Antiproliferative and antimigratory actions of synthetic long chain n-3 monounsaturated fatty acids in breast cancer cells that overexpress cyclooxygenase-2

Article abstract of DOI:10.1021/jm300673z

Cyclooxygenase-2 (COX-2) is overexpressed in many human cancers and converts the n-6 polyunsaturated fatty acid (PUFA) arachidonic acid to prostaglandin E₂ (PGE₂), which drives tumorigenesis; in contrast, n-3 PUFA inhibit tumorigenesis. We tested the hypothesis that these antitumor actions of n-3 PUFA may involve the n-3 olefinic bond. n-3 Monounsaturated fatty acids (MUFAs) of chain length C16-C22 were synthesized and evaluated in MDA-MB-468 breast cancer cells that stably overexpressed COX-2 (MDA-COX-2 cells). Longer chain (C19-C22) n-3 MUFAs inhibited proliferation, activated apoptosis, decreased PGE₂ formation, and decreased cell invasion; C16-C18 analogues were less active. Molecular modeling showed that interactions of Arg120, Tyr355, and several hydrophobic amino acid residues in the COX-2 active site with C19-C22 MUFA analogues were favored. Thus, longer-chain n-3 MUFAs may be prototypes of novel anticancer agents that decrease the formation of PGE₂ in tumor cells that contain high levels of COX-2.

Full text of DOI:10.1021/jm300673z

Article
pubs.acs.org/jmc
Antiproliferative and Antimigratory Actions of Synthetic Long Chain
n‑3 Monounsaturated Fatty Acids in Breast Cancer Cells That
Overexpress Cyclooxygenase‑2
†
†
†
†
‡
‡
Pei H. Cui, Tristan Rawling, Kirsi Bourget, Terry Kim, Colin C. Duke, Munikumar R. Doddareddy,
‡
†
§
¶
,†
David E. Hibbs, Fanfan Zhou, Bruce N. Tattam, Nenad Petrovic, and Michael Murray*
†
‡
§
Pharmacogenomics and Drug Development Group, Pharmaceutical Chemistry, and Thomas R. Watson Mass Spectrometry
Facility, Faculty of Pharmacy, University of Sydney, NSW 2006, Australia
¶
Pharmacy and Medical Sciences, University of South Australia, GPO Box 2471, Adelaide SA 5001, Australia
S Supporting Information
*
ABSTRACT: Cyclooxygenase-2 (COX-2) is overexpressed in many human
cancers and converts the n-6 polyunsaturated fatty acid (PUFA) arachidonic
acid to prostaglandin E (PGE ), which drives tumorigenesis; in contrast, n-
2
2
3
PUFA inhibit tumorigenesis. We tested the hypothesis that these
antitumor actions of n-3 PUFA may involve the n-3 olefinic bond. n-3
Monounsaturated fatty acids (MUFAs) of chain length C16−C22 were
synthesized and evaluated in MDA-MB-468 breast cancer cells that stably
overexpressed COX-2 (MDA-COX-2 cells). Longer chain (C19−C22) n-3
MUFAs inhibited proliferation, activated apoptosis, decreased PGE2
formation, and decreased cell invasion; C16−C18 analogues were less
active. Molecular modeling showed that interactions of Arg120, Tyr355, and
several hydrophobic amino acid residues in the COX-2 active site with
C19−C22 MUFA analogues were favored. Thus, longer-chain n-3 MUFAs
may be prototypes of novel anticancer agents that decrease the formation of PGE in tumor cells that contain high levels of COX-
2
2.
INTRODUCTION
breast cancer cells is decreased by EPA and docosahexaenoic
■
4
,5
acid. Dietary PUFA are not only important risk factors for
tumor progression but may also offer opportunities for the
development of new cancer therapeutics.
The major cause of death from breast, prostate, and other
cancers is metastatic disease in which cells migrate from the
original tumor site to distant organs and establish secondary
1
n-3 and n-6 PUFAs are substrates for cyclooxygenase
tumors. Metastasis also frequently renders the tumor
(
COX), lipoxygenase, and cytochrome P450 enzymes and
unresponsive to subsequent drug treatment and, because of
their disseminated nature, surgical removal of metastases is not
undergo biotransformation to parallel series of eicosanoid
metabolites, including prostaglandins (PGs), leukotrienes, and
2
feasible. The metastatic process is complex, involving the
epoxides. n-6-Derived PGs, especially PGE , are strongly
2
interplay of several discrete processes, including increased cell
proliferation, invasion and angiogenesis.
n-3 and n-6 Polyunsaturated fatty acids (PUFAs) are two
major classes of essential dietary fatty acids, exemplified by
eicosapentaenoic acid (EPA; n-3, C20:5) and arachidonic acid
6
,7
implicated in tumor growth and metastasis. There are two
major COX genes in man: COX-1 is constitutively expressed in
a wide range of tissues, while the constitutive expression of
COX-2 is negligible but is strongly induced by proinflammatory
8,9
cytokines and extracellular stress stimuli. COX-2 is overex-
pressed to a much greater extent than COX-1 in many human
breast and epithelial cell cancers and is associated with
(
AA; n-6, C20:4), respectively. EPA and AA are structurally
analogous, but EPA has an additional olefinic double bond at
the n-3 position located between carbons 17 and 18. Numerous
epidemiological studies in man and experimental studies in cells
and animals have suggested that n-3 PUFAs like EPA offer
potential benefits in cancer treatment by decreasing tumor
growth and metastasis, whereas n-6 PUFAs are tumor-
promoting. Thus, in a large cohort of French patients, increased
content of n-3 docosahexaenoic or α-linolenic acids in breast
adipose tissue is associated with improved survival compared
increased rates of PGE production, which augments tumor
2
7
,8,10
progression and metastasis, leading to a poor prognosis.
COX-2 inhibitors such as celecoxib have been shown to
decrease mammary tumor burden in animal models by
increasing tumor cell apoptosis, decreasing proliferation, and
11−13
suppressing angiogenesis.
However, while nonsteroidal
3
with women in the lowest intake group. In xenograft mouse
models the spread of lung metastases produced by human
Received: May 15, 2012
Published: July 23, 2012
©
2012 American Chemical Society
7163
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of n-3 MUFAs
a
Reagents and conditions: (i) Acetyl chloride, EtOH, rt, 4 h; (ii) PEPPSI-iPr, LiBr, THF/N-methyl-2-pyrrolidone (2:1), rt, 18 h; (iii) Raney Ni,
NaH PO , pyridine/H O/AcOH (2:1:1), 40 °C, 2 h; (iv) [Ph PPr]Br, NaN(TMS) , −78 °C to rt, 2 h; (v) 1.5 M NaOH, ethanol, rt, 3 h.
2
2
2
3
2
16
anti-inflammatory drugs effectively inhibit tumorigenesis, the
IPr; Scheme 1). Prior to cross-coupling, carboxylic acid
groups were protected by esterification with acetyl chloride and
ethanol (Scheme 1). Subsequent alkyl−alkyl Negishi coupling
of ethyl esters 4a−e with the alkyl zinc reagent 3 proceeded in
use of these agents in longer-term preventive treatment for
cancer is impractical because of toxicity. The identification of
14
well-tolerated alternate agents that decrease PGE -dependent
2
16
tumor cell migration by selective inhibition of COX-2 may
assist the development of novel antimetastatic therapies. From
studies in animals, the antimetastatic actions of n-3 PUFA are
mediated in part by inhibiting COX-2 activity and PGE2
good yield (Scheme 1). Selective reduction of the nitrile
groups of 5a−e, without concurrent reduction of the ester
group, was achieved with Raney nickel/sodium hypophos-
17,18
phite,
affording the aldehyde intermediates 6a−e in good
5
formation. Thus, increasing the dietary intake of n-3 PUFAs
yields (Scheme 1). The Wittig reaction, using THF as solvent
and a coupling temperature of −78 °C, generated the n-3
bonds in 7a−e with excellent (Z)-selectivities (Z:E ≥ 97:3).
Double bond configurations and isomeric purities of 7a−e were
established using 13C NMR and GC-MS due to overlapping
could be a viable strategy to impair PGE production by tumors
2
and improve health outcomes. However, because long-term
patient compliance with altered dietary regimen is rarely
successful, there is a need to develop effective new chemical
entities. In the present study we synthesized a series of novel n-
1
15
olefinic proton resonances in the H NMR spectra of 7a−e.
3
monounsaturated fatty acid (MUFA) derivatives based on
Briefly, (Z)-geometries were confirmed from the chemical shifts
of the resonances of carbon atoms that were adjacent to the
olefinic bond. In n-3 double bonds these occur at ∼17 and 20
ppm and are diagnostic for the n-3 (Z)-configuration. Isomeric
purity was accurately determined by GC-MS on a Zebron ZB-
Wax column. Base-catalyzed hydrolysis of 7a−e at room
temperature then afforded the required n-3 MUFAs without
the need for chromatography.
naturally occurring n-3 PUFAs and evaluated their potential as
inhibitors of breast cancer cell growth and migration. The
principal finding to emerge was that the longer chain n-3
MUFA analogues (C20−C22) effectively decreased the growth
and invasion capacity of breast cancer cells that overexpress
COX-2 and markedly decreased PGE production by these
2
cells.
Antiproliferative and Apoptotic Actions of n-3 MUFA
in Breast Cancer Cells. To evaluate the actions of the n-3
MUFAs against tumor cells in which COX-2 expression was
high, we engineered the moderately invasive human breast
cancer cell line MDA-MB-468 (MDA-CON) to overexpress
COX-2 (MDA-COX-2 cells). As shown in Figure 1, basal
expression of COX-2 immunoreactive protein in the parental
cell line was low but was strongly expressed in the MDA-COX-
2 stable transfectants. The effects of 1a−g on the proliferation
and viability of MDA-COX-2 cells were assessed by
mitochondrial reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) and trypan blue
exclusion. Basal mitochondrial MTT reduction measured after
48 h in culture was increased in MDA-COX-2 cells to 1.5 ±
0.1-fold of MDA-CON (p < 0.01), consistent with increased
viability of the transfectants. Compounds 1a−g elicited
RESULTS
■
Synthesis of n-3 MUFA Analogues. We have previously
described the facile synthesis of the C18 and C19 n-3 MUFA
1
5
1
c and 1d. The key C15 and C16 oxo-fatty acids (analogous
to 6, Scheme 1), that were used in Wittig reactions to construct
the n-3 olefinic bonds in 1c and 1d, were obtained by
esterification and subsequent oxidation of 15-hydroxypentade-
canoic acid and 16-hydroxyhexadecanoic acid, respectively.
However, because precursor ω-hydroxy-fatty acids required for
synthesis of the analogues 1a,b,e−g by that route are not
commercially available, we developed a general strategy for the
preparation of long chain oxo-fatty acids that utilized alkyl−
alkyl Negishi cross-coupling reactions catalyzed by the N-
heterocyclic carbene [1,3-bis(2,6-diisopropylphenyl)imidazol-2-
ylidene](3-chloropyridyl)palladium(II) dichloride (PEPPSI-
7
164
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
The relationship between the growth inhibitory actions of
the n-3 MUFAs and the activation of cellular apoptosis was
assessed. Consistent with the observed decreases in cell
proliferation the C21 analogue 1f potently activated the key
proapoptotic enzyme caspase-3 (EC50 12 ± 3 μM; Table 1),
while the C19, C20, and C22 analogues were somewhat less
effective (EC range 26 to 43 μM); the shorter chain analogues
50
(
C16−C18), as well as the naturally occurring PUFAs EPA and
AA, were less active.
Inhibition of COX-2-Dependent Cell Invasion and
PGE Formation by n-3 MUFAs. COX-2 overexpression is
2
associated with an increase in tumor cell invasion. Using the
19
Matrigel droplet assay developed in this laboratory, basal
invasion activity of MDA-COX-2 cells was markedly increased
to 300 ± 15% of MDA-CON cells that expressed only low
levels of COX-2 (57 ± 3 versus 19 ± 2 migrated cells/20 h, p <
0
.001; Figure 2A). The migration of MDA-COX-2 cells was
Figure 1. Construction of MDA-MB-468 cells that overexpressed
COX-2 (MDA-COX-2 cells). (A) Plasmids encoding blank (GFP) and
COX-2 used in transfections. (B) Colony formation in the presence of
G418 antibiotic. (C) Immunoblot analysis of COX-2 protein
expression in control (CON) and COX-2 cells.
concentration-related decreases in cell proliferation that were
apparent after 24 h of treatment and IC50 values were
determined after 48 h of treatment (Table 1). The C21 n-3
Table 1. Antiproliferative and Proapoptotic Activities of n-3
MUFAs in MDA-COX-2 Breast Cancer Cells
b
carbon chain
length
antiproliferative
a
proapoptotic activity
Figure 2. Effects of fatty acids on the number of cells that migrated in
the Matrigel droplet assay over 20 h. (A) number of migrated MDA-
COX-2 cells relative to MDA-MB-468 (CON) cells and effect of
treatment with AA or EPA (20 μM) on migration. (B) Effect of
treatment with n-3 MUFA (10 or 20 μM) on migration of MDA-
COX-2 cells. Different from DMSO-treated MDA-COX-2 control: *p
analogue
activity (μM)
(μM)
1
1
1
1
1
1
1
a
b
c
d
e
f
16
17
18
19
20
21
22
20
>100
>100
>100
69 ± 19
54 ± 12
43 ± 9
43 ± 0.2
12 ± 3
26 ± 5
>100
>100
36 ± 4
31 ± 8
9.3 ± 2.9
21 ± 1
>100
†
†
< 0.05, **p < 0.01, ***p < 0.001; different from MDA-CON cells:
p
†
††
<
0.01, p < 0.001.
g
AA
further increased when the media was supplemented with
exogenous AA (20 μM; p < 0.05) but was decreased by EPA
EPA
20
b
>100
>100
a
MTT activity. Caspase-3 activity.
(
20 μM; p < 0.05). We tested the capacity of the novel n-3
MUFA derivatives to inhibit the migration of MDA-COX-2
cells. The C22 n-3 MUFA 1g (10 μM) effectively decreased cell
invasion to 48 ± 4% of the uninhibited rate, while compounds
1e−g, when tested at a concentration of 20 μM, inhibited
invasion to 75 ± 8%, 59 ± 3%, and 48 ± 1% of control (P <
0.001); 1a−d were inactive (Figure 2B).
MUFA analogue (1f) exhibited the greatest antiproliferative
potency (IC50 for inhibition of MTT reduction was 9.3 ± 2.9
μM), whereas the C22 (1g), C20 (1e), and C19 (1d)
analogues were slightly less potent (IC s 21, 31, and 36 μM,
5
0
respectively). In contrast, the short-chain analogues 1a−c
(
C16−C18) were essentially inactive (IC >100 μM), while
PGE is the most abundant eicosanoid that is generated by a
5
0
2
the naturally occurring n-3 and n-6 PUFAs EPA and AA also
range of tumors and has been shown to contribute significantly
6
,7,20,21
exhibited IC s that exceeded 100 μM.
to cancer progression.
Basal secretion of PGE into the
50
2
To test whether the cytotoxic actions of the n-3 MUFAs
were mediated by membrane disruption, MDA-COX-2 cells
were treated with 1a−g (20 and 50 μM) for 48 h, and viable
cells were identified by trypan blue exclusion. However, only
small decreases in viable cell number (by 25 ± 2% and 20 ± 2%
from control) were apparent with the analogues 1a (C16) and
media of cultured MDA-CON cells was undetectable by LC-
MS-MS but was markedly increased in the medium from
cultured MDA-COX-2 cells (68 ± 5 pmol/10 cells/48 h). In
4
the present study, the effects of n-3 MUFAs (50 μM) on basal
PGE2 formation by MDA-COX-2 cells were investigated
(Figure 3). Treatment with the longer chain n-3 MUFA 1e−
1
d (C19), and only at the higher concentration; none of the
other analogues induced membrane disruption.
g (50 μM; 48 h) decreased basal PGE formation to 46 ± 4%,
2
49 ± 8%, and 45 ± 4% of control, respectively, while
7
165
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
Figure 3. Effect of n-3 MUFAs (50 μM) on the secretion of PGE into the media of MDA-COX-2 cells over 48 h in culture. (A, left) Basal PGE
2
2
secretion, and (B, right) AA-augmented PGE secretion. Different from control: *p < 0.05, ***p < 0.001.
2
Figure 4. Inhibition of COX-2 activity in lysates of MDA-COX-2 cells by n-3 MUFA (50 μM). (A) Inhibition of PGE formation, and (B) inhibition
2
of peroxidase activity. Different from control: *p < 0.05, **p < 0.01, ***p < 0.001.
compounds 1a−d were inactive. Co-incubation of MDA-COX-
cells with AA (20 μM) enhanced PGE formation to 16-fold
of the basal rate (1.08 ± 0.12 nmol/10 cells/48 h; Figure 3B).
The longer chain n-3 MUFAs (1f, 1g) effectively decreased AA-
MUFAs to inhibit COX-2-dependent PGE formation, which
results in impaired cell viability, increased apoptosis and
decreased invasion potential.
2
2
2
4
Modeling of n-3 MUFA in the Active Center of COX-2.
To understand the interactions of the n-3 MUFAs with COX-2
in greater detail, compounds 1a−g were docked into the active
site of the enzyme (PDB code: 3HS5) (see Experimental
Section for details of the computational modeling). As shown
in Figure 5, the carboxylate groups of the n-3 MUFA analogues
1a and 1g are H-bonded to the Tyr355 (1.79 Å) and in a
bifurcated manner to Arg120 (1.93, 1.92 Å). These are
important residues in COX-2 that are adjacent to the entrance
of the substrate access channel to the enzyme binding center
and are involved in binding of PUFA substrates and COX
augmented PGE secretion to 40−52% of control (Figure 3B).
2
Because the longer chain n-3 MUFA inhibited the formation
of PGE by MDA-COX-2 cells we directly tested their capacity
2
to inhibit COX-2 activity in cell lysates. These analogues
decreased both basal and AA-supplemented PGE formation in
2
MDA-COX-2 breast cancer cells. In the presence of AA (20
μM), basal PGE formation in lysates was 11.6 ± 3.1 fmol/mg
2
protein/min that was readily suppressed by 1e−g (50 μM) to
4
2−62% of AA-stimulated formation (P < 0.01); compound 1d
was less effective (Figure 4A). This finding was substantiated by
testing the longer chain n-3 MUFA analogues against COX-2-
mediated peroxidation of N,N,N′,N′-tetramethyl-p-phenylene-
2
3,24
inhibitors.
Additional stabilization of the binding of the
longer chain analogue 1g is provided by the extension of the ω-
end of the molecule into a hydrophobic groove enclosed by the
residues Phe205, Phe209, Gly227, Val228, Ile377, Gly533, and
Leu534 and that is located above the catalytic Ser530 residue
(Figure 5B). However, in the case of the shorter chain n-3
MUFAs, exemplified by the C16 analogue 1a, extension into
the hydrophobic groove was minimal, and bonding interactions
with Tyr355 (2.58 Å) and Arg120 (1.96, 2.18 Å) are slightly
weaker than for 1g (Figure 5A). This decreased the number of
2
2
diamine. As anticipated, the activity was increased in MDA-
COX-2 cell lysates to 2.3-fold of that in lysates from MDA-
CON cells (24 ± 5 U/mL compared to 10 ± 2 U/mL).
Compounds 1f and 1g (50 μM) decreased peroxidase activity
to 54 ± 4% and 58 ± 2% of control, while 1d and 1e were
somewhat less effective (71 ± 6% and 62 ± 6% decreases in
activity; P < 0.05; Figure 4B). Together these findings are
consistent with the capacity of the synthetic longer chain n-3
7
166
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
Figure 5. Docking of the n-3 MUFA analogues (A, upper) 1a and (B, lower) 1g into the COX-2 active site.
47
a
Table 2. Summary of XP Docking Results and Prime MMGBSA ΔG Values
b
c
compound
GScore
Lipophilic EvdW
PhobEn
HBond
Electro
LowMW
RotPenal
LSE
ΔGbind‑gs
ΔGbind‑gbsa
1
1
1
1
1
1
1
a
b
c
d
e
f
−10.11
−10.44
−11.11
−11.32
−13.12
−12.97
−13.48
−5.61
−5.88
−6.50
−6.29
−6.82
−6.90
−7.14
−2.51
−2.70
−2.70
−2.70
−2.70
−2.70
−2.70
−1.79
−1.65
−1.65
−1.98
−1.84
−1.74
−1.96
−0.51
−0.48
−0.50
−0.55
−0.62
−0.51
−0.58
−0.50
−0.50
−0.50
−0.50
−0.47
−0.42
−0.37
2.81
2.78
2.74
2.70
1.33
1.30
1.28
8.58
7.10
5.30
6.59
8.07
7.85
10.70
−7.53
−7.85
−8.44
−10.35
−10.55
−11.12
−11.33
−31.02
−40.61
−41.40
−46.88
−45.76
−45.00
−53.56
g
a
Description of XP terms: GScore: Total Glide Score; LipophilicEvdW: ChemScore lipophilic pair term and fraction of the total protein−ligand vdW
energy; PhobEn: hydrophobic enclosure reward; HBond: ChemScore H-bond pair term; Electro: electrostatic rewards; LowMW: reward for ligands
b
with low molecular weight; RotPenal: rotatable bond penalty; LSE: ligand strain energy in kcal/mol. ΔG
in kcal/mol calculated using Glide
bind‑gs
c
‘
score in place’ method. More details on scoring function employed can be found in the Supporting Information. ΔG
in kcal/mol calculated
bind‑gbsa
using Prime MMGBSA method. More details on scoring function employed can be found in the Supporting Information.
contacts made with the protein and the quality of the
interaction (as reflected by the parameters derived using the
Glide program; Table 2). Because the COX-2/MUFA
interactions are principally hydrophobic in nature it would be
anticipated that the van der Waals (vdW) contribution would
strongly influence the overall binding energy. Thus, compared
to shorter chain analogues, the longer carbon chains exhibited
superior lipophilic character that improved vdW interactions
(lipophilic EvdW) and negative effects on bond rotation
(RotPenal) that were major components of Gscore calculations.
7
167
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
Indeed, the Gscore values for the longer chain analogues (1e−g)
were higher (range −12.97 to −13.48) than those of shorter
chain length (1a−1d) (range −10.11 to −11.32).
term use as antimetastatic agents, and there is a need for
alternative well-tolerated molecules that selectively inhibit
COX-2-dependent formation of PGE and minimize tumor
2
The calculation of absolute values for ΔG is a challenging
spread.
bind
task in the absence of experimental values. Two methods are
presented here: Glide’s “score in place” method and Prime’s
molecular mechanics and surface-generalized Born solvation
n-3 PUFA, such as EPA, offer apparent benefits in the
management of invasive cancers. Their antitumor actions have
been attributed to decreased formation of COX-2-derived
34
model (MM-GBSA). The ΔG values obtained by using the
PGE2. However, the potencies of naturally occurring n-3
PUFA are limited by their intracellular availability, which would
decline rapidly in cells because they undergo biotransformation
by COX-2 and other enzymes. In addition, PUFAs have the
potential for lipid peroxidation, which increases with the
number of olefinic bonds but is minimized in saturated fatty
bind
(
MM-GBSA) model gave values of ΔG
that were
bind‑gbsa
unrealistically large, compared to those obtained from Glide
ΔGbind‑gs), although they follow the same overall pattern, also
favoring ligands with longer chain length. Thus, a ΔG
(
bind‑gbsa
value of −53.56 kcal/mol was calculated for 1g, whereas the
3
5,36
ΔG for the C16 analogue 1a was only −31.02 kcal/mol
acids.
To retain the antitumor actions of the n-3 olefinic
bind‑gbsa
(
Table 2). ΔG
gave in comparison, more typical values in
bond without promoting lipid peroxidation, we prepared and
evaluated a series of novel n-3 MUFA as potential anticancer
agents. The longer chain n-3 MUFA analogues effectively
inhibited the proliferation and invasion potential of breast
cancer cells that were engineered to overexpress COX-2; these
agents also activated apoptotic cell death. The potencies of the
more active longer-chain n-3 MUFA analogues in the inhibition
of proliferation and activation of apoptosis resembled those
reported for clinically effective COX-2 inhibitors in cell-based
assays. Thus, ibuprofen and naproxen directly inhibited wild-
bind‑gs
the −7 to −11 kcal/mol range; these values again followed a
similar trend to those obtained from the more rigorous MM-
GBSA method (Table 2). Gscores and ΔG values were well
correlated (r 0.81 and 0.93, for ΔG
respectively; Table 2).
and ΔG
,
bind‑gs
bind‑gbsa
The protein and MUFA strain energies are 20.56 and 8.58
kcal/mol, respectively, in the case of the interaction of 1g with
COX-2, whereas for 1a there is a 10-fold increase in the protein
strain energy (213.61 kcal/mol) and a more modest increase in
MUFA strain (10.70 kcal/mol). Taken together, the parameters
calculated from modeling studies are consistent with the
experimentally observed inhibitory effects of the longer chain n-
24
type COX-2 with IC s of 10 and 18 μM, respectively, and a
50
number of nonsteroidal anti-inflammatory agents inhibited the
formation of PGE in lipopolysaccharide-activated monocytes
2
37
3
MUFAs on COX-2 activity and the diminished potencies of
with IC s between 0.1 and 4 μM.
50
the short chain analogues.
Longer chain n-3 MUFAs effectively decreased COX-2-
mediated PGE formation and peroxidase activity in cells.
When the n-3 MUFAs 1a−g were docked into the active site of
COX-2, they adopted an L-shaped configuration similar to
2
DISCUSSION
■
COX-2 is overexpressed in ∼40% of human invasive breast
cancers, and COX-2-derived PGE is implicated in breast
those found with the naturally occurring n-6 PUFA AA and n-3
2
1
0,12,25
38−40
cancer metastasis.
In clinical studies, PGE formation is
PUFA EPA.
The carboxylate groups in n-6 and n-3 PUFAs
2
associated with a high potential for metastasis by driving tumor
proliferation and inhibiting apoptosis and results in adverse
form an ion pair with Arg120 and a hydrogen bond with
Tyr355 that are located at the entrance of the channel that
controls the access of substrate to the active region of the
enzyme. The ω-ends of the PUFA extend above Ser530 into an
upper channel adjacent to Gly533 that is stabilized by Phe205,
2
6
prognoses. In mammary tissue of female COX-2-transgenic
mice, several changes, including hyperplasia of terminal ductal
lobules, epithelial neoplasia, and increased tumor cell invasive
behavior, were noted that are consistent with accelerated
38
Phe209, Val228, and Leu534. Binding is apparently stabilized
by vdW interactions with methylene units in the aliphatic
7
tumorigenesis due to overproduction of PGE2. In vitro
39
findings in cells are consistent with these observations. Thus,
chain. Thus, 54 and 56 contacts have been identified in
modeling studies for interaction of AA and EPA, respectively,
with residues that line the COX-2 active site. Similar to these
findings, the aliphatic ω-ends of the n-3 MUFAs also extend
into the hydrophobic groove that is adjacent to Ser530. The
longer chain analogues (C20−C22) projected into the groove,
whereas these contacts were minimal in the shorter chain
analogues. In accord with these findings, Dong et al. found that
MUFA analogues of short to intermediate chain length (C16−
COX-2 overexpression promoted the growth and invasion
27,28
potential of breast cancer cells
Overexpression of COX-2
enhanced the secretion of PGE by tumor cells and activated
2
angiogenesis in adjacent endothelial cells by augmenting the
7,21
angiopoietin/vascular endothelial growth factor axis.
Con-
sidered together, inhibition of COX-2-derived PGE formation
2
may be a strategy that could be exploited to minimize tumor
growth, metastasis, and angiogenesis.
9
11
Nonsteroidal anti-inflammatory drugs and COX-2-selective
inhibitors have been found to inhibit tumor metastasis in vitro
C20), that carried Δ or Δ olefinic bonds, did not inhibit
COX-2 activity, which suggests that they are unable to
coordinate effectively into the substrate binding site of COX-
7
,29−32
and in vivo.
Thus, celecoxib treatment inhibited COX-2
41
activity and PGE secretion and arrested the growth of breast
2.
2
2
0
cancer cell lines. In highly invasive MDA-MB-231 breast
cancer cells, the mechanism of celecoxib-induced growth arrest
involved induction of apoptosis, as reflected by caspase
activation. In mouse orthotopic and xenograft models, the
daily administration of celecoxib significantly decreased the
growth and migration of tumor cells and activated apoptosis;
Previous modeling studies with a range of clinically effective
and experimental COX-2 inhibitors identified interactions with
the active region of the enzyme that were similar to those of the
present n-3 MUFA. Thus, from crystallography and molecular
modeling, carboxylate substituents in these agents also
interacted with Arg120 and Tyr355 and projected inward
into hydrophobic clefts in the enzyme, which enabled
interactions with hydrophobic amino acid residues that include
1
2,13,33
tumor vascularization was also decreased.
Unfortunately,
however, the toxicity of most nonsteroidal anti-inflammatory
agents and COX-2-selective inhibitors precludes their long-
4
2,43
Val349, Val523, Ala527, Ser530, and Leu534.
Together, the
7
168
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
findings from modeling studies account for the experimentally
observed inhibitory effects of the longer chain n-3 MUFAs on
COX-2 activity as well as the lower potencies of the shorter
chain analogues. The present molecules emerge as prototypes
of a novel class of anticancer agents that modulate cell
metastasis; further structural modification may produce
analogues with enhanced potency and efficacy in vivo.
202.98, 173.94, 60.16, 43.92, 34.39, 29.48, 29.40, 29.38, 29.34, 29.24,
+
2
9.16, 29.13, 24.98, 22.08, 14.26. CI-MS: m/z (%): 257 ([M + H] ).
General Method for Preparation of n-3 MUFA Ethyl Esters
7
a−e: Ethyl cis-13-Hexadecenoate (7a). To a suspension of n-
propyltriphenylphosphonium bromide (0.885 g, 2.03 mmol) in
anhydrous THF (4 mL) at 0 °C under a nitrogen atmosphere was
added 1 M NaN(TMS) (2.09 mL, 2.09 mmol). The resulting orange
2
mixture was stirred for 40 min at room temperature and cooled to −78
°
C, and 6a (0.270 g, 1.04 mmol) in THF (3 mL) was added dropwise
EXPERIMENTAL SECTION
by syringe. Stirring at −78 °C was continued for 30 min after which
■
the reaction mixture was warmed to room temperature. After further
General Chemistry. All chemicals, including 4a,b, and anhydrous
solvents were purchased from Sigma Aldrich (Castle Hill, NSW,
Australia). 9-Bromononanoic acid was prepared by the method of
stirring for 1.5 h, the reaction was quenched with saturated NH Cl (20
4
mL) and extracted with dichloromethane (3 × 40 mL). The combined
44
extracts were dried over Na SO and concentrated in vacuo. The
2
4
Duffy et al. Dry column vacuum chromatography (DCVC) was used
to purify reaction products on silica gel with gradient elution by
hexane and dichloromethane. TLC was performed on silica gel 60 F254
plates, and products were visualized with ceric ammonium molybdate.
residue was purified by silica gel DCVC, affording 7a (0.244 g, 83%) as
1
a colorless oil. H NMR (400 MHz, CDCl ): 5.40−5.28 (m, 2H), 4.11
3
(
q, J = 7.2 Hz, 2H), 2.28 (t, J = 7.2 Hz, 2H), 2.07−1.97 (m, 4H),
HH
HH
1
13
1.61 (quin, JHH = 7.2 Hz, 2H), 1.37−1.20 (m, 19H), 0.95 (t, JHH = 7.6
The 400 MHz H NMR and C NMR spectra were recorded using a
Hz, 3H). 13C NMR (100 MHz, CDCl ): δ 173.91, 131.49, 129.33,
3
Varian 400-MR instrument and were referenced internally to CDCl
3
1
13
60.12, 34.39, 29.76, 29.59, 29.56, 29.52, 29.44, 29.27, 29.25, 29.14,
(
H δ 7.26, C δ 77.10). Melting points were measured on a Stuart
+
2
7.08, 24.98, 20.49, 14.38, 14.24. CI-MS: m/z (%): 283 ([M + H] ). t
R
SMP10 melting point apparatus. GC-MS was performed on a PolarisQ
GC-MS-MS ion trap mass spectrometer coupled to a Trace GC, as
=
24.14 min. Z:E ratio = 98.0: 2.0.
General Method for Preparation of n-3 MUFAs 1a,b,e−g: cis-
15
described elsewhere. Elemental analysis (C, H, N) was carried out by
the Microanalytical Unit at the Research School of Chemistry,
Australian National University (ACT, Australia), and all values were
within ±0.4% of the calculated values.
13-Hexadecenoic Acid (1a). To a solution of 7a (0.099 g, 0.35
mmol) in ethanol (15 mL) was added 1.5 M NaOH (7.5 mL). The
suspension was stirred at 45 °C until a solution was obtained. After the
solution was stirred at room temperature for 3 h, the ethanol was
removed in vacuo, and water (10 mL) was added. The mixture was
adjusted to pH 2 with 1 M HCl, and the product was extracted with
General Method for Preparation of Bromo Ethyl Esters 4c−
e: Ethyl 10-Bromodecanoate (4c). A solution of 10-bromodecanoic
acid (2.500 g, 9.95 mmol) and acetyl chloride (3.907 g, 49.77 mmol)
in ethanol (150 mL) was stirred at room temperature for 4 h. The
solvent was removed in vacuo, and the residue was dissolved in diethyl
ether (100 mL) and washed with saturated NaHCO (2 × 70 mL),
water (70 mL), and brine (70 mL). The organic phase was dried with
dichloromethane (3 × 20 mL), dried over Na SO , and concentrated
2
4
1
in vacuo to afford 1a as a white solid, mp 27−29 °C. H NMR (400
MHz, CDCl ): 5.40−5.23 (m, 2H), 2.35 (t, J = 7.2 Hz, 2H), 2.07−
3
HH
3
1
0
1
.98 (m, 4H), 1.63 (quin, J = 7.6 Hz, 2H), 1.40−1.20 (m, 16H),
HH
.95 (t, JHH = 7.2 Hz, 3H). 13C NMR (100 MHz, CDCl ): δ 179.06,
Na SO , and concentrated in vacuo. The crude product was purified
3
2
4
31.52, 129.36, 33.88, 29.78, 29.60, 29.57, 29.53, 29.43, 29.29, 29.24,
by silica gel DCVC, affording 4c (2.456 g, 88%) as a colorless oil. The
1
29.06, 27.10, 24.69, 20.51, 14.41. CI-MS (FAME): m/z (%): 269 ([M
H NMR and MS spectra of the product are in agreement with partial
+
45
1
+ H] ). R (FAME) = 22.27 min. Z:E ratio =98.0: 2.0. Anal. Calcd for
t
data previously reported. H NMR (400 MHz, CDCl ): δ 4.11 (q,
3
C H O : C 75.54, H 11.89. Found: C 75.84, H 11.69.
16
30
2
JHH = 7.2 Hz, 2H), 3.39 (t, JHH = 6.8 Hz, 2H), 2.27 (t, JHH = 7.6 Hz,
Cell Culture. Human MDA-MB-468 breast cancer cells were
2
2
H), 1.84 (p, JHH = 7.2 Hz, 2H), 1.61 (p, JHH = 7.6 Hz, 2H), 1.41 (m,
H), 1.28 (m, 8H), 1.24 (t, JHH = 7.2 Hz, 3H). 13C NMR (100 MHz,
obtained from ATCC (Manassas, VA). COX-2 and green fluorescent
protein expression plasmids (OriGene, Rockville, MD, and pEGFP-
N1, Invitrogen, Mulgrave, VIC, Australia, respectively) under the
control of the CMV promoter were transfected into MDA-MB-468
cells using Lipofectamine 2000. Media containing increasing
concentrations of G418 (100−400 μg/mL) was used to select for
stable transfectants, with resistance conferred by the plasmid-encoding
green fluorescent protein. Positive clones were selected by G418
resistance, fluorescence microscopy, and immunoblotting for COX-2.
Cells were cultured in Dulbecco’s Modified Eagle Medium
supplemented with 10% fetal bovine serum (Thermo Fischer
Scientific, Waltham, MA) and 1% penicillin/streptomycin (Invitro-
CDCl ): δ 173.86, 60.14, 34.35, 34.01, 32.78, 29.21, 29.12, 29.06,
3
+
2
8.67, 28.11, 24.92, 14.25. CI-MS: m/z (%): 281 ([M + H] ).
General Method for Preparation of Cyano Ethyl Esters 5a−e:
Ethyl 12-Cyanododecanote (5a). Anhydrous LiBr (2.084 g, 24.00
mmol) and PEPPSI-IPr (0.096 g, 0.14 mmol) were dissolved in
anhydrous THF (5 mL) and N-methyl-2-pyrrolidone (10 mL) under
nitrogen. 4a (1.575 g, 7.06 mmol) and then 6-cyanohexylzinc bromide
solution (24 mL, 12 mmol) were added, and the solution was stirred at
room temperature for 18 h. Diethyl ether (70 mL) was added, and the
organic phase was washed with 1 M Na EDTA solution (50 mL),
water (50 mL), and brine (50 mL). After drying with Na SO , the
solvent was removed in vacuo, and the residue was purified on silica
gel DCVC, affording 5a (1.037 g, 58%) as a colorless oil. The H and
3
2
4
gen), and grown at 37 °C in a humidified atmosphere of 5% CO in
2
1
air. Confluent cells (80−90%) were harvested using trypsin/EDTA
after washing in phosphate-buffered saline (Amresco, Solon, OH).
Test compounds were dissolved in DMSO (Sigma-Aldrich)/2-
propanol (Merck; Kilsyth, VIC, Australia) (final concentration 0.1%).
Cell Viability and Apoptosis. For MTT and caspase-3 activity
1
3
C NMR spectra of the product are in good agreement with
46
+
previously reported data. CI-MS: m/z (%): 254 ([M + H] ).
General Method for Preparation of Oxo Ethyl Esters 6a−e:
Ethyl 13-Oxotridecanoate (6a). To a solution of 5a (0.600 g, 2.37
mmol) in pyridine (20 mL) were added water (10 mL), acetic acid (10
mL), sodium hypophosphite (2.007 g, 18.93 mmol), and Raney nickel.
The suspension was then stirred at 40 °C for 2 h. The catalyst was
removed by filtration and washed with ethanol (2 mL). Water (150
mL) and diethyl ether (40 mL) were added to the filtrate, and the
organic layer was separated. The aqueous phase was further extracted
with diethyl ether (2 × 40 mL), and the combined extracts were
washed with water (150 mL) and brine (80 mL). The extracts were
concentrated in vacuo, and the residue was purified by silica gel
4
assays, cells were seeded in 96-well plates at a density of 1 × 10 cells/
mL. Cells were treated with MUFAs (0.1−200 μM) for 48 h with a
change of media at 24 h. MTT activity was determined spectrophoto-
metrically. IC50 values were determined from plots of “percent
remaining MTT activity” versus “log [n-3 MUFA].” Caspase-3 activity
was measured in cell lysates by luminometry (Caspase-Glo 3/7 kit,
Promega; Alexandria, NSW, Australia). EC50 values were determined
from plots of “percent caspase-3 activation” versus “log [n-3 MUFA].”
6
Matrigel Droplet Cell Migration Assay. Cells (3.5 × 10 cells/
1
DCVC, affording 6a (0.432 g, 71%) as a colorless oil. H NMR (400
mL) were mixed with matrigel (Trevigen; Gaithersburg, MD), and 20
μL was loaded carefully onto the surface of six-well plates (Nunc;
Roskilde, Denmark) to produce well-defined droplets. Plates were
MHz, CDCl ): 9.76 (t, J = 1.6 Hz, 1H), 4.11 (q, JHH = 7.2 Hz, 2H),
3
HH
2
(
.41 (td, JHH = 7.2, 1.6 Hz, 2H), 2.28 (t, J = 7.6 Hz, 2H), 1.67−1.36
HH
13
m, 4H), 1.36−1.22 (m, 17H). C NMR (100 MHz, CDCl ): δ
incubated for 5 min at 37 °C under 5% CO to enable the matrigel
3
2
7
169
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
1
9
droplets to semisolidify. Endothelial cell growth medium-2 that was
supplemented with Single Quote and 10% fetal bovine serum
(
underwent subsequent energy minimizations. A total of ten induced-fit
protein conformations were generated for each of the test ligands.
Stage 3 involved redocking the test molecules into the corresponding
structures that were within 30.0 kcal/mol of the lowest energy
structure. Finally, the ligand poses were scored in stage 4 using a
combination of Prime and GlideScore scoring functions in which the
top ranked pose for each MUFA was chosen as the final result. The XP
Cambrex Bioscience; Mt Waverley, VIC, Australia) was added to
plates and incubated for 20 h under a 5% CO atomosphere at 37 °C,
after which cells that migrated out of matrigel droplets were counted
by inverted phase-contrast microscopy (Olympus CKX41).
COX-2 Activity of Cell Lysates. The effects of n-3 MUFAs on
2
8
47
COX-2 activity in confluent MDA-COX-2 cells (1 × 10 ) were
scoring function was used in all docking stages.
determined by two approaches. Heme (10 μM), arachidonic acid (20
μM), and the n-3 MUFA analogues (50 μM) were incubated for 5 min
at 25 °C in Tris-HCl buffer (100 mM, pH 8.0). COX-2-dependent
peroxidase activity was measured spectrophotometrically at 590 nm by
Statistical Analysis. All measurements were performed at least in
duplicate with experiments replicated on at least three occasions. Data
are presented throughout as mean ± SE and were analyzed by one-way
ANOVA in combination with Fisher’s Protected Least Significant
Difference test to detect differences between multiple treatments.
2
2
the oxidation of N,N,N′,N′-tetramethyl-p-phenylenediamine. COX-
2
-dependent PGE formation in cell lysates was determined similarly,
2
except that quantification was by LC-MS-MS. After a 10 min
incubation at 37 °C, ice-cold acetonitrile (0.8 mL) and internal
standard PGE -d (50 ng) were added and the mixture was loaded
onto Oasis HLB SPE cartridges (Waters, Rydalmere, NSW, Australia).
Eicosanoids were eluted with acetonitrile (4 mL) and ethyl acetate (2
mL), dried under nitrogen, and reconstituted in acetonitrile (100 μL)
for LC-MS analysis.
ASSOCIATED CONTENT
Supporting Information
Analytical and spectroscopic data for all compounds, and G
■
*
S
2
4
score
and ΔG calculations. This material is available free of charge
bind
via the Internet at http://pubs.acs.org.
Quantification of PGE Formation. LC-MS-MS was performed
on a Thermo Scientific TSQ Quantum Access Max System (Waltham,
MA). Eicosanoids were separated using a Gemini-NX C18 110A 3 μm
50 × 4.6 mm LC column (Phenomenex, Torrance, CA). Gradient
elution was with acetonitrile:water:acetic acid (A: 15:85:0.02, B:
AUTHOR INFORMATION
Corresponding Author
Tel: 61-2-9351-2326. Fax: 61-2-9351-4391. E-mail: michael.
murray@sydney.edu.au.
2
■
*
1
9
3
3
0:10:0.02) at an initial ratio of 65% A:35% B for 15 min, increasing to
5% A:65% B for 10 min, and then followed by a linear gradient from
5% A:65% B to 10% A:90% B over 10 min. The flow rate was 0.4
Notes
The authors declare no competing financial interest.
mL/min, the sample injection volume was 10 μL, and the MS was
operated in the electrospray negative ion mode with nitrogen as the
cone gas. The temperatures of the heated capillary and the vaporizer
were 261 °C and 475 °C, respectively. The scan time was 2 s, and the
interscan delay was 0.1 s. All eicosanoids were detected according to
parent and fragmentation ions using Thermo Xcalibur 2.1 software
ACKNOWLEDGMENTS
This study was supported by a grant from the Australian
National Health and Medical Research Council.
■
ABBREVIATIONS USED
■
(
Thermo Fischer Scientific Inc., Waltham, MA).
Computational Modeling. MUFA structures were built,
PUFA, polyunsaturated fatty acid; EPA, eicosapentaenoic acid;
AA, arachidonic acid; COX, cyclooxygenase; PG, prostaglandin;
MUFA, monounsaturated fatty acid; PEPPSI-IPr, [1,3-bis(2,6-
diisopropylphenyl)imidazol-2-ylidene](3-chloropyridyl)-
palladium(II) dichloride; MDA-CON, MDA-MB-468-control
cells; MDA-COX-2, MDA-MB-468 cell-stable transfectants that
overexpress COX-2; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide; MM-GBSA, molecular mechan-
ics and surface-generalized Born solvation models; vdW, van
der Waals; DCVC, dry column vacuum chromatography; XP,
extra precision
manipulated and adjusted for chemical correctness using Maestro
v9.1, Schrodinger, LLC, New York, NY) graphical user interface
employing MacroModel (v9.1, Schrodinger, LLC). Geometry
minimizations were performed on all ligands using the OPLS_2005
MacroModel, v9.1, Schrodinger, LLC) force field and the Truncated
(
̈
̈
(
̈
Newton Conjugate Gradient. Optimizations were converged to a
gradient rmsd below 0.012 kJ/mol or continued to a maximum of 500
iterations, at which point there were negligible changes in rmsd
gradients.
MUFAs were independently docked into COX-2 with Glide
version 5.7, Schrodinger, LLC) utilizing the extra precision (XP)
̈
(
scoring function to estimate protein−MUFA binding affinities. The
site targeted in the docking calculations was defined by the position of
the AA molecule observed in complex with COX-2 as part of the
crystal structure (PDB code: 3HS5). Protein preparation and
refinement protocols directed by the protein preparation facility
REFERENCES
■
(1) Mehlen, P.; Puisieux, A. Metastasis: a question of life or death.
Nat. Rev. Cancer 2006, 6, 449−458.
(2) Borst, P.; Jonkers, J.; Rottenberg, S. What makes tumors
multidrug resistant? Cell Cycle 2007, 6, 2782−2787.
(
̈
Glide v5.7 Schrodinger, LLC) were performed on both target and
molecule to achieve chemical correctness. Briefly, this included
deleting crystallographic waters, adding hydrogens, adjusting bond
orders and formal charges, and alleviating potential steric clashes via
protein minimization with the OPLS_2005 force field. The shape and
properties of the binding site were characterized and setup for docking
using the receptor grid generation panel (Glide v5.7 Schrodinger,
̈
LLC). A Coulomb-vdW scaling of 1.0/0.8 was set for protein/MUFA
vdW radii, respectively.
The induced-fit docking protocol was run from the graphical user
interface accessible within Maestro 9.1. The overall procedure has four
stages. Briefly, during stage 1, initial softened-potential Glide docking
was performed on a vdW scaled-down rigid-protein; a scaling of 0.7/
(3) Maillard, V.; Bougnoux, P.; Ferrari, P.; Jourdan, M. L.; Pinault,
M.; Lavillonniere, F.; Body, G.; Le Floch, O; Chajes, V. N-3 and N-6
fatty acids in breast adipose tissue and relative risk of breast cancer in a
case-control study in Tours, France. Int. J. Cancer 2002, 98, 78−83.
(4) Rose, D. P.; Connolly, J. M.; Rayburn, J.; Coleman, M. Influence
of diets containing eicosapentaenoic or docosahexaenoic acid on
growth and metastasis of breast cancer cells in nude mice. J. Natl.
Cancer Inst. 1995, 87, 587−592.
(5) Rose, D. P.; Connolly, J. M.; Coleman, M. Effect of omega-3 fatty
acids on the progression of metastases after the surgical excision of
human breast cancer cell solid tumors growing in nude mice. Clin.
Cancer Res. 1996, 2, 1751−1756.
0
.5 was set for protein/MUFA vdW radii, respectively. In stage 2,
(6) Schrey, M. P.; Patel, K. V. Prostaglandin E production and
2
sampling and refinement were performed on residues within 5.0 Å of
docked MUFA for each of the seven MUFA:protein complexes.
Amino acid side chains, as well as the protein backbone and MUFA,
metabolism in human breast cancer cells and breast fibroblasts.
Regulation by inflammatory mediators. Br. J. Cancer 1995, 72, 1412−
1419.
7
170
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
(
7) Chang, S. H.; Liu, C. H.; Conway, R.; Han, D. K.; Nithipatikom,
(26) Subbaramaiah, K.; Norton, L.; Gerald, W.; Dannenberg, A. J.
Cyclooxygenase-2 is overexpressed in HER-2/neu-positive breast
cancer: evidence for involvement of AP-1 and PEA3. J. Biol. Chem.
2002, 277, 18649−18657.
K.; Trifan, O. C.; Lane, T. F.; Hla, T. Role of prostaglandin E2-
dependent angiogenic switch in cyclooxygenase 2-induced breast
cancer progression. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 591−596.
(
8) Hwang, D.; Scollard, D.; Byrne, J.; Levine, E. Expression of
(27) Liu, X. H.; Rose, D. P. Differential expression and regulation of
cyclooxygenase-1 and -2 in two human breast cancer cell lines. Cancer
Res. 1996, 56, 5125−5127.
cyclooxygenase-1 and cyclooxygenase-2 in human breast cancer. J.
Natl. Cancer Inst. 1998, 90, 455−460.
(
9) Hla, T.; Bishop-Bailey, D.; Liu, C. H.; Schaefers, H.; Trifan, O. C.
(28) Takahashi, Y. K. F.; Noguchi, M.; Miwa, K.; Sato, H.; Seiki, M.;
Inoue, H.; Tanabe, T.; Yoshimoto, T. Activation of matrix metal-
loproteinase-2 in human breast cancer cells overexpressing cyclo-
oxygenase-1 or -2. FEBS Lett. 1999, 460, 145−148.
(29) Sharpe, C. R.; Collet, J. P.; McNutt, M.; Belzile, E.; Boivin, J. F.;
Hanley, J. A. Nested case-control study of the effects of non-steroidal
anti-inflammatory drugs on breast cancer risk and stage. Br. J. Cancer
2000, 83, 112−120.
(30) Davies, G.; Salter, J.; Hills, M.; Martin, L. A.; Sacks, N.; Dowsett,
M. Correlation between cyclooxygenase-2 expression and angiogenesis
in human breast cancer. Clin. Cancer Res. 2003, 9, 2651−2656.
(31) Harris, R. E.; Chlebowski, R. T.; Jackson, R. D.; Frid, D. J.;
Ascenseo, J. L.; Anderson, G.; Loar, A.; Rodabough, R. J.; White, E.;
McTiernan, A. Breast cancer and nonsteroidal anti-inflammatory
drugs: prospective results from the Women’s Health Initiative. Cancer
Res. 2003, 63, 6096−6101.
Cyclooxygenase-1 and -2 isoenzymes. Int. J. Biochem. Cell Biol. 1999,
1, 551−557.
10) Half, E.; Tang, X. M.; Gwyn, K.; Sahin, A.; Wathen, K.;
3
(
Sinicrope, F. A. Cyclooxygenase-2 expression in human breast cancers
and adjacent ductal carcinoma in situ. Cancer Res. 2002, 62, 1676−
1681.
(
11) Rozic, J. G.; Chakraborty, C.; Lala, P. K. Cyclooxygenase
inhibitors retard murine mammary tumor progression by reducing
tumor cell migration, invasiveness and angiogenesis. Int. J. Cancer
2
(
001, 93, 497−506.
12) Basu, G. D.; Pathangey, L. B.; Tinder, T. L.; Lagioia, M.;
Gendler, S. J.; Mukherjee, P. Cyclooxygenase-2 inhibitor induces
apoptosis in breast cancer cells in an in vivo model of spontaneous
metastatic breast cancer. Mol. Cancer Res. 2004, 2, 632−642.
(
13) Connolly, E. M.; Harmey, J. H.; O’Grady, T.; Foley, D.; Roche-
Nagle, G.; Kay, E.; Bouchier-Hayes, D. J. Cyclo-oxygenase inhibition
reduces tumor growth and metastasis in an orthotopic model of breast
cancer. Br. J. Cancer 2002, 87, 231−237.
(32) Giovannucci, E.; Egan, K. M.; Hunter, D. J.; Stampfer, M. J.;
Colditz, G. A.; Willett, W. C.; Speizer, F. E. Aspirin and the risk of
colorectal cancer in women. N. Engl. J. Med. 1995, 333, 609−614.
(33) Basu, G. D.; Pathangey, L. B.; Tinder, T. L.; Gendler, S. J.;
Mukherjee, P. Mechanisms underlying the growth inhibitory effects of
the cyclo-oxygenase-2 inhibitor celecoxib in human breast cancer cells.
Breast Cancer Res. 2005, 7, R422−R435.
(
14) Rao, C. V.; Reddy, B. S. NSAIDs and chemoprevention. Curr.
Cancer Drug Targets 2004, 4, 29−42.
(
15) Rawling, T.; Duke, C. C.; Cui, P. H.; Murray, M. Facile and
stereoselective synthesis of (Z)-15-octadecenoic acid and (Z)-16-
nonadecenoic acid: monounsaturated omega-3 fatty acids. Lipids 2010,
(34) Minoura, T.; Takata, T.; Sakaguchi, M.; Takada, H.; Yamamura,
M.; Hioki, K.; Yamamoto, M. Effect of dietary eicosapentaenoic acid
on azoxymethane-induced colon carcinogenesis in rats. Cancer Res.
1988, 48, 4790−4794.
4
5, 159−165.
16) Organ, M. G.; Avola, S.; Dubovyk, I.; Hadei, N.; Kantchev, E. A.;
O’Brien, C. J.; Valente, C. A user-friendly, all-purpose Pd-NHC-
NHC=N-heterocyclic carbene) precatalyst for the Negishi reaction: a
step towards a universal cross-coupling catalyst. Chem.Eur. J. 2006,
2, 4749−4755.
17) Backeberg, O. G.; Staskun, B. A novel reduction of nitriles to
aldehydes. J. Chem. Soc. 1962, 3961−3963.
18) Heffner, R. J.; Jiang, J.; Joullie, M. M. Total synthesis of
−)-nummularine F. J. Am. Chem. Soc. 1992, 114, 10181−10189.
19) Szymczak, M.; Murray, M.; Petrovic, N. Modulation of
(
(
́
(35) Begin, M. E.; Ells, G.; Horrobin, D. F. Polyunsaturated fatty
acid-induced cytotoxicity against tumor cells and its relationship to
lipid peroxidation. J. Natl Cancer Inst. 1988, 80, 188−194.
(36) Falconer, J. S.; Ross, J. A.; Fearon, K. C.; Hawkins, R. A.;
O’Riordain, M. G.; Carter, D. C. Effect of eicosapentaenoic acid and
other fatty acids on the growth in vitro of human pancreatic cancer cell
lines. Br. J. Cancer 1994, 69, 826−832.
1
(
(
(
(
(37) Palomer, A.; Cabre,
Entrena, A.; Gallo, M. A.; García, L; Mauleon
́
F.; Pascual, J.; Campos, J.; Trujillo, M. A.;
, D.; Espinosa, A.
angiogenesis by ω-3 polyunsaturated fatty acids is mediated by
cyclooxygenases. Blood 2008, 111, 3514−3521.
́
Identification of novel cyclooxygenase-2 selective inhibitors using
pharmacophore models. J. Med. Chem. 2002, 45, 1402−1411.
(38) Rowlinson, S. W.; Crews, B. C.; Lanzo, C. A.; Marnett, L. J. The
binding of arachidonic acid in the cyclooxygenase active site of mouse
prostaglandin endoperoxide synthase-2 (COX-2). A putative L-shaped
binding conformation utilizing the top channel region. J. Biol. Chem.
1999, 274, 23305−23310.
(
20) Horia, E.; Watkins, B. A. Comparison of stearidonic acid and α-
linolenic acid on PGE production and COX-2 protein levels in MDA-
2
MB-231 breast cancer cell cultures. J. Nutr. Biochem. 2005, 16, 184−
192.
(
21) Wang, D.; DuBois, R. N. Cyclooxygenase 2-derived
prostaglandin E regulates the angiogenic switch. Proc. Natl. Acad.
2
Sci. U.S.A. 2004, 101, 415−416.
(39) Vecchio, A. J.; Simmons, D. M.; Malkowski, M. G. Structural
basis of fatty acid substrate binding to cyclooxygenase-2. J. Biol. Chem.
2010, 285, 22152−22163.
(40) Kiefer, J. R.; Pawlitz, J. L.; Moreland, K. T.; Stegeman, R. A.;
Hood, W. F.; Gierse, J. K.; Stevens, A. M.; Goodwin, D. C.; Rowlinson,
S. W.; Marnett, L. J.; Stallings, W. C.; Kurumbail, R. G. Structural
insights into the stereochemistry of the cyclooxygenase reaction.
Nature 2000, 405, 97−101.
(41) Dong, L.; Vecchio, A. J.; Sharma, N. P.; Jurban, B. J.; Malkowski,
M. G.; Smith, W. L. Human cyclooxygenase-2 is a sequence
homodimer that functions as a conformational heterodimer. J. Biol.
Chem. 2011, 286, 19035−19046.
(42) Picot, D.; Loll, P. J.; Garavito, R. M. The X-ray crystal structure
of the membrane protein prostaglandin H2 synthase-1. Nature 1994,
367, 243−249.
(43) Kurumbail, R. G.; Stevens, A. M.; Gierse, J. K.; McDonald, J. J.;
Stegeman, R. A.; Pak, J. Y.; Gildehaus, D.; Miyashiro, J. M.; Penning,
T. D.; Seibert, K.; Isakson, P. C.; Stallings, W. C. Structural basis for
(
22) Petrovic, N.; Murray, M. Using N,N,N′,N′-tetramethyl-p-
phenylenediamine (TMPD) to assay cyclooxygenase activity in vitro.
Methods Mol. Biol. 2010, 594, 129−140.
(
23) Bhattacharyya, D. K.; Lecomte, M.; Rieke, C. J.; Garavito, M.;
Smith, W. L. Involvement of arginine 120, glutamate 524, and tyrosine
55 in the binding of arachidonate and 2-phenylpropionic acid
3
inhibitors to the cyclooxygenase active site of ovine prostaglandin
endoperoxide H synthase-1. J. Biol. Chem. 1996, 271, 2179−2184.
(
24) Rowlinson, S. W.; Kiefer, J. R.; Prusakiewicz, J. J.; Pawlitz, J. L.;
Kozak, K. R.; Kalgutkar, A. S.; Stallings, W. C.; Kurumbail, R. G.;
Marnett, L. J. A novel mechanism of cyclooxygenase-2 inhibition
involving interactions with Ser-530 and Tyr-385. J. Biol. Chem. 2003,
2
(
78, 45763−45769.
25) Ristimaki, A.; Sivula, A.; Lundin, J.; Lundin, M.; Salminen, T.;
̈
Haglund, C.; Joensuu, H.; Isola, J. Prognostic significance of elevated
cyclooxygenase-2 expression in breast cancer. Cancer Res. 2002, 62,
632−635.
7
171
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172

Journal of Medicinal Chemistry
Article
selective inhibition of cyclooxygenase-2 by anti-inflammatory agents.
Nature 1996, 384, 644−648.
(
44) Duffy, P. E.; Quinn, S. M.; Roche, H. M.; Evans, P. Synthesis of
trans-vaccenic acid and cis-9-trans-11-conjugated linoleic acid.
Tetrahedron 2006, 62, 4838−4843.
(
45) Joshi, B. S.; Viswanathan, N.; Gawad, D. H.; Von Philipsborn,
13
W. Piperaceae alkaloids: Part I. Structure of piperstachine; C- and
1
H-NMR studies. Helv. Chim. Acta 1975, 58, 1551−1559.
(
46) Zhou, J.; Fu, G. C. Palladium-catalyzed Negishi cross-coupling
reactions of unactivated alkyl iodides, bromides, chlorides, and
tosylates. J. Am. Chem. Soc. 2003, 125, 12527−12530.
(
47) Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.;
Greenwood, J. R.; Halgren, T. A.; Sanschagrin, P. C.; Mainz, D. T.
Extra precision glide: Docking and scoring incorporating a model of
hydrophobic enclosure for protein−ligand complexes. J. Med. Chem.
2
006, 49, 6177−6196.
7
172
dx.doi.org/10.1021/jm300673z | J. Med. Chem. 2012, 55, 7163−7172