Isoxazole-Based-Scaffold Inhibitors Targeting Cyclooxygenases (COXs)

Article abstract of DOI:10.1002/cmdc.201500439

A new set of cyclooxygenase (COX) inhibitors endowed with an additional functionality was explored. These new compounds also contained either rhodamine 6G or 6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline, two moieties typical of efflux pump substrates and

Full text of DOI:10.1002/cmdc.201500439

DOI: 10.1002/cmdc.201500439
Full Papers
Isoxazole-Based-Scaffold Inhibitors Targeting
Cyclooxygenases (COXs)
Maria Grazia Perrone,[a] Paola Vitale,[a] Andrea Panella,[a] Savina Ferorelli,[a]
Marialessandra Contino,[a] Antonio Lavecchia,[b] and Antonio Scilimati*[a]
A new set of cyclooxygenase (COX) inhibitors endowed with
an additional functionality was explored. These new com-
pounds also contained either rhodamine 6G or 6,7-dimethoxy-
1,2,3,4-tetrahydroisoquinoline, two moieties typical of efflux
pump substrates and inhibitors, respectively. Among all the
synthesized compounds, two new COX inhibitors with oppo-
site selectivity were discovered: compound 8 [N-(9-{2-[(4-{2-[3-
(5-chlorofuran-2-yl)-4-phenylisoxazol-5-yl]acetamido}butyl)car-
bamoyl]phenyl-6-(ethylamino)-2,7-dimethyl-3H-xanthen-3-ylide-
ne}ethanaminium chloride] was found to be a selective COX-
1 inhibitor, whereas 17 (2-[3,4-bis(4-methoxyphenyl)isoxazol-5-
yl]-1-[6,7-dimethoxy-3,4-dihydroisoquinolin-2-(1H)-yl]ethanone)
was found to be a sub-micromolar selective COX-2 inhibitor.
However, both were shown to interact with P-glycoprotein.
Docking experiments helped to clarify the molecular aspects of
the observed COX selectivity.
Introduction
Cyclooxygenase (COX)-1 and -2 isoforms share 60–65% se-
quence identity within species, and about 85–90% sequence
identity between different species.[1] They are bifunctional en-
zymes that carry out two sequential reactions in spatially dis-
tinct but mechanistically coupled active sites: the bisoxygena-
tion of arachidonic acid (AA) to prostaglandin G2 (PGG2), and
the reduction of PGG2 to PGH2. COX-1 and COX-2 are homo-
dimers of 70 kDa subunits, and dimerization is required for
structural integrity and catalytic activity. Each monomer con-
tains a COX and a peroxidase active site and consists of three
structural domains: a short N-terminal epidermal growth factor
domain, a membrane-binding domain, and a large, globular C-
terminal catalytic domain. COX monomers forming the dimer,
although identical in the resting enzyme, differ from one an-
other during catalysis. hCOXs function, in practice, as confor-
mational heterodimers having a catalytic monomer (E cat) with
a bound heme and an allosteric monomer (E allo) lacking
heme. Studies of AA substrate turnover at high enzyme-to-sub-
strate ratios indicate that non-substrate-like inhibitors bind the
COX site of E allo to modulate the properties of E cat. The non-
functioning subunit may provide structural support, enabling
its partner monomer to catalyze COX reactions.[2]
tions to determine the binding mode. For more complex sys-
tems, as in the case of COX E allo/E cat, the chance of observ-
ing a useful interaction for a randomly chosen compound de-
creases dramatically. We further explored these implications in
the design of novel compounds in an attempt to potentiate
the activity of our COX-1-selective inhibitors to develop thera-
nostic agents for in vitro and in vivo experiments targeting
specific tissues and cells.[3,4]
Traditional nonsteroidal anti-inflammatory drugs (tNSAIDs)
target both COX isoenzymes. Coxibs, mainly targeting COX-2,
have been the focus of several studies aimed at developing
drugs lacking gastrointestinal toxicity. Conversely, few selective
COX-1 inhibitors have been described, although a number of
pharmacological investigations have proven the consolidated
involvement of this enzyme in the carcinogenesis, pathogene-
sis of neuroinflammation, and pain.[1] Nowadays, selective COX-
1 inhibition is considered a promising theranostic strategy in
cancer and neuroinflammation-derived neurodegenerative dis-
eases.[1,5–7]
In many type of cancers and neurodegenerative diseases,
drug efficacy strongly affects multidrug resistance (MDR),[8]
a phenomenon that limits drug cell accumulation due to efflux
pump action, such as by P-glycoprotein (P-gp), a key member
of the superfamily of ATP-binding cassette (ABC) efflux trans-
porters. This prompted us to design suitable selective COX-1 in-
hibitors[1,3–5,9] that were also capable of residing for a longer
time inside the cell. Until now, no attempt to correlate COX-
1 and P-gp activity has been successful.
Interactions of varying complexities between ligands and
(large) biomolecules (i.e., enzymes, DNA, receptors) are often
studied using a simple model of ligand–biomolecule interac-
[a] Dr. M. G. Perrone,+ Dr. P. Vitale,+ Dr. A. Panella, Prof. S. Ferorelli,
Dr. M. Contino, Prof. A. Scilimati
Dipartimento di Farmacia—Scienze del Farmaco
Università degli Studi di Bari “A. Moro”, Via E. Orabona 4, 70125 Bari (Italy)
E-mail: antonio.scilimati@uniba.it
Hence, the new projected compounds have a COX inhibitor
recognition moiety, constituted by one of three well-known se-
lective inhibitors such as indomethacin,[10] P6,[3,11] or mofezo-
lac[12] (Figure 1), and a P-gp-recognized fragment as a further
functionality. Notably, rhodamine 6G (Rh6G) is a P-gp substrate,
and 6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline (6,7-DM-
[b] Prof. A. Lavecchia
Dipartimento di Farmacia, “Drug Discovery” Laboratory
Università di Napoli “Federico II”, Via D. Montesano 49, 80131 Napoli (Italy)
[+] These authors contributed equally to this work.
ChemMedChem 2016, 11, 1172 – 1187
1172
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
derived from leads such as indomethacin (COX-1 IC50 =
0.099 mm; COX-2 IC50 =4.2 mm),[9] P6 (COX-1 IC50 =22 mm; COX-
2 IC50 >50 mm),[3] and mofezolac[11] (COX-1 IC50 =0.024 mm;
COX-2 IC50 >50 mm) (Figure 1) to further identify crucial chemi-
cal entities responsible for COX selective inhibition.
Indomethacin was chosen as one of the most prescribed an-
algesic and antipyretic tNSAIDs. Mofezolac is the active ingredi-
ent of the analgesic drug Disopain, commercially available in
Japan. P6 represents a prototype of the isoxazole scaffold we
have been studying.[5,19,20]
Figure 1. Structures of indomethacin, P6, and mofezolac.
THIQ) is a structural key element of elacridar and tariquidar
(two P-gp saturating substrates or inhibitor-like com-
pounds).[13–16] Indomethacin, P6, and mofezolac were then
linked to either a Rh6G or 6,7-DM-THIQ moiety to first explore
the COX inhibitor recognition and, secondly, to verify if such
new molecules retain P-gp substrate/inhibitor activity.
P6 was converted into P6-COOH (Figure 2; 32 and 58% in-
hibition of COX-1 and COX-2 activity, respectively, at 50 mm) by
adding the carboxylic group to the P6 isoxazole side chain.[19]
A series of conformationally restricted and regioisomeric in-
domethacin analogues were already described as multidrug-re-
sistant modulators. Evaluation of the inhibitory effects of these
compounds on COX, P-gp, and multidrug resistance (MDR) in-
dicated that modulation of multidrug-resistant P-gp and multi-
drug-resistant protein 1 by a number of NSAIDs is not associat-
ed with COX-1 and COX-2 inhibitory activities.[17] Ibuprofen
overcame MDR in human doxorubicin-resistant uterine sarco-
ma cells (MES-SA/Dx-5), expressing high levels of efflux pump
P-gp.[18] Rh6G was also conjugated with indomethacin, P6, and
mofezolac in an attempt to produce fluorescent molecules tar-
geting COX-1 expressed in healthy and pathological tissues/
cells.
Figure 2. Structures of P6-COOH and Br-P6.
All three COX inhibitors (indomethacin, P6-COOH, and mofezo-
lac) were linked to the fluorescent fragment Rh6G[21] or to the
efflux-pump-interacting moiety 6,7-DM-THIQ,[14] obtaining com-
pounds 7–9, 14–17, and 20 (Tables 1 and 2, and Schemes 2
and 4–6).
The COX-1 inhibitor portion of the compounds and the
Rh6G or 6,7-DM-THIQ moieties were coupled by a linker of the
appropriate length[21,22] to first allow the entrance of the COX-
1-inhibitor portion into the COX active site to reach the tyrosyl
radical (Tyr385) (Figure 3)[21,22] and hence block the auto-cata-
lytic oxidation of AA to PGH2. Second, the linker length was se-
lected so as to allow the additive Rh6G or 6,7-DM-THIQ func-
tionality to remain outside of the constriction site formed by
Tyr355, Arg120, and Glu524 (Figure 3a). The same amino acid
residues allow easier access to the COX-2 active site, which has
a larger side pocket than the COX-1 isoform (Figure 3b). Target
Herein we report the synthesis of these original compounds,
their ability to inhibit COX activity, and their extent of interac-
tion with the P-gp efflux pump expressed in the MDCK-MDR1
cell line. Results of the docking experiments are also discussed.
Results and Discussion
Among tNSAIDs, other than low-dose Aspirin, only a few exam-
ples of selective COX-1 inhibitors have been investigated in
depth thus far.[1] For this study, we identified novel compounds
Figure 3. Working hypothesis: target molecules in the a) COX-1 and b) COX-2 active sites. COX-1 and COX-2 active sites are schematically depicted as narrow
hydrophobic channels separated from the residual protein, partly consisting of the wide lobby of the membrane-binding domain with a constriction site com-
posed of three amino acids (Arg120, Tyr355, and Glu524). The channel is located at the interface between the region that penetrates the membrane and the
catalytic region. For the COX-1 isoform, Ile523 (blue) closes the hydrophilic side pocket located laterally. In the COX-2 isoform, there is a more accessible
pocket. There is a wider channel opening in COX-2, with narrowing caused by Arg120 and Tyr355 in COX-1.
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1173
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 1. Fragment-based construction of target compounds 7–9 and 14–17, derived from indomethacin, P6-COOH, and mofezolac.
Compound
COX-1-recognized fragment
R1 (linker)
R2 (additive functionality)
7
8
9
14
15
16
17
–
compound 17, linking the COX inhibitory and 6,7-DM-THIQ
moieties, was used to test the role of the linker in this series
and to identify structural features to design novel selective
COX inhibitors.
to target compounds 7–9 (Scheme 2). Rh6G-COOH was ob-
tained by hydrolysis of the commercially available Rh6G (ethyl
ester) in the presence of NaOH.[24]
Target compounds 14–16 were prepared from 5-amino-1-
[6,7-dimethoxy-3,4-dihydroisoquinolin-2-(1H)-yl]pentan-1-one
hydrochloride (13) by reacting the 6,7-DM-THIQ (11) with N-
Boc-protected 5-aminovaleric acid (10),[25] activating in situ
with EDCl/HOBt, followed by deprotection with HCl(g)
(Scheme 3). The subsequent conjugation of 13 with indome-
thacin, P6-COOH, or mofezolac led to target compounds 14–
16 (Scheme 4). Target compound 17, obtained by conjugating
mofezolac and 6,7-DM-THIQ (11), was prepared to verify the
importance of the distance between the COX-1 inhibitor por-
tion and 6,7-DM-THIQ (Scheme 5). The first synthetic condi-
tions (Scheme 6) used to prepare 14 resulted in cleavage of
the 4-chlorobenzoyl group, thus obtaining compound 20,
which was included in the pharmacological characterization
(Table 2).
Chemistry
Syntheses of 1–3 were accomplished by an amide bond forma-
tion between the corresponding one of the three COX inhibi-
tors and a diamine, which was used as a linker. Mono-N-Boc
protection of 1,4-diaminobutane afforded the corresponding
tert-butyl (4-aminobutyl)carbamate[23] (Scheme 1), which in
turn reacted with in situ EDCl/HOBt-activated indomethacin,
P6-COOH, and mofezolac (Scheme 1) to give mono-N-Boc ad-
ducts 1–3.
The N-Boc deprotection of 1–3, precursors of 4–6, was per-
formed in high yield (92–98%) by HCl(g). The subsequent con-
jugation of 4–6 with the rhodamine moiety (Rh6G-COOH) led
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1174
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Scheme 1. Reagents and conditions: a) di-tert-butyl dicarbonate, dioxane, RT; b) DIEA, EDCl, HOBt, dry DMF, RT, overnight; c) HCl(g), dry CH2Cl2, RT, 1.5 h.
Scheme 2. Reagents and conditions: a) NaOH(aq), EtOH, reflux, overnight; b) PyBOP, TEA, CH2Cl2/DMSO, RT, 24 h.
Biology
Indomethacin selectively inhibited COX-1 with SI=0.02
(Table 2). Rh6G-COOH did not inhibit COXs activity at 50 mm. A
combination of indomethacin and Rh6G-COOH, with a 1,4-dia-
minobutane linker, afforded 7, which was not able to inhibit
COX-1, although it decreased COX-2 activity at 50 mm by 28%.
On the other hand, 8 was found to be highly COX-1-selective,
with IC50 =6.0 mm and a total lack of efficacy toward COX-2 at
All synthesized compounds were tested to determine their in-
hibitory activity toward the COX-1 and COX-2 enzymes, respec-
tively (Table 2), using a colorimetric COX (ovine) inhibitor
screening assay kit. COX is a bifunctional enzyme that exhibits
both cyclooxygenase and peroxidase activity.
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1175
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Scheme 3. Reagents and conditions: a) di-tert-butyl dicarbonate, dioxane, RT; b) 6,7-DM-THIQ (11), DIPEA, EDCl, HOBt, dry DMF, RT, 20 h; c) HCl(g), dry CH2Cl2,
RT, 1.5 h.
Scheme 4. Reagents and conditions: a) DIPEA, EDCl, HOBt, dry DMF, RT, 20 h.
Scheme 5. Reagents and conditions: a) DIEA, EDCl, HOBt, dry DMF, RT, overnight.
concentrations up to 50 mm, although it was found that conju-
gation of the Rh6G moiety with P6 to afford compound 8 re-
sults in a loss of fluorescence (data not shown). Compound 9
was found to be a poor inhibitor, with activity only toward
COX-1 (16% inhibition at 50 mm).
Within the series of 6,7-DM-THIQ derivatives, indomethacin
derivative 14 was completely inactive toward both COX iso-
forms, and its analogue (20) showed only 10% inhibition of
COX-1. Compound 15, derived from P6-COOH, slightly affected
only COX-1 activity (23% inhibition at 50 mm). Mofezolac deriv-
ative 16, at a concentration of 50 mm, showed 33 and 13% in-
hibition of COX-1 and COX-2, respectively. The common struc-
tural element in all of the target compounds was the approxi-
mate distance between the COX and P-gp binding fragments.
This was chosen based on the results obtained from building
fluorescent imaging agents for the selective visualization of
COX-2 in inflammation and cancer.[21,22] Such a linker length
proved to be suitable to allow 8 to have the P6 moiety close
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1176
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Scheme 6. Reagents and conditions: a) SOCl2, 08C; b) anhydrous CH3OH, reflux; c) indomethacin, HOBt, DCC, DIEA/anhydrous CH2Cl2; d) LiOH, THF, RT; e) DCC,
DMAP (08C, 30 min)/anhydrous CH2Cl2; f) 6,7-DM-THIQ (11), 16 h, RT.
to the Tyr385 residue in the COX-1 catalytic site and the bulky
Rh6G moiety in proximity of the entrance of the long hydro-
phobic channel of the COX membrane-binding domain
(Figure 4). In the absence of a linker, as with compound 17,
48% and 69% inhibition of COX-1 and COX-2 activity, respec-
tively, were observed at a concentration of 50 mm (COX-2
IC50 =0.95 mm).
Ser530 side chains were allowed to move during the docking
experiments.
An initial validation of the docking protocol was performed
by comparing the conformation, position, and orientation
(pose) of the inhibitor indomethacin-(R)-a-ethyl-ethanolamide,
as obtained from docking with the one experimentally deter-
mined X-ray crystal structure. Correctly re-docking the crystal-
lographically observed inhibitor was a minimum requirement
to determine whether the program was applicable to this
system. The top ligand conformation predicted by the GOLD
program was very close to the crystal structure-bound confor-
mation. The root-mean-square deviation (RMSD) between the
docked pose and its bound conformation in the crystal struc-
ture was 0.63 Š, indicating that GOLD was able to reproduce
the correct pose.
Docking studies
The COX active site of both COX-1 and COX-2 is located at the
terminus of a long hydrophobic channel that runs from the
protein surface to the interior of the protein (Figure 3). The ini-
tial portion of the channel has a large volume, or lobby, which
narrows at a constriction site composed of residues Arg120,
Tyr355, and Glu524. The constriction site constitutes a gate
that must open and close for substrates and inhibitors to pass
into or out of the above-located COX active site. All COX inhib-
itors bind in the active site above the constriction site, and the
constriction site residues play an important role in binding car-
boxylic acid-containing inhibitors through a combination of
ion pairing and hydrogen bonding.[26]
As illustrated in Figure 4a, the P6 moiety of 8 fully inserts
into the binding pocket of COX-1, while the bulky secondary
amide Rh6G functionality projects through the constriction site
at the base of the active site and into the wide lobby in the
membrane-binding domain. The position and orientation of
the P6 moiety in the COX-1 active site closely resembles bind-
ing mode B previously obtained for the COX-1-selective isoxa-
zole 3-(5-bromofuran-2-yl)-5-methyl-4-phenylisoxazole (Br-P6).[9]
In particular, the N2 isoxazole nitrogen of 8 is within hydrogen
bonding distance from the g-OH group of Ser530, which
adopted a down position during the flexible docking simula-
tion.[3] The N2···O atoms are separated by 3.2 Š and display a fa-
vorable geometry for hydrogen bonding. The phenyl ring at
position 4 of the isoxazole is in close contact with various hy-
drophobic side chains, including Leu352, Ile523, Tyr355, and
Phe518. The chlorofuryl moiety is oriented toward the apex of
the COX-1 active site and forms van der Waals interactions
with residues Met522, Phe518, Leu352, Trp387, Tyr385,
Leu384, and Phe381.
To gain more insight into the probable nature of the binding
interactions of compound 8 within the active site of COX-1,
docking studies were conducted using the ovine COX-1 in
complex with bound inhibitor indomethacin-(R)-a-ethyl-etha-
nolamide (PDB ID: 2OYE).[27] This crystal structure was selected,
because the side chain of Arg120 adopts a significantly altered
conformation to accommodate the inhibitor, relative to the
more extended conformation observed in most other COX-
1 crystal structures. One consequence of this structural read-
justment is the disruption of the Arg120–Tyr355–Glu524 con-
striction site and potential expansion of the entrance to the
main hydrophobic channel.
Docking investigations were carried out using the
GOLD 5.2.2 automated docking program[28,29] in combination
with the ChemPLP scoring function, followed by rescoring with
ChemScore.[3,19,30–34] Alternative conformations of the Arg120,
Tyr355, Glu524, and Ser530 side chains have been previously
observed in COX crystal structures complexed with several in-
hibitors, highlighting the innate plasticity of the active
site.[26,35,36] Accordingly, the Arg120, Tyr355, Glu524, and
The amide bond of the linker conjugating the P6 moiety
makes two hydrogen bonds, one with the OH group of Tyr355
(dN···O =2.9 Š) and the other with the guanidinium moiety of
Arg120 (dO···N =2.7 Š), which is known to play an important
role in binding substrates and carboxylic-acid-containing inhib-
itors. The linker is also engaged in hydrophobic interactions
with Ile89, Leu93, Val116, Val349, Leu359, and Leu531. In ad-
dition, the Rh6G portion interacts with multiple hydrophobic
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1177
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 2. COX-inhibitory activity of target compounds 7–9, 14–17, and 20.
Compound
Inhibition [%][a] (IC50 [mm])[b]
COX-1
COX-2
indomethacin
Rh6G-COOH
81 (0.1Æ0.012)
86 (4.2Æ0.014)
n.a.
n.a.
7
n.a.
28
8
9
69 (6.0Æ0.7)
n.a.
n.a.
16
14
n.a.
n.a.
15
16
23
33
n.a.
13
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1178
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 2. (Continued)
Compound
Inhibition [%][a] (IC50 [mm])[b]
COX-1
COX-2
17
20
48
10
69 (0.95Æ0.08)
n.a.
[a] Inhibitory activity toward COX-1 and COX-2 determined at compound concentrations of 50 mm using a colorimetric COX (ovine) inhibitor screening
assay kit. [b] Values are the meanÆSEM of three separate experiments; n.a.: not active, i.e., no inhibitory activity was observed at 50 mm.
Figure 4. Binding modes of compound 8 (a, yellow) and compound 17 (c, violet) in the COX-1 and COX-2 active sites, respectively. Two-dimensional projec-
tion of b) 8–COX-1 and d) 17–COX-2 interactions generated with MOE.[42] Only residues located within 4.5 Š of the bound ligand are displayed and labeled.
Hydrogen bonds discussed in the text are depicted as dashed black lines. The constriction site residues (Arg120, Tyr355, and Glu524), which separate the
lobby from the COX active site located above it, are underlined and colored red.
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1179
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
residues, including Ile89, Pro84, Pro86, Leu115, Val119,
Leu123, and Gly471, which provide considerable binding
strength. The O2 carboxylate oxygen of Glu524 is 2.5 Š from
the positive charge on the ethyl ammonium nitrogen of the
Rh6G moiety, thereby neutralizing both buried charges. The
Rh6G nucleus is further anchored to the lobby region of COX-
1 through a favorable cation–p interaction between its benzo-
fused ring and Arg79, as well as through an associated hydro-
gen bond also formed by the same amino acid residue with
the central ring O1 atom of Rh6G.
a hydrogen bond with the guanidinium group of Arg120
(dNH2···O =3.0 Š). Finally, the 6,7-DM-THIQ moiety extends into
the lobby region of the enzyme and makes hydrophobic con-
tacts with residues Val89, Leu93, Ile112, Tyr115, Val116,
Tyr355, and Leu359.
Comparison of the 8/COX-1 and 17/COX-2 complexes re-
veals striking similarities and differences in the binding of the
ligands (Figure 5). Interestingly, the two inhibitors are bound in
very similar conformations, with their N2 isoxazole nitrogen hy-
drogen bonding to Ser530 into both isoenzymes. However, 17
Notably, in the COX-1 structure with 8 bound, although the
side chain of Arg120 reorients to form a hydrogen bond with
the ligand, it is in position to preserve the salt bridge with
Glu524. In this conformation, Arg120 no longer interacts with
the OH group of Tyr355, which in turn interacts with the
ligand, leaving the bottom of the COX-1 binding site in an
open conformation. Our predicted structural readjustment of
Arg120 is in agreement with that observed in the crystal struc-
ture of indomethacin-(R)-a-ethyl-ethanolamide complexed
with COX-1, in which this residue adopts a significantly altered
conformation to accommodate the ligand in comparison with
crystal structures of different COX-1-ligand complexes (includ-
ing NSAIDs and fatty acid substrates).[27]
To rationalize the high inhibitory activity of compound 17
toward COX-2, docking experiments with the COX-2 binding
site (PDB ID: 6COX)[26] were also carried out. The main differ-
ence between the two COX active sites is the replacement of
two isoleucine residues (Ile434 and Ile523) in COX-1 by two
less bulky amino acids (Val434 and Val523) in COX-2. This
double substitution opens a polar side pocket, enlarging the
volume of the COX-2 active site. Additionally, the substitution
of His513 in COX-1 by Arg513 in COX-2 allows for hydrogen
bonding with the sulfonyl moiety of COX-2 inhibitors. A
second difference, at the top of the channel, is the substitution
of Phe503 in COX-1 by Leu503 in COX-2. The smaller residue
in COX-2 allows Leu384 to create an extra space at the top of
the binding site in COX-2, thus permitting larger inhibitors to
bind.
Figure 5. Superimposition of the best poses of 8 (yellow) and 17 (violet)
into the COX-1 and COX-2 active sites, respectively.
makes two hydrogen bonds with the binding cleft of COX-2:
one between the methoxy oxygen on the phenyl at C3 of the
isoxazole and the Tyr385 hydroxy group and the other be-
tween the methoxy oxygen on the phenyl at C4 of the isoxa-
zole and the nitrogen atom of the His90 imidazole ring. Ac-
cording to previous studies, Tyr385 is involved in the abstrac-
tion of 13-pro-S-hydrogen from arachidonic acid,[37] whereas
His90 is considered to control the access of ligands to the ad-
junct pocket, which is created by the replacement of Ile523 in
COX-1 by valine in the COX-2 active site and is responsible for
selectivity.[38] These hydrogen bonding interactions contribute
to stabilize the binding of 17 to the COX-2 active site. More-
over, 17 shows only partial entry into the COX-1 enzyme active
site, owing to the more restricted secondary internal pocket,
and therefore does not exhibit significant interactions with the
COX-1 active site residues. This is in agreement with the pre-
dominant COX-2 inhibitory potency and selectivity of com-
pound 17. On the other hand, the selectivity of compound 8
for COX-1 is likely to be due to the favorable interactions of
the Rh6G moiety with the residues of both the constriction
and lobby sites, contributing significantly to the complex sta-
bility.
GOLD successfully predicted a plausible binding pose for 17
within the binding pocket of COX-2. As illustrated in Figure 4c,
the mofezolac functionality of the ligand fully inserts into the
binding pocket of COX-2, while the bulky 6,7-DM-THIQ moiety
projects through the constriction site at the base of the active
site and into the wide lobby in the membrane-binding
domain. In particular, the N2 isoxazole nitrogen of 17 is engag-
ed in a hydrogen bond with the OH group of Ser530 (dN2···O
=
2.7 Š). The p-methoxyphenyl ring at position 3 of the isoxazole
ring is directed toward the top of the COX-2 active site and
forms hydrophobic interactions with residues Leu352, Phe381,
Leu384, Tyr385, Trp387, Phe518, Met522, and Gly526. The p-
methoxy group accepts a hydrogen bond from the Tyr385 OH
group (dO···O =3.3 Š). Importantly, the p-methoxyphenyl ring at
position 4 projects toward the side pocket of COX-2 and
makes hydrophobic contacts with residues His90, Tyr355,
Ala516, Phe518, and Val523. The p-methoxy group is within
hydrogen bonding distance of the nitrogen atom of the His90
residue (dO···Ne2 =3.2 Š). In addition, the C=O amide of 17 forms
Target compounds 7–9 and 15–17, each containing a P-gp-
interacting fragment, were also studied for their ability to
modulate efflux pump activity by measuring calcein accumula-
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1180
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
recorded on a 300 MHz Varian Mercury-VX spectrometer, and
chemical shifts (d) are reported in parts per million. Absolute
values of the coupling constants are reported. FT-IR spectra were
recorded on a PerkinElmer 681 spectrometer. Thin-layer chroma-
tography (TLC) was performed on silica gel sheets with fluorescent
indicator, the spots on the TLC were observed under ultraviolet
light (l=254 or 315 nm). Column chromatography was performed
using silica gel 60 with a particle size distribution of 40-63 mm and
230-400 ASTM. GC–MS analyses were performed on an HP 5995C
instrument. ESI-MS analyses were performed on an Agilent 1100
LC–MSD trap system VL. All synthesized compounds were analyzed
by HPLC, performed on an Agilent 1260 Infinity instrument
equipped with a 1260 DADVL+ detector, and their purity was
found to be higher than 95%. Indomethacin, Rh6G, 6,7-DM-THIQ,
and other reagents and solvents were purchased from Sigma–Al-
drich.
tion in MDCK-MDR1 cells overexpressing P-gp and by deter-
mining apparent permeability through a Caco-2 cell monolay-
er. Compounds 7 and 9, despite bearing the Rh6G moiety, did
not interact with the efflux transporter, as demonstrated by
their low EC50 values (Table 3). EC50 values were determined by
fitting the percent fluorescence increase versus log[dose].[39] All
of the other compounds (8 and 15–17) showed a significant
interaction with P-gp (EC50 values ranging from 0.42 to
4.61 mm). The permeability assay performed on the same com-
pounds (Table 3) provided BA/AB (Papp, apparent permeability)
values> 2, indicating that they are P-gp substrates.[40]
Table 3. Calcein-AM assay results and Papp determination.
[b]
Compound
EC50 [mm][a]
Papp [nmsÀ1
]
BA/AB
tert-Butyl (4-aminobutyl)carbamate (Scheme 1):[23] Boc2O (4.49 g,
0.0196 mol) in dioxane (80 mL) was added dropwise to a stirred so-
lution of 1,4-diaminobutane (10.1 g, 0.115 mol) in dioxane (80 mL)
at room temperature. The reaction mixture was stirred overnight
at room temperature. The solvent was then evaporated under re-
duced pressure. H2O (100 mL) was added, and the aqueous sus-
pension was extracted with CH2Cl2 (5”30 mL). The combined or-
ganic layers were dried over anhydrous sodium sulfate, and the
solvent was evaporated under reduced pressure. The product was
isolated as a yellow oil (3.61 g, 17% yield): 1H NMR (400 MHz,
CDCl3): d=5.14 (bs, 1H, NH: exch. with D2O), 2.94 (q, 2H, J=
6.4 Hz, CH2NH), 2.53 (t, 2H, J=6.4 Hz, CH2NH2), 1.36–1.26 (m, 13H,
CH2(CH2)2CH2 and C(CH3)3), 1.07 ppm (s, 2H, NH2: exchange with
D2O); FT-IR (neat): n˜ =3359, 2976, 2933, 2866, 1694, 1528, 1453,
BA
AB
n.t.
292
n.t.
321
613
442
484
7
8
9
15
16
17
24Æ4.6
4.61Æ0.7
n.a.
3.78Æ0.5
2.80Æ0.1
0.42Æ0.03
20
n.t.
n.t.
6.1
n.t.
3.9
3.7
4.0
2.3
1802
n.t.
1283
2293
1789
1130
rhodamine 6G
[a] Values are the meanÆSEM of three separate experiments with MDCK-
MDR1; n.a.: not active. [b] n.t.: not tested.
Conclusions
1391, 1365, 1276, 1252, 1174, 1042, 989, 867, 781 cmÀ1
.
In this study, a set of new molecules were prepared to enable
knowledge advancement of the COX active site molecular rec-
ognition, complemented by docking simulation experiments
that added insight into the binding mode of these com-
pounds. Two compounds were identified as highly selective
COX-1 and COX-2 inhibitors, respectively (8, COX-1 IC50 =
6.0 mm; 17, COX-2 IC50 =0.95 mm). Molecular docking studies
rationalized this strong COX-1 and COX-2 inhibitory activity.
The diarylisoxazole moiety (P6) of 8 fully inserts into the bind-
ing pocket of COX-1, whereas the bulky Rh6G functionality
projects through the constriction site into the wide lobby in
the membrane-binding domain. The Rh6G portion interacts
with multiple hydrophobic residues, providing considerable
binding strength. On the other hand, the selective COX-2 in-
hibitor 17 breaches the constriction site at the mouth of the
active site and projects its 6,7-DM-THIQ moiety into the steri-
cally uncongested lobby region. Rh6G and 6,7-DM-THIQ, pres-
ent in 8 and 17, respectively, were expected to confer opposite
P-gp activity modulation to the two compounds. On the con-
trary, 8 and 17 retained COX-1 or COX-2 inhibitory activity, re-
spectively, but both behaved as P-gp substrates (BA/AB>2,
Table 3). All of these results have been taken into account for
further studies.
N-[9-(2-Carboxyphenyl)-6-(ethylamino)-2,7-dimethyl-3H-xanthen-
3-ylidene]ethanaminium (Rh6G-COOH) (Scheme 2):[24] A solution
of NaOH (75.6 mg, 1.89 mmol) in H2O (60 mL) was added to
a stirred solution of Rh6G (500 mg, 1.04 mmol) in absolute EtOH
(50 mL) at room temperature. The reaction mixture was stirred at
reflux overnight. The EtOH was then evaporated under reduced
pressure, and 37% HCl was added until an acidic pH was achieved.
The product was isolated after filtration as a pink solid (258 mg,
1
55% yield); mp: >3008C. H NMR (300 MHz, [D6]acetone): d=8.41
(bs, 1H, NH+: exch. with D2O), 7.91 (d, 1H, J=7.6 Hz, arom.), 7.75–
7.59 (m, 2H, arom.), 7.17 (d, 1H, J=7.6 Hz, arom.), 6.36 (s, 2H,
arom.), 6.33 (s, 2H, arom.), 4.67 (bs, 1H, NH: exch. with D2O), 3.24
(q, 4H, J=7.2 Hz, 2 NCH2CH3), 2.90 (bs, 1H, COOH: exch. with D2O),
1.93 (s, 6H, 2 ArCH3), 1.27 ppm (t, 6H, J=7.2 Hz, 2NCH2CH3); FT-IR
(KBr): n˜ =3415, 2924, 1608, 1530, 1501, 1365, 1307, 1187, 1016,
793 cmÀ1; ESI-MS (m/z): C26H27N2O3, 415 [M]+, 437 [M+Na]+.
tert-Butyl-(4-{2-[1-(4-chlorobenzoyl)-5-methoxy-2-methyl-1H-
indol-3-yl]acetamido}butyl)carbamate (1): tert-Butyl (4-aminobu-
tyl)carbamate (320 mg, 1.7 mmol), HOBt (151 mg, 0.81 mmol), DIEA
(225 mg, 1.7 mmol), and EDC·HCl (153 mg, 0.798 mmol) were
added to nitrogen-flushed, three-necked flask equipped with
a magnetic stirrer, nitrogen inlet, and two dropping funnels, con-
taining a solution of indomethacin (210 mg, 0.588 mmol) solubi-
lized in DMF (14 mL). The mixture was stirred at room temperature
for 48 h. Then, H2O (20 mL) was added, and the aqueous solution
was extracted with EtOAc (3”30 mL). The combined organic layers
were washed with brine and dried over anhydrous sodium sulfate,
and the solvent was evaporated under reduced pressure. Column
chromatography of the crude residue on silica gel using CHCl3/
MeOH (95:5) as a mobile phase afforded the target compound as
Experimental Section
Chemistry
General methods: Melting points were taken on an electrothermal
apparatus and are uncorrected. H NMR and 13C NMR spectra were
1
1
a yellow solid (140 mg, 45% yield). H NMR (300 MHz, CDCl3): d=
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1181
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
7.67 (d, 2H, J=8.4 Hz, arom.), 7.49 (d, 2H, J=8.4 Hz, arom.), 6.90–
6.85 (m, 2H, arom.), 6.71–6.68 (m, 1H, arom.), 5.78 (bs, 1H,
CONHCH2: exch. with D2O), 4.58 (bs, 1H, NHCOO: exch. with D2O),
3.82 (s, 3H, OCH3), 3.63 (s, 2H, ArCH2CONH), 3.25–3.18 (m, 2H,
NHCH2(CH2)3), 3.15–2.98 (m, 2H, (CH2)3CH2NH), 2.39 (s, 3H, ArCH3),
1.50–1.37 ppm [m, 13H, CH2CH2 and C(CH3)3].
(4-{2-[3-(5-chlorofuran-2-yl)-4-phenylisoxazol-5-yl]acetamido}butyl)-
carbamate (2) as a reagent to give a yellow oil (98.8 mg, quantita-
tive yield); 1H NMR (300 MHz, [D6]DMSO): d=8.50–8.40 (m, 1H,
CONHCH2: exch. with D2O), 8.14 (bs, 3H, NH3+: exchange with
D2O), 7.46–7.38 (m, 2H, arom.), 7.36-7.30 (m, 3H, arom.), 6.54 (d,
1H, J=3.5 Hz, furyl), 6.38 (d, 1H, J=3.5 Hz, furyl), 3.61 (s, 2H,
CH2CO), 3.20–2.96 (m, 2H, CH2NH3+), 2.74–2.64 (m, 2H, CONHCH2),
1.54–1.36 ppm (m, 4H, CH2CH2); 13C NMR (75 MHz, [D6]DMSO): d=
166.6, 165.9, 165.4, 152.4, 143.6, 137.6, 130.4, 130.2, 129.4, 129.2,
128.9, 116.8, 114.8, 109.5, 94.6, 70.4, 63.5, 38.9, 33.2, 26.5,
tert-Butyl
(4-{2-[3-(5-chlorofuran-2-yl)-4-phenylisoxazol-5-yl]-
acetamido}butyl)carbamate (2): Compound 2 was prepared using
the same procedure used to synthesize 1, replacing indomethacin
with P6-COOH to give a yellow solid (116.8 mg, 42% yield); mp:
126.5–1278C; 1H NMR (300 MHz, CDCl3): d=7.48–7.44 (m, 3H,
arom.), 7.38-7.35 (m, 2H arom.), 6.30 (d, 1H, J=3.5 Hz, furyl), 6.16
(d, 1H, J=3.5 Hz, furyl), 6.04 (bs, 1H, CONHCH2: exch. with D2O),
4.58 (s, 1H, NHBoc: exch. with D2O), 3.66 (s, 2H, isoxazole-CH2CO),
3.27 (q, 2H, J=6.8 Hz, CONHCH2), 3.18–3.10 (m, 2H, CH2NHCO),
1.54–1.48 (m, 4H, CH2CH2), 1.44 ppm (s, 9H, C(CH3)3); 13C NMR
(75 MHz, CDCl3): d=166.1, 130.2, 129.1, 129.06, 114.4, 108.2, 59.7,
39.9, 38.3, 34.0, 32.1, 31.4, 31.1, 29.9, 29.5, 28.6, 27.8, 22.9,
14.3 ppm; FT-IR (KBr): n˜ =3346, 2941, 1683, 1658, 1539, 1521, 1437,
1366, 1261, 1172, 1019, 941, 892, 787, 699 cmÀ1; ESI-MS (m/z):
C24H28ClN3O5, 496 [M+Na]+.
25.0 ppm; FT-IR (KBr): n˜ =3405, 1633, 1266, 1134, 1017, 793 cmÀ1
;
ESI-MS (m/z): C19H20N3O337Cl, 374 [M+H]+, (37) C19H20N3O335Cl, 374
[M+H]+, (100).
N-(4-Aminobutyl)-2-(3,4-bis(4-methoxyphenyl)isoxazol-5-yl)acet-
amide hydrochloride (6): Dry HCl (obtained by dripping H2SO4
conc. in 37% HCl and NaCl)) was bubbled in a solution of tert-
butyl (4-{2-[3,4-bis(4-methoxyphenyl)isoxazol-5-yl]acetamido}butyl)-
carbamate (3) (170 mg, 0.33 mmol) in dry CH2Cl2 (15 mL). After 2 h,
the reaction mixture became a suspension. The solvent was then
evaporated under vacuum, and the product was isolated as
a brown solid (quantitative yield) and was used without any further
purification. 1H NMR (300 MHz, CDCl3): d=7.35 (d, 2H, J=8.8 Hz,
arom.), 7.15 (d, 2H, J=8.8 Hz, arom.), 6.88 (d, 2H, J=8.8 Hz, arom.),
6.81 (d, 2H, J=8.8 Hz, arom.), 6.80–6.74 (bs, 1H, NH: exch. with
D2O), 3.80 (s, 3H, OCH3), 3.77 (s, 3H, OCH3), 3.64 (s, 2H, ArCH2CON),
3.25 (q, 2H, J=6.1 Hz, NHCH2), 2.71 (t, 2H, J=6.6 Hz, CH2NH2),
2.05-1.95 (bs, 2H, NH2: exchange with D2O), 1.57–1.43 ppm (m, 4H,
CH2CH2); 13C NMR (75 MHz, CDCl3): d=166.9, 163.2, 161.3, 160.8,
159.6, 131.30, 131.27, 130.0, 129.9, 121.6, 121.2, 117.7, 114.5, 114.2,
55.5, 55.4, 41.7, 40.0, 34.3, 30.8, 27.1 ppm; ESI-MS (m/z): C23H27N3O4,
410 [M+H]+.
tert-Butyl (4-{2-[3,4-bis(4-methoxyphenyl)isoxazol-5-yl]acetami-
do}butyl)carbamate (3). tert-Butyl (4-aminobutyl)carbamate
(478 mg, 2.54 mmol), HOBt (185 mg, 1.21 mmol), dry DIEA (328 mg,
2.54 mmol), and EDC·HCl (229 mg, 1.19 mmol) were added to a so-
lution of mofezolac (300 mg, 0.88 mmol) solubilized in dry DMF
(15 mL) in a nitrogen-flushed, three-necked flask equipped with
a magnetic stirrer, nitrogen inlet, and two dropping funnels. The
mixture was stirred at room temperature for 48 h. Then, H2O
(20 mL) was added, and the aqueous solution was extracted with
EtOAc (3”30 mL). The combined organic layers were washed with
brine and dried over anhydrous sodium sulfate, and the solvent
was evaporated under vacuum. The product was isolated as
a yellow solid (184 mg, 41% yield) by column chromatography
N-(9-(2-((4-(2-(1-(4-Chlorobenzoyl)-5-methoxy-2-methylindolin-3-
yl)acetamido)butyl)carbamoyl)phenyl-6-(ethylamino)-2,7-di-
methyl-3H-xanthen-3-ylidene)ethanaminium chloride (7): Bben-
zotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate
(PyBOP; 20.2 mg, 0.039 mmol) and triethylamine (50.3 mg, 69 mL,
0.498 mmol) were added to a stirred solution of Rh6G-COOH (N-[9-
(2-carboxyphenyl)-6-(ethylamino)-2,7-dimethyl-3H-xanthen-3-ylide-
ne]ethanaminium chloride) (18 mg, 0.037 mmol) in CH2Cl2/DMSO
(3:1, 4 mL). After 2 h, a solution of 4-{2-[1-(4-chlorobenzoyl)-5-me-
thoxy-2-methyl-1H-indol-3-yl]acetamido}butan-1-aminium chloride
(4) in CH2Cl2/DMSO (3:2) (5 mL) was added dropwise, and the mix-
ture was stirred for 24 h at room temperature. The solvent was
then removed under reduced pressure. Column chromatography
of the crude residue on silica gel using hexane/EtOAc (1:1) as the
mobile phase afforded the target compound as a yellow semi-solid
(15 mg, 47% yield): 1H NMR (300 MHz, CDCl3): d=8.52–8.30 (bs,
3H, 2NHCO+NH+, exchange with D2O), 7.87–7.85 (m, 1H, arom.),
7.69–7.67 (m, 2H, arom.), 7.49–7.46 (m, 2H, arom.), 7.45–7.43 (m,
1H, arom.), 7.05-7.02 (m, 1H, arom.), 6.92 (d, 1H, J=2.4 Hz, arom.),
6.87 (d, 1H, J=9.2 Hz, arom.), 6.66 (dd, 1H, J=9.2 and 2.4 Hz,
arom.), 6.34 (s, 2H, arom.), 6.25-6.23 (bs, 1H, NH: exch. with D2O),
6.18 (s, 2H, arom.), 3.78 (s, 3H, OCH3), 3.60 (s, 2H, ArCH2CO), 3.21
(q, 4H, J=7.2 Hz, 2 NHCH2CH3), 3.05 (m, 4H, CH2(CH2)2CH2), 2.38 (s,
3H, CH3), 1.88 (s, 6H, 2 CH3), 1.32 (t, 6H, J=7.2 Hz, NHCH2CH3),
1.29–1.20 ppm (m, 4H, CH2(CH2)2CH2); 13C NMR (100 MHz, CDCl3):
d=170.1, 168.5, 156.4, 151.9, 147.6, 139.5, 136.4, 134.1, 132.6,
131.4, 131.1, 130.8, 129.3, 128.63, 128.2, 124.0, 122.8, 118.0, 115.2,
113.5, 112.3, 110.9, 106.2, 101.2, 96.7, 94.6, 65.2, 60.6, 55.9, 39.5,
39.4, 38.6, 32.4, 31.4, 29.9, 26.1, 25.9, 21.2, 16.8, 14.9, 14.4,
13.6 ppm; FT-IR (neat): n˜ =3429, 3330, 3051, 2963, 2923, 2854,
1675, 1621, 1517, 1468, 1322, 1262, 1218, 1149, 1090, 1016, 925,
1
(silica gel, CHCl3/MeOH=95:5). H NMR (300 MHz, CDCl3): d=7.38
(d, 2H, J=8.9 Hz, arom.), 7.16 (d, 2H, J=8.8 Hz, arom.), 6.90 (d, 2H,
J=8.8 Hz, arom.), 6.84 (d, 2H, J=8.9 Hz, arom.), 6.09 (bs, 1H, NH:
exch. with D2O), 4.60 (bs, 1H, NH: exch. with D2O), 3.83 (s, 3H,
OCH3), 3.80 (s, 3H, OCH3), 3.68 (s, 2H, ArCH2CON), 3.28 (q, 2H J=
6.3 Hz, NHCH2), 3.15–3.05 (m, 2H, CH2NH), 1.54–1.37 ppm [m, 13H,
CH2CH2 and C(CH3)3]; 13C NMR (75 MHz, CDCl3): d=166.8, 162.9,
161.4, 160.8, 159.7, 156.3, 131.3, 130.0, 121.6, 121.2, 117.7, 114.6,
114.2, 79.5, 55.6, 55.4, 40.2, 39.9, 34.4, 29.9, 28.6, 27.8, 27.6,
26.7 ppm; ESI-MS (m/z): C28H35N3O6, 532 [M+Na]+.
4-{2-[1-(4-Chlorobenzoyl)-5-methoxy-2-methyl-1H-indol-3-yl]acet-
amido}butan-1-aminium chloride (4): Dry HCl (obtained by drip-
ping conc. H2SO4 into 37% HCl and NaCl) was bubbled in a solution
of
tert-butyl-(4-{2-[1-(4-chlorobenzoyl)-5-methoxy-2-methyl-1H-
indol-3-yl]acetamido}butyl)carbamate (1; 140 mg, 0.265 mmol) in
dry CH2Cl2 (10 mL) at 08C (ice bath). After 3 h, the solvent of the
brown suspension was evaporated under reduced pressure, and
the product was isolated as a brown solid (118 mg, 96% yield).
1H NMR (400 MHz, [D6]DMSO): d=8.20–8.11 (bs, 1H, NH: exch. with
D2O), 8.00–7.80 (bs, 3H, NH3+: exchange with D2O), 7.66–7.58 (m,
4H, arom.), 7.10 (d, 1H, J=2.2 Hz, arom.), 6.88 (d, 1H, J=9.2 Hz,
arom.), 6.66 (dd, 1H, J=9.2 and 2.2 Hz, arom.), 3.72 (s, 3H, OCH3),
3.46 (s, 2H, ArCH2CON), 3.06–2.97 (m, 2H, CH2NH3+), 2.78-2.66 (m,
2H, NHCH2), 2.20 (s, 3H, ArCH3), 1.60–1.19 ppm (m, 4H, CH2CH2).
N-(4-Aminobutyl)-2-[3-(5-chlorofuran-2-yl)-4-phenylisoxazol-5-
yl]acetamide hydrochloride (5): Compound 5 was prepared fol-
lowing the same procedure used for compound 4 using tert-butyl
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1182
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
879, 801, 743, 693 cmÀ1; ESI-MS (m/z): C49H5037ClN5O5, 824 [M+H]+
(38), C49H5035ClN5O5 824 [M+H]+ (100).
9 mmol) in H2O (25 mL) at 08C. The mixture was stirred overnight
at room temperature. The reaction mixture was then acidified with
3n HCl and extracted with EtOAc. The combined organic layers
were dried over anhydrous sodium sulfate, and the solvent was
evaporated under reduced pressure to afford the crude solid,
which was recrystallized by CHCl3/hexane. The pure compound
was obtained as a white solid (2.1 g, 97%): mp: 53–568C (lit. mp:
45–488C);[25] 1H NMR (300 MHz, CDCl3): d=4.56 (bs 1H, NH: exch.
with D2O), 3.10-3.05 (m, 2H, CH2NH), 2.34 (t, 2H, J=7.2 Hz, CH2CO),
1.66–1.59 (m, 2H, CH2CH2CO), 1.54–1.46 (m, 2H, CH2CH2CH2CO),
1.40 ppm (s, 9H, tBu); 13C NMR (75 MHz, CDCl3): d=179.0, 156.3,
79.5, 40.3, 33.8, 29.6, 28.6, 22.1 ppm; ESI-MS (m/z): C10H19NO4, 240
[M+Na]+.
N-(9-{2-[(4-{2-[3-(5-Chlorofuran-2-yl)-4-phenylisoxazol-5-yl]acet-
amido}butyl)carbamoyl]phenyl-6-(ethylamino)-2,7-dimethyl-3H-
xanthen-3-ylidene}ethanaminium chloride (8): EDC·HCl (30 mg,
0.158 mmol), HOBt (31.5 mg, 0.206 mmol), DIEA (114 mg,
0.735 mmol), and compound 5 (60 mg, 0.147 mmol) were added
to a stirred solution of N-[9-(2-carboxyphenyl)-6-(ethylamino)-2,7-
dimethyl-3H-xanthen-3-ylidene]ethanaminium (77 mg, 0.158 mmol)
in DMF (6 mL). The reaction mixture was stirred overnight at room
temperature. The solvent was then removed under reduced pres-
sure. Column chromatography of the crude residue on silica gel
using hexane/EtOAc (1:1), followed by CHCl3/MeOH (95:5), as the
mobile phase afforded the target compound as a brown oil
(18 mg, 15% yield): 1H NMR (300 MHz, CDCl3): d=7.99–7.97 (bs,
1H, NH+: exch. with D2O), 7.82–7.79 (m, 1H, arom.), 7.43–7.35 (m,
7H, arom.), 7.04–6.96 (m, 1H, arom.), 6.72–6.64 (bs like triplet, 1H,
CONHCH2: exch. with D2O), 6.31 (s, 2H, arom.), 6.22 (d, 1H, J=
2.8 Hz, furanyl), 6.18 (s, 2H, arom.), 6.10 (d, 1H, J=2.8 Hz, furanyl),
3.61 (s, 2H, CH2CO), 3.60–3.40 (bs, 2H, 2 NHCH2CH3: exchange with
D2O), 3.17 (q, 4H, J=7.1 Hz, 2 NHCH2CH3), 3.10–3.04 (m, 4H,
CH2(CH2)2CH2), 1.87 (s, 6H, 2 ArCH3), 1.30–1.22 ppm (m, 10H, 2
NHCH2CH3 and (CH2)2CH2CH2 and CH2CH2(CH2)2); 13C NMR (100 MHz,
CDCl3): d=168.5, 166.0, 163.7, 162.5, 153.6, 152.4, 151.6, 147.4,
143.3, 138.5, 132.4, 131.0, 130.0, 128.8, 128.6, 128.57, 128.47, 128.0,
123.7, 122.6, 117.9, 116.9, 113.9, 107.9, 106.0, 96.5, 65.1, 39.6, 39.2,
38.3, 33.5, 29.7, 25.9, 25.2, 16.6, 14.7 ppm; ESI-MS (m/z):
C45H4437ClN5O5 792 [M+H]+ (39), C45H4435ClN5O5 792 [M+H]+ (100).
tert-Butyl {5-[6,7-dimethoxy-3,4-dihydroisoquinolin-2-(1H)-yl]-5-
oxopentyl}carbamate (12): EDC·HCl (0.42 g, 2.20 mmol), HOBt
(1.26 g, 9.33 mmol), drained DIPEA (0.061 mL, 3.60 mmol), and 5-
[(tert-butoxycarbonyl)amino]pentanoic
acid
(10)
(0.385 g,
1.77 mmol) were added at room temperature to a solution of 6,7-
DM-THIQ (11) (1.3 g, 6.73 mmol) in dry DMF (10 mL). The mixture
was stirred overnight, then NaHCO3 and brine were added, and the
aqueous phase was extracted with EtOAc. The combined organic
layers were dried over anhydrous sodium sulfate, and the solvent
was evaporated under reduced pressure to afford a brown oil.
Column chromatography of the crude reaction mixture on silica
gel using EtOAc/MeOH (9:1) as the mobile phase to give the target
1
compound as a yellow oil (160 mg, 23% yield): H NMR (300 MHz,
CDCl3): d=6.60–6.57 (m, 2H, arom.), 4.70 (s, 1H, NH: exch. with
D2O), 4.63 (s, slightly broad, 1H, ArCHHNCO, conformer signal),
4.52 (s, slightly broad, 1H, ArCHHNCO, conformer signal), 3.84-3.82
(m, 6H, 2 OCH3, conformers signals), 3.79 (t, 1H, J=5.9 Hz,
ArCH2CHHNCO, conformer signal), 3.64 (t, 1H, J=5.9 Hz,
ArCH2CHHNCO, conformer signal), 3.12 (q, 2H, J=7.0 Hz, NHCH2),
2.79 (t, 1H, J=5.9 Hz, ArCHHCH2NCO, conformer signal), 2.73 (t,
1H, J=5.9 Hz, ArCHHCH2NCO, conformer signal), 2.40 (t, 2H, J=
7.0 Hz, (CH2)3CH2CON), 1.73–1.63 (m, 2H, CH2CH2CO), 1.57–1.48 (m,
2H, (CH2)2CH2CH2NH), 1.41 ppm (s, 9H, tBu); 13C NMR (75 MHz,
CDCl3): d=171.9, 156.3, 148.1, 148.0, 125.9, 125.6, 111.4, 109.6,
79.2, 56.1, 47.2, 44.1, 40.3, 33.8, 29.9, 28.6, 28.2, 22.4 ppm; FT-IR
(neat): n˜ =3348, 2936, 1694, 1518, 1453, 1366, 1272, 1166, 1116,
1037, 754 cmÀ1; ESI-MS (m/z): C21H32N2O5, 415 [M+Na]+.
N-(9-{2-[(4-{2-[3,4-Bis(4-methoxyphenyl)isoxazol-5-yl]acetamido}-
butyl)carbamoyl]phenyl}-6-(ethylamino)-2,7-dimethyl-3H-xanth-
en-3-ylidene)ethanaminium chloride N-(4-{2-[3,4-bis(4-methoxy-
phenyl)isoxazol-5-yl]acetamido}butyl)-2-[6-(ethylamino)-3-(ethyl-
imino)-2,7-dimethyl-3H-xanthen-9-yl]benzamide
(9):
PyBOP
(338 mg, 0.650 mmol) and triethylamine (379 mg, 3.75 mmol,
0.052 mL) were added to a stirred solution of (Z)-2-(6-(ethylamino)-
3-(ethylimino)-2,7-dimethyl-3H-xanthen-9-yl)benzoic acid (146 mg,
0.03 mmol) in CH2Cl2/DMSO (3:1, 22 mL) at 08C (ice bath). After 2 h
stirring at 08C, a solution of compound 6 (133 mg, 0.3 mmol) in
CH2Cl2/DMSO (3:2, 22 mL) was added dropwise, and the reaction
mixture was stirred for 24 h at room temperature. The solvent was
then removed under reduced pressure. Column chromatography
of the crude residue on silica gel using hexane/EtOAc (1:1) as the
mobile phase afforded the target compound as a brown oil
(21.7 mg, 90% yield): 1H NMR (300 MHz, CDCl3): d=7.90–7.84 (m,
1H, arom.), 7.46–7.41 (m, 2H, arom.), 7.39–7.32 (m, 2H, arom.),
7.22–7.16 (m, 2H, arom.), 7.07–7.01 (m, 1H, arom.), 6.90–6.80 (m,
3H, arom.), 6.45–6.30 (bs, 2H, 2 NH: exchange with D2O), 6.34 (s,
2H, arom.), 6.21 (s, 2H, arom.), 3.80 (s, 3H, OCH3), 3.79 (s, 3H,
OCH3), 3.62 (s, 2H, ArCH2NCO), 3.52 (bs, 2H, 2 NHCH2CH3: exchange
with D2O), 3.23–3.05 (m, 8H, 2 NHCH2CH3 and CH2(CH2)2CH2), 1.89
(s, 6H, 2 ArCH3), 1.37–1.24 [m, 8H, 2NHCH2CH3 and (CH2)2CH2CH2],
1.14–1.05 ppm (m, 2H, CH2CH2(CH2)2); 13C NMR (75 MHz, CDCl3):
d=168.7, 166.8, 163.2, 161.2, 160.7, 159.6, 153.8, 151.9, 147.6,
132.7, 131.4, 130.0, 128.8, 128.7, 128.3, 124.0, 122.9, 121.9, 121.5,
118.1, 117.7, 114.5, 114.1, 106.3, 96.8, 96.7, 65.3, 55.5, 55.4, 42.9,
39.9, 39.6, 34.2, 26.0, 25.8, 16.9, 14.9 ppm; FT-IR (neat): n˜ =3429,
3330, 3051, 2963, 2923, 2854, 1675, 1621, 1517, 1468, 1322, 1262,
1218, 1149, 1090, 1016, 925, 879, 801, 743, 693 cmÀ1; ESI-MS (m/z):
C49H51N5O6, 828 [M+Na]+.
5-Amino-1-[6,7-dimethoxy-3,4-dihydroisoquinolin-2-(1H)-yl]pen-
tan-1-one hydrochloride (13): Dry HCl (obtained by dripping
H2SO4 conc. into 37% HCl and NaCl) was bubbled into a solution
of 12 (600 mg, 1.53 mmol) in dry CH2Cl2 (10 mL). The solvent was
evaporated under reduced pressure to afford the crude solid,
which was recrystallized by CHCl3/hexane. The pure compound
was obtained as a yellow solid (451 mg; 90% yield). 1H NMR
(300 MHz, CDCl3): d=8.15–8.07 (bs, 3H, NH3+: exchange with D2O),
6.76 (slightly broad singlet, 1H, arom.), 6.71 (slightly broad singlet,
1H, arom.), 4.53 (slightly broad singlet, 1H, ArCHHNCO, conformer
signal), 4.48 (slightly broad singlet, 1H, ArCHHNCO, conformer
signal), 3.70–3.68 (m, 6H, 2 OCH3, conformer signal), 3.65–3.60 (m,
+
2H, ArCH2CH2NCO), 2.82–2.68 (m, 3H, CH2NH3 and ArCH-
HCH2NCO), 2.65–2.60 (m, 1H, ArCHHCH2NCO), 2.49–2.38 (m, 2H,
CH2CON), 1.65–1.48 ppm (m, 4H, (CH2)2CH2NH3+); FT-IR (KBr): n˜ =
3405, 2938, 2254, 1618, 1518, 1466, 1369, 1257, 1216, 1113, 912,
738, 669 cmÀ1
.
2-[1-(4-Chlorobenzoyl)-5-methoxy-2-methyl-1H-indol-3-yl]-N-{5-
[6,7-dimethoxy-3,4-dihydroisoquinolin-2-(1H)-yl]-5-oxopentyl}a-
cetamide (14): Compound 13 (164 mg, 0.5 mmol), HOBt (104 mg,
0.77 mmol), EDC·HCl (112 mg, 0.58 mmol), and dry DIPEA (0.17 mL,
0.97 mmol) were added to a stirred solution of 2-[1-(4-chloroben-
5-[(tert-Butoxycarbonyl)amino]pentanoic acid (10): NaHCO3
(1.44 g, 0.017 mmol) and Boc2O (2.8 g, 0.013 mol) in dioxane
(10 mL) were added to a solution of 5-aminovaleric acid (1.0 g,
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1183
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
zoyl)-5-methoxy-2-methylindolin-3-yl]acetic acid (indomethacin;
300 mg, 0.84 mmol) in dry DMF (10 mL) under argon atmosphere
at room temperature. The resulting yellow reaction mixture was
stirred overnight. Then, saturated aq. NaHCO3 (100 mL) was added,
and the aqueous solution was extracted with EtOAc (3”30 mL).
The combined organic layers were washed with brine and dried
over anhydrous sodium sulfate, and the solvent was evaporated
under reduced pressure. Column chromatography of the crude res-
idue on silica gel using CHCl3/MeOH (15:1) as the mobile phase af-
1551, 1518, 1435, 1257, 1115, 941, 773, 701 cmÀ1; ESI-MS (m/z):
C31H3237ClN3O6, 600 [M+H]+ (30), C31H3235ClN3O6 600 [M+H]+ (100).
2-[3,4-Bis(4-methoxyphenyl)isoxazol-5-yl]-N-{5-[6,7-dimethoxy-
3,4-dihydroisoquinolin-2-(1H)-yl]-5-oxopentyl}acetamide
(16):
Compound 13 (200 mg, 0.61 mmol), HOBt (150 mg, 0.98 mmol),
EDC·HCl (140 mg, 0.730 mmol), and drained DIPEA (134 mg,
1.04 mmol) were added to a solution of 2-[3,4-bis(4-methoxyphe-
nyl)isoxazol-5-yl]acetic acid (mofezolac; 340 mg, 1.04 mmol) in dry
DMF (10 mL) at room temperature. The reaction mixture was
stirred overnight, then saturated aq. NaHCO3 (20 mL) was added,
and the aqueous solution was extracted with EtOAc (3”30 mL).
The combined organic layers were washed with brine and dried
over anhydrous sodium sulfate, and the solvent was evaporated
under reduced pressure. The crude residue was purified by column
chromatography on a silica gel column with CHCl3/MeOH (95:5)
and EtOAc as the mobile phases to give the target compound as
a white solid (150 mg, 40% yield): mp: 90–928C; 1H NMR (300 MHz,
CDCl3): d=7.42–7.38 (m, 2H, arom.), 7.19–7.15 (m, 2H, arom.), 6.95-
6.88 (m, 2H, arom.), 6.82–6.78 (m, 2H, arom.), 6.62–6.57 (m, 2H,
isoquinoline), 6.38–6.25 (bs, 1H, NH: exch. with D2O), 4.59 (slightly
broad singlet, 1H, ArCHHNCO, conformer signal), 4.51 (slightly
broad singlet, 1H, ArCHHNCO, conformer signal), 3.83–3.73 (m,
13H, 4 OCH3 and 1H, ArCH2CHHNCO, conformers signals), 3.68–
3.58 (m, 3H, ArCH2CONH and ArCH2CHHNCO, conformers signals),
3.27 (q, 2H, J=6.2 Hz, NHCH2(CH2)3), 2.77 (t, 2H, J=5.8 Hz, ArCH-
HCH2NCO, conformer signal), 2.71 (t, 1H, J=5.8 Hz, ArCHHCH2NCO,
conformer signal), 2.40 (t, 1H, J=6.9 Hz, (CH2)3CH2CON), 1.71–
1.53 ppm (m, 4H, CH2(CH2)2CH2); 13C NMR (75 MHz, CDCl3, conform-
ers signals): d=171.80, 171.76, 167.0, 163.1, 161.2, 160.7, 159.6,
148.1, 148.0, 147.9, 147.8, 131.3, 130.0, 127.1, 126.0, 125.5, 124.4,
121.7, 121.4, 117.7, 114.5, 114.1, 111.77, 111.73, 111.42, 111.36, 109.6,
109.5, 109.1, 109.08, 56.2, 56.1, 55.5, 55.4, 47.2, 44.1, 43.5, 39.9,
39.7, 34.2, 33.2, 32.9, 29.2, 29.0, 28.2, 22.2, 22.1 ppm; FT-IR (KBr):
n˜ =3300, 3074, 3000, 2934, 2837, 1612, 1515, 1455, 1433, 1304,
1251, 1177, 1114, 1021, 967, 936, 836, 799, 750 cmÀ1; ESI-MS (m/z):
C35H39N3O7, 636 [M+Na]+.
1
forded the target compound as an oil (278 mg, 50% yield): H NMR
(300 MHz, CDCl3): d=7.71–7.62 (m, 2H, arom.), 7.50–7.39 (m, 2H,
arom.) 6.90-6.82 (m, 2H, arom.), 6.67–6.56 (m, 3H, arom.), 6.23–6.09
(two broad triplets, 1H, NH: exch. with D2O, two conformers sig-
nals), 4.52 (slightly broad singlet, 1H, ArCHHNCO, conformer
signal), 4.48 (slightly broad singlet, 1H, ArCHHNCO, conformer
signal), 3.84–3.81 (m, 6H, 2 OCH3, conformers signals), 3.78 (s, 3H,
OCH3), 3.73 (t, 1H, J=5.9 Hz, ArCH2CHHNCO, conformer signal),
3.64–3.54 (m, 3H, CH2CONH and ArCH2CHHNCO, conformers sig-
nals), 3.21 (q, 2H, J=6.24 Hz, NHCH2(CH2)3, conformer signal), 2.76
(t, 1H, J=5.7 Hz, ArCHHCH2NCO, conformer signal), 2.69 (t, 1H, J=
5.7 Hz, ArCHHCH2NCO, conformer signal), 2.40–2.28 (m, 5H, CH3
and CH2CO-isoquinoline), 1.62–1.42 ppm (m, 4H, NHCH2(CH2)2CH2);
13C NMR (75 MHz, CDCl3, conformers signals): d=171.6, 170.26,
170.23, 168.6, 156.4, 148.15, 148.06, 147.91, 147.87, 139.63, 139.56,
136.6, 136.5, 133.9, 131.4, 131.2, 130.73, 130.69, 129.3, 127.1, 125.8,
125.5, 124.3, 115.3, 113.2, 112.4, 111.8, 111.7, 111.4, 111.3, 109.6,
109.5, 109.0, 101.0, 56.27, 56.21, 56.12, 56.0, 55.9, 47.1, 44.0, 43.4,
39.9, 39.2, 33.0, 32.6, 32.5, 29.9, 29.2, 28.2, 21.86, 21.81, 13.6 ppm;
FT-IR (KBr): n˜ =3306, 3069, 2930, 2835, 1675, 1645, 1518, 1477,
1455, 1358, 1323, 1257, 1223, 1148, 1115, 1088, 1064, 1034, 1014,
925, 851, 832, 754 cmÀ1; ESI-MS (m/z): C35H3837ClN3O6 654 [M+H]+
(33), C35H3835ClN3O6 654 [M+H]+ (100).
2-[3-(5-Chlorofuran-2-yl)-4-phenylisoxazol-5-yl]-N-[5-(3,4-dihy-
dro-6,7-dimethoxyisoquinolin-2-(1H)-yl)]-5-oxopentylacetamide
(15): Compound 13 (26 mg, 0.08 mmol), HOBt (16 mg, 0.12 mmol),
EDC·HCl (18 mg, 0.09 mmol), and dry DIPEA (0.023 mL,
0.135 mmol) were added to a solution of 2-[3-(5-chlorofuran-2-yl)-
4-phenylisoxazol-5-yl]acetic acid (P6-COOH; 41 mg, 0.14 mmol) in
dry DMF (3 mL) at room temperature. The reaction mixture was
stirred for 24 h, then H2O (20 mL) was added, and the aqueous so-
lution was extracted with EtOAc (3”30 mL). The combined organic
layers were washed with brine and dried over anhydrous sodium
sulfate, and the solvent was evaporated under reduced pressure.
The crude residue was purified by column chromatography on
a silica gel column with EtOAc as the mobile phase to give the
target compound as a yellow solid (16.2 mg, 35% yield): mp: 225–
2288C; 1H NMR (300 MHz, CDCl3): d=7.43–7.39 (m, 5H, arom.),
6.62–6.58 (m, 2H, isoquinoline), 6.44 (bs, 1H, NH: exch. with D2O),
6.27 (d, 1H, J=6.1 Hz, furanyl): 6.15 (d, 1H, J=6.1 Hz, furyl), 4.61 (s,
1H, ArCHHNCO, conformer signal), 4.53 (s, 1H, ArCHHNCO, confor-
mer signal), 3.85 (s, 6H, OCH3), 3.77 (t, 1H, J=5.8 Hz, ArCH2CHHN,
conformer signal), 3.65 (s, 2H, CH2CONH, and t, 1H, J=5.8 Hz,
ArCH2CHHN, conformers signals), 3.26 (q, 2H, J=6.2 Hz,
NHCH2(CH2)3), 2.80 (t, 1H, J=6.2 Hz, ArCHHCH2N, conformer signal),
2.72 (t, 1H, J=6.2 Hz, ArCHHCH2N, conformer signal), 2.41 (t, 2H,
J=6.9 Hz, NH(CH2)3CH2CO, conformer signal), 1.72–1.51 ppm (m,
4H, NHCH2(CH2)2CH2CO); 13C NMR (75 MHz, CDCl3, conformers sig-
nals): d=171.75, 171.71, 166.4, 163.8, 152.8, 148.2, 148.1, 147.92,
147.89, 143.4, 138.9, 130.3, 129.1, 129.0, 128.6, 127.1, 125.9, 125.5,
124.3, 117.3, 114.4, 114.3, 111.79, 111.73, 111.39, 111.33, 109.6, 109.0,
108.3, 56.24, 56.14, 47.2, 44.1, 43.5, 39.8, 34.0, 33.0, 32.8, 29.2, 28.9,
28.2, 22.0, 21.9 ppm; FT-IR (KBr): n˜ =3435, 2923, 2851, 1722, 1629,
2-[3,4-Bis(4-methoxyphenyl)isoxazol-5-yl]-1-[6,7-dimethoxy-3,4-
dihydroisoquinolin-2-(1H)-yl]ethanone (17): 6,7-Dimethoxy-3,4-di-
hydroisoquinoline (50.2 mg, 0.26 mmol), HOBt (61 mg, 0.40 mmol),
EDC·HCl (57.5 mg, 0.30 mmol), and dry DIPEA (58.1 mg, 0.023 mL,
0.45 mmol) were added to a solution of mofezolac (144 mg,
0.44 mmol) in dry DMF (14 mL) at room temperature. The reaction
mixture was stirred for 24 h, then saturated aq. NaHCO3 (20 mL)
was added, and the aqueous solution extracted with EtOAc (3”
30 mL). The combined organic layers were washed with brine and
dried over anhydrous sodium sulfate, and the solvent was evapo-
rated under reduced pressure. Column chromatography of the
crude residue on silica gel with EtOAc as the mobile phase gave
the target compound as a yellow solid (37.4 mg, 30% yield): mp:
70–738C. 1H NMR (300 MHz, CDCl3): d=7.42–7.35 (m, 2H, arom.),
7.23–7.15 (m, 2H, arom.), 6.91–6.78 (m, 4H, arom.), 6.62–6.58 (m,
2H, isoquinoline), 4.66 (s, 1H, ArCHHNCO, conformer signal), 4.54
(s, 1H, ArCHHNCO, conformer signal), 3.91–3.61 (m, 16H, 4 OCH3,
ArCH2CON and ArCH2CH2NCO), 2.82 (t, 1H, J=5.7 Hz, ArCH-
HCH2NCO, conformer signal), 2.75 ppm (t, 1H, J=5.7 Hz, ArCH-
HCH2NCO, conformer signal); 13C NMR (75 MHz, CDCl3, conformers
signals): d=166.1, 165.9, 163.1, 161.1, 160.7, 159.6, 148.2, 148.1,
147.9, 131.5, 131.4, 130.1, 129.9, 126.9, 125.8, 125.0, 123.9, 122.0,
121.9, 121.5, 121.4, 117.6, 117.5, 114.4, 114.1, 111.75, 111.67, 111.40,
111.32, 109.56, 109.48, 109.0, 108.9, 56.3, 56.1, 55.7, 55.5, 47.9, 44.6,
44.3, 40.5, 32.3, 31.9, 29.1, 28.1 ppm; FT-IR (KBr): n˜ =3069, 2997,
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1184
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
2935, 2836, 1651, 1610, 1574, 1516, 1363, 1112, 1024, 968, 902,
10H, 3 OCH3 and ArCH2CHHN, conformer signal), 3.60–3.56 (m, 3H,
CH2CONH and ArCH2CHHN, conformer signal), 3.18 (q, 2H, J=
6.7 Hz, NHCH2(CH2)3), 2.78 (t, 1H, J=5.8 Hz, ArCHHCH2N, conformer
signal), 2.73 (t, 1H, J=5.8 Hz, ArCHHCH2N, conformer signal), 2.35-
2.31 (m, 5H, ArCH3 and CH2CO-isoquinoline), 1.61–1.38 ppm (m,
4H, CH2(CH2)2CH2); 13C NMR (75 MHz, CDCl3, conformers signals):
d=171.8, 171.7, 154.5, 148.13, 148.06, 147.91, 147.87, 134.4, 130.6,
128.9, 127.1, 126.0, 125.5, 124.4, 111.7, 111.6, 111.5, 111.4, 109.5,
104.7, 100.0, 56.21, 56.16, 56.08, 47.2, 44.1, 43.5, 39.9, 39.2, 33.2,
32.9, 32.5, 29.4, 29.2, 28.2, 22.3, 22.2, 11.9; FT-IR (KBr): n˜ =3298,
3069, 2933, 1640, 1518, 1486, 1455, 1362, 1311, 1257, 1217, 1114,
1031, 913, 853, 799, 753 cmÀ1; ESI-MS (m/z): C28H35N3O5, 516 [M+
Na]+.
835 cmÀ1; ESI-MS (m/z): C30H30N2O6, 537 [M+Na]+.
Methyl 5-{2-[1-(4-chlorobenzoyl)-5-methoxy-2-methyl-1H-indol-
3-yl]acetamido}pentanoate (18): DCC (144.2 mg, 0.7 mmol) and
HOBt (81 mg, 0.6 mmol) were added to a stirred solution of indo-
methacin (100 mg, 0.5 mmol) in dry CH2Cl2 (10 mL) at 08C. DIEA
(0.2 mL, 1.1 mmol) and methyl 5-aminopentanoate (100 mg,
0.6 mmol) in dry CH2Cl2 (10 mL) were then dropwise added at 08C.
The reaction was stirred overnight at room temperature. The reac-
tion mixture was then filtered, and the solvent was evaporated
under reduced pressure. Column chromatography of the crude res-
idue on silica gel and EtOAc/hexane (2:8) as a mobile phase afford-
ed the target compound as a yellow solid (180 mg, 77% yield);
mp: 126–1288C; 1H NMR (300 MHz, CDCl3): d=7.67 (d, 2H, J=
8.5 Hz, arom.), 7.48 (d, 2H, J=8.5 Hz, arom.), 6.89–6.85 (m, 2H,
arom.), 6.69 (dd, 1H, J=9.0 and 2.5 Hz, arom.), 5.70–5.66 (bs, 1H,
NH: exch. with D2O), 3.82 (s, 3H, ArOCH3), 3.63 (s, 2H, CH2CONH),
3.62 (s, 3H, COOCH3), 3.25–3.16 (m, 2H, NHCH2(CH2)3), 2.38 (s, 3H,
ArCH3), 2.27 (t, 2H, J=6.8 Hz, (CH2)3CH2), 1.58–1.38 ppm (m, 4H,
CH2(CH2)2CH2); 13C NMR (75 MHz, CDCl3): d=174.0, 170.1, 168.6,
156.5, 139.8, 136.6, 133.8, 131.4, 131.1, 130.5, 129.4, 115.4, 113.1,
112.6, 100.9, 55.9, 51.8, 39.3, 34.1, 33.5, 32.5, 29.1, 25.8, 25.1, 22.0,
13.5 ppm; ESI-MS (m/z): C25H27N2O5, 469 (M À1), 493 [M+Na]+.
Computational chemistry
Molecular modeling and graphics manipulations were performed
using the molecular operating environment (MOE)[42] and UCSF-
CHIMERA[43] software packages, running on a E4 Computer Engi-
neering E1080 workstation with an Intel Core i7-930 Quad-Core
processor. GOLD 5.2.2[28,29] was used for all docking calculations.
Figure were generated using PyMOL 1.0.[44]
5-[2-(5-Methoxy-2-methyl-1H-indol-3-yl)acetamido]pentanoic
acid (19):[41] A solution of 0.5N LiOH (200 mg, 8 mmol, in 7 mL of
Ligand and receptor preparation: Model building and geometry
optimization of compounds 8 and 17 was accomplished with the
MMFF94X force field, available within MOE. The crystal structures
of ovine COX-1 in complex with bound inhibitor indomethacin-(R)-
a-ethyl-ethanolamide (PDB ID: 2OYE)[27] and murine COX-2 com-
plexed with SC-558 (PDB ID: 6COX)[26] were used in the docking ex-
periments, as the murine and ovine COX-2 sequences share a high
level of identity (87.3%). Database searching (SWISS-PROT), se-
quence alignment, and analysis of murine and ovine COX-2 se-
quences were carried out using the BLAST program.[45] Moreover,
all COX-2 active-site residues are largely conserved across the two
species. Bound ligands and water molecules were removed. Cor-
rect atom assignment for Asn, Gln, and His residues was per-
formed, and hydrogen atoms were added using standard MOE ge-
ometries. Partial atomic charges were computed by MOE using the
Amber99 force field. All heavy atoms were then fixed, and hydro-
gen atoms were minimized using the Amber99 force field and a di-
electricÀc1onstant equal to 1, terminating at a gradient of 0.001 kcal
H2O) was added dropwise to
a solution of 18 (170 mg,
0.362 mmol) in THF (5 mL) at room temperature. After 1.5 h, 1N
HCl (18 mL) was added, and the aqueous phase was extracted with
EtOAc (3”30 mL). The combined organic layers were dried over
anhydrous sodium sulfate, and the solvent was evaporated under
reduced pressure. Column chromatography of the crude residue
on silica gel with EtOAc (100%) to CHCl3/CH3OH (95:5) as the
mobile phase afforded the target compound as a white solid
(100 mg, 87% yield); mp: 122–1258C; 1H NMR (300 MHz,
[D4]MeOH): d=7.50–7.35 (bs, 1H, NH: exch. with D2O), 7.13 (d, 2H,
J=8.8 Hz, arom.), 6.90 (d, 2H, J=2.3 Hz, arom.), 6.67 (dd, 1H, J=
8.8 and 2.3 Hz, arom.), 3.78 (s, 3H, OCH3), 3.54 (s, 2H, CH2CONH),
3.20–3.10 (m, 2H, NHCH2), 2.34 (s, 3H, ArCH3), 2.23 (t, 2H, J=
7.1 Hz, CH2COOH), 1.58–1.40 ppm (m, 4H, CH2(CH2)2CH2); 13C NMR
(75 MHz, CDCl3): d=176.1, 173.8, 154.0, 134.3, 131.1, 128.9, 110.9,
110.3, 103.8, 99.7, 55.1, 38.9, 33.2, 31.7, 28.7, 22.0, 10.3 ppm; FT-IR
(KBr): n˜ =3403, 3292, 2924, 2854, 1699, 1622, 1540, 1488, 1463,
molÀ1
Š .
1434, 1355, 1311, 1203, 1111, 1037, 864, 668, 556 cmÀ1
.
Docking simulations: Docking of 8 and 17 to COX-1 and COX-2,
respectively, was performed using the GOLD software, which uses
a genetic algorithm (GA) for determining the docking modes of li-
gands and proteins. The coordinates of the co-crystallized ligand
(indomethacin-(R)-a-ethyl-ethanolamide for COX-1 and SC-558 for
COX-2) were chosen as the active site origin. The active site radius
was set to 10 Š. Mobility of the Arg120, Tyr355, Glu524, and
Ser530 side chains was established using the flexible side chains
option in the GOLD front end, which incorporates the Lovell rota-
mer library.[46] Each GA run used the default parameters of 100000
genetic operations on an initial population of 100 members divid-
ed into five subpopulations, with weights for crossover, mutation,
and migration set to 95, 95, and 10, respectively. GOLD allows
a user-definable number of GA runs per ligand, each of which
starts from a different orientation. For these experiments, the
number of GA runs was set to 200 without the option of early ter-
mination, and scoring of the docked poses was performed with
the ChemPLP scoring function, followed by rescoring with Chem-
Score.[3,19,30–34] The final receptor–ligand complex for each ligand
was chosen interactively by selecting the highest scoring pose that
N-{5-[6,7-Dimethoxy-3,4-dihydroisoquinolin-2-(1H)-yl]-5-oxopen-
tyl}-2-(5-methoxy-2-methyl-1H-indol-3-yl)acetamide (20): DCC
(338 mg, 1.64 mmol) and DMAP (36 mg, 0.296 mmol) were added
to a stirred solution of 19 (265 mg, 0.83 mmol) in dry CH2Cl2
(20 mL) under argon atmosphere at 08C. After 30 min, a solution
of 6,7-DM-THIQ (317 mg, 1.64 mmol) in dry CH2Cl2 (10 mL) was
added dropwise at room temperature. The reaction mixture was
stirred overnight, then 37% HCl was added, and the aqueous solu-
tion was extracted with CH2Cl2. The combined organic phases were
dried over anhydrous sodium sulfate, and the solvent was evapo-
rated under reduced pressure. Column chromatography of the
crude residue on silica gel using EtOAc/hexane (70:30) to CHCl3/
CH3OH (97: 3) as the mobile phase afforded the target compound
as a yellow solid (409 mg, 23% yield); mp: 57–608C; 1H NMR
(300 MHz, CDCl3): d=8.10 (s, 1H, NH indole: exch. with D2O), 7.17
(d, 1H, J=8.8 Hz, arom.), 6.86 (s, 1H, arom.), 6.79 (dd, 1H, J=8.8
and 2.4 Hz, arom.), 6.62–6.59 (m, 2H, isoquinoline), 5.84–5.81 (m,
1H, CONH: exch. with D2O), 4.59 (s, 1H, ArCHHNCO, conformer
signal), 4.49 (s, 1H, ArCHHNCO, conformer signal), 3.86–3.70 (m,
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1185
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
was consistent with experimentally derived information regarding
the binding mode of the ligand.
(HBSS, pH 7.4) and were sterile-filtered. After 21 days of cell
growth, the medium was removed from filter wells and from the
receiver plate. The filter wells were each filled with 75 mL of fresh
HBSS buffer and the receiver plate was filled with 250 mL per well
of the same buffer. This procedure was repeated twice, and the
plates were incubated at 378C for 30 min. After incubation, the
HBSS buffer was removed, and in some wells, drug solutions were
added to the filter wells (75 mL), and HBSS without drug was
added to the corresponding receiver plate (250 mL); in other wells,
the drug solutions were added to the receiver plate (250 mL), and
HBSS without drug was added to the corresponding filter wells.
Biology
Cyclooxygenase inhibition assays: The target compounds were eval-
uated for their ability to inhibit ovine COX-1 and COX-2 enzyme
catalytic activity (percent inhibition at 50 mm, unless otherwise in-
dicated). IC50 values were determined only for compounds that
showed a reasonable COX-1 inhibitory activity (>50%) at 50 mm
(Table 2). Each reported IC50 value and the percentage of inhibition
(measured at 50 mm as the tested compound concentration) is the
average of the results of three separate assays (triplicate). Enzyme
inhibition was determined using a colorimetric COX (ovine) inhibi-
tor screening assay kit (Cayman Chemicals, Ann Arbor, MI, USA) fol-
lowing the manufacturer’s instructions. Stock solutions of test com-
pounds were dissolved in a minimum volume of DMSO.
The plates were incubated at 378C for 120 min. After incubation,
the samples were removed both from the apical (filter well) and
basolateral (receiver plate) side of the monolayer and were stored
at À208C pending analysis. The concentrations of compounds
were analyzed using UV/Vis spectroscopy. The Papp, in units of
nmsÀ1, was calculated using Equation (1):
Calcein-AM assays: Calcein-AM is a lipophilic pro-fluorescent probe
and a P-gp substrate. In the presence of a P-gp inhibitor, calcein-
AM is able to permeate the cell membrane and is hydrolyzed by
cytosolic esterases to fluorescent calcein. As calcein is hydrophilic
and is not a P-gp substrate, it cannot cross the cell membrane, and
a rapid increase in fluorescence can be measured. These experi-
ments were carried out as previously described[39] with minor
modifications. Briefly, an MDCK-MDR1 cell line was seeded (50000
cells per well) into black 96-well plates with 100 mL of medium,
and cells were allowed to become confluent overnight. Test com-
pounds were solubilized in 100 mL of culture medium and were
added to the cell monolayers. The plates were then incubated at
378C for 30 min. Calcein-AM was added in 100 mL of phosphate-
buffered saline (PBS) to yield a final concentration of 2.5 mm, and
the plate was incubated for another 30 min. Each well was washed
three times with ice-cold PBS. Saline buffer (100 mL) was added to
each well, and the plates were read with a Victor3 fluorimeter (Per-
kinElmer) at excitation and emission wavelengths of l 485 and
535 nm, respectively. Under these experimental conditions, calcein
cell accumulation in the absence and presence of tested com-
pounds was evaluated, and basal-level fluorescence was estimated
from the fluorescence of untreated cells. In treated wells, the in-
crease in fluorescence was measured relative to that of the basal
level. EC50 values were determined by fitting the percent fluores-
cence increase versus log[dose].
Papp ¼ ½VA=ðarea  tފ  ð½drugŠacceptor=½drugŠinitÞ
ð1Þ
where VA is the volume (mL) in the acceptor well, area is the sur-
face area of the membrane (0.11 cm2 of the well), t is the total
transport time in seconds (7200 s), [drug]acceptor is the concentration
of the drug as measured by UV spectroscopy, and [drug]init is the
initial drug concentration (1”10À4 m) in the apical or basolateral
wells.
Analytical methods: ESI-MS analyses were performed by
10 mLminÀ1 direct infusion of a MeOH sample solution on an Agi-
lent 1100 LC–MSD trap system VL (ESI ion source type, 4000 V ion
source voltage). Samples from in vitro permeation studies were an-
alyzed using a Shimadzu UV-1800 spectrophotometer. For each
compound, a calibration curve was constructed at the opportune
wavelength (Table 4).
Table 4. Absorption wavelength and molar extinction coefficient of each
compound tested.
Compound
l [nm]
e [mÀ1 cmÀ1
]
8
230
280
230
230
44725
8248
35831
33109
15
16
17
Permeability assays: Caco-2 cells were harvested with trypsin-EDTA
and seeded into a MultiScreen Caco-2 assay system at a density of
10000 cells per well. The culture medium was replaced every 48 h
for the first six days and every 24 h thereafter. After 21 days in cul-
ture, the Caco-2 cell monolayer was used for the permeability ex-
periments. The transepithelial electrical resistance (TEER) of the
monolayers was measured daily before and after the experiment
using an epithelial voltohmmeter (Millicell-ERS; Millipore, Billerica,
MA, USA). Generally, the obtained TEER values were greater than
1000 W for a 21-day culture.
Statistical analysis: The IC50 and EC50 values of the compounds re-
ported in Tables 2 and 3 were determined by nonlinear curve fit-
ting using GraphPad Prism version 3.0 (GraphPad Software Inc.,
San Diego, CA, USA).[47]
Acknowledgements
Drug transport assays: The permeability assay allows two fluxes
through the cell monolayer to be studied, from the basolateral to
the apical (B!A) and from the apical to the basolateral (A!B)
compartments, thus predicting the interacting mechanism of the
tested compound with the efflux pump. In particular, if a studied
This work was supported financially by the Ministero dell’Istru-
zione, dell’Università e della Ricerca Scientifica e Tecnologica,
Rome, Italy (MIUR-PRIN 2010-2011, prot. 2010W7YRLZ_003 to
A.L.).
compound is a P-gp inhibitor has an apparent permeability (Papp
)
BA/AB<2, whereas a P-gp substrate shows BA/AB>2.[40] Apical to
basolateral (Papp A!B) and basolateral to apical (Papp B!A) perme-
ability of drugs were measured at 120 min at a concentration of
100 mm. Drugs were dissolved in Hank’s balanced salt solution
Keywords: molecular docking
cyclooxygenase · isoxazoles · P-glycoprotein
· decomposition analysis ·
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1186
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[1] M. G. Perrone, A. Scilimati, L. Simone, P. Vitale, Curr. Med. Chem. 2010,
17, 3769–3805.
[2] C. Yuan, R. S. Sidhu, D. V. Kuklev, Y. Kado, M. Wada, I. Song, W. L. Smith,
J. Biol. Chem. 2009, 284, 10046–10055.
[3] P. Vitale, M. G. Perrone, P. Malerba, A. Lavecchia, A. Scilimati, Eur. J. Med.
Chem. 2014, 74, 606–618.
[4] M. G. Perrone, P. Vitale, A. Panella, C. G. Fortuna, A. Scilimati, Eur. J. Med.
Chem. 2015, 94, 252–264.
[26] R. G. Kurumbail, A. M. Stevens, J. K. Gierse, J. J. McDonald, R. A. Stege-
man, J. Y. Pak, D. Gildehaus, J. M. Miyashiro, T. D. Penning, K. Seibert,
P. C. Isakson, W. C. Stallings, Nature 1996, 384, 644–648.
[27] C. A. Harman, M. V. Turman, K. R. Kozak, L. J. Marnett, W. L. Smith, R. M.
Garavito, J. Biol. Chem. 2007, 282, 28096–28105.
[28] GOLD version 5.2.2, CCDC Software Ltd., Cambridge (UK), 2008.
[29] G. Jones, P. Willett, R. C. Glen, A. R. Leach, R. Taylor, J. Mol. Biol. 1997,
267, 727–748.
[30] M. L. Verdonk, J. C. Cole, M. J. Hartshorn, C. W. Murray, R. D. Taylor, Pro-
teins Struct. Funct. Genet. 2003, 52, 609–623.
[31] N. Micale, R. Ettari, A. Lavecchia, C. Di Giovanni, K. Scarbaci, V. Troiano,
S. Grasso, E. Novellino, T. Schirmeister, M. Zappalà, Eur. J. Med. Chem.
2013, 64, 23–34.
[32] G. Fracchiolla, A. Laghezza, L. Piemontese, P. Tortorella, F. Mazza, R.
Montanari, G. Pochetti, A. Lavecchia, E. Novellino, S. Pierno, D. Conte
Camerino, F. Loiodice, J. Med. Chem. 2009, 52, 6382–6393.
[33] L. Porcelli, F. Gilardi, A. Laghezza, L. Piemontese, N. Mitro, A. Azzariti, F.
Altieri, L. Cervoni, G. Fracchiolla, M. Giudici, U. Guerrini, A. Lavecchia, R.
Montanari, C. Di Giovanni, A. Paradiso, G. Pochetti, G. M. Simone, P. Tor-
torella, M. Crestani, F. Loiodice, J. Med. Chem. 2012, 55, 37–54; G. Frac-
chiolla, M. Giudici, U. Guerrini, A. Lavecchia, R. Montanari, C. Di Giovan-
ni, A. Paradiso, G. Pochetti, G. M. Simone, P. Tortorella, M. Crestani, F.
Loiodice, J. Med. Chem. 2012, 55, 37–54.
[5] M. G. Perrone, P. Malerba, Md. J. Uddin, P. Vitale, A. Panella, B. C. Crews,
C. K. Daniel, K. Ghebreselasie, M. Nickels, M. N. Tantawy, H. C. Manning,
L. J. Marnett, A. Scilimati, Eur. J. Med. Chem. 2014, 80, 562–568.
[6] S. H. Choi, S. Aid, L. Caracciolo, S. S. Minami, T. Niikura, Y. Matsuoka, R. S.
Turner, M. P. Mattson, F. Bosetti, J. Neurochem. 2013, 124, 59–68.
[7] R. Calvello, M. A. Panaro, M. L. Carbone, A. Cianciulli, M. G. Perrone, P.
Vitale, P. Malerba, A. Scilimati, Pharmacol. Res. 2012, 65, 137–148.
[8] N. A. Colabufo, F. Berardi, M. Cantore, M. Contino, C. Inglese, M. Niso, R.
Perrone, J. Med. Chem. 2010, 53, 1883–1897.
[9] P. Vitale, S. Tacconelli, M. G. Perrone, P. Malerba, L. Simone, A. Scilimati,
A. Lavecchia, M. Dovizio, E. Marcantoni, A. Bruno, P. Patrignani, J. Med.
Chem. 2013, 56, 4277–4299.
[10] J. J. Prusakiewicz, A. S. Felts, B. S. Mackenzie, L. J. Marnett, Biochemistry
2004, 43, 15439–15445.
[11] L. Di Nunno, P. Vitale, A. Scilimati, S. Tacconelli, P. Patrignani, J. Med.
Chem. 2004, 47, 4881–4890.
[12] M. Kajitani, M. Tanaka, Y.-I. Hagiwara, M. Yasumoto (Taiho Pharmaceuti-
cal Co. Ltd.), US Pat. No. US5142091, 1992.
[34] K. Scarbaci, V. Troiano, N. Micale, R. Ettari, C. Di Giovanni, C. Cerchia, S.
Grasso, E. Novellino, T. Schirmeister, A. Lavecchia, M. Zappalà, Eur. J.
Med. Chem. 2014, 76, 1–9.
[35] A. J. Vecchio, D. M. Simmons, M. G. Malkowski, J. Biol. Chem. 2010, 285,
22152–22163.
[36] K. Gupta, B. S. Selinsky, C. J. Kaub, A. K. Katz, P. J. Loll, J. Mol. Biol. 2004,
335, 503–518.
[37] E. D. Thuresson, K. M. Lakkides, C. J. Rieke, Y. Sun, B. A. Wingerd, R. Mi-
cielli, A. M. Mulichak, M. G. Malkowski, R. M. Garavito, W. L. Smith, J. Biol.
Chem. 2001, 276, 10347–10357.
[38] O. Llorens, J. J. Perez, A. Palomer, D. Mauleon, Bioorg. Med. Chem. Lett.
1999, 9, 2779.
[39] B. Feng, J. B. Mills, R. E. Davidson, R. J. Mireles, J. S. Janiszewski, M. D.
Troutman, S. M. de Morais, Drug Metab. Dispos. 2008, 36, 268–275.
[40] J. W. Polli, S. A. Wring, J. E. Humphreys, L. Huang, J. B. Morgan, L. O.
Webster, C. S. Serabjit-Singh, J. Pharmacol. Exp. Ther. 2001, 299, 620–
628.
[13] N. A. Colabufo, F. Berardi, M. G. Perrone, E. Capparelli, M. Cantore, C. In-
glese, R. Perrone, Curr. Top. Med. Chem. 2010, 10, 1703–1714.
[14] K. McHugh, R. Callaghan, Multidrug Resistance: Biological and Pharma-
ceutical Advances in Antitumor Treatment (Ed.: N. A. Colabufo), Research
Signpost, Kerala, India, 2008.
[15] E. Capparelli, L. Zinzi, M. Cantore, M. Contino, M. G. Perrone, G. Luurtse-
ma, F. Berardi, R. Perrone, N. A. Colabufo, J. Med. Chem. 2014, 57, 9983–
9994.
[16] M. Contino, L. Zinzi, M. Cantore, M. G. Perrone, M. Leopoldo, F. Berardi,
R. Perrone, N. A. Colabufo, Bioorg. Med. Chem. Lett. 2013, 23, 3728–
3731.
[17] M. Arisawa, Y. Kasaya, T. Obata, T. Sasaki, T. Nakamura, T. Araki, K. Yama-
moto, A. Sasaki, A. Yamano, M. Ito, H. Abe, Y. Ito, S. Shuto, J. Med. Chem.
2012, 55, 8152–8163.
[41] T. Y. Shen, L. R. Ellis, T. B. Windholz, A. R. Matzuk, A. Rosegay, S. Lucas,
B. E. Witzel, C. H. Stammer, A. N. Wilson, F. W. Holly, J. D. Willett, L. H.
Sarett, W. J. Holtz, J. Am. Chem. Soc. 1963, 85, 488–489.
[42] Molecular Operating Environment (MOE) version 2013.08, Chemical
Computing Group Inc., 1010 Sherbooke St. West, Suite #910, Montreal,
QC, H3A 2R7 (Canada), 2012.
[43] C. C. Huang, G. S. Couch, E. F. Pettersen, T. E. Ferrin, Pac. Symp. Biocom-
put. 1996, 724, http://www.cgl.ucsf.edu/chimera.
[44] W. L. DeLano, The PyMOL Molecular Graphics System, DeLano Scientific
LLC, San Carlos, CA (USA), http://www.pymol.org/.
[45] Y. J. Pak, S. Wang, J. Phys. Chem. B 2000, 104, 354–359.
[46] S. C. Lovell, J. M. Word, J. S. Richardson, D. C. Richardson, Proteins Struct.
Funct. Bioinf. 2000, 40, 389–408.
[47] Prism version 3.0 for Windows, GraphPad Software Inc., San Diego, CA
(USA), 1998.
[18] A. Angelini, M. Iezzi, C. Di Febbo, C. Di Ilio, F. Cuccurullo, E. Porreca,
Oncol. Rep. 2008, 20, 731–735.
[19] M. G. Perrone, P. Vitale, P. Malerba, A. Altomare, R. Rizzi, A. Lavecchia, C.
Di Giovanni, E. Novellino, A. Scilimati, ChemMedChem 2012, 7, 629–641.
[20] M. G. Perrone, D. Lofrumento, P. Vitale, F. De Nuccio, V. La Pesa, A. Panel-
la, R. Calvello, A. Cianciulli, M. A. Panaro, A. Scilimati, Pharmacology
2015, 95, 22–28.
[21] Md. J. Uddin, B. C. Crews, K. Ghebreselasie, L. J. Marnett, Bioconjugate
Chem. 2013, 24, 712–723.
[22] Md. J. Uddin, B. C. Crews, A. L. Blobaum, P. J. Kingsley, D. L. Gorden, J. O.
McIntyre, M. L. Matrisian, K. Subbaramaiah, A. J. Dannenberg, D. W.
Piston, L. J. Marnett, Cancer Res. 2010, 70, 3618–3627.
[23] M. A. Cinelli, B. Cordero, T. S. Dexheimer, Y. Pommier, M. Cushman,
Bioorg. Med. Chem. 2009, 17, 7145–7155.
[24] L. Huang, F.-p. Hou, P. Xi, D. Bai, M. Xu, Z. Li, G. Xie, Y. Shi, H. Liu, Z.
Zeng, J. Inorg. Biochem. 2011, 105, 800–805.
[25] I. Carlescu, H. M. I. Osborn, J. Desbrieres, D. Scutaru, M. Popa, Carbohydr.
Res. 2010, 345, 33–40.
Received: September 24, 2015
Published online on May 2, 2016
ChemMedChem 2016, 11, 1172 – 1187
www.chemmedchem.org
1187
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim