Structure-Based design, synthesis and SAR of a novel series of thiopheneamidine urokinase plasminogen activator inhibitors

Article abstract of DOI:10.1016/S0960-894X(01)00247-5

The serine protease urokinase plasminogen activator (uPA) is thought to play a central role in tumor metastasis and angiogenesis. Molecular modeling studies suggest that 5-thiomethylthiopheneamidine inhibits uPA by binding at the S1 pocket of the active site. Further structure based elaboration of this residue resulted in a novel class of potent and selective inhibitors of uPA.

Full text of DOI:10.1016/S0960-894X(01)00247-5

Bioorganic& Medicinal Chemistry Letters 11 (2001) 1379–1382
Structure-Based Design, Synthesis and SAR of a Novel Series of
Thiopheneamidine Urokinase Plasminogen Activator Inhibitors
Nalin L. Subasinghe,* Carl Illig, James Hoffman, M. Jonathan Rudolph,
Kenneth J. Wilson, Richard Soll, Troy Randle, David Green, Frank Lewandowski,
Marie Zhang, Roger Bone, John Spurlino, Renee DesJarlais, Ingrid Deckman,
Christopher J. Molloy, Carl Manthey, Zhau Zhou, Celia Sharp, Diane Maguire,
Carl Crysler and Bruce Grasberger
3-Dimensional Pharmaceuticals Inc., 665 Stockton Drive, Exton, PA 19341, USA
Received 6 March 2001; accepted 23 March 2001
Abstract—The serine protease urokinase plasminogen activator (uPA) is thought to play a central role in tumor metastasis and
angiogenesis. Molecular modeling studies suggest that 5-thiomethylthiopheneamidine inhibits uPA by binding at the S1 pocket of
the active site. Further structure based elaboration of this residue resulted in a novel class of potent and selective inhibitors of uPA.
# 2001 Elsevier Science Ltd. All rights reserved.
The serine protease urokinase plasminogen activator
(uPA) is thought to play a central role in tumor metas-
tasis and angiogenesis.1 Proteolytically active uPA is a
disulfide-linked two-chain protein, which is generated
from a largely proteolytically inactive 54-kDa pro-uPA
by hydrolysis of the Lys158-Ile159 peptide bond.2 The
primary function of uPA is to convert plasminogen to
plasmin,3 which can digest components of the extra-
cellular matrix and basement membrane either directly
or indirectly by activating pro-MMPs. Extracellular
matrix remodeling via proteolysis is a key step in tumor
metastasis and angiogenesis. An up-regulation of uPA
in tumor versus normal tissue has been observed for a
wide variety of human cancers.4,5 Furthermore, many
studies have shown that the inhibition of urokinase
enzyme activity can reduce tumor growth and/or
metastasis.5 Given these observations, inhibition of
plasminogen activation by uPA appears to be an
attractive approach for the therapeutic intervention of
tumor growth and metastasis.
phenylguanidines, acylguanidines and bisbenzami-
dines.6 The best of these early uPA inhibitors have mM
potencies and poor selectivity. Several novel uPA inhi-
bitors with nM potency and selectivity towards uPA
have been described in the recent literature. These
include benzothiopheneamidines7 (Eisai), cyclohehyl-
thiopheneamidines8 (Fujisawa), naphthylamidines9
(Abbott), and isoquinolynylguanidines10 (Pfizer). Here,
we describe the design and synthesis of a novel series of
potent and selective uPA inhibitors.
Our efforts began with the search for a novel basic resi-
due that would potentially bind to the S1 site11 of uPA.
After screening several hundred amidines, guanidines,
and amines, 5-methylthiothiopheneamidine 1 (Ki=6
mM) was selected as a suitable P1 residue for structure-
based design. Molecular modeling of 1 and the ben-
zothiophene B623 (Eisai, Ki=0.53 mM) within the
enzyme active site suggested a common binding mode
for both compounds. The arylamidine portion of both
compounds appears to bind within the S1 pocket where
the amidine moiety forms a salt bridge with Asp189 of
uPA. Furthermore, these studies suggested that the
styryl side chain of B623 would fit into a small hydro-
phobicpocket formed by Gly218, Ser146, and Cys191–
Cys220 disulfide bridge.12 Substitutions at the 4-posi-
tion of 1 occupy the same area of space as the styryl side
chain of B623 and could potentially bind within this
There are reports of nonpeptidic, reversible inhibi-
tors of uPA from as far back as the late 1950’s.
These small molecule inhibitors include benzamidines,
*Corresponding author. Tel.: +1-610-458-6066; fax: +1-610-458-
8249; e-mail: nalin@3dp.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00247-5

1380
N. L. Subasinghe et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1379–1382
proximal aryl-binding pocket. To explore this possibi-
lity a series of 4-substituted thiopheneamidines was
synthesized (Table 1). The more promising 4-substituted
thiopheneamidines were further elaborated to identify
structure–activity relationships.
was synthesized by treating the bromoketone 14 with
phenylthioamide (Scheme 2).16
The nitrile 13 was selectively hydrolyzed to the acid by
heating with tetrafluorophthalicacid in a sealed tube for
72 h.17 The acid was converted to the acid chloride and
then treated with trimethylsilyldiazomethane to form
the diazoketone.18 The diazoketone was treated with
HBr in AcOH to give the bromoketone 14. 4-Alkyl and
4-arylthiazolylthiopheneamidines listed in Tables 2 and 3
were synthesized according to Scheme 3. The nitrile 13
was treated with H2S19 to form the thioamide 15, which
was reacted with bromoketones to give a variety of
thiazolylthiophene esters. These thiazolylthiophenes
were converted to the corresponding amidines in a
4-(2-Phenyl-6-pyridyl)thiopheneamidine 2 and 4-(m-
biphenyl)-thiopheneamidine 7 were synthesized as sum-
marized in Scheme 1. The dilithium salt, generated by
treating methyl 4-bromo-2-methylthiothophene-2-car-
boxylate 8 with 2 equiv of nBuLi, was converted to the
stannane 9 by treating with trimethyltin chloride.
The stannane was treated with 3-phenylbromobenzene
and 2,6-dibromopyridine under Stille13 coupling condi-
tions to give the biphenylthiophene 10 and (6-bromo-
pyridyl)thiophene 11, respectively. The (6-bromo-
pyridyl)thiophene 11 was treated with phenylboronic
acid under Suzuki14 coupling conditions to give com-
pound 12. The esters 11 and 12 were converted to the
corresponding amidines 2 and 7 by treating with tri-
methylaluminum and ammonium chloride in refluxing
toluene.15
Scheme 1. Reagents and conditions: (i) BuLi (2 equiv), ClSnMe3; (ii)
TMSCHN2; (iii) RBr, Pd; (iv) PhB(OH)2, Pd(PPh3)4, K2CO3, aq
DMF, 80 ꢀC; (v) AlMe3/NH4Cl, toluene, reflux.
4-Phenylthiopheneamidine 3 and 4-(4-methylthiazo-
lyl)thiopheneamidine 4 were prepared by subjecting the
corresponding esters (Maybridge Chemical Company,
Cornwall, UK) to the amidination conditions described
in Scheme 1. 4-(2-Phenythiazolyl)-thiopheneamidine 5
Scheme 2. Reagents: (i) Tetrafluorophthalicaicd,
Á; (ii) (CO)2Cl2,
DMF cat, CH2Cl2, 3 h; (iii) (a) TMSCHN2, CH3CN 4 h; (b) 30%
HBr/AcOH 1 h; (iv) PhCSNH2, reflux, acetone.
Table 1. uPA inhibitory activity of 4-substituted thiopheneamidines
Table 2. SAR of thiazole substitution
Compound
R
Ki (mM)
>28
2
Compound
R1
R2
Ki (nM)
3
4
5
6
7
0.6
1.0
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Thiophene-2-yl
Benzothiophen-3-yl
2-Naphthyl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
Ph
89
141
138
145
91
108
100
154
164
120
815
178
641
470
305
982
5,6,7,8-Tetrahydronaphth-2-yl
3,4-(OCH2O)Ph
3,4-(OCH2CH2O)Ph
3,4-(OCH2CH2CH2O)Ph
2-Cl-3-Pyridyl
4-Cl-3-Pyridyl
2-Cl-4-Pyridyl
CH3
0.2
0.09
>34
PhOCH2
Cyclohexyl
PhCH2
Ph
Ph

N. L. Subasinghe et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1379–1382
Table 3. SAR of phenyl substitution for compound 6 Table 4. Selectivity for representative compounds
1381
Compound
Ki (mM)
Thrombin Plasmin Chymotrypsin tPA Trypsin
20
41
42
>29a
>30a
>25a
10.8
6.9
32.1
—
1.7
1.9
1.7
0.88
0.86
0.93
>30a
>31a
Compound
R1
R2
Ki (nM)
aNo observable inhibition at this screening concentration.
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
4-CH3
4-CH3PhNHO2S
4-Cl
4-Phenyl
4-NO2
4-CH3O
4-CH3SO2NH
2-CH3O
2-CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
94
272
99
1740
270
327
94
130
277
115
86
442
546
44
substitution on the phenyl ring of compound 6 was also
evaluated (Table 3). Small hydrophobicresidues such as
Me (32) and Cl (34) in the para position can be accom-
modated with no loss in potency.
There is no significant difference between electron-
withdrawing and electron-donating residues. Even
though bulky arylsulfonamides at the para position (38)
retain activity, the rigid phenyl substitution (35) results
in a 10-fold loss of activity. While hydrophobic groups
at the meta position attenuated activity, m-aryl-carbo-
nylamino (46) and m-(methoxycarbonyl)-methoxy (45)
substitutions gave a 2-fold increase in activity.
3-CH3O
3-OH
3-Br
3-CH3
3-CH3OCOCH2O
3-(4-F-PhCONH)
2-CH3O
2-Cl
47
4-CH3O
4-Cl
4-OH
4-Cl
152
287
44
3-CH3
3-NO2
161
Selected compounds were also tested for their inhibitory
activity against other serine proteases. Activities for
three representative compounds are shown in Table 4.
Compound 42 has approximately 300-fold or greater
selectivity over thrombin, plasmin, and chymotrypsin,
and also has 20- and 10-fold selectivity over tPA and
trypsin, respectively.
In conclusion, starting from a weakly active P1 scaffold,
we have constructed a novel series of potent and selec-
tive inhibitors of human uPA. These compounds have
nanomolar potencies and show inhibitory activity in
cell-based assays21 of tumor metastasis. Further results
from structure-based optimization and in vivo biological
results will be reported in due course.
Scheme 3. Reagents: (i) H2S, Et3N, anhyd MeOH, 18 h; (ii) acetone,
reflux, 5 h.
manner similar to that described in Scheme 1. Bromo-
ketones, which were not commercially available, were
generated either by brominating the corresponding
ketones18 or by treating the corresponding carboxylic
acid in a manner similar to that shown in Scheme 2
(steps ii and iii).
References and Notes
1. Rosenberg, S. Annu. Rep. Med. Chem. 1999, 34, 121.
2. Gunzler, W. A.; Steffens, G. J.; Otting, F.; Buse, G.; Flohe,
L. Hoppe-Seyler’s Z. Physiol. Chem. 1982, 363, 133.
3. Sobel, G. W.; Mohler, S. R.; Jones, N. W.; Dowdy, A. B. C.;
Guest, M. M. Am. J. Physiol. 1952, 171, 768.
4. (a) Schmitt, M.; Narbeck, N.; Thomssen, C.; Wilhelm, O.;
Magdolen, V.; Reuning, U.; Ulm, K.; Hofler, H.; Janicke, F.;
Graeff, H. Thromb. Haemostasis 1997, 78, 285. (b) Duffy, M.
J. Clin. Cancer Res. 1996, 2, 613.
5. Andreasen, P. A.; Kjoller, L.; Christensen, L.; Duffy, M. J.
Int. J. Cancer 1997, 72, 1.
6. Doe, J. S.; Smith, J. J.; Roe, R. P. J. Am. Chem. Soc. 1968,
90, 8234.
7. Towle, M. J.; Lee, A.; Maduakor, E. C.; Schwartz, C. E.;
Bridges, A. J.; Littlefield, B. A. Cancer Res. 1993, 53, 2553.
8. Tanaka, A.; Mizuno, H.; Sakurai, M. International Patent
Application WO 9811089; Chem. Abstr. 1998, 128, 230238.
9. Geyer, A. G.; Mcclellan, W. J.; Rockway, T. W.; Stew-
art, K. D.; Weitzberg, M.; Wendt, M. D. International
Patent Application WO 99/05096; Chem. Abstr. 1999, 130,
153476.
The inhibition constants for several 4-substituted thio-
pheneamidines are shown in Table 1.20 Phenyl (3) and
thiazolyl (4) substitution at the 4-position of the thio-
phene ring increases activity by 10- and 8-fold, respec-
tively. Biphenyl (7) substitution resulted in an inactive
compound. Replacing the methyl group at the 4-posi-
tion of the thiazole in compound 4 with a phenyl group
(6) resulted in a further 8-fold increase in activity. Based
on these results, 4-(2-thiazolyl)thiopheneamidine was
chosen as a scaffold for further elaboration. Results in
Table 2 show the effect of substitutions on the thiazole
ring. 4-Thiophene (16) and 4-(3,4-dioxalanyl)phenyl
(20) substitution gave potency that was equivalent to
the phenyl analogue (6). Other aryl substitutions did not
provide a significant enhancement in activity. In con-
trast, 4-alkyl substitutions gave less active compounds.
Alkyl and aryl groups at the 5-position of the thiazole
ring also gave less active compounds. The effect of further

1382
N. L. Subasinghe et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1379–1382
10. Barber, C. G.; Fish, P. V.; Dickinson, R. P. International
Patent Application WO 99/20608; Chem. Abstr. 1999, 130,
311705.
16. Hantzsch, A. R.; Weber, J. H. Chem. Ber. 1887, 20, 3118.
17. Rounds, W. D.; Gribble, G. W. Tetrahedron Lett. 1988,
29, 6557.
11. The binding site for a polypeptide substrate consists of a
series of subsites within the enzyme, which are occupied by the
amino acid residues. By convention these sites are labeled S1,
S2, etc., counting from the carboxy residue of the scis0sile amide
18. Illig, C. R.; Subasinghe, N. L.; Hoffman, J. B.; Wilson, K.
J.; Rudolph, M. J. PCT Int. Appl. WO 0047578, 2000; Chem.
Abstr. 2000, 133, 177162.
19. Ren, W. Y.; Rao, K. V. B.; Klein, R. S. J. Heterocycl.
Chem. 1986, 23, 1757.
0
bond towards the N-terminus of the substrate and S1 , S2 , etc.,
counting from the amino residue of the scissile bond towards
the C-terminus of the substrate. The corresponding amino acid
residues of the substrate are labeled P and P0 in the same way.
20. Enzyme assays were carried out as described in ref 17.
Human kidney cell urokinase was purchased from Sigma
Chemical Co. (St. Louis, MO, USA). N-CBz-Val-Gly-Arg-p-
nitroanilide was used as the fluorogenicsubstrate.
21. Molloy, C. J.; Sharp, C.; Manthey, C. L.; Zhou, Z.; Ran-
dle, T.; Green, D.; Hoffman, J.; Subasinghe, N.; Rudolph, J.;
Wilson, K.; Deckman, I.; Bone, R.; Illig, C. Inhibition of
Pericellular Urokinase Plasminogen Activator Activity and
Basement Membrane Cell Invasion by Novel Small Molecule
Urokinase Inhibitors. In Proceedings of the AACR-NCI-
EORTC New Drugs Meeting, Washington, DC, November
16, 1999, p 113.
12. This hydrophobicpocket in uPA has been termed the S1
b
site in the literature Nienaber, V.; Wang, J.; Davidson, D.;
Henkin, J. J. Biol. Chem. 2000, 275, 7239.
13. Stille, J. K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508.
14. Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457.
15. (a) Bridges, A. J.; Lee, A.; Schwartz, C. E.; Towle, M. J.;
Littlefield, B. A. Bioorg. Med. Chem. 1993, 1, 403. (b) Sidler,
D. R.; Lovelace, T. C.; McNamara, J. M.; Reider, P. J. J. Org.
Chem. 1994, 59, 1231.