Selective urokinase-type plasminogen activator inhibitors. 4. 1-(7-Sulfonamidoisoquinolinyl)guanidines

Article abstract of DOI:10.1021/jm061066t

1-Isoquinolinylguanidines were previously disclosed as potent and selective inhibitors of urokinase-type plasminogen activator (uPA). Further investigation of this template has revealed that incorporation of a 7-sulfonamide group furnishes a new series of potent and highly selective uPA inhibitors. Potency and selectivity can be achieved with sulfonamides derived from a variety of amines and is further enhanced by the incorporation of sulfonamides derived from amino acids. The binding mode of these 1-isoquinolinylguanidines has been investigated by X-ray cocrystallization studies. uPA inhibitor 26 was selected for further evaluation based on its excellent enzyme potency (K_i 10 nM) and selectivity profile (4000-fold versus tPA and 2700-fold versus plasmin). In vitro, compound 26 is able to inhibit exogenous uPA in human chronic wound fluid (IC₅₀ = 0.89 μM). In vivo, in a porcine acute excisional wound model, following topical delivery, compound 26 is able to penetrate into pig wounds and inhibit exogenous uPA activity with no adverse effect on wound healing parameters. On the basis of this profile, compound 26 (UK-371,804) was selected as a candidate for further preclinical evaluation for the treatment of chronic dermal ulcers.

Full text of DOI:10.1021/jm061066t

J. Med. Chem. 2007, 50, 2341-2351
2341
Selective Urokinase-Type Plasminogen Activator Inhibitors. 4.
1-(7-Sulfonamidoisoquinolinyl)guanidines†
Paul V. Fish,*,‡ Christopher G. Barber,‡ David G. Brown,§ Richard Butt,| Michael G. Collis,| Roger P. Dickinson,‡
Brian T. Henry, Valerie A. Horne,§ John P. Huggins,| Elizabeth King, Margaret O’Gara,§ Dawn McCleverty,
Fraser McIntosh,| Christopher Phillips,§ and Robert Webster#
Departments of DiscoVery Chemistry, Molecular Informatics Structure and Design, DiscoVery Biology, Pharmaceutical Sciences, and
Pharmacokinetics and Drug Metabolism, Pfizer Global Research and DeVelopment, Sandwich, Kent, CT13 9NJ, U.K.
ReceiVed September 8, 2006
1-Isoquinolinylguanidines were previously disclosed as potent and selective inhibitors of urokinase-type
plasminogen activator (uPA). Further investigation of this template has revealed that incorporation of a
7-sulfonamide group furnishes a new series of potent and highly selective uPA inhibitors. Potency and
selectivity can be achieved with sulfonamides derived from a variety of amines and is further enhanced by
the incorporation of sulfonamides derived from amino acids. The binding mode of these 1-isoquinoli-
nylguanidines has been investigated by X-ray cocrystallization studies. uPA inhibitor 26 was selected for
further evaluation based on its excellent enzyme potency (Ki 10 nM) and selectivity profile (4000-fold
versus tPA and 2700-fold versus plasmin). In Vitro, compound 26 is able to inhibit exogenous uPA in
human chronic wound fluid (IC50 ) 0.89 µM). In ViVo, in a porcine acute excisional wound model, following
topical delivery, compound 26 is able to penetrate into pig wounds and inhibit exogenous uPA activity with
no adverse effect on wound healing parameters. On the basis of this profile, compound 26 (UK-371,804)
was selected as a candidate for further preclinical evaluation for the treatment of chronic dermal ulcers.
Scheme 1. General Synthesis of
Introduction
1-Guanidino-7-isoquinolinesulfonamidesa
Urokinase-type plasminogen activator (uPA)a (urokinase or
urinary-type plasminogen activator; International Union of
Biochemistry classification number EC.3.4.21.31) is a serine
protease of the trypsin family produced by a large variety of
cell types (smooth muscle cells, fibroblasts, endothelial cells,
macrophages, and tumor cells). It has been implicated in cellular
invasion and tissue remodeling. A principal substrate for uPA
is plasminogen, which is converted by cell surface-bound uPA
to the serine protease plasmin. Locally produced, high plasmin
concentrations mediate cell invasion by breaking down the
extracellular matrix. Important processes involving cellular
invasion and tissue remodeling include wound repair, bone
remodeling, angiogenesis, tumor invasiveness, and spread of
metastases.1,2
Conditions of particular interest to us include chronic dermal
ulcers which are a major cause of morbidity in the aging
population. Chronic dermal ulcers are characterized by excessive
uncontrolled proteolytic degradation resulting in ulcer extension,
loss of functional matrix molecules (e.g., fibronectin), and
retardation of epithelialization and ulcer healing. A number of
groups have investigated the enzymes responsible for the
excessive degradation in the wound environment, and the role
of plasminogen activators has been highlighted.3-7 Normal
human skin demonstrates low levels of plasminogen activators
which are localized to blood vessels and identified as tissue-
a (a) R1R2NH, NEt3, CH2Cl2; (b) HNdC(NH2)2‚HCl, NaH or t-BuOK,
DMSO or DME, 90 °C.
† PDB ID: 2jde and 2uwk.
type plasminogen activator (tPA). In marked contrast, chronic
ulcers demonstrate high levels of uPA localized diffusely
throughout the ulcer periphery and the lesion and readily
detectable in wound fluids.
Thus, overexpression of uPA in the wound environment has
the potential to promote uncontrolled matrix degradation and
inhibition of tissue repair. Inhibitors of the enzyme thus have
the potential to promote healing of chronic wounds. Several
* Tel: +44 (0)1304 644589. Fax: +44 (0)1304 651987. E-mail:
paul.fish@pfizer.com.
‡ Department of Discovery Chemistry.
§ Department of Molecular Informatics Structure and Design.
| Department of Discovery Biology.
Department of Pharmaceutical Sciences.
# Department of Pharmacokinetics and Drug Metabolism.
a Abbreviations: urokinase-type plasminogen activator, uPA; tissue-type
plasminogen activator, tPA.
10.1021/jm061066t CCC: $37.00 © 2007 American Chemical Society
Published on Web 04/21/2007

2342 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10
Chart 1. Nonpeptidic uPA Inhibitors
Fish et al.
Scheme 2. Synthesis of 7-Isoquinolinesulfuryl Chlorides 46 and
47. Method 1a
a (a) SOCl2, 80 °C; (b) NaN3, acetone-H2O, 0 °C; (c) 270 °C, Ph2O
[CAUTION]; (d) for 53: POCl3, 110 °C; (e) for 54: PCl5, 140 °C; (f), i.
n-BuLi, THF-Et2O, -78 °C, ii. SO2Cl2, hexane.
Scheme 3. Synthesis of 1,4-Dichloro-7-isoquinolinesulfuryl
Chloride (47). Method 2a
related enzymes such as tPA, which also acts Via production of
plasmin, play a key role in the fibrinolytic cascade. Hence, it is
important that an inhibitor has adequate potency and selectivity
for uPA relative to both tPA and plasmin to avoid the possibility
of antifibrinolytic side effects.
Two types of nonpeptidic, low molecular weight inhibitors
of uPA are emerging (Chart 1).8 Various aromatic amidines,
e.g., BABIM, have been reported to inhibit uPA.9 These
compounds are generally weak and/or nonselective for uPA
relative to related serine proteases. Bridges disclosed a series
of benzothiophene amidines (e.g., B-623) with significantly
greater potency and selectivity with respect to tPA and plas-
min,10 and Wendt reported a series of potent naphthyl amidines
(e.g., 1).11 There are few reports of guanidine derivatives as
uPA inhibitors. Amiloride is a weak but selective inhibitor of
uPA,12 and various substituted phenylguanidines are reported
to have a similar level of potency (e.g., 2).13 Our own efforts
identified 1-isoquinolinylguanidine 3 (UK-356,202) as a potent
and selective inhibitor of uPA.14
The compounds described herein are potent reversibly
competitive inhibitors of uPA enzymatic activity, with selectivity
for uPA relative to the fibrinolytic enzymes tPA and plasmin.
Chemistry. The methods for the synthesis of inhibitors
disclosed in Tables 1-5 (compounds 4-45) are described in
Schemes 1-5. The general method for the synthesis of
1-guanidino-7-isoquinolinylsulfonamides is outlined in Scheme
1. Sulfuryl chlorides 46 and 47 were reacted with a number of
amines to give the corresponding sulfonamides 48 and 49
respectively, and the guanidine group was then introduced by
reaction with guanidine free base to give compounds 4-10, 36,
40, 41, 44, and 45.
a (a) NCS, MeCN; (b) ClSO3H, 100 °C; (c) POCl3, MeCN.
a Curtius-like rearrangement15 of the cinnamyl azide 51 to give
the isoquinolone 52 which was then converted to the 1-chlor-
oisoquinoline 53 by treatment with POCl3, or to the 1,4-
dichloroisoquinoline 54 by reaction with PCl5. 7-Bromoiso-
quinolines 53 and 54 were then converted to the 7-sulfonyl
chlorides 46 and 47 respectively by low-temperature lithiation
(BrfLi exchange) with n-BuLi and then reaction with sulfuryl
chloride. Sulfuryl chlorides 46 and 47 prepared by this method
were used in situ without isolation. An improved synthesis of
47 is shown in Scheme 3. It had been shown that 1(2H)-
isoquinolinone (55) undergoes electrophilic nitration at the 5
and 7 positions16 and so it was proposed that introduction of
the 4-Cl atom prior to electrophilic sulfonylchlorination should
promote attack at the 7-position by sterically blocking the
5-position; this strategy proved successful. Thus, chlorination
of 55 with N-chlorosuccinimide (NCS) gave 56, reaction with
neat ClSO3H then furnished 57, and finally treatment of 57 with
POCl3 gave 47.
The preparation of sulfonamides derived from amino acids
is shown in Scheme 4. A few sec-sulfonamides (58: R1 ) H)
were further functionalized at this stage by N-alkylation to give
7-Isoquinolinylsulfonyl chlorides 46 and 47 were prepared
by two methods (Schemes 2 and 3). The first method employed

SelectiVe Urokinase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10 2343
Scheme 4. Synthesis of 7-Isoquinolinesulfonamides Derived
Scheme 5. Synthesis of 1-Guanidino-7-isoquinolinesulfonamides
from Amino Acidsa
Derived from Amino Acidsa
a (a) RO2C-X-NH-R1, NEt3, CH2Cl2; (b) cyclopentylmethanol, DEAD,
PPh3, THF; (c) for 38b: K2CO3, MeI; (d) for 39b: K2CO3, Me2NCH2CH2Cl,
DMF.
tert-sulfonamides. Glycine derivative 13b (58: X ) CH2, R )
Et, R1 ) H) was reacted with cyclopentylmethanol under
Mitsunobu conditions to give 19b, and cycloleucine derivative
28a (58: X ) C(CH2)4, R ) Et, R1 ) H) was converted to
38b by treatment with K2CO3-MeI, or to 39b by reaction with
K2CO3-Me2NCH2CH2Cl. The guanidine group was then in-
troduced into the template by reaction of 58 with guanidine free
base to give 1-guanidinoisoquinolines 59 (Scheme 5). For the
inhibitors which incorporate a protected carboxy group, the acid
was unmasked at this stage. Methyl and ethyl esters 59 (R )
Me or Et) were hydrolyzed with dilute NaOH and deprotection
of tert-butyl esters 59 (R ) t-Bu) was performed with HCl or
TFA. Acids 26 and 33 were converted to the amides 42, 43,
and 37 respectively by initial treatment with oxalyl chloride to
give the corresponding acid chloride and then reaction with the
appropriate amine.
a (a) HNdC(NH2)2‚HCl, NaH or t-BuOK, DMSO or DME or DMA, 90
°C; (b) NaOH, MeOH; (c) HCl or TFA; (d) (COCl)2, DMF, CH2Cl2; (e)
NH3, CH2Cl2; (f), amine, CH2Cl2.
ample scope to further explore the 7-position of the 1-isoquino-
linyl guanidine template, and we decided to investigate 7-sul-
fonamides. We reasoned that an appropriately positioned
sulfonamide group would promote an opportunity to gain
additional affinity through a backbone H-bonding interaction
with Gly-216 of uPA in a similar manner to those interactions
observed with other small, noncovalent serine protease inhibi-
tors.20 In addition, we hypothesized that a polar sulfonamide
group would also confer improved solubility and lower lipo-
philicity for this series of uPA inhibitors, and these combined
properties may aid delivery of compound into the wound
environment.
Compound Design
In our earlier studies with heterocyclic guanidines as uPA
inhibitors,17,18 we showed that 1-isoquinolinylguanidine had
moderate affinity for inhibition of uPA (1900 nM). This template
was then further developed by the incorporation of a m-benzoic
acid at the 7-position of the isoquinoline ring to yield compound
3 (17 nM) which is a potent and selective inhibitor of uPA.14
We believe the 100-fold increase in potency is due to a binding
interaction between the benzoic acid group and solvent-exposed
Arg-217 of the enzyme active site in a similar manner to the
binding between the glutamic acid of H-Glu-Gly-Arg-CH2Cl
(EGR-cmk) and Arg-217 as described in the first crystal
structure of uPA.19 This hypothesis may also explain the high
degree of selectivity demonstrated by 3 for uPA compared to
tPA and plasmin, as the corresponding residue in both these
enzymes is Leu-217, and no similar interaction is possible.
The synthesis of 7-isoquinolinylsulfonyl chlorides 46 and 47,
and precedented chemistry for the introduction of the guanidine
group, allowed the rapid exploration of SAR in this series. The
first analogue, primary sulfonamide 4, demonstrated encouraging
potency for uPA (Ki 280 nM) and selectivity over tPA and
plasmin, and both potency and selectivity were further enhanced
by the incorporation of a 4-chloro atom in the isoquinoline
template 6 (Ki 140 nM) (Table 1). This observation is consistent
with all previous examples in the heterocycle guanidine series,
and so the 4-Cl atom was then retained for all further analogues.
Secondary sulfonamides 5 and 7 and tertiary sulfonamide 8,
all derived from simple lipophilic amines, also gave an
enhancement in uPA inhibitory potency compared to 4 and 6,
With compound 3 selected for clinical development for the
treatment of chronic dermal ulcers, our objective was to discover
an agent that would have improved potency, high selectivity
over tPA and plasmin, and physicochemical properties suitable
for the topical delivery to an open chronic wound. There was

2344 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10
Table 1. 1-Guanidino-7-sulfonamidoisoquinolines
Fish et al.
Ki (nM)a
compound
R1
NR2R3
uPA
tPA
plasmin
4
5
6
7
8
9
10
11
12
13
14
15
H
H
NH2
NHPh
NH2
NH-cyclopentyl
pyrrolidinyl
morpholinyl
4-methyl-1-piperazinyl
2-NHC6H4CO2H
3-NHC6H4CO2H
NHCH2CO2H
NHCH2CH2CO2H
NHCH2C(Me)2CO2H
280
160
140
71
130
330
550
86
59
48
58
43
24000
15000
13000
3200
>100 µM
>100 µM
>30 µM
ND
ND
33000
9800
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
>100 µMb
>30 µMc
>30 µM
ND
ND
>100 µM
>100 µM
>30 µM
>100 µM
ND
>30 µM
>100 µM
ND
a Calculated Ki (see Experimental Section). b <50% inhibition @ 100 µM. c <50% inhibition @ 30 µM. d ND ) not determined.
Table 2. Substituted Glycine Derivatives
Ki (nM)a
compound
R1
R2
uPA
tPA
plasmin
16
17
18
19
20
21
22
23
24
25
Me
Ph
CH2Ph
CH2cyclopentyl
H
H
H
H
H
H
H
H
H
R-Me
S-Me
R-CHMe2
S- CHMe2
R-CMe3
S-Me
54
29
35
31
28
21
19
66
22
27
>100 µMb
>100 µM
>100 µM
>100 µM
NDc
ND
38000
>100 µM
>100 µM
>100 µM
>100 µM
>100 µM
>100 µM
23000
ND
15000
16000
45000
>100 µM
ND
CH2Ph
a Calculated Ki. b <50% inhibition @ 100µM. c ND ) not determined.
whereas sulfonamides 9 and 10, derived from polar amines, were
inferior to 6 but still better than having no substituent at the
7-position.
with the larger groups. Substitution on the amino acid R-carbon
proved to be a more successful with sulfonamides derived from
alanine, valine, and tert-leucine 21, 22, 24 (Ki 20-30 nM) hav-
ing potency approaching that of compound 3. Although inhibi-
tors derived from the two enantiomers of alanine (20 and 21)
could be accommodated equally, when amino acids incorporat-
ing larger alkyl R-groups were introduced there was a preference
for the R-isomer (22 versus 23). Compound 25 incorporated
both N-alkyl and R-carbon groups (i.e., a hybrid of 18 and 21)
and again demonstrated that the N-alkyl group contributed little
to the binding of the inhibitor (compare 21 and 25).
The results of the alanine isomers, 20 and 21, prompted us
to explore R,R-disubstituted glycine derivatives (Table 3). The
geminal dimethyl compound 26 proved to be our most potent
(Ki 10 nM) and selective inhibitor to date, and potency could
be further enhanced by incorporation of a small alkyl ring 27-
29 (Ki 6-8 nM). Comparison of cycloleucine derivative 28 (Ki
5.8 nM) with cyclopentyl sulfonamide 7 (Ki 71 nM) shows that
the acid group bestows a 10 fold increase in potency. Further
modification of the cyclohexyl spiro-fused ring system (29)
demonstrated that both polar groups, such as tetrahydropyran
These preliminary results gave encouragement that our
objective should be achievable in this series. We then prepared
a number of 7-sulfonamides derived from amino acids in an
attempt to further improve affinity for uPA through (i) an
interaction between the carboxylate with Arg-217, (ii) a sul-
fonamide H-bond interaction with Gly-216, and (iii) additional
nonspecific lipophilic binding between the amino acid carbon
framework and the lipophilic residues of the active site of uPA.
The first examples derived from aminobenzoic acids, glycine,
and â-alanines, compounds 11-15 respectively (Table 1), once
again gave a modest enhancement in affinity (Ki 40-80 nM)
while retaining selectivity for tPA and plasmin. Of these, glycine
derivative 13 (Ki 48 nM) was selected for further modification,
and the effects of substitution on the sulfonamide N atom, on
the amino acid R-carbon, and on both of these positions
simultaneously were investigated (Table 2). The incorporation
of a small selection of N-alkyl and N-aryl sulfonamides, com-
pounds 16-19, gave at best less than a 2-fold increase in affinity

SelectiVe Urokinase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10 2345
Table 3. R,R-Disubstituted Glycine Derivatives
Ki (nM)a
compound
R1; R2
uPA
tPA
plasmin
26
27
28
29
30
31
Me, Me
CH2CH2CH2
CH2CH2CH2CH2
CH2CH2CH2CH2CH2
CH2CH2OCH2CH2
CH2CH2N(Me)CH2CH2
10
7.4
5.8
8
7.4
11
40000
27000
16000
14000
39000
41000
71000
>100 µMb
38000
>100 µM
>100 µM
NDc
a Calculated Ki. b <50% inhibition @ 100µM. c ND ) not determined.
Table 4. Cyclic Derivatives
a calculated Ki. b <50% inhibition @ 100µM. c ND ) not determined.
30, and basic groups, such as N-methyl piperidine 31, could be
accommodated at this position.
accommodated at the R-carbon. Alcohol 36 and carboxamide
37 were prepared as the preferred R-enantiomers. Comparison
of 36 (Ki 160 nM) and 37 (Ki 54 nM) with proline isomer 33
(Ki 9.9 nM) and the unsubstituted pyrrolidinyl sulfonamide 8
(Ki 130 nM) confirmed the importance of the acid for affinity
but also highlighted that appropriately functionalized carboxa-
mides may also be potent inhibitors of uPA.
A number of sulfonamides derived from cyclic amino acids
were prepared (Table 4). The analogues 32-35 demonstrate
that the position of the acid group is important for gaining good
affinity and the sulfonamides derived from the two enantiomers
of proline, 33 and 34, again demonstrated a preference for the
R-isomer (cf. 22 and 23). The proline derivative 33 was then
modified to see if the carboxylic acid was essential for good
potency and selectivity, or if an alternative group could be
The final set of sulfonamides were derived from the cyclo-
leucine derivative 28 (Ki 5.8 nM), our most potent uPA inhibitor
prepared so far (Table 5). Substitution of the sulfonamide N

2346 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10
Table 5. Cycloleucine Derivatives
Fish et al.
Ki (nM)a
compound
R1
R2
uPA
tPA
plasmin
ND
38
39
40
41
42
43
44
45
Me
CO2H
CO2H
CO2Et
CH2OH
11
2.6
14
67
30
10
1.6
1.4
NDc
23000
23000
ND
CH2CH2NMe2
32000
H
H
H
H
H
H
>100 µMb
ND
CONHCH2CH2NMe2
CONHCH2CH2OH
CON(CH2CH2) 2O
CON(CH2CH2)2NMe
ND
>100 µM
>100 µM
>30 µM
65000
108000
34400
34700
a Calculated Ki. b <50% inhibition @ 100µM. c ND ) not determined.
Table 6. Comparison of Physicochemical Properties of 3 and 26
Table 7. Selectivity vs Enzymes of the Coagulation Cascade for 3 and
26
physicochemical propertya
3
26
385.8
Ki /nM or % inhib
MW
Log D7.4
340.8
2.3
0.0
enzymea
3
26
pKa
2.35, 8.4
60-120 µM
<1.0 µg/mL
99.8
3.4, 7.7
1500 µM
2.3 µg/mL
96.7
thrombin (b)
Factor VIIa (h)
Factor IXa (h)
Factor Xa (h)
2% inhib @ 10 µM
3% inhib @ 10 µM
1240 nM
ND
ND
solubility: human plasma
PBS (pH 6.5)
PPB: human
13% inhib @ 12.5 µM
0% inhib @ 10 µM
0% inhib @ 10 µM
a
PBS: phosphate buffered saline. PPB: plasma protein binding.
a Human (h), bovine (b). b ND ) not determined.
atom was revisited in this series (cf. Table 2), and, as before,
the introduction of a Me group (38) had no beneficial effect.
However, the incorporation of a pendent amino group 39 gave
a very potent (Ki 2.6 nM) and selective inhibitor, once again
showing an additional basic center could be accommodated at
an appropriate position. The search for a nonacid group was
also more extensively explored in this series with compounds
40-45. Ethyl ester 40 had good affinity (Ki 14 nM), but once
again compounds that incorporated an alcohol (41, Ki 67 nM)
contributed little to binding (cf. 7, Ki 71 nM). Following the
encouraging result with proline carboxamide 37, a small series
of carboxamides were prepared derived from 28. Secondary
carboxamides with a pendent amine 42 or, more preferably, a
pendent alcohol 43 were tolerated; however, our most potent
and selective uPA inhibitors were derived from tertiary car-
boxamides. Morpholine amide 44 and N-methylpiperazine amide
45 are both potent inhibitors of uPA (Ki ca. 1-2 nM) and
demonstrate >20000-fold selectivity over tPA and plasmin.
Throughout the course of analogue preparation, compounds
were evaluated for their physicochemical properties, and selected
results for 26, along with 3, are shown in Table 6. Sulfonamide
26 is of an acceptable size (as measured by MW) and reasonable
lipophilicity (Log D7.4) and is within our target range for topical
delivery. The ionization constant of the guanidine group in 26
(pKa 7.7) is somewhat lower than usual and warrants additional
comment. There are a number of factors that contribute to the
weak basicity of the guanidine: (i) the electron-deficient nature
of the isoquinoline ring, (ii) a putative internal H-bond between
an NH of the guanidine and the N atom of the isoquinoline
ring; this interaction serves to hold the guanidine essentially in
the plane of the isoquinoline,17 and (iii) the electron-withdrawing
effects of the 7-sulfonamide substituent.
Table 8. Comparison of in Vitro Potency of 3 and 26
3
26
in Vitro primary potency, measured Ki/µM 0.017 ( 0.0007 0.012 ( 0.002
in Vitro potency in porcine wound
8.35 ( 2.5
0.39 ( 0.05
homogenate, IC50/µM
in Vitro potency in human wound fluid,
IC50/µM
6.91 ( 1.18
0.89 ( 0.26
solubility of 26 compared to 3 (>10-fold) is likely to aid delivery
of the compound from a suitable formulation into the wound
environment.21,22 The lower lipophilicity of 26, compared to 3,
translated to a lowering of plasma protein binding (PPB) and a
higher free drug fraction (16-fold).
The selectivity of 3 and 26 for uPA over several serine
proteases which have important functions in the coagulation
cascade was assessed (Table 7).23 No significant inhibition of
these enzymes was observed as measured by chromogenic
enzyme assays.24 Furthermore, the effects of 3 upon thrombin
time (TT), activated partial thromboplastin time (APTT), and
prothrombin (PT) times of human plasma were evaluated as
functional measures of the potential effects upon the coagulation
cascade. Compound 3 did not significantly affect TT, APTT,
or PT over the concentration range 0.3-100 µM where clinically
relevant changes in clotting times are regarded as >2-fold
increase in normal values.
On the basis of the enhanced potency and selectivity profile,
and attractive physicochemical properties, 26 was selected for
further evaluation using two in Vitro models for uPA inhibition
in physiological wound samples (Table 8). The ability of 26 to
inhibit exogenous uPA added to porcine acute wound tissue
homogenate, and exogenous uPA added to human chronic
wound fluid, was measured. Compound 26 demonstrated an
ability to inhibit uPA in both these systems to a greater extent
than compound 3. This enhanced ability of 26 to inhibit
uPA is explained by the improved affinity for uPA in conjunc-
tion with the significantly greater free fraction. The drop off in
The solubility of the inhibitors was routinely measured in
phosphate-buffered saline (PBS), and then the solubility of
preferred compounds was measured in human plasma as a
surrogate for human chronic wound fluid. The increased

SelectiVe Urokinase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10 2347
Figure 1. Effect of 26 on wound healing parameters. Compound 26:
white bars, n ) 7 (mean ( SEM); vehicle: gray bars, n ) 8 (mean (
SEM).
Figure 2. Inhibitor 33 (blue) in uPA X-ray showing enzyme surface
and key residues.
potency of 3 and 26 in these assays compared to the uPA
primary assay is simply a function of the reduction of the free
drug levels by nonspecific binding to tissue homogenate or
plasma proteins.
It should be noted that there is no animal model of chronic
wound healing that demonstrates uPA upregulation which is
predictive of efficacy in human chronic dermal ulcers. Therefore,
26 was evaluated in an acute (normal) wound model.25,26 It was
not expected that a uPA inhibitor would have a beneficial effect
on wound healing parameters in this model as uPA levels are
not elevated. However, as uPA is present in acute wounds, it
was important to demonstrate that uPA inhibition by 26 is not
detrimental to normal processes of dermal repair. Thus, the aim
of this in ViVo study was to determine the ability of topically
applied 26 to inhibit exogenous uPA, to determine topical and
systemic pharmacokinetics, and to demonstrate safety with
respect to wound closure parameters. A porcine acute excisional
wound model was employed as pig is a ‘tight skinned’ mammal
like human and has similar acute healing processes. An
excisional wound model would allow quantification of healing
parameters over a period of several days.
Two female pigs were subjected to eight excisional wounds.
The wounds were dressed and treated daily for 10 days with
either 1 mL of a 10 mg/mL formulation of 26 in hydrogel
vehicle, or hydrogel vehicle alone (control). On day 11 the
animals were sacrificed, terminal blood samples were taken to
assess any systemic exposure of the compounds, and the wounds
were excised from the surrounding normal skin. The central
portion of each wound was fixed and mounted for histological
analysis of wound re-epithelialization. The remaining wound
tissue was homogenized and analyzed for its ability to inhibit
uPA activity.
Compound 26 demonstrated no adverse effects on acute
wound closure in this model. Comparison of results from
wounds treated with 26 and vehicle, or with vehicle alone, as
measured by wound deficit, wound contraction, and wound re-
epithelialization showed no significant difference (Figure 1).
Compound 26 was shown to penetrate into the pig wounds
following topical delivery. Concentrations of 26 in the dermis
were 41.8 ( 9.0 µM. This result was supported by the ex ViVo
assay results which demonstrated 82.3 ( 7.6% inhibition of
exogenous uPA. Hence, compound 26 has the correct physi-
cochemical properties to penetrate wounds and is able to inhibit
uPA activity. In addition, topical dosing of 26 caused no detected
systemic exposure (limit of detection ) 20 nM).
Figure 3. Overlay of the crystal structures of 33 (blue), amiloride
(green), and APC-11421 (magenta). The trajectory of exit from P1 is
maintained.
to cocrystallize one of these inhibitors in the active site domain
of uPA. Compound 33 was cocrystallized with uPA, and X-ray
data to 2.3 Å resolution was collected for the uPA-33 complex
(Figure 2). Compound 33 was unambiguously modeled into the
initial difference maps, and the binding mode for 33 in uPA
was defined. The structure shows the guanidine group making
the previously defined interactions in the S1 pocket. The
4-chlorine atom packs against the side chain of Ser-195, which
has rotated to accommodate the inhibitor perturbing the catalytic
triad. Further direct interactions between uPA and inhibitor 33
are limited to the proline acid moiety which is in good hydrogen-
bonding distance to the Nꢀ of Arg-217 and the main-chain NH
of Gly-218. Also of note, 33 does not gain any affinity through
an interaction with Gly-216, and the excellent selectivity is
clearly not gained through an interaction with the 97 insertion
loop (Thr-97A-Leu-97B) (which is unique to uPA and nonexist-
ent in tPA or plasmin), as 33 does not contact that region of
the uPA active site.
Following the initial crystal structure of uPA in 1995,19
a
number of high resolution, noncovalent inhibitor-uPA complex
crystal structures have been determined and are now available
in the protein data bank (PDB).27-30 The complex with
amiloride,28 which maintains the guanidine group interactions,
shows the same trajectory out of P1 as 33 (Figure 3). Additional
conserved interactions within P1 include the position of a
H-bond acceptor: the nitrogen of the isoquinoline ring of 33
overlaps with the carbonyl moiety of amiloride, accepting a good
H-bond from the conserved water within P1; this water in turn
forms an H-bond to the side chain of Ser190. The significance
of this H-bond network in selectivity of inhibitors over serine
protease members which have an Ala at position 190 (e.g., tPA
Enzyme-Inhibitor Binding
As inhibitor synthesis was underway and our understanding
of SAR was evolving, contemporaneous studies were undertaken

2348 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10
Fish et al.
1
solid. H (DMSO-d6, 400 MHz) δ 7.45 (1H, s), 7.8 (1H, d), 8.0
(1H, d), 8.5 (1H, s), 11.5 (1H, br s) ppm.
and factor Xa) has been established by a series of selective
inhibitors from Axys30 in which a halogen atom displaces the
P1 water. Selectivity of 33 and amiloride may be due to
enhancing this hydrogen-bonding network to the serine, an
interaction not available when an alanine is present.
1,4-Dichloro-7-isoquinolinesulfonyl Chloride (47). POCl3 (9.65
mL, 103.5 mmol) was added to a stirred suspension of 57 (22.1 g,
79.6 mmol) in MeCN (500 mL) at room temperature, and the
mixture was then heated at reflux for 15 h. On cooling, the MeCN
solution was decanted from the insoluble sludge and evaporated in
Vacuo. The residue was extracted with hot EtOAc and evaporated
to leave a solid which was stirred with Et2O (1.2 L) at room
temperature overnight. The ethereal solution was decanted from
the insoluble material and evaporated in Vacuo to give 47 (20 g,
Conclusion
We have shown that introduction of a 7-sulfonamide sub-
stituent into the 1-isoquinolinyl guanidine template has furnished
a new series of potent and highly selective uPA inhibitors.
Potency and selectivity can be achieved with sulfonamides
derived from amines and is further enhanced by the incorpora-
tion of sulfonamides derived from amino acids (e.g., 26-31
and 33). The most potent and selective examples are sulfona-
mides derived from cycloleucine which has either the free
carboxylate (28 and 39) or a carboxamide (44 and 45) as part
of the amino acid moiety. The binding mode of these 1-iso-
quinolinylguanidines in uPA was confirmed by X-ray cocrys-
tallization studies. Sulfonamide ligand 33 adopts a similar
trajectory of exit from the P1 pocket as has been seen previously
although this binding mode could not have been predicted from
the uPA-EGR-cmk structure.
Finally, on the basis of the excellent enzyme potency and
selectivity profile, ability to inhibit exogenous uPA in human
chronic wound fluid, physicochemical properties suitable for
topical application, and safety profile in the porcine excisional
wound model, 26 (UK-371,804) was selected as a candidate
for further preclinical evaluation for the treatment of chronic
dermal ulcers.
1
67 mmol) as a pale yellow solid. H (DMSO-d6, 400 MHz) δ 8.2
(2H, s), 8.5 (1H, s), 8.55 (1H, s) ppm.
Synthesis of 7-Sulfonamides (49). General Method. A solution
of 47 (2.0 mmol) in CH2Cl2 (3 mL) was added to a stirred solution
of the amine (2.4 mmol) and NEt3 (0.70 mL, 5.0 mmol) in CH2Cl2
(3 mL), and the mixture was stirred at room temperature for 24 h.
The mixture was diluted with CH2Cl2 (50 mL), washed with dilute
HCl (2 M), saturated aqueous NaHCO3, and brine, dried (Na2SO4),
and evaporated in Vacuo. The residue was purified by column
chromatography upon silica gel.
Synthesis of 1-Guanidino-7-isoquinolinesulfonamides. Gen-
eral Method. NaH (30 mg, 80% dispersion by weight in mineral
oil, 1.01 mmol) was added in one portion to a stirred suspension
of guanidine hydrochloride (154 mg, 1.61 mmol) in DME (5.0 mL),
and the mixture was heated at 60 °C under N2 for 45 min. A solution
of the 1-chloroisoquinoline (49) (0.40 mmol) in DME (2.0 mL)
was added and the mixture heated at 95 °C for 4 h. The solvents
were evaporated in Vacuo, and the residue was purified by column
chromatography upon silica gel.
Complete experimental procedures, along with spectroscopic and
analytical data for the preparation of all intermediates and target
compounds for the uPA inhibitors 4-45 can be found in the
Supporting Information.
Experimental Section
Preparation of uPA Inhibitors 3, 26, and 33. (4-Chloro-7-
(3-carboxyphenyl)isoquinolin-1-yl)guandine Monohydrate (3).
The preparation of 3 has been reported previously.14 off-white solid.
mp >300 °C; 1H (DMSO-d6, 300 MHz) δ 7.3 (3H, br s), 7.63 (1H,
t), 7.98 (4H, m), 8.11 (1H, d), 8.27 (1H, s), 8.96 (1H, s) ppm;
LRMS 341, 343 (MH+); Anal. (C17H13ClN4O2S·H2O) C, H, N.
2-{[(4-Chloro-1-guanidino-7-isoquinolinyl)sulfonyl]amino}-
isobutyric Acid Hydrochloride (26). Methyl 2-{[(1,4-Dichloro-
7-isoquinolinyl)sulfonyl]amino}isobutyrate (26b). A mixture of
methyl 2-aminoisobutyrate (1.24 g, 8.07 mmol), NEt3 (2.34 mL,
16.9 mmol), and 47 (2.00 g, 6.74 mmol) in CH2Cl2 (120 mL) was
stirred at 23 °C for 15 h. The mixture was diluted with CH2Cl2
(100 mL), washed with dilute HCl (2 M), saturated aqueous
NaHCO3, and brine, dried (Na2SO4), and evaporated in Vacuo The
residue was purified by column chromatography upon silica gel
using hexane-EtOAc (70:30) as eluant to give 26b (1.13 g, 3.00
Chemistry. General Details. Melting points (mp) were deter-
mined using open glass capillary tubes and either Gallenkamp or
Electrothermal melting point apparatus and are uncorrected. Proton
nuclear magnetic resonance (1H NMR) data were obtained using a
Varian Unity 300 or a Varian Inova 400. Low-resolution mass
spectral (LRMS) data were recorded on a Fisons Instruments Trio
1000 (thermaspray) or a Finnigan Mat. TSQ 7000 (APCI). The
calculated and observed ions quoted refer to the isotopic composi-
tion of lowest mass. Elemental combustion analyses (Anal.) were
determined by Exeter Analytical UK. Ltd. Column chromatography
was performed using Merck silica gel 60 (0.040-0.063 mm). Log
D
7.4 determinations were similar to those described by Stopher and
McClean.31 pKa were determined by Sirius Analytical Instruments
Ltd (UK). The following abbreviations are used: N-chlorosuccin-
imide, NCS; 1,2-dimethoxyethane, DME; dimethyl sulfoxide,
DMSO; trifluoroacetic acid, TFA.
1
mmol) as a white solid. mp 159.5-161 °C (EtOAc); H (CDCl3,
General Methods for the Synthesis of 1-Guanidino-7-iso-
quinolinesulfonamides. 4-Chloro-1(2H)-isoquinolone (56). A
solution of NCS (9.66 g, 72 mmol) in MeCN (80 mL) was added
dropwise to a stirred solution of 1-(2H)-isoquinolone (55) (10 g,
69 mmol) in MeCN (250 mL) which was being heated under reflux.
The mixture was heated under reflux for an additional 1.5 h and
then cooled to room temperature. The resulting precipitate was
collected by filtration, with MeCN rinsing, and then dried in Vacuo
to give 56 (11.3 g, 62.9 mmol) as a pale pink solid. 1H (DMSO-d6,
300 MHz) δ 7.5 (1H, s), 7.6 (1H, dd), 7.8-7.9 (2H, m), 8.25 (1H,
d), 11.5 (1H, br s), ppm; LRMS 180, 182 (MH+), 359, 361, 363
(M2H+).
4-Chloro-1-oxo-1,2-dihydro-7-isoquinolinesulfonyl Chloride
(57). 56 (20.62 g, 115 mmol) was added portionwise to stirred
chlorosulfonic acid (61 mL, 918 mmol) at 0 °C. The mixture was
heated at 100 °C for 3.5 d and then cooled to room temperature.
The reaction mixture was added in small portions onto ice-water
[CAUTION], and the resulting precipitate was collected by filtra-
tion. The solid was washed with water, triturated with MeCN, and
then dried in Vacuo to give 57 (18.75 g, 67.4 mmol) as a cream
400 MHz) δ 1.5 (6H, s), 3.7 (3H, s), 5.55 (1H, s), 8.25 (1H, d),
8.35 (1H, d), 8.5 (1H, s), 8.9 (1H, s) ppm; LRMS 377 (MH+).
1-{[(4-Chloro-1-guanidino-7-isoquinolinyl)sulfonyl]amino}-
isobutyric Acid Methyl Ester (26a). NaH (199 mg, 80% dispersion
by weight in mineral oil, 6.63 mmol) was added in one portion to
a stirred solution of guanidine hydrochloride (1.05 g, 10.6 mmol)
in DMSO (20 mL), and the mixture was heated at 50 °C under N2
for 20 min. 26b (1.00 g, 2.65 mmol) was added in one portion and
the mixture heated at 80 °C for 6.5 h. The cooled mixture was
poured into water (200 mL) and extracted with EtOAc (3 × 75
mL), and the combined organic extracts were washed with water
and brine, dried (Na2SO4), and evaporated in Vacuo. The residue
was purified by column chromatography upon silica gel using CH2-
Cl2-MeOH-0.880NH3 (95:5:0.5 then 90:10:01) as eluant to give
the product. Recrystallization with EtOAc gave 26a (270 mg, 0.675
mmol) as yellow solid. mp >170 °C (dec); 1H (CD3OD, 300 MHz)
δ 1.4 (6H, s), 3.5 (3H, s), 8.15-8.25 (3H, m), 9.1 (1H, s) ppm;
LRMS 400, 402 (MH+).
2-{[(4-Chloro-1-guanidino-7-isoquinolinyl)sulfonyl]amino}-
isobutyric Acid (26). A solution of NaOH (1 mL, 2 M, 2 mmol)

SelectiVe Urokinase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10 2349
was added to a solution of 26a (204 mg, 0.51 mmol) in MeOH (8
mL), and the mixture was heated at 40-50 °C for 16 h. The cooled
mixture was neutrilized with dilute HCl (1 mL, 2 M) to give a
precipitate. The solid was collected by filtration, with copious water
washing, and then dissolved in concd HCl. The solvents were
evaporated in Vacuo, azeptroping with PhMe, and then dried under
high vacuum to give 26·HCl (68 mg, 0.16 mmol) as a pale cream
solid. mp 258 °C (dec); 1H (CD3OD, 400 MHz) δ 1.45 (6H, s), 8.4
(1H, d), 8.4 (1H, s), 8.45 (1H, d), 8.9 (1H, s) ppm; LRMS 386,
388 (MH+); Anal. (C14H16ClN5O4S·HCl·0.8H2O) C, H, N.
and following the 30 min incubation with substrate using a
SPECTRAMax microplate reader (Molecular Devices Corporation.
Background values were subtracted from the final absorbance
values. Percentage inhibition was calculated and plotted against
compound concentration to generate IC50 values. The enzymatic
Ki was calculated from the known Km of the substrate, 90 µM using
the equation
Ki ) IC50/((1 + ([S]/Km))
N-[(4-Chloro-1-guanidino-7-isoquinolinyl)sulfonyl]-D-pro-
line Hydrochloride (33). N-[(1,4-Dichloro-7-isoquinolinyl)sul-
fonyl]-D-proline tert-Butyl Ester (33b). A mixture of d-proline
tert-butyl ester hydrochloride (340 mg, 1.64 mmol), NEt3 (0.50 mL,
3.6 mmol), and 47 (400 mg, 1.35 mmol) in CH2Cl2 (30 mL) was
stirred at 23 °C for 20 h. The mixture was diluted with CH2Cl2 (50
mL), washed with dilute HCl (2 M), saturated aqueous NaHCO3,
and brine, dried (MgSO4), and evaporated in Vacuo to give 33b
(550 mg, 1.28 mmol) as a white solid. mp 80-82 °C; 1H (CDCl3,
400 MHz) δ 1.4 (9H, s), 1.9-2.0 (3H, m), 2.2 (1H, m), 3.4-3.6
(2H, m), 4.4 (1H, m), 8.3 (1H, d), 8.4 (1H, d), 8.5 (1H, s), 8.9
(1H, s) ppm; LRMS 431 (MH+), 448 (MNH4+).
The method for analysis of tPA inhibition was similar to that
for uPA inhibition. The assay utilized final concentrations of tPA
at 0.4 µg/mL with 0.1 mg/mL tPA stimulator, 0.4 mM H-D-Ile-
Pro-Arg-p-nitroanilide dihydrochloride substrate (S-2288), and
various concentrations of inhibitors made up in uPA assay buffer.
Preincubation was carried out with compound, enzyme, and enzyme
stimulator, as for uPA, prior to the incubation with substrate.
Incubation time was 60 min at performed at 37 °C. Data analysis
was identical to that described above for uPA, using a known Km
for tPA of 250 µM.
Plasmin inhibition was assayed by incubating human plasmin at
0.7 µg/mL with 0.2 mM Chromozym-PL (substrate) and various
concentrations of inhibitors in uPA assay buffer. Preincubation was
carried out as for uPA, and the incubation was performed at 37 °C
for 30 min. Data manipulation and percentage inhibition was
calculated as for uPA, using a known Km for plasmin of 200 µM.
The measured Ki of compounds for uPA was determined using
a number of different substrate (30-300µM) and inhibitor (0-
150 nM) concentrations. Compounds were diluted from a 10 mM
stock in DMSO to the required concentration in uPA assay buffer.
The assay format was identical to that carried out for IC50
determination. Following substrate addition the cleavage rate was
measured at 30 s intervals for 30 min at 37 °C at 405 nm). The
rates of substrate cleavage were used to determine Ki for the
compound:enzyme complex.
Determination of Potency in Porcine Wound Homogenate
and Human Wound Fluid. The uPA bioassay used a peptide probe
to quantify the amount of active uPA inhibitor compound in a
particular biological sample. The probe biotin-glutamate-glycine-
arginine-chloromethyl ketone (PFBL-1) molecule is directed to
the active site of uPA (Via glutamate-glycine-arginine) and binds
irreversibly to the enzyme (Via the chloromethyl ketone moiety).
The degree of probe:uPA enzyme complex was then detected and
quantified by colorimetric analysis (Via biotin-streptavidin label).
Our studies have demonstrated that the probe:uPA enzyme interac-
tion, although irreversible, could be blocked in a concentration-
dependent fashion by the presence of competitive uPA inhibitors.
Therefore, the concentration of compound in wound sample governs
the amount of probe-uPA enzyme complex, which is formed at
any particular time. As the glutamate-glycine-arginine moiety was
not exclusively specific for uPA, the amount of specific probe-
uPA enzyme complex formed was separated from other probe-
enzyme complexes by protein electrophoresis.
Assay of uPA Inhibitor Activity in Human Wound Fluid.
Human chronic wound fluids 9001, 9003, and 9004 were obtained
from Dr. M. Stacey. Inhibition of exogenous uPA added to wound
fluid samples was assayed by incubating 0.5 µL of 375 ng/µL
human uPA with 0.5 µL of various concentrations (0.01-100 mL)
of compound dissolved in DMSO, with 2 µL 100% human chronic
wound fluid, at 37 °C for 15 min. This was followed by incubation
with 0.5 µL of 7 mM biotin-Glu-Gly-Arg-chloromethyl ketone
(PLFB-1) in PBS at 37 °C for 10 min. Samples were prepared for
gel electrophoresis by the addition of 15 µL of sample buffer
containing 5% 2-mercaptoethanol and boiling for 5 min.
N-[(4-Chloro-1-guanidino-7-isoquinolinyl)sulfonyl]-D-pro-
line tert-Butyl Ester (33a). Guanidine hydrochloride (220 mg, 2.3
mmol) was added in one portion to a stirred suspension of NaH
(55 mg, 80% dispersion by weight in mineral oil, 1.83 mmol) in
DME (8 mL) and the mixture was heated at 60 °C under N2 for 30
min. 33b (250 mg, 0.58 mmol) was added and the mixture heated
at reflux for 5 h. The cooled mixture was diluted with EtOAc,
washed with water and brine, and dried (MgSO4), and the solvents
were evaporated in Vacuo. The residue was purified by column
chromatography upon silica gel using CH2Cl2-MeOH-0.880NH3
(97:3:0.3) as eluant to give 33a (200 mg, 0.44 mmol) as a yellow
1
solid. mp >170 °C (dec); H (CDCl3, 400 MHz) δ 1.45 (9H, s),
1.7-1.8 (1H, m), 1.8-2.05 (3H, m), 3.3-3.45 (1H, m), 3.5-3.6
(1H, m), 4.3 (1H, dd), 6.3-6.6 (4H, br), 8.05 (1H, d), 8.1 (1H, d),
8.1 (1H, s), 9.2 (1H, s) ppm; LRMS 454, 456 (MH+).
N-[(4-Chloro-1-guanidino-7-isoquinolinyl)sulfonyl]-D-pro-
line (33). 33a (50 mg, 0.11 mmol) was dissolved in a solution of
EtOAc saturated with HCl (10 mL) and the mixture stirred at room
temperature for 2.5 h. The mixture was concentrated in Vacuo,
azeotroping with CH2Cl2, to give 33·HCl (40 mg, 0.092 mmol) as
1
a white powder. mp >200 °C (dec); H (CD3OD, 400 MHz) δ
1.7-1.85 (1H, m), 1.9-2.2 (3H, m), 3.4-3.5 (1H, m), 3.5-3.6
(1H, m), 4.4 (1H, dd), 8.4 (1H, d), 8.45 (1H, s), 8.5 (1H, d), 8.9
(1H, s) ppm; LRMS 397, 399 (MH+). Anal. (C15H16ClN5O4S·
1.0HCl·0.2H2O·0.25CH2Cl2) C, H, N.
Biology. Determination of Inhibitor Potency and Selectivity.
High molecular weight human uPA from urine, 3000 IU/vial
(Calbiochem, 672081), was reconstituted in H2O to give 30000 IU/
mL stock and stored frozen (-18 °C). Chromogenic urokinase
substrate pyro-Glu-Gly-Arg-p-nitroanilide (S-2444), 25 mg/vial
(Quadratech, 820357), was reconstituted in H2O to give 3 mM stock
and stored at 4 °C. Human tPA stimulator (Chromogenix 822130-
63/9) was reconstituted to 1 mg/mL in buffer and used fresh. Human
tPA (one chain), 10 µg/vial (Chromogenix, 821157-039/0), was
reconstituted to 4 µg/mL in buffer and used fresh. S-2288,
chromogenic substrate for serine proteases, 25 mg/vial (Chromoge-
nix, 820852-39), was reconstituted in H2O to give 10 mM stock
and stored at 4 °C. Human plasmin, 2 mg/vial (Quadratech,
810665), was reconstituted to 1 mg/mL in buffer and stored frozen
(-18 °C). Chromozym-PL (Boehringer Mannheim, 378 461), 1 mM
stock in buffer, was prepared fresh.
IC50 and Ki values for compounds were calculated by incubation
of 33 IU/mL uPA with 0.18 mM S-2444 (substrate) and various
compound concentrations, all diluted in uPA assay buffer (75 mM
Tris, pH 8.1, 50 mM NaCl). A preincubation of compound with
enzyme was carried out for 15 min at 37 °C, followed by substrate
addition and further incubation for 30 min at the same temperature.
The final assay volume was 200 µL. Absorbance was read at 405
nM following preincubation (background, time zero measurement)
Assay of uPA Inhibitor Activity in Pig Granulation Tissue
Homogenate. Twelve full thickness 15 mm punch biopsies were
taken from the back of a pig. Granulation tissue was allowed to
accumulate for 3 days, followed by excision of the wound. Pig
wounds were stored at -70 °C and then kept on ice. An 8 mm
punch biopsy was made through the wound bed, and the dermis
was homogenized (30% w/v) in uPA assay buffer (75 mM Tris,

2350 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10
Fish et al.
pH 8.1, 50 mM NaCl) with a hand-held glass homogenizer.
Homogenates were stored at -20 °C.
Christine Ridden, Gary Salmon, Nick Smith, Ross Strang, Dania
Tesei, and Kathleen Welch. We also wish to thank members of
the Department of Physical Sciences for spectroscopic services
and HPLC analyses.
Inhibition of exogenous uPA added to pig wound homogenate
supernatants was assayed by incubating 10 mL of granulation tissue
homogenate supernatant with 1 µL of 375 ng/mL human uPa and
1 µL of PFBL-1 (final concentration of 1µM). This assay mixture
was incubated at 37 °C for 10 min, shaking throughout. Samples
were prepared for gel electrophoresis by the addition of 12 µL of
sample buffer containing 5% 2-mercaptoethanol and boiling for 5
min.
Electrophoresis and Western Blotting. Human wound fluid
samples were analyzed with a gel loading of 5.5 µL/lane and pig
granulation tissue homogenates with a gel loading of 15 µL/lane.
The separation was carried out at 125 V for ∼120 min using an
XCellII Mini-Cell). The gel was removed to transfer buffer (48
mM Tris base, 39 mM glycine, 10% (v/v) methanol) for 15 min
before blotting onto a nitrocellulose membrane using a Trans Blot
SD cell ‘semi-dry’ system (BioRad).
Supporting Information Available: Complete experimental
details, along with spectroscopic and analytical data, for the
preparartion of all intermediates and target compounds for the uPA
inhibitors 4-45. A reaction scheme for a synthesis of amino acid
derivatives which are not readily available. A detailed account of
the enzyme-inhibitor modeling and X-ray structure determination
of both uPA and trypsin. Experimental details for the uPA and
trypsin X-ray structure determinations. Materials and methods for
solubility measurements in PBS and human plasma. This material
is available free of charge via the Internet at http://pubs.acs.org.
References
(1) Epstein, F. H. Cutaneous wound healing. New Eng. J. Med. 1999,
341, 738-746.
(2) Frankenne, F.; Noel, A.; Bajou, K.; Sounni, N. E.; Goffin, F.; Masson,
V.; Munaut, C.; Remacle, A.; Foidart, J. M. Molecular interactions
interactions urokinase plasminogen activator (uPA), its receptor
(uPAR) and its inhibitor, plasminogen activator inhibitor-1 (PAI-1),
as new targets for tumour therapy. Emerg. Ther. Targets 1999, 3,
469-481.
(3) Stacey, M. C.; Burnand, K. G.; Mahmoud-Alexandroni, M.; Gaffney,
P. J.; Bhogal, B. S. Tissue and urokinase plasminogen activators in
the environs of venous and ischaemic leg ulcers. Br. J. Surg. 1993,
80, 596-599.
(4) Palolahti, M.; Lauharanta, J.; Stephens, R. W.; Kuusela, P.; Vaheri,
A. Proteolytic activity in leg ulcer exudate. Exp. Dermatol. 1993, 2,
29-37.
Detection of uPA Inhibition. The nitrocellulose membrane was
blocked with 3% bovine serum albumin (BSA) in water at room
temperature for 30 min and washed twice for 5 min in 2 mM Tris-
HCl, 50 mM NaCl (TBS)/0.05% Tween (TTBS). The membrane
was then incubated with Streptavidin-AP conjugate (1 U/mL in TBS
for 30 min) and washed twice in TTBS for 7.5 min before the
addition of BCIP/NBT (one tablet in 10 mL water) for 5 min. The
color development was halted by washing with water. The
nitrocellulose was dried, and the intensity of staining of the uPA
bands was assessed using the imaging densitometer (BioRad Model
GS-700). The % inhibitions and 50% inhibition values of the added
uPA were calculated.
Topical Application of compound 26 to Porcine Acute
Excisional Wounds. Compound 26 was assessed in a model of
porcine acute excisional wound healing to ensure no adverse effects
or inhibition of any measureable aspect of wound healing and a
measure of topical and systemic pharmacokinetics.
Formulation. The vehicle gel for this study was made from the
following reagents: Glycerol (5%), Lutrol F127 (2%), Blanose
Carboxy Methyl Cellulose (CMC) Hydrogel (7HF) (3%), and water
(to 100%). Lutrol F127 was dissolved and the glycerol dispersed
in water at 4 °C. Once at room temperature, CMC was added during
vortex mixing until fully dissolved. For production of gel containing
26, solid compound was passed through a 180 µM pore sieve and
added to the glycerol/water/F127 mixture, prior to addition of CMC.
The final 26 concentration in the suspension was 10 mg/mL. Vehicle
or compound containing gel was then loaded into 2 mL sterile
syringes and stored at 4 °C prior to use. Dose levels: 26; 1 mL of
10 mg/mL (3.18 mg/cm2, 10 mg/day) applied topically at daily
intervals.
Animal Study. Female pigs (crossbreed of Danish country,
Duroc, and Yorkshire) were subjected to eight full thickness
excisional skin wounds of 20 mm diameter (four on each side of
the spine), using a circular knife. Wounds were dressed and treated
daily for 10 days with 1 mL of a 10 mg/mL formulation of
compound formulated in hydrogel or hydrogel alone (control). Each
wound was assessed daily for inflammation, hemorrhage exudation,
necrosis, hypergranulation, granulation tissue deposition, and wound
area by planimetry. On day 11 the animals were sacrificed and
terminal blood samples were taken to assess any systemic exposure
of the compounds. Following treatment the wounds were excised
from the surrounding normal skin. The central portion of each
wound was fixed and mounted for histological analysis of wound
re-epithelialization. The remaining wound tissue was homogenized
as described above and analyzed for its ability to inhibit uPA activity
using the assay described above.
(5) Rogers, A. A.; Burnett, S.; Moore, J. C.; Shakespeare, P. G.; Chen,
W. Y. J. Involvement of proteolytic enzymes - plasminogen activators
and matrix metalloproteinases - in the pathology of pressure ulcers.
Wound Rep. Reg. 1995, 3, 273-283.
(6) Rogers, A. A.; Burnett, S.; Lindholm, C.; Bjellerup, M.; Christensen,
O. B.; Zederfeldt, B.; Peschen, M.; Chen, W. Y. J. Expression of
tissue-type and urokinase-type plasminogen activator activities in
chronic venous leg ulcers. VASA, J. Vasc. Dis. 1999, 28, 101-105.
(7) Wysocki, A. B.; Kusakabe, A. O.; Chang, S.; Tuan, T.-L. Temporal
expression of urokinase plasminogen activator and gelatinase-B in
chronic wound fluid switches from a chronic to acute profile with
progression to healing. Wound Rep. Reg. 1999, 7, 154-165.
(8) For reviews see: (a) Steinmetzer, T. Synthetic urokinase inhibitors
as potential antitumor drugs. IDrugs 2003, 6, 138-146. (b) Sturze-
becher, J.; Schweinitz, A.; Schmalix, W. A.; Wikstrom, P. Synthetic
urokinase inhibitors as potential anti-invasive drugs IDrugs 2001, 4,
677-683. (c) Magill, C.; Katz, B. A.; Mackman, R. L. Emerging
therapeutic targets in oncology: urokinase-type plasminogen activator
system. Emerg. Ther. Targets 1999, 3, 109-133.
(9) Geratz, J. D.; Shaver, S. R.; Tidewell, R. R. Inhibitory effect of
amidino-substituted heterocyclic compounds on the amidase activity
of plasmin and high and low molecular urokinase and on urokinase-
induced plasminogen activation. Thromb. Res. 1981, 24, 73-83.
(10) Bridges, A. J.; Lee, A.; Schwartz, E.; Towle, M. J.; Littlefield, B.
A. The synthesis of three 4-substituted benzo[b]thiophene-2-car-
boxamides as potent and selective inhibitors of urokinase. Bioorg.
Med. Chem. 1993, 1, 403-410.
(11) Wendt, M. D.; Geyer, A.; McClellan, W. J.; Rockway, T. W.;
Weitzberg, M.; Zhao, X.; Mantei, R.; Stewart, K.; Nienaber, V.;
Klinghofer, V.; Giranda, V. L. Interaction with the S1b-pocket of
urokinase: 8-heterocycle substituted and 6,8-disubstituted 2-naph-
thamidine urokinase inhibitors Bioorg. Med. Chem. Lett. 2004, 14,
3063-3068.
(12) Vassalli, J.-D.; Belin, D. Amiloride selectively inhibits the urokinase-
type plasminogen activator. FEBS Lett. 1987, 214, 187-191.
(13) Yang, H.; Henkin, J.; Kim, K. H.; Greer, J. Selective inhibition of
urokinase by substituted phenylguanidines: quantitative structure-
activity relationship analyses. J. Med. Chem. 1990, 33, 2956-2961.
(14) Barber, C. G.; Dickinson, R. P.; Fish, P. V. Selective urokinase-
type plasminogen activator (uPA) inhibitors. Part 3: 1-Isoquinoli-
nylguanidines. Bioorg. Med. Chem. Lett. 2004, 14, 3227-3230.
(15) Eloy, F.; Deryckere, A. Synthese d’isocarbostyriles et de chloro-1-
isoquinoleines. HelV. Chim. Acta 1969, 52, 1755-1761.
(16) Kawazoe, Y.; Yoshioka, Y. Studies on hydrogen exchange IX.
Electrophilic deuteration of 2-pyridinol, 2-quinolinol and 1-isoquino-
linol. Chem. Pharm. Bull. 1968, 16, 715-720.
Acknowledgment. We would like to thank the following
colleagues for their expert and enthusiastic technical as-
sistance: Yvonne Ailwood, Stephane Billotte, Gerwyn Bish,
David Bull, Usa Datta, Gwen Easter, Cecile Glairet, Keith
Holmes, Stephen Irving, Linda Kitching, Emma Newstead,
(17) Barber, C. G.; Dickinson, R. D.; Horne, V. A. Selective urokinase-
type plasminogen activator (uPA) inhibitors. Part 1: 2-Pyridi-
nylguanidines. Bioorg. Med. Chem. Lett. 2002, 12, 181-184.

SelectiVe Urokinase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 10 2351
(18) Barber, C. G.; Dickinson, R. D. Selective urokinase-type plasminogen
activator (uPA) inhibitors. Part 2: (3-Substituted-5-halo-2-pyridi-
nylguanidines. Bioorg. Med. Chem. Lett. 2002, 12, 185-187.
(19) Spraggon, G.; Phillips, C.; Nowak, U. K.; Ponting, C. P.; Saunders,
D.; Dobson, C. M.; Stuart, D. I.; Jones, E. Y. The crystal structure
of the catalytic domain of human urokinase-type plasminogen
activator. Structure 1995, 3, 681-691. Available in the Brookhaven
Protein Data Bank as pdb 1lmw.ent.
(20) For reviews see: (a) Sanderson, P. E. J. Small, noncovalent serine
protease inhibitors. Med. Res. ReV. 1999, 19, 179-197. (b) Vacca,
J. P. New advances in the discovery of thrombin and factor Xa
inhibitors. Curr. Opin. Chem. Biol. 2000, 4, 394-400.
(21) Sloan, K. B.; Wasdo, S. Designing for topical delivery: prodrugs
can make the difference. Med. Res. ReV. 2003, 23, 763-793.
(22) Benson, H. A. E. Transdermal drug delivery: Penetration enhance-
ment techniques. Curr. Drug DeliVery 2005, 2, 23-33.
(23) Rang, H. P.; Dale, M. M.; Ritter, J. M.; Moore, P. K. Haemostasis
and Thrombosis. In Pharmacology, 5th ed.; Elsevier: New York,
2003; pp 314-329.
(26) Watcher, M. A.; Wheeland, R. G.; The role of topical agents in the
healing of full-thickness wounds. J. Dermatol. Surg. Oncol. 1989,
15, 1188-1195.
(27) Nienaber, V. L.; Davidson, D.; Edalji, R.; Giranda, V. L.; Klinghofer,
V.; Henkin, J. Structure-directed discovery of potent non-peptidic
inhibitors of human urokinase that access a novel binding subsite.
Struct. Fold. Des. 2000, 8, 553-563.
(28) Zeslawska, E.; Schweinitz, A.; Karcher, A.; Sondermann, P.; Sperl,
S.; Strurzebecker, J.; Jacob, U. Crystals of the urokinase type
plasminogen activator variant bc-uPA in complex with small molecule
inhibitors open the way towards structure based drug design. J. Mol.
Biol. 2000, 301, 465-475.
(29) Zeslawska, E.; Jacob, U.; Schweinitz, A.; Coombs, G.; Bode, W.;
Madison, E. Crystals of the urokinase type plasminogen activator
complexes reveal the binding mode of peptidomimetic inhibitors. J.
Mol. Biol. 2003, 328, 109-118.
(30) Katz, B. A.; Sprengeler, P. A.; Luong, C.; Verner, E.; Elrod, K.;
Kirtley, M.; Janc, J.; Spencer, J. R.; Breitenbucher, J. G.; Hui, H.;
McGee, D.; Allen, D.; Martelli, A.; Mackman, R. L. Engineering
inhibitors highly selective for the S1 sites of Ser190 trypsin-like serine
protease drug targets. Chem. Biol. 2001, 8, 1107-1121.
(31) Stopher, D.; McClean, S. An Improved Method for the Determination
of Distribution Co-efficient. J. Pharm. Pharmacol. 1990, 42, 144.
(24) Danilewicz, J. C.; Abel, S. M.; Brown, A. D.; Fish, P. V.;
Hawkeswood, E.; Holland, S. J.; James, K.; McElroy, A. B.;
Overington, J.; Powling, M. J.; Rance, D. J. Design of Selective
Thrombin Inhibitors Based on the D-Phe-Pro-Arg Sequence. J. Med
Chem. 2002, 45, 2432-2453.
(25) Singer, A. J.; McClain, S. A. Development of a porcine excisional
wound model. Acad. Emerg. Med. 2003, 10, 1029-1033.
JM061066T