Secondary amides of sulfonylated 3-amidinophenylalanine. New potent and selective inhibitors of matriptase

Article abstract of DOI:10.1021/jm051272l

Matriptase is an epithelium-derived type II transmembrane serine protease and has been implicated in the activation of substrates such as pro-HGF/SF and pro-uPA, which are likely involved in tumor progression and metastasis. Through screening, we have identified bis-basic secondary amides of sulfonylated 3-amidinophenylalanine as matriptase inhibitors. X-ray analyses of analogues 8 and 31 in complex with matriptase revealed that these inhibitors occupy, in addition to part of the previously described S4-binding site, the cleft formed by the molecular surface and the unique 60 loop of matriptase. Therefore, optimization of the inhibitors included the incorporation of appropriate sulfonyl substituents that could improve binding of these inhibitors into both characteristic matriptase subsites. The most potent derivatives inhibit matriptase highly selective with K_i values below 5 nM. Molecular modeling revealed that their improved affinity results from interaction with the S4 site of matriptase. Analogues 8 and 59 were studied in an orthotopic xenograft mouse model of prostate cancer. Compared to control, both inhibitors reduced tumor growth, as well as tumor dissemination.

Full text of DOI:10.1021/jm051272l

4116
J. Med. Chem. 2006, 49, 4116-4126
Secondary Amides of Sulfonylated 3-Amidinophenylalanine. New Potent and Selective
Inhibitors of Matriptase†
Torsten Steinmetzer,*, Andrea Schweinitz, Anne Stu¨rzebecher, Daniel Do¨nnecke, Kerstin Uhland,‡ Oliver Schuster,‡
Peter Steinmetzer,§ Friedemann Mu¨ller, Rainer Friedrich,# Manuel E. Than,#,| Wolfram Bode,#,| and Jo¨rg Stu¨rzebecher§
Curacyte Chemistry GmbH, Winzerlaer Strasse 2, D-07745 Jena, Germany, Curacyte DiscoVery GmbH, Deutscher Platz 5d, D-04103 Leipzig,
Germany, Institut fu¨r Vaskula¨re Medizin, Klinikum der UniVersita¨t Jena, Nordha¨user Strasse 78, D-99089 Erfurt, Germany,
Max-Planck-Institut fu¨r Biochemie, Forschungsgruppe Proteinasen, Am Klopferspitz 18, D-82152 Martinsried, Germany, and
Ludwig-Maximilians-UniVersita¨t Mu¨nchen, Department of Chemistry and Biochemistry, Genzentrum, Feodor-Lynen-Strasse 25,
81377 Mu¨nchen, Germany
ReceiVed December 21, 2005
Matriptase is an epithelium-derived type II transmembrane serine protease and has been implicated in the
activation of substrates such as pro-HGF/SF and pro-uPA, which are likely involved in tumor progression
and metastasis. Through screening, we have identified bis-basic secondary amides of sulfonylated
3-amidinophenylalanine as matriptase inhibitors. X-ray analyses of analogues 8 and 31 in complex with
matriptase revealed that these inhibitors occupy, in addition to part of the previously described S4-binding
site, the cleft formed by the molecular surface and the unique 60 loop of matriptase. Therefore, optimization
of the inhibitors included the incorporation of appropriate sulfonyl substituents that could improve binding
of these inhibitors into both characteristic matriptase subsites. The most potent derivatives inhibit matriptase
highly selective with Ki values below 5 nM. Molecular modeling revealed that their improved affinity results
from interaction with the S4 site of matriptase. Analogues 8 and 59 were studied in an orthotopic xenograft
mouse model of prostate cancer. Compared to control, both inhibitors reduced tumor growth, as well as
tumor dissemination.
Introduction
tumor-associated serine proteases of the plasmin/plasminogen
activator system and matrix metalloproteinases are closely
connected in a proteolytic cascade. uPA is assigned to be the
major activator of plasmin, which in turn generates more uPA
in a feedback reaction and also activates MMPs, which are
responsible for the degradation of extracellular matrix proteins.
Cancer is a leading cause of mortality worldwide. Although
single tumors can often be removed by surgery or treated with
chemotherapy or radiotherapy, the patients often die because
of formation of metastases. Metastasis formation includes
complex processes, which also require a series of proteolytic
steps for degradation and remodeling of the extracellular matrix.
After penetration through the basal lamina around the primary
tumor, which mainly consists of collagen, glycoproteins, and
proteoglycans, the tumor cells have to migrate through extra-
cellular matrix proteins in the connective tissue and intravasate
through a basal membrane into the blood or lymph system. After
distribution over the whole body, tumor cells extravasate from
the vascular system and form metastases at different loci.
Proteases of different families, such as matrix metalloproteases
and aspartate, cysteine, and serine proteases, are implicated in
these processes and are also responsible for growth factor
activation and angiogenesis, which are required for normal tumor
growth and progression. Therefore, the development of specific
inhibitors for these proteases is one possible strategy to prevent
metastasis. During the past decade, extensive work has focused
on the development of inhibitors for matrix metalloproteases
(MMP). However, all clinical cancer trials with relatively
nonspecific MMP inhibitors showed only poor efficacy.1 In
addition, the urokinase-type plasminogen activator (uPA), a
serine protease of the plasmin/plasminogen activator system,
emerged as a potential target for antimetastatic drugs.2,3 The
The zymogen pro-uPA can also be activated in vitro by
several proteases other than plasmin, among them the cysteine
proteases cathepsin B and cathepsin L, the aspartate protease
cathepsin D, and the trypsin-like serine proteases plasma kalli-
krein, prostate-specific antigen or tryptase. Recently, it was also
shown that matriptase, an epithelial-derived type II transmem-
brane serine protease,4 converts pro-uPA into enzymatically
active uPA with high efficacy.5 In addition, matriptase is a very
potent activator of hepatocyte growth factor (HGF, also named
scatter factor), which acts as a major motility factor and induces
scattering of cells through binding and activation of its cell sur-
face receptor c-Met. A preferred cleavage sequence of matriptase
also contains the protease-activated receptor-2 (PAR2), and it
has been shown that a soluble form of matriptase can trigger
calcium signaling from PAR2-expressing Xenopus oocytes.5
Enzymatically active matriptase was originally isolated from
human breast cancer cells7,8 and in an inhibited form through
binding to endogenous hepatocyte growth factor activator inhib-
itor-1 (HAI-1) from human milk.9 Independently, it was cloned
as MT-SP1 (membrane-type serine protease 1) from a human
prostatic cancer cell line.10 Matriptase (76 kDa) has a reduced
size when compared with MT-SP1 (95 kDa) because of a
truncation in the integral N-terminal anchor peptide. However,
the catalytic serine protease domain of both enzymes is identical.
Recently, the crystal structures of this catalytic domain in
complex with the small inhibitor benzamidine and with the
bovine pancreatic trypsin inhibitor have been published.11
† Coordinates of 8/matriptase and 31/matriptase have been deposited to
the Protein Data Bank under PDB codes 2gv6 and 2gv7, respectively, and
will be released upon publication of the manuscript.
* To whom correspondence should be addressed. Telephone: +49-3641-
508516. Fax: +49-3641-508507. E-mail: torsten.steinmetzer@curacyte.com.
Curacyte Chemistry GmbH.
‡ Curacyte Discovery GmbH.
§ Klinikum der Universita¨t Jena.
Matriptase emerged as a potential target for the development
of anti-invasiveness drugs to prevent metastasis due to its
# Max-Planck-Institut fu¨r Biochemie, Forschungsgruppe Proteinasen.
| Ludwig-Maximilians-Universita¨t Mu¨nchen.
10.1021/jm051272l CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/13/2006

NoVel Matriptase Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4117
Scheme 1. Synthesis of Inhibitor 8a
potential substrates pro-uPA and pro-HGF and due to its
overexpression in epithelial tumors.12-14 Recently it was
demonstrated that a modest orthotopic overexpression of
matriptase in transgenic mice caused spontaneous squamous cell
carcinoma and potentiated carcinogen-induced tumor formation,
whereas increased expression of HAI-1 completely negated the
oncogenic effects of matriptase.15 Presently, only a few matriptase
inhibitors have been described, among them several bis-
benzamidines with inhibition constants between 0.1 and 1 µM16
and a 14-amino acid peptide isolated from sunflower seeds
having a Ki value of 0.92 nM.17 Recently, a selective arginal-
derived matriptase inhibitor (CVS-3983, Ki ) 3.3 nM) has been
used to suppress the growth of androgen-independent prostate
tumor xenografts.18
One of the archetypical inhibitors of almost all trypsin-like
serine proteases is 3-TAPAP (1),19 which also inhibits matriptase
with an inhibition constant of 14 µM. A thorough screening of
other available sulfonylated 3-amidinophenylalanine derivatives
led to the identification of racemic compound 2 from a
previously described series of uPA inhibitors.20 Compound 2
has a Ki value of 82 nM for matriptase and reduced affinity
toward related trypsin-like serine proteases, such as the clotting
proteases thrombin and factor Xa or the fibrinolytic enzymes
plasmin and uPA.
a Reagents and conditions: (a) 2.0 equiv of NH2OH‚HCl and DIPEA in
ethanol; (b) 2 equiv of (Ac)2O in acetic acid; (c) 2 equiv of NH4Cl in water/
methanol; (d) (Boc)2O in ethyl acetate; (e) H2 and PtO2 as catalyst in 90%
acetic acid; (f) PyBop/DIPEA in DMF; (g) 1 N HCl in acetic acid;
(h) H2 and Pd/C in 90% acetic acid; (i) 1.5 equiv of 1H-pyrazole-1-
carboxamidine‚HCl/DIPEA in DMF, preparative reversed-phase HPLC.
For further optimization of inhibitor 2, we maintained the
central 3-amidinophenylalanine as a key moiety and modified
the N-terminal sulfonyl group and the C-terminal secondary
amide residue. These modifications resulted in a series of new
compounds, which show improved potency and selectivity as
matriptase inhibitors. Two analogues were selected for X-ray
analysis in complex with the catalytic domain of matriptase to
obtain information about their binding mode. This structural
information was used for further inhibitor design. In addition,
two inhibitors were investigated for their anti-invasive potential
and inhibition of tumor growth in an orthotopic xenograft mouse
model of prostate cancer. The results are summarized in this
publication.
led to 4-aminoethylpyridine 4,22,23 which was Boc-protected and
hydrogenated using platinium(IV) oxide as catalyst to give
intermediate 5. Coupling of 3 and 5 using PyBop/diisopropy-
lethylamine resulted in 6. The intermediate obtained after Boc
cleavage was hydrogenated and treated with pyrazole-1-car-
boxamidine24 to give inhibitor 8.
Inhibitors containing a phenoxyarylsulfonyl group (53, 54)
were obtained alternatively by reduction of the appropriate
acetylhydroxyamidino derivative with zinc powder in acetic acid
(3 h, room temperature), because of the instability of such
sulfonyl residues under catalytic hydrogenation conditions using
Pd/C as catalyst.
Inhibitors 56-63 were obtained by mixed-anhydride-mediated
coupling of Boc- or Cbz-â-Ala-OH and its glycine and
γ-aminobutyric acid analogues to 3- or 4-aminobenzenesulfonyl-
3-cyanophenylalanyl-(Boc-amidoethyl)piperidide, which was
obtained by reduction of the appropriate nitrobenzenesulfonyl
derivative with zinc powder in acetic acid. For compounds 62
Results
Chemistry. The compounds were synthesized according to
published procedures. Scheme 1 exemplarily summarizes the
synthesis of inhibitor 8. The cyano group of an appropriate
sulfonylated 3-cyanophenylalanine was converted into an acety-
lated hydroxyamidine 3 by the method of Judkins.21 The
C-terminal part of the molecule was obtained from 4-vinylpy-
ridine as starting material. Treatment with ammonium chloride

4118 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Scheme 2. Synthesis of Inhibitor 12a
Steinmetzer et al.
a Reagents and conditions: (a) compound 5 and PyBop/DIPEA in DMF; (b) zinc powder in acetic acid, 2 h, room temp; (c) Cbz-â-Ala-OH, isobutyl
chloroformate, N-methylmorpholin in DMF; (d) 2 equiv of NH2OH‚HCl and DIPEA in ethanol; (e) 2 equiv of (Ac)2O in acetic acid; (f) zinc powder in
acetic acid, 3 h, room temp; (g) 90% TFA, 1 h, room temp; (h) 1.5 equiv of 1H-pyrazole-1-carboxamidine‚HCl/DIPEA in DMF; (i) H2 and Pd/C in 90%
acetic acid, preparative reversed-phase HPLC.
and 63 the appropriate Boc-amidopropylpiperidide derivative
was used. The nitrile moiety was converted into an amidine by
the procedures described in Scheme 1 for all compounds with
identical N- and C-terminal amino or guanidino groups (e.g.,
59 and 58). In contrast, the acetylhydroxyamidino group was
converted to the amidine using zinc powder in acetic acid (3 h,
room temperature) for inhibitors with a single C-terminal
guanidino group (12 and 63). This strategy retained the
N-terminal Cbz protection and allowed the selective guanylation
of the C-terminus using pyrazole-1-carboxamidine24 as shown
in Scheme 2 for the synthesis of inhibitor 12.
Enzyme Kinetics and X-ray Crystallography. To elucidate
the binding geometry of compound 2, we superimposed the
known X-ray structure of uPA with compound 13 (PDB code
1F92),25 which is closely related to 2, using the matriptase
structure as deposited under the PDB entry 1EAX.11 On the
basis of this model, we hypothesized that the C-terminal
guanidino group of inhibitor 2 might interact with the superficial,
solvent-exposed side chain of matriptase residue Asp96 (chy-
motrypsinogen numbering11), which is harbored in the 99 loop
“balcony” delimiting the active-site cleft to the north and extends
toward the S2 subsite. To probe this interaction, further
analogues of 2 with variable lengths of the guanidinoalkyl chain
and different hydrophobic sulfonyl residues were prepared. Their
affinity was measured toward matriptase and the closely related
clotting proteases thrombin and factor Xa as well as toward
the fibrinolytic enzymes plasmin and uPA.
Most of the inhibitors summarized in Table 1 show a
preference for matriptase over the other enzymes. In analogy
to known uPA or thrombin inhibitors, the affinity of racemic 2
could be improved by using 3-amidinophenylalanine in the
S-configuration (2S). Inhibitors with a C-terminal guanidino
group have been found to be 2-3 times more potent than the
analogous amino derivatives in most cases. There is only a
marginal difference in matriptase affinity between compounds
2S and 14, containing a guanidinopropionyl or butyryl group,
respectively. A slightly enhanced potency toward matriptase was
observed after incorporation of a cyclic N-(amidino)isonipecotyl
residue (31). In contrast, the exchange of the piperazine ring
by homopiperazine has a negligible effect on the inhibition of
matriptase (e.g., inhibitor 13 vs 19, and 2S vs 18).
The elimination of the basic group in the C-terminal part of
the molecule (35, 36) reduces the potency toward matriptase.
However, the previously described nonselective uPA inhibitor
35 (WX-UK1), containing a neutral ethyloxycarbonyl group,
still inhibits matriptase with an inhibition constant of 0.32 µM,
which is comparable to the inhibition value toward uPA.26 Some
of the described effects of 35 in suppression of rat breast cancer
metastasis and tumor growth reduction could also possibly be
related to matriptase inhibition.
In previous work with closely related thrombin inhibitors
containing a central 3-amidinophenylalanylpiperazide, it was
found that the piperazine moiety could be replaced by other
secondary amines, among them isonipecotic acid.27,28 Therefore,
isonipecotic acid analogues were synthesized, which allow for
a simple C-terminal elongation with basic residues derived from
diaminoethane or piperazine (Table 2).
The incorporation of isonipecotic acid resulted in selective
and potent matriptase inhibitors. Also in this series, compounds
containing a C-terminal guanidino moiety were more potent than
the corresponding amino analogues. In contrast to the piperazine
derivatives (Table 1), the 2,4,6-triisopropylphenylsulfonyl de-
rivatives are less effective than the anthraquinonsulfonyl- or
4-cyclohexylphenylsulfonyl analogues. Only small differences in
matriptase affinity were found after replacement of the C-ter-
minal alkylguanidino group by a cyclic carbamidoylpiperazine.
In an additional series, more simple bis-basic secondary amines
were incorporated to eliminate one of the peptide bonds in these
molecules. The kinetic results are summarized in Table 3.
Several inhibitors (8, 50, 52) inhibit matriptase with Ki values
of approximately 50 nM and maintain some selectivity toward
the other enzymes. To obtain information of the binding mode,
two inhibitors (8, 31) were selected for X-ray analysis. Figure
1 shows the complex between matriptase and these inhibitors.
Some characteristic interactions are very similar in both
structures. The inhibitors adopt a compact Y-shaped conforma-

NoVel Matriptase Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4119
Table 1. Inhibition Constants for Inhibitors of the General Compound Showna
a The asterisk (/) indicates that the compound is racemic at the CR-atom of 3-amidinophenylalanine.
tion similar to that previously found for related arylsulfonyl-
3-amidinophenylalanine-derived inhibitors in complex with uPA,
thrombin, or trypsin.25,29,30 The central benzamidine moiety is
sandwiched, like the free benzamidine,11 between matriptase
segments Ser190-Gln192 and Trp215-Gly216. The two amidino
nitrogens juxtapose the two carboxylate oxygens of Asp189
situated at the bottom of the S1 specificity pocket under
formation of favorable twin ionic bonds and form additional
hydrogen bonds to the carbonyl of Gly219, the side chain
hydroxyl of Ser190, and a conserved water molecule.
The 3-amidinophenylalanyl moiety is positioned, similar to
that found in the complexes of inhibitor 1 in trypsin31 and
inhibitor 13 in uPA,25 such that its NH and carbonyl groups
form hydrogen bonds to the carbonyl and the NH group of
Gly216, respectively. In addition, one of the two sulfonamide
oxygens is in hydrogen bond distance to Gly219 NH. In both
structures, the sulfonylated aryl ring and the proximal piperidinyl
or piperazine group are folded together into a compact entity,
making intense contacts with the hydrophobic cleft formed by
Trp215, the His57 imidazole ring, and the Phe99 phenyl group.
Presumably, this conformation is close to the preferred confor-
mation in solution, resulting in improved binding due to
entropically favorable conditions.29
The piperidine group of compound 8, on the C-terminal side,
can adopt both a boot and a chair conformation, with the
equatorial guanidinoethyl group extending into the negatively
charged “cation cleft” formed by the initial part of the
characteristically exposed 60 insertion loop and the active-site
cleft of matriptase and lined by the Asp96 carboxylate, the
carbonyl groups of His57 and Ile60, and the carboxylate group

4120 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Steinmetzer et al.
Table 2. Inhibition Constants for Isonipecotic Acid Containing
Table 3. Inhibition Constants for Inhibitors of the General Compound
Inhibitors of the General Compound Shown
Shown
Orthotopic Xenograft Mouse Model of Prostate Cancer.
A nonselective macromolecular inhibitor of matriptase, the
serine protease inhibitor ecotin, has previously been reported
to inhibit the growth and dissemination of tumors derived from
the matriptase-expressing human prostate cancer cell line PC-3
in immune-compromised mice.10 The anti-invasive potential of
the matriptase inhibitors 8 and 44 on PC-3 cells was previously
demonstrated in an in vitro scatter assay.32 To detect the anti-
invasive activity of the matriptase inhibitors in vivo as well,
we investigated the effects of the initial lead compound 8 and
the more potent and more selective derivative 59 on tumor
growth and metastasis formation in nude mice bearing ortho-
topically implanted PC-3 prostate carcinoma. The inhibitors
were administered intraperitoneally at a daily dose of 5 mg/kg
for 4 weeks to nude mice with established orthotopic prostate
tumors that were transplanted from the invasive part of human
PC-3 tumor tissue pregrown subcutaneously in nude mice. The
inhibitors 8 and 59 were both well tolerated by the PC-3 tumor-
bearing mice and reduced the median primary tumor growth
by 17% and 40%, respectively, compared to a vehicle-treated
control group (Figure 2A).
of Asp60b. With some of these hydrogen bond acceptors, the
guanidyl group makes water-mediated hydrogen bonds. The
N-terminal naphthyl moiety of inhibitor 8 is, in contrast to the
equivalent naphthyl group of NAPAMP-thrombin,29 curiously
abducted from the active-site cleft, making only weak van der
Waals contacts with Trp215. This might be due to a concerted
movement done together with the C-terminal moiety but might
also be caused by the Arg60c side chain of a symmetry-related
matriptase molecule extending into the S4 subsite. Both the
proximal piperazine and the distal piperidine group, located on
the C-terminal side of compound 31, are confined to the chair
conformation. The distal amidino group extends into the
aforementioned “cation cleft”, making hydrogen bonds with
Ile60 O (strong), Asp60a, and Cys58 O (weak). The triisopropyl-
phenylsulfonyl group of inhibitor 31, on the N-terminal side,
is, as in the urokinase complex with the closely related inhibitor
13,25 abducted from the active-site cleft, making only a weak
contact with Trp215 via one isopropyl group. Again, this loose
contact might in part be due to the extending Arg60c side chain
of a neighboring matriptase molecule.
Therefore, in solution, the N-terminal sulfonyl groups might
further approach the S2/S4 subsite and extend into the hydro-
phobic surface cleft formed by the side chains of Phe97, Phe99,
Gln175, and Trp215 and by the carbonyl groups of Phe97 and
Thr98 of matriptase. Molecular modeling revealed that it might
be possible to use this binding site of matriptase by incorporation
of appropriate sulfonyl substituents. Kinetic results for the first
derivatives of this type are summarized in Table 4, showing
that several inhibitors have improved affinity and selectivity
toward matriptase.
In addition, the inhibitors further reduced substantially
metastasis formation. Whereas 30% of the control mice had
metastases in the thorax (heart, lung, pleura), metastases were
only found in 14% and 20% of the thoraxes of the animals
treated with compounds 8 and 59, respectively. The effect was
even more pronounced with abdominal metastases. Whereas
metastases were found in 50% of the abdomens of the control
mice, abdominal metastasis formation was completely abolished
in both of the treatment groups (Figure 2B). These results

NoVel Matriptase Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4121
Figure 1. Stereoview of inhibitors 8 (A) and 31 (B) bound to the active site of matriptase. Selected matriptase residues are labeled, and strong
hydrogen bonds are shown in green. Only water molecules that are involved in strong hydrogen bonding are shown (red balls).
indicate that matriptase might be a relevant target in the
progression of prostate cancer.
(compounds shown in Tables 3 and 4) was well tolerated.
Although the elimination of a peptide bond should increase their
conformational flexibility, which often leads to less potent
derivatives, this was not the case for these matriptase inhibitors.
Originally, we had assumed that the C-terminal basic group
interacts with the side chain of the matriptase residue Asp96.
However, in both X-ray crystal structures the C-terminal distal
guanidyl group is directed into the aforementioned “cation cleft”
between the initial 60 insertion loop and the matriptase surface
rather than to the relatively flexible Asp96 side chain carboxy-
late, which is placed at the outer edge of this hole. The outward
location of the N-terminal arylsulfonyl groups, in contrast, might
be an artifact of the distinct matriptase crystal packing. We
speculate that in solution the arylsulfonyl groups might be placed
Discussion
Several series of matriptase inhibitors based on secondary
amides of 3-amidinophenylalanine have been developed. Their
structures are closely related to previously described thrombin
or urokinase inhibitors of the 3-amidinophenylalanine type.20,28
In contrast to these compounds, however, the matriptase
inhibitors presented in this study contain an additional basic
group in the C-terminal part of the molecule, which is important
for selectivity and contributes to potency. The replacement of
the basic acylated piperazide segment, used for all compounds
summarized in Table 1, by simple basic alkylpiperidines

4122 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Steinmetzer et al.
Table 4. Inhibition Constants for Inhibitors of the General Compound
Shown
Figure 2. Effect of the matriptase inhibitors 8 and 59 on median
primary tumor growth (A) and the metastatic potential (B) of PC-3-
derived tumors in an orthotopic mouse model of pancreatic cancer.
The inhibitors were administered intraperitoneally at a daily dose of 5
mg/kg for 4 weeks into nude mice with orthotopically transplanted PC-
3-derived tumors. Abdominal metastases (black) were only found in
the vehicle-treated control group but not in the treatment groups. The
dissemination into the thorax (gray) was reduced by 53% and 33% for
inhibitors 8 and 59, respectively. Three mice treated with inhibitor 8
died because of postoperative suffering during the first treatment days.
determined for the less selective analogues 8 and 54, probably
because of a moderate thrombin inhibition. The estimated IC200
values in the TT, aPTT, and PT assays are 0.32, 2.7, and 3.8
µM for inhibitor 8 and 0.27, 1.5, and 1.6 µM for analogue 54,
respectively.
The potency of matriptase inhibitors 8 and 44 has previously
been demonstrated in a series of in vitro assays.32 Both inhibitors
are able to reduce the matriptase-catalyzed processing of pro-
HGF into HGF with IC50 values between 1 and 3 µM and inhibit
the pro-HGF/SF-induced scattering of matriptase-expressing
DLD-1 and PC-3 cells at micromolar concentrations, as well
as their subsequent invasion in a matrigel assay.
Matriptase has been isolated as a major protease of the human
prostate carcinoma cell line PC-3.10 Two additional recent
studies showed that matriptase expression is elevated in human
prostate carcinoma compared to benign prostate tissue.33,34
Because the protein level of matriptase increases with increasing
tumor grade, matriptase was suggested to be a novel predictive
biomarker for human prostate adenocarcinoma. The increase
of matriptase expression comes along with a progressive loss
of the expression of its cognate inhibitor, hepatocyte growth
factor activator inhibitor type 1 (HAI-1), implicating that
matriptase may be functionally involved in the progression of
prostate cancer and giving a rationale for inhibiting matriptase
in anticancer therapy. Therefore, we have studied the effects of
two of our inhibitors on primary tumor growth and metastasis
closer to the S4 subsite and nestle into the hydrophobic S2/S4
cleft of matriptase.
Therefore, inhibitors were synthesized that contain an aryl
ring bridged via an oxygen or methylene to the para position
of an arylsulfonyl group or that possesed an additional acyl
residue attached to the aryl moiety via a peptide bond. Several
of these inhibitors have enhanced matriptase affinity and
improved selectivity (see Table 4). Molecular modeling in
complex with matriptase revealed that the terminal phenoxy ring
of inhibitor 54 and the â-Ala moiety of compound 59 may
occupy the S4-binding site of matriptase (Figure 3).
The binding mode of inhibitor 59 may explain its improved
affinity and selectivity as matriptase inhibitor. As suggested
previously, the S4 subsite of matriptase seems to be well suited
to accommodate residues containing positively charged groups.5,11
To detect potential side effects for in vivo testing, selected
matriptase inhibitors 8, 54, 59, and 63 were evaluated for their
influence on blood clotting. The very specific matriptase
inhibitors 59 and 63 did not show any significant activity (IC200
> 10 µM) in the standard coagulation assays, such as activated
partial thromboplastin time (aPTT), prothrombin time (PT), and
thrombin time (TT). In contrast, some anticoagulant activity was

NoVel Matriptase Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4123
Figure 3. Models of inhibitors 54 (gray carbon atoms) and 59 (orange carbon atoms) in complex with matriptase, visualized as a surface indicating
the atom charges. The compounds were manually modeled into the active site cleft of matriptase under consideration of the known interactions
between the sulfonylated peptidic backbone of the inhibitor and the Trp215-Gly219 segment of matriptase as well as the most probable interactions
of these compounds with the S2/S4 subsite and the S1 specificity pocket, respectively. The models were generated from the X-ray structure of
inhibitor 8 in matriptase by replacement of the sulfonyl substituent in both analogues and elimination of the C-terminal carbamidine in the case of
compound 59.
2-Naphthylsulfonyl-3-acetyloxamidinophenylalanine (2-Nas-
Phe(3-AcOxam)-OH) (3). An amount of 3.8 g (10 mmol) of 2-Nas-
Phe(3-CN)-OH, obtained by coupling of 2-Nas-Cl to 3-cyanophe-
nylalanine under Schotten-Baumann conditions,28 was dissolved
in 100 mL of absolute ethanol and treated with 1.39 g (20 mmol)
of hydroxylamine‚HCl and 3.48 mL (20 mmol) of DIPEA (diiso-
propylethylamine). The mixture was refluxed for 6 h and stirred
overnight at room temperature. The solvent was evaporated. The
remaining oil was dissolved in 50 mL of acetic acid and 2.83 mL
(30 mmol) of acetic anhydride. The mixture was stirred at room
temperature for 30 min. The solvent was evaporated, and the
remaining oil was dissolved in ethyl acetate and washed 3× with
brine. The organic layer was dried over Na2SO4 and evaporated.
The product was crystallized from ethyl acetate/petroleum ether
(bp 40-60 °C) to give a white solid (yield, 3.02 g ) 6.64 mmol;
TLC, Rf ) 0.67; HPLC, retention time 34.04 min; MS 456 (M +
H)+; 1H NMR (DMSO-d6) δ 2.13 (s, 3H), 2.70-3.05 (m, 2H), 4.01
(m, 1H), 6.68 (s, br. 2H), 7.12 (m, 1H), 7.20 (d, 1H), 7.42 (d, 1H),
7.52 (s, 1H), 7.54-7.69 (m, 3H), 7.96 (m, 2H), 8.04 (d, 1H), 8.24
(s, 1H), 8.34 (d, 1H), 12.67 (s, br, 1H)).
4-(2-Aminoethyl)pyridine (4). A slurry of 76.23 g (1.42 mol)
of ammonium chloride and 74.71 g (0.71 mol) of 4-vinylpyridine
in 225 mL of water and 100 mL of methanol was refluxed
overnight. The solution was cooled. At 0 °C, an amount of 59.02
g (1.47 mol) of NaOH was added in small portions. The aqueous
layer was extracted 3× with chloroform, and the organic layer was
dried over Na2SO4 and evaporated. The remaining oil was purified
by distillation (22 mbar, 121 °C, oil; yield, 46.9 g ) 0.38 mol;
TLC, Rf ) 0.11; MS 123 [M + H]+; 1H NMR (CDCl3) δ 0.94 (s,
br, 2H), 2.53 (t, 2H), 2.78 (t, 2H), 6.90 (d, 2H), 8.30 (d, 2H)).
4-(2-Boc-amidoethyl)piperidine (5). An amount of 29.5 g (135
mmol) of di-tert-butyl dicarbonate, dissolved in 50 mL of tert-
butanol, was added within 30 min to 15 g (123 mmol) of 4-(2-
aminoethyl)pyridine (4) in 250 mL of tert-butanol. The mixture
was stirred at room temperature for 12 h and washed twice with
saturated NaHCO3 solution and once with brine. The organic layer
was dried over Na2SO4 and evaporated (oil, 23.9 g, 107 mmol, Rf
)0.43). The remaining oil was dissolved in 800 mL of 90% acetic
acid and treated with 250 mg of catalyst (PtO2). The mixture was
stirred under an atmosphere of hydrogen at 35 °C for 96 h. The
catalyst was removed by filtration, and the solvent was evaporated.
The residue was dissolved in water adjusted to pH 11 with 1 N
NaOH. The aqueous layer was extracted 3× with DCM. The
formation in an orthotopic xenograft model of prostate cancer
using PC-3-derived tumor tissue. The compounds were well
tolerated and inhibited both the primary growth and the
metastatic capacity of these tumors. These results confirm a
previously published study with an arginal-derived matriptase
inhibitor, which inhibited the growth of tumors established from
human androgen independent prostate cell lines CWR22R and
CWRSA6.18 Also in this study, no side effects due to matriptase
inhibition have been described, although experiments with
knockout mice have demonstrated that matriptase is required
for postnatal survival, and one might argue that the inhibition
of matriptase could lead to side effects in vivo. Matriptase
exhibits its pleiotropic functions in epidermal differentiation due
to profilaggrin-processing, hair follicle development, and thymic
homeostasis.35,36 Obviously, other matriptase substrates are
involved in tumor progression and invasion, such as pro-HGF
and pro-uPA.
Our data provide an additional proof of concept for the
targeting of matriptase in the design of new anti-invasive drugs.
Experimental Section
Synthesis. All reagents, solvents for synthesis, and amino acids
were purchased from Aldrich, Lancaster, Acros, Fluka, or Bachem
and were used without further purification. 3-Cyanophenylalanine
was obtained from Senn Chemicals, Switzerland. The molecular
mass of the inhibitors was determined using a Finnigan ESI-MS
LCQ (Bremen, Germany). Analytical HPLC experiments were
performed on a Shimadzu LC-10A system (Phenomenex Luna C18,
5 µm column, 4.6 mm × 250 mm) with a linear gradient of
acetonitrile (10-70% in 60 min, detection at 220 nm) containing
0.1% TFA at a flow rate of 1 mL/min. The final inhibitors were
purified to more than 97% purity by preparative HPLC on a
Shimadzu LC-8A system (Phenomenex Luna C18, 5 µm column,
30 mm × 250 mm) with a linear gradient of acetonitrile (45%
increase of acetonitrile within 120 min, detection at 220 nm)
containing 0.1% TFA at a flow rate of 10 mL/min and obtained as
TFA salts after lyophilization. Thin-layer chromatography was
performed on silica gel plates (silica gel 60 F254, Merck, Darmstadt,
1
Germany) in n-butanol/acetic acid/water 4/1/1 (v/v/v). H NMR
spectra were recorded on a Bruker DPX-300 spectrometer at 300.13
MHz, and chemical shifts are referenced to the residual proton signal
of the solvent.

4124 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Steinmetzer et al.
organic layer was dried over Na2SO4, and the solvent was
evaporated. The product crystallized from the remaining oil at 4
°C (white solid; yield, 21.9 g ) 96 mmol; TLC, Rf ) 0.38; MS
1.75 (m, 2H), 2.23 (m, 0.5H), 2.44 (m, 0.5H), 2.66 (m, 0.5H), 2.91
(m, 2.5H), 3.07 (m, 2H), 3.62 (m, 1H), 4.27 (m, 1H), 4.49 (m,
1H), 4.55 (m, 0.5H), 6.47 (m, 1H), 7.37 (m, 3H), 7.46 (m, 1H),
7.66 (m, 1H), 8.04 (m, 1H), 8.35 (m, 1H), 8.47 (m, 1H), 8.66 (broad,
1H); one resonance hidden by the large tBu signal at 1.42 ppm).
N-(3-Cbz-â-Ala)amidophenylsulfonyl-Phe(3-CN)-4-(2-Boc-
amido)ethylpiperidide (10). An amount of 5 g of the crude 9 was
dissolved in 100 mL of acetic acid at room temperature and treated
with zinc powder. The mixture was stirred for 1.5 h, the remaining
zinc was removed by filtration, and the solvent was evaporated.
The residue was dissolved in ethyl acetate and washed 4× with
saturated NaHCO3 solution and 2× with brine. The organic layer
was dried over Na2SO4 and evaporated (yellow foam; yield, 4.5 g;
TLC, Rf ) 0.86; HPLC, retention time 40.29 min; MS [M - H]-
554.2). An amount of 2.74 g of the crude 3-aminophenylsulfonyl-
Phe(3-CN)-4(2-Boc-amido)ethylpiperidide was added to a mixed
anhydride, which was prepared from 1.0 g (4.48 mmol) of Cbz-
â-Ala-OH, 583 µL (4.48 mmol) of isobutyl chloroformate, and 493
µL (4.48 mmol) of N-methylmorpholine in DMF at -15 °C within
10 min. The mixture was stirred for 1 h at -15 °C and at room
temperature overnight. The solvent was evaporated, and the residue
was dissolved in ethyl acetate, washed 3× with 5% KHSO4 solution,
1× with brine, 3× with saturated NaHCO3 solution, and 3× with
brine. The organic layer was dried over Na2SO4 and evaporated
(brown oil). The product was obtained as a TFA salt after
preparative reversed-phase HPLC (white lyophilized powder; yield,
2.3 g ) 3.02 mmol; TLC, Rf ) 0.96; HPLC, retention time 51.88
1
229.2 [M + H]+; H NMR (DMSO-d6): δ 1.31 (m, 4H), 1.36 (s,
9H), 1.50 (m, 1H), 1.75 (d, 2H), 2.74 (m, 2H), 2.93 (m, 2H), 3.17
(m, 2H) 6.75 (br, 1H)).
2-Nas-Phe(3-AcOxam)-4-(2-Boc-amido)ethylpiperidide (6).
An amount of 500 mg (1.1 mmol) of 2-Nas-Phe(3-AcOxam)-OH
(3) and 251 mg (1.1 mmol) of compound 5 were dissolved in 10
mL of DMF. At 0 °C 571 mg (1.1 mmol) of PyBop and 383 µL
(2.2 mmol) of DIPEA were added. The mixture was stirred for 20
min at 0 °C and for an additional period of 3 h at room temperature.
The solvent was evaporated, and the remaining oil was dissolved
in ethyl acetate, washed 3× with 5% KHSO4 solution, 1× with
brine, 3× with saturated NaHCO3 solution, and 2× with brine. The
organic layer was dried over Na2SO4, and the solvent was
evaporated. The remaining brown foam was purified by preparative
reversed-phase HPLC, and the product was dried in vacuo (white
solid; yield, 583 mg ) 0.88 mmol; TLC, Rf ) 0.85; HPLC, retention
1
time 48.07 min; MS 664.3 [M - H]-; H NMR (CDCl3-d), two
conformers in a relative ratio of 55:45, δ -0.15 (m, 1H), 0.54 (m
0.55H), 0.64 (m, 1H), 0.98-1.25 (m, 4H), 1.37 (s, 4.05H), 1.39 (s,
4.95H), 1.50 (m, 0.45H), 2.15 (s, 3H), 2.20 (m, 2H), 2.70-2.98
(m, 4H), 3.26 (m, 1H), 3.90-4.06 (m, 1H), 4.26-4.49 (m, 1H),
5.58 (br, 2H), 6.50 (m, 1H), 7.18 (m, 2H), 7.48-7.60 (m, 4H),
7.70 (m, 1H), 7.83 (m, 3H), 8.29 (s, 1H)).
2-Nas-Phe(3-Am)-4-(2-aminoethyl)piperidide‚2TFA (7). An
amount of 5 mL of 1 N HCl in glacial acetic acid was added to
500 mg (0.75 mmol) of compound 6 and was dissolved in 5 mL of
acetic acid. The mixture was shaken at room temperature for 1 h.
The solvent was evaporated, and the remaining product (TLC: Rf
) 0.35; HPLC, retention time 25.66 min; MS 566.3 [M + H]+)
was dissolved in 50 mL of 90% acetic acid and treated with 15 mg
of catalyst (10% Pd/C). The mixture was stirred under an
atmosphere of hydrogen at 40 °C overnight. The catalyst was
removed by filtration, and the solvent was evaporated. The crude
product was purified by reversed-phase HPLC and lyophilized from
water (white powder; yield, 70.4 mg ) 0.096 mmol; TLC, Rf )
1
min; MS 759.2 [M - H]-; H NMR, (CDCl3-d), two conformers
in a relative ratio of 0.62:0.38, multiplicity not assigned because
of broad resonances at room temperature, δ 0.21 (1H), 0.58 (0.62H),
0.73 (0.38H), 1.21 (2H), 1.40 (3.42H), 1.41 (5.58H), 1.50 (1H),
2.17 (0.38H), 2.29 (0.62H), 2.44 (0.62H), 2.57 (2H), 2.63 (0.38H),
2.85 (2H), 2.98 (2H), 3.35 (1.38H), 3.49 (2.62H), 4.24 (1.62H),
4.41 (0.38H), 4.58 (0.38H), 4.75 (0.62H), 5.05 (2H), 5.73 (1H),
6.67 (1H), 7.23-7.45 (11.76H), 7.68 (0.62H), 7.84 (1H), 8.09
(0.62H), 8.75 (0.38H), 9.10 (0.62H)).
N-(3-Cbz-â-Ala)amidophenylsulfonyl-Phe(3-Am)-4-(2-Boc-
amido)ethylpiperidide‚TFA (11). The intermediate N-(3-Cbz-â-
Ala)amidophenylsulfonyl-Phe(3-AcOxam)-4(2-Boc-amido)ethylpi-
peridide, prepared from 242 mg (0.318 mmol) of the nitrile 10 as
described for compound 3, was reduced with zinc powder in acetic
acid over a period of 3 h at room temperature. The remaining zinc
was removed by filtration, the solvent was evaporated, and the
product was obtained by preparative HPLC (white lyophilized
powder; yield, 121 mg ) 0.135 mmol; TLC, Rf ) 0.76; HPLC,
retention time 36.41 min; MS 778.5 [M + H]+; 1H NMR, (CDCl3-
d), two conformers in a relative ratio of 1:1, multiplicity not
assigned because of very broad resonances at room temperatur,
dilute samples also gave very broad signals, δ -0.31 (0.5H), 0.39
(0.5H), 0.62 (1H), 1.21 (2H), 1.41 (9H), 2.15 (2H), 2.33 (0.5H),
2.61 (3H), 2.88 (4H), 3.49 (3H), 4.04 (0.5H), 4.22 (1H), 4.31 (0.5H),
4.52 (0.5H), 5.02 (2H), 5.87 (0.5H), 6.01 (0.5H), 7.25 (10H), 7.65
(3H), 8.03 (1H), 8.23 (1H), 8.53 (1H), 9.38 (0.5H), 9.54 (0.5H),
10.00 (1H), 10.07 (1H)).
1
0.34; HPLC, retention time 19.03 min; MS 508.4 [M + H]+; H
NMR (DMSO-d6), two conformers in a relative ratio of 40:60, δ
-0.02 (m, 1H), 0.43 (m, 0.4H), 0.58 (m, 0.6H), 1.09-1.45 (m,
5H), 1.69 (m, 0.4H), 2.14 (m, 0.6H), 2.59-2.83 (m, 4H), 2.90 (m,
1H), 3.70 (s, br, 2H), 3.70-3.94 (m, 2H), 4.53 (m, 1H), 7.37-
7.50 (m, 2H), 7.58 (d, 1H), 7.62-7.75 (m, 6H), 7.97-8.10 (m,
3H), 8.17 (d, 0.6H), 8.27 (d, 0.4H), 8.30 (s, 1H), 9.25 (s, br, 3H)).
2-Nas-Phe(3-Am)-4-(2-guanidinoethyl)piperidide‚2TFA (8).
An amount of 50 mg (0.068 mmol) of compound 7 was dissolved
in 2 mL of DMF and treated with 20 mg (0.136 mmol) of 1H-
pyrazole-1-carboxamidine‚HCl and 35.5 µL (0.204 mmol) of
DIPEA. The mixture was stirred for 24 h, the solvent was
evaporated, and the remaining crude product was purified by
reversed-phase HPLC and lyophilized from water (white powder;
yield, 42 mg ) 0.054 mmol; TLC, Rf ) 0.40; HPLC, retention
1
time 21.19 min; MS 550.4 [M + H]+; H NMR (DMSO-d6), two
conformers in a relative ratio of 40:60, δ -0.10 (m, 0.6H), 0.08
(m, 0.4H), 0.45 (m, 1H), 0.95 (m, 1H), 1.20 (m, 3H), 1.35 (m,
1H), 1.70 (m, 0.4H), 2.12 (m, 0.6H), 2.45 (m, 0.6H), 2.66 (m, 0.4H),
2.78 (m, 1H), 2.94 (m, 3H), 3.79 (m, 2H), 4.54 (m, 1H), 5.60 (s,
br, 1H), 7.23 (s, br, 3H), 7.44 (m, 2H), 7.63 (m, 5H), 7.75 (s, 1H),
8.02 (m, 3H), 8.16 (d, 0.6H), 8.24 (d, 0.4H), 8.30 (d, 1H), 9.27 (s,
br, 2H), 9.41 (s, br, 2H)).
N-(3-â-Ala)amidophenylsulfonyl-Phe(3-Am)-4-(2-guanidino)-
ethylpiperidide‚3TFA (12). An amount of 120 mg of compound
11 was dissolved in 3 mL of 90% TFA and shaken for 1 h at room
temperature. The solvent was evaporated and lyophilized from water
(124 mg; TLC, Rf ) 0.27; HPLC, retention time 21.76 min; MS
678.3 [M + H]+). An amount of 120 mg (0.132 mmol) of the TFA
salt was dissolved in 2 mL of DMF and treated with 38.7 mg (0.264
mmol) of 1H-pyrazole-1-carboxamidine‚HCl and 69 µL (0.396
mmol) of DIPEA. The mixture was stirred for 24 h. The solvent
was evaporated, and the remaining oil (TLC, Rf ) 0.36; HPLC,
retention time 23.04 min; MS 720.3 [M + H]+) was dissolved in
50 mL of 90% acetic acid and treated with 10 mg of catalyst (10%
Pd/C). The mixture was hydrogenated at room temperature over-
night. The catalyst was removed by filtration, the solvent was
evaporated, and the product was purified by reversed-phase HPLC
and lyophilized from water (white powder; yield, 48 mg ) 0.052
mmol; TLC, Rf ) 0.04; HPLC, retention time 10.38 min; MS 586.3
3-Nitrophenylsulfonyl-Phe(3-CN)-4-(2-Boc-amido)ethylpiperi-
dide (9). The synthesis was performed by PyBop mediated coupling
as described for compound 6 using 3.9 g (10.39 mmol) of
3-nitrophenylsulfonyl-Phe(3-CN)-OH and 2.37 g (10.39 mmol) of
5 as starting materials (orange oil; yield, 5.3 g; purity 81% based
on analytical HPLC at 220 nm). A sample was purified by
preparative reversed-phase HPLC for analysis (TLC, Rf ) 0.96;
1
HPLC, retention time 50.34 min; MS 584.3 [M - H]-; H NMR,
(CDCl3-d), two conformers in a relative ratio of 1:1, δ 0.32 (m,
0.5H), 0.68 (m, 0.5H), 0.82 (m, 1H), 1.30 (m, 2H), 1.42 (s, 9H),

NoVel Matriptase Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4125
1
[M + H]+; H NMR (D2O-d2), two conformers in a relative ratio
Table 5. Refinement Statistics for the Matriptase Complexes with
Inhibitors 8 and 31 Determined by X-ray Crystallography
of 0,54:0,46, δ 0.07 (m, 0.46H), 0.69 (m, 1.54H), 1.35 (m, 1H),
1.45 (m, 1H), 1.50 (m, 1H), 1.62 (m, 2H), 2.24 (m, 0.46H), 2.54
(m, 0.54H), 2.77 (m, 1H), 2.95 (m, 2.46H), 3.08 (m, 1.54H), 3.15
(m, 1H), 3.21 (m, 1H), 3.41 (m, 2H), 3.73 (m, 1H), 4.06 (m, 1H),
4.56 (m, 1H), 7.57 (m, 5H), 7.71 (m, 2H), 7.96 (m, 1H)).
Cloning of the Serine Protease Domain of Matriptase. The
serine protease domain was amplified from I.M.A.G.E. clone
IMAGp998N139560 (RZPD, Berlin, Germany) by PCR using
5′-GGCAATTCCATATGAAACATCACCATCATCACCATGT-
TGTTGGGGGCACGGATGCG-3′ and 5′-GCATGAATTCTTA-
TACCCCAGTGTTCTCTTTGATCCA-3′ as sense and antisense
primers, respectively. The primer were chosen to introduce an Nde1
site at the 5′ end of the protease domain followed by a MK(His)6
sequence and an EcoR1 site at the 3′ end of matriptase. The ∼750
base pair long amplification product was purified, digested with
Nde1 and EcoR1, and subcloned into pET24 (Novagen, Merck
Biosciences, Bad Soden, Germany) for expression in Escherichia
coli. The catalytic domain of matriptase was expressed in inclusion
bodies and then subsequently denatured, purified, refolded, and
activated as described below.
Expression, Purification, Refolding, and Activation of the
Catalytic Domain of Matriptase. BL21(DE3) cells (Novagen)
containing the expression vector were incubated in LB and 30 µg/
mL kanamycin at 37 °C and 250 rpm. The expression was induced
by the addition of 1 mM IPTG at a cell density of OD600 ) 0.8,
and the incubation continued for 1 h. The cells were harvested and
lysed with BugBuster protein extraction reagent (Novagen), and
the DNA was digested with Benzonase (25 U/g cell pellet,
Novagen). The inclusion bodies were washed and denatured with
denaturation buffer (6 M guanidine HCl, 10 mM Tris-HCl, 100
mM sodium phosphate, pH 8.0, 5 mL per 1 mL of pellet). The
denatured protein was freed from insoluble constituents by cen-
trifugation and filtration (0.2 µm), and the his-tagged matriptase
was purified by metal chelate chromatography (NiNTA Agarose,
Qiagen, Hilden, Germany). Matriptase-containing fractions were
pooled, derivatized with glutathione, and then renatured by dilution
to a final protein concentration of 50 µg/mL in refolding buffer
(50 mM Tris-HCl, 0.5 M L-arginine, 20 mM CaCl2, 1 mM EDTA,
0,1 M NaCl, 1 mM cysteine, pH 7,5). After 3 days of incubation
at room temperature, the refolding solution was filtered and
concentrated to >300 µg/mL by ultrafiltration (Centricon Plus-80,
PL-10, Millipore, Schwabach, Germany) and the buffer was
exchanged to activation buffer (20 mM sodium phosphate, 150 mM
NaCl, pH 7.0).
Because correct processing of the N-terminus is required for the
activity of serine proteases, the refolded matriptase had to be
activated by removal of the N-terminal peptide sequence MK(His)6.
This was achieved by incubating the protease for 2 h with 2.5 mU/
mL of activated DAPase (Qiagen) and by subsequent separation
of activated from nonactivated matriptase and his-tagged DAPase
via metal chelate chromatography. The yield of this procedure was
approximately 2 mg of active catalytic domain per 1 L of culture.
Enzyme Kinetics. The kinetic measurements with thrombin,
factor Xa, uPA, and plasmin were performed as described previ-
ously.28 The matriptase activity was determined at 405 nM using
the chromogenic substrate Pefachrome tPA (Loxo GmbH, Dos-
senheim, Germany) in 0.05 M Tris-HCl buffer, pH 8.0, containing
0.154 M NaCl, 0.1% BSA, and 5% (v/v) ethanol.
Animal Study. An orthotopic xenograft mouse model of prostate
cancer was performed by Oncotest GmbH (Freiburg, Germany)
according to the guidelines of the German Animal Health and
Welfare Act. Small tumor fragments of 0.5-1 mm in diameter that
had been established by subcutaneous inoculation of the human
prostate tumor cell line PC3 in nude mice were orthotopically
transplanted into 5-week-old to 6-week-old male anesthetized NMRI
nu/nu mice. After an induction period of 4 days that also allowed
healing of the surgery wound, treatment of the established tumors
was started. Animals were randomized into groups of 7-10 tumor-
bearing mice and intraperitoneally treated daily, 7 days a week,
with inhibitor 8 or 59 (5 (mg/kg)/day) or vehicle (0.9% saline),
parameter
8
31
resolution range (Å)
space group
molecules per asymmetric unit
a (Å)
14.6-2.1
C222
14.6-2.2
C222
1
1
66.84
141.02
51.80
90, 90, 90
59, 899
99.3
67.07
140.15
51.73
90, 90, 90
85, 118
98.0
b (Å)
c (Å)
R, â, γ (deg)
no. of reflections
completeness (%)
multiplicity
4.1
10.6
3.7
R
merge (%)
15.8
no. of unique reflections
14,540
18.19
22.93 (4.9)
0.008
12,728
19.69
23.87 (5.1)
0.008
R
R
fac (%)
free (%) (test set, %)
rmsd bonds (Å)
respectively. After a treatment period of 4 weeks, animals were
sacrificed and the antitumor activity was determined as tumor
weight relative to the tumors in the vehicle control group. The
antimetastatic capacity was determined by macroscopic determi-
nation of metastases during a pathological examination.
X-ray Analysis. Structure determination and crystallographic
refinement were performed as described previously.11 Briefly,
platelike crystals of the benzamidine-matriptase complex were
grown at 18 °C from a solution containing 5 mg/mL matriptase,
10 mM benzamidine, 0.1 M Tris-HCl, pH 8.5, 20% PEG 8000,
and 200 mM MgCl2, using the hanging drop vapor diffusion
technique. These crystals were transferred to the same buffer lacking
benzamidine and containing 1 mM of compound 8 or 31. After
soaking for 3 h, the crystals were transferred to the same buffer
containing 16% glycerol and were flash-frozen in a nitrogen stream
of 100 K.
Complete data sets to 2.1 and 2.2 Å resolution were collected
in-house using a 345 mm MAR Research image plate (MAR
Research, Norderstedt, Germany) and monochromatic Cu KR
radiation from a RIGAKU rotating anode X-ray generator (Rigaku/
MSC, Sevenoaks, U.K.). The data were evaluated with the
MOSFLM package37 and scaled using SCALA from the CCP4
program suite38 (see Table 5). Starting with the matriptase
coordinates 1eax, the structures were crystallographically refined
using CNS, version 1.1.39 In between, the inhibitor compounds were
built with Sybyl 7.0 (Tripos, Inc.) and placed into the appropriate
electron density using MAIN.40 In the final rounds of refinement,
water molecules were added and individual restrained atomic B
values were refined. The Rfree was calculated from 5.0% of the
reflections not used for refinement. The final Rfac/Rfree values are
18.2/23.0 and 19.7/23.9% (see Table 5).
Acknowledgment. The authors thank Ute Altmann, Rainer
Sieber, Angelika Grau, Claudia Neuwirth, Cathleen Naumann,
and Martin Korsonewski for their excellent technical help. The
authors also thank Chris Privalle for correcting the English
version of the manuscript. This work was supported by the
Thuringian Ministry of Economy, Technology and Labor
(TMWTA).
Supporting Information Available: A table containing MS and
HPLC data of all final inhibitors and 13C NMR spectra of
compounds described in the Experimental Section of the manuscript.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Coussens, L. M.; Fingleton, B.; Matrisian, L. M. Matrix metallo-
proteinase inhibitors and cancer: Trials and tribulations. Science
2002, 295, 2387-2392.
(2) Rockway, T. W.; Nienaber, V.; Giranda, V. L. Inhibitors of the
protease domain of urokinase-type plasminogen activator. Curr.
Pharm. Des. 2002, 8, 2541-2558.

4126 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Steinmetzer et al.
(3) Steinmetzer, T. Synthetic urokinase inhibitors as potential antitumor
drugs. IDrugs 2003, 6, 137-145.
(24) Bernatowicz, M. S.; Wu, Y.; Matsueda, G. R. 1H-Pyrazole-1-
carboxamidine hydrochloride: An attractive reagent for guanylation
of amines and its application to peptide synthesis. J. Org. Chem.
1992, 57, 2497-2502.
(25) Zeslawska, E.; Schweinitz, A.; Karcher, A.; Sondermann, P.; Sperl,
S.; Stu¨rzebecher, J.; Jacob, U. Crystals of the urokinase type
plasminogen activator variant beta(c)-uPA in complex with small
molecule inhibitors open the way towards structure-based drug design.
J. Mol. Biol. 2000, 301, 465-475.
(26) Setyono-Han, B.; Stu¨rzebecher, J.; Schmalix, W. A.; Muehlenweg,
B.; Sieuwerts, A. M.; Timmermanns, M.; Magdolen, V.; Schmitt,
M.; Klijn, J. G. M.; Foekens, J. A. Suppression of rat breast cancer
metastasis and reduction of primary tumour growth by the small
synthetic urokinase inhibitor WX-UK1. Thromb. Haemostasis 2005,
93, 779-786.
(27) Stu¨rzebecher, J.; Prasa, D.; Wikstro¨m, P.; Vieweg, H. Structure-
activity relationships of inhibitors derived from 3-amidinophenyla-
lanine. J. Enzyme Inhib. 1995, 9, 87-99.
(28) Stu¨rzebecher, J.; Prasa, D.; Hauptmann, J.; Vieweg, H.; Wikstro¨m,
P. Synthesis and structure-activity relationships of potent thrombin
inhibitors: Piperazides of 3-amidinophenylalanine. J. Med. Chem.
1997, 40, 3091-3099.
(29) Bergner, A.; Bauer, M.; Brandstetter, H.; Stu¨rzebecher, J.; Bode, W.
The X-ray structure of thrombin in complex with NR-2-naphthyl-
sulfonyl-l-3-amidino-phenylalanyl-4-methylpiperidide: The benefical
effect of filling out an empty cavity. J. Enzyme Inhib. 1995, 9, 101-
110.
(30) Renatus, M.; Bode, W.; Huber, R.; Stu¨rzebecher, J.; Stubbs, M. T.
Structural and functional analyses of benzamidine-based inhibitors
in complex with trypsin: Implications for the inhibition of factor
Xa, tPA, and urokinase. J. Med. Chem. 1998, 41, 5445-5456.
(31) Turk, D.; Stu¨rzebecher, J.; Bode, W. Geometry of binding of the N
alpha-tosylated piperidides of m-amidino-, p-amidino- and p-guani-
dino phenylalanine to thrombin and trypsin. X-ray crystal structures
of their trypsin complexes and modeling of their thrombin complexes.
FEBS Lett. 1991, 287, 133-138.
(32) Fo¨rbs, D.; Thiel, S.; Stella, M. C.; Stu¨rzebecher, A.; Schweinitz, A.;
Steinmetzer, T.; Stu¨rzebecher, J.; Uhland, K. In vitro inhibition of
matriptase prevents invasive growth of cell lines of prostate and colon
carcinoma. Int. J. Oncol. 2005, 27, 1061-1070.
(33) Riddick, A. C. P.; Shukla, C. J.; Pennington, C. J.; Bass, R.; Nuttall,
R. K.; Hogan, A.; Sethia, K. K.; Ellis, V.; Collins, A. T.; Maitland,
N. J.; Ball, R. Y.; Edwards, D. R. Identification of degradome
components associated with prostate cancer progression by expression
analysis of human prostatic tissues. Br. J. Cancer 2005, 92, 2171-
2180.
(34) Saleem, M.; Adhami, V. M.; Zong, W.; Longley, B. J.; Lin, C.-Y.;
Dickson, R. B.; Reagan-Shaw, S.; Jarrad, D. F. Mukhtar, H. A novel
biomarker for staging human prostate adenocarcinoma: Overexpres-
sion of matriptase with concomitant loss of its inhibitor, hepatocyte
growth factor activator inhibitor-1. Cancer Epidemiol., Biomarkers
PreV. 2006, 15, 217-227.
(35) List, K.; Haudenschild, C. C.; Szabo, R.; Chen, W.; Wahl, S. M.;
Swaim, W.; Engelhol, L. H.; Behrendt, N.; Bugge, T. H. Matriptase/
MT-SP1 is required for postnatal survival, epidermal barrier function,
hair follicle development, and thymic homeostasis. Oncogene 2002,
21, 3765-3779.
(36) List, K.; Szabo, R.; Wertz, P. W.; Segre, J.; Haudenschild, C. C.;
Kim, S.-Y.; Bugge, T. H. Loss of proteolytically processed filaggrin
caused by epidermal deletion of matriptase/MT-SP1. J. Cell Biol.
2003, 163, 901-910.
(37) Leslie A. G. W. Molecular data processing. In Crystallographic
Computing 5; Moras, D., Podjarny, A. D., Thierry, J. C., Eds.; Oxford
University Press: Oxford, U.K., 1991; pp 50-61.
(38) Collaborative Computational Project, Number 4. The CCCP4 suite:
programs for protein crystallography. Acta Crystallogr. 1994, D50,
760-763.
(39) Bru¨nger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros,
P.; Grosse-Kunstleve, R. W.; Jiang, J. S.; Kuszewski, J.; Nilges, M.;
Pannu, N. S.; Read, R. J.; Rice, L. M.; Simonson, T.; Warren, G. L.
Crystallography & NMR system: A new software suite for macro-
molecular structure determination. Acta Crystallogr. 1998, D54, 905-
921.
(4) Hopper, J. D.; Clements, J. A.; Quigley, J. P.; Antalis, T. M. Type
II transmembrane serine proteases. J. Biol. Chem. 2001, 276, 857-
860.
(5) Takeuchi, T.; Harris, J. L.; Huang, W.; Yan, K. W.; Coughlin, S.
R.; Craik, C. S. Cellular localization of membrane-type serine protease
1 and identification of protease-activated receptor-2 and single-chain
urokinase-type plasminogen activator as substrates. J. Biol. Chem.
2000, 275, 26333-26342.
(6) Lee, S.-L.; Dickson, R. E.; Lin, C.-Y. Activation of hepatocyte growth
factor and urokinase/plasminogen activator by matriptase, an epithelial
membrane serine protease. J. Biol. Chem. 2000, 275, 36720-36725.
(7) Shi, Y. E.; Torri, J.; Yieh, L.; Wellstein, A.; Lippman, M. E.; Dickson,
R. B. Identification and characterization of a novel matrix-degrading
protease from hormone-dependent human breast cancer cells. Cancer
Res. 1993, 53, 1409-1415.
(8) Lin, C.-Y.; Wang, J.-K.; Torri, J.; Dou, L.; Sang, Q. A.; Dickson, R.
B. Characterization of a novel, membrane-bound, 80-kDa matrix-
degrading protease from human breast cancer cells. J. Biol. Chem.
1997, 272, 9147-9152.
(9) Lin, C.-Y.; Anders, J.; Johnson, M.; Dickson, R. B. Purification and
characterization of a complex containing matriptase and a kunitz-
type serine protease inhibitor from human milk. J. Biol. Chem. 1999,
274, 18237-18242.
(10) Takeuchi, T.; Shuman, M. A.; Craik, C. A. Reverse biochemistry:
Use of macromolecular protease inhibitors to dissect complex
biological processes and identify a membrane-type serine protease
in epithelial cancer and normal tissue. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 11054-11061.
(11) Friedrich, R.; Fuentes-Prior, P.; Ong, E.; Coombs, G.; Hunter, M.;
Oehler, R.; Pierson, D.; Gonzalez, R.; Huber, R.; Bode, W.; Madison,
E. Catalytic domain structures of MT-SP1/matriptase, a matrix-
degrading transmembrane serine proteinase. J. Biol. Chem. 2002, 277,
2160-2168.
(12) Oberst, M. D.; Johnson, M. D.; Dickson, R. B.; Lin, C.-Y. Singh,
B.; Stewart, M.; Williams, A.; al-Nafussi, A.; Smyth, J. F.; Gabra,
H.; Sellar, G. C. Expression of the serine protease matriptase and its
inhibitor HAI-1 in epithelial ovarian cancer: Correlation with clinical
outcome and tumor clinicopathological parameters. Clin. Cancer Res.
2002, 8, 1101-1107.
(13) Bhatt, A. S.; Takeuchi, T.; Yistra, B.; Ginzinger, D.; Albertson, D.;
Shuman, M. A.; Craik, C. S. Quantification of membrane type serine
protease 1 (MT-SP1) in transformed and normal cells. Biol. Chem.
2003, 384, 257-266.
(14) Hoang, C. D.; D’Cunha, J. D.; Kratzke, M. G.; Casmey, C. E.;
Frizelle, S. P.; Maddaus, M. A.; Kratzke, R. A. Gene expression
profiling identifies matriptase overexpression in malignant mesothe-
lioma. Chest 2004, 125, 1843-1852.
(15) List, K.; Szabo, R.; Molinolo, A.; Sriuranpong, V.; Redeye, V.;
Murdock, T.; Burke, B.; Nielson, B. S.; Gutkind, J. S.; Bugge, T. H.
Deregulated matriptase causes ras-independent multistage carcino-
genesis and promotes ras-mediated malignant transformation. Genes
DeV. 2005, 19, 1934-1950.
(16) Enyedy, I.; Lee, S.-L.; Kuo, A. H.; Dickson, R. B.; Lin, C.-Y.; Wang,
S. Structure-based approach for the discovery of bis-benzamidines
as novel inhibitors of matriptase. J. Med. Chem. 2001, 44, 1349-
1355.
(17) Long, Y.-Q.; Lee, S.-L.; Lin, C.-Y.; Enyedy, I.; Wang, S.; Li, P.;
Dickson, R. B.; Roller, P. P. Synthesis and evaluation of the sunflower
derived trypsin inhibitor as a potent inhibitor of the type II
transmembrane serine protease, matriptase. Bioorg. Med. Chem. Lett.
2001, 11, 2515-2519.
(18) Galkin, A. V.; Mullen, L.; Fox, W. D.; Brown, J.; Duncan, D.;
Moreno, O.; Madison, E. L.; Agus, D. B. CVS-3983, a selective
matriptase inhibitor, suppresses the growth of androgen independent
prostate tumor xenografts. Prostate 2004, 61, 228-235.
(19) Markwardt, F.; Wagner, G.; Stu¨rzebecher, J.; Walsmann, P. N Alpha-
arylsulfonyl-omega-(4-amidinophenyl)-alpha-aminoalkylcarboxylic
acid amides. Novel selective inhibitors of thrombin. Thromb. Res.
1980, 17, 425-431.
(20) Stu¨rzebecher, J.; Vieweg, H.; Steinmetzer, T.; Schweinitz, A.; Stubbs,
M. T.; Renatus, M.; Wikstro¨m, P. 3-Amidinophenylalanine-based
inhibitors of urokinase. Bioorg. Med. Chem. Lett. 1999, 9, 3147-
3152.
(21) Judkins, B. D.; Allen, G. D.; Cook, T. A.; Evans, B.; Sardharwala,
T. S. A versatile synthesis of amidines from nitriles via amidoximes.
Synth. Commun. 1996, 26, 4351-4367.
(40) Turk, D. Weiterentwicklung eines Programmes fu¨r Moleku¨lgraphik
und Elektronendichte-Manipulation und seine Anwendung auf ver-
schiedene Protein-Strukturaufkla¨hrung (Advancement of a program
for molecular graphics and electron density manipulation and its
application to different protein structures). Ph.D. Thesis, Technische
Universita¨t Mu¨nchen, Germany, 1992.
(22) Kirchner, F. K.; McCormick, J. R.; Cavallito, C. J.; Miller, L. C.
The use of diazo compounds in the preparation of some benzylpeni-
cillin esters. J. Org. Chem. 1949, 14, 388-393.
(23) Phillips, A. P. Synthetic hypotensive agents. V. 4-(2′-Aminoethyl)-
piperidine derivatives. J. Am. Chem. Soc. 1957, 79, 2836-2838.
JM051272L