Design, synthesis, biochemical studies, cellular characterization, and structure-based computational studies of small molecules targeting the urokinase receptor

Article abstract of DOI:10.1016/j.bmc.2012.06.002

The urokinase receptor (uPAR) serves as a docking site to the serine protease urokinase-type plasminogen activator (uPA) to promote extracellular matrix (ECM) degradation and tumor invasion and metastasis. Previously, we had reported a small molecule inhibitor of the uPAR·uPA interaction that emerged from structure-based virtual screening. Here, we measure the affinity of a large number of derivatives from commercial sources. Synthesis of additional compounds was carried out to probe the role of various groups on the parent compound. Extensive structure-based computational studies suggested a binding mode for these compounds that led to a structure-activity relationship study. Cellular studies in non-small cell lung cancer (NSCLC) cell lines that include A549, H460 and H1299 showed that compounds blocked invasion, migration and adhesion. The effects on invasion of active compounds were consistent with their inhibition of uPA and MMP proteolytic activity. These compounds showed weak cytotoxicity consistent with the confined role of uPAR to metastasis.

Full text of DOI:10.1016/j.bmc.2012.06.002

Bioorganic & Medicinal Chemistry 20 (2012) 4760–4773
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Design, synthesis, biochemical studies, cellular characterization,
and structure-based computational studies of small molecules
targeting the urokinase receptor
Fang Wang a, W. Eric Knabe a, Liwei Li a,b, Inha Jo a, Timmy Mani a, Hartmut Roehm a, Kyungsoo Oh c,
Jing Li a, May Khanna a, Samy O. Meroueh a,b,c,
⇑
a Indiana University, Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, IN 46202, United States
b Center for Computational Biology and Bioinformatics, Indiana University School of Medicine, Indianapolis, IN 46202, United States
c Department of Chemistry and Chemical Biology, Indiana University Purdue University Indianapolis (IUPUI), Indianapolis, IN 46202, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The urokinase receptor (uPAR) serves as a docking site to the serine protease urokinase-type plasminogen
activator (uPA) to promote extracellular matrix (ECM) degradation and tumor invasion and metastasis. Pre-
viously, we had reported a small molecule inhibitor of the uPARꢀuPA interaction that emerged from struc-
ture-based virtual screening. Here, we measure the affinity of a large number of derivatives from
commercial sources. Synthesis of additional compounds was carried out to probe the role of various groups
on the parent compound. Extensive structure-based computational studies suggested a binding mode for
these compounds that led to a structure–activity relationship study. Cellular studies in non-small cell lung
cancer (NSCLC) cell lines that include A549, H460 and H1299 showed that compounds blocked invasion,
migration and adhesion. The effects on invasion of active compounds were consistent with their inhibition
of uPA and MMP proteolytic activity. These compounds showed weak cytotoxicity consistent with the con-
fined role of uPAR to metastasis.
Received 31 March 2012
Revised 25 May 2012
Accepted 1 June 2012
Available online 12 June 2012
Keywords:
Urokinase receptor
Small molecule inhibitors
Protein–protein interaction
Virtual screening
Structure–activity relationship
Metastasis
Ó 2012 Elsevier Ltd. All rights reserved.
Lung cancer
Synthesis
Structure-based drug design
1. Introduction
sive cells including lung cancer.1 uPAR is able to associate with
multiple binding partners at the cell surface to promote extra-
The urokinase receptor (uPAR) is a cell-surface GPI-anchored
protein that has been strongly implicated with tumor invasion
and metastasis. The receptor is up-regulated in highly aggres-
cellular matrix (ECM) degradation and signaling,2 such as the
urokinase-type plasminogen activator (uPA) and vitronectin. De-
spite its lack of transmembrane domain, the receptor has been
shown to promote signaling through integrins,3–7 receptor tyro-
sine kinases (RTKs)8,9 and GPCRs.10 In lung cancer cells, uPAR
not only mediates signaling via integrins,11 it also promotes
invasion and degradation of the ECM by serving as a docking
site to uPA and focusing proteolysis to the pericellular milieu.12
In vivo, studies have shown that inhibition of uPAR activity re-
duces lung tumor growth and metastasis.13–15 In patient sam-
ples, uPAR serves as a strong prognostic factor for non-small
cell lung cancer (NSCLC).16
The urokinase system has been implicated in nearly every step
of tumor formation and progression, including tumorigenesis,17
cell proliferation,18–20 cell migration,21,22 adhesion,2,19 angiogene-
sis,23,24 and invasion.19,20,25,26 Hence, blocking the uPARꢀuPA inter-
action is expected to result in significant impairment of metastasis.
Peptides and antibodies have been developed to inhibit the tight
protein interaction between uPAR and uPA.27 However, the devel-
opment of small molecules that abrogate this interaction remains
Abbreviations: uPAR, urokinase receptor; uPA, urokinase-type plasminogen
activator; suPAR, soluble urokinase receptor; MMP, matrix metalloproteinase; RTK,
receptor tyrosine kinase; GPCR, G-protein coupled receptor; ECM, extracellular
matrix; MAPK, mitogen-activated protein kinase; FAK, focal adhesion kinase;
DMSO, dimethyl sulfoxide; FP, fluorescence polarization; FDA, food and drug
administration; DNA, deoxyribonucleic acid; PBS, phosphate buffered saline; Src,
sarcoma; MDA-MB-231, MD Anderson Metastatic Breast; FBS, fetal bovine serum;
DMEM, Dulbecco’s Modified Eagle Medium; SDS, sodium dodecyl sulfate; CBB,
Coomassie brilliant blue; BSA, bovine serum albumin; FITC, fluorescein isothiocy-
anate; TBST, Tris-Buffered Saline and Tween 20; IB, immunoblot; NMR, nuclear
magnetic resonance; DCM, dichloromethane; HRMS, high-resolution mass spec-
trometry; THF, tetrahydrofuran; PDB, protein databank; DMAP, 4-dimethylamino-
pyridine; DCC, N,N0-dicyclohexylcarbodiimide; FAM, 6-carboxyfluorescein; GPI,
glycosylphosphatidylinositol; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltet-
razolium bromide; HPLC, high-performance liquid chromatography; FN, fibronec-
tin; ELISA, enzyme-linked immunosorbent assay.
⇑
Corresponding author. Tel.: +1 317 274 8315; fax: +1 317 278 9217.
E-mail address: smeroueh@iupui.edu (S.O. Meroueh).
0968-0896/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2012.06.002

F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
4761
challenging given the tight nature of the binding.28–30 Recently, we
reported the discovery of a small molecule (IPR-456; 1) that binds
ingly, these two compounds were the weakest inhibitors. Compari-
son of 3b (IPR-660) and 3k (IPR-824) reveals that the total number
of methyl groups on the six-membered piperidine ring had no effect
on affinity.
To gain insight into the direct binding of compounds to uPAR, we
take advantage of the red-shifted fluorescence of the compounds
and employ fluorescence polarization to study their interaction
with uPAR. A concentration-dependent study was performed for
eight compounds (Fig. 1D). Among them, seven (3b–d, 3f, 3i, 3m,
and 3n) bind to uPAR with equilibrium constants ranging from
to uPAR at sub-micromolar affinity (0.3
interaction with the amino-terminal fragment of uPA (uPAATF) with
an IC50 of 10
M.31 The compound was discovered by virtual
lM) and blocks its
l
screening of a commercial library docked to multiple structures
obtained from MD simulations. Cellular studies revealed that the
compound inhibited MDA-MB-231 breast cancer cell invasion.
Here, we characterize a large number of derivatives of 1 from (i)
commercial sources, and (ii) synthesis driven by rational design.
Structure-based computational studies using molecular docking,
explicit-solvent molecular dynamics (MD) simulations, and free
energy calculations afforded a structure–activity study that uncov-
ered the structural basis for inhibition. Cellular studies in NSCLC
cell lines were carried out to assess the effect of two derivatives
on cell invasion, migration, adhesion and proliferation. Further in-
sight into the mechanism of action for the compound was assessed
with protease activity and signaling studies probing ERK, Src, and
FAK signaling.
0.4 to 0.7
for 1 in the past (0.3
l
M. These values are similar to those that we measured
l
M).31 Compound 3o (IPR-632) showed no
binding at the submicromolar concentrations considered in the FP
assay. Despite its lack of activity in the ELISA, 3n showed strong
activity in the FP assay. This suggests that the compound is binding
to uPAR, but is not able to displace the full protein interaction. This
is due to the lack of a carboxylate moiety at C4 of the anthracene
ring system, which we had earlier found to be critical for inhibition
of the protein interaction. The benzoic acid ester lacks the charge on
the benzoate group of the parent compound, which is likely respon-
sible for its lack of activity.
2. Results
2.2. Design and synthesis of derivatives
2.1. A cheminformatics search for derivatives
To study the role of the functional groups that did not vary in the
commercial set of inhibitors, we resorted to synthesis of a set of
derivatives ( Schemes 2–6). Analogs of 1-hydroxyanthraquinone
were prepared (Scheme 5). The Diels–Alder cyclization of 5-hydro-
xy-1, 4-napthoquinone (12) and substituted butadienes afforded
the cycloaddition products (13a–c) employing aluminum chloride
as a catalyst.33 Bromination of the Diels–Alder adduct was carried
out using standard conditions of sodium acetate in refluxing acetic
acid.34 Arylamination of brominated adduct (14a–c) at the
p-bromine in relation to the hydroxyl group was accomplished with
Ullmann coupling conditions to give the required derivatives
(15a–c).35 Various alkylations (Schemes 2–4) at the hydroxyl group
were accomplished using standard conditions, namely benzyl (6),
isopropyl (9), and methyl (11) protection in order to probe the S4
pocket.36,37
The replacement of the piperidine ring of 1 by a bromine group
in 16 (IPR-630) resulted in little effect on activity. Further probing
of the parental structure by opening the isoxazole ring of 1 also led
to little impact on inhibition. The role of the primary amine that is
created as a result of the ring opening was investigated in 10 (IPR-
861) by replacing the amine with a hydroxyl moiety. The activity of
10 was identical to that of 17 (IPR-855) and slightly higher than 16.
This indicates that the amine group is neutralized upon binding, as
it is unlikely that the change from a charged to a neutral group
would not have an effect on activity. Neutral amines are often
observed in a hydrophobic environment. To further probe the
pocket occupied by this group, we designed and synthesized addi-
tional derivatives of 17, namely 6 (IPR-863), 9 (IPR-864), and 11
(IPR-860). 11 bears a methoxy group, and an o-carboxylate. 6
and 9 possess a benzyloxy, and an isopropoxy group instead of
the hydroxyl moiety. A concentration-dependent study showed
significant loss of activity for 6 by nearly an order of magnitude,
suggesting that the pocket for this group is not large enough to
accommodate the benzyloxy group (Fig. 1C). However, the isopro-
Starting with the structure of 1 (Scheme 1) a ligand-based sim-
ilarity search of the ZINC chemical library32 identified 127 deriva-
tives that share the core structure of the parental molecules. These
derivatives were purchased and tested for inhibition of uPAR bind-
ing to uPAATF, using an ELISA that we previously developed.31
Among the 127 compounds tested at 50 lM, 15 had inhibitory
activity (Fig. 1A). A concentration-dependent study confirmed
these results (Fig. 1B). IC50 values for 13 compounds that showed
concentration-dependent inhibition—ranged from 5
(IPR-824) to 100 M for 3i (IPR-809) (Table 1).
lM for 3k
l
The aforementioned 127 derivatives had a common anthraqui-
none core with rings A, B and C (Scheme 1). Among these, the sub-
stituents at C2 and C4 varied considerably. It is interesting that
the active derivatives possessed a carboxylate group at C4, either
as part of a benzoic acid substituent (3a–d, 3i, and 3k), or appended
to a chain that ranges from one to three carbon atoms in length (3e–
h, 3j and 3l)). The highest potency compound in the former category
is 3k (IPR-824), and in the latter category it is 3e (IPR-664). ortho-,
meta-, and para-carboxylates showed similar activity suggesting
that the carboxylate can be positioned anywhere on the ring. But
the m-carboxylate showed higher inhibition, for example, 3k is
more potent than 3c (IPR-661). The C2 substituents for active com-
pounds consisted primarily of hydrophobic five- and six-membered
pyrrolidine and piperidine rings, except for 2 (IPR-803), which bears
a seven-membered azepane ring. The piperidine rings have one or
two methyl appendages, but for 3h (IPR-804), which contains an
ethylester moiety, while 3i is a piperidinone derivative. Interest-
O
N
N
O
O
8
10
1
N
N
7
6
2
3
A
5
C
B
9
4
poxy bearing 9 did show some inhibition (IC50 = 30
still weaker by threefold than its parent 17. It was a surprise that
11 (IPR-860) showed no inhibition up to 100 M despite the smal-
lM), but was
HN
O
HN
HO
l
ler methoxy substituent compared with the isopropoxy group of 9.
It is possible that the o-carboxylate of 11 combined with the meth-
oxy group resulted in a loss of activity in a cooperative manner.
Finally, the role of the aromatic A-ring was studied through the
introduction of methyl groups at A6 and A7. Two derivatives,
O
OH
O
IPR-456 (1)
IPR-803 (2)
Scheme 1.

4762
F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
A
B100
120
100
80
60
40
20
0
3h
3k
3j
80
60
40
20
0
3g
3m
3a
3b
3d
3c
3e
3l
3f
3i
-20
-0.5
0.0
0.5
1.0
1.5
2.0
Compound
Log [Compound (μM)]
120
C
D
Compound
K (μM)
D
Compound IC50 (μM)
250
200
150
100
50
1
11
13
10
na
10
67
na
30
100
80
60
40
20
0
3n 0.6
3d 0.4
3o NA
3b 0.5
3c 0.5
3f 0.7
3i 0.5
3m 0.6
16
17
11
10
15c
6
9
15b
146
0
-20
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
-1000
0
1000
2000
3000
Log [Compound (μM)]
[uPAR] (nM)
Figure 1. (A) Screening of 127 compounds that emerged from a similarity search of the ZINC chemical library based on the structure of 1. Compounds were tested at a single
concentration of 50 M. (B) A concentration-dependent study of hit compounds obtained from commercial databases. Structure and inhibition constants are provided in
l
Table 1. (C) A concentration-dependent study using our ELISA for compounds that were synthesized in this work. (D) binding affinity (KD) of compounds using fluorescence
polarization measured at a fixed concentration of compound and increasing concentration of uPAR.
namely 15b (IPR-865) and 15c (IPR-862), were prepared to contain
a methyl group at A7 (15b), or two methyl groups, one at A6 and
another at A7 (15c). It is interesting both of these compounds
showed weaker inhibition than the parent, with IC50 of 146 and
the deeper S1 pocket in the cavity is occupied by the unsubstituted
anthraquinone benzene A-ring.
To identify the most likely binding mode, we resorted to expli-
cit-solvent MD simulations. The complex between the protein and
the small molecule are placed into a large box of solvent molecules,
and the trajectory of each atom is followed with respect to time,
providing a detailed account of the motion of the complex. The
snapshots along the trajectory not only provide insight on the sta-
bility of the ligand, but they can also be used to determine the free
energy of binding using the MM-GBSA method as we have previ-
ously done.31 Starting with each binding mode of 1, 2 and 3k, we
carried out a 5 ꢁ 10 ns explicit-solvent molecular simulation using
the PMEMD module within the AMBER9 suite of programs.32 The struc-
tures collected over the course of the trajectory are used to mea-
sure the root-mean-square deviation (RMSD) of the atoms within
the compound with respect to the first snapshot. As illustrated
for 2 inFigure 2B, BP1 and BP3 showed lower RMSDs than BP2, sug-
gesting that the binding modes are more stable. The RMSDs of the
other compounds, namely 1 and 3k, also revealed greater stability
for BP1 and BP3 over BP2 (Supplementary data, Fig. S1).
67 lM, respectively (Fig. 1C). This suggests that the methyl groups
cause steric clashes that prevent the compound from adopting the
binding mode required for tight binding and inhibition of the pro-
tein interaction. This suggests that small alterations to the struc-
ture at the A ring significantly impair the compound’s ability to
inhibit the protein–protein interaction.
2.3. Structure-based computational study of binding mode
An analysis of the three dimensional structure of compounds in
complex with uPAR could help explain changes in activity that were
observed as a result of chemical modification of the parent com-
pound. Short of a crystal structure, structure-based computational
tools such as docking, molecular dynamics simulations (MD), and
free energy calculations can be used not only to predict binding
mode, but also to provide insight into the physical basis for interac-
tion, and inform future structure-based design efforts. Docking of 1
along with two of its derivatives (2 and 3k) to uPAR revealed three
possible binding poses (BP1, BP2, and BP3 as illustrated in Fig. 2A
for 2). The first binding mode (BP1) places the dimethyl piperidine
(1 and 3k), and the azepane (2) substituent into the S2 cavity, while
the aminobenzoic group of each compound is located at S3 (Fig. 2A
and Fig. S1). In the second binding mode (BP2), the core anthraqui-
none structure is positioned similarly, but the carboxylate moiety
of the aminobenzoic acid is buried into S3. In the third binding
mode (BP3) the core anthraquinone structure adopts a similar
position as BP1 and BP2, but the C2 and C4 substituents are flipped:
C2 is ensconced into S4 and C3 is located in S2 (Fig. 2C). In each case,
To gain further insight on the stability of the binding mode of
each compound, MM-GBSA calculations were performed to deter-
mine the free energy of binding and its components (Table 2):
These include (i) a non-polar component (
which comprises the sum of the van der Waals potential energy
and non-polar solvation energy; (ii) a polar component ( EGBELE
ECOUL EGB), which consists of the Coulomb electrostatic energy
and the Generalized-Born solvation energy; (iii) an entropy com-
ponent SNM (Table 2). Interestingly, BP1 showed the most favor-
DENP = DEVDW + DESA),
D
=
D
+ D
D
able free energy of binding for each compound, namely ꢂ16, ꢂ11,
and ꢂ15 kcal/mol for 1, 2, and 3k, respectively. BP2, the binding
mode with the buried carboxylate, exhibited less stable values:

F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
4763
M)
Table 1
Table 1 (continued)
Activity of purchased derivatives
R1
R2
Compound
IC50
NA
(l
N
O
O
R1
N
O
3n (IPR-646)
O
R2
H
N
3o (IPR-632)
NA
R1
R2
Compound
IC50
21
(lM)
HN
N
3a (IPR-658)
H
O
ꢂ12, ꢂ6, and ꢂ10 kcal/mol. Even less stable energies were found
for BP3, with binding energies of ꢂ8, ꢂ10, and ꢂ8 kcal/mol for 1,
2, and 3k. These results, combined with the lower RMSDs that
were observed for this binding mode suggest that BP1 is the most
likely binding mode for 1 and its derivatives.
O
HN
N
3b (IPR-660)
7
MD simulations were also carried out for 3a, 3i, 6, 9 and 15c.
RMSDs for 3a (IPR-658) were lower than 3i (Supplementary data
Fig. S1). The greater stability of 3a was reflected by the lower
MM-GBSA binding free energy for 3i (Table 2). This trend is consis-
tent with our experimental data showing a higher affinity for 3a.
Compound 15c (and 15b) was conceived to probe the deep S1 cav-
ity (Fig. 2A). However, the docked structures of these compounds
reveal that the methyl groups at A6 and A7 prevented the com-
pounds from adopting a similar binding mode to that of the active
compounds. This was reflected in the higher free energy of binding
(ꢂ8 kcal/mol) for this compound.
HO
O
3c (IPR-661)
3d (IPR-662)
N
N
11
10
OH
O
HN
H
O
O
O
Compound 6 and 9 were designed to probe the cavity that
accommodates nitrogen group in 16 and the hydroxyl group of
10. The predicted structure of the compounds in the BP1 binding
pose is shown in Figure S2. The anthracene core of 9 adopts a similar
binding mode that is observed in 1 and other compounds. The iso-
propoxy group is ensconced into the smaller hydrophobic S4 pock-
et. The binding mode of 9 provided an explanation for the order of
magnitude loss of affinity observed for this compound. The large
benzyl group of 6 simply cannot be accommodated into the small
S4 cavity (Fig. S2) and the aromatic ring is found to instead occupy
the S3 binding pocket. This will likely result in severe clashes for
compounds that have a large substituent at S3, such as 1 and 2.
The trend in the MM-GBSA free energies was consistent with our
experimental data. Compound 6 was found to have the most unsta-
HN
3e (IPR-664)
3f (IPR-790)
18
36
N
N
HO
HN
O
OH
OH
HN
HN
HN
3g (IPR-796)
3h (IPR-804)
24
51
N
N
O
H
O
O
O
ble free energy (ꢂ7.5 kcal/mol; IC50 = 30
lM) compared with 9
(ꢂ8.2 kcal/mol; IC50 = 150 M). Finally, despite its lack of activity
O
O
l
in our biochemical studies, we attempted to predict a binding pose
for 15c (Fig. S2). Interestingly, we found that the two methyl groups
on the compound prevented the anthracene core from adopting the
same structure of the BP1 binding mode. Initially, these two methyl
groups were added to extend into the deeper pocket of the S1 cavity
of uPAR. Instead, it appears that the methyl groups clashed with the
side chains of the residues at the entrance of the pocket, preventing
the compound from adopting a productive binding mode for inhibi-
tion of the protein–protein interaction.
N
3i (IPR-809)
3j (IPR-811)
>100
20
H
O
O
O
HN
HN
O
O
N
N
H
3k (IPR-824)
3l (IPR-663)
3m (IPR-853)
5
2.4. Inhibitors impair lung cancer invasion, migration and
adhesion
H
O
HN
H
O
The importance of uPAR in promoting invasion and metastasis
is well-documented. To study its role in NSCLC cell invasion, small
interfering RNA (siRNA) knockdown of the receptor was carried out
in A549, H1299 and H460, and confirmed by Western blot analysis
(Fig. 3A). The effect on invasion was assessed using the Boyden
chamber apparatus, revealing that the knockdown of uPAR im-
paired NSCLC invasion through the Matrigel layer (Fig. 3B and C).
Nearly 40% inhibition of invasion was observed for each cell line.
When 3k is added to cells with silenced uPAR, an additional 30%
21
O
N
N
HN
>100
O
O

4764
F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
O
O
OH
Br
O
O
O
O
O
O
Br
Br
a
Br
b
H2N
H
CO2
Br
HN
H
CO2
4
5
6
Scheme 2. Reagents and conditions: (a) K2CO3, MEK/DMF (3:1), benzyl bromide; (b) Cu(OAc)2, Cu, KOAc amyl alcohol, 150–160 °C.
O
O
O
O
O
OH
O
O
Br
Br
Br
b
a
O
O
HN
HN
H
CO2
HN
O
OH
O
9
8
15a
Scheme 3. Reagents and conditions: (a) Ag2O, CHCl3, isopropyl iodide, rt; (b) 2 M NaOH (aq), MeOH, 70–80 °C.
were also evaluated as we have done previously by measuring their
effect on cell attachment to wells pre-coated with fibronectin and
vitronectin.38 Compound 2 inhibited adhesion in a concentration-
dependent manner (Fig. 4E). The inhibition plateaued at about
40% for both vitronectin and fibronectin. The less significant effect
on adhesion is not unexpected given that the compounds target
the uPARꢀuPA interface and do not directly impact the uPAR-inte-
grin interface, which is located at a different site on uPAR.
O
O
OH
O
O
O
Br
Br
a
HN
HN
HO
2C
HO
2C
11
10
2.5. Effect on proteolytic activity and signaling
Scheme 4. Reagents and conditions: (a) Ag2O, CHCl3, iodomethane, rt.
The effect of 2 and its derivatives on invasion prompted us to as-
sess whether the compounds affected uPA proteolytic activity in
cells using a fluorimetric assay described previously.39 While the
small molecule does not directly bind to uPA, it is believed that
the blocking of the uPARꢀuPA interaction at the cell surface will
likely result in less activation of uPA. Chloromethylketone (CMK)
showed irreversible inhibition of uPA activity with nearly 50% and
inhibition of invasion is observed (Fig. 3B and C). Previously, we
had shown that 1 caused no additional effects on invasion when
added to MDA-MB-231 cells with silenced uPAR.31 In this study,
however, the silencing of uPAR in the lung cancer cell lines was
not as complete, so that there may be some uPAR present at the
surface. Hence, the greater level of inhibition observed for com-
pound only could be due to inhibition of the residual uPAR recep-
tors. Another possibility is that 3k may bind to additional proteins
given the structural differences with 1. Hence 3k may modulate
the function of other proteins within the cell.
Further testing of invasion was done in a concentration-depen-
dent study in the three NSCLC cell lines using compound 2, whose
binding and inhibition profile was thoroughly characterized previ-
ously.31 As shown in Figure 4A and B, 2 inhibited H1299, A549, and
H460 invasion in a concentration-dependent manner with an IC50
60% inhibition at 10 and 50 lM of inhibitor, respectively (Fig. 5A).
Compound 2 was also tested at these concentrations and showed
20% and 50% inhibition, similar to the activity seen for chloromethyl
ketone. Another compound, namely UK122, which is thought to in-
hibit uPA catalytic activity, was also tested. However, significantly
less activity of the compound was detected in our hands. A concen-
tration of 400
tion. It is interesting that IPR-803 inhibited the uPA proteolytic
activity by nearly 50% at a concentration of 50 M (Fig. 5A). Even
at a concentration of 10 M, IPR-803 inhibited by 25%. Unlike
lM had to be attained to detect nearly 50% of inhibi-
l
that is estimated at about 10
in H460 cells as evidenced by nearly complete inhibition at
50 M. It showed less potency in H1299 cells, as illustrated by only
20% inhibition at 25 M, compared with 80% for the other cell
l
M. The inhibitor was more potent
l
CMK or UK122, IPR-803 is not expected to directly inhibit uPA
catalytic activity. Its effects are indirect, by blocking the binding
of pro-uPA to uPAR, and therefore inhibiting its activation at the cell
surface.uPA is known to lead to the activation of a number of pro-
teinases, including matrix metalloproteinases (MMPs), which are
well-known to degrade components of the ECM such as collagen.
Numerous studies have targeted MMPs with small molecules as a
strategy to block invasion.40 To further probe the effect of 2 on pro-
teolytic activity, matrix metalloproteinase activity is monitored
with gelatin zymography in H1299 cells (Fig. 5B and C). Interest-
l
l
lines. The IC50 for inhibition of cell proliferation by 2 (Fig. 4F),
which is nearly four times as large as the IC50 in our ELISA and
invasion assays, suggesting that the effects on invasion are unre-
lated to cytotoxicity.
Studies have shown that uPAR engages cell surface integrins,
which are responsible for cell attachment to ECM components and
migration.3,5,7 First, we evaluated the effects of 2 on H1299 migra-
tion using the Boyden chamber apparatus (Fig. 4C and D). The com-
pound inhibited migration in a concentration-dependent manner
ingly, an IC50 of nearly 50 lM is observed for 2, which correlated
well with the inhibition levels observed for uPA activity. It has been
with an IC50 = 50 lM. The effects of the compounds on adhesion
suggested in the past that activation of uPA results in plasmin acti-

F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
4765
O
OH
H
O
OH
O
O
O
R1
R2
R1
R2
Br
b
a
substituted
butadiene
O
Br
O
12
=
H
13a: R1
R
,
=
H
14a: R1
R
2
,
2
=
=
H
13b: R1 Me R
,
=
=
14b: R1 Me R
H
2
,
2
=
13c: R1
R
,
Me
=
Me
2
14c: R1
R
2
,
O
O
OH
R1
R2
Br
c
H2N
H
CO2
H
CO2
HN
=
15a: R1
R
H
,
2
=
=
H
15b: R1 Me R
,
2
=
Me
2
15c: R1
R
,
Scheme 5. Synthesis of IPR-456 derivatives. Reagents and conditions: (a) AlCl3, DCM, rt; (b) Br2, NaOAc, AcOH, reflux; (c) Cu(OAc)2, Cu, KOAc amyl alcohol, 150–160 °C.
NH2
O
O
O
O
OH
NH2
O
O
r
B
Br
Br
H
CO2
HN
HN
HN
HO
2C
H
O2C
IPR-861 (10)
IPR-630 (16)
IPR-855 (17)
Ph
H
O
O
O
O
O
O
Br
O
O
O
Br
Br
H
CO2
HN
H
CO2
HN
H
CO2
HN
IPR-863 (6)
IPR-864 (9)
IPR-865 (15b)
O
O
O
OH
Br
Br
H
CO2
HN
O
HN
O
H
O2C
IPR-862 (15c)
IPR-860 (11)
Scheme 6. Synthesis of IPR-456 derivatives.
vation, which in turn leads to activation of MMPs. Hence, inhibition
of uPA activity by 2 may affect MMP activation. Another possibility
is that 2 inhibits uPAR-mediated signaling through integrins and
other cell surface receptors that have been shown to promote
activation. It is also possible that 2 and its derivatives inhibit other
targets that affect MMP activity.
The interaction of uPAR with integrins and other cell surface
receptors suggests that blocking these interactions may result in
impairment of cell signaling. Three pathways in particular, namely
ERK, FAK and Src have been implicated with integrin signaling. ERK
signaling is responsible for promoting a number of cellular pro-
cesses that include proliferation, differentiation and survival. FAK
is constitutively associated with b-integrin subunits of integrin
receptors. Src, which is upstream of ERK, is also implicated in inte-
grin signaling. Despite the fact that 2 inhibits the uPARꢀuPA inter-
action, the effect of the compound on cell migration and adhesion
prompted us to test its effect on cell signaling mediated by inte-
grins by Western blot analysis. As shown in Figure 5D, there did
not appear to be any effects on ERK phosphorylation. This is consis-
tent with the fact that 2 binds to the uPA cavity, which is far from
the integrin-binding site on uPAR.3,47 Therefore 2 is unlikely to im-
pair the uPAR-integrin interaction, which has been shown to affect
ERK signaling in previous studies using peptides and antibod-
ies.3,12,41–49

4766
F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
Structure-Based analysis
5
A
B
4
3
2
1
0
BP1
BP2
BP3
0
2
4
6
8
10
Time (ns)
Leu-168
O
Leu-144
O
C
O
His-166
O
Val-125
NH
NH
H
N
NH
Asn-157
O
H
N
S4
H
N
O
O
O
N
N
NH
N
Leu-55
S2
S1
HN
His-251
HN
HN
O
HN
2
N
N
H
O
Leu-66
O
OH
S3
O
HO
NH2
Asp-254
-O
N
O
HN
O
H
NH
Leu-66
HN
O
N
H
O
Arg-53
Glu-68
Figure 2. Structure-based analysis of uPAR/2 interactions. (A) Stereoview of the top three binding modes of compound 2 docked onto the uPA binding pocket on uPAR. The
compound and the residues it interacts with were shown in sticks (carbon atoms in compound and protein were colored in yellow and green, respectively; nitrogens in blue
and oxygens in red). The uPAR backbone was shown as grey ribbons. (B) The MD structures were aligned onto the first snapshot in the trajectory. The RMSD was calculated
based on the heavy atoms of the compound. It was reported as the average of five independent trajectories in length of 10 ns. (C) The 2D schematic illustration of compound 2
interacting with residues on the uPAR pocket. Compound explores four sub-pockets (depicted with colored curves and named as S1, S2, S3 and S4, respectively) on the uPA
binding pocket.
Table 2
Compound binding energy to uPAR calculated by MM-GBSA (in unit of kcal molꢂ1
)
Compound
Pose
D
EVDW
D
EGBELE
D
ESA
D
EGBTOT
TD
SNM
DGGB
1
1
1
3k
3k
3k
2
2
2
3a
3i
15c
6
BP1
BP2
BP3
BP1
BP2
BP3
BP1
BP2
BP3
—
ꢂ53.1 0.2
ꢂ49.8 0.3
ꢂ48.1 0.2
ꢂ57.4 0.2
ꢂ46.7 0.5
ꢂ48.6 0.2
ꢂ53.8 0.2
ꢂ42 0.4
21.0 0.0
21.9 0.0
26.1 0.0
24.8 0.0
21.2 0.0
25.2 0.0
25.6 0.0
20.2 0.1
25.1 0.0
20.0 0.0
16.8 0.0
24.0 0.0
27.1 0.0
23.3 0.0
ꢂ6.2 0.0
ꢂ6.0 0.0
ꢂ5.8 0.0
ꢂ6.2 0.0
ꢂ5.8 0.0
ꢂ5.8 0.0
ꢂ6.0 0.0
ꢂ5.6 0.0
ꢂ5.7 0.0
ꢂ5.5 0.0
ꢂ5.5 0.0
ꢂ5.7 0.0
ꢂ6.1 0.0
ꢂ5.7 0.0
ꢂ39.2 0.2
ꢂ34.6 0.3
ꢂ28.6 0.2
ꢂ39.6 0.2
ꢂ31.9 0.4
ꢂ29.9 0.2
ꢂ35.0 0.2
ꢂ28.1 0.3
ꢂ31.1 0.1
ꢂ31.7 0.2
ꢂ29.9 0.3
ꢂ29.5 0.1
ꢂ29.3 0.2
ꢂ31.4 0.2
ꢂ22.9 0.6
ꢂ22.9 0.6
ꢂ20.8 0.6
ꢂ25.0 0.5
ꢂ22.3 0.6
ꢂ22.4 0.6
ꢂ24.0 0.6
ꢂ21.7 0.6
ꢂ21.4 0.6
ꢂ21.2 0.6
ꢂ21.8 0.6
ꢂ21.2 0.6
ꢂ21.8 0.6
ꢂ23.3 0.6
ꢂ16.3 0.6
ꢂ11.7 0.6
ꢂ7.8 0.6
ꢂ14.7 0.5
ꢂ9.6 0.7
ꢂ7.5 0.6
ꢂ11.0 0.6
ꢂ6.4 0.7
ꢂ9.7 0.6
ꢂ10.5 0.6
ꢂ8.1 0.7
ꢂ8.3 0.6
ꢂ7.5 0.6
ꢂ8.2 0.6
ꢂ49.8 0.1
ꢂ45.6 0.2
ꢂ40.6 0.3
ꢂ47.1 0.2
ꢂ49.5 0.2
ꢂ48.4 0.2
—
—
—
—
9
D
EVDW, van der Waals potential energy;
Born equation; ESA, nonpolar solvent contribution to solvation free energy;
analysis; GGB, the calculated free energy of binding using GB model.
D
EGBELE, electrostatic contributions to the binding energy, of which the polar solvent contributions were calculated with Generalized
D
D
EGBTOT, the sum of EVDW EGBELE and ESA; T SNM, entropy calculated with normal mode
D
,
D
D
D
D
2.6. Discussion
sult of the ring opening does not play a significant role in binding.
This was evidenced by the minimal loss of activity upon its con-
version to a hydroxyl group. Further functionalization of this hy-
droxyl group to methoxy, isopropoxy and benzyloxy resulted in
loss of inhibition suggesting that these groups engage a smaller
pocket on uPAR. This was confirmed by extensive structure-based
computational studies involving molecular docking, molecular
dynamics simulations and free energy calculations. Functionaliza-
tion of the A ring of 1 derivatives by introduction of methyl
groups resulted in significant loss of activity. The predicted bind-
We build on our previous discovery of 1, a small molecule that
binds to uPAR and inhibits its protein interactions, to acquire a
series of derivatives from commercial sources. These compounds
varied strictly at C2 and C4, so synthesis was used to prepare an
additional series of derivatives to probe the role of other groups
on 1. Beyond confirming the importance of the carboxylate moi-
ety at C2, our activity data showed that the isoxazole ring of 1
is not critical for binding, and the amine group generated as a re-

F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
4767
A
B
C
120
uPAR
β-actin
A549
H460
H1299
100
80
H460
A549
H1299
60
40
20
0
DMSO
siRNA
siRN
A+
uPA
3
k
R
DMSO
siRNAuPAR
siRNA+3k
Figure 3. Invasion in the absence (control) or with 3k for lung cancer cells with siRNA knockdown of uPAR. (A) H1299, H460 and A549 cells were transfected with siRNAs for
uPAR or control siRNA (control). After 48 h, cells were lysed and analyzed by immunoblotting. (B) Representative experimental cells from control and in the presence of uPAR
siRNA with or without 3k were photographed for the Boyden chamber assays (ꢁ200) to illustrate the effect of 3k on invasion as quantified in (C).
ing mode suggested that these groups clash with residues at the
opening of the deeper S1 cavity on uPAR and prevent the com-
pounds from adopting the necessary binding mode to inhibit the
protein interaction. The computational studies revealed that
the benzoic acid is ensconced within a hydrophobic pocket near
the mouth of uPA binding cavity at the S3 binding pocket, engag-
ing nearby Arg53 in a salt-bridge interaction. The substituent at
C2 occupied another hydrophobic, but more solvent-exposed
pocket on uPAR (S2). Cellular assays were performed to assess
the activity of 2 in NSCLC cell lines. Compound 2 was shown to
block lung cancer cell invasion in a concentration-dependent
manner. The compound also inhibited lung cancer cell migration
in a concentration-dependent manner and moderately inhibited
adhesion suggesting a potential effect on the interaction between
uPAR and integrins. However, signaling studies revealed that the
compounds did not affect ERK, FAK or Src signaling, signifying
that 2 may not be affecting integrin signaling. The effects ob-
served on adhesion and migration could therefore be indirect, per-
haps due to inhibition of the breakdown of the ECM, providing
less space for the cells to migrate. The lack of effect on ERK signal-
ing, which is affected by a large number of proteins, suggests that
the compounds are specific to the uPARꢀuPA interaction. Future
efforts are concentrating on further improving the potency of 2
in tumor invasion and metastasis.
3.2. Reagents
Phospho-p44/42 MAPK and p44/42 MAPK, phospho-FAK
(Tyr397), FAK; phospho-Src family (Tyr416) and Src antibodies
were purchased from Cell Signaling Technology, Inc. (Danvers,
MA). Actin antibody, uPAR siRNA and control siRNA were from
Santa Cruz Biotechnology (Santa Cruz, CA). UK122 and Glu-Gly-
Arg-chloromethyl ketone were from Calbiochem (San Diego, CA)
and Z-Gly-Gly-Arg-AMC.HCl was from Bachem (Torrance, CA).
3.3. Microtiter-based ELISA
We used our previously developed microtiter-based ELISA.31
Briefly, medium- to high-binding microplates were coated and
incubated for 2 h at 4 °C with 100 lL of 2 l
g mLꢂ1 of uPA ATF
in 1ꢁ PBS for immobilization. A 1:1 mixture of Superblock buffer
in PBS (Thermo Fisher Scientific, Inc.) with 0.04 M NaH2PO4 and
0.3 M NaCl buffer was used for blocking. Following incubation
and washing steps, serially diluted compounds were added with
suPAR277 mixtures for 30 min. Following washing, biotinylated
uPAR antibody (R&D Systems) in 1% BSA 1ꢁ PBS buffer was added
to the wells (100 lL/well) and incubated for 1 h to allow for the
detection of bound uPAR. After that, streptavidin-peroxidase in
1% BSA 1ꢁ PBS buffer was added for 20 min. The signal obtained
in the presence of TMB in phosphate-citrate buffer (0.2 M Na2H-
PO4 and 0.1 M citrate, pH 5) and hydrogen peroxide was stopped
by adding H2SO4 solution and detected using an EnVisionÒ Multi-
label Plate Reader (PerkinElmer).
3. Materials and methods
3.1. Cell culture
3.4. Proliferation assay
H1299 and H460 cells were cultured in RPMI-1640 medium
(Cellgro, Manassas, VA). A549 cells were cultured in Dulbecco’s
Modified Eagle Medium (Cellgro, Manassas, VA). Each medium
was supplemented with 10% FBS and 1% penicillin/streptomycin
in a 5% CO2 atmosphere at 37 °C.
Cells were cultured at 37 °C in 10% FBS-DMEM or RPMI-1640
medium containing various amounts of compounds. 20 mM com-
pound stock in 100% DMSO was serially diluted and added into

4768
F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
120
100
80
60
40
20
0
A
H1299
A549
H460
0
25
50
100
[2] (μM)
120
100
80
60
40
20
0
C
E
H1299 Migration
0
25
[2] (µM)
50
100
180
160
140
120
100
80
H1299 Adhesion
2
FN
VN
70
60
50
40
30
20
10
0
F
60
40
20
0
0
12.5
25
[2] (µM)
50
100
H460
H1299
A549
Figure 4. Compound 2 blocks invasion, migration and adhesion of H1299 cancer cells. (A, B) Invasion of H1299, H460 and A549 cells at increasing concentration of IPR-803, as
indicated. Ten fields of each membrane were counted for cell numbers (ꢁ200). These data represent the average (SD of three independent experiments). (C, D) Migration of
H1299 cells at increasing concentration of 2, as indicated. Ten fields on each membrane were counted for cell numbers (ꢁ200). These data represent the average (SD of three
independent experiments). (E) The adhesion of H1299 cells to ECM components fibronectin (FN) or vitronectin (VN) in the absence or presence of 2 are shown as indicated.
The numbers of attached cells were quantified by MTT assay. (F) IC50 for inhibition of A549, H460, and H1299 cell proliferation by 2.
each well of a 96-well plate. Then treated cells were incubated for
3 days. Viable cells were quantified by MTT assay at absorbance of
570 and 630 nm as described previously.50,51
5 ꢁ 104 cells in 500
lL (for invasion) or 250 lL (for migration) of
0.1% FBS media containing the same compounds or 1% DMSO (as
control) were added to the upper chambers. We incubated the
chambers for 16 h at 37 °C, 5% CO2. Non-invaded or non-migrated
cells were removed from the upper chamber with a cotton swab.
The invaded or migrated cells were fixed with 100% methanol and
then stained with Hematoxylin Stain Harris Modified Method
(Fisher, SH30-500D). We washed the filters with water three times.
Filters were air-dried, and the invaded or migrated cells were
counted in ten randomly selected microscopic fields (200 magnifi-
cation). The experiment was performed in triplicate per group and
shown by mean SD.
3.5. Invasion and migration assays
Assays were performed using BD BioCoat Matrigel Invasion
Chamber (cat. 354480, BD Biosciences, Bedford, MA) and Corning
Transwell Permeable Support (Cat 3422, Corning Incorporated,
Corning, NY) as described previously.31,50,52,53 In brief, the under-
surfaces of the inserts were coated with 30 lg/mL of fibronectin
(Sigma, F2006) in PBS at 4 °C overnight. The inserts were washed
with PBS once. Serum-free medium (0.5 mL) was separately added
to the upper and lower chambers to equilibrate at 37 °C, 5% CO2 for
2 h. After starvation with serum-free DMEM for 4 h, subconfluent
cells were trypsinized and resuspended in 0.1% FBS media. Five
hundred microliters of 10% FBS media containing various concen-
trations of compounds or DMSO were added to the lower chambers.
3.6. Adhesion assay
Quantitative cell adhesion assays were carried out as described
previously7,31,50 in 96-well microtiter plates, which were coated
with either 15 lg/mL fibronectin (Sigma) or 15 lg/mL vitronectin

F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
4769
A
C
100
80
60
40
20
0
A549
Chloromethyl ketone
UK122
2
D
200
400
10
50
10
50
Concentration (µM)
B
120
H1299
100
80
60
40
20
0
0
10
20
50
100
200
[2] (µM)
Figure 5. Proteolytic activity and signaling. (A) Fluorescence assays of cell surface-bound uPA activity in A549 cells. Cells were pre-incubated for 30 min with a range of
concentrations of UK122, chloromethyl ketone or 2 on ice. Cells were washed, fluorogenic uPA-specific substrate was added and fluorescence measurements were recorded
immediately. Initial rates of change in fluorescence after subtraction of background fluorescence (cells only) are presented as a percentage of control (no test compound
added). Values represent means SD (n = 3). (B, C) Effect on ECM degradation. Gelatin zymography analysis for H1299 cells with increasing concentration of 2 (D) Cell
signaling study. H460 and A549 cells were treated with 1, 10 and 50 lM compound(s) for 30 min, then immunoblotted with phospho-p44/42 MAPK and p44/42 MAPK,
phospho-FAK (Tyr397), FAK; phospho-Src family (Tyr416), Src and actin, respectively.
(R&D Systems, Minneapolis, MN) overnight at 4 °C. Coated and un-
coated control wells were blocked for 1 h with 2% bovine serum
albumin (BSA) at 37 °C. Cells were split 1 day prior to the experi-
ment to achieve a subconfluent culture. Briefly, H1299 cells were
collected with trypsin, washed twice with serum-free medium,
atinolytic degradation appeared as transparent bands on the blue
stained background of the gel. Data were quantified using Image J.
3.8. siRNA knockdown
and 2 ꢁ 104 cells in 100
lL of serum-free medium containing var-
H1299, H460 and A549 cells were transfected with uPAR siRNA
or control siRNA for 48 h. Cells were then collected, and total cell
lysates were prepared in standard RIPA extraction buffer contain-
ing aprotinin and phenyl-methyl-sulfonyl-fluoride as previously
ious compounds were added to each well, quadruplicate per group
and incubated for 90 min at 37 °C. The wells were washed, and the
number of adherent cells was quantified by MTT assay at 570 nm
and 630 nm. The results were shown by means SD.
described.31 A 20
lg sample of protein from these samples was
separated by 12% SDS–PAGE and transferred to nitrocellulose
membranes (Amersham, Arlington Heights, IL). The membranes
were immunoprobed with antibodies against biotin-uPAR at 4 °C
overnight. Next, membranes were treated with the appropriate
HRP-conjugated secondary antibody and then developed according
to enhanced chemi-luminescence protocol (Thermo Scientific,
Rockford, IL). Membranes were stripped and reprobed with a
monoclonal antibody against actin as a loading control.
3.7. Gelatin zymography
Zymography experiments were performed as described previ-
ously.50,54 H1299 cells were treated with different concentrations
of 2 in serum free medium for 24 h, the conditioned medium
was collected and concentrated by Amicon Ultra centrifugal filter
units (Millipore, #UFC500324), and proteins were normalized
and electrophoresed on 7.5% sodium dodecyl sulfate (SDS) poly-
acrylamide gels containing 1 mg/mL gelatin. After electrophoresis,
the gel was washed twice in 50 mM Tris–HCl (pH 7.6) containing
5 mM CaCl2 and 2.5% Triton X-100 for 30 min at room temperature
and incubated in buffer that contained 50 mM Tris–HCl (pH 7.6),
200 mM NaCl, 10 mM CaCl2, 1 mM ZnCl2 at 37 °C for 36 h. Then,
the gels were stained with 0.05% Coomassie brilliant blue (CBB)
and destained with 30% methanol in 10% acetic acid. Areas of gel-
3.9. Immunoblots
H460 and A549 cells were treated with 1, 10 and 50 lM of
IPR803 and 1% DMSO for 30 min. Cell lysates were harvested and
separated by SDS–PAGE as mentioned above. For immunoblotting,
the first antibody was phospho-p44/42 MAPK (Thr202/Tyr204),
p44/42 MAPK; phospho-FAK (Tyr397), FAK; and phos-pho-Src

4770
F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
family (Tyr416) or Src mAb (Cell Signaling Technology, Inc.,
Danvers, MA) in 5% nonfat milk-TBST at 4 °C overnight. Secondary
antibody consisted of goat anti-rabbit IgG or goat anti-mouse IgG
HRP 1:3000 in 5% nonfat milk-TBST at room for 1 h and then
developed as mentioned above.
culations, and normal mode analyses for energy and entropy
calculation. In total, 500 snapshots were extracted evenly from the
production trajectories for energy and entropy analysis. The
MM-PBSA perl scripts in Amber56 were employed to compute bind-
ing energy components and to decompose the binding energy on a
per residue basis. The latter provides a useful insight on the relative
importance of residues on the pocket to the binding of a ligand to
uPAR.
3.10. Fluorometric HMW-uPA activity assays
Cell surface-bound uPA activity was measured using the fluoro-
genic substrate Z-Gly-Gly-Arg-AMC as previously described (De
Souza 2011). The excitation wavelength of the substrate is
340 nm and the emission wavelength range is 460 nm. A549 cells
were trypsinized, resuspended in cold PBS at 1 ꢁ 106 cells mLꢂ1
and incubated with test compounds for 30 min at 4 °C. Cells were
then washed twice, transferred to a fluor plate, and PBS containing
a final concentration of 0.1 mM Z-Gly-Gly-Arg-AMC as the fluoro-
genic substrate was added. Fluorescence emission was measured
immediately using an EnvisionÒ Multilabel Plate Reader (PerkinEl-
mer). Data was recorded at 2 min intervals over a period of 30 min.
A control sample with no test compound was included. We calcu-
lated the rate of change in fluorescence minꢂ1 using the linear
region of a graph where fluorescence was plotted against time.
3.12. Cloning, expression and purification of uPAR
We successfully cloned, expressed and purified uPAR. From 1 L
of culture, we express nearly 12 mg of protein. Briefly, a truncated,
soluble form of human uPAR (suPAR, amino acids 1–277) was ex-
pressed in stably transfected Drosophila S2 cells using the Drosoph-
ila Expression System (Invitrogen) and purified as described
previously.
3.13. Fluorescence polarization
For direct binding studies of compounds using FP, varying con-
centrations of suPAR277 protein were titrated against intrinsically
fluorescent compound with a final concentration of 1
l
M in 1ꢁ PBS
3.11. MD simulations and analysis
with 0.01% Triton X-100. The inhibitor-protein mix was incubated
for 15 min to allow sufficient binding. Polarized fluorescence
intensities were then measured using EnVisionÒ Multilabel Plate
Reader (PerkinElmer) with excitation and emission wavelengths
of 531 and 595 nm, respectively. The reactions were carried out
in duplicates.
The uPAR 3D structure was extracted from a uPAR crystal com-
plex (PDB code: 3BT2). It was then prepared with the protein mod-
eling package from Schrödinger, LLC (New York, NY) to add missing
residues and atoms. Hydrogen atoms were added and optimized to
a neutral environment. Disulfide bonds were created among
Cysteine residues with short distance between sulfur atoms. The
compounds were docked onto the uPA binding pocket on uPAR
with autodock vina (version 1.1.2). The most favorable binding
modes were returned. The compound binding modes were then
visually inspected in PyMOL.55 For 3 compounds (1, 2 and 3k),
three most probable binding modes for each compound were
selected to subjected to explicit-solvent MD simulations. We
retained the most commonly observed binding mode for other
compounds. The following protocol for setting up and running
the molecular dynamics simulations applies to all cases. The
AMBER956 ff99SB force field for protein and GAFF57 for small mole-
cules were assigned. The atomic charges of the compound were
determined by using the AM1-BCC methodology implemented in
the antechamber program. The program Leap from the AMBER pack-
age was used to neutralize the complexes. The complexes of uPAR/
compound were immersed in a box of TIP3P58 water molecules
such that no atom in the complex was within 14 Å from any side
of the box. All bonds involving hydrogen atoms were constrained
3.14. Synthesis
All basic chemicals were purchased from commercially avail-
able sources and used as received. 1-Amino-2,4-dibromoanthr-
aquinone and 5-hydroxy-1,4-naphthoquinone were purchased
from Alfa Aesar. 1-hydroxyanthraquinone was purchased from
TCI America. Column chromatography was carried out with silica
gel (25–63 TCI Ament 6520 Accurate Mass Q-TOF instrument. 1H
NMR was recorded in CDCl3, MeOH, or DMSO on a Bruker
500 MHz spectrometer. HP-LCMS was carried out on a Agilent
1100 LC/MSD fitted with a Eclipse XBD-C18 (4.6 ꢁ 150 mm) col-
umn eluting at 1.0 ml/min employing a gradient of (acetonitrile/
methanol)/water (each containing 5 mM NH4OAc) from 70% to
100% acetonitrile/methanol over 15 min and holding at 100% ace-
tonitrile/methanol for 2 min. Chemical shifts are reported in ppm
using residual CHCl3, MeOH, or DMSO as internal references. Pre-
parative HPLC was carried out using a X-Bridge Ost C18 2.5 lm col-
umn on a Waters 1525 Binary HPLC pump. All final compounds are
95% or greater purity from LC/MS, except for some intermediates or
as otherwise indicated.
59
by using the SHAKE algorithm, and a 2 fs time step was used. The
particle mesh Ewald60 method was used to treat long-range
electrostatics. Water molecules were first energy minimized and
equilibrated by running a short simulation with the complex fixed
by using Cartesian restraints. This was followed by a series of
energy minimizations in which the Cartesian restraints were grad-
ually relaxed from 500 kcal Åꢂ2 to 0 kcal Åꢂ2, and the system was
subsequently gradually heated to 300 K via a 48 ps molecular
dynamics run. Production simulations were carried out in 5 inde-
pendent runs, starting from the same structure but different initial
atomic velocities. The PMEMD in AMBER9 was employed for production
runs. MD snapshots were saved every 2 ps. A trajectory of 10 ns
was collected for each run. The initial 2 ns on each trajectory was
discarded to ensure only equilibrated structures are used in the
analysis.
3.14.1. 1-Hydroxy-7-methylanthracene-9,10-dione and 1-
hydroxy-6-methylanthracene-9,10-dione (13b)
5-Hydroxy-1,4-naphthoquinone (1.72 mmol, 300 mg) was dis-
solved in dry DCM (9 mL). The substituted butadiene (1.5 equiv)
was added. Under Argon the solution was cooled to 0 °C and AlCl3
(0.34 mmol, 45 mg) was added. The reaction was stirred at 0 °C for
1 h then warmed to ambient temperature. The reaction mixture
was then heated to reflux overnight (ꢃ16 h). After the reaction
was complete as judged by TLC, the reaction was cooled to ambient
temperature and triethylamine (8.6 mmol, 1.20 mL) was added
and left to stir vigorously overnight. The reaction was then washed
with 1 N HCl, satd aq K2CO3 and brine. The organic layer was dried
over anhydrous MgSO4 and concentrated in vacuo. The crude solid
was purified by column chromatography (eluting with 20% Et2O/
hexanes) to give the Diels–Alder product.33 Yield: 123 mg (30%)
The method of MM-PBSA/GBSA for determining the binding free
energy has been described in the past.61–63 It combines molecular
mechanics, Generalized-Born electrostatics, surface-accessible cal-

F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
4771
as a yellow solid. HRMS calcd for C15H9O3 (MꢂH)ꢂ 237.0557, found
237.0569. 1H NMR (500 MHz, CDCl3) d 2.44 (s, 3H), 7.18 (ddd,
J = 8.51, 1.26, 0.63 Hz, 1H), 7.47 (m, 1H), 7.56 (m, 1H), 7.68 (dt,
J = 7.57, 1.73 Hz, 1H), 7.92 (s, 1H), 8.02 (dd, J = 7.88, 1.58 Hz, 1H),
12.47, 12.52 (syn and anti with respect to Me, 2s, 1H). 13C NMR
(100 MHz, CDCl3) d 188.6, 188.2, 182.3, 181.8, 162.3, 145.8,
145.2, 136.5, 136.3, 135.3, 134.8, 133.3, 133.2, 132.8, 131.1,
130.7, 127.5, 127.4, 126.9, 124.1, 119.2, 116.0, 34.6, 31.5, 25.2,
22.6, 21.8, 20.6, 14.0.
J = 8 Hz, 1H), 7.84–7.95 (m, 5H), 8.39 (t, J = 8.2 Hz, 2H). LC/MS:
rt = 4.62 min.
3.14.7. 2-((3-Bromo-4-hydroxy-9,10-dioxo-9,10-dihydroanthra
cen-1-yl)amino)benzoic acid (10)
Crude yield: 125 mg (37%) as a purple/blue solid. HRMS calcd
for C21H11BrNO5 (MꢂH)ꢂ 435.9826, found 435.9855. 1H NMR
(500 MHz, CDCl3) d 7.12–7.19 (m, 1H), 7.43–7.49 (m, 2H), 7.82–
7.97 (m, 3H), 8.04 (s, 1H), 8.30–8.43 (m, 2H). LC/MS:
rt = 3.68 min.
3.14.2. 1-Hydroxy-6,7-dimethylanthracene-9,10-dione (13c)
Yield: 115 mg (26%) as a yellow solid. HRMS calcd for C16H11O3
(MꢂH)ꢂ 251.0714, found 251.0726. 1H NMR (500 MHz, CDCl3) d
2.43 (s, 6H), 7.28 (dd, J = 8.51, 0.95 Hz, 1H), 7.64 (dd, J = 7.88,
7.88 Hz, 1H), 7.80 (dd, J = 7.57, 1.26 Hz, 1H), 8.03 (d, J = 5.04 Hz,
2H), 12.68 (s, 1H).
3.14.8. 3-((3-Bromo-4-hydroxy-6-methyl-9,10-dioxo-9,10-
dihydroanthracen-1-yl)amino)benzoic acid (15b)
Crude yield: 68 mg (60%) as a purple solid. HRMS calcd for
C
22H13BrNO5 (MꢂH)ꢂ 449.9983, found 450.0000. 1H NMR
(500 MHz, CDCl3) d 2.55, 2.56 (2s syn and anti with respect to hy-
droxyl group, 4H), 7.36–7.41 (m, 1H), 7.48 (t, J = 7.8 Hz, 1H), 7.69
(d, J = 8 Hz, 0.58H), 7.74 (d, J = 8 Hz, 1H), 7.81–7.90 (m, 3H),
8.16–8.20 (m, 1H), 8.26 (t, J = 7.6 Hz, 1H). LC/MS: rt = 5.04 min.
3.14.3. 2,4-Dibromo-1-hydroxy-7-methylanthracene-9,10-dione
and 2,4-dibromo-1-hydroxy-6-methylanthracene-9,10-dione
(14b)
To a stirred slurry of anthraquinone (1.0 equiv) and acetic acid
(2 M) at ambient temperature was added NaOAc (3.0 equiv) fol-
lowed by bromine (3.0 equiv). The reaction mixture was refluxed
overnight (ꢃ16 h). The mixture was cooled to ambient tempera-
ture and diluted with water. The precipitate was filtered off and
washed with additional water. The solid was dried in vacuo to give
the desired brominated product.34 Yield: 320 mg (quantitative) as
3.14.9. 3-((3-Bromo-4-hydroxy-6,7-dimethyl-9,10-dioxo-9,10-
dihydroanthracen-1-yl)amino)benzoic acid (15c)
Crude yield: 39 mg (35%) as a purple solid. HRMS calcd for
C
23H15BrNO5 (MꢂH)ꢂ 464.0139, found 464.0159. 1H NMR
(500 MHz, MeOD) d 2.21 (s, 1H), 2.46 (d, J = 4 Hz, 6H), 7.36–7.38
(m, 1H), 7.48 (t, J = 8 Hz, 1H), 7.78 (s, 1H), 7.84 (d, J = 7.5 Hz, 1H),
7.88 (s, 1H), 8.07 (app d, J = 3.5 Hz, 2H). LC/MS: rt = 6.43 min.
a
reddish-yellow solid. HRMS calcd for C
15H9Br2O3 (M+H)+
394.8913, found 394.8928. 1H NMR (500 MHz, CDCl3) d 2.54 (d,
J = 3.78 Hz, 3H), 7.60 (d, J = 7.88 Hz, 1H), 7.64 (d, J = 7.88 Hz, 1H),
8.06 (s, 1H), 8.15 (d, J = 3.15 Hz, 1H), 8.16 (d, J = 3.15 Hz, 1H) 8.19
(m, 1H), 8.21 (m, 1H), 14.04 (s, 1H), 14.09 (s, 1H).
3.14.10. 2-((4-Amino-3-bromo-9,10-dioxo-9,10-dihydroanthra
cen-1-yl)amino)benzoic acid (16)
1-Amino-2,4-dibromoanthraquinone (0.13 mmol, 50 mg), cop-
per carbonate (0.03 mmol, 4 mg), potassium acetate (0.46 mmol,
45 mg), and anthranilic acid (0.14 mmol, 19 mg) were refluxed
for 12 h at 140 °C. The reaction was cooled to 80 °C and metha-
nol was added. The blue solid was filtered off, washed with
methanol, and air dried. The crude solid was recrystallized from
acetic acid to give a bluish solid (46 mg, 81%).64 HRMS calcd for
3.14.4. 2,4-Dibromo-1-hydroxy-6,7-dimethylanthracene-9,10-
dione (14c)
Yield: 228 mg (97%) as a reddish-yellow solid. HRMS calcd for
C
16H11Br2O3 (M+H)+ 408.9069, found 408.9075. 1H NMR
(500 MHz, CDCl3) d 2.42 (d, J = 4.10 Hz, 6H), 7.97 (d, J = 4.41 Hz,
2H), 8.16 (s, 1H), 14.09 (s, 1H).
C
21H14BrN2O4 (M+H)+ 437.0137, found 437.0143. 1H NMR
(500 MHz, DMSO) 3.37 (s, 1H), 5.74 (s, 1H), 7.07 (t,
d
J = 7.09 Hz, 1H), 7.41 (d, J = 7.88 Hz, 1H), 7.52 (t, J = 6.94 Hz,
1H), 7.84 (s, 2H), 7.95 (m, 2H), 8.19 (d, J = 8.20 Hz, 2H), 12.17
(s, 1H), 13.13 (s, 1H). 13C NMR (100 MHz, DMSO) d 54.9, 110.5,
115.4, 119.7, 120.1, 120.5, 121.6, 126.0, 130.1, 131.8, 133.2,
133.3, 133.9, 133.5, 133.6, 136.9, 141.8, 144.5, 167.8, 182.7,
182.8. LC/MS: rt = 2.06 min.
3.14.5. 2,4-Dibromo-1-hydroxyanthracene-9,10-dione (14a)
Yield: 2.93 g (86%) as an orange solid. HRMS calcd for
C
14H5Br2O3 (MꢂH)ꢂ 378.8611, found 378.8617. 1H NMR
(500 MHz, CDCl3) d 7.65–7.78 (m, 2H), 8.08 (s, 1H), 8.11–8.21 (m,
2H), 13.87 (s, 1H). 13C NMR (100 MHz, CDCl3) d 187.7, 180.2,
159.2, 145.6, 135.2, 133.9, 133.5, 131.2, 127.4, 126.5, 118.9,
117.3, 112.4, 104.4.
3.14.11. 3-((4-Amino-3-bromo-9,10-dioxo-9,10-dihydroanthra
cen-1-yl)amino)benzoic acid (17)
3.14.6. 3-((3-Bromo-4-hydroxy-9,10-dioxo-9,10-dihydroanthra
cen-1-yl)amino)benzoic acid (7)
To 1-amino-2,4-dibromoanthraquinone (2.62 mmol, 1 g) was
added copper (II) acetate (0.79 mmol, 143 mg), copper dust
(0.79 mmol, 50 mg), potassium acetate (5.24 mmol, 514 mg), 3-
aminobenzoic acid (3.93 mmol, 539 mg), and amyl alcohol
(5 mL). The reaction mixture was heated to 150–160 °C for 24 h.
After the reaction was complete via TLC, the reaction was cooled
to ambient temperature, ethanol was added with vigorous stirring,
the precipitate was filtered off and washed with ethanol, then the
solid was transferred to a dilute aqueous HCl solution and heated
to 80–90 °C for 10 min and filtered again. The crude solid was left
to air dry for 1 h and then dried in vacuo. The crude solid can be
recrystallized in acetic acid to remove most of the impurities.
Crude yield: 413 mg (36%), bluish solid. HRMS calcd for
To the dibromo anthraquinone derivative 14 (1.0 equiv) was
added copper(II) acetate (0.3 equiv), copper dust (0.3 equiv), potas-
sium acetate (2.0 equiv), aniline (3.0 equiv), and amyl alcohol
(0.5 M). The reaction mixture was heated to 150–160 °C for
20–48 h. After the reaction was complete via TLC, the reaction
was cooled to ambient temperature, ethanol was added with vigor-
ous stirring, the precipitate was filtered off and washed with etha-
nol, then the solid was transferred to a dilute aqueous HCl solution
and heated to 80–90 °C for 10 min and filtered again. The crude so-
lid was left to air dry for 1 h and then dried in vacuo. The crude so-
lid can be recrystallized in acetic acid to remove most of the
impurities, but in certain cases prep HPLC is required for further
purity.35 Crude yield 1.01 g (88%) as a purple solid 3-((3-bromo-
4-hydroxy-9,10-dioxo-9,10-dihydroanthracen-1-yl)amino)benzoic
acid (15a). HRMS calcd for C21H11BrNO5 (MꢂH)ꢂ 435.9826, found
435.9851. 1H NMR (500 MHz, CDCl3) d 7.38–7.43 (m, 1H), 7.49 (t,
C
21H12BrNO5 (MꢂH)ꢂ 434.9986, found 435.0011. 1H NMR
(500 MHz, CDCl3) d 7.33 (d, J = 7 Hz, 1H), 7.47 (d, J = 8 Hz, 1H),
7.76–7.93 (m, 7H), 8.31–8.38 (m, 2H). LC/MS: rt = 3.28 min,
purity = 87%.

4772
F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
3.14.12. 2-((3-Bromo-4-methoxy-9,10-dioxo-9,10-dihydro
anthracen-1-yl)amino)benzoic acid (11)
To a mixture of 2-((3-bromo-4-hydroxy-9,10-dioxo-9,10-dihy-
droanthracen-1-yl)amino)benzoic acid (0.11 mmol, 50 mg) and sil-
ver(I) oxide (0.22 mmol, 51 mg) in chloroform (1 mL) was added
dried in vacuo. A portion of the crude reddish solid (43 mg, 74%)
was purified via prep HPLC. HRMS calcd for C28H17BrNO5 (MꢂH)
ꢂ
526.0296, found 526.0321. 1HNMR (500 MHz, MeOH) d 5.05 (s,
2H), 7.34–7.36 (m, 1H), 7.39–7.41 (m, 2H), 7.49 (t, J = 8 Hz, 1H),
7.68 (app d, J = 7 Hz, 2H), 7.45 (s, 1H), 7.82–7.90 (m, 3H), 7.90–
7.91 (m, 1H), 8.22 (app d, J = 7 Hz, 1H), 8.30 (app d, J = 7 Hz, 1H),
8.55 (br s, 2H). LC/MS: rt = 5.89 min.
iodomethane (0.33 mmol, 21 lL). The mixture was stirred for 20 h
at ambient temperature in the dark. Additional silver(I) oxide
(0.11 mmol) and iodomethane (0.11 mmol) was then added and
the stirring continued for another 20 h. After the reaction was com-
plete as judged by TLC, the mixture was filtered through celite and
the solvent removed in vacuo to give the crude product (31 mg,
62%) as a reddish solid.37 HRMS calcd for C22H13BrNO5 (MꢂH)ꢂ
449.9983, found 450.0007. 1H NMR (500 MHz, CDCl3) d 3.93 (s,
3H), 7.57–7.61 (m, 2H), 7.83–7.92 (m, 4H), 7.99 (br s, 1H), 8.04–
8.08 (m, 1H), 8.20–8.34 (m, 3H). LC/MS: rt = 11.81 min, purity = 91%.
Acknowledgments
The research was supported by the National Institutes of Health
(CA135380) (SOM), the INGEN grant from the Lilly Endowment,
Inc. (SOM), by The Indiana University Melvin and Bren Simon
Cancer Center Translational Research Acceleration Collaboration
(ITRAC) (SOM), the Showalter Trust (SOM), by the Indiana Univer-
sity Biomedical Research Fund (SOM) and by the American Cancer
Society Institutional Grant (SOM). Computer time on the Big Red
supercomputer at Indiana University is funded by the National
Science Foundation and by Shared University Research.
3.14.13. 3-((3-Bromo-4-isopropoxy-9,10-dioxo-9,10-dihydro
anthracen-1-yl)amino)benzoic acid (9)
To a mixture of 3-((3-bromo-4-hydroxy-9,10-dioxo-9,10-dihy-
droanthracen-1-yl)amino)benzoic acid (0.23 mmol, 100 mg) and
silver(I) oxide (0.46 mmol, 107 mg) in chloroform (2 mL) was
added 2-iodopropane (0.69 mmol, 69 lL). The mixture was stirred
Supplementary data
for 20 h at ambient temperature in the dark. Additional silver(I)
oxide (0.23 mmol) and 2-iodopropane (0.23 mmol) was then
added and stirred for another 20 h. After the reaction was complete
as judged by TLC, the mixture was filtered through Celite and the
solvent removed in vacuo to give the dialkylated crude product
(50 mg, 45%). The dialkylated crude material was hydrolyzed in
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2012.06.002.
References and notes
2 M aqueous NaOH (140 lL) and methanol (1 mL). Methanol was
1. Werle, B.; Kotzsch, M.; Lah, T. T.; Kos, J.; Gabrijelcic-Geiger, D.; Spiess, E.;
Schirren, J.; Ebert, W.; Fiehn, W.; Luther, T.; Magdolen, V.; Schmitt, M.; Harbeck,
N. Anticancer Res. 2004, 24, 4147.
2. Andreasen, P. A.; Kjoller, L.; Christensen, L.; Duffy, M. J. Int. J. Cancer 1997, 72, 1.
3. Wei, Y.; Tang, C. H.; Kim, Y.; Robillard, L.; Zhang, F.; Kugler, M. C.; Chapman, H.
A. J. Biol. Chem. 2007, 282, 3929.
4. Wei, Y.; Czekay, R. P.; Robillard, L.; Kugler, M. C.; Zhang, F.; Kim, K. K.; Xiong, J.
P.; Humphries, M. J.; Chapman, H. A. J. Cell Biol. 2005, 168, 501.
5. Simon, D. I.; Wei, Y.; Zhang, L.; Rao, N. K.; Xu, H.; Chen, Z. P.; Liu, Q. M.;
Rosenberg, S.; Chapman, H. A. J. Biol. Chem. 2000, 275, 10228.
6. Wei, Y.; Lukashev, M.; Simon, D. I.; Bodary, S. C.; Rosenberg, S.; Doyle, M. V.;
Chapman, H. A. Science 1996, 273, 1551.
7. Wei, Y.; Eble, J. A.; Wang, Z.; Kreidberg, J. A.; Chapman, H. A. Mol. Biol. Cell 2001,
12, 2975.
8. Kiyan, J.; Kiyan, R.; Haller, H.; Dumler, I. EMBO J. 2005, 24, 1787.
9. Liu, D.; Aguirre Ghiso, J.; Estrada, Y.; Ossowski, L. Cancer Cell 2002, 1, 445.
10. Resnati, M.; Pallavicini, I.; Wang, J. M.; Oppenheim, J.; Serhan, C. N.; Romano,
M.; Blasi, F. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 1359.
removed in vacuo and the solid residue was acidified to pH 2 to
give the monoalkylated crude product (35 mg, 76%) as a reddish
solid. HRMS calcd for C24H17BrNO5 (MꢂH)ꢂ 478.0296, found
478.0320. 1HNMR (500 MHz, MeOH) d 1.29 (s, 1H), 1.36 (d,
J = 6 Hz, 6H), 4.41–4.46 (m, 1H), 7.40–7.41 (m, 1H), 7.49 (t,
J = 8 Hz, 1H), 7.73 (s, 1H), 7.81–7.89 (m, 3H), 7.91 (s, 1H), 8.19
(app d, J = 8 Hz, 1H), 8.29 (app d, J = 7 Hz, 1H). LC/MS: rt = 4.18 min.
3.14.14. 1-(Benzyloxy)-2,4-dibromoanthracene-9,10-dione (5)
To a solution of 2,4-dibromo-1-hydroxyanthracene-9,10-dione
(0.51 mmol, 200 mg) in 3:1 methyl ethyl ketone/DMF (5 mL),
potassium carbonate (0.77 mmol, 106 mg) was added followed
by benzyl bromide (1.0 mmol, 119 lL). The solution was refluxed
for 20 h followed by quenching with slow addition of 1 M HCl
(2.5 mL). Precipitate formed after cooling to 0 °C. The solid was fil-
tered, washed with water, satd aqueous sodium carbonate, water,
and brine to give 211 mg (88%) of a reddish solid. 1H NMR
(500 MHz, CDCl3) d 5.15 (s, 2H), 7.37 (t, J = 7.25 Hz, 1H), 7.43 (t,
J = 7.25 Hz, 2H), 7.70 (d, J = 6.94 Hz, 2H), 7.77 (m, 2H), 8.18 (ddd,
J = 8.99, 3.78, 3.63 Hz, 2H), 8.28 (s, 1H). 13C NMR (100 MHz, CDCl3)
d 76.1, 117.6, 126.8, 126.9, 127.3, 128.5, 128.9, 130.0, 131.9, 133.4,
133.7, 134.0, 134.1, 136.0, 144.5, 155.5, 181.3, 181.9. HRMS calcd
for C21H13Br2O3 (M+H)+ 470.9226, found 470.9232.
11. Tang, C. H.; Hill, M. L.; Brumwell, A. N.; Chapman, H. A.; Wei, Y. J. Cell Sci. 2008,
121, 3747.
12. Duriseti, S.; Goetz, D. H.; Hostetter, D. R.; LeBeau, A. M.; Wei, Y.; Craik, C. S. J.
Biol. Chem. 2010, 285, 26878.
13. Henneke, I.; Greschus, S.; Savai, R.; Korfei, M.; Markart, P.; Mahavadi, P.;
Schermuly, R. T.; Wygrecka, M.; Sturzebecher, J.; Seeger, W.; Gunther, A.;
Ruppert, C. Am. J. Respir. Crit. Care Med. 2010, 181, 611.
14. Carriero, M. V.; Longanesi-Cattani, I.; Bifulco, K.; Maglio, O.; Lista, L.; Barbieri,
A.; Votta, G.; Masucci, M. T.; Arra, C.; Franco, R.; De Rosa, M.; Stoppelli, M. P.;
Pavone, V. Mol. Cancer Ther. 2009, 8, 2708.
15. Rao, J. S.; Gondi, C.; Chetty, C.; Chittivelu, S.; Joseph, P. A.; Lakka, S. S. Mol.
Cancer Ther. 2005, 4, 1399.
16. Almasi, C. E.; Høyer-Hansen, G.; Christensen, I. J.; Pappot, H. APMIS. 2009, 117,
755.
17. Shapiro, R. L.; Duquette, J. G.; Nunes, I.; Roses, D. F.; Harris, M. N.; Wilson, E. L.;
Rifkin, D. B. Am. J. Pathol. 1997, 150, 359.
18. Gandhari, M.; Arens, N.; Majety, M.; Dorn-Beineke, A.; Hildenbrand, R. Int. J.
Oncol. 2006, 28, 1463.
19. Liang, X.; Yang, X.; Tang, Y.; Zhou, H.; Liu, X.; Xiao, L.; Gao, J.; Mao, Z. Oral Oncol.
2008, 12, 23.
20. Kondraganti, S.; Gondi, C. S.; McCutcheon, I.; Dinh, D. H.; Gujrati, M.; Rao, J. S.;
Olivero, W. C. Int. J. Oncol. 2006, 28, 1353.
21. Prager, G. W.; Breuss, J. M.; Steurer, S.; Olcaydu, D.; Mihaly, J.; Brunner, P. M.;
Stockinger, H.; Binder, B. R. Circ. Res. 2004, 94, 1562.
22. Schiller, H. B.; Szekeres, A.; Binder, B. R.; Stockinger, H.; Leksa, V. Mol. Biol. Cell
2009, 20, 745.
23. Mignatti, P.; Rifkin, D. B. Enzyme Protein 1996, 49, 117.
24. Rabbani, S. A.; Mazar, A. P. Surg. Oncol. Clin. N. Am. 2001, 10, 393.
25. Subramanian, R.; Gondi, C. S.; Lakka, S. S.; Jutla, A.; Rao, J. S. Int. J. Oncol. 2006,
28, 831.
3.14.15. 3-((4-(Benzyloxy)-3-bromo-9,10-dioxo-9,10-dihydro
anthracen-1-yl)amino)benzoic acid (6)
To 1-(benzyloxy)-2,4-dibromoanthracene-9,10-dione (0.11 mmol,
50 mg) was added copper(II) acetate (0.033 mmol, 6 mg), copper
dust (0.033 mmol, 2 mg), potassium acetate (0.22 mmol, 22 mg),
3-aminobenzoic acid (0.33 mmol, 45 mg), and amyl alcohol
(1 mL). The reaction mixture was refluxed for 20 h at 150–160 °C.
After 20 h the reaction was cooled to ambient temperature, ethanol
was added with vigorous stirring, the precipitate was filtered off
and washed with ethanol, then the solid was transferred to a dilute
aqueous HCl solution and heated to 80–90 °C for 10 min and fil-
tered again. The crude solid was left to air dry for 1 h and then

F. Wang et al. / Bioorg. Med. Chem. 20 (2012) 4760–4773
4773
26. Kunigal, S.; Lakka, S. S.; Gondi, C. S.; Estes, N.; Rao, J. S. Int. J. Cancer 2007, 121,
2307.
27. Mazar, A. P. Clin. Cancer Res. 2008, 14, 5649.
48. Ma, Z.; Webb, D. J.; Jo, M.; Gonias, S. L. J. Cell Sci. 2001, 114, 3387.
49. Wei, Y.; Yang, X.; Liu, Q.; Wilkins, J. A.; Chapman, H. A. J. Cell Biol. 1999, 144,
1285.
28. Ploug, M.; Ellis, V. FEBS Lett. 1994, 349, 163.
29. Gardsvoll, H.; Gilquin, B.; Le Du, M. H.; Menez, A.; Jorgensen, T. J.; Ploug, M. J.
Biol. Chem. 2006, 281, 19260.
50. Wang, F.; Li, J.; Sinn, A. L.; Knabe, W. E.; Khanna, M.; Jo, I.; Silver, J. M.; Oh, K.; Li,
L.; Sandusky, G. E.; Sledge, G. W.; Nakshatri, H.; Jones, D. R.; Pollok, K. E.;
Meroueh, S. O. J. Med. Chem. 2011, 54, 7193.
30. Reuning, U.; Sperl, S.; Kopitz, C.; Kessler, H.; Kruger, A.; Schmitt, M.; Magdolen,
V. Curr. Pharm. Des. 2003, 9, 1529.
51. Twentyman, P. R.; Luscombe, M. Br. J. Cancer 1987, 56, 279.
52. Albini, A.; Iwamoto, Y.; Kleinman, H. K.; Martin, G. R.; Aaronson, S. A.;
Kozlowski, J. M.; McEwan, R. N. Cancer Res. 1987, 47, 3239.
53. Parkkila, S.; Rajaniemi, H.; Parkkila, A. K.; Kivela, J.; Waheed, A.; Pastorekova,
S.; Pastorek, J.; Sly, W. S. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 2220.
54. Toth, M.; Fridman, R. Methods Mol. Med. 2001, 57, 163.
55. DeLano, W. L. DeLano Scientific; Palo Alto: CA, USA, 2002.
56. Case, D. A.; Darden, T. A., Cheatham, T.E., III; Simmerling, C. L.; Wang, J.; Duke,
R. E.; Luo, R.; Merz, K. M.; Wang, B.; Pearlman, D. A.; Crowley, M.; Brozell, S.;
Tsui, V.; Gohlke, H.; Mongan, J.; Hornak, V.; Cui, G.; Beroza, P.; Schafmeister, C.;
Caldwell, J. W.; Ross, W. S.; Kollman, P. A. University of California: San
Francisco, 2004.
31. Khanna, M.; Wang, F.; Jo, I.; Knabe, W. E.; Wilson, S. M.; Li, L.; Bum-Erdene, K.;
Li, J.; Sledge, G. W.; Khanna, R.; Meroueh, S. O. ACS Chem. Biol. 2011, 6, 1232.
32. Irwin, J. J.; Shoichet, B. K. J. Chem. Inf. Model. 2005, 45, 177.
33. Elkazaz, S.; Jones, P. B. J. Org. Chem. 2010, 75, 412.
34. Jones, P. B.; Brinson, R. G.; Sarma, S.; Elkazaz, S. Org. Biomol. Chem. 2008, 6,
4204.
35. Ullmann, F.; Eiser, O. Ber. Dtsch. Chem. Ges. 1916, 49, 2154.
36. Sarma, S. J.; Jones, P. B. J. Org. Chem. 2010, 75, 3806.
37. Tietze, L. F.; Gericke, K. M.; Singidi, R. R.; Schuberth, I. Org. Biomol. Chem. 2007,
5, 1191.
38. Khanna, M.; Chelladurai, B.; Gavini, A.; Li, L.; Shao, M.; Courtney, D.; Turchi, J. J.;
Matei, D.; Meroueh, S. Mol. Cancer Ther. 2011, 10, 626.
39. De Souza, M.; Matthews, H.; Lee, J. A.; Ranson, M.; Kelso, M. J. Bioorg. Med.
Chem. 2011, 19, 2549.
40. Brown, S.; Meroueh, S. O.; Fridman, R.; Mobashery, S. Curr. Top. Med. Chem.
2004, 4, 1227.
41. Aguirre Ghiso, J. A.; Kovalski, K.; Ossowski, L. J. Cell Biol. 1999, 147, 89.
42. Aguirre-Ghiso, J. A.; Estrada, Y.; Liu, D.; Ossowski, L. Cancer Res. 2003, 63, 1684.
43. Aguirre-Ghiso, J. A.; Liu, D.; Mignatti, A.; Kovalski, K.; Ossowski, L. Mol. Biol. Cell
2001, 12, 863.
57. Wang, R.; Lu, Y.; Fang, X.; Wang, S. J. Chem. Inf. Comput. Sci. 2004, 44, 2114.
58. Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. J.
Chem. Phys. 1983, 79, 10.
59. Ryckaert, J. P.; Ciccotti, G.; Berendsen, J. J. C. J. Comput. Phys. 1977, 23, 15.
60. Darden, T.; York, D.; Pedersen, L. J. Chem. Phys. 1993, 98, 10089.
61. Srinivasan, J.; Cheatham, T. E.; Cieplak, P.; Kollman, P. A.; Case, D. A. J. Am.
Chem. Soc. 1998, 120, 9401.
62. Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S. H.; Chong, L.; Lee, M.;
Lee, T.; Duan, Y.; Wang, W.; Donini, O.; Cieplak, P.; Srinivasan, J.; Case, D. A.;
Cheatham, T. E. Acc. Chem. Res. 2000, 33, 889.
44. Ahmed, N.; Oliva, K.; Wang, Y.; Quinn, M.; Rice, G. Br. J. Cancer 2003, 89, 374.
45. Chaurasia, P.; Aguirre-Ghiso, J. A.; Liang, O. D.; Gardsvoll, H.; Ploug, M.;
Ossowski, L. J. Biol. Chem. 2006, 281, 14852.
46. Chaurasia, P.; Mezei, M.; Zhou, M. M.; Ossowski, L. PloS One 2009, 4, e4617.
47. Geetha, N.; Mihaly, J.; Stockenhuber, A.; Blasi, F.; Uhrin, P.; Binder, B. R.;
Freissmuth, M.; Breuss, J. M. J. Biol. Chem. 2011, 286, 25663.
63. Wang, J. M.; Morin, P.; Wang, W.; Kollman, P. A. J. Am. Chem. Soc. 2001, 123,
5221.
64. Ukponmwan, D. O.; Greenhaigh, M.; Peters, A. T. J. Chem. Eng. Data 1984, 29,
482.