Discovery of new small molecules targeting the vitronectin ...€¦ · 22/05/2013 · Discovery of new small molecules targeting the vitronectin binding site of the urokinase receptor

Discovery of new small molecules targeting the vitronectin binding site of the urokinase

receptor that block cancer cell invasion

Vincenza Elena Anna Rea1^, Antonio Lavecchia2^, Carmen Di Giovanni2, Francesca Wanda

Rossi1, Anna Gorrasi3, Ada Pesapane1, Amato de Paulis1, Pia Ragno3 and Nunzia Montuori1

1From the Department of Translational Medical Sciences, 2Department of Pharmacy, “Drug

Discovery” Laboratory, “Federico II” University, Naples, Italy; 3Department of Chemistry

and Biology, University of Salerno, Salerno, Italy;

^V.E.A. Rea and A. Lavecchia equally contributed to this work

*Running title: Small molecules inhibiting uPAR binding to VN

To whom correspondence should be addressed:

Nunzia Montuori, MD PhD,

Department of Translational Medical Sciences, University “Federico II”

via S. Pansini, 5

80131 Naples, Italy.

Phone: +39 0817463309

FAX: +39 0817463308

e-mail: [email protected]

Note: The authors declare no conflicts of interests.

Research. on September 11, 2020. © 2013 American Association for Cancermct.aacrjournals.org Downloaded from

Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on May 22, 2013; DOI: 10.1158/1535-7163.MCT-12-1249

http://mct.aacrjournals.org/

ABSTRACT

Besides focusing urokinase (uPA) proteolytic activity on the cell membrane, the urokinase

receptor (uPAR) is able to bind vitronectin (VN), via a direct binding site. Furthermore,

uPAR interacts with other cell surface receptors, such as integrins, receptor tyrosin kinases

and chemotaxis receptors, triggering cell-signalling pathways that promote tumor

progression. The ability of uPAR to coordinate binding and degradation of extracellular

matrix and cell signalling makes it an attractive therapeutic target in cancer.

We used structure-based virtual screening (SB-VS) to search for small molecules targeting

the uPAR binding site for VN. 41 compounds were identified and tested on uPAR-negative

HEK-293 epithelial cells transfected with uPAR (uPAR-293 cells), using the parental cell

line transfected with the empty vector (V-293 cells), as a control.

Compounds, 6 and 37 selectively inhibited uPAR-293 cell adhesion to VN and the resulting

changes in cell morphology and signal transduction, without exerting any effect on V-293

cells. 6 and 37 inhibited uPAR-293 cell binding to VN with IC50 values of 3.6 and 1.2 µM,

respectively.

Compounds 6 and 37 targeted S88 and R91, key residues for uPAR binding to VN but also

for uPAR interaction with the f-MLF family of chemotaxis receptors (fMLF-Rs). As a

consequence, 6 and 37 impaired uPAR-293 cell migration toward FCS, uPA and f-MLF,

likely by inhibiting the interaction between uPAR and FPR1, the high affinity fMLF-R.

Both compounds blocked in vitro extracellular-matrix invasion of several cancer cell types

and could represent new promising leads for pharmaceuticals in cancer.




3

INTRODUCTION

The urokinase (uPA)-mediated plasminogen activation system, which generates the potent

serine-protease plasmin, is involved in various pathological processes, including

angiogenesis, inflammation, wound healing and metastasis (1,2).

The key molecule of this system, the uPA receptor (uPAR), is anchored to the plasma

membrane by a glycosylphosphatidylinositol moiety and is formed by three homologous

domains (DI, DII and DIII, from the N-terminus). The uPA-binding site is located in the DI

domain, but the full-length molecule is required for an efficient binding (3,4). uPAR

enhances pericellular proteolysis by serving as a docking site to uPA, thus triggering a

cascade of proteolytic events that leads to the active degradation of extracellular matrix

(ECM) components (5).

Despite the lack of a transmembrane domain, uPAR can activate intracellular signalling

through lateral interactions with other cell surface receptors, such as integrins, receptor

tyrosin kinases and G-protein coupled chemotaxis receptors (6).

uPAR ability to regulate integrin activity plays a key role in cell adhesion, migration,

proliferation and survival (7,8). Recently, integrin binding sites have been identified in

uPAR domain DII (residues 130-142) (9) and in uPAR domain DIII (residues 240-248) (10).

Moreover, uPAR crosstalk with EGF receptor (EGFR) is extensive and may regulate the

shift from tumor cell dormancy to proliferation (11,12).

uPAR interaction with receptors for fMet-Leu-Phe (fMLF-Rs) (13) is required for both uPA- and

fMLF-dependent cell migration and occurs through a chemotactic domain located in the DI-DII

linker region, the SRSRY sequence (a.acids 88-92) (14). Since a soluble cleaved form of uPAR,

exposing at the N-terminus the SRSRY sequence is a ligand for fMLF-Rs (15), it has been

proposed that uPA binding to uPAR determines a conformational modification of the receptor




4

with the exposure of the chemotactic SRSRY domain that, in turn, binds and activates members

of the fMLF-R family, thus inducing chemotaxis (16).

Furthermore, uPAR itself is an adhesion receptor; indeed, it binds vitronectin (VN), an abundant

component of provisional ECM (17,18). uPAR interactions with integrins and VN are positively

regulated by uPA (19,20), and both uPA and VN can induce uPAR-mediated cytoskeletal

reorganization and cell migration (21,22).

Towards the goal of discovering uPAR inhibitory compounds, we focused on uPAR binding

to VN. Indeed, it has been recently reported that uPAR promotes metastasis of human

malignancies by engaging VN, through the activation of a cell signaling to Rac1 (23). In

epithelial cell lines, uPAR induces phenotypic changes consistent with hypoxia-induced

EMT, through a direct binding to VN (24). Abnormal uPAR levels, like occur in cancer, may

encourage EMT through VN binding, thus facilitating tumor invasion and metastasis (25).

The X-ray structure of the ternary complex between uPAR, the aminoterminal fragment of

uPA (ATF) and the somatomedin B domain of VN (SMB) has been recently determined

(26). There is now evidence that uPAR:VN interaction is entirely mediated by a composite

epitope exposed on the DI/DII interface of uPAR (residues R30, W32, S56, R58, I63, S65,

S88, R91, R116 and Q114) (27).

Therefore, we used structure-based virtual screening (SB-VS) (28) of the National Cancer

Institute (NCI) Diversity Set II to identify small molecules targeting this site and disrupting

uPAR binding to VN.

Here, we describe the successful outcome of this search and the initial biological evaluation

of the two most promising compounds from this effort.




5

MATERIALS AND METHODS

Computational chemistry - Molecular modeling and graphics manipulations were performed

using Maestro software (Maestro, version 9.2, Schrödinger, LLC, New York, NY, 2011) and

Pymol packages (29) running on a E4 Computer Engineering E1080 workstation provided

of a Intel Core i7-930 Quad-Core processor.

Protein preparation - The X-ray coordinates of human uPAR–ATF–SBM ternary complex

(PDB code: 3BT1) (27) were extracted from the Protein Data Bank (30). The structure was

then prepared using the Protein Preparation Wizard of the Schrödinger graphical interface

Maestro. Hydrogen atoms were added to the protein consistent with the neutral physiologic

pH (7.0). Thus, the R and K side chains were cationic while the E and D side chain

carboxylates were anionic. The protonation and flip states of the imidazole rings of the H

residues were adjusted together with the side chain amides of N and Q residues and the OH

and SH orientations to optimize such interactions. X-ray water molecules and ATF and SMB

ligands were removed during protein preparation, the last step of which was energy

minimization of the entire structure. The minimization was terminated when the root mean

square deviation (rmsd) of the heavy atoms in the energy minimized structure relative to the

starting (X-ray) coordinates exceeded 0.3 Angstrom. This ensures that the integrity of the X-

ray structure is preserved in further modeling studies while eliminating potential

stereochemical short contacts that may exist in the protein structure. Furthermore, this

process also ensures that the hydrogen atoms are placed in optimized geometries.

Ligand preparation and filtering - 3D structures of NCI Diversity Set II (1364 compounds)

were extracted by the web site (31) and prepared using LigPrep software v2.5 (LigPrep,

version 2.5, Schrödinger, LLC, New York, NY, 2011) with Epik to expand protonation and




6

tautomeric states at 7.0 ± 2.0 pH units. In post LigPrep steps unspecified stereoisomers were

retained up to 4 low energy stereoisomers and sample 5/6 membered rings retained up to one

conformation per ligand as default parameters suggested in Maestro virtual screening

workflow. High energy ionization/tautomer states were removed from the generated

conformations. So, about 2000 structures including stereoisomers, tautomers, and ionization

states were ready to be submitted to the subsequent docking runs.

SB-VS Protocol - The GLIDE virtual screening application in Schrödinger Molecular

Modeling Suite was used to screen compounds using two levels of docking precision. A

modified version of the Chemscore function is employed by GLIDE to assign a score to each

ligand in all poses. In the first step, GLIDE was run in standard Precision (SP) mode. The

20% of the top-scoring ligands (272 compounds) were kept and redocked using the GLIDE

extra precision (XP) mode, which similarly retained the 20% of the best ranked compounds

(54 compounds). The XP docking procedure of GLIDE, which incorporates a more accurate,

finer-grained docking algorithm, was designed to eliminate false positives that survive the

SP stage. After visual inspection of the top ranked compounds, 41 hits were chosen for

biological evaluation. Flexible docking was allowed in all stages and default parameters

from the Virtual Screening Workflow were used in all docking studies, in addition to the

aforementioned modifications to the percent of compounds entering each stage. All final

scores and poses came from GLIDE XP. The grid for docking studies was chosen

sufficiently large to enclose all residues involved in the VN/uPAR interactions within a

cubic box of dimensions 46Å x 46Å x 46Å. The enclosing box was centered on the VN

binding site setting the bounding box with the sizes of 14 Å x 14 Å x14 Å. A van der Waals

radius scaling factor of 0.80 for atoms with a partial atomic charge (absolute value) less than

0.15 was used in order to soften the potential for nonpolar parts of the receptor. Compounds




7

6 and 37 identified by SB-VS were flexibly docked using the same protein grid prepared for

virtual screening protocol. Ten poses were collected for each ligand and ranked according to

predicted GLIDE XP score.

Chemical inhibitors - All compounds were obtained from the NCI/DTP Open Chemical

Repository (32). The compounds identified by Virtual Screening were dissolved in dimethyl

sulfoxide (DMSO) and stored at -20 °C, at a concentration of 0.01 M.

Cell cultures and transfections - The uPAR-negative (17) human embryonic kidney cell line

HEK-293 (ATCC Certified from LGC Standards, Milan, Italy) was grown in Dulbecco's

modified Eagle's medium (DMEM) (GIBCO, Gaithersburg, MD, USA) supplemented with

10% fetal bovine serum (FBS). The sarcoma-derived HT1080, the PC3 prostate cancer, the

HCT colon cancer, the MDAMB231 breast cancer cell lines (ATCC Certified from LGC

Standards, Milan, Italy) were grown in DMEM (GIBCO) supplemented with 10% FBS.

Authentication of cell lines was conducted by ATCC using a short tandem repeat assay.

Upon receipt, all cell lines were stored in liquid nitrogen and passaged for less than 6 months

before use in this study.

uPAR cDNA was cloned in a pcDNA3 vector with resistance to Geneticin (Invitrogen,

Carlsbad, CA, USA), and the resulting plasmid was named uPAR-pcDNA3. HEK-293 cells

were stably transfected with uPAR-pcDNA3 or with the empty vector pcDNA3, as described

(14).

Western Blot - Cells were lysed in PBS (0.08 M NaCl, 0.002 M KCl, 0.0115 M Na2HPO4,

0.002 M KH2PO4, pH 7.2) containing 1% Triton X-100, in the presence of a protease

inhibitor cocktail containing AEBSF, Aprotinin, Bestatin, E-64, Leupeptin and Pepstatin A




8

(Sigma-Aldrich, St. Louis, MO, USA) and a phosphatase inhibitor cocktail containing

microcystin LR, cantharidin, and bromotetramisole (Sigma-Aldrich). Protein concentration

of lysates was determined using a colorimetric assay (Biorad, Richmond, CA, USA). Equal

amounts of protein were subjected to SDS-polyacrylamide gel electrophoresis and

transferred to poly-vinyl difluoride (PVDF) filters (Millipore, Windsor, MA, USA).

Membranes were incubated for 20 h at 4 °C with the 399 polyclonal anti-uPAR antibody

(American Diagnostica, Greenwich, CT, USA), mouse monoclonal anti-p-Erk or rabbit

polyclonal anti-Erk2 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA), or with

non-immune immunoglobulins (Igs) (Jackson ImmunoResearch, Suffolk, England), as a

negative control. Filters were further incubated for 30 min at room temperature with HRP-

conjugated anti-rabbit or anti-mouse antibodies (Biorad). The reaction was revealed by an

enhanced Chemo-Luminescence (ECL) detection kit (Amersham Biosciences, Little

Chalfont, England).

Cell adhesion assay - The adhesion assays were carried out on 96-well flat-bottomed plates

for cell (Nunc, Roskilde, Denmark). Wells were coated with 1 µg of VN (Becton Dickinson

Biosciences, Franklin Lakes, NJ, USA), fibronectin (FN) (Roche, Indianapolis, USA) or

with 100 ml of heat-denatured 1% Bovine serum albumin (BSA) in PBS, as a negative

control, and incubated overnight at 4 °C. The plates were then blocked 1 h at room

temperature with 1% heat-denatured BSA in PBS. Cells were detached with trypsin,

resuspended in DMEM containing 10% FBS and incubated for 1 h at 37 °C, 5% CO2, to

allow receptor recovery. Cells were then washed with serum-free DMEM, counted,

distributed into the wells at a density of 105 cells/well and incubated for 1 h at 37 °C in the

presence of inhibitors, or DMSO, as a vehicle control. Attached cells were fixed with 3%

paraformaldehyde in PBS for 10 min and then incubated with 2% methanol for 10 min. Cells




9

were finally stained for 10 min with 0.5% crystal violet in 20% methanol. Stain was eluted

by 0.1 M sodium citrate in 50% ethanol, pH 4.2, and the absorbance at 540 nm was

measured by a spectrophotometer.

Cytotoxicity assay – V-293 and uPAR-293 cells were plated, at 5x104 cells/well, in 96-well

plates and grown for 24h. Cells were serum-starved overnight using DMEM containing

0.1% BSA, and then incubated for 1 and 24 hours at 37 °C, 5% CO2 in the same medium

containing the selected small molecules or DMSO, as a vehicle control. 20 µl/well of

CellTiter 96 AQueous One Solution Reagent (Promega, Madison, WI, USA) was added and,

after incubation at 37°C for 4 h, the absorbance was determined by an ELISA reader (Bio-

Rad) at a wavelength of 490 nm.

Cell Migration and Invasion assays - Cell migration assays were performed in Boyden

chambers using 8 µm pore size polyvinylpyrrolidone (PVPF)-free chemotaxis filters

(Wathman Int., Kent, UK), coated with 50 µg/ml collagen (CG) or 5 µg/ml vitronectin (VN)

V-293 and uPAR-293 cells (2x105) were plated in the upper chamber in DMEM 0.1% BSA

containing the selected small molecules or DMSO, as a vehicle control. DMEM 0.1% BSA

alone or containing the Aminoterminal fragment of uPA (ATF) (1x10-8 M), fMLF (1x10-7

M) or 10% FBS was added in the lower chamber and the cells were allowed to migrate for

4h at 37°C, 5% CO2.

For the invasion assay, filters were coated with 50 µg/ml MatrigelTM (BD Biosciences, San

Jose, CA) and incubated for 30 min at 37 °C for gelling. 2x105 cells, plated in the upper

chamber in DMEM 0.1% BSA containing the selected small molecules or DMSO, were

allowed to migrate toward DMEM medium supplemented with 10% FCS, or toward DMEM

medium supplemented with 0.1% BSA, as a control, for 18 h at 37°C, 5% CO2.




10

At the end of both experiments, cells on the lower surface of the filter were fixed in ethanol,

stained with hematoxylin, and counted at 200x magnification (10 random fields/filter). Cell

migration and invasion were expressed as a percent increase over the control.

Coimmunoprecipitation - uPAR-293 cells (5x106/sample) were plated in 100mm dishes for

24h; then, the cells were incubated for 16h in DMEM 0.1% BSA containing the selected

small molecules or DMSO, as a vehicle control, in serum-free medium. Cells were lysed in

RIPA buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.5, 1% deoxycholate acid, 0.1% SDS,

1% Triton X-100, protease and phosphatases inhibitors) and incubated with non immune

serum and 10% protein A-conjugated Sepharose (GE Healthcare, Uppsala, Sweden) for 2

hours at 4 °C. After centrifugation, the supernatants were incubated with 2 µg/ml of the R4

monoclonal anti-uPAR antibody, kindly provided by Dr. G. Hoyer-Hansen (The Finsen

Laboratory, Copenhagen, Denmark), or with non-immune Igs, for 2 hours at 4 °C and then

with 10% protein A-Sepharose for 30 minutes at room-temperature. The immunoprecipitates

were washed in RIPA buffer, subjected to 10% SDS-PAGE, and analyzed by Western blot

using a polyclonal antibody directed against the high affinity fMLF receptor, FPR1 (Santa

Cruz Biotechnology, Santa Cruz, CA), or non immune Igs, as a negative control.

Rac1 pull-down assay - uPAR-293 cells were starved for 24 h and then plated on VN-coated

wells (5 mg/ml) for the indicated time, in the presence of compounds 6 and 37, or DMSO as

a vehicle control. After a quick wash with ice-cold PBS, cells were lysed with GST-Fish

buffer (50mM Tris–HCl ph 7.4, 2mM MgCl2, 1% NP-40, 10% glycerol, 100mM NaCl,

1mg/ml leupeptin, 1mg/ml pepstatin, 1mg/ml aprotinin, 1mM PMSF, and 2mM DTT). After

10 min at 4°C under agitation, cells were scraped and lysates were cleared by centrifugation

in a precooled rotor. 500 mg of total protein extract was mixed with 10 mg of GST-PAK-




11

CRIB domain, which specifically recognizes the GTP-bound forms of Rac1, coupled to

glutathione-sepharose beads (Upstate Biotechnology, Danvers, MA) and incubated 30 min at

4°C under agitation. Beads were then rinsed three times rapidly with 1ml of ice-cold GST-

Fish buffer. The amounts of total Rac and Rac-GTP were estimated by immunoblot against

Rac1 (Upstate Biotechnology).

uPAR biotinylation - Recombinant soluble uPAR (American Diagnostica) was biotinylated

and purified with the Amersham ECL protein biotinylation module according to

manufacturer instructions (GE Healthcare).

Binding of soluble uPAR to immobilized VN – High binding plates with 96 flat-bottomed

wells (Corning, Amsterdam, ND) were coated with 0.5 µg/well VN diluted in PBS, or BSA

as a negative control, and incubated at 4 °C overnight. After a wash in PBS, residual

binding sites were blocked with 200 µl of 1% BSA in PBS, for 1 h at room temperature. 25

nM biotinylated s-uPAR (diluted in PBS, 1 mg/ml BSA), alone or in the presence of

compounds 6 and 37, was placed into coated wells. The plates were kept at 4 °C for 1 h,

washed with PBS containing 0.1% Tween 20 and then peroxidase-labeled avidin

(Amersham), diluted 1:1500 in PBS, 10 mg/ml BSA, was added. After additional washings,

the peroxidase substrate o-phenylenediamine (Sigma) was added and allowed to react for 3

min. The reactions were terminated with l M H2SO4 and the product was quantified by

measuring absorbance at 492 nm using an automated plate reader (Rio-Rad), as described

(17). Results were expressed as a percent decrease over the control, i.e. the binding in the

absence of compounds.




12

RESULTS

Identification by SB-VS of small molecules directed to the uPAR binding site for VN. uPAR

overexpression functions as a biomarker for cancer progression and metastasis in many

forms of human malignancy (33). It has been recently proposed that uPAR can promote

metastasis not only by a uPA-dependent mechanism but also through a direct binding to VN

followed by activation of a specific signal transduction (23-25).

In order to identify new potential small molecules capable of interfering in the VN/uPAR

interaction, we performed a SB-VS experiment using the X-ray crystal structure of the

human uPAR-ATF-SBM ternary complex (27) determined at 2.8 Å resolution (3BT1 in the

Protein Data Bank).

Before starting the search, the binding site for VN was identified on the uPAR crystal

structure; it is exposed on the DI/DII interface of the receptor and comprises residues R30,

W32, S56, R58, I63, S65, S88, R91, R116, and Q114 (27). Among these, W32, R58, I63,

R91 and Y92 were identified as key residues for uPAR/VN binding by alanine-scanning

mutagenesis (18,25,26,27). In details, residues F13, Y28 and D22 of VN form an open

pocket to bind R91 of uPAR. Y27 and Y28 of VN insert into a large cavity of uPAR,

showing shape complementarity of this interface. The phenyl ring of Y28 contacts uPAR

residues I63 and W32, mainly through hydrophobic interactions. Interestingly, the VN

binding site contains some residues included in the uPAR chemotactic sequence (SRSRY,

aminoacids 88-92), required for cell migration toward fMLF and uPA (14-16).

Docking simulations were conducted through the GLIDE (Grid-based Ligand Docking from

Energetics) software (available from Schrödinger LLC, New York) (34,35), selecting the

NCI Diversity Set II as virtual library of compounds. Then, GLIDE SP (Standard Precision

mode) docked each chemical structure into the uPAR:VN binding site retaining the 20% of

the top-scoring ligands. The resulting 272 compounds were then redocked and scored with




13

Glide XP (Extra Precision) in order to estimate binding affinity and rank the ligands. The

resulting 54 top-ranked compounds were analyzed by visual inspection because it has

broadly demonstrated that docking scoring functions are often more successful at predicting

a binding pose than the actual binding affinity (36).

Therefore, compounds were prioritized by taking into account their binding mode as well as

the overall match among binding modes of all the stereoisomers, tautomers, and ionization

states of each compound and, second, their docking score. As for the visual inspection,

compounds were checked for a good protein-ligand complementarity. Moreover, ligands

able to make interactions with residues known to be important by mutagenesis studies or

because they interact with known substrates such as VN or f-MLF were prioritized. To

promote the selection of structurally diverse compounds, potential hits were grouped into

chemical classes by visual inspection, and one molecule was selected for each class.

Finally, compounds were checked for readily sample availability from the compound

provider, and 41 structurally diverse compounds (Fig.1) were requested and tested in a cell-

based assay.

Four compounds selected by SB-VS inhibited uPAR-transfected epithelial cell binding to VN.

Human embryonic kidney (HEK-293) cells, unlike most cancer cells, do not express uPAR

or uPA endogenously (14,17,18,23). Therefore, as a model to mimic the effects of uPAR

overexpression in cancer cells and to investigate the ability of SB-VS selected compounds to

inhibit uPAR-mediated cell binding to VN, HEK-293 cells were stably transfected with a

human uPAR cDNA (uPAR-293 cells) or with an empty vector (V-293 cells).

uPAR expression was then evaluated by Western blot analysis, using an anti-uPAR

polyclonal antibody, in uPAR-293 and V-293 cell lysates. uPAR-293 cells expressed high

levels of full-length uPAR whereas V-293 cells were uPAR-negative (Fig.2A).




14

In uPAR-293 cells, uPAR and integrins form stable complexes that both inhibit the native

integrin-mediated cell adhesion to FN and promote adhesion to VN, via the direct VN

binding site on uPAR (17,14). Therefore, we evaluated uPAR-293 cell adhesion to FN and

VN, using the uPAR-negative V-293 cell line as a control. As expected, uPAR-293 cells

showed a strongly increased cell adhesion to VN, as compared to V-293 cells; on the

contrary, the native cell adhesion to FN was inhibited by uPAR overexpression in uPAR-293

cells, as compared to V-293 cells (Fig.2B).

To identify potential molecules able to inhibit uPAR binding to VN, uPAR-293 cells were

subjected to adhesion experiments on VN-coated wells in the presence of 41 SB-VS selected

compounds (50 µM) or DMSO, as a negative control. While compounds 9-12 and 14,

increased cell adhesion to VN, only 4 compounds significantly inhibited uPAR-293 cell

adhesion to VN; namely, compounds 6 [4,4'-dimethyl-[1,1'-biphenyl]-2,2',5,5'-tetraol], 13

[1-(phenanthren-9-yl)guanidine], 24 [8-((4-(tert-butyl) phenoxy)methyl)-1,3-dimethyl-1H-

purine-2,6(3H,9H)-dione] and 37 [piperidin-2-yl(2-(trifluoromethyl)-6-(4-(trifluoromethyl)

phenyl)pyridin-4-yl)methanol] (Fig.2C).

Two compounds specifically inhibited uPAR-mediated cell binding to VN independently of

cell cytotoxicity. To ensure that inhibition of uPAR-293 cell adhesion to VN was not due to

cell killing, the toxicity of the above selected compounds was measured by a cell viability

assay. Serum-starved uPAR-293 cells were incubated for 1h and 24h with compounds 6, 13,

24 and 37, at the same concentration used in the adhesion assay (50 µM), or with DMSO, as

a control. Cell viability was then measured by a MTS-PES assay. Compounds 13 and 24

showed a significant level of cell toxicity (Fig.3A), also confirmed on uPAR-negative V-293

cells (not shown), and were withdrawn from the study for their unspecific toxic effects.




15

To demonstrate the specificity of uPAR as a target, the ability of compounds 6 and 37 to

inhibit in vitro binding of recombinant soluble uPAR (suPAR) to immobilized VN was

evaluated. Compounds 6 and 37 inhibited suPAR binding to VN, albeit with a lower

efficiency (Fig.3B). Indeed, suPAR differs from membrane GPI-linked uPAR in the

conformation and flexibility of the molecule, being the linker region between domains D1

and D2 less accessible (37). The possibility of conformational effects, imposed by the C-

terminal hydrophobic group, is a widespread phenomenon among GPI anchored proteins

(37); therefore, we decided to study compounds efficiency in uPAR-293 cells, using the

uPAR-negative V-293 cell line as a control. Compounds 6 and 37 inhibited uPAR-293 cell

adhesion to VN without affecting V-293 cell adhesion to the same substrate (Fig.3C).

Indeed, in the absence of uPAR, HEK-293 cell binding to VN is mediated by specific

integrins (38).

To further demonstrate the specificity of compounds 6 and 37 for uPAR-VN interaction,

their ability to inhibit uPAR-293 cell adhesion was evaluated both on VN and FN.

Compounds 6 and 37 specifically inhibited uPAR-293 cell adhesion to VN without affecting

adhesion to FN of the same cells (Fig.3D).

Thus, compounds 6 and 37 are specific inhibitors of uPAR binding to VN.

Compounds 6 and 37 inhibits uPAR-dependent cell binding to VN in a dose-dependent

manner with a sub-micromolar affinity. uPAR-923 cell adhesion to VN was also evaluated

in the presence of decreasing concentration of compounds 6 and 37, in order to evaluate the

dose-dependency of their inhibitory activity. IC50 values of 6 and 37 were 3.6 µM and 1.2

µM, respectively, as calculated by non-linear regression curves using the sigmoidal dose-

response analysis of the GraphPad Prism software (Fig.4A,B). Ki of 6 and 37 were 0.58 µM




16

and 0.19 µM, respectively, as calculated by the Cheng and Prusoff Equation from EC50,

using GraphPad Prism (39).

Thus, two novel small molecules specifically inhibiting uPAR-mediated cell binding to VN

were identified by SB-VS (Table 1).

Compounds 6 and 37 abolished the changes in cell morphology and the signal transduction

induced by uPAR-mediated cell adhesion to VN. uPAR triggers cell attachment to the matrix

through a direct interaction with the SMB domain of VN (27). This initial adhesion is

followed by engagement of integrins, changes in cell morphology, migration and signal

transduction (18,23,25).

When uPAR-293 cells were seeded on VN, they underwent marked changes in morphology,

including cell flattening and extensive lamellipodia formation. In the presence of 6 and 37, at

2 x IC50, uPAR-293 cells retained a rounded cell body, failed to form lamellipodia and only a

round adhesion patch could be observed under the cell body (Fig.5A).

The limited cell contact with matrix in the presence of compounds 6 and 37 determined a

strongly reduced cell adhesion to VN. Indeed, uPAR-293 cell adhesion to increasing

concentration of VN, in the absence of inhibitors, showed a KD of 26 nM, similar to that

reported for in vitro uPAR:VN binding (30 nM) (17); the addition of compounds 6 and 37, at

their 2 x IC50, resulted in a significant loss of uPAR-293 cell binding to VN (Fig.5B).

After a direct binding to VN, uPAR can activate ERKs, most probably through its lateral

interaction with integrins (18); therefore, we investigated whether compounds 6 and 37 were

able to block ERK activation in response to VN. Serum starved uPAR-293 cells were plated

on VN-coated wells for 20 minutes in the presence of the two molecules, at 2 x IC50, or

DMSO, as a negative control; then, cell lysates were analyzed by Western blot with an anti-

phospho-ERK antibody. Both compounds strongly reduced VN-mediated ERK activation, as




17

compared to DMSO-treated cells, in which a strong VN-induced ERK activation was

observed (Fig.5C).

It has been reported that uPAR, through its binding to VN, activates the small GTPase Rac-

1, thus stimulating cell migration (23); indeed uPAR-293 cells plated on VN show extensive

lamellipoida formation (18,23,25 and Fig.5A, arrows). Therefore, we investigated, by pull-

down assays, whether compounds 6 and 37 were able to block Rac-1 activation in response

to VN. Serum starved uPAR-293 cells were plated on VN-coated wells for 20 minutes in the

presence of both ligands, at 2xIC50, or DMSO, as a negative control. Active Rac-1 (Rac-

1/GTP) was precipitated from cell lysates using the p21 binding domain (PBD) of its target,

PAK1, bound to agarose beads; Rac1/GTP was eluted from beads and estimated by Western

blot with a monoclonal anti-Rac1 antibody. Both compounds strongly reduced VN-

mediated Rac-1 activation, as compared to DMSO-treated cells, in which Rac-1 was strongly

activated (Fig.5D).

Therefore, compounds 6 and 37 specifically inhibited uPAR-mediated cell binding to VN

and its downstream signals.

Structural basis for uPAR inhibition by Compounds 6 and 37. To understand the structural

basis of the binding of the newly discovered compounds 6 and 37 to uPAR, we scrutinized

their binding poses by means of GLIDE program with the procedure described in the

Experimental Section. Fig.6 depicts the predicted binding modes of the two compounds in

the VN binding site of uPAR.

All the top-ranked poses found by GLIDE for compound 6 show a twisted conformation of

the ligand, which is stabilized by an intramolecular H-bond between the two phenolic OH

groups. The molecule extends deep into the VN binding site of uPAR, making several H-

bonds and hydrophobic interactions with key residues of the site. In particular, the OH group




18

at position 2 is predicted to form two H-bonds with both S88 OH group and the guanidinyl

Ne of R116. The OH group at position 5 also engages a H-bond with R91 side chain.

Finally, the ligand 5' OH group establishes a bidentate H-bond with the guanidinium group

of R30. In addition, the uPAR/6 complex is further stabilized by hydrophobic contacts which

involve the 4'-methyl group of the ligand and residues I63 and W32 of the receptor.

Interestingly, the docking results showed that, although with a different arrangement, the two

phenyl rings of compound 6 nicely mimic the VN residues Y27 and Y28, which in the

crystal structure of the uPAR–ATF–SMB complex insert into a large cavity on uPAR's

surface making H-bonding and hydrophobic interactions (27).

It is also noteworthy that 6 interacts with R91 and S88, two key residues of the uPAR

chemotactic epitope, which could interfere with the fMLF-R recognition and affect uPAR-

dependent cell migration.

Since compound 37 has two chiral centers and the chemical sample from the NCI is a

racemic mixture with undetermined diastereomeric ratio, we docked the four diastereomers

(R,S)-37, (R,R)-37, (S,S)-37 and (S,R)-37 generated by Ligprep software (see Matherials and

Methods) into the uPAR-VN binding site and scored them by GLIDE XP. Whether one or

all four of the diastereomers of compound 37 contributed to the inhibition is not entirely

clear, since the top-ranked poses for each isomer within the VN-binding site were not

different and gave comparable G-scores (Table 1).

As depicted in Fig.6B,C,D,E, one of the 4-CF3 fluorine atoms of the diastereomers interacts

tightly with the cationic terminus of both R30 and R116 side chains. The remaining two

fluorine atoms of (R,S)-37 and (R,R)-37 also experience a slightly elongated H-bond with

both S56 and S65 OH group. The 4-CF3 phenyl ring of the four isomers forms hydrophobic

contacts with W32 and I63 and is stabilized by edge-on-face aromatic stacking interactions

with W32. The 2-CF3 group of (R,S)-37, (R,R)-37 and (S,S)-37 is optimally oriented to make




19

H-bonds with the cationic terminus of R91side chain. On the contrary, the 2-CF3 group of

(S,R)-37 engages a H-bond with R58 residue. The F···H-NH2 and F···H-O distances are

below 4 Å, respectively, and the corresponding F···H-NH2 and F···H-O angles are comprised

between 107.0° and 151.0°. All these values fall in the range considered acceptable for

F···H-X (X = O, N or S) H-bonds (36,37). It has been argued that fluorine rarely engages in

H-bonds in small molecule X-ray crystal structures (40,41). However, in protein pockets

where ligands are immobilized by a variety of forces, they appear to be more common. F···X

distances beyond 3.0 Å can be regarded as dipole-dipole interactions that most likely provide

small stabilizing contributions (≤ 1 kcal/mol) for the observed binding poses.

The alcoholic OH group of (R,S)-37 and (R,R)-37 forms H-bonds with the R58 NH2 group

and the S112 CO backbone, while in (S,S)-37 and (S,R)-37 it H-bonds the Q114 CO group

and the R116 NH2 group. With the exception of (R,R)-37, the protonated piperidine N1 of

(R,S)-37, (S,S)-37 and (S,R)-37 donates a H-bond to Q114 CO side chain.

The predicted G-scores could suggest that the (R,S)-37 isomer binds slightly more strongly

than the other ones. But further studies are needed to determine which isomer is the more

active form for binding to uPAR.

From the docking results appears evident that compound 37 entirely fills the VN recognition

pocket and engages numerous H-bonding and hydrophobic contacts with the receptor,

explaining the slightly better affinity in comparison with compound 6 (Table 1). However,

compound 37, differently from compound 6, contacts only partially the chemotactic SRSRY

sequence interacting only with the R91 side chain.

Compounds 6 and 37 inhibit the structural and functional interaction betweeen uPAR and

fMLF-Rs, thus blocking tumor cell invasion. Membrane-bound uPAR functionally interacts

with fMLF-Rs through its SRSRY sequence (residues 88-92) and this interaction is required




20

for both fMLF- and uPA-dependent cell migration (14-16,23,42,43). Therefore, we tested

whether compounds 6 and 37 were able to affect uPAR-dependent cell migration, since they

also target key residues of the uPAR chemotactic domain, namely S88 and R91.

To this aim, in vitro cell migration experiments were performed on uPAR-293 cells. These

cells acquire the ability to migrate toward uPA and fMLF by expressing uPAR; on the

contrary, uPAR negative V-293 cells do not migrate toward uPA and f-MLF, albeit they

express fMLF-Rs (14).

Both compounds strongly inhibited uPAR-293 cell migration toward FCS, whose a main

chemotactic component is VN, as well as toward uPA and fMLF, when plated on collagen-

coated membranes; as a control, V-293 cell migration toward FCS was not affected (Fig.

7A). As expected, given the ability of compounds 6 and 37 to inhibit uPAR-mediated cell

adhesion to VN, migration of uPAR-293 cells on membranes coated with VN confirmed the

results obtained with collagen-coated membranes (Fig. 7A).

Therefore, compounds 6 and 37, targeting R91 of uPAR, not only block cell adhesion to VN,

but also impair cell migration toward serum, uPA and f-MLF by specifically inhibiting the

functional interaction between uPAR and fMLF-Rs.

In order to demonstrate that compounds 6 and 37 acted by inhibiting the structural

interaction between uPAR and fMLF-Rs, coimmunoprecipitation experiments were

performed in uPAR-293 cells, constitutively expressing high levels of FPR1, the high

affinity receptor for f-MLF.

uPAR-293 cells were incubated for 20h in DMEM containing 0.1% BSA in the presence of

compounds 6 and 37, at 2xIC50, or DMSO, as a negative control. Cells lysates were

incubated with the R4 monoclonal anti-uPAR antibody or with non-immune Igs, as a

control. Immunocomplexes were purified with protein-A sepharose and subjected to Western

blot with a polyclonal antibody recognizing FPR1.




21

In DMSO-treated uPAR-293 cells, immunoprecipitation with anti-uPAR antibodies and

Western blot with anti-FPR1 antibodies revealed a band corresponding to FPR1 that was

absent in the same lysate immunoprecipitated with non immune Igs, thus indicating the

existence of a structural interaction between uPAR and FPR1. This interaction was

completely blocked by compound 6 and slighly reduced by compound 37. As a control, no

bands were evidenced in uPAR-293 lysates imunoprecipitated with non-immune Igs and

subjected to Western blot with anti-FPR1 antibodies.

The inhibition of uPAR/FPR1 interaction in cells treated with compounds 6 and 37 was

specific, as evidenced by the presence of the same amount of FPR1 in all cell lysates tested

by Western blot with the anti-FPR1 antibody, as a loading control (Fig.7B).

A growing body of evidence suggests that the uPA/uPAR system promotes tumor metastasis

by several different mechanisms, and not solely through the breakdown of the ECM (30,41).

Among uPAR partners, the fMLP-R family of G-protein-coupled chemotaxis receptors is

crucial in mediating cancer cell invasion and metastasis (42,43). Therefore, we sought to

investigate whether compounds 6 and 37 were able to inhibit the invasion through

reconstituted basal membranes of the highly invasive human fibrosarcoma HT1080 cell line,

constitutively expressing high uPAR levels (42). As shown in Fig. 7C, both compounds

strongly reduced matrigel invasion, as compared to DMSO-treated cells.

The same result was obtained on HCT human colon cancer cells, PC3 human prostate cancer

cells and on MDAMB231 breast cancer cells (Fig. 7D).

Therefore, compounds 6 and 37, two newly identified small molecule inhibitors of uPAR,

are able to block cancer cell invasion by targeting the direct uPAR binding to VN and its

structural and functional interaction with fMLF-Rs.




22

DISCUSSION

uPAR expression and function have been implicated in nearly every step of tumor formation

and progression (30,44). Therefore, the identification of reagents with favorable

pharmacokinetic characteristics capable of interfering with uPAR-mediated signalling is an

area of great interest.

Many of the events of uPAR signalling are dependent on its binding to extracellular ligands,

such as uPA and VN. The main goal of this study was to identify small molecules that, by

inhibiting uPAR binding to VN, would block tumor cell migration and ECM invasion.

We based our work on previous studies demonstrating that, although uPAR plays a pivotal

role in the activation of protease cascades at the cell surface, it promotes metastasis of

human malignancies mostly by engaging VN, through the activation of a cell signaling to

Rac1 (23,24,45) and by encouraging EMT (18,24,25).

We posed that inhibiting uPAR binding to VN could be strongly effective in cancer therapy.

Indeed, uPA/uPAR complex formation enhances pericellular proteolysis through the

activation of plasminogen and, in turn, matrix metalloproteinases (MMPs). However, recent

experimental evidence indicates that some members of the MMP family behave as tumour-

suppressor enzymes and may therefore be regarded as anti-targets in cancer therapy (46).

Moreover, small molecule inhibitors of uPA binding to uPAR have been recently described

(47,48).

Another important advantage of our approach is the specificity of the target; indeed, uPAR

binding to VN takes place when uPAR is overexpressed (Fig.2), a condition mostly of

malignant tumors. Moreover, VN, a circulating adhesive protein, becomes abundantly

associated with extracellular matrix sites upon tissue remodelling, injury/repair, or under

disease conditions (49).




23

Combined biological and biochemical approaches identified the composite site of interaction

between uPAR and VN (26,27). This information facilitated the in silico screening of

compounds targeting the uPAR binding site for VN and the computational evaluation of the

binding to the target protein. Through this approach, which has already been used to

successfully identify inhibitors of Cdc25B dual specificity phosphatases (50), we identified

two small molecules able to inhibit the uPAR-dependent cell binding to VN and, when

applied to cancer cells, to block ECM invasion.

Compounds 6 and 37 disrupted cell adhesion to VN of uPAR transfected HEK-293 cells, and

stopped the activation of signals to ERK and Rac1 in the same cells (Fig.3 and 4). Cells that

do not express uPAR, such as non transfected HEK-293 cells, were insensitive to the effects

of compounds 6 and 37.

Docking studies showed that both compounds make several H-bonds and hydrophobic

interactions with many residues of the VN binding site of uPAR (Fig.6). Compounds 6

deeply extends into the VN binding site of uPAR, nicely mimicking VN itself; compounds

37 entirely fills the VN recognition pocket of the receptor, explaining the slightly better

affinity in comparison with compound 6 (Table 1).

Among others, compounds 6 and 37 also target residues comprised within the uPAR

chemotactic domain (a.acids 88-92), able to mediate uPAR interaction with the fMLF family

of chemotaxis receptors (fMLF-Rs). Indeed, both compounds 6 and 37 do not blocked only

cell adhesion to VN, but also impaired the uPAR and fMLF-Rs dependent cell migration

(Fig.7A).

Compound 6, which interacts with R91 and S88, completely abolished also the structural

interaction between uPAR and the high affinity fMLF receptor, FPR1. Compound 37 which,

differently from compound 6, partially contacts the chemotactic SRSRY sequence

interacting only with the R91 side chain, slightly inhibited the structural interaction between




24

uPAR and FPR (Fig.7B), even though it exerted the same functional effect of compound 6

on cell migration. It is noteworthy that R91, the crucial residue in uPAR binding to VN, as

shown by Ala-scanning studies (18), is also, as demonstrated by our results, the key residue

in uPAR interaction with fMLF-Rs.

It was previously thought that, due to the large surfaces area involved in protein/protein

interface (51), small molecules would not be optimal candidates for disruption of these

interactions. More recently, it was concluded that ‘‘hotspots’’ might be responsible for the

high affinity protein/protein interaction (52). To the best of our knowledge the disruption by

a small molecule of uPAR binding to VN and its consequence on intracellular signaling and

biological outcomes has not been already described, adding novelty to our approach.

Moreover, compounds 6 and 37 also inhibited uPAR interaction with fMLF-Rs that plays a

fundamental role in tumor cell migration, invasion and metastasis (18,22,23,42,43). Indeed,

both compounds did not cause acute cancer cell death (Fig.3A), but blocked ECM invasion

by cells from several cancer types (Fig.7C and D); thus, they may be effective in preventing

the occurrence of metastasis.

In summary, this study employed an integrated drug discovery pipeline consisting of

molecular modeling approaches followed by experimental validation. We performed a

screening of a full diversity compound library based on predicted binding to the uPAR:VN

binding site and identified two novel putative compounds for the treatment of cancer

diseases. In addition, the functional selectivity and specificity of the selected compounds will

allow for further insights in uPAR function and signaling.




25

REFERENCES

1. Blasi F, Carmeliet P. uPAR: a versatile signalling orchestrator. Nat Rev Mol Cell Biol

2002;3:932-43.

2. Ragno P. The urokinase receptor: a ligand or a receptor? Story of a sociable molecule.

Cell Mol Life Sci 2006; 63: 1028-37.

3. Llinas P, Le Du MH, Gårdsvoll H, Danø K, Ploug M, Gilquin B, et al. Crystal

structure of the human urokinase plasminogen activator receptor bound to an

antagonist peptide. EMBO J 2005;24:1655-63.

4. Barinka C, Parry G, Callahan J, Shaw DE, Kuo A, Bdeir K, et al. Structural basis of

interaction between urokinase-type plasminogen activator and its receptor. J Mol Biol

2006;363:482-95.

5. Montuori N, Cosimato V, Rinaldi L, Rea VE, Alfano D, Ragno P. uPAR regulates

pericellular proteolysis through a mechanism involving integrins and fMLF-receptors.

Thromb Haemost. 2013;109:309-18.

6. Montuori N. and Ragno P. Multiple activities of a multifaceted receptor: roles of

cleaved and soluble uPAR. Front Biosci 2009;14:2494-503.

7. Van der Pluijm G, Sijmons B, Vloedgraven H, van der Bent C, Drijfhout JW,

Verheijen J, et al. Urokinase-receptor/integrin complexes are functionally involved in

adhesion and progression of human breast cancer in vivo. Am J Pathol 2001;159:971-

82.

8. Tang CH, Wei Y. The urokinase receptor and integrins in cancer progression. Cell Mol

Life Sci 2008;65:1916-32.

9. Degryse B, Resnati M, Czekay RP, Loskutoff DJ, Blasi F. Domain 2 of the urokinase

receptor contains an integrin-interacting epitope with intrinsic signaling activity:

generation of a new integrin inhibitor. J Biol Chem 2005;280:24792-803.




26

10. Chaurasia P, Aguirre-Ghiso JA, Liang OD, Gardsvoll H, Ploug M, Ossowski L. A

region in urokinase plasminogen receptor domain III controlling a functional

association with alpha5beta1 integrin and tumor growth. J Biol Chem 2006;281:14852-

63.

11. Liu D, Aguirre Ghiso J, Estrada Y, Ossowski L. EGFR is a transducer of the urokinase

receptor initiated signal that is required for in vivo growth of a human carcinoma.

Cancer Cell 2002;1:445-57.

12. Jo M, Thomas KS, Marozkina N, Amin TJ, Silva CM, Parsons SJ, et al. Dynamic

assembly of the urokinase-type plasminogen activator signaling receptor complex

determines the mitogenic activity of urokinase-type plasminogen activator. J Biol

Chem 2005;280:17449-457.

13. Le Y, Murphy PM, Wang JM. Formyl-peptide receptors revisited. Trends Immunol

2002;23:541-48.

14. Montuori N, Carriero MV, Salzano S, Rossi G, Ragno P. The cleavage of the

urokinase receptor regulates its multiple functions. J Biol Chem 2002;277:46932-39.

15. Resnati M, Pallavicini I, Wang JM, Oppenheim J, Serhan CN, Romano M, et al. The

fibrinolytic receptor for urokinase activates the G protein-coupled chemotactic receptor

FPRL1/LXA4R. Proc Nat Acad Sci USA 2002;99:1359-64.

16. de Paulis A, Montuori N, Prevete N, Fiorentino I, Rossi FW, Visconte V, et al.

Urokinase induces basophil chemotaxis through a urokinase receptor epitope that is an

endogenous ligand for formyl peptide receptor-like 1 and -like 2. J Immunol

2004;173:5739-48.

17. Wei Y, Waltz DA, Rao N, Drummond RJ, Rosenberg S, Chapman HA. Identification

of the urokinase receptor as an adhesion receptor for vitronectin. J Biol Chem

1994;269:32380-88.




27

18. Madsen CD, Ferraris GM, Andolfo A, Cunningham O, Sidenius N. uPAR-induced cell

adhesion and migration: vitronectin provides the key. J Cell Biol 2007;177:927-39.

19. Sidenius N, Andolfo A, Fesce R, Blasi F. Urokinase regulates vitronectin binding by

controlling urokinase receptor oligomerization. J Biol Chem 2002;277:27982–90.

20. Carriero MV, Del Vecchio S, Capozzoli M, Franco P, Fontana L, Zannetti A, et al.

Urokinase receptor interacts with alpha(v)beta5 vitronectin receptor, promoting

urokinase-dependent cell migration in breast cancer. Cancer Res 1999;59:5307-14.

21. Degryse B, Orlando S, Resnati M, Rabbani SA, Blasi F. Urokinase/urokinase receptor

and vitronectin/alpha(v)beta(3) integrin induce chemotaxis and cytoskeleton

reorganization through different signaling pathways. Oncogene 2001;20:2032-43.

22. Gargiulo L, Longanesi-Cattani I, Bifulco K, Franco P, Raiola R, Campiglia P, et al.

Cross-talk between fMLP and vitronectin receptors triggered by urokinase receptor-

derived SRSRY peptide. J Biol Chem 2005;280:25225-32.

23. Jo M, Takimoto S, Montel V, Gonias SL. The urokinase receptor promotes cancer

metastasis independently of urokinase-type plasminogen activator in mice. Am J

Pathol 2009;175:190-200.

24. Lester RD, Jo M, Montel V, Takimoto S, Gonias SL. uPAR induces

epithelialmesenchymal transition in hypoxic breast cancer cells. J Cell Biol

2007;178:425-36.

25. Madsen CD, Sidenius N. The interaction between urokinase receptor and vitronectin in

cell adhesion and signalling. Eur J Cell Biol 2008;87:617-29.

26. Gardsvoll H, Ploug M. Mapping of the vitronectin-binding site on the urokinase

receptor: involvement of a coherent receptor interface consisting of residues from both

domain I and the flanking interdomain linker region. J Biol Chem 2007;282:13561–72.




28

27. Huai Q, Zhou A, Lin L, Mazar AP, Parry GC, Callahan J, et al. Crystal structures of

two human vitronectin, urokinase and urokinase receptor complexes. Nat Struct Mol

Biol 2008;15:422-23.

28. Lyne PD. Structure-based virtual screening: an overview. Drug Discovery Today

2002;7:1047–55.

29. Warren L. DeLano "The PyMOL Molecular Graphics System." DeLano Scientific

LLC, San Carlos, CA, USA. http://www.pymol.org/.

30. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The

Protein Data Bank. Nucleic Acids Res 2000;28:235–42.

31. http://dtp.nci.nih.gov/branches/dscb/diversity explanation.htm.

32. http://dtp.cancer.gov.

33. Mazar AP. Urokinase plasminogen activator receptor choreographs multiple ligand

interactions: implications for tumor progression and therapy. Clin Cancer Res

2008;14:5649–55.

34. Friesner RA, Banks JL, Murphy RB, Halgren TA, Klicic JJ, Mainz DT, et al. Glide: A

New Approach for Rapid, Accurate Docking and Scoring. 1. Method and Assessment

of Docking Accuracy. J Med Chem 2004;47:1739–49.

35. Halgren TA, Murphy RB, Friesner RA, Beard HS, Frye LL, Pollard WT, et al. Glide:

A New Approach for Rapid, Accurate Docking and Scoring. 2. Enrichment Factors in

Database Screening. J Med Chem 2004;47:1750–59.

36. Jacobsson M, Gäredal M, Schultz J, Karlén A. Identification of Plasmodium

falciparum spermidine synthase active site binders through structure-based virtual

screening. J Med Chem 2008;51:2777–86.




29

37. Høyer-Hansen G, Pessara U, Holm A, Pass J, Weidle U, Danø K, et al. Urokinase-

catalysed cleavage of the urokinase receptor requires an intact glycolipid anchor.

Biochem J. 2001;358:673-9.

38. Bodary SC, McLean JW. The integrin beta 1 subunit associates with the vitronectin

receptor alpha v subunit to form a novel vitronectin receptor in a human embryonic

kidney cell line. J Biol Chem 1990;265:5938–41.

39. Cheng Y, Prusoff WH. Relationship between the inhibition constant (K1) and the

concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic

reaction. Biochem Pharmacol 1973;22:3099-108.

40. Dunitz JD, Taylor R. Organic Fluorine Hardly Ever Accepts Hydrogen Bonds. Chem

Eur J 1997;3:89.

41. Dunitz JD. Organic Fluorine: Odd Man Out. Chem Bio Chem 2004;5:614.

42. Carriero MV, Longanesi-Cattani I, Bifulco K, Maglio O, Lista L, Barbieri A, et al.

Structure-based design of an urokinase-type plasminogen activator receptor-derived

peptide inhibiting cell migration and lung metastasis. Mol Cancer Ther 2009;8:2708-

17

43. Bifulco K, Longanesi-Cattani, I, Franco P, Pavone V, Mugione P, Di Carluccio G, et

al. Single amino Acid substitutions in the chemotactic sequence of urokinase receptor

modulate cell migration and invasion. PLoS One 2012;7:e44806.

44. Laufs S, Schumacher J, Allgayer H. Urokinase-receptor (u-PAR): an essential player in

multiple games of cancer: a review on its role in tumor progression, invasion,

metastasis, proliferation/dormancy, clinical outcome and minimal residual disease.

Cell Cycle 2006;5:1760–71.




30

45. Kjoller L, Hall A. Rac mediates cytoskeletal rearrangements and increased cell

motility induced by urokinase-type plasminogen activator receptor binding to

vitronectin. J Cell Biol 2001;152:1145–57.

46. Decock J, Thirkettle S, Wagstaff L, Edwards DR. Matrix metalloproteinases:

protective roles in cancer. J Cell Mol Med 2011;15:1254-65.

47. Wang F, Li J, Sinn AL, Knabe WE, Khanna M, Jo I, et al. Virtual screening targeting

the urokinase receptor, biochemical and cell-based studies, synthesis, pharmacokinetic

characterization, and effect on breast tumor metastasis. J Med Chem 2011;54:7193-

205.

48. Wang F, Eric Knabe W, Li L, Jo I, Mani T, Roehm H, et al. Design, synthesis,

biochemical studies, cellular characterization, and structure-based computational

studies of small molecules targeting the urokinase receptor. Bioorg Med Chem

2012;20:4760-73.

49. Preissner KT, Reuning U. Vitronectin in vascular context: facets of a multitalented

matricellular protein. Semin Thromb Hemost 2011;37:408-24.

50. Lavecchia A, Di Giovanni C, Pesapane A, Montuori N, Ragno P, Martucci NM, et al.

Discovery of new inhibitors of Cdc25B dual specificity phosphatases by structure-

based virtual screening. J Med Chem 2012;55:4142-58.

51. Lo Conte L, Chothia C, Janin J. The atomic structure of protein-protein recognition

sites. J Mol Biol 1999;285:2177–98.

52. Arkin MR, Wells JA. Small-molecule inhibitors of protein-protein interactions:

progressing towards the dream. Nat Rev Drug Discov 2004;3:301–17.




31

Table 1. The two low-micromolar uPAR/VN inhibitors indentified by SB-VS.a

Cpd Structure NCI ID isomer

GLIDE GScore

(kcal/mol)

uPAR IC50 (µM)

uPAR Ki

(µM)

LogP

6

2805

-6.54

2.96

0.48

2.93

37

305798

1.23

0.20

4.28

R,S -6.83 R,R -6.81 S,S -6.74 S,R -6.70

a The first and second columns show compound structures, ID used throughout the text and NSC number. The third column shows the four diastereoisomers of compound 37. The fourth column shows the estimated free energy of binding (kcal/mol). The fifth and sixth columns show the IC50 and Ki values as measured in adhesion assays of uPAR-923 cell to VN. The final column shows the predicted LogP value of each compound.




32

FIGURE LEGENDS

FIGURE 1. Chemical structures of compounds selected from NCI Diversity Set II by SB-

VS.

FIGURE 2. Increased cell adhesion to vitronectin in uPAR-transfected HEK-293 cells

can be inhibited by Structrure Based Virtual Screening selected compounds.

A. HEK-293 cells were transfected with uPAR cDNA (uPAR-293) or with empty vector (V-

293) and their lysates were analyzed by 10% SDS-PAGE and Western blot with a polyclonal

anti-uPAR antibody. uPAR-293 cells expressed high amounts of uPAR.

B. V-293 ( ) and uPAR-293 ( ) cells were plated onto vitronectin (VN)- and fibronectin

(FN)-coated wells for 1h, attached cells were fixed and stained with crystal violet; the stain

was eluted and the absorbance (OD) at 540 nm was measured by a spectrophotometer.

Values represent the mean ± SD of three experiments performed in triplicate (*: p< 0.05).

uPAR-293 cells showed increased cell adhesion to VN and decreased cell adhesion to FN, as

compared to V-293 cells.

C. uPAR-293 cells were plated on VN-coated wells for 1 h in the presence of 41 compounds

identified by structure based virtual screening (50 µM) or DMSO (-), as a vehicle control.

Attached cells were fixed and stained with crystal violet; the stain was eluted and the

absorbance (OD) at 540 nm was measured by a spectrophotometer. Values represent the

mean ± SD of three experiments performed in triplicate (*: p< 0.05). Four compounds (6, 14,

24 and 37) significantly inhibited uPAR-293 cell adhesion to VN.

FIGURE 3. Two top-scoring molecules from the in silico screening show specificity for

uPAR-VN interaction and are not toxic to the cells.




33

A. Serum-starved uPAR-293 cells were incubated for 1h and 24 h with 50 µM compound 6

( ), 13 ( ), 24 ( ) and 37 ( ) or DMSO ( ), as a vehicle control; cell survival was then

evaluated by adding MTS-PES and, after incubation at 37°C for 4 h, the absorbance (OD)

was determined at 490 nm by a spectrophotometer. Values are the mean ± SD of three

experiments performed in triplicate (*: p< 0.05). Compounds 6 and 37 did not affect uPAR-

293 cell viability whereas compounds 13 and 24 were cytotoxic.

B. Biotinylated suPAR was placed for 1h at 4°C on VN-coated wells in the presence of 50

µM and 100 µM compounds 6 ( ) and 37 ( ) or DMSO ( ), as a vehicle control. Bound

suPAR was revealed by HRP-avidin and OPD staining; the absorbance at 490 nm was

measured by a spectrophotometer. Values of suPAR binding in BSA-coated wells were

subtracted to obtain specific binding. Values represent the mean ± SD of three experiments

performed in triplicate (*: p< 0.05). 100% values represent suPAR binding to VN in the

absence of compounds.

C. V-293 ( ) and uPAR-293 ( ) cells were plated on VN-coated wells in the presence of 50

µM compounds 6 or 37, or DMSO, as negative control. Attached cells were fixed and

stained with crystal violet; the stain was eluted and the absorbance (OD) at 540 nm was

measured by a spectrophotometer. Values represent the mean ± SD of three experiments

performed in triplicate (*: p< 0.05). Compounds 6 and 37 inhibited uPAR-293 cell adhesion

to VN without exerting any effect on uPAR-negative V-293 cells.

D. uPAR-293 cells were plated on FN- and VN-coated wells in the presence of 50 µM

compounds 6 ( ) or 37 ( ) or DMSO ( ), as a negative control. The attached cells were

fixed and stained with crystal violet; the stain was eluted and the absorbance at 540 nm was

measured by a spectrophotometer. Values represent the mean ± S.D. of three experiments

performed in triplicate (*: p< 0.05). Compounds 6 and 37 inhibited uPAR-293 cell adhesion

to VN without affecting their adhesion to FN.




34

FIGURE 4. Compounds 6 and 37 inhibit uPAR-mediated cell adhesion to VN in a dose-

dependent manner. uPAR-293 cells were plated for 1h on VN-coated wells in the presence

of decreasing concentrations of compounds 6 (A) or 37 (B); attached cells were fixed and

stained with crystal violet, the stain was eluted and the absorbance at 540 nm was measured

by a spectrophotometer. Values represent the mean±SD of experiments performed in

triplicate.

FIGURE 5. Compounds 6 and 37 abolish the changes in cell morphology and the signal

transduction induced by uPAR-mediated cell adhesion to VN.

A. uPAR-293 cells were plated on VN-coated wells for 1h in the presence of compounds 6

or 37, at their 2x IC50, or DMSO, as a vehicle control; attached cells were fixed with 3%

paraformaldehyde for 10 min and photographed by a contrast phase microscope. In the

presence of compounds 6 and 37, uPAR-293 cells did not spread on VN and failed to form

lamellipodia.

B. uPAR-293 cells were plated for 1h on wells coated with increasing concentrations of VN

in the presence of compounds 6 ( ) or 37 ( ), at their 2x IC50, or DMSO ( ), as a vehicle

control; attached cells were fixed and stained with crystal violet, the stain was eluted and the

absorbance at 540 nm was measured by a spectrophotometer. Values represent the mean±SD

of experiments performed in triplicate. The KD of uPAR-293 cell binding to VN in the

presence of vehicle alone is similar to that reported for in vitro uPAR-VN interaction (26

nM) and is increased two-fold by 6 and four-fold by 37.

C. Serum starved uPAR-293 cells were plated on VN-coated wells in the presence of

compounds 6 or 37, at their 2x IC50, or DMSO, as negative control. The cells were harvested

at the indicated times and lysed for Western blot analysis with anti-phospho-ERKs and anti-




35

ERK 2 (as a loading control) antibodies. uPAR-293 cell adhesion to VN activated ERKs;

this effect was abolished by both compounds.

D. Serum starved uPAR-293 cells were plated on VN-coated wells in the presence of

compounds 6 or 37, at their 2x IC50, or DMSO, as negative control. The cells were harvested

at the indicated times and were subjected to Rac-1-GTP pull down assay, as described in

Materials and Methods. The amounts of total (Rac1) and active (Rac1-GTP) protein were

estimated by immunoblotting with an anti-Rac1 antibody. uPAR-293 cell adhesion to VN

activated Rac1; this effect was abolished by both 6 and 37.

FIGURE 6. Binding modes of compounds 6 (A), (R,S)-37 (B), (R,R)-37 (C), (S,S)-37 (D)

and (S,R)-37 (E) into the VN uPAR binding site as calculated by GLIDE.

For clarity, only interacting residues are displayed and labeled. Compound 6 (cyan),

compound 37 (green) and interacting key residues (yellow) are represented as stick models,

while the protein is shown as a sky-blue ribbon model. Chemotactic epitope residues S88,

R91, and Y92 are displayed in magenta. H-bonds are shown as dashed red lines.

FIGURE 7. Compounds 6 and 37 inhibit the structural and functional interaction

betweeen uPAR and fMLF-Rs, thus blocking tumor cell invasion.

A. V-293 ( ) and uPAR-293 ( ) cells were incubated with DMSO, as a vehicle control, or

with compounds 6 ( ) and 37 ( ), at 2x IC50, plated in Boyden chambers coated with

collagen (CG) or vitronectin (VN) and allowed to migrate toward FBS, the aminoterminal

fragment of urokinase (ATF) or fMLF. 100% values represent cell migration in the absence

of chemoattractants. The values are the mean ± S.D. of three experiments performed in

triplicate (*: p< 0.05). The FBS-, ATF- and fMLF-dependent uPAR-293 cell migration was

strongly inhibited by compounds 6 and 37 both on CG and VN.




36

B. Serum starved uPAR-293 cells were incubated for 16h with compounds 6 or 37, at 2x

IC50, or DMSO, as negative control. Cell lysates were immunoprecipitated with the R4 anti-

uPAR monoclonal antibody or with nonimmune serum. The immunoprecipitated samples

were electrophoresed on 10% SDS-PAGE and analyzed by Western-Blot with an anti-FPR1

antibody (right upper panel) or with nonimmune serum (left upper panel). Non

immunoprecipitated cell lysates were subjected to Western blot with the anti-FPR1 antibody

(right lower panel) or with nonimmune serum (left lower panel), as a loading control. uPAR-

co-immunoprecipitates with FPR1; treatment with compounds 6 or 37 significantly reduced

uPAR association to FPR1.

C. Highly invasive HT1080 sarcoma cells were treated with DMSO (-) or with compounds 6

or 37, at their 2x IC50, and allowed to invade Matrigel, using as chemoattractants cell culture

medium without serum (-FCS), to evaluate spontaneous invasion, or with 10% serum

(+FBS). Invading cells on the lower side of the membrane were fixed, stained and

photographed; a representative image is shown.

D. HT1080 sarcoma cells ( ), HCT human colon cancer cells ( ), PC3 human prostate

cancer cells ( ) and MDAMB231 human breast cancer cells ( ) were treated with DMSO

(-) or with compounds 6 or 37, at their 2x IC50, and allowed to invade Matrigel, using as

chemoattractants cell culture medium without serum (-FCS) or with 10% serum (+FBS).

Invading cells on the lower side of the membrane were fixed, stained, counted

(magnification x 200) and the results were expressed as a percentage of the invasion in the

absence of chemoattractant (-FBS). The values are the mean ± S.D. of three experiments

performed in triplicate (*: p< 0.05). FBS-induced cancer cell invasiveness was significantly

reduced by compounds 6 and 37.




FIG.1Research.

on September 11, 2020. © 2013 American Association for Cancermct.aacrjournals.org Downloaded from



FIG.2

A B

C

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

VN FN

OD

OD

0

0.5

1.0

2.0

1.5

**

* *

INH - 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

V-293 uPAR-293

- uPAR

- actin

*

*




FIG. 30

0.2

0.4

0.6

0.8

1

1.2

DMSO 6 37

OD

**

VN

C D

A B

0

1

2

3

1h 24h

*O

D

50 µM 100 µM 0

20

40

60

80

100

120

% o

f inh

ibiti

on

*

FN VN

OD

* *

0

0.2

0.4

0.6

0.8

1

1.2

uPAR-293

** *

**




A B

FIG. 4

Log [6 (M)]

% in

hibi

tion

of c

ell b

indi

ng to

VN

-8-7-6-5-40

50

100

Log [37 (M)]%

inhi

bitio

n of

cel

l bin

ding

to V

N-8-7-6-5-4

0

50

100




A

B C

D

-5 .5 - 5 . 0 -4 .5 -4 .0 -3 .5

0 .0

0 .5

1 .0

1 .5

Log [VN (µM)]

OD

INH: - - 6 37uPAR-293

VN: 0 min 20 min

-pERKs

-ERK2

FIG. 5

DMSO 6 37

uPAR-293

-GTP/rac-1

-rac-1

INH: - - 6 37uPAR-293

VN: 0 min 20 min




A

B C

D E

FIG.6




FIG.7

uPAR-293

IP: Igs anti-uPARWB: non immune Ig

Igs anti-uPAR anti -FPR

INH: - - 6 37 - - 6 37-FPR

-FPR

Stimulus FBS FBS ATF fMLF FBS ATF fMLF

* *%

OF

CEL

L M

IGR

ATI

ON

(ove

r con

trol

) **

**

0

50

100

150

200

250

CG VN

**

* * * *

HT1080

INH: - - 6 37- FBS +FBS

- 6 37 - 6 37 - 6 370

50

100

150

200

250

* * ** * *

INH - 6 37

+ FBS

0

150

300

450

600

750

% O

F C

ELL

INVA

SIO

N (o

ver c

ontr

ol)

**




Published OnlineFirst May 22, 2013.Mol Cancer Ther Vincenza Elena Anna Rea, Antonio Lavecchia, Carmen Di Giovanni, et al. invasionbinding site of the urokinase receptor that block cancer cell Discovery of new small molecules targeting the vitronectin

Updated version

10.1158/1535-7163.MCT-12-1249doi:

Access the most recent version of this article at:

Manuscript

Authorbeen edited. Author manuscripts have been peer reviewed and accepted for publication but have not yet

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://mct.aacrjournals.org/content/early/2013/05/22/1535-7163.MCT-12-1249To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-12-1249

http://mct.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://mct.aacrjournals.org/content/early/2013/05/22/1535-7163.MCT-12-1249


Documents

Discovery of new small molecules targeting the vitronectin ...€¦ · 22/05/2013 · Discovery of new small molecules targeting the vitronectin binding site of the urokinase receptor