New metabolically stable fatty acid amide ligands of cannabinoid receptors: Synthesis and receptor affinity studies

Article abstract of DOI:10.1016/j.bmcl.2005.09.023

We investigated the structure-activity relationships for the interactions of fatty acid amide analogs of the endocannabinoid anandamide with human recombinant cannabinoid receptors. Thirty-five novel fatty acid amides were synthesized using five different

Full text of DOI:10.1016/j.bmcl.2005.09.023

Bioorganic & Medicinal Chemistry Letters 16 (2006) 138–141
New metabolically stable fatty acid amide ligands of
cannabinoid receptors: Synthesis and receptor affinity studies
Paolo Urbani,a,b Paolo Cavallo,a,b Maria Grazia Cascio,a,b Mariafrancesca Buonerba,b
Giovanni De Martino,b Vincenzo Di Marzoa,* and Carmela Saturninob
aEndocannabinoid Research Group, Institute of Biomolecular Chemistry, Consiglio Nazionale delle Ricerche,
via Campi Flegrei 34, Comprensorio Olivetti, Bldg. 70, 80078 Pozzuoli (NA), Italy
`
Department of Pharmaceutical Science, Universita degli Studi di Salerno, via Ponte Don Melillo, 84084 Fisciano (SA), Italy
b
Received 7 August 2005; revised 6 September 2005; accepted 12 September 2005
Available online 6 October 2005
Abstract—We investigated the structure–activity relationships for the interactions of fatty acid amide analogs of the endocannab-
inoid anandamide with human recombinant cannabinoid receptors. Thirty-five novel fatty acid amides were synthesized using five
different types of acyl chains and 11 different aromatic amine Ôheads.Õ Although none of the new compounds was a more potent
ligand than anandamide, we identified three amine groups capable of improving the metabolic stability of arachidonoylamides
and their CB1/CB2 selectivity ratio to over 20-fold, and several aromatic amines capable of improving the affinity of short chain
or monosaturated fatty acids for cannabinoid CB1 receptors. For the first time a tertiary amide of arachidonic acid was found
to possess moderate affinity (Ki = 300 nM) for cannabinoid CB1, but not CB2, receptors.
Ó 2005 Elsevier Ltd. All rights reserved.
Anandamide (N-arachidonoyl-ethanolamine) is one of
the two best studied endogenous ligands of cannabinoid
receptors. This compound binds both the CB1 and CB2
subtypes of cannabinoid receptors with sub-micromolar
affinity (Ki ranging between 40 and 250 nM depending
on the assay conditions),1,2 although it behaves as a par-
tial agonist at CB1 receptors and is almost functionally
inactive at CB2 receptors.3 The structure–activity rela-
tionships for the interactions of anandamide with both
cannabinoid receptors have been widely investigated,2
and have led, among other things, to the only selective
CB1 receptor ligands identified to date,4 N-arachido-
nyl-2-chloroethylamide and N-arachidonyl-cyclopro-
pylamide. However, these compounds lack metabolic
stability.4 With the present study we aimed at: (1)
expanding our knowledge of the structure–activity rela-
tionships for anandamide interactions with human CB1
and CB2 receptors, and (2) developing CB1-selective li-
gands with higher metabolic stability, by studying the
affinity of 35 novel N-acyl-amides obtained from the
condensation of 5 different fatty acids with 6 or, as in
the case of arachidonic acid, 11 different aromatic amine
Ôheads.Õ We report for the first time the development of
weak non-arachidonoyl-containing CB1 ligands with
some selectivity over CB2, and of metabolically stable
and moderate affinity, CB1-selective, arachidonic acid
amide ligands.
As shown in Scheme 1, the synthesis of the amides was
accomplished by direct condensation between trans-2-un-
decenoic acid (tail 1), 10-undecenoic acid (tail 2), erucic
acid (tail 3), oleic acid (tail 4), or arachidonic acid (tail
5), and several amines, that is, 4-methoxyaniline (head
A), 4-methoxybenzylamine (head B), 4-methoxy-phen-
ethylamine (head C), 3,5-di-methoxybenzylamine (head
D), 2,4-di-methoxybenzylamine (head E), 4-(4-morpholi-
no)aniline (head F), using 1-propylphosphonic acid cyclic
anhydride (PPAA) as catalyst, with 65–75% yields5 (see
Supplementary materials). For the arachidonoyl tail, we
utilized five more amines, that is, 2,5-dimethoxybenzyl-
amine (UP61), 2,4-di-methoxy-benzylamine (UP63),
1-(3,4-dimethylphenyl) piperazine (UP67), 1-(4-chloro-
phenyl)piperazine (UP69), and 1-(4-fluoro-phenyl)piper-
azine (UP70). The affinity for cannabinoid receptors of
the synthesized amides was assessed by binding assays,6
and the results obtained are reported in Table 1.
Keywords: Cannabinoid; Anandamide; CB1; CB2; Receptor.
*
Data obtained from the binding assays allow us to make
the following considerations regarding the importance
Corresponding author. Tel.: +39 081 8675093; fax: +39 081
8041770; e-mail: vdimarzo@icmib.na.cnr.it
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.09.023

P. Urbani et al. / Bioorg. Med. Chem. Lett. 16 (2006) 138–141
139
O
vates (UP61, 63, 67, 69 and 70), which also showed quite
good affinity and selectivity for CB1 receptors.
O
+ H2N-Y
a
Y
65-75%
X
OH
X
N
These findings are in full agreement with previous find-
ings on the role of the alkyl chain in fatty acid amide
interactions with the cannabinoid receptors.2,8 However,
we have also shown here that the affinity of alkyl chains
(i.e., trans-2-undecenoic, oleic and erucic) that, when
amidated with ethanolamine, were previously found to
be inactive at these receptors (Ki > 5 lM, V. Di MarzoÕs
unpublished observations) can be significantly improved
when using certain aromatic amines.
H
“Tails”
“Heads”
1
A
2
3
4
5
B
C
O
O
O
O
N
O
N
H
H
O
UP61 (CB1 Ki = 0.5 µM;
UP63 (CB1 Ki
CB2 Ki = 4.0 µM)
=
0.2 µM;
CB2 Ki = 4.1 µM)
D
O
O
N
N
N
N
Cl
UP67 (CB1 Ki = 0.4 µM;
Ki = 4.2 µM)
UP69 (CB1 Ki = 0.6 µM; CB2
CB2 Ki > 10 µM)
E
F
O
N
a= PPAA, Et3N,
CH2Cl2
N
F
UP70 (CB1 Ki = 0.3 µM; CB2 Ki = 5.1 µM)
Scheme 1. General scheme for the synthesis of the new compounds.
We also identified new structural requirements for the
design of fatty acid amides not only with improved affin-
ity with cannabinoid receptors, but also with higher
selectivity over CB2 than anandamide. In particular:
of the alkyl chain in fatty acid amide ligands of cannab-
inoid receptors:
(1) In the trans-2-undecenoic acid series, some notable
affinity for CB1 receptors was observed and the most
active compounds were UP3-238 and UP4-14, which,
however, exhibited only a ꢀ2-fold selectivity over CB2
receptors. UP1-13 showed a slightly higher affinity for
CB2. Interestingly, the affinity for both CB1 and CB2
receptors was lost if the double bond in the acyl chain
was moved to D10. This indicates that a certain rigidity
near the amide group is necessary for these ÔshortÕ fatty
acid amides to interact with the two receptors.
(1) The addition of a methylene group between the
amide and the para-methoxy-phenyl group increases
affinity for CB1 receptors in the oleic acid series, decreas-
es affinity for both CB1 and CB2 in the trans-2-undece-
noic acid series and, unlike the para-hydroxy-phenyl
ÔheadÕ,7 has no effect in the case of the arachidonoyl
derivative.
(2) The addition of a second methylene group between
the amide and the para-methoxy-phenyl group increases
affinity for both CB1 and CB2 receptors of the trans-2-
undecenoic acid derivative, and worsens it in the case
of arachidonate derivative, with no effect on the oleic
acid derivative.
(2) When long-chain mono-unsaturated acyl chains were
used, notable affinity for CB1 receptors was observed
mostly for oleic acid derivatives, which were almost all
more potent than the corresponding trans-2-undecenoic
acid derivatives. In particular, we observed the following
rank of activity: UP28 > UP27-18 > UP30 > UP29 >
UP26-7 > UP31-8. A higher than 20-fold selectivity over
CB2 receptors was also found, for example, with UP28
and UP27-18. Interestingly, the same order of activity
was not observed for those erucic acid derivatives that
exhibited some affinity for cannabinoid receptors.
(3) The addition of an ortho-methoxy group to the para-
methoxy-phenylethyl derivatives decreases affinity for
CB1 receptors of the trans-2-undecenoic and oleic acid
derivatives and also for CB2 receptors in the case of
trans-2-undecenoic acid; it improves affinity for both
receptors in the case of the arachidonic acid derivative;
and it improves affinity only for CB1 receptors for the
erucic acid derivative. Addition of a ortho-methoxy-
group (UP63) to the para-methoxy-benzyl derivative of
arachidonic acid (UP58) also slightly improves affinity
for CB1. Methylation of the para-phenyl group results
(3) The derivatives of arachidonic acid yielded, as
expected, the most potent ligands. They exhibited not
only the highest affinity for CB1 receptors, but also
selectivity for this receptor type over CB2 receptors.
For this reason we tested five more arachidonate deri-

140
P. Urbani et al. / Bioorg. Med. Chem. Lett. 16 (2006) 138–141
Table 1. Affinity constants for human recombinant cannabinoid CB1
and CB2 receptors as calculated in binding assays
Table 2. Affinity constants of some of the most potent ligands
developed in this study as calculated in binding assays carried out in
the absence or presence of PMSF (200 lM)
Compounds
X
Y
Human CB1
Ki (lM)
Human CB2
Ki (lM)
Compounds
Human CB1
FAAH
IC50 (lM)
UP1-13
UP2
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
4
4
4
5
5
5
5
5
5
A
B
C
D
E
F
A
B
C
D
E
F
A
B
C
D
E
F
A
B
C
D
E
F
A
B
C
D
E
F
4.9
>10
2.4
2.5
>10
5.1
Ki (lM) without
PMSF
Ki (lM) with
PMSF
UP3-238
UP4-14
UP5
UP26-7
UP27-18
UP28
1.6
0.5
0.4
1.3
0.8
0.2
0.15
0.3
1.5
0.6
0.3
1.0
0.6
0.2
0.15
0.3
>25
>25
>25
>25
>25
>25
>25
>25
2.8
5.2
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
5.4
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
4.0
UP7-229
UP10
UP29
UP30
UP11-5
UP12
UP13
UP63
UP66
UP70
UP14
UP16-28
UP19
Data are means of three measurements. SD values were 610%.
UP20-2
UP21
(6) Again for the first time, tertiary amides of arachidon-
ic acid were found to significantly bind to CB1 but not
CB2 receptors; the affinity seems to improve in the pres-
ence of not too hindering para electronegative groups in
the aromatic moiety.
UP22-3
UP23
3.7
6.4
UP24-11
UP26-7
UP27-18
UP28
1.6
0.5
0.4
1.3
The stability of these compounds to enzymatic hydroly-
sis was assessed in two ways: (1) by testing the effect of
the thirty-five compounds on [14C]anandamide hydroly-
sis by rat brain membranes, where a very abundant fatty
acid amide hydrolase (FAAH) activity is present;9 (2) by
assaying some of the highest affinity ligands in binding
assays carried out in the presence of the non-selective
hydrolase inhibitor PMSF (200 lM). The fact that none
of the new compounds was capable of inhibiting
anandamide hydrolysis up to a 25 lM concentration
(Table 2 and data not shown) indicates that they are
not substrates for FAAH, and hence not hydrolysed
by this enzyme. The fact that the affinity of some of
the most potent compounds did not improve in the pres-
ence of PMSF (Table 2) indicates that, unlike the other
previously identified CB1-selective fatty acid amides,4
they were not easily hydrolysed during the binding assay
conditions.
UP29
UP30
0.8
UP31-8
UP57
UP58
>10
0.5
0.6
4.5
UP59
UP62
2.4
>10
0.5
>10
>10
4.0
UP60
UP66
0.15
0.072
0.021
0.013
1.767
4.3
Anandamide
WIN55,212
CP55,940
AM404
0.18
0.002
0.0007
—
Ki values of reference compounds tested here are also shown. Data are
means of three measurements. SD values were never higher than 10%.
in a 3-fold higher affinity for CB1 receptors (compare
UP57 and AM404).
In conclusion, we have identified here: (1) at least three
novel (UP63, UP66 and UP70) moderate affinity
(Ki = 150–300 nM) ligands for cannabinoid CB1 recep-
tors with 17- to 28-fold selectivity over CB2 receptors
and metabolic stability to enzymatic hydrolysis; addition-
ally, one of these compounds, UP70, by containing a ter-
tiary amine, might yield an ammonium salt under acidic
conditions, which might result in more water-soluble
CB1 ligand than those currently available; and (2) new
chemical features necessary to fatty acid amides to inter-
act with cannabinoid CB2 and, particularly, CB1 recep-
tors; in particular, three amine groups, when amidated
to oleic acid, which is less expensive and less sensitive to
oxidation than arachidonic acid, yield fatty acid amides
(UP27-18, UP28 and UP30) that exhibit metabolic stabil-
ity, sub-micromolar affinity for CB1 receptors (Ki =
400–800 nM) and 10- to >20-fold selectivity over CB2
receptors. To the best of our knowledge these three com-
pounds are the first ligands of cannabinoid receptors to be
developed from oleic acid and might therefore be suitable
for in vivo use. The agonist or antagonist activity of these
(4) The shift of the two methoxy groups from ortho–para
to meta–meta in di-methoxyphenylethyl derivatives im-
proves affinity for both CB1 and CB2 receptors with
trans-2-undecenoic acid, decreases affinity for both
receptors in the case of arachidonic acid, and decreases
affinity for CB1 receptors in the case of the oleoyl and
erucoyl derivatives. In the case of the dimethoxybenzyl
derivatives of arachidonic acid, also the shift of just
one methoxy group from para to meta causes a slight de-
crease of affinity for CB1 receptors (compare UP63 and
UP61).
(5) The amidation of arachidonic acid, but not of the
other fatty acids, with 2,5-dimethoxybenzylamine
(UP61), 2,4-dimethoxybenzylamine (UP63), 1-(3,4-dim-
ethylphenyl)piperazine (UP67), 1-(4-chloro-phenyl)pi-
perazine (UP69), 1-(4-fluoro-phenyl)piperazine (UP70)
and, particularly, 4-(4-morpholino)aniline (UP 66) was
shown here for the first time to result in moderate to
high affinity ligands for CB1 but not CB2 receptors.

P. Urbani et al. / Bioorg. Med. Chem. Lett. 16 (2006) 138–141
141
Column chromatography was carried out using Merck
silica gel 60 (230–400 mesh).
new compounds will have to be tested next, using the
appropriate functional assays.
6. For CB1 and CB2 receptor binding assays, the new
compounds were tested by using P2 membranes from
HEK cells transfected with either the human CB1 or CB2
receptor and [3H]-(À)-cis-3-[2-hydroxy-4-(1,1-dimethyl-
heptyl)-phenyl]-trans-4-(3-hydroxy-propyl)-cyclohexanol
([3H]CP-55,940) (Kd = 0.27 nM) as the high affinity
ligand, as described by the manufacturer (Perkin-Elmer,
Italia). In some experiments, membranes were pre-incu-
bated for 10 min with PMSF (200 lM), which was also left
during the incubation. Displacement curves were generat-
ed by incubating drugs with 0.27 nM [3H]CP-55,940. In all
cases, Ki values were calculated by applying the Cheng–
Prusoff equation to the IC50 values (obtained by Graph-
Pad) for the displacement of the bound radioligand by
increasing concentrations of the test compounds.
7. Khanolkar, A. D.; Abadji, V.; Lin, S.; Hill, W. A.;
Taha, G.; Abouzid, K.; Meng, Z.; Fan, P.; Makriyannis,
A. J. Med. Chem. 1996, 39, 4515.
8. Di Marzo, V.; Bisogno, T.; De Petrocellis, L.; Melck, D.;
Martin, B. R. Curr. Med. Chem. 1999, 6, 721.
9. The effect of increasing concentrations of the synthetic
compounds on the enzymatic hydrolysis of anandamide
was studied as described previously10 by using membranes
prepared from rat brain. Membranes were incubated with
increasing concentrations (1–5–10–25–50 lM) of the test
compounds and [14C]anandamide, 4 lM (10,000 cpm) in
50 mM Tris–HCl, pH 9, for 30 min at 37 °C. [14C]Etha-
nolamine produced from [14C]anandamide hydrolysis was
used to calculate FAAH activity and was measured by
scintillation counting of the aqueous phase after extraction
of the incubation mixture with 2 volumes of CHCl3/
MeOH 1:1 (by vol.).
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.2005.
09.023.
References and notes
1. Devane, W. A.; Hanus, L.; Breuer, A.; Pertwee, R. G.;
Stevenson, L. A.; Griffin, G.; Gibson, D.; Mandelbaum,
A.; Etinger, A.; Mechoulam, R. Science 1992, 258, 1946.
2. Martin, B. R.; Mechoulam, R.; Razdan, R. K. Life Sci.
1999, 65, 573.
3. McAllister, S. D.; Glass, M. Prostaglandins Leukot.
Essent. Fatty Acids 2002, 66, 161, Review.
4. Hillard, C. J.; Manna, S.; Greenberg, M. J.; DiCamelli, R.;
Ross, R. A.; Stevenson, L. A.; Murphy, V.; Pertwee, R. G.;
Campbell, W. B. J. Pharmacol. Exp. Ther. 1999, 289, 1427.
5. To a solution of fatty acid in DCM, TEA (4.0 equiv) and
PPAA (1.2 equiv) were added with stirring. After 15 min
at room temperature, the appropriate amine (1.2 equiv)
was added and the mixture was stirred at room temper-
ature for 2 h. The organic phase was evaporated under
reduced pressure. The residue was purified by column
chromatography on silica gel using hexane/AcOEt as
eluants. Melting points were determined with a Gallenk-
amp hot stage apparatus. Infrared spectra were recorded
with a Shimadzu 800 spectrophotometer. 1H NMR
spectra (see Supplementary materials) were measured with
a Bruker 300 MHz spectrometer using CDCl3 as solvent.
10. Maurelli, S.; Bisogno, T.; De Petrocellis, L.; Di Luccia, A.;
Marino, G.; Di Marzo, V. FEBS Lett. 1995, 377, 82.