Indomethacin analogues that enhance doxorubicin cytotoxicity in multidrug resistant cells without cox inhibitory activity

Article abstract of DOI:10.1021/ml100292y

Conformationally restricted indomethacin analogues were designed and prepared from the corresponding 2-substituted indoles, which were synthesized by a one-pot isomerization/enamide-ene metathesis as the key reaction. Conformational analysis by calculations, NMR studies, and X-ray crystallography suggested that these analogues were conformationally restricted in the s-cis or the s-trans form due to the 2-substituent as expected. Their biological activities on cyclooxygenase-1 (COX-1) inhibition, cyclooxygenase-2 (COX-2) inhibition, and modulation of MRP-1-mediated multidrug resistance (MDR) are described. Some of these indomethacin analogues enhanced doxorubicin cytotoxicity, although they do not have any COX inhibitory activity, which suggests that the MDR-modulating effect of an NSAID can be unassociated with its COX-inhibitory activity. This may be an entry into the combination chemotherapy of doxorubicin with a MDR modulator.

Full text of DOI:10.1021/ml100292y

LETTER
pubs.acs.org/acsmedchemlett
Indomethacin Analogues that Enhance Doxorubicin Cytotoxicity
in Multidrug Resistant Cells without Cox Inhibitory Activity
Mitsuhiro Arisawa,*,† Yayoi Kasaya,† Tohru Obata,‡ Takuma Sasaki,‡ Mika Ito,†,§ Hiroshi Abe,§
Yoshihiro Ito,§ Akihito Yamano,|| and Satoshi Shuto*,†
†Faculty of Pharmaceutical Sciences, Hokkaido University, Kita 12, Nishi 6, Kita-ku, Sapporo 060-0812, Japan
‡School of Pharmacy, Aichi Gakuin University, 1-100 Kusumoto-cho, Chikusa-ku, Nagoya 464-8650, Japan
§Nano Medical Engineering Laboratory, RIKEN Advanced Science Institute, 2-1, Hirosawa, Wako 351-0198, Japan
Rigaku Corporation, X-ray Research Laboratory, 3-9-12 Matsubara-cho, Akishima, Tokyo 196-8666, Japan
S Supporting Information
b
ABSTRACT: Conformationally restricted indomethacin analogues were designed and
prepared from the corresponding 2-substituted indoles, which were synthesized by a
one-pot isomerization/enamide-ene metathesis as the key reaction. Conformational
analysis by calculations, NMR studies, and X-ray crystallography suggested that these
analogues were conformationally restricted in the s-cis or the s-trans form due to the
2-substituent as expected. Their biological activities on cyclooxygenase-1 (COX-1)
inhibition, cyclooxygenase-2 (COX-2) inhibition, and modulation of MRP-1-mediated multidrug resistance (MDR) are described.
Some of these indomethacin analogues enhanced doxorubicin cytotoxicity, although they do not have any COX inhibitory activity,
which suggests that the MDR-modulating effect of an NSAID can be unassociated with its COX-inhibitory activity. This may be an
entry into the combination chemotherapy of doxorubicin with a MDR modulator.
KEYWORDS: Cancer, conformation analysis, cytotoxicity, drug design, nitrogen heterocycles
ultidrug resistance (MDR) continues to be a serious
problem in cancer therapy and is one of the limiting factors
conformations of indomethacin in complexes with COX-1 and
COX-2 are the s-trans form7 and the s-cis form,8 respectively
(4COX.pdb and 1PGG.pdb). We thought that the introduction
of an alkyl substituent other than a methyl group at the 2-position
of indomethacin might restrict its conformation in the s-cis or
s-trans form, respectively, due to steric repulsion or πꢀπ stacking
interaction by the introduced substituent with the N-4-chloroben-
zoyl side chain. When the substituent at the 2-position (R) is an
alkyl group bulkier than methyl, the s-cis form can be more stable
than the s-trans form due to steric repulsion. However, introduc-
tion of a phenyl substituent at the 2-position may make the s-trans
form more stable due to πꢀπ stacking (Figure 2). That these
conformationally restricted analogues may be selectively active
toward COX-1 or COX-2 offers some insight into the relationship
betweenthe COX-1 and/or COX-2inhibitoryeffect and theMDR
property of cancer cells. Since we have just developed a novel
synthetic method for various substituted indoles using a combina-
tion of isomerization and enamide-ene metathesis,9,10 conse-
quently we decided to design and synthesize the indomethacin
analogues 1bꢀd substituted at the 2-position.11
M
in the development and application of antitumor drugs. Doxor-
ubicin is a chemotherapeutic drug widely used for the treatment of
various solid tumors. However, its efficacy is often limited by the
development of MDR, which has been linked to the up-regulation
of P-glycoprotein (P-gp) in cancer cells.1 It is known that tumor
progression often is accompanied by increased COX-2 expression,
and selective COX-2 inhibitors prevent the formation of multiple
tumor types in experimental animals.2 In fact, it is also known that
MDR caused by multidrug resistance associated protein 1 (MRP-1)
is modulated by nonsteroidal antiinflammatory drugs (NSAIDs),
nonselective COX inhibitors such as indomethacin 1a, or COX-2
selective inhibitors.3,4 Mazzanti reported that the MDR phenotype
might be associated with the expression of COX-2 in a human
hepatocellular carcinoma cell line.5 Cho showed a mechanism by
which COX inhibitors exerted an enhancing effect on the cytotoxic
effect of doxorubicin via direct inhibition of P-gpATPase activity in
human esophageal squamous cell carcinoma cell lines.6 However, it
is less certain whether COX-1 and/or COX-2 activity is directly
related to the MDR property of cancer cells.2
The present study was carried out to investigate whether the
COX activity of a series of our conformationally restricted
Indomethacin (1a, Figure 1) is a clinically useful NSAID with
an indole structure, which inhibits both COX-1 and COX-2. In
indomethacin, the two stable forms, the s-trans form and the s-cis
form, around the amide bond are known (Figure 1). The COX
inhibitory activities are closely related to the conformation of
indomethacin itself. It has been reported that the bioactive
Received: December 7, 2010
Accepted: March 3, 2011
Published: March 17, 2011
r
2011 American Chemical Society
353
dx.doi.org/10.1021/ml100292y ACS Med. Chem. Lett. 2011, 2, 353–357
|

ACS Medicinal Chemistry Letters
LETTER
Figure 1. Structure of indomethacin (1a).
Figure 3. Compounds stable in the s-cis conformer by molecular field
calculation. Numbers in parentheses indicate energy differences be-
tween the two comformers.
Figure 4. Compounds stable in the s-trans conformer by molecular field
calculation. Numbers in parentheses indicate energy differences be-
tween the two comformers.
Figure 2. Designed indomethacin derivatives (1bꢀd).
indomethacin analogues was related to their doxorubicin cyto-
toxicity using MDR cell lines.
was more stable than the s-trans form. In C, the model of 1c, a
huge energy increase was observed at 330°, probably due to the
steric bulk of the ethyl.
’ RESULTS AND DISCUSSION
Conformational Analysis by Calculations. Since conforma-
tional restriction is an important point in this research project, we
decided to investigate the conformational stability of the de-
signed indomethacin derivatives by theoretical calculations.
Thus, we first investigated the stable conformation of com-
pounds 1aꢀd by molecular field calculations using Macromodel.
As a result, it appeared that 1aꢀc would favor the s-cis form
(Figure 3) while 1d would favor the s-trans form (Figure 4).
Energy differences between the two forms in these compounds
were quite high, as indicated in the figures.12
Next, the rotational barrier energy around the C7aꢀN1ꢀ
C10ꢀO dihedral angle of the model compounds AꢀD, in which
the carboxymethyl group at the 3-position of the indomethacin
analogues is replaced with a methyl group (Figure 5), was
calculated on the basis of density functional theory (DFT).
The dihedral angle was rotated from 0° to 360° at intervals of
10°, and the single point energies of the optimized forms were
calculated with B3LYP/6-31G* to obtain the energy profile. The
dihedral angle defines the s-cis form at 180° and the s-trans form
at 0°, respectively. As shown in Figure 5, in D (the model of 1d),
the s-trans form was relatively more stable than the s-cis form
while in A, B, and C (the models of 1a, 1b, and 1c), the s-cis form
Hence, both molecular field and density functional theory
calculations led to almost the same results, which indicated that
the substituents at the 2-position on the indole ring can affect the
conformation of indomethacin derivatives, as expected.
Chemistry. Most of the known indomethacin analogues are
modified at the carboxylic acid moiety or at the 5-position of the
indole ring, which were prepared by the traditional Fisher indole
synthesis.13ꢀ15 However, the designed target compounds
(especially 1c and 1d) could not be prepared by the traditional
methods. The synthetic route used to construct the target
compounds (1aꢀd) is depicted in Scheme 1. The substituted
indoles (3aꢀd), the key synthetic compounds, were obtained
from the corresponding N-allyl-N-toluenesulfonyl-2-vinylaniline
derivatives by a one-pot isomerization/enamide-ene metathesis
and subsequent removal of the p-toluenesulfonyl group on the
nitrogen. The introduction of a substituent at the 3-position on
the indole ring was accomplished by a conventional 3-alkylation
or Mannich reaction to yield 4aꢀd. 4-Chlorobenzylation of
4aꢀd and subsequent hydrogenation led to 1aꢀd, which have
not been prepared so far except for 1b,16 probably due to
limitations in the traditional synthetic methods as described
above.
354
dx.doi.org/10.1021/ml100292y |ACS Med. Chem. Lett. 2011, 2, 353–357

ACS Medicinal Chemistry Letters
LETTER
Figure 5. DFT calculation.
Scheme 1. Synthesis of Indomethacin 1a and Its Derivatives (1bꢀ1d)a
a Conditions: (a) KOH, EtOH, THF, reflux, 78ꢀ98%. (b) ICH2CO2Bn, nBuLi, ZnCl2, THF, 0 °C to rt, 30ꢀ88%. (c) (i) Me2NH, HCHO, AcOH. (ii)
MeI, toluene. (iii) KCN, MeCN/H2O = 1:3, reflux. (iv) cHCl, dioxane, reflux. (v) BnBr, CsCO3, DMF, 0 °C, rt, 76% (5 steps). (d) 4-ClC6H4COCl,
tBuOK, THF, ꢀ78 °C, 69ꢀ95%. (e) 10% Pd/C, H2, AcOEt, rt or ꢀ15 °C, 76ꢀ95%.
Conformational Analysis by NMR. NOE experiments for 1a
and 1c were carried out to investigate the stable conformation of
the compounds in solution. Irradiation of the aromatic methine
proton at the ortho-position of the benzoyl carbonyl of 1a led to
an NOE with the methine proton at the 7 position of the indole
ring of 1.7% and with the methyl proton at the 2 position of
0.54% (Figure 6). These results suggest that the aromatic
methine proton at the ortho-position of the benzoyl carbonyl
and the methine proton at the 7 position of the indole ring are
close to each other. Thus, 1a would be restricted in the s-cis form,
at least to some extent, as speculated. Furthermore, as shown in
Figure 6. NOE correlations of 1a and c. Arrows start from the
equivalent methine.
Figure 6, the NOE in 1c was observed only at the 7-position
when the same aromatic methine proton was irradiated, which
suggested that 1c would be significantly restricted in the s-cis
form. The ethyl group appeared to be more effective in restricting
the conformation than the methyl group (1a vs 1c). These results
support our hypothesis that the introduction of a bulkier substituent
at the 2-position of indomethacin can restrict its conformation in
355
dx.doi.org/10.1021/ml100292y |ACS Med. Chem. Lett. 2011, 2, 353–357

ACS Medicinal Chemistry Letters
LETTER
Figure 7. ORTEP drawings of (a) 1a and (b) 1d.
Figure 8. Influence of the indole compounds 1aꢀd on the cytotoxicity
of doxorubicin. T98G cells were pretreated with the compounds [50
μM] for 2 h and then cotreated with doxorubicin [0, 0.3, 1, or 3 μM] for
3 days. Survival was measured by crystal violet staining, as described in
the Supporting Information. The data shown are the mean percentages
( SD for a minimum of three experiments.
Table 1. SAR Studies on the COXs Inhibition by 1aꢀd (n = 3)
COX-1
COX-2
ihibition
ihibition
compound
(100 μM, %) IC50 (μM) (100 μM, %) IC50 (μM)
As shown in Figure 8, doxorubicin showed moderate
cytotoxicity in the cell line because of the MRP-1 expression,
e.g., about 40% inhibition of the cell growth at 1 μM. In accord
with the previously reported results,21 the coadministration of
indomethacin (1a) effectively enhanced the cytotoxicity of
doxorubicin by about two times at 1 μM. In this system, 1c
and 1d, having only insignificant COX-1 and COX-2 inhibi-
tory activity (IC50 > 400 μM), surprisingly enhanced the
cytotoxicity of doxorubicin, as was the case with 1a. Also, 1c
and 1d without doxorubicin did not show any cytotoxicity in
this cell line. The previous biological studies suggested that
the COX inhibitory effect of the NSAIDs, including indo-
methacin 1a, may be related to the MDR of cancer cells.3,4,6
Accordingly, these results, demonstrating that the MDR-modulat-
ing effect of the NSAID indomethacin would not be associated with
its COX-inhibitory activity, are of significant importance.
1a (synthesized)
1ba
81 ( 11
79 ( 6.4
<0
0.01
0.01
89 ( 0.7
76 ( 1.4
14 ( 0.9
13 ( 2.0
83 ( 0.45
2 ( 1.5
18
39
1c
>400
>400
0.12
>100
>400
>400
20
1d
<0
1a (purchased)
aspirin
110 ( 11
38 ( 3.1
>400
a These data are in accord with those reported previously.20
the s-cis form due to steric repulsion of the substituent at the
2-position with the N-acyl side chain.17
Conformational Analysis by X-ray Crystallography. The
crystal structures of indomethacin 1a and its derivative 1d were
determined by single-crystal X-ray diffraction methods
(Figure 7).18 Crystals of 1a and 1d suitable for X-ray analysis were
grown in 50% aqueous ethanol and in ethyl acetate, respectively.
The analysis showed that a crystal of 1a consists of units of the s-cis
and the s-trans forms in a ratio of 2:1, which is in accord with the
previously reported R type polymorphism of indomethacin.14,19
On the other hand, compound 1d crystallized only in the s-trans
form, as we had hypothesized, which was identical with the stable
form suggested by the theoretical calculations.
Biological Activity. COX Inhibition. Table 1 shows the COX-
1 and COX-2 inhibitory activities of indomethacin 1a and its
derivatives 1bꢀd at a concentration of 100 μM and also their
IC50 values. Although 1a and 1b showed potent but nonselective
COX-inhibitory activity, compounds 1c and 1d were almost
inactive versus COX-1 and COX-2. It is possible that the
2-substituent introduced in 1c and 1d might affect the binding
to COX-1 and COX-2 due to steric hindrance.
Effect on Multidrug Resistance (MDR). The development of
potent and nontoxic agents, which have minimal side effects, to
prevent MRP-1-induced MDR is one of the major aims of medical
and pharmacological research. Thus, according to the reported
procedure,21 compounds 1aꢀd were evaluated in a cell biological
cytotoxicity assay employing the MRP-1 expressing human glio-
blastoma cell line T98G as a model system of MDR tumors at a
concentration of 50 μMwithorwithout0.3, 1, or3μM doxorubicin.
’ CONCLUSION
We designed and synthesized the conformationally restricted
indomethacin analogues 1c and 1d, whose conformations were
restricted due to the effect of the 2-substituent as expected.
Although these analogues were inactive versus COX-1 and COX-
2, they surprisingly enhanced the cytotoxicity of doxorubicin in
spite of their non-COX inhibitory activities. These compounds
may become a prototype for developing a useful MDR modulator
without cytotoxicity.
’ ASSOCIATED CONTENT
S
Supporting Information. Detailed experimental proce-
b
dures. This material is available free of charge via the Internet at
http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Telephone: þ81-11-706-3976. Fax: þ81-11-706-3769. E-mail:
arisawa@pharm.hokudai.ac.jp (M.A.) or shu@pharm.hokudai.
ac.jp (S.S.).
356
dx.doi.org/10.1021/ml100292y |ACS Med. Chem. Lett. 2011, 2, 353–357

ACS Medicinal Chemistry Letters
LETTER
Author Contributions
Nuss, G. W.; Winter, C. A. Non-Steroid Anti-Inflammatory Agents.
J. Am. Chem. Soc. 1963, 85, 488–489.
(14) Yamamoto, H. 1-Acyl-indoles. II. A New Syntheses of 1-
(p-chlorobenzoyl)-5-methoxy-3-indolylacetic Acid and Its Polymorph-
ism. Chem. Pharm. Bull. 1968, 16, 17–21.
(15) Mukai, C.; Takahashi, Y. A. New Entry to the Synthesis of 2,3-
Disubstituted Indoles. Org. Lett. 2005, 7, 5793–5796.
(16) Compound 1b is a known.
(17) An NOE experiment of 1d could not performed, since the
aromatic proton signals of the 4-chlorobenzoyl and the 2-phenyl groups
were overlapped and could not be assigned.
(18) CCDC 814524 and CCDC 814525 contain the supplementary
crystallographic data for 1a and 1d, respectively. These data can be
obtained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
M.A., Y.K., and S.S. contributed to the design, synthesis, calcula-
tion by Macromodel and study on COX-inhibitory activity. T.O.
and T.S. contributed to the study on modulation of MRP-1
mediated multidrug resistance. M.I., H.A., and Y.I. contributed to
the DFT calculation. A.Y. contributed to X-ray crystallography
analysis. M.A. and S.S. contributed to writing of the manuscript,
and are also the corresponding authors.
Funding Sources
This research was supported by a Grant-in-Aid from the Ministry
of Education, Culture, Sports, Science and Technology, Japan.
’ REFERENCES
(19) Hꢀedoux, A.; Guinet, Y.; Capet, F.; Paccou, L.; Descamps, M.
Rayleigh Scattering, Long-time Tails, and the Harmonic Spectrum of
Topologically Disordered Systems. Phys. Rev. B 2008, 77, 094205.
(20) Touhey, S.; O’Connor, R.; Plunkett, S.; Maguire, A.; Clynes, M.
Structureꢀactivity Relationship of Indomethacin Analogues for MRP-1,
COX-1 and COX-2 Inhibition. Eur. J. Cancer 2002, 38, 1661–1670.
(21) Rosenbaum, C.; R€ohrs, S.; M€uller, O.; Waldmann, H. Modula-
tion of MRP-1-Mediated Multidrug Resistance by Indomethacin Ana-
logues. J. Med. Chem. 2005, 48, 1179–1187.
(1) Mineault, M.; Hauke, R.; Batra, S. K. Recent Advances on
the Molecular Mechanisms Involved in the Drug Resistance of Cancer
Cells and Novel Targeting Therapies. Clin. Pharmacol. Ther. 2008, 83,
673–691.
(2) Dannenberg, A. J.; Lippman, S. M.; Mann, J. R.; Subbaramaiah,
K.; Dubois, R. N. Cyclooxygenase-2 and Epidermal Growth Factor
Receptor: Pharmacologic Targets for Chemoprevention. J. Clin. Oncol.
2005, 23, 254–266.
(3) Matsunaga, S.; Asano, T.; Asano, A. T.; Fukunaga, Y. Indo-
methacin Overcomes Doxorubicin Resistance with Inhibiting Multi-
drug Resistance Protein 1 (MRP1). Cancer Chemother. Pharmacol. 2006,
58, 348–353.
(4) Roy, K. R.; Reddy, G. V.; Maitreyi, L.; Agarwal, S.; Achari, C.;
Vali, S.; Reddanna, P. Celecoxib Inhibits MDR1 Expression through
COX-2-dependent Mechanism in Human Hepatocellular Carcinoma
(HepG2) Cell Line. Cancer Chemother. Pharmacol. 2010, 65, 903–911.
(5) Fantappie, O.; Masini, E.; Sardi, I.; Raimondi, L.; Bani, D.;
Solazzo, M.; Cannacci, A.; Mazzanti, R. The MDR Phenotype is
Associated with the Expression of COX-2 and iNOS in a Human
Hepatocellular Carcinoma Cell Line. Hepatology 2002, 35, 843–852.
(6) Yu, L.; Wu, W. K. K.; Li, Z. J.; Liu, Q. C.; Li, H. T.; Wu, Y. C.;
Cho, C. H. Enhancement of Doxorubicin Cytotoxicity on Human
Esophageal Squamous Cell Carcinoma Cells by Indomethacin and
4-[5-(4-Chlorophenyl)-3-(trifluoromethyl)-1H-pyrazol-1-
yl]benzenesulfonamide (SC236) via Inhibiting P-Glycoprotein Activity.
Mol. Pharmacol. 2009, 75, 1364–1373.
(7) Loll, P. J.; Picot, D.; Ekabo, O.; Garavito, R. M. Synthesis and Use
of Iodinated Nonsteroidal Antiinflammatory Drug Analogs as Crystal-
lographic Probes of the Prostaglandin H2 Synthase Cyclooxygenase
Active Site. Biochemistry 1996, 35, 7330–7340.
(8) Kurumbail, R. G.; Stevens, A. M.; Gierse, J. K.; Mcdonald, J. J.;
Stegeman, R. A.; Pak, D.; Gildehaus, J. Y.; Miyashiro, J. M.; Penning,
T. D.; Seibert, K.; Isakson, P. C.; Stallings, W. C. Structural Basis for
Selective Inhibition of Cyclooxygenase-2 by Anti-Inflammatory Agents.
Nature 1996, 384, 644–648.
(9) Arisawa, M.; Terada, Y.; Takahashi, K.; Nishida, A. Development
of Isomerization and Cycloisomerization with Use of a Ruthenium
Hydride with N-Heterocyclic Carbene and Its Application to the
Synthesis of Heterocycles. J. Org. Chem. 2006, 71, 4255–4261.
(10) Kasaya, Y.; Hoshi, K.; Terada, Y.; Nishida, A.; Shuto, S.;
Arisawa, M. Aromatic Enamide/Ene Metathesis toward Substituted
Indoles and Its Application to the Synthesis of Indomethacins. Eur.
J. Org. Chem. 2009, 4606–4613.
(11) Demethoxy indomethacine 1b would provide some informa-
tion concerning an effect of the methoxy group.
(12) Energy difference of 1d is lower than those of 1aꢀc, which is
enough to restrict the conformation. Energy difference of 1c is higher
than that of 1b, probably due to the influence of the steric bulk of
the ethyl.
(13) Shen, T. Y.; Ellis, R. L.; Windholz, T. B.; Matzuk, A. R.;
Rosegay, A.; Lucas, S.; Witzel, B. E.; Stammer, C. H.; Wilson, A. N.;
Holly, F. W.; Willett, J. D.; Sarett, L. H. J.; Holtz, W. J.; Risley, E. A.;
357
dx.doi.org/10.1021/ml100292y |ACS Med. Chem. Lett. 2011, 2, 353–357