Acylamido analogs of endocannabinoids selectively inhibit cancer cell proliferation

Article abstract of DOI:10.1016/j.bmc.2008.10.015

A series of amide derivatives of long-chain fatty acids has been studied for their effects on the proliferation of cancer cells in vitro. Fatty acids ranged from palmitic to higher polyunsaturated types containing 22 carbon atoms. The amino portions of the molecules included ammonia, ethanolamine, various amino acids and dopamine. Several cell lines were used as models and these included HTB-125 (normal human breast cells), HTB-126 (human breast cancer cells), HeLa (cervical cancer cells), WI-38 (human embryonic lung cells), RAW264.7 (mouse macrophage tumor cells) and RBL-2H3 (rat basophilic leukemia cells). The HTB lines were obtained from the same donor, so, could be considered a matched pair, that is, normal control versus cancer cells and thus, provide a model for testing specificity of action for the acylamido analogs. While many compounds were efficacious in inhibiting the proliferation of various cell lines, only two analogs showed a high degree of specificity in the matched HTB cell lines. N-palmitoyl dopamine and N-palmitoyl tyrosine each demonstrated complete specificity of action at a concentration of 10 μM and were highly efficacious in both cases. No clear structure-activity pattern could be derived from these studies since the intensity of the inhibitory action seemed to depend on three factors, namely, the fatty acid, the amine group and the cell type.

Full text of DOI:10.1016/j.bmc.2008.10.015

Bioorganic & Medicinal Chemistry 16 (2008) 9644–9651
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Acylamido analogs of endocannabinoids selectively inhibit
cancer cell proliferation
*
Sumner Burstein , Rebecca Salmonsen
Department of Biochemistry and Molecular Pharmacology, The University of Massachusetts Medical School, Worcester, MA 01605, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of amide derivatives of long-chain fatty acids has been studied for their effects on the prolifer-
ation of cancer cells in vitro. Fatty acids ranged from palmitic to higher polyunsaturated types containing
22 carbon atoms. The amino portions of the molecules included ammonia, ethanolamine, various amino
acids and dopamine. Several cell lines were used as models and these included HTB-125 (normal human
breast cells), HTB-126 (human breast cancer cells), HeLa (cervical cancer cells), WI-38 (human embryonic
lung cells), RAW264.7 (mouse macrophage tumor cells) and RBL-2H3 (rat basophilic leukemia cells). The
HTB lines were obtained from the same donor, so, could be considered a matched pair, that is, normal
control versus cancer cells and thus, provide a model for testing specificity of action for the acylamido
analogs. While many compounds were efficacious in inhibiting the proliferation of various cell lines, only
two analogs showed a high degree of specificity in the matched HTB cell lines. N-palmitoyl dopamine and
Received 28 August 2008
Revised 3 October 2008
Accepted 4 October 2008
Available online 11 October 2008
Keywords:
Cell proliferation
Endocannabinoids
HTB-126
HeLa
N-palmitoyl tyrosine each demonstrated complete specificity of action at a concentration of 10 lM and
WI-38
RAW264.7
Mast cells
Human
Breast cancer
Macrophage
Embryonic
Fibroblast
were highly efficacious in both cases. No clear structure–activity pattern could be derived from these
studies since the intensity of the inhibitory action seemed to depend on three factors, namely, the fatty
acid, the amine group and the cell type.
Ó 2008 Elsevier Ltd. All rights reserved.
O
1. Introduction
O
C
C
NH2
NH
The discovery of an endogenous cannabinoid system (ECS) has
fueled an intense interest in identifying the roles that this complex
may have in the regulation of cell function and the possible signif-
icance for health care strategies.1 Various studies have addressed
this question in both physiological and pathological processes at
central as well as peripheral sites. Thus far, the endocannabinoid
system has been shown to be involved in the immune system,
energy balance, stress recovery, food intake, metabolic homeosta-
sis, cardiovascular depressive effects, endocrine axes, fertility, pain
modulation and neuroprotection. The ECS may also have a role in
cancer biology because of its effects on cell proliferation.2
Many of the substances, and related analogs, that modulate this
system are lipid in nature and generally are comprised of a long-
chain fatty acid coupled to a more polar group such as ethanol-
amine, glycerol, dopamine or one of a number of amino acids
(Fig. 1). The fatty acid component of the structure can range from
a fully saturated acid such as palmitic to a higher polyunsaturated
type such as arachidonic. To date, fairly extensive groups of these
ligands or mediators have been identified and to some degree
R
R
d
R
NH
a
O
C
OH
OH
OH
b
O
O
R'
C
C
COOH
NH
R
COOH
NH
R
e
c
Figure 1. Chemical structures. (a) N-acyl amides. (b) N-acyl ethanolamides. (c)
N-acyl amino acids (elmiric acids) R0 = H, CH3, (d) N-acyl dopamine. (e) N-acyl GABA
(c-aminobutyric acid).
characterized.3 Thus, it is probably safe to say that the extent of
this family of compounds is far greater than what has thus far been
discovered.
Our own interest is focused primarily on the amino acid conju-
gates that we have named elmiric acids (EMA).4 These analogs do
not produce typical cannabimimetic actions in animal models,
however, they do show both analgesic5–7 and anti inflammatory8
properties in vivo. In addition, they exhibit a number of in vitro
* Corresponding author. Tel.: +1 508 856 2850; fax: +1 508 856 2003.
E-mail address: sumner.burstein@umassmed.edu (S. Burstein).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.10.015

S. Burstein, R. Salmonsen / Bioorg. Med. Chem. 16 (2008) 9644–9651
9645
properties such as inhibition of fatty acid amide hydrolase (FAAH),9
40000
35000
30000
25000
20000
15000
10000
5000
DMSO
effects on prostaglandin profiles,4 calcium mobilization and firing
Rx
N-palmitoyl tyrosine
Ajulemic Acid
rates in dorsal root ganglia,10,11 and cell proliferation.4
*
*
The plant derived cannabinoids, or phytocannabinoids, have
been studied as possible anti cancer agents for some time.12 Vari-
ous mechanisms have been suggested to help explain their anti
proliferative actions and these may have relevance for the effects
of the endocannabinoids as well.13–16 Tetrahydrocannabinol
(THC), the principle psychotropic ingredient in Cannabis, has long
been used in the form of the drug dronabinol, to reduce the side ef-
fects of chemotherapy, but, its impact on the course of the disease
in humans is not clear. Synthetic analogs of THC have been pre-
pared and studied for a possible use in cancer therapy. Such a com-
pound, ajulemic acid (AJA) is a peripheralized,17 non psychoactive
analog with moderate anti cancer actions in preclinical models.18
Nabilone, another THC analog, is used clinically to control nausea
and emesis in patients receiving chemotherapy, however, its possi-
ble effects on tumor progression has not been reported.19
*
*
p<0.005 vs DMSO
0
0
10 20 30 40 50 60 70 80 90 100 110
TIME (hrs)
Figure 2. Time course in HTB-126 breast cancer cells. Cells derived from a human
tumor were obtained from ATCC and seeded in 18 wells of four 24 well plates at
2000 cells/well under the conditions described in Section 6.1.1. Following a 24 h
incubation period, each plate was treated with vehicle, N-palmitoyl tyrosine and
AJA (N = 6). After 60 min, one plate was subjected to the cell proliferation assay (see
Section 6.1.1), a second plate was assayed after 24 h, a third after 48 h and the last
after 72 h. Both agents were given at a final concentration of 10
stock solution in DMSO. Values shown are in relative fluorescence units.
lM from a 100ꢀ
We previously carried out a limited study on the inhibition of
proliferation by several EMAs in a mouse macrophage tumor-
derived cell line, RAW264.7.8 Full dose–response data were not
obtained in that study; however, a complete inhibition of prolifer-
human tumor-derived breast cancer cells (HTB-126) were plated
at time zero. All three treatments were applied at 24 h and cell
numbers measured two hours later. Subsequent measurements,
on parallel plates, were made at 48, 72 and 96 h post plating time.
The EMA showed an immediate anti proliferative effect whereas
the AJA showed no effect initially. At 48 and 72 h the two agents
produced the same response while at 96 h the beginning of a diver-
gence was observed in which the inhibitory effect of AJA appeared
to be increasing.
ation was seen between 1 and 10 lM. Only the effects of glycine
and alanine conjugates were reported and their polyunsaturated
derivatives appeared to be more active than the saturated
analogs. The limited nature of the study did not allow any detailed
conclusions regarding structure–activity relationships (SAR) to be
made.
In subsequent experiments reported here, we sought to extend
the previous work to a series of new analogs and to expand the
number and variety of cell types studied with these compounds.
In addition, more complete relationships between drug concentra-
tion and cell proliferation have been obtained. Of particular inter-
est, is an observed selectivity of inhibition when the effect on
tumor cells is compared with a matched normal control, which
was observed for certain compounds in a model of human breast
cancer.
3.2. Selective inhibition of human breast cancer cells (HTB-126)
by acylamido analogs
Figure 3(A–E) shows dose–response data for several acylamido
analogs in which the effects on human tumor-derived breast can-
cer cells (HTB-126) were compared with normal human breast
cells (HTB-125). The best separation of activity was obtained with
N-palmitoyl dopamine (Fig. 3A), which had no effect on the normal
(HTB-125) cells at concentrations up to 10 lM but produced an
80% inhibition of tumor cell (HTB-126) proliferation. The other
dopamine analogs were also inhibitory under the same conditions,
however, they did exhibit moderate anti proliferative effects on
normal cells (Fig. 3B–D).
N-palmitoyl tyrosine, an example of the EMA family, gave a re-
sponse pattern similar to that of N-palmitoyl dopamine in the
same model (Fig. 3A and E). Here, a vigorous inhibition of tumor
2. Chemistry
Only one of the compounds used in this study, N-arachidonoyl-
L-tyrosine; L-EMA-10 (16:0) (see Ref. 8 for nomenclature), was syn-
thesized in the university laboratories using the procedures de-
scribed below in Section 6.1.2. Briefly this involved condensation
of the mixed anhydride palmitoyl-i-butylchloroformate with L-
tyrosine methyl ester in the presence of triethylamine followed
by saponification with lithium hydroxide to yield the EMA. The rest
of the compounds came from a commercially available library of
endocannabinoids obtained from Biomol International LP (Plym-
outh Meeting PA 19462) that contains sixty examples, half of
which would fall into the category of EMAs. They are composed
cell proliferation was seen at 10
lM with no changes at 0.1–
1.0 M concentrations, similar to the dopamine analog responses.
l
All of the values presented in Figure 3A–E were normalized to
100% for the vehicle treated controls.
Some insight into the mechanism of action of N-palmitoyl tyro-
sine was obtained using a receptor antagonist approach. Panel F
shows the effect of a CB1 antagonist, rimonabant (SR141716), on
the anti proliferative response of N-palmitoyl tyrosine in HTB-
126 tumor cells. The profound decrease in proliferation caused
of glycine, L-alanine and c-aminobutyric acid (GABA) conjugates
with ten different naturally occurring long-chain fatty acids rang-
ing from palmitic to docosahexaenoic acid. Also included are sets
of simple amine, ethanolamine and N-dopamine derivatives of
the same fatty acids.
by treatment with N-palmitoyl tyrosine (10
lM) was totally re-
versed by rimonabant (10 M). The antagonist alone produced a
l
small increase in tumor cell number. The values in this panel are
given as fluorescence units, which are directly proportional to cell
number.
3. Results
3.1. Time course
3.3. Effects of acyl dopamide analogs on proliferation of
cervical cancer (HeLa) cells
In Figure 2 the time course for the anti proliferative effect of one
of the EMAs, N-palmitoyl tyrosine (10
l
M) was compared to vehi-
The anti proliferation effects of a series of dopamine-unsatu-
rated fatty acid conjugates in a different tumor cell type was stud-
ied and the results are shown in Figure 4. In cervical cancer (HeLa)
cle (DMSO) treated cells and AJA (10
lM), a cannabinoid analog
with anti tumor activity.18 Equal numbers (2000 cells/well) of

9646
S. Burstein, R. Salmonsen / Bioorg. Med. Chem. 16 (2008) 9644–9651
140
120
100
80
175
B
D
F
A
ARACHIDONOYL DOPAMINE/125
ARACHIDONOYL DOPAMINE/126
150
125
100
75
60
50
40
PALMITOYL DOPAMINE/125
PALMITOYL DOPAMINE/126
25
20
0
0
0.01
0.1
1
10
100
0.01
0.1
1
10
100
CONC [uM]
CONC [uM]
120
100
80
120
100
80
60
40
20
0
C
60
40
g-LINOLENOYL DOPAMINE/125
g-LINOLENOYL DOPAMINE/126
20
0
OLEOYL DOPAMINE/125
OLEOYL DOPAMINE/126
0.01
0.1
1
10
100
0.01
0.1
1
10
100
CONC [uM]
CONC [uM]
140
120
100
80
32000
24000
16000
8000
0
PALMITOYL TYROSINE/125
E
p<0.0001
*
60
40
20
DMSO
PalTyr
Rimonabant
+
-
-
+
+
-
+
-
+
+
+
+
PALMITOYL TYROSINE/126
0
0.01
0.1
1
10
100
CONC [uM]
Figure 3. Selective inhibition of human breast cancer cells (HTB-126) versus normal (HTB-125) cells by acylamido analogs. Cells were grown and treated under conditions
similar to those described in Figure 2 except that 96 well plates were used and four replicates were done for each point. The cell proliferation assay was performed at 48 h
following drug treatment and the values in (A–E) are expressed as a percentage of the vehicle treated wells. In (F), the antagonist rimonabant (10
lM) was added 30 min prior
to the treatment with N-palmitoyl tyrosine (10
lM) and the assay performed 48 h later. Here the values are shown as fluorescence units since there are two control
treatments.
cells, all of the analogs tested produced strong inhibition at 10
l
M
A similar study was done with a possibly more relevant series of
analogs in which the polar group is ethanolamine (Table 2). Again,
in these two cell types most of the compounds were not especially
potent with one possible exception, namely, docosatetra-
7Z,10Z,13Z,16Z-enoyl ethanolamide in the WI-38 model. Not
shown are the results of a very similar experiment done in the
same models where a series of N-fatty acid GABA conjugates were
tested. In almost every case, modest stimulatory actions were seen
with one exception. N-oleoyl GABA produced a small decrease in
cell proliferation in RAW cells.
concentrations. A tendency toward stimulation was seen at lower
concentrations analogous to the N-palmitoyl tyrosine effect shown
in Figure 3E. A negative control agent, N-palmitoyl glycine, had no
effect as expected and a positive control, AJA, appeared to be less
active than the dopamine derivatives, however, this may be due
to a time course effect (see Fig. 2).
3.4. Effects of acylamides on human embryonic lung (WI-38)
and mouse macrophage tumor (RAW264.7) cell proliferation
The simplest examples of acylamido analogs are those where
the polar group is NH2. In Table 1 are data from a single concentra-
tion (10 lM) treatment of human embryonic lung fibroblasts and
3.5. The effects of N-acyl dopamide conjugates on human
embryonic lung (WI-38) and mouse macrophage tumor
(RAW264.7) cell proliferation
mouse macrophage cancer cells with a series of such fatty acid
amides ranging from palmitamide to docosahexaeneamide. None
of the analogs was particularly effective in either cell type, thus,
there were no clear structure–activity trends.
One other class of compounds, the N-fatty acid dopamine type,
was studied in the WI-38/RAW cell model and the findings
are shown in Table 3. Here, in almost every example, a robust

S. Burstein, R. Salmonsen / Bioorg. Med. Chem. 16 (2008) 9644–9651
9647
Docosatetraenoyl Dopamine [10]
Docosatetraenoyl Dopamine [3]
Docosatetraenoyl Dopamine [1]
Docosatetraenoyl Dopamine [0.1]
Eicosapentaenoyl Dopamine [10]
Eicosapentaenoyl Dopamine [3]
Eicosapentaenoyl Dopamine[1]
Eicosapentaenoyl Dopamine [0.1]
DH-g-linolenoyl Dopamine [10]
DH-g-linolenoyl Dopamine [3]
DH-g-linolenoyl Dopamine [1]]
DH-g-linolenoyl Dopamine [0.1]
g-Linolenoyl Dopamine [10]
g-Linolenoyl Dopamine [3]
g-Linolenoyl Dopamine [1]
g-Linolenoyl Dopamine [0.1]
Oleoyl Dopamine [10]
Oleoyl Dopamine [3]
Oleoyl Dopamine [1]
Oleoyl Dopamine [0.1]
Ajulemic acid [10]
Ajulemic acid [4]
Palmitoyl glycine [20]
DMSO
0
25
50
75 100 125 150 175
% DMSO
Figure 4. Effects of acyl dopamide analogs on proliferation of cervical cancer (HeLa) Cells. The cells were seeded in 96 well plates and cultured as described in Section 6.1.1.
Treatments were made at 24 h after seeding and the assays performed at 48 h post treatment. Four replicates were done for each and the concentrations are shown as the
numbers in brackets in lM units. DH = dihomo. Values are mean SEM.
Table 1
3.6. Effects of palmitoyl and arachidonoyl derivatives on LPS-
stimulated, rat basophilic leukemia (RBL-2H3) cell proliferation
Effects of acylamides on human embryonic lung (WI-38) and mouse macrophage
tumor (RAW264.7) cell proliferation.
Treatment (c = 10
lM)
WI-38 cells
Control (%)
RAW cells
Control (%)
Finally, a study was done using rat basophyllic leukemia cells
(RBL-2H3) with examples of both EMAs and N-acyl dopamine ana-
SD
SD
DMSO
Palmitamide
Oleamide
100
118
94
85
33
55
58
58
61
61
89
17
17
22
35
14
18
7.6
17
30
23
15
100
100
134
126
94
74
82
91
64
16
7.3
12
28
13
7.5
9.5
10
13
10
22
logs (Table 4). The amino acid groups consisted of glycine, L-alanine
and GABA while in each case palmitoyl and arachidonoyl fatty acid
groups were included. All cells were stimulated with lipopolysac-
Linolenamide
charide (LPS) (10 ng/ml). Concentrations of 0.1, 1.0 and 10 lM
c
-Linolenamide
Eicosa-1Z,14Z-dienamide
Dihomo- -Linolenamide
were used for each compound; the results are shown both as
unprocessed data and also as the ratio of treated over control val-
ues. All of the compounds except for N-palmitoyl glycine, which
had no inhibitory action, were effective at 10 lM and both of the
dopamine derivatives seemed to be more potent than any of the
c
Arachidonamide
Eicosapentaenamide
Docosatetra-7Z,10Z,13Z,16Z-enamide
Docosahexaenamide
91
112
EMAs.
Cell culture conditions and treatment procedures are described in Section 6.1.1 and
in Figure 3.
4. Discussion
Table 2
Almost a century has passed since Paul Ehrlich used the term
‘‘silver bullet” to describe salvarsan, his drug for the treatment of
syphilis. Unfortunately, salvarsan did not exhibit the specificity
of action he sought, but this has not discouraged generations of sci-
entists since then from continuing the search for agents that will
only target disease sites leaving healthy tissue unaffected. This
has been particularly true in the area of cancer chemotherapy
where specificity of action remains a principle consideration in
designing anti cancer agents. In this report, we describe a family
of molecules, the acylamido analogs of the endocannabinoids,
which may provide an approach for advancing toward this goal.
In an earlier study, we presented preliminary findings showing
that a subset of the acylamido analogs called EMAs inhibited the
proliferation of the RAW264.7 murine leukemia virus-induced tu-
mor cell line.4 In that report, a number of the EMAs in the glycine
Effects of acylethanolamides on human embryonic lung (WI-38) and mouse
macrophage tumor (RAW264.7) cell proliferation.
Treatment (c = 10
lM)
WI-38 cells
RAW cells
Control (%) SD Control (%) SD
DMSO
100
130
140
63
54
58
54
74
92
17 100
32 66
31 117
16 100
23 87
13 54
6.5 70
38 110
27 83
7.4 65
17 36
16
5.2
23
10
23
21
30
35
18
11
15
Palmitoyl ethanolamide
Oleoyl ethanolamide
Linoleoyl ethanolamide
c
-Linoleoyl ethanolamide
Eicosa-1Z,14Z-dienoyl ethanolamide
Dihomo- -linolenoyl ethanolamide
c
Arachidonoyl ethanolamide
Eicosapentaenoyl ethanolamide
Docosatetra-7Z,10Z,13Z,16Z-enoyl ethanolamide 24
Docosahexaenoyl ethanolamide 50
Cell culture conditions and treatment procedures are described in Section 6.1.1 and
in Figure 3.
and
L
-alanine groups showed a profound inhibition of RAW cell
proliferation at
a
10 concentration. These observations
lM
prompted us to carry out further experiments to better character-
ize this effect and to examine the possibility of selectivity between
normal and cancer cells. It was also decided to look beyond the
EMAs and include other analogs (Fig. 1), for example, the N-dopa-
mine-fatty acid derivatives, because of reports on their effects in
cancer cells.21–23
To test for selectivity, we utilized two commercially available
cell lines (HTB-125 and HTB-126) that were derived from human
breast tissue (see Section 6.1.1 for details). The former was
inhibition of cell proliferation was found in both cell lines. Thus, as
a group, the dopamine conjugates were more active than any of the
others described in this report. Indomethacin has been reported to
inhibit glioma cell proliferation,20 and was included here as a posi-
tive control. However, it actually caused an increase in both cell
types suggesting that it may be very tissue specific in its possible
anti cancer activity.

9648
S. Burstein, R. Salmonsen / Bioorg. Med. Chem. 16 (2008) 9644–9651
Table 3
N-palmitoyl tyrosine (Fig. 3E). All of the data shown were normal-
ized to the control values obtained from vehicle treated cells under
identical conditions. Based on these limited data, we estimate a
rank order of selective inhibition to be: N-palmitoyl dopa-
The effects of N-acyl dopamide conjugates on human embryonic lung (WI-38) and
mouse macrophage tumor (RAW264.7) cell proliferation.
Treatment (c = 10
lM)
WI-38 cells
RAW cells
mine = N-palmitoyl tyrosine > N-arachidonoyl dopamine = N-c-lin-
Control (%) SD Control (%) SD
oleoyl dopamine > N-oleoyl dopamine. While no clear SAR can be
discerned from this small group of compounds, it seems reasonable
to conclude that selective action can be achieved with palmitoyl
derivatives. It would be of great interest to see whether similar re-
sults would be obtained in other matched models of anti cancer
activity. Moreover, these observations suggest that some of the
acylamido analogs would show a high therapeutic index in vivo.
The question of receptor mediation in the above experiments
has been briefly addressed and the results are shown in Figure
3F. Using the same model for which the other data in Figure 3 were
obtained, the effect of pretreatment with a CB1 receptor antago-
nist, rimonabant24,25 was studied. A complete reversal of the inhi-
bition of proliferation by N-palmitoyl tyrosine was found
suggesting that the CB1 receptor has an important role in this pro-
cess. There are no data on the levels of CB1 in either of these breast
tissue-derived cell lines, however, it is interesting to note that a re-
cent report gives data that cancer cells over express both CB1 and
CB2 receptors and that receptor activation inhibits prolifera-
tion.26,27 A greater expression level of CB1 in HTB-126 cells versus
HTB-125 cells might help to explain the selectivity found in our
study. Since N-arachidonyl glycine has been reported to have no
affinity for CB1,28 it may be inferred that N-palmitoyl tyrosine like-
wise does not bind. Thus, our observation with rimonabant could
be the result of an indirect CB1 activation mechanism through
some type of ‘‘cross talk” effect involving a different receptor such
as GPR1829 for which N-arachidonoyl glycine is a ligand (vide
infra).
DMSO
Palmitoyl dopamine
Oleoyl dopamine
100
18 100
9.5 11.1
5.5 7.02
15 7.39
25 6.10
26 3.95
10 5.72
6.6 2.33
23 10.5
9.5 4.32
23 6.60
32 177
26
1.4
0.61
7.1
1.0
0.93
0.94
0.87
12
0.92
0.77
35
41.6
24.0
21.6
38.3
66.6
25.5
22.1
48.0
Linoleoyl dopamine
c
-Linoleoyl dopamine
Eicosa-1Z, 14Z-dienoyl dopamine
Dihomo- -linolenoyl dopamine
c
Arachidonoyl dopamine
Eicosapentaenoyl dopamine
Docosatetra-7Z,10Z,13Z,16Z-enoyl dopamine 25.7
Docosahexaenoyl dopamine
Indomethacin
40.7
144
Cell culture conditions and treatment procedures are described in Section 6.1.1 and
in Figure 3.
Table 4
Effects of N-palmitoyl and N-arachidonoyl derivatives on rat basophilic leukemia
(RBL-2H3) cell proliferation.
Treatment (n = 3)
Concn (
—
lM)
Response
4665
SD
601
Control (%)
100
DMSO
Palmitoyl glycine
Palmitoyl glycine
Palmitoyl glycine
Arachidonoyl glycine
Arachidonoyl glycine
Arachidonoyl glycine
0.1
1.0
10
0.1
1.0
10
5906
5346
5291
6224
6203
959
611
1246
948
524
1016
392
130
110
110
130
130
21
Palmitoyl
Palmitoyl
Palmitoyl
L
L
L
-alanine
-alanine
-alanine
0.1
1.0
10
5067
4355
614
1866
1005
35
110
90
10
Arachidonoyl
Arachidonoyl
Arachidonoyl
L
L
L
-alanine
-alanine
-alanine
0.1
1.0
10
5822
7315
2312
819
1454
1042
120
160
50
Several other agents were tested for their effects on the breast
cancer model (unpublished observations). A second CB1 antago-
nist, AM251, completely reversed the anti proliferative effect of
10 lM N-palmitoyl tyrosine in agreement with the data shown in
Fig. 3F. However, the CB2 antagonist, AM630, gave a weaker rever-
Palmitoyl GABA
Palmitoyl GABA
Palmitoyl GABA
Arachidonoyl GABA
Arachidonoyl GABA
Arachidonoyl GABA
0.1
1.0
10
0.1
1.0
10
5459
4636
4461
4531
5605
674
1544
732
151
808
1048
30
120
100
100
100
120
15
sal of the N-palmitoyl tyrosine response suggesting a lesser role for
CB2 in our model. A possible role for PPAR-
AJA has been reported to bind and activate this important nuclear
receptor.30,31 GW9662, a potent, selective and irreversible PPAR-
c was investigated since
c
Palmitoyl dopamine
Palmitoyl dopamine
Palmitoyl dopamine
Arachidonoyl dopamine
Arachidonoyl dopamine
Arachidonoyl dopamine
0.1
1.0
10
0.1
1.0
10
5882
2365
704
5254
3447
648
658
696
13
601
670
4
130
50
15
110
70
14
antagonist had no effect on N-palmitoyl tyrosine inhibition in our
model suggesting the absence of a role for this nuclear receptor. Fi-
nally, the cells were treated with a competitive inhibitor of FAAH,
N-arachidonoyl serotonin, since other EMAs were shown to inhibit
this catalyst of endocannabinoid metabolism9 and an elevation of
anandamide levels could, in some cases, result in a decrease in cell
proliferation. However, no effect was found compared with vehicle
treated cells suggesting that the observed inhibitions were not due
to an increase in endocannabinoid concentration in HTB-126 cells.
Ninety-six-well plates with 500 mast cells/well/100
incubated for 24 h. Media were then changed to 100
incubated for 4 h. Cells were treated as indicated in the table and incubated for 1 h.
LPS (10 ng/ml) was then added to each well and the incubation continued for 24 h.
Cell numbers were then obtained using the CellTiter-Glo assay.
l
l
l media with serum were
l of serum free RPMI and
On the other hand, direct treatment with anandamide at 10 lM
obtained from healthy breast tissue while the latter was derived
from a breast tumor present in the same donor. We reasoned that
this would provide a good model to test our analogs for selective
action where most extraneous factors would be eliminated. The
optimal time point for measuring changes in proliferation was
found to be not less than 48 hours as shown in Figure 2. In this
experiment N-palmitoyl tyrosine, an example of an EMA, was com-
pared with a positive control substance, AJA that has anti cancer
activity.18 In passing, it is interesting to note that although effica-
cies between the two were similar, the onset of action was differ-
ent suggesting the possibility of different mechanisms.
caused a significant 38% (p < 0.006) decrease in HeLa cell prolifer-
ation pointing to possible differences in mechanism in other cell
types. Thus, the question of involvement of FAAH in the anti prolif-
erative actions of the EMAs remains open.
Another possible mechanism needs to be considered in future
studies. The orphan receptor GPR 18 has recently been shown to
be a target for N-arachidonoyl glycine, the prototypic EMA.29 It
was reported that this receptor is abundant in monocytic and lym-
phoid cell lines, and was increased after stimulation with the mito-
genic substance phytohemagglutinin. If GPR 18 is also expressed in
breast cancer cells, it might have a role in the action of the acylam-
ido analogs on cell proliferation and provide the basis for a putative
mechanism.
Five analogs were tested using this breast cancer model and the
findings are shown in Figure 3A–E. The results ranged from
virtually no selectivity with N-oleoyl dopamine (Fig. 3C) to com-
pletely selective effects with N-palmitoyl dopamine (Fig. 3A) and
In addition to the subject of selective actions for the acylamido
analogs, the scope of the inhibitory effects in terms of cell types

S. Burstein, R. Salmonsen / Bioorg. Med. Chem. 16 (2008) 9644–9651
9649
and molecular structure has been studied. Figure 4 and Tables 1–4
address these questions. A series of dopamine-fatty acid analogs
was studied in cervical cancer cells (HeLa) and the results are
shown in Figure 4 in which N-palmitoyl glycine, a saturated fatty
acid analog, and AJA are the controls as well as vehicle treated cells.
palmitoyl and arachidonoyl derivatives in a rat basophilic leukemia
line (RBL-2H3). The findings with the glycine and GABA analogs
might suggest that saturated fatty acid derivatives are inactive.
However, in the L-alanine and dopamine groups both types of fatty
acid analogs produced roughly equal degrees of inhibition. Thus,
the nature of an SAR for the acylamido analogs remains unclear
and the only conclusion at this point is that activity is dependent
on both the fatty acid and amino containing components of the
molecule.
Concentrations of 1, 3, and 10 lM were used for each of the analogs
all of which were unsaturated fatty acid derivatives. Significant
inhibition of proliferation was seen at the highest concentration
for each compound similar to the data shown for the HTB-126 cells
(Fig. 3). In each case, at 3 and 1 lM no significant inhibition and
even trends toward stimulation of cell proliferation were observed
indicating a steep dose–response curve. Since no matched normal
control cells are available for HeLa cells, it was not possible to
study the selectivity of these compounds in this model. Neverthe-
less, the data suggest that these analogs may be effective in more
than one type of cancer cell. Our findings are in agreement with
an earlier report on a series of acyl dopamides in the MCF-7 human
breast cancer cell line.21 The authors identified N-arachidonoyl
dopamine (NADA) as the most active derivative and showed that
its anti proliferative action is mediated by CB1.
Structurally, the simplest acylamido analogs are the fatty acid
amides (Fig. 1a) that we have studied in two different cell lines
(Table 1). The WI-38 human embryonic lung cell is a rapidly prolif-
erating line with fibroblast morphology while the RAW264.7 cell is
a murine macrophage tumor-derived line. Almost all of the amides
were essentially ineffective as inhibitors of cell proliferation
5. Conclusions
Future studies with the types of fatty acid amido compounds re-
ported on here should be aimed at three important questions. First,
the mechanism(s) responsible for the specificity of the anti prolif-
erative actions should be addressed. Second, a more detailed SAR
should be obtained that would allow the design of more potent
selective analogs. And, finally, the in vivo potential of the more
promising compounds should be studied with a view toward their
becoming drug candidates. The data presented here justifies carry-
ing out such studies.
6. Materials and methods
6.1.1. Materials
when tested at a single 10
lM concentration. One exception was
All of the acylamido analogs studied were obtained from
BIOMOL International, L.P., Plymouth Meeting, PA 19462 (Screen-
WellTM Endocannabinoid Library). The only exception was
N-palmitoyl tyrosine that was synthesized using a published pro-
cedure.8 All other chemicals were purchased from Sigma-Aldrich,
St. Louis, MO 63178. All cell lines were obtained from American
Type Culture Collection (ATCC), Manassas, VA 20108. Cell prolifer-
ation assay reagents were purchased from Promega Corporation,
Madison, WI 53711. Cell culture supplies were obtained from Gib-
co, Carlsbad, CA.
c
-linolenamide that reduced WI-38 cell proliferation by 67%. A
previously published report stated that oleamide suppressed the
proliferation of human breast carcinoma EFM-18 cells.32 However,
generally speaking the fatty acid amides do not appear to be a
promising source of novel anti cancer agents.
A similar series of ethanolamine derivatives (Fig. 1b) was studied
under the same conditions as the simple amine derivatives de-
scribed above. As the data show (Table 2), the inhibitory effects were
variable with no clear pattern emerging. The greatest inhibitory ac-
tion was found with N-docosahexaenoyl ethanolamide, which re-
duced RAW cell proliferation by 64%. The most widely studied
compound in this group is N-arachidonoyl ethanolamide otherwise
known as anandamide, one of the important endocannabinoids. A
wide range of actions has been ascribed to anandamide, including
anti proliferative activity,32–34 so it is surprising that it showed little
effect in our model. This finding would also tend to be inconsistent
with any FAAH inhibition mechanisms for the anti proliferative ac-
tions of the acylamido analogs. Another common ethanolamide, N-
palmitoyl ethanolamide, was not effective in our models, however,
others have reported that it is effective in reducing proliferation in
human breast cancer cells34 probably by some indirect mechanism.
Several of the acyl dopamine derivatives whose actions in HeLa
cells are shown in Figure 4 were also tested in the WI-38 and RAW
cell lines and the results are given in Table 3. In each example, they
showed appreciable activity especially in the RAW cell line where
every analog produced 90% or better inhibition. Four of these mol-
ecules were also studied for possible selectivity of action in the
matched HTB cell line and the findings are in Figure 3. These com-
bined results point to the N-acyl dopamides as promising template
structures for the discovery of novel and safe anti cancer drug can-
didates. It is worth mentioning that some of these compounds are
potent activators of the vanilloid receptor TRPV1,35 however, an
exception is N-palmitoyl dopamine that does not activate this
receptor. In our studies, N-palmitoyl dopamine is both selective
and efficacious in reducing cancer cell proliferation suggesting that
the TRPV1 receptor is not a mediator for the inhibition of cell pro-
liferation by the acylamido analogs studied here.
6.1.2. N-arachidonoyl-L-tyrosine; L-EMA-10 (16:0)
Palmitic acid (500 mg) was dissolved in acetonitrile (5 ml) that
contained triethylamine (1 ml) and cooled in ice. Isobutylchloro-
formate (0.5 ml) was then added and the reaction allowed to pro-
ceed for 30 min in the ice bath. The reaction mixture was then
partitioned between water and ethyl acetate, the organic phase
dried over sodium sulfate and evaporated under vacuum. L-tyro-
sine methyl ester (400 mg) in triethylamine (1 ml)/dimethylform-
amide (5 ml) was added to the residue and the reaction allowed to
proceed for 18 h at 0 °C. The mixture was again partitioned be-
tween water and ethyl acetate, washed with 1 N potassium car-
bonate and then washed with 1 N HCl. The crude methyl ester
(190 mg) was then dissolved in a mixture of tetrahydrofuran
(5 ml) and 1N LiOH (2 ml) and stirred at room temperature for
18 h. The mixture was acidified with 1 N HCl, dried and evaporated
to an oily residue. Trituration with hexane gave a waxy white
material (130 mg). 1H NMR (400 MHz CDCl3) 6.94 (d, J = 8.0 Hz,
2H), 6.88 (d, J = 8.0Hz1H), 6.68 (d, J = 8.0 Hz, 2H), 5.81 (m, 1H),
4.55 (m, 1H), 3.10–3.00 (m, 2H), 2.12 (m, 2H), 1.51 (m, 2H), 1.18
(m, 24H) and 0.81 (t, J = 8.0 Hz, 3H). HRMS calcd C25H42N1O4
(M+H); 420.3114, found; 420.3119 (
D mass +0.0005).
6.2. Proliferation assay
Cell numbers were estimated using a commercially available kit
(Promega). The CellTiter-GloTM Luminescent Cell Viability Assay kit
used here is based on the measurement of ATP, which signals the
Lastly, in Table 4, the question of saturated versus polyunsatu-
rated fatty acids is addressed by directly comparing four classes of

9650
S. Burstein, R. Salmonsen / Bioorg. Med. Chem. 16 (2008) 9644–9651
presence of metabolically active cells. The assay uses luciferase as
the detection enzyme because of the absence of endogenous
luciferase activity in mammalian cells. An equal volume of CellTit-
er-GloTM Reagent is added to the cell culture and, after 15 min,
luminescence is measured. The light signal is proportional to the
amount of ATP present, which correlates with the number of viable
cells present. The VeritasTM Microplate Luminometer used in this
study detects as little as 1.5 ꢀ 10ꢁ15 moles ATP and values are
linear from 760 fg to 5.1 ng of ATP. A study was done to determine
the relationship between the experimentally obtained fluorescence
response and the actual cell numbers by counting the HTB-125 and
HTB-126 cell lines and comparing the assay values (data not
shown). In each, the relationship is linear over a wide range;
however, the intensity of the luminescence is different for each cell
type.
bars represent S.E.M. when not indicated otherwise. Statistical
comparisons were made using an unpaired, two-tailed t-test with
95% confidence limits or by one-way ANOVA.
Acknowledgments
This publication was made possible by grant DA17969 from Na-
tional Institute on Drug Abuse, National Institutes of Health,
Bethesda, MD. Its contents are solely the responsibility of the
authors and do not necessarily represent the official views of the
National Institute on Drug Abuse. We thank Dr. Robert E. Zurier,
Dr. Ethan S. Burstein and Ronald Rossetti for helpful discussions
throughout the course of this study and during the preparation
of the manuscript and we are also grateful to Dr. Akbar Ali for help
with the NMR data.
6.3. Cell culture
References and notes
1. Pacher, P.; Batkai, S.; Kunos, G. Pharmacol. Rev. 2006, 58, 389.
2. Flygare, J.; Sander, B. Semin. Cancer Biol. 2008, 18, 176.
3. Tan, B.; Bradshaw, H. B.; Rimmerman, N.; Srinivasan, H.; Yu, Y. W.; Krey, J. F.;
Monn, M. F.; Chen, J. S.; Hu, S. S.; Pickens, S. R.; Walker, J. M. AAPS J. 2006, 8,
E461.
4. Burstein, S. Neuropharmacology 2007. doi:10.1016/j.neuropharm.2007.11.011.
5. Huang, S. M.; Bisogno, T.; Petros, T. J.; Chang, S. Y.; Zavitsanos, P. A.; Zipkin, R.
E.; Sivakumar, R.; Coop, A.; Maeda, D. Y.; De Petrocellis, L.; Burstein, S.; Di
Marzo, V.; Walker, J. M. J. Biol. Chem. 2001, 276, 42639.
6. Succar, R.; Mitchell, V. A.; Vaughan, C. W. Mol. Pain 2007, 3, 24.
7. Vuong, L. A.; Mitchell, V. A.; Vaughan, C. W. Neuropharmacology 2008, 54, 189.
8. Burstein, S. H.; Adams, J. K.; Bradshaw, H. B.; Fraioli, C.; Rossetti, R. G.;
Salmonsen, R. A.; Shaw, J. W.; Walker, J. M.; Zipkin, R. E.; Zurier, R. B. Bioorg.
Med. Chem. 2007, 15, 3345.
9. Cascio, M.; Minassi, A.; Ligresti, A.; Appendino, G.; Burstein, S.; Di Marzo, V.
Biochem. Biophys. Res. Commun. 2004, 314, 192.
10. Rimmerman, H.; Bradshaw, D. K.; O’Dell, D. K.; Walker, J. M. Neurosci. Abstr.
2006, 2454, N3.
11. Rimmerman, N.; Bradshaw, H. B.; Hughes, H. V.; Chen, J. S.; Hu, S. S.; McHugh,
D.; Vefring, E.; Jahnsen, J. A.; Thompson, E. L.; Masuda, K.; Cravatt, B. F.;
Burstein, S.; Vasko, M. R.; Prieto, A. L.; O’Dell, D. K.; Walker, J. M. Mol.
Pharmacol. 2008, 10, 10.
12. Sarfaraz, S.; Adhami, V. M.; Syed, D. N.; Afaq, F.; Mukhtar, H. Cancer Res. 2008,
68, 339.
The culturing conditions are generally those suggested by the
supplier, ATTC; however, a brief description for each cell type is gi-
ven below.
HTB-126TM (Hs 578T). The Hs 578T cell strain was derived from a
carcinoma of the breast. It was originated by Hackett et al.36 along
with the Hs 578Bst (see HTB-125 below), which is a normal fibro-
blast-like line from the same patient. As with Hs 578Bst, no estro-
gen receptors or endogenous viruses were detected. The donor was
a 74-year-old female Caucasian. The cells were not tumorigenic in
immunosuppressed mice but were in semisolid medium. The cells
are cultured in Dulbecco’s Modified Eagle’s Medium complete
growth medium; the following components are added to the base
medium: 0.01 mg/ml bovine insulin and fetal bovine serum to a fi-
nal concentration of 10%.
HTB-125TM (Hs 578Bst). This is a diploid human cell line that ex-
presses the epidermal growth factor (EGF) receptor. Hs 578Bst was
derived by Hackett et al. from normal breast tissue peripheral to an
infiltrating ductal carcinoma which was the source for HTB-126.36
The base medium for this cell line is ATCC Hybri-Care Medium. To
make a complete growth medium, the following components are
added to the base medium: 1.5 g/l sodium bicarbonate, 30 ng/ml
EGF and fetal bovine serum to a final concentration of 10%.
HeLa. The base medium is Gibco Dulbecco’s Modified Eagle‘s
Medium with 10% fetal bovine serum added. Cells are grown in a
T-75 flask in 15 ml of medium that is replaced on day four and
sub cultured on day seven using trypsin to remove the cells. The
sub cultivation ratio is approximately 1:10.
WI-38. The base medium is Gibco Dulbecco’s Modified Eagle‘s
Medium with 10% fetal bovine serum added. Cells are grown in a
T-75 flask in 15 ml of medium, which is replaced on day four and
sub cultured on day seven using trypsin to remove the cells. The
sub cultivation ratio is approximately 1:10.
RAW264.7. The base medium is Gibco RPMI with 10% fetal bo-
vine serum added. Cells are grown in a T-75 flask in 15 ml of med-
ium; medium is replaced on day four and sub cultured on day
seven. Cells are removed by scraping without the aid of trypsin. A
sub cultivation ratio of 1:3 to 1:6 was used.
13. Caffarel, M. M.; Moreno-Bueno, G.; Cerutti, C.; Palacios, J.; Guzman, M.;
Mechta-Grigoriou, F.; Sanchez, C. Oncogene 2008, 10, 10.
14. Blazquez, C.; Salazar, M.; Carracedo, A.; Lorente, M.; Egia, A.; Gonzalez-Feria, L.;
Haro, A.; Velasco, G.; Guzman, M. Cancer Res. 2008, 68, 1945.
15. Ruiz, L.; Miguel, A.; Diaz-Laviada, I. FEBS Lett. 1999, 458, 400.
16. Greenhough, A.; Patsos, H. A.; Williams, A. C.; Paraskeva, C. Int. J. Cancer 2007,
121, 2172.
17. Pryce, G.; Giovannoni, G.; Baker, D. Int. Cannabinoid Res. Soc. 2008, 25.
18. Recht, L. D.; Salmonsen, R.; Rosetti, R.; Jang, T.; Pipia, G.; Kubiatowski, T.;
Karim, P.; Ross, A. H.; Zurier, R.; Litofsky, N. S.; Burstein, S. Biochem. Pharmacol.
2001, 62, 755.
19. Ahmedzai, S.; Carlyle, D. L.; Calder, I. T.; Moran, F. Br. J. Cancer 1983, 48, 657.
20. Bernardi, A.; Jacques-Silva, M. C.; Delgado-Canedo, A.; Lenz, G.; Battastini, A. M.
Eur. J. Pharmacol. 2006, 532, 214.
21. Bisogno, T.; Melck, D.; Bobrov, M.; Gretskaya, N. M.; Bezuglov, V. V.; De
Petrocellis, L.; Di Marzo, V. Biochem. J. 2000, 351, 817.
22. Di Marzo, V.; Melck, D.; De Petrocellis, L.; Bisogno, T. Prostaglandins Other Lipid
Mediat. 2000, 61, 43.
23. Melck, D.; Bisogno, T.; De Petrocellis, L.; Chuang, H.; Julius, D.; Bifulco, M.; Di
Marzo, V. Biochem. Biophys. Res. Commun. 1999, 262, 275.
24. Rinaldi-Carmona, M.; Barth, F.; Heaulme, M.; Alonso, R.; Shire, D.; Congy, C.;
Soubrie, P.; Breliere, J. C.; Le Fur, G. Life Sci. 1995, 56, 1941.
25. Rinaldi-Carmona, M.; Barth, F.; Heaulme, M.; Shire, D.; Calandra, B.; Congy, C.;
Martinez, S.; Maruani, J.; Neliat, G.; Caput, D., et al FEBS Lett. 1994, 350, 240.
26. Gustafsson, K.; Wang, X.; Severa, D.; Eriksson, M.; Kimby, E.; Merup, M.;
Christensson, B.; Flygare, J.; Sander, B. Int. J. Cancer 2008, 123, 1025.
27. Ligresti, A.; Bisogno, T.; Matias, I.; De Petrocellis, L.; Cascio, M. G.; Cosenza, V.;
D’Argenio, G.; Scaglione, G.; Bifulco, M.; Sorrentini, I.; Di Marzo, V.
Gastroenterology 2003, 125, 677.
RBL-2H3. The base medium is Eagle’s Minimal Essential Medium
with Earle’s BSS and 2 mM L-glutamine (EMEM). In addition to the
base medium, the following constituents are added; 1.0 mM so-
dium private, 0.1 mM nonessential amino acids, 1.5 g/l sodium
bicarbonate and 15% heat inactivated fetal bovine serum.
28. Sheskin, T.; Hanus, L.; Slager, J.; Vogel, Z.; Mechoulam, R. J. Med. Chem. 1997, 40,
659.
29. Kohno, M.; Hasegawa, H.; Inoue, A.; Muraoka, M.; Miyazaki, T.; Oka, K.;
Yasukawa, M. Biochem. Biophys. Res. Commun. 2006, 347, 827.
30. Ambrosio, A. L.; Dias, S. M.; Polikarpov, I.; Zurier, R. B.; Burstein, S. H.; Garratt,
R. C. J. Biol. Chem. 2007, 282, 18625.
6.4. Statistical analysis
31. Liu, J.; Li, H.; Burstein, S. H.; Zurier, R. B.; Chen, J. D. Mol. Pharmacol. 2003, 63,
All of the plotting, curve fitting and statistical comparisons of
data were done with Prism 4 by GraphPad, Inc. All of the data
points shown represent the means of 4–6 replicates and the error
983.
32. Bisogno, T.; Katayama, K.; Melck, D.; Ueda, N.; De Petrocellis, L.; Yamamoto, S.;
Di Marzo, V. Eur. J. Biochem. 1998, 254, 634.

S. Burstein, R. Salmonsen / Bioorg. Med. Chem. 16 (2008) 9644–9651
9651
33. Patsos, H. A.; Hicks, D. J.; Dobson, R. R.; Greenhough, A.; Woodman, N.; Lane, J.
D.; Williams, A. C.; Paraskeva, C. Gut 2005, 54, 1741.
35. De Petrocellis, L.; Chu, C. J.; Moriello, A. S.; Kellner, J. C.; Walker, J. M.; Di Marzo,
V. Br. J. Pharmacol. 2004, 143, 251.
34. De Petrocellis, L.; Melck, D.; Palmisano, A.; Bisogno, T.; Laezza, C.; Bifulco, M.;
Di Marzo, V. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 8375.
36. Hackett, A. J.; Smith, H. S.; Springer, E. L.; Owens, R. B.; Nelson-Rees, W. A.;
Riggs, J. L.; Gardner, M. B. J. Natl. Cancer Inst. 1977, 58, 1795.