4-Aminoarylguanidine and 4-aminobenzamidine derivatives as potent and selective urokinase-type plasminogen activator inhibitors

Article abstract of DOI:10.1016/S0960-894X(02)00312-8

The structure-based design of potent and selective urokinase-type plasminogen activator (uPA) inhibitors with 4-aminoarylamidine or 4-aminoarylguanidine S1 binding groups, is described.

Full text of DOI:10.1016/S0960-894X(02)00312-8

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2023–2026
4-Aminoarylguanidine and 4-Aminobenzamidine Derivatives
as Potent and Selective Urokinase-type Plasminogen
Activator Inhibitors
Jeffrey R. Spencer, Danny McGee,y Darin Allen,{ Bradley A. Katz, Christine Luong,
Martin Sendzik,* Neil Squires and Richard L. Mackman*,x
Celera, 180 Kimball Way, South San Francisco, CA 94080, USA
Received 8 November 2001; accepted 29 March 2002
Abstract—The structure-based design of potent and selective urokinase-type plasminogen activator (uPA) inhibitors with 4-
aminoarylamidine or 4-aminoarylguanidine S1 binding groups, is described. # 2002 Elsevier Science Ltd. All rights reserved.
The serine protease, urokinase-type plasminogen
activator (uPA) plays a key role in several biological
processes, including tissue remodeling, cell migration,
and matrix degradation.1À3 High levels of uPA are
associated with many different tumor types and a strong
correlation between the level of uPA and poor prog-
nosis has been noted.1 Furthermore, a small, moder-
ately potent uPA inhibitor, B428 (1; Ki=0.21 mM), has
also shown efficacy in models of prostate and breast
cancer.4À7 Based on these promising findings there has
been an increased interest in the development of potent
and selective uPA inhibitors as cancer therapeutics.
these inhibitors to those based on benzimidazole 2 or
indole 3.8
The key reaction in preparing the inhibitors described in
Table 1 is the formation of an amide bond between an
arylamine and a salicylic acid derivative (Scheme 1).
The inhibitor 2 (Fig. 1), identified from screening a
library of amidine-bearing heterocycles, was previously
developed into a series of potent and selective uPA
inhibitors based on the 5-carboxamidine substituted
benzimidazole or indole, (e.g., 3) heterocycle.8 It was
discovered that the S1 binding group could be replaced
by a simple amide-linked 4-aminobenzamidine group
(e.g., 4a) without a loss in affinity toward uPA. The
more facile synthesis of 4a prompted the development
of this lead in parallel to the development of 2. This
report discusses the results of a structure-based design
approach toward the optimization of 4a and contrasts
Figure 1. Selected inhibitors of uPA.
*Corresponding authors. Tel.: +1-650-866-6533; fax: +1-650-866-
6655; e-mail: martin.sendzik@celera.com
yCurrent address: Corixa, 600 Gateway Blvd., South San Francisco,
CA 94080, USA.
{Current address: Sunesis Pharmaceuticals, 341 Oyster Pt. Blvd.,
South San Francisco, CA 94080, USA.
xCurrent address: Gilead Sciences, 333 Lakeside Dr., Foster City, CA
94404, USA.
Scheme 1. (a) Ac2O, H2SO4; (b) (COCl)2, cat. DMF, EtOAc; (c) aryl-
amine, DMA; (d) NH4OH (or satd NaHCO3), THF.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00312-8

2024
J. R. Spencer et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2023–2026
Table 1. Inhibition profiles of amidine- and guanidine-based uPA inhibitors
No.
R1
X
R2
R
Ki (mM)
30
40
50
uPA
t-PA
Factor Xa
Thrombin
Plasmin
4a
4b
4c
4d
4e
4f
4g
4h
4i
Am
Am
Am
Am
Am
Am
Am
Am
Am
Am
Am
Am
Am
Am
Am
Am
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
Gu
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
H
H
H
H
H
H
H
H
H
H
H
H
H
H
F
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
Me
Ph
H
Br
I
I
I
H
H
H
H
H
H
H
H
Cl
Me
Ph
H
Br
I
Br
I
OMe
OiPr
H
H
H
H
H
H
Cl
Me
OEt
H
H
H
-Naphthyl-
H
H
H
H
H
H
H
H
H
NO2
Br
Me
OMe
H
H
H
H
H
2.7
44
8.0
2.6
3.1
6.0
1.0
0.65
0.60
1.7
10
94
21
6.9
9.4
24
3.6
2.0
1.5
3.9
53
1.8
0.36
0.24
1.7
12
99
15
9.4
5.2
2.8
2.3
37
1.0
5.7
188
12
0.44
0.46
0.52
115
22
185
29
14
12
9.5
10
34
1.0
8.0
120
23
0.28
0.38
0.41
245
90
N.D.
N.D.
N.D.
N.D.
27
6.0
N.D.
N.D.
18
N.D.
4.0
4.4
4j
4k
4l
H
8.5
0.19
0.15
0.080
0.27
6.0
4m
4n
6a
7a
8a
8b
8c
8d
8e
8f
8g
8h
8i
8j
8k
8l
8m
8n
8o
8p
8q
8r
9a
9b
9c
9d
9e
Me
Me
Me
Me
H
NO2
Br
Me
OMe
H
2.9
5.5
256
>75
>75
N.D.
N.D.
>75
N.D.
N.D.
>75
>75
N.D.
N.D.
N.D.
283
24
>75
27
13
20
27
5.5
10
42
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
287
>563
>95
>95
181
>563
338
506
104
>450
315
>470
>470
>470
>470
>470
198
>470
>470
350
>470
>470
151
>470
>470
>470
13
>450
>450
>450
>450
>450
>450
>450
>450
>450
>450
125
>450
>450
>450
205
>450
43
330
H
Cl
Me
OEt
H
H
H
-Naphthyl-
H
H
H
H
H
H
60
>75
H
0.30
8.0
8.5
1.6
Me
Me
205
>75
>75
>75
55
-Naphthyl-
-Naphthyl-
-Naphthyl-
-Naphthyl-
-Naphthyl-
-Naphthyl-
-Naphthyl-
-Naphthyl-
-Naphthyl-
4.6
0.25
0.23
0.085
0.22
0.70
0.28
0.18
240
55
235
68
>78
88
75
>75
>75
>75
>75
>75
N
N
N
N
Br
I
OMe
OiPr
225
150
300
310
195
>450
N.D., not determined; Am, C(¼NH)NH2; Gu, NHC(¼NH)NH2.
Figure 2. (a) X-ray structure of uPA–4a complex. The inhibitor co-binds with a water molecule H2O1S1 near Ser190 and also makes direct hydrogen
bonds with Asp189. The hydroxyl group binds in the oxyanion hole forming H-bonds to NaSer195 and NaGly193 and OgSer195. (b) X-ray structure of
uPA–8l complex. The position ortho to the hydroxyl group is directed toward S20 affording an avenue for further optimization.

J. R. Spencer et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2023–2026
2025
Clean and efficient amide bond formation required the
mixing of the acetyl (or benzyl) protected salicylic
chloride derivative and the aryl amine in dimethyl-
acetamide (DMA).9 The aryl amines that have not been
previously reported were prepared according to Scheme
2. The acids were either commercially available, readily
prepared from commercial materials through standard
halogenation procedures, or synthesized according to
Scheme 3.10
Despite the enhanced activity toward uPA in some
analogues, especially 4n, the selectivity against the other
enzymes is only marginal. In the series of inhibitors
based on 2 a small halo group at the 6-position adjacent
to the amidine, designed to displace the water molecule
(H2O1S1) that co-binds in S1 (Fig. 2a), introduced
excellent selectivity against Ala190 proteases (tPA, fac-
tor Xa, plasma-kallikrein, and thrombin).12,13 Displace-
ment of the water molecule in the Ala190 proteases
leaves the amidine in an environment, lacking a full
complement of hydrogen bonding partners thus leading
to reduced potency. The Ser190 proteases can compen-
sate for the loss of the water molecule by the formation
of a full hydrogen bond between the amidine and
OgSer190. The introduction of a chlorine group at the
6-position on the most potent analogue 4n was found
to generate with 7a a compound with good selectivity
toward uPA and against the Ala190 enzymes. Unfor-
tunately, the potency of 7a toward uPA was reduced by
up to 126-fold. The smaller fluorine group in analogue
6a only marginally improved selectivity against one
enzyme, tPA. In the analogue series based on 2, the
chlorine analogue was also found to be superior to
fluorine in selectivity, but importantly, potency toward
uPA was not significantly reduced by the chlorine sub-
stitution.13 The alternative binding mode between the two
series of molecules is once again believed to be the reason
for the subtly different effects in potency and selectivity.
The lead molecule 4a can display two modes of binding
depending on the serine protease.11 One binding mode is
similar to uPA–2 and involves the formation of an array
of unusually short (<2.3 A) hydrogen bonds between the
inhibitor hydroxyl group, Ser195 and a water molecule
trapped in the oxyanion hole. The alternative mode of
binding, observed in the uPA–4a complex places the
hydroxyl group in the oxyanion hole, hydrogen bonding
to Ser195 (Fig. 2a). The inhibitors were tested for affi-
nity toward a panel of five serine proteases and the
results are shown in Table 1. Substitutions at 50, ana-
logues 4b–e, did not enhance potency toward any of the
enzyme targets despite the fact that this substitution
affects the pKa of the phenol group. This result was
consistent with the results found in a similar series based
on the benzimidazole scaffold 2.8 Potency increased
with select groups at either the 30 or 40 sites (e.g., 4h,i) or
naphthyl analogue 4l. The most noteable inhibitors
were the 30-halo derivatives (4m,n) that enhanced
potency by 5- to 34-fold toward uPA and by up to 79-
fold toward thrombin. This trend compares well with
that observed for similar substititions on the inhibitor
2.8 However, in contrast to 2, the 30-phenyl derivative
4k was a weakuPA inhibitor. This difference is most
likely due to the oxyanion-hole binding mode of these
inhibitors in uPA, which directs the 30-substituent into
residue Ser195 and Cys42 rather than toward the S10
subsite. The size of the group adjacent to the hydroxyl is
therefore limited to sterically small groups in this mode
of binding.
It has been reported that aryl guanidines favor binding
to uPA but not plasmin, tissue-type plasminogen acti-
vator (tPA) and other proteases.14,15 The basis for this
preference is not fully understood but may involve the
Ser190/Ala190 difference within S1. However, plasmin,
a Ser190 enzyme, prefers lysine in S1 and also disfavors
guanidines in general, so other factors must contribute
to this selectivity preference. Replacement of the ami-
dine on inhibitors generated a complementary series of
inhibitors, 8a–m. As predicted, potency against all the
Ala190 enzymes and plasmin was markedly reduced but
in the majority of examples, the uPA activity was also
reduced especially for the promising 30-halides (8m,n).
One analogue, 8l, however, was equipotent to its ami-
dine counterpart 4l and highly selective, >500-fold,
against all the anti-targets.
A high-resolution X-ray structure of 8l in uPA was
solved (Fig. 2b) and reveals that the mode of binding is
very similar to that of the uPA–4a complex. The site
adjacent to the hydroxyl, like the 30-position in analo-
gues 4a–n, is directed toward residue 195 and is pre-
sumably the reason for the reduced activity of the
bromine and iodine analogues 8o and 8p, respectively.
Scheme 2. (a) Guanidine, BuOH, 120 ꢀC; (b) H2, 10% Pd/C, MeOH;
t
(c) HCl, dioxane; (d) Me3Al, NH4Cl.
It was reasoned that oxygen linked groups may afford
better results by providing a sharp directional change.
Analogues 8q,r were indeed found to retain excellent
potency and selectivity toward uPA.16 This result now
opens an avenue for S20 directed modifications to fur-
ther increase potency. Finally, substitution of the phenyl
S1 binding ring for a pyridyl ring on the optimal inhi-
bitors 8l and 8o–r, analogues 9a–e, respectively, led to a
slight increase in potency toward uPA. Inhibitor 9a
Scheme 3. (a) MeOH, 18 N H2SO4; (b) MgCl2, (CH2O)n, Et3N;
(c) BnBr, Cs2CO3; (d) H2O2, H2SO4; (e) alkyl-I, Cs2CO3; (f) KOH.

2026
J. R. Spencer et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2023–2026
(Ki=85 nM) is the most potent and selective inhibitor in
the series and has several sites or vectors directed
towards the S1–S20 binding regions for further optimi-
zation.
acetate and treated with oxalyl chloride (10 equiv) and a few
drops of DMF. After 60 min, the mixture was concentrated,
dissolved in DMA and treated with the arylamine (1.2 equiv).
After 17 h, the mixture was treated with concentrated ammo-
nium hydroxide. The product was collected by filtration. Puri-
fication if necessary was carried out by HPLC methods.
Intermediates having hydroxyl groups protected as benzyl
ethers were coupled using the same procedure except without
the acetylation step. Benzyl removal was performed by hydro-
genation over 10% palladium on carbon catalyst (1 atm).
10. Preparation of 17 and 18: A methanolic solution of 16 was
treated with HCl (g) at 0 ꢀC for 15 min, refluxed for 2 h, and
then concentrated. The methyl ester product was formylated
(see ref 17) to yield an equimolar mixture of the 4-formyl and
8-formyl isomers. The desired product methyl, 4-formyl-3-
hydroxynaphthoic acid was soluble in hot methanol and could
be separated by filtration from the methyl, 8-formyl-3-hydroxy-
naphthoic acid isomer. The product was treated with 1 equiv
of Cs2CO3 in DMF for 60 min followed by 1 equiv of benzyl
bromide to give the benzyl ether. The ether, dissolved in
methanol: 18 N H2SO4 (25:1), was stirred with 35% (w/w)
hydrogen peroxide solution (6.0 equiv) for 19 h and con-
centrated. Further treatment of the product with 1.1 equiv of
Cs2CO3 in DMF for 60 min followed by addition of the
necessary alkyl halide (3.0 equiv) gave the desired benzyl pro-
tected ether product. Treatment with KOH in MeOH removed
the methyl ester to give the desired acid.
In summary, a series of amidine-based, nonselective
uPA inhibitors and a series of guanidine-based, selective
inhibitors of uPA have been developed. The inhibitors
display a unique mode of binding toward uPA, which
results in a subtly different SAR to a closely related
series of uPA inhibitors. The accumulated structural
information and SAR for the selective inhibitors now
provides a solid platform for further optimization in the
S1–S20 subsites.
Acknowledgements
The authors thankP. Sprengeler for modeling assis-
tance and also J. Janc, K. Radika and J. Wang for Ki
determinations. We are also indebted to J. Knolle for
advice and support in the course of this research.
References and Notes
11. Katz, B. A.; Elrod, K. C.; Luong, C.; Rice, M.; Mackman,
R. L.; Sprengeler, P. A.; Spencer, J. R.; Hataye, J.; Janc, J.;
Link, J.; Litvak, J.; Rai, R.; Rice, K.; Sideris, S.; Verner, E.;
Young, W. B. J. Mol. Biol. 2001, 307, 1451.
12. Katz, B. A.; Sprengeler, P. A.; Luong, C.; Verner, E.;
Elrod, K.; Kirtley, M.; Janc, J.; Spencer, J. R.; Breitenbucher,
J. G.; Hui, H. C.; McGee, D.; Allen, D.; Martelli, A.; Mack-
man, R. L. Chem. Biol. 2001, 8, 1107.
1. Andreasen, P. A.; Kjøller, L.; Christensen, L.; Duffy, M. J.
Int. J. Cancer 1997, 72, 1.
2. Magill, C.; Katz, B. A.; Mackman, R. L. Emerg. Ther.
Targets 1999, 3, 109.
3. Rabbani, S. A.; Xing, R. H. J. Int. Oncol. 1998, 12, 911.
4. Rabbani, S. A.; Harakidas, P.; Davidson, D. J.; Henkin, J.;
Mazar, A. P. Int. J. Cancer 1995, 63, 840.
5. Alonso, D. F.; Tejera, A. M.; Farias, E. F.; Joffe, E. B.; De,
K.; Gomez, D. E. Anticancer Res. 1998, 18, 4499.
6. Xing, R. H.; Mazar, A.; Henkin, J.; Rabbani, S. A. Cancer
Res. 1997, 57, 3585.
13. Mackman, R. L.; Katz, B. A.; Breitenbucher, J. G.; Hui,
H. C.; Verner, E.; Luong, C.; Sprengeler, P. A. J. Med. Chem.
2001, 44, 3856.
14. Sperl, S.; Jacob, U.; Arroya De Prada, N.; Sturzebecher,
¨
7. Evans, D. M.; Sloan-Stakleff, K. D. Invasion Metastasis
1998/99, 18, 252.
J.; Wilhelm, O. G.; Bode, W.; Magdolen, V.; Huber, R.;
Moroder, L. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 5113.
15. Yang, H.; Henkin, J.; Kim, K. H.; Greer, J. J. Med.
Chem. 1990, 33, 2961.
16. Squires N.; McGee, D.; Allen, D.; Hruzewicz, W.; Spencer,
J. R.; Sprengeler, P.; Katz, B.; Luong, C., Mackman, R.
Abstract of Posters, 221st National Meeting of the American
Chemical Society, San Diego, CA, April 1–5, 2001; MEDI 293.
17. Hofsløkken, N. U.; Skattebøl, L. Acta Chem. Scand. 1999,
53, 258.
8. Verner, E.; Katz, B. A.; Spencer, J. R.; Allen, D.; Hataye,
J.; Hruzewicz, W.; Hui, H. C.; Kolesnikov, A.; Li, Y.; Luong,
C.; Martelli, A.; Radika, K.; Rai, R.; She, M.; Shrader, W.;
Sprengeler, P. A.; Trapp, S.; Wang, J.; Young, W. B.; Mack-
man, R. L. J. Med. Chem. 2001, 44, 2753.
9. The salicylic acids were first acetylated by dissolution in
acetic anhydride and a catalytic amount of 18 N H2SO4. The
product after concentration and drying, was dissolved in ethyl