tested com-
pound, vehicle (negative control) or CP55940 (positive control).
13C NMR (100 MHz, CDCl3, 25 ꢁC, TMS)
d
¼ 14.9, 22.3, 23.0, 25.2,
Non-specific binding was measured in presence of 10 mM GTPyS
25.6, 26.1, 28.7, 31.4, 33.4, 78.6, 103.8, 118.1, 122.0, 122.9, 123.1, 131.6,
(Sigma). The tubes were incubated at 30 ꢁC for 90 min under
shaking and then they were put on ice to stop the reaction. An
151.5, 153.2. GCeMS m/z: 272, 257, 243, 230, 215, 202, 136, 115, 94.
C
18H24O2 HRESI-MS: [MꢀH]ꢀ, m/z 271.07982 calctd ¼ 272.17763,
aliquot (185 mL) of the reaction mixture was rapidly filtered through
purity ¼ 98%.
a 96-well microplate bonded with GF/B glass fiber filters (UniFilter-
96 GF/B, PerkinElmer Life Sciences) previously pre-soaked with ice-
cold washing buffer (50 mM of TriseHCl pH 7.4 plus of 0.1% BSA
fatty acid free). The filters were washed 6 times with 180 mL of
washing buffer under vacuum and let them dry under the air drier
flow. The radioactivity was measured with the 1450 Microbeta
WallacTrilux Top counter after the addition of 45
cocktail. Specific binding was calculated by subtracting the residual
radioactivity signal obtained in presence of an excess of GTP
the results were expressed as % of vehicle control.
4.2.6. (9R)-4,6,6,9-tetramethyl-7,8,9,10-tetrahydro-6H-benzo[c]
chromen-3-ol (4e)
24.5
Yield 8%. Yellow oil, [
a]
þ 49.3ꢁ (c 0.25 in CHCl3), 1H NMR
¼ 1.07 (d, J ¼ 6.5 Hz, 3H, CH3), 1.27
D
(400 MHz, CDCl3, 25 ꢁC, TMS)
d
(m, 2H, CH2), 1.30 (s, 3H, CH3), 1.44 (s, 3H, CH3), 1.79 (m, 2H, CH2),
1.95, (m, 1H, CH), 2.12 (s, 3H, CH3), 2.13 (m, 1H, CH), 2.40 (m, 1H,
CH), 4.77 (bs, 1H, OH), 6.38 (d, J ¼ 8.3 HZ, 1H, ArH), 6.87 (d,
mL of scintillation
gS and
J ¼ 8.3 Hz, 1H, ArH). 13C NMR (100 MHz, CDCl3, 25 ꢁC, TMS)
¼ 7.9,
d
21.9, 24.9, 25.1, 25.7, 28.3, 30.9, 33.0, 78.2, 106.7, 111.4, 117.7, 119.4,
123.0, 131.0, 150.9, 153.5. GCeMS m/z:258, 243, 232, 227, 216, 202,
189, 176, 112, 94, 70, 58. C17H22O2 HRESI-MS: [MꢀH]ꢀ, m/z
257.15311 calctd ¼ 258.161980, purity ¼ 98%.
Acknowledgments
The authors gratefully acknowledge Prof. Maria Carla Marco-
tullio and Dr. Daniela Lanari for helpful discussions. The biological
work was supported by grants from the Lifelong Learning Pro-
gramme (LLP) at the Institute of Biochemistry and Molecular
Medicine, University of Bern, CH. The authors would also like to
thank Prof. Andrea Temperini at Department of Pharmaceutical
Sciences, University of Perugia,IT for chiral HPLC analysis and Prof.
Sonia Piacente, University of Salerno, IT for HRESI-MS analysis.
4.2.7. (3R)-3,6,9,9-tetramethyl-2,3,4,9-tetrahydro-1H-xanthen-8-
ol (3a)
Yield 13%, (9R)-3,6,6,9-tetramethyl-7,8,9,10-tetrahydro-6H-
benzo[c]chromen-3-ol (4a) yield 43%, (R)-(þ)-
D
3-THC (4d) yield
49%, (S)-(ꢀ)
D
3-THC (4d) yield 47%: data in accordance to the
literature [18,20,38]. Purity of 3a, 4a, (R)-(þ) and 4d (S)-(ꢀ)4d
resulted to be 98%.
4.3. Radioligand binding assays on cannabinoid receptors
Appendix A. Supplementary data
Receptor-binding experiments were performed with membrane
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2014.07.062.
preparations as previously reported [52]. Briefly, 18
membrane expressing hCB1 or hCB2 receptors (prepared as
described earlier) were re-suspended in 300 L of binding buffer
mg of crude
m
References
(50 mM TriseHCl, 2.5 mM EDTA, 5 mM MgCl2, 0.5 mg/mL fatty acid
free BSA, pH 7.4) in silanized glass tubes and co-incubated different
[1] R. Pertwee, Cannabinoids, Springer-Verlag, Heidelberg, Germany, 2005.
[2] E.L. Abel, Marijuana: the First Twelve Thousand Years, Plenum Press, New
York and London, 1980.
concentrations (10ꢀ13e10ꢀ4 M) of
D D
3-THCs, 4a-THXs or vehicle
and 0.5 nM of [3H]-CP-55940 (168 Ci/mmol) for 2 h at 30 ꢁC. Non-
[3] Y. Gaoni, R. Mechoulam, Isolation, structure, and partial synthesis of an active
constituent of hashish, J. Am. Chem. Soc. 86 (1964) 1646e1647.
[4] B.M. Trost, K. Dogra, Synthesis of (ꢀ)-Delta9-trans-tetrahydrocannabinol:
stereocontrol via mo-catalyzed asymmetric allylic alkylation reaction, Org.
Lett. 9 (2007) 861e863.
specific binding of the radioligand was determined in presence of
10 mM of WIN55,512-2. Non-specific binding was around 5%. After
the incubation time, membrane suspensions were rapidly filtered
through a 0.5% polyethyleneimine pre-soaked 96-well microplate
equipped with GF/B glass fiber filters (UniFilter-96 GF/B, Perki-
nElmer Life Sciences) under vacuum and washed 12 times with
[5] E. Ballerini, L. Minuti, O. Piermatti, High-pressure Diels-Alder cycloadditions
between benzylideneacetones and 1,3-butadienes: application to the syn-
thesis of (R,R)-(ꢀ)- and (S,S)-(þ)-
D8-tetrahydrocannabinol, J. Org. Chem. 75
150 mL of ice-cold washing buffer (50 mM TriseHCl, 2.5 mM EGTA,
(2010) 4251e4260.

86
O. Rosati et al. / European Journal of Medicinal Chemistry 85 (2014) 77e86
[6] W.A. Devane, F.A. Dysarz 3rd, M.R. Johnson, L.S. Melvin, A.C. Howlett, Deter-
[31] A. Mahadevan, C. Siegel, B.R. Martin, M.E. Abood, I. Beletskaya, R.K. Razdan,
Novel cannabinol probes for CB1 and CB2 cannabinoid receptors, J. Med. Chem.
43 (2000) 3778e3785.
mination and characterization of a cannabinoid receptor in rat brain, Mol.
Pharmacol. 34 (1988) 605e613.
[7] S. Munro, K.L. Thomas, M. Abu-Shaar, Molecular characterization of a pe-
ripheral receptor for cannabinoids, Nature 365 (1993) 61e65.
[8] S.P. Cohen, Cannabinoids for chronic pain, Br. Med. J. 336 (2008) 167e168.
[9] R.D. Hosking, J.P. Zajicek, Therapeutic potential of cannabis in pain medicine,
Br. J. Anaesth. 101 (2008) 59e68.
[10] P. Papathanasopoulos, L. Messinis, E. Lyros, A. Kastellakis, G. Panagis, Multiple
sclerosis, cannabinoids and cognition, J. Neur. Clin. Neurosci. 20 (2008) 36e51.
[11] I.R. Sutton, P. Daeninck, Cannabinoids in the management of intractable
chemotherapy-induced nausea and vomiting and cancer-related pain,
J. Supp.Oncol. 4 (2006) 531e535.
[12] R.W. Foltin, J.V. Brady, M.W. Fischman, Behavioral analysis of marijuana ef-
fects on food intake in humans, Pharm. Biochem. Behav. 25 (1986) 577e582.
[13] L.M. Eubanks, C.J. Rogers, A.E.T. Beuscher, G.F. Koob, A.J. Olson, T.J. Dickerson,
K.D. Janda, A molecular link between the active component of marijuana and
Alzheimer's disease pathology, Mol. Pharmacol. 3 (2006) 773e777.
[14] G. Appendino, S. Gibbons, A. Giana, A. Pagani, G. Grassi, M. Stavri, E. Smith,
M.M. Rahman, Antibacterial cannabinoids from cannabis sativa: a structur-
eꢀactivity study, J. Nat. Prod. 71 (2008) 1427e1430.
[32] O. Rosati, M. Curini, F. Messina, M. Marcotullio, G. Cravotto, Ferrier rear-
rangement and 2-deoxy sugar synthesis from D-glycals mediated by layered
a
-zirconium sulfophenylphosphonate-methanphosphonate as heterogeneous
catalyst, Cat. Lett. 143 (2013) 169e175.
[33] J.S. Yadav, B.V.S. Reddy, K.B. Reddy, M. Satyanarayana, CeCl3$7H2O: a novel
reagent for the synthesis of 2-deoxysugars from d-glycals, Tetrahedron Lett.
43 (2002) 7009e7012.
[34] H. Yamamoto, K. Futatsugi, “Designer acids”: combined acid catalysis for
asymmetric synthesis, Angew. Chem. Int. Ed. 44 (2005) 1924e1942.
[35] H. Yamamoto, K. Futatsugi, Recent Advance of ‘Combined Acid’ Strategy for
Asymmetric Catalysis, Dep. Chem. Univ. Chicago, Chicago, IL USA, 2008.
[36] H. Yamamoto, From designer Lewis acid to designer Bronsted acid towards
more reactive and selective acid catalysis, Proc. Jap. Acad. Ser. B Phys. Biol. Sci.
84 (2008) 134e146.
[37] A. de la Hoz, A. Díaz-Ortiz, A. Moreno, Microwaves in organic synthesis.
Thermal and non-thermal microwave effects, Chem. Soc. Rev. 34 (2005)
164e178.
[38] L.E. Hollister, H.K. Gillespie, R. Mechoulam, M. Srebnik, Human pharmacology
of 1S and 1R enantiomers of delta-3-tetrahydrocannabinol, Psychopharma-
cology 92 (1987) 505e507.
[15] R. Ghosh, A.R. Todd, D.C. Wright, Cannabis indica. Part VI. The condensation of
pulegone with alkyl resorcinols. A new synthesis of cannabinol and of a
product with hashish activity, J. Chem. Soc. 1 (1941) 137e140.
[16] G. Leaf, A.R. Todd, S. Wilkinson, Cannabis indica. Part IX. The isolation of 30:40:
50: 60-tetrahydrodibenzopyran derivates from pulegone-orcinol and
pulegone-olivetol condensation products. Synthesis of d-tetrahydrocannab-
inol, J. Chem. Soc. 1 (1942) 185e188.
[17] R. Adams, S. Loewe, C.M. Smith, W.D. McPhee, Tetrahydrocannabinol homo-
logs and analogs with marihuana activity. XIII1, J. Am. Chem. Soc. 64 (1942)
694e697.
[39] J.W. Huffman, S.G. Duncan Jr., J.L. Wiley, B.R. Martin, Synthesis and pharma-
cology of the 10,20-dimethylheptyl-
D8-THC isomers: exceptionally potent
cannabinoids, Bioorg. Med. Chem. Lett. 7 (1997) 2799e2804.
[40] D.P. Papahatjis, V.R. Nahmias, S.P. Nikas, T. Andreou, S.O. Alapafuja, A. Tsotinis,
J. Guo, P. Fan, A. Makriyannis, C10-cycloalkyl side chain pharmacophore in
tetrahydrocannabinols, J. Med. Chem. 50 (2007) 4048e4060.
[41] R.K. Razdan, Structure-activity relationships in cannabinoids, Pharmacol. Rev.
38 (1986) 75e149.
[18] J.B. Chazan, G. Ourisson, Tetrahydrocannabinol analogs and related com-
pounds. I. Condensation of pulegone and orcinol in the presence of phos-
phorus oxychloride, Bull. Soc. Chim. Fr. 4 (1968) 1374e1384.
[19] U. Claussen, P. Mummenhoff, F. Korte, HaschischdXVIII: Zur tetrahydrocan-
nabinol-synthese aus olivetol und pulegon, Tetrahedron 24 (1968)
2897e2898.
[20] M. Srebnik, N. Lander, A. Breuer, R. Mechoulam, Base-catalysed double-bond
isomerizations of cannabinoids: structural and stereochemical aspects,
J. Chem. Soc. Perkin Trans. 1 (1984) 2881e2886.
[42] Y. Gareau, C. Dufresne, M. Gallant, C. Rochette, N. Sawyer, D.M. Slipetz,
N. Tremblay, P.K. Weech, K.M. Metters, M. Labelle, Structure activity re-
lationships of tetrahydrocannabinol analogues on human cannabinoid re-
ceptors, Bioorg. Med. Chem. Lett. 6 (1996) 189e194.
[43] J.W. Huffman, S. Yu, V. Showalter, M.E. Abood, J.L. Wiley, D.R. Compton,
B.R. Martin, R.D. Bramblett, P.H. Reggio, Synthesis and pharmacology of a very
potent cannabinoid lacking a phenolic hydroxyl with high affinity for the CB2
receptor, J. Med. Chem. 39 (1996) 3875e3877.
[44] J.W. Huffman, S.M. Bushell, J.R. Miller, J.L. Wiley, B.R. Martin, 1-Methoxy-, 1-
deoxy-11-hydroxy- and 11-hydroxy-1-methoxy-Delta(8)-tetrahydrocannab-
inols: new selective ligands for the CB2 receptor, Bioorg. Med. Chem. 10
(2002) 4119e4129.
[21] V. Di Marzo, Cannabinoids, first ed., Plenum Publishers, Georgetown, Texas,
USA, 2004.
[22] O. Rosati, M. Curini, F. Montanari, M. Nocchetti, S. Genovese,
a-Zirconium
sulfophenylphosphonate as catalyst for the synthesis of 3,4-
dihydropyrimidin-2(1H)-one derivatives under solvent free conditions, Cat.
Lett. 141 (2011) 850e853.
a
[45] H.H. Seltzman, Structure and receptor activity for classical cannabinoids, Curr.
Med. Chem. 6 (1999) 685e704.
[46] A. Poso, J.W. Huffman, Targeting the cannabinoid CB2 receptor: modelling and
structural determinants of CB2 selective ligands, Br. J. Pharmacol. 153 (2008)
335e346.
[23] F. Epifano, C. Pelucchini, O. Rosati, S. Genovese, M. Curini, Ytterbium triflate
promoted one-pot three component synthesis of 3,4,5-trisubstituted-3,6-
dihydro-2 H-1,3-oxazines, Cat. Lett. 141 (2011) 844e849.
[47] S.J. Govaerts, E. Hermans, D.M. Lambert, Comparison of cannabinoid ligands
affinities and efficacies in murine tissues and in transfected cells expressing
[24] J.F. Zhou, X. Chen, Q.B. Wang, B. Zhang, L.Y. Zhang, A. Yusulf, Z.F. Wang,
J.B. Zhang, J. Tang, H2SO4-SiO2: highly efficient and novel catalyst for the
Ferrier-type glycosylation, Chin. Chem. Lett. 21 (2010) 922e926.
human recombinant cannabinoid receptors, Eur. J. Pharm. Sci.
3 (2004)
233e243.
[25] P. Lidstrom, J. Tierney, B. Wathey, J. Westman, Microwave assisted organic
[48] C.A. Paronis, S.P. Nikas, V.G. Shukla, A. Makriyannis, D
9-Tetrahydrocannabinol
€
synthesisda review, Tetrahedron 57 (2001) 9225e9283.
[26] F. Messina, O. Rosati, Superheated water as solvent in microwave assisted
organic synthesis of compounds of valuable pharmaceutical interest, Curr.
Org. Chem. 17 (2013) 1158e1178.
[27] R. Brenneisen, in: M. ElSohly (Ed.), Marijuana and the Cannabinoids, Humana
Press, 2007, pp. 17e49.
acts as a partial agonist/antagonist in mice, Behav. Pharmacol. 8 (2012)
802e805.
[49] G. Murineddu, F. Deligia, A. Dore, G. Pinna, B. Asproni, G.A. Pinna, Different
classes of CB2 ligands potentially useful in the treatment of pain, Recent Pat.
CNS Drug Discov. 1 (2013) 42e69.
[50] T. Kenakin, A. Christopoulos, Signalling bias in new drug discovery: detection,
[28] K. Worm, Q. Zhou, G. Stabley, R. DeHaven, R. Dolle, Biaryl cannabinoid mi-
metics e synthesis and structure e activity relationship, Bioorg. Med. Chem.
Lett. 17 (2007) 3652e3656.
[29] A.C. Howlett, T.M. Champion, G.H. Wilken, R. Mechoulam, Stereochemical
effects of 11-OH-delta 8-tetrahydrocannabinol-dimethylheptyl to inhibit
adenylate cyclase and bind to the cannabinoid receptor, Neuropharmacol 29
(1990) 161e165.
quantification and therapeutic impact, Nat. Rev. Drug. Discov.
205e216.
3 (2013)
[51] R. Chen, J. Zhang, N. Fan, Z. Teng, Y. Wu, H. Yang, Y. Tang, H. Sun, Y. Song,
C. Chen,
D
9-THC-caused synaptic and memory impairments are mediated
through COX-2 signaling, Cell 155 (2013) 1154e1165.
[52] A. Chicca, J. Marazzi, J. Gertsch, The antinociceptive triterpene b-amyrin in-
hibits 2-arachidonoylglycerol (2-AG) hydrolysis without directly targeting
[30] M. Rhee, Z. Vogel, J. Barg, M. Bayewitch, R. Levy, L. Hanus, A. Breuer,
R. Mechoulam, Binding to cannabinoid receptors and inhibition of adeny-
lylcyclase, J. Med. Chem. 40 (1997) 3228e3233.
cannabinoid receptors, Br. J. Pharmacol. 167 (2012) 1596e1608.