Potent bivalent thrombin inhibitors: Replacement of the scissile peptide bond at P1-P1' with arginyl ketomethylene isosteres

Article abstract of DOI:10.1021/jm9807297

We have designed highly potent synthetic bivalent thrombin inhibitors, which consist of an active site blocking segment, a fibrinogen recognition exosite blocking segment, and a linker connecting these segments. The bivalent inhibitors bind to the active

Full text of DOI:10.1021/jm9807297

J . Med. Chem. 1999, 42, 3109-3115
3109
P oten t Biva len t Th r om bin In h ibitor s: Rep la cem en t of th e Scissile P ep tid e
Bon d a t P 1-P 1′ w ith Ar gin yl Ketom eth ylen e Isoster es3
Torsten Steinmetzer,†,§ Bing Yan Zhu,‡,| and Yasuo Konishi*,†
Biotechnology Research Institute, National Research Council Canada, 6100 Royalmount Avenue,
Montreal, Quebec, Canada H4P 2R2, and BioChem Therapeutic Inc., 275 Armand-Frappier Boulevard,
Laval, Quebec, Canada H7V 4A7
Received December 28, 1998
We have designed highly potent synthetic bivalent thrombin inhibitors, which consist of an
active site blocking segment, a fibrinogen recognition exosite blocking segment, and a linker
connecting these segments. The bivalent inhibitors bind to the active site and the fibrinogen
recognition exosite simultaneously. As a result, the inhibitors showed much higher affinity for
thrombin than the individual blocking segments. Various arginyl ketomethylene isosteres ArgΨ-
[CO-CH2-X]P1′ were incorporated into the bivalent inhibitors as P1-P1′ segment to eliminate
the scissile bond. The P1′ residue is a natural or unnatural amino acid; specifically, the
incorporation of mercaptoacetic acid exhibited superiority in synthesis and affinity for thrombin.
Inhibitor 16, (D-cyclohexylalanine)-Pro-ArgΨ[CO-CH2-S]Gly-(Gly)4-Asp-Tyr-Glu-Pro-Ile-Pro-
Glu-Glu-Tyr-cyclohexylalanine-(D-Glu)-OH, showed the lowest Ki value of 3.5 ( 0.5 × 10-13 M,
which is comparable to that (Ki ) 2.3 × 10-13 M) of recombinant hirudin. Consequently we
successfully reduced the size of the inhibitor from ∼7 kDa of recombinant hirudin to ∼2 kDa
without losing the affinity.
In tr od u ction
bolytic therapy seems to be due to the inaccessibility of
heparin to the thrombin molecules in the thrombus.8,11,12
The trypsin-like serine protease thrombin (EC 3.4.21.5)
is a primary target in developing successful anticoagu-
lants due to its central functions in the process of
hemostasis and thrombosis. Thrombin converts soluble
circulating fibrinogen into clottable fibrin and amplifies
its own generation through the activation of other
coagulation enzymes such as factors V and VIII.1,2 In
addition, thrombin activates factor XIII, which stabilizes
the clot by cross-linking fibrin, and stimulates platelet
secretion and aggregation through a proteolytic activa-
tion of the thrombin receptor on the platelets.3 Throm-
bin also mediates a negative-feedback regulation of the
coagulation cascade by activating protein C upon bind-
ing to the endothelial cell surface protein thrombo-
modulin.1
The importance of blocking thrombin activity in
prophylaxis of thromboembolic disorders in high-risk
patients was described elsewhere.4,5 Anticoagulants
such as heparin and antiplatelet agents such as aspirin
are routinely used during coronary angioplasty.6 They
are also used in combination with thrombolytic therapy
induced by urokinase, streptokinase, or tissue plasmi-
nogen activator7 to prevent rethrombosis.8 Thrombin
released from thrombus seems to be one of the major
components responsible for the reocclusion observed in
10-20% of patients despite heparinization.9,10 The
modest benefit of heparin in combination with throm-
Various small active-site thrombin inhibitors for
chronic dosing have been extensively designed over
nearly three decades. However, there has been no
compound in clinical use that clearly satisfies all of the
necessary criteria of high and reproducible oral bio-
availability, long duration, selectivity, and satisfactory
pharmacokinetic properties.13 In acute treatment, hiru-
din, a 65-amino acid protein from the leech Hirudo
medicinalis, may have some advantages over the low-
molecular-weight inhibitors or heparin.14 Hirudin, due
to its bivalent binding mode, has a high selectivity and
a high affinity for human R-thrombin (Ki ) 2.2 × 10-14
M):15 the N-terminal globular hirudin(1-48) contacts
the thrombin surface adjacent to the active site, and
simultaneously, the C-terminal hirudin(55-65) inter-
acts with the fibrinogen recognition exosite (FRE) of
thrombin.16 Another advantage of hirudin is that hiru-
din reduces fibrin polymerization rates and the content
of fibrin-fibrinogen in the blood clots more efficiently
than the covalent small active-site inhibitor PPACK.17
This suggests that the FRE accelerates the fibrin clot
assembly as a cofactor. Furthermore, hirudin, but not
heparin, neutralized both thrombin in solution and
fibrin-bound thrombin in the coagulation activated
during thrombolysis, which was induced by a plasmi-
nogen activator.18,19
The high selectivity and affinity of hirudin motivated
the design of synthetic bivalent inhibitors in which the
bulky hirudin(1-48) was replaced by a small active-site-
directed inhibitor segment, e.g., D-Phe-Pro-Arg-Pro,
which was then connected to the FRE blocking segment
r-hirudin(55-65) through an oligopeptide spacer.20,21
Substitution of the N-terminal D-Phe by D-cyclohexyl-
alanine (D-Cha) improved not only affinity but also
3 NRCC Publication no. 42921.
* Author for correspondence.
† National Research Council Canada.
‡ BioChem Therapeutic Inc.
§ Present address: Biological-Pharmaceutical Department, Institute
of Biochemistry & Biophysics, Friedrich-Schiller-University J ena,
Philosophenweg 12, 07743 J ena, Germany.
| Present address: COR Therapeutics Inc., 256 East Grand Ave,
South San Francisco, CA 94080.
10.1021/jm9807297 CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/16/1999

3110 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Steinmetzer et al.
Sch em e 1
The type I pseudopeptide bonds ArgΨ[CO-CH2-NR]-
Xaa, where R is H, CH3, or acetyl, were used in
inhibitors 6-13. They were generated by coupling Boc-
Arg(Z2)-CH2I (2) to the N-terminus of P1′-linker-(the
FRE blocking segment)-Pam resin, while the side chains
of the peptides on the resin were still protected. The
reaction was never completed, even with a large excess
of the iodomethyl ketone (10 times), at high temperature
(37 °C), and with a prolonged coupling time of 1 week.
The unreacted amino groups were capped with acetic
anhydride. No capping was performed for peptides 6 and
11 in order to avoid acetylation of the secondary amine
at the P1′ residue. The coupling yield of 2 was 15-55%
based on the HPLC profiles of the crude products. A low
yield (∼25%) of alkylation was also reported in the
solution synthesis of a furine inhibitor with a similar
ArgΨ[CO-CH2-NH]Ala building unit.33 The coupling of
the P3-P2 dipeptide to H-Arg(Z2)Ψ[CO-CH2-NR]Xaa-
linker-(the FRE blocking segment)-Pam resin avoided
the formation of the N-terminal unprotected dipeptidyl
ketone intermediate, which has a strong tendency to
undergo intramolecular cyclization.31,34,35
proteolytic stability against thrombin.22 Pro at the P1′
residue also improved proteolytic stability against throm-
bin; however, the Arg-Pro peptide bond was still slowly
cleaved by thrombin.21 Complete proteolytic stability
against thrombin was achieved by introducing ho-
moarginine ArgΨ[CH2-CO-NH]Gly or a reduced bond
ArgΨ[CH2-NH]Gly, although they also reduced the
affinity of the inhibitors 3-10-fold compared to the
corresponding inhibitor with Arg-Pro.23 Ketomethylene
pseudopeptide bonds ArgΨ[CO-(CH2)n]Gly (n ) 1-4),
on the other hand, were successfully introduced with
full proteolytic stability against thrombin and no loss
of affinity. This class of inhibitors achieved a Ki value
of as little as 1.4 × 10-10 M.24,25 Ketoamide transition-
state mimetics ArgΨ[CO-CO-NH]Gly were further in-
troduced to stimulate the affinity of the inhibitor, i.e.,
the Ki value of the inhibitor CVS995 in this class, which
reached 1 × 10-12 M; however, it should be recalled that
synthesis of the building unit of the transition-state
analogue, 6-nitroguanidino-3(S)-t-Boc-amido-2(R,S)-hy-
droxyhexanoic acid, is labor-intensive.26
In this article, we introduced a series of ketomethyl-
ene pseudopeptide bonds containing substituted or
nonsubstituted heteroatoms (N or S)27-29 into the biva-
lent thrombin inhibitors. These are ArgΨ[CO-CH2-NH]-
Gly, ArgΨ[CO-CH2-N(CH3)]Gly, ArgΨ[CO-CH2-N(acetyl)]-
Gly, ArgΨ[CO-CH2-NH-CO-CH2]Gly, and ArgΨ[CO-
CH2-S]Gly. The incorporation of ArgΨ[CO-CH2-S]Gly
has advantages in the yield of the synthesis as well as
in the affinity for thrombin.
The type II pseudopeptide bond ArgΨ[CO-CH2-NH-
CO-CH2]Gly has an interesting feature in that -NH-
CH2- of the original P1′ glycine residue is reversed and
the NH group of the normal P2′ glycine residue is
replaced by a methylene group. Unlike the type I
pseudopeptide bond, there is no insertion of a methylene
group at the P1′ residue. The chloromethyl ketone 1 was
slowly converted to the aminomethyl ketone 4 with an
excess of NH3 in absolute diethyl ether. Compound 4
was reacted with succinic anhydride to give compound
5 (Scheme 1). Building unit 5 was then incorporated into
the inhibitors by using conventional solid-phase peptide
synthesis. The HPLC profile of the crude peptide 17
showed two major peaks (∼1:1 ratio) at 24.6 and 25.0
min (peptides 17a and 17b). The peptides showed the
same molecular mass (2096, MH+) and the same profile
in the amino acid analysis. Peptide 17a inhibited
thrombin in the nanomolar range of the Ki value,
whereas 17b showed no significant inhibition even at a
concentration of 1 µM. Based on the results of the HPLC
profile, the molecular mass, the amino acid composition,
and the inhibition constant, inhibitor 17 seems to be
racemized, most likely during the synthesis of the
aminomethyl ketone 4 because of the large excess of
NH3 used.
The type III pseudopeptide bond has a formula of
ArgΨ[CO-CH2-S]Gly. Unlike the low alkylating yield of
the type I pseudopeptide bond, chloromethyl ketones
easily alkylate mercapto groups such as dithioerythri-
tol.36 Thus, the intermediate Boc-Arg(Z2)Ψ[CO-CH2-S]-
Gly (3) was quantitatively produced in the presence of
diisopropylethylamine (Scheme 1). Compound 3 was
incorporated in the bivalent inhibitors by standard solid-
phase peptide synthesis procedure. Similar type III
pseudopeptide bonds have been reported in active-site-
directed thrombin inhibitors37 and angiotensin-convert-
ing enzyme (ACE) inhibitors.27,29
Ch em istr y
Three types of pseudopeptide bonds were introduced
in the bivalent thrombin inhibitors. The Arg side chain
was protected by two benzyloxycarbonyl (Z) groups in
all syntheses to eliminate side reaction during the
synthesis. The P2 and P3 amino acids were coupled as
a dipeptide, e.g., Boc-(D-Phe)-Pro-OH, for all inhibitors
in this article. Compound 1, Boc-Arg(Z2)-CH2Cl, was
synthesized by procedures described elsewhere30-32 with
slight modifications. Compound 1 was further converted
to 2, 3, and 5 (Scheme 1), which were then used in solid-
phase synthesis.
Table 1 lists the sequences of the bivalent thrombin
inhibitors used in this article.
Resu lts a n d Discu ssion
Slow In h ibition . Slow binding inhibition was ob-
served in some inhibitors such as 9-11 (Figure 1).38

Potent Bivalent Thrombin Inhibitors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3111
Ta ble 1. Sequences of the Bivalent Thrombin Inhibitors
inhibitor
active site blocking segment
linker
FRE blocking segment
6
7
8
(D-Phe)-Pro-ArgΨ[COCH2NH]Gly
(D-Phe)-Pro-ArgΨ[COCH2NCH3]Gly
(D-Phe)-Pro-ArgΨ[COCH2NCH3]Gly
(D-Cha)-Pro-ArgΨ[COCH2NCH3]Gly
(D-Cha)-Pro-ArgΨ[COCH2NAc]Gly
(D-Cha)-Pro-ArgΨ[COCH2NH]Ala
(D-Cha)-Pro-ArgΨ[COCH2NCH3]Gly
(D-Cha)-Pro-ArgΨ[COCH2N]Pro
(D-Phe)-Pro-ArgΨ[COCH2S]Gly
(Gly)4
(Gly)4
(Gly)4
(Gly)4
(Gly)4
(Gly)4
(Gly)4-Asn-Gly
(Gly)4-Asn-Gly
(Gly)4
Asp-Phe-Glu-Glu-Ile-Pro-Glu-Glu-Tyr-Leu-Gln-OH
Asp-Phe-Glu-Glu-Ile-Pro-Glu-Glu-Tyr-Leu-Gln-OH
Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Tyr-Cha-Asp-OH
Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Tyr-Cha-Asp-OH
Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Tyr-Cha-Asp-OH
Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Tyr-Cha-Asp-OH
Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Ala-Cha-(D-Glu)-OH
Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Ala-Cha-(D-Glu)-OH
Asp-Phe-Glu-Glu-Ile-Pro-Glu-Glu-Tyr-Leu-Gln-OH
Asp-Phe-Glu-Glu-Ile-Pro-Glu-Glu-Tyr-Leu-Gln-OH
Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Ala-Cha-(D-Glu)-OH
Asp-Phe-Glu-Glu-Ile-Pro-Glu-Glu-Tyr-Leu-Gln-OH
9
10
11
12
13
14
15
16
17
Ac-(D-Phe)-Pro-ArgΨ[COCH2S]Gly
(D-Cha)-Pro-ArgΨ[COCH2S]Gly
(D-Phe)-Pro-ArgΨ[COCH2NHCOCH2]Gly
(Gly)4
(Gly)4
ꢀAhx-Gly
F igu r e 2. Observed first-order rate constant kobs at various
concentrations of inhibitor 9.
inhibitor 9 was estimated as 9.5 ( 1.4 × 108 M-1 s-1
,
which is 7 times faster than that (kon ) 1.37 ( 0.03 ×
108 M-1 s-1) of r-hirudin.39 Site-directed mutagenesis
of r-hirudin at the FRE blocking segment, e.g., substitu-
tions of GluH57, GluH58, GluH61, and GluH62 by Gln,
predominantly reduced the kon value with little effect
on the koff value.39 Similarly, the improved kon value of
9 may be due to its 10-fold higher affinity for the FRE
blocking segment compared to that of r-hirudin(55-
65).40 The kon values, which are close to a diffusion-
controlled process, were also observed in inhibitors 10
and 11. Thus, the slow inhibition of these inhibitors is
simply due to the low inhibitor concentrations (5-50 ×
10-11 M) in the assays.
F igu r e 1. Slow inhibition of human R-thrombin by inhibitor
9 as monitored by the hydrolysis of the fluorogenic substrate
tosyl-Gly-Pro-Arg-AMC (40 µM). The solid lines are the
nonlinear least-squares fitting of the data (b at [I] ) 0 M, 1
at [I] ) 5.1 × 10-11 M, 3 at [I] ) 8.5 × 10-11 M, * at [I] ) 1.19
× 10-10 M, 0 at [I] ) 1.70 × 10-10 M, 9 at [I] ) 2.6 × 10-10 M,
and 4 at [I] ) 5.1 × 10-10 M) using eq 2.
Ta ble 2. Inhibition Constants and Inhibitor Class of the
Bivalent Thrombin Inhibitorsa
inhibitor
Ki (M)
inhibitor classb
6
7
8
8.3 × 1.5 × 10-11
8.3 × 0.8 × 10-11
3.5 × 0.7 × 10-11
2.0 × 0.3 × 10-12
1.7 × 0.3 × 10-12
1.5 × 0.2 × 10-12
8.7 × 0.3 × 10-13
5.7 × 0.7 × 10-12
9.4 × 1.2 × 10-12
1.06 × 0.10 × 10-9
3.5 ( 0.5 × 10-13
7.3 ( 0.4 × 10-9
classical
classical
classical
slow binding
slow binding
slow binding
slow, tight binding
slow binding
slow binding
hyperbolicc
9
F RE In h ibitor Segm en t. Krstenansky et al.40 ex-
tensively made single or multiple substitutions of the
FRE inhibitor fragment of r-hirudin. For example, single
substitutions of PheH56 by Tyr, of GluH58 by Pro, and
of GlnH65 by Asp or D-Glu improved the IC50 of the
r-hirudin-based FRE inhibitors by 10%, 70%, and 60%,
respectively. Although the effect of each substitution
was relatively small, the combined multiple substitu-
tions improved the IC50 value up to 10-fold, e.g., a
multiple substitution of PheH56, GluH58, LeuH64, and
GlnH65 by Tyr, Pro, Cha, and Asp, respectively, and a
multiple substitution of PheH56, GluH58, TyrH63,
LeuH64, and GlnH65 by Tyr, Pro, Ala, Cha, and D-Glu,
respectively. The IC50 values of the N-terminal acety-
lated FRE inhibitors 18-21, which were incorporated
into the bivalent inhibitors in this article, are listed in
Table 3.
10
11
12
13
14
15
16
17
slow, tight binding
hyperbolicc
a
b
Values are shown as mean ( SE at pH 7.8 and 23 °C. The
inhibitors are classified according to Morrison and Walsh38 except
hyperbolic inhibition. c The hyperbolic inhibition was analyzed as
described elsewhere.47 Kr values are 0.0078 and 0.013 for inhibitors
15 and 17, respectively.
Each progress curve of the slow inhibition was fitted to
eq 2 (see the Experimental Section). The steady-state
velocity vS was then incorporated into eq 1 to estimate
the Ki value (Table 2). The parameter kobs increased
linearly with the inhibitor concentration (Figure 2). The
N-Ter m in a l P 3 Resid u e. X-ray structures of throm-
bin in complex with D-Phe-Pro-Arg-based inhibitors
showed that the side chain of the P3 residue D-Phe binds
slope and the intercept are kon/(1 + [S]/Km) and koff
,
respectively (eq 3). The association rate constant kon of

3112 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Ta ble 3. Inhibition (IC50) and Sequences of the FRE Inhibitorsa
Steinmetzer et al.
inhibitor
sequence
IC50 (M)
18
19
20
21
Ac-Asp-Phe-Glu-Glu-Ile-Pro-Glu-Glu-Tyr-Leu-Gln-OH
Ac-Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Tyr-Cha-Asp-OH
Ac-Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Ala-Cha-(D-Glu)-OH
Ac-Asn-Gly-Asp-Tyr-Glu-Pro-Ile-Pro-Glu-Glu-Ala-Cha-(D-Glu)-OH
1.46 ( 0.10 × 10-6
2.8 ( 0.2 × 10-7
1.35 ( 0.10 × 10-7
1.59 ( 0.22 × 10-7
a
Values are shown as mean ( SE at pH 7.52 and 37 °C.
to a hydrophobic cavity, the so-called aryl binding site,41
which is formed by Leu99, Ile174, Trp215, and Tyr60A.
Inhibitor 9 improved the Ki value ∼20-fold by substitut-
ing N-terminal D-Phe of 8 with D-Cha. This is consistent
with the ∼30-fold improvement of hirulog-B2 by replac-
ing N-terminal D-Phe in hirulog-1 with D-Cha.22 The
bivalent thrombin inhibitors with an N-terminal acetyl-
D-Phe- (hirutonin-2)42 or D-Cha residue (P596)43 insert
the side chain of D-Cha or D-Phe into the same S3 subsite
according to their crystal structures; however, the
cyclohexyl ring was slightly rotated with respect to the
plane of the phenyl ring. Rehse et al.43 suggested that
the improved affinity of D-Cha is due to the stronger
van der Waals interactions of the saturated and non-
planar ring, which points directly toward the aliphatic
side chains of Ile174 and Leu99. Another drastic effect
of the P3 residue is the acetylation of the N-terminal
amino group. Inhibitor 15 with an N-terminal acetyl
group has low affinity (Ki ) 1.06 ( 0.10 × 10-9 M) for
thrombin, and the inhibition was hyperbolic. Removal
of the N-terminal acetyl group resulted in inhibitor 14,
with ∼100-fold improved Ki value (Ki ) 9.4 ( 1.2 × 10-12
M) and slow inhibition mode. Similar large effects of
the N-terminal acetylation were reported in active-site-
directed inhibitors based on the D-Phe-Pro-Arg se-
quence44 and in the kcat/Km values of the chromogenic
substrates.45 In the crystal structures of hirutonin-242
and P596,43 both the N-terminal amide and the amino
protons of hirutonin-2 and P596, respectively, form a
hydrogen bond to the carbonyl oxygen of Gly216 of
thrombin; however, the amide hydrogen bond is gener-
ally weaker than that of the free amino group.45,46 In
addition, the N-terminal amino group in the P596-
thrombin complex was well-hydrated by several bound
water molecules, whereas no bound water molecule was
observed around the N-terminal acetyl group in the
hirutonin-2-thrombin complex. The weaker hydrogen
bond and the larger dehydration energy may contribute
to the weaker affinity of inhibitor 15 compared to that
of inhibitor 14.
In h ibitor s w ith a n Ar gΨ[CO-CH2-NR]Gly Seg-
m en t. Peptides 6 and 7 with ArgΨ[CO-CH2-NH]Gly or
ArgΨ[CO-CH2-N(CH3)]Gly at the P1-P1′ residues, re-
spectively, were classical competitive inhibitors with Ki
values of 8.3 ( 1.5 × 10-11 and 8.3 ( 0.8 × 10-11 M,
respectively. Both 6 and 7 bound to thrombin 25 times
tighter than hirulog-1, D-Phe-Pro-Arg-Pro-(Gly)4-(r-hiru-
din(53-64)) (Ki ) 2.3 × 10-9 M).20 It should be noticed
that the P1′ residues of inhibitors 6 and 7 are basic due
to their secondary and tertiary amines, respectively. If
these amines were protonated in the thrombin-inhibi-
tor complex, their electron-withdrawing effect on the
carbonyl carbon of the P1 residue would stabilize the
transition-state analogue binding mode with enhanced
affinity for thrombin. On the other hand, the corre-
sponding secondary or tertiary amines of the ACE
inhibitors containing PheΨ[CO-CH2-NH]Gly or PheΨ-
[CO-CH2-N(CH3)]Gly were suggested to be a hydrogen
bond acceptor; i.e., the amines were not protonated in
the enzyme-inhibitor complexes.29 We investigated the
amines further by preparing a reference inhibitor 10 in
which the amine of the P1′ residue is acetylated with
no possibility of protonation. The acetylation had no
effect on affinity, with a Ki value of 1.7 ( 0.3 × 10-12
M
compared to that (2.0 ( 0.3 × 10-12 M) of the reference
inhibitor 9. Thus, it is unlikely that inhibitors 6 and 7
are protonated at the amine of the P1′ residue and act
as transition-state mimetics.
In h ibitor w ith a n Ar gΨ[CO-CH2-NH-CO-CH2]Gly
Segm en t. Inhibitors 6-16 have a methylene group
inserted between the P1 and P1′ residues, shifting the
peptide chain of the P′ residues by one atom. This shift
was eliminated in inhibitor 17, where the ArgΨ[CO-
CH2-NH-CO-CH2]Gly segment corresponding to the P1-
P1′-P2′ residues have the same number of main chain
atoms as the natural amino acid residues. The Ki value
(7.3 ( 0.4 × 10-9 M) of inhibitor 17 was surprisingly
high, and the inhibition was hyperbolic. This is ∼100-
fold less active than the similar inhibitor 6. Despite the
significant affinity loss of 17, there is no obvious atom
in inhibitor 17 which may disturb the binding to
thrombin; i.e., the NH of the P1′ residue of inhibitor 17
is located at the same position as that in inhibitors 6
and 11; the carbonyl of the P1′ residue is located at the
same position as that of the substrates; and 6-amino-
hexanoic acid (ꢀAhx) was well-accepted in the thrombin
S′ subsites with almost no effect on the affinity for
thrombin.47
In h ib it or s w it h a n Ar gΨ[CO-CH 2-S]Gly Seg-
m en t. Inhibitor 14 replaces the ArgΨ[CO-CH2-NH]Gly
pseudopeptide bond of inhibitor 6 by ArgΨ[CO-CH2-S]-
Gly and improves the Ki value ∼9-fold, in contrast to
the negative effects reported previously where incorpo-
ration of the corresponding PheΨ[CO-CH2-S]Gly at the
P1-P1′ residues of ACE inhibitors reduced in vitro
activity significantly or inactivated the inhibitor in
vivo.27,29 Further substitutions of the P3 residue and of
the FRE blocking segment led to a potent bivalent
thrombin inhibitor 16 with a Ki value of 3.5 ( 0.5 ×
10-13 M, comparable to that (2.3 × 10-13 M) of r-
hirudin.39 The advantages of R-thiomethyl ketone are
not only the improved affinity of the inhibitor but also,
and more importantly, the straightforward synthesis of
Boc-Arg(Z)2Ψ[CO-CH2-S]Gly-OH (3) with a high yield
of 89%, compared to the synthesis of other pseudopep-
tide bonds which are often tedious with low yield.
Furthermore, it provides an option to incorporate amino
acid mimetics other than Gly mimetic at the P1′ residue;
e.g., 2-mercaptopropionic acid is a mimetic of Ala,
3-mercaptopropionic acid substitutes for â-Ala, and
mercaptosuccinic acid mimics Asp.

Potent Bivalent Thrombin Inhibitors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3113
Boc-D-Ch a -P r o-OH. The dipeptide was synthesized using
Con clu sion
the same method as that for Boc-D-Phe-Pro-OH described
Arginyl methyl ketone derivatives were incorporated
as P1-P1′ residues in the bivalent thrombin inhibitors
by a combination of solution and solid-phase peptide
synthesis and eliminated the scissile peptide bond. The
use of a pseudopeptide bond based on an R-aminomethyl
ketone moiety (ArgΨ[CO-CH2-NR]Xaa) resulted in po-
tent inhibitors with picomolar Ki values. Although
inhibitors 9 and 11 contain basic amines of ArgΨ[CO-
CH2-N(CH3)]Gly and ArgΨ[CO-CH2-NH]Gly, respec-
tively, it is unlikely that these amines are protonated
in the enzyme-inhibitor complex or that the inhibitors
act as transition-state analogues. Pseudopeptide bond
ArgΨ[CO-CH2-S]Gly exhibited dual advantages in in-
hibitor design; i.e., Boc derivative of ArgΨ[CO-CH2-S]-
Gly was easily synthesized with a high yield of 89%,
and inhibitor 16 containing this pseudopeptide bond
bound to human R-thrombin with the highest affinity
(Ki ) 3.5 ( 0.5 × 10-13 M), comparable to that (Ki ) 2.3
× 10-13 M) of r-hirudin.
above: mp 158-162 °C; [R]23 ) -48.3° (c 0.63, acetic acid);
D
tR 38.9 min; 13C NMR (CDCl3) 174.27, 172.76, 155.65, 79.98,
59.98, 49.94, 47.37, 40.30, 34.04, 33.89, 32.55, 28.25, 27.93,
26.35, 26.21, 26.00, 24.63.
Boc-Ar g(Z)2-CH2Cl (1). Isobutyl chloroformate (2.21 mL,
17 mmol) was added to a solution of Boc-Arg(Z2)-OH (8.68 g,
16 mmol) and N-methylmorpholine (1.87 mL, 17 mmol) in 75
mL of dry tetrahydrofuran (THF) at -15 °C. After 15 min,
the precipitated salt was removed by filtration and the mixed
anhydride solution was cooled to -70 °C. The diazomethane
generated from Diazald was distilled at 30-35 °C into the
cooled (-70 °C) mixed anhydride solution. The reaction
mixture was then kept overnight at 4 °C. The solvent and
the excess of diazomethane were removed in vacuo. The
remaining crystalline diazomethyl ketone was redissolved in
250 mL of dry THF and cooled to 0 °C. The diazomethyl ketone
was converted to 1 by adding 8 mL of 4 N HCl/dioxane
dropwise over a period of 10 min at 0 °C. The reaction solution
was further incubated for 10 min at 0 °C before the solvent
was evaporated. 1 was redissolved in ethyl acetate and washed
three times with half-saturated NaHCO3 solution and three
times with saturated NaCl solution. The solution was dried
over Na2SO4, and the solvent was evaporated. The product was
crystallized from ethyl acetate/hexane and recrystallized from
Exp er im en ta l Section
Ma ter ia ls. 7-Aminomethylcoumarin (AMC) was obtained
from Aldrich (Milwaukee, WI). Human R-thrombin (3000 NIH
units/mg), bovine fibrinogen (70% protein, 85% of protein
clottable), tosyl-Gly-Pro-Arg-AMC‚HCl salt, poly(ethylene gly-
col) 8000, and Tris were purchased from Sigma (St. Louis,
MO). Boc-Cha-OH, Boc-D-Cha-OH, Boc-sarcosine-OH, and Boc-
Arg(Z2)-OH were purchased from BaChem BioScience Inc.
(Philadelphia, PA). All other amino acid derivatives for peptide
synthesis were obtained from Advanced ChemTech (Louisville,
KY). The side chain protecting groups for Boc-amino acids are
benzyl for Glu, cyclohexyl for Asp, and 2-bromobenzyloxycar-
bonyl for Tyr. Boc-Gln-Pam resin and Boc-Asp(OBzl)-Pam
resin were purchased from Applied Biosystems Inc. (Foster
City, CA). Boc-D-Glu(OBzl)-Pam resin was purchased from
Peninsula Laboratories Inc. (Belmont, CA). The solvents for
peptide synthesis were purchased from Anachemia Chemical
Inc. (Rouses Point, NY) and Applied Biosystems Inc. HF and
TFA were purchased from Matheson Gas Products Inc. (Se-
caucus, NJ ) and Halocarbon Products Co. (North August, SC),
respectively. All other chemicals were from Aldrich (Milwau-
kee, WI).
Gen er a l. Melting points were determined on a micro hot
plate according to BOETIUS and were uncorrected. Optical
rotations were measured with a Polamat (Carl Zeiss J ena,
Germany). The molecular mass of the compounds was deter-
mined using a SCIEX API III mass spectrometer (Sciex,
Concord, ON, Canada). A Beckman model 6300 amino acid
analyzer (Beckman, Fullerton, CA) was used to determine the
amino acid composition and the peptide content of the inhibi-
tors. The bivalent inhibitors were purified by a preparative
HPLC (Vydac C4 column, 46 × 250 mm, 33 mL/min flow rate,
linear gradient of 15-45% acetonitrile in 0.1% TFA in 80 min,
monitored at 220 nm) and, if necessary, further by a semi-
preparative HPLC with the same gradient (Vydac C18 column,
20 × 250 mm, 13 mL/min flow rate). Analytical HPLC
experiments were performed on a Waters HPLC (Vydac C18
column, 4.6 × 250 mm) with a linear gradient of acetonitrile
(10-70% in 60 min) containing 0.1% TFA at a flow rate of 1
mL/min. The 13C NMR spectra were recorded on an AC-250
(Bruker) at 62 MHz and referenced to internal tetramethyl-
silane.
ethanol: yield 59%; mp 67-69 °C; [R]23 ) -17.1° (c 0.50,
D
acetic acid); tR 53.5 min; MS 575 (MH+); 13C NMR (CDCl3)
201.61, 163.55, 160.52, 155.67, 155.56, 136.54, 134.51, 128.92,
128.83, 128.46, 128.36, 128.11, 128.03, 80.14, 69.03, 67.14,
57.2446.85, 44.02, 28.26, 26.92, 24.55.
Boc-Ar g(Z)2-CH2I (2). NaI (15 g, 100 mmol) was added to
compound 1 (5 g, 8.7 mmol) solution in 200 mL of anhydrous
dimethylformamide (DMF). The reaction mixture was incu-
bated with stirring at 35-40 °C for 8 h in the dark before the
solvent was evaporated. The product was then redissolved in
ethyl acetate, washed three times with half-saturated NaCl
solution, and dried over Na2SO4. After evaporating the solvent,
the crude oily product, which still contained 5-10% of the
starting material 1 (determined by analytical HPLC), was used
in the solid-phase peptide synthesis without further purifica-
tion: MS 667 (MH+), tR 55.4 min.
Boc-Ar g(Z)2Ψ[CO-CH2-S]Gly-OH (3). Compound 1 (1 g,
1.74 mmol), diisopropylethylamine (DIEA) (0.45 mL, 2.6
mmol), and mercaptoacetic acid (0.185 mL, 2.6 mmol) were
added to 15 mL of anhydrous THF and incubated under
stirring at room temperature for 24 h. After evaporation of
THF in vacuo, product 3 was dissolved in ethyl acetate and
washed three times with saturated NaHCO3 solution to
remove the excess of mercaptoacetic acid. The product was
further washed three times with saturated NaCl solution and
dried over Na2SO4. After evaporation of the solvent in vacuo,
product 3 was obtained as a white foam: yield 89%; tR 46.5
min; MS 631 (MH+); 13C NMR (CDCl3) 204.47, 173.47, 163.53,
160.48, 155.71, 136.60, 134.57, 128.88, 128.82, 128.42, 128.33,
128.06, 127.95, 80.11, 69.00, 67.11, 57.89, 44.16, 38.63, 33.91,
28.26, 27.51, 24.67.
Boc-Ar g(Z)2Ψ[CO-CH2-NH-CO-CH2]Gly-OH (5). Com-
pound 1 (2 g, 3.5 mmol) was dissolved in 150 mL of anhydrous
ether and mixed with 7 mL of 2 M ammonia in methanol at 0
°C. The reaction was continued with stirring at room tem-
perature. The same volume of 2 M ammonia in methanol was
further added twice at 24 and 48 h, respectively. The reaction
was terminated after 3 days when more than 95% of the
starting material 1 had reacted (based on the HPLC profile).
The solvents were then evaporated. The aminomethyl ketone
product was redissolved in ethyl acetate, washed three times
with saturated NaCl solution, and dried over Na2SO4. After
evaporation of the solvent, product 4 was characterized as MS
556 (MH+) and tR 39.7 min. Product 4 was redissolved in 25
mL of anhydrous THF and treated with succinic anhydride
(3.5 g, 35 mmol) and DIEA (6.1 mL, 35 mmol) for 24 h. The
unreacted anhydride was then converted to succinic acid by
adding 15 mL of saturated NaHCO3 solution and 25 mL of
Boc-(D-P h e)-P r o-OH. The dipeptide was prepared from
Boc-D-Phe-OH and H-Pro-OBzl in solution using a mixed
anhydride procedure, followed by hydrogenation catalyzed by
palladium on activated carbon (Pd 5%).48 The product was
crystallized from ethyl acetate/hexane: mp 171-174 °C; [R]23
D
) -91.0° (c 0.56, acetic acid); tR 31.6 min; 13C NMR (CDCl3)
172.86, 172.66, 155.14, 135.96, 129.34, 128.53, 127.18, 80.17,
59.70, 53.84, 47.33, 39.76, 28.27, 27.66, 24.25.

3114 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Steinmetzer et al.
THF with stirring for 12 h. After evaporation of the solvent
in vacuo, the residue was redissolved in ethyl acetate, washed
three times with saturated NaHCO3 and NaCl solution, and
dried over Na2SO4. The crude product 5 was used for solid-
phase peptide synthesis without further purification; however,
a small amount of product 5 was further purified by flash
chromatography for identification (silica gel 60, benzene/
acetone/acetic acid, 27/10/0.5): tR 43.8 min; MS 656 (MH+);
13C NMR (CDCl3) 201.93, 176.03, 171.43, 163.55, 160.18,
155.55, 154.43, 136.73, 134.51, 128.91, 128.81, 128.41, 128.35,
128.01, 127.83, 80.73, 73.24, 68.94, 66.95, 43.96, 30.59, 29.05,
28.06, 22.33, 21.91, 20.65.
substrate concentration, the inhibitor concentration, the Michae-
lis constant, and the inhibition constant, respectively. Inhibi-
tors 9, 10, 11, 13, and 14 with Ki values of 1-20 × 10-12
M
were analyzed at concentrations of 5 × 10-11 M or higher, and
the progress curve characteristic of slow binding inhibition was
observed. The progress curves were fitted to eq 2 by nonlinear
least-squares fitting in Microsoft Excel:
[P] ) vSt + (v0 - vS)(1 - exp(-kobst))/kobs + c
kobs ) kon[I]/(1 + [S]/Km) + koff
(2)
(3)
Solid -P h a se P ep tid e Syn th esis. The bivalent thrombin
inhibitors were prepared as described elsewhere,49 except the
N-terminal P3, P2, P1, and P1′ residues. When the P1′ residue
was -CH2-S-CH2-CO- or -CH2-NH-CO-CH2-CH2-CO-, the P1-
P1′ building unit 3 or 5 was coupled to the side chain-protected
linker-(the FRE blocking segment)-Pam-resin using conven-
tional Boc chemistry. When the P1′ residue was -CH2-sarcosine,
-CH2-Gly, -CH2-N(acetyl)Gly, -CH2-Pro, or -CH2-Ala, the ap-
propriate amino acid was first attached to the peptide resin
using a standard Boc chemistry protocol. After deprotection
of the Boc group, the resin was washed five times with
dichloromethane (DCM). The peptide resin (∼300 mg) was
then transferred into a polypropylene reaction column (5 mL
volume, Abimed, Germany) and washed twice with 10% DIEA
in DCM and three times with DCM. After being washed three
times with DMF, the resin was treated with a 10-fold excess
of compound 2 and shaken at room temperature in the dark
for 3 days, followed by an extensive six-times-repeated wash
to remove compound 2. The resin was then capped by acetic
anhydride, except for peptides 6 and 11, where the resin was
treated with a 10-fold excess of Z-Cl in the presence of DIEA.
The P3-P2 residues were then coupled as a Boc-dipeptide. It
should be mentioned that the neutralization and washing of
the peptide resin after deprotecting the P1 residue must be
performed as quickly as possible in order to get a reasonable
yield. All peptides were cleaved with HF containing 10%
anisole (v/v) and 10% dimethyl sulfide (v/v) at -5 °C for 1 h.
After evaporation of HF, the peptides were washed twice with
ethyl ether, dissolved in 10% acetic acid, and lyophilized. The
peptides were purified by the preparative HPLC. Only the
fractions with >97% purity based on the analytical HPLC
profiles were collected and lyophilized. All the peptides used
have correct amino acid compositions and molecular masses.
F ibr in Clottin g. The fibrin clotting assay was performed
in 50 mM Tris‚HCl buffer (pH 7.52 at 37 °C) containing 0.1
M NaCl and 0.1% poly(ethylene glycol) 8000 with 9.0 × 10-10
M (0.1 NIH unit/mL) and 0.03% (w/v) of the final concentra-
tions of human R-thrombin and bovine fibrinogen, respectively,
as reported elsewhere.49 The clotting time was plotted against
the inhibitor concentrations, and IC50 was estimated as the
inhibitor concentration required to double the clotting time
relative to the control.
where [P], v0, kobs, c, kon, and koff represent the concentration
of 7-(aminomethyl)coumarin product, the velocity observed in
the absence of the inhibitor at the defined substrate concentra-
tion [S], the apparent first-order rate constant, the displace-
ment of [P] from zero at t ) 0, the association rate constant of
the enzyme and the inhibitor, and the dissociation rate
constant of the enzyme-inhibitor complex, respectively. The
steady-state velocity vS estimated from eq 2 was then incor-
porated into eq 1 to estimate the Ki values of the inhibitors by
nonlinear least-squares fitting as described above. The as-
sociation rate constant kon of the inhibitors was estimated from
eq 3. Inhibitors 12 and 16 with Ki values below 1 × 10-12
M
were analyzed at concentrations of 1 × 10-11 M or higher, and
the progress curve characteristic of slow tight binding inhibi-
tion was observed. The progress curves were fitted to the slow
tight binding inhibition eq 4 by nonlinear least-squares fitting
in Microsoft Excel:
[P] )
vSt + ((v0 - vS)(1 - γ)/γk) ln((1 - γe-kt)/(1 - γ)) + c (4)
vS ) v0([E]T - [I]T - Ki′ + Q)/(2[E]T)
(5)
where γ ) (Ki′ + [E]T + [I]T - Q)/(Ki′ + [E]T + [I]T + Q); Q )
((Ki′ + [E]T + [I]T)2 - 4[E]T[I]T)1/2, and Ki′ ) Ki(1 + [S]/Km),
where k is the observed second-order rate constant for the
interaction of the inhibitor with thrombin and v0 is the velocity
observed in the absence of the inhibitor at the defined
substrate concentration [S].15,51 The steady-state velocity vS
was then used to calculate the Ki values according to eq 5 using
nonlinear least-squares fitting in Microsoft Excel, where [E]T
and [I]T are the total enzyme and the total inhibitor concentra-
tions, respectively.15 For the regression, the observed steady-
state velocities were weighted according to the reciprocal of
their value. The hyperbolic inhibition of inhibitors 15 and 17
was analyzed using eq 6 and nonlinear least-squares fitting
in Microsoft Excel as described elsewhere:47
vS ) Vmax([S]/Km + 1.6[I][S]Kr/KmKi)/(1 + [S]/Km
+
(1 + 1/Kr)[I]Kr/Ki + 1.6[I][S]Kr/KmKi) (6)
Am id olytic Assa ys. Inhibition of the amidolytic activity
of human R-thrombin was measured on a Hitachi F2000
fluorescence spectrophotometer (and Shimadzu RF-5001PC
spectrofluorometer for inhibitor 16) (λex ) 383 nm, λem ) 455
nm) which was calibrated with AMC. The assay was performed
with tosyl-Gly-Pro-Arg-AMC as a fluorogenic substrate in 50
mM Tris‚HCl buffer (pH 7.8) containing 0.1 M NaCl and 0.1%
poly(ethylene glycol) 8000 at 23 °C.49 The final concentrations
of the substrate and human R-thrombin were 2-40 × 10-6 and
1.0 × 10-11 M, respectively. The final inhibitor concentrations
varied depending on the Ki values (Table 2). Inhibitors 6-8
with Ki values of 2-10 × 10-11 M were analyzed at concentra-
tions of 1 × 10-10 M or higher and showed classical competitive
inhibition. The linear steady-state velocities (vS) were analyzed
by using the method (eq 1) described by Segel50 and a nonlinear
squares fitting in UltraFit (Biosoft) or Microsoft Excel (Mi-
crosoft):
where Kr represents an equilibrium constant between the
enzyme-inhibitor complex formed by simultaneous bindings
of the FRE and the active site inhibitor segments to thrombin
and the enzyme-inhibitor complex formed only through the
interaction of the FRE inhibitor segment with thrombin (the
active site inhibitor segment is exposed to the solvent).
Ack n ow led gm en t. The authors thank J ean Lefeb-
vre and Bernard F. Gibbs for their technical assistance
in peptide synthesis and amino acid analysis and also
M. Arshad Siddiqui for useful discussion.
Refer en ces
(1) Davie, E. W.; Fujikawa, K.; Kisiel, W. The Coagulation Cas-
cade: Initiation, Maintenance, and Regulation. Biochemistry
1991, 30, 10363-10370.
(2) Fenton, J . W. H.; Ofosu, F. A.; Moon, D. G.; Maraganore, J . M.
Thrombin Structure and Function: Why Thrombin is the
Primary Target for Antithrombotics. Blood Coagulation Fibrin-
olysis 1991, 2, 69-75.
vS ) Vmax[S]/(Km(1 + [I]/Ki) + [S])
(1)
where Vmax, [S], [I], Km, and Ki are the maximum velocity, the

Potent Bivalent Thrombin Inhibitors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3115
(3) Vu, T.-K. H.; Hung, D. T.; Wheaton, V. I.; Coughlin, S. R.
Molecular Cloning of a Functional Thrombin Receptor Reveals
a Novel Proteolytic Mechanism of Receptor Activation. Cell 1991,
64, 1057-1068.
(29) Gordon, E. M.; Natarajan, S.; Pluscek, J .; Weller, H. N.; Godfrey,
J . D.; Rom, M. B.; Sabo, E. F.; Engebrecht, J .; Cushman, D. W.
Ketomethyldipeptides II. Effect of Modifications of the R-Ami-
noketone Portion on Inhibition of Angiotensin Converting En-
zyme. Biochem. Biophys. Res. Commun. 1984, 124, 148-155.
(30) Aplin, R. T.; Christiansen, J .; Young, G. T. Amino acids and
Peptides: Part 48. Synthesis of Bradykinyl-chloromethane. Int.
J . Pept. Protein Res. 1983, 21, 555-561.
(31) Kettner, C.; Shaw, E. Inactivation of Trypsin-like Enzymes with
Peptides of Arginine Chloromethyl Ketone. Methods Enzymol.
1981, 80, 826-842.
(32) Fittkau, S. alpha-Aminochloromethyl ketone aus Aminosa¨uren
und Peptiden als Substratanaloge Inhibitoren der Leucinami-
nopeptidase. J . Prakt. Chem. 1973, 315, 1037-1044.
(33) Angliker, H. Synthesis of Tight Binding Inhibitors and Their
Action on the Proprotein-Processing Enzyme Furin. J . Med.
Chem. 1995, 38, 4014-4018.
(34) Fittkau, S.; J ahreis, G.; Peters, K.; Bala´spiri, L. Aminoketone
- Ein Beitrag zur Synthese Optisch Aktiver Derivate von
Aminosa¨uren und Peptiden. J . Prakt. Chem. 1986, 328, 529-
538.
(35) Steinmetzer, T.; Silberring, J .; Mrestani-Claus, C.; Fittkau, S.;
Demuth, H.-U. Peptidyl Ammonium Methyl Ketones as Sub-
strate Analogue Inhibitors of Proline Specific Peptidases. J .
Enzyme Inhib. 1993, 7, 77-85.
(36) Bro¨mme, D.; Barthels, B.; Kirschke, H.; Fittkau, S. Peptide
Methyl Ketones as Reversible Inhibitors of Cysteine Proteinases.
J . Enzyme Inhib. 1989, 3, 13-21.
(37) J ones, D. M.; Atrash, B.; Ryder, H.; Teger-Nilsson, A.-C.;
Gyzander, E.; Szelke, M. Thrombin inhibitors Based on Keton
Derivatives of Arginine and Lysine. J . Enzyme Inhib. 1995, 9,
43-60.
(38) Morrison, J . F.; Walsh, C. T. The Behavior and Significance of
Slow-binding Enzyme Inhibitors. Adv. Enzymol. 1988, 61, 201-
301.
(39) Braun, P. J .; Dennis, S.; Hofsteenge, J .; Stone, S. R. Use of Site-
directed Mutagenesis to Investigate the Basis for the Specificity
of Hirudin. Biochemistry 1988, 27, 6517-6522.
(40) Krstenansky, J . L.; Broersma, R. J .; Owen, T. J .; Payne, M. H.;
Yates, M. T.; Mao, S. J . T. Development of MDL 28,050, a Small
Stable Antithrombin Agent Based on a Functional Domain of
the Leech Protein, Hirudin. Thromb. Haemostasis 1990, 63,
208-214.
(41) Bode, W.; Turk, D.; Karshikov, A. The Refined 1.9-Å X-ray
Crystal Structure of D-Phe-Pro-Arg Chloromethylketone-inhib-
ited Human alpha-Thrombin: Structure Analysis, Overall Struc-
ture, Electrostatic Properties, Detailed Active-site Geometry, and
Structure-function Relationships. Protein Sci. 1992, 1, 426-
471.
(42) Zdanov, A.; Wu, S.; DiMaio, J .; Konishi, Y.; Li, Y.; Wu, X.;
Edwards, B. F. P.; Martin, P. D.; Cygler, M. Crystal Structure
of the Complex of Human alpha-Thrombin and Nonhydrolyzable
Bifunctional Inhibitors, Hirutonin-2 and Hirutonin-6. Proteins:
Struct. Funct. Genet. 1993, 17, 252-265.
(43) Rehse, P. H.; Steinmetzer, T.; Li, Y.; Konishi, Y.; Cygler, M.
Crystal Structure of a Peptidyl Pyridinium Methyl Ketone
Inhibitor with Thrombin. Biochemistry 1995, 34, 11537-11544.
(44) Steinmetzer, T.; Konishi, Y. Tripeptidyl Pyridinium Methyl
Ketones as Potent Active Site Inhibitors of Thrombin. Bioorg.
Med. Chem. Lett. 1996, 6, 1677-1682.
(4) Wallis, R. B. Inhibition of Thrombin, a Key Step in Thrombosis.
Drugs Today 1989, 9, 596-605.
(5) Hauptmann, J .; Markwardt, F. Pharmacologic Aspects of the
Development of Selective Synthetic Thrombin Inhibitors as
Anticoagulants. Semin. Thromb. Hemostasis 1992, 18, 200-217.
(6) Popma, J . J .; Coller, B. S.; Ohman, E. M.; Bittl, J . A.; Weitz, J .;
Kuntz, R. E.; Leon, M. B. Antithrombotic Therapy in Patients
Undergoing Coronary Angioplasty. Chest 1995, 108, 486S-501S.
(7) Granger, C. B.; Califf, R. M.; Topol, E. J . Thrombolytic Therapy
for Acute Myocardial Infarction. A Review. Drugs 1992, 3, 293-
325.
(8) Becker, R. C. Reocclusion Following Successful Thrombolysis.
Cardiology 1993, 82, 265-273.
(9) Sherry, S. Appraisal of Various Thrombolytic Agents in the
Treatment of Acute Myocardial Infarction. Am. J . Med. 1987,
83, 31-46.
(10) Gold, H. K.; Leinbach, R. C.; Garabedian, S. M.; Yasuda, T.;
J ohns, J . A.; Grossbard, E. B.; Palacios, I.; Collen, D. Acute
Coronary Reocclusion after Thrombolysis with Recombinant
Tissue-type Plasminogen Activator: Prevention by a Mainte-
nance Infusion. Circulation 1987, 73, 347-352.
(11) Hogg, P. J .; J ackson, C. M. Fibrin Monomer Protects Thrombin
from Inactivation by Heparin-antithrombin III: Implications for
Heparin Efficacy. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 3619-
3623.
(12) Weitz, J . I.; Hudoba, M.; Massel, D.; Maraganore, J .; Hirsh, J .
Clot-bound Thrombin is Protected from Inhibition by Heparin-
Antithrombin III but is Susceptible to Inactivation by Anti-
thrombin III-independent Inhibitors. J . Clin. Invest. 1990, 86,
385-391.
(13) Kimball, D. S. Thrombin Active Site Inhibitors. Curr. Pharm.
Des. 1995, 1, 441-468.
(14) Markwardt, F. Hirudin is an Inhibitor of Thrombin. Methods
Enzymol. 1970, 19, 924-932.
(15) Stone, R. S.; Hofsteenge, J . Kinetics of the Inhibition of Thrombin
by Hirudin. Biochemistry 1986, 25, 4622-4628.
(16) Stubbs, M. T.; Bode, W. A Player of Many Parts: The Spotlight
Falls on Thrombin Structure. Thromb. Res. 1993, 69, 1-58.
(17) Kaminski, M.; Siebenlist, K. R.; Mosesson, M. W. Evidence for
Thrombin Enhancement of Fibrin Polymerization that is Inde-
pendent of 1st Catalytic Activity. J . Lab. Clin. Med. 1991, 117,
218-225.
(18) Mirshani, M.; Soria, J .; Soria, C.; Faivre, R.; Lu, H.; Courtney,
M.; Roitsch, C.; Tripier, D.; Caen, J . P. Evaluation of the
Inhibition by Heparin and Hirudin of Coagulation Activation
During r-tPA-induced Thrombolysis. Blood 1989, 74, 1025-1030.
(19) Ru¨bsamen, K.; Eschenfelder, V. Reocclusion after Thromboly-
sis: A Problem Solved by Hirudin? Blood Coagulation Fibrin-
olysis 1991, 2, 97-100.
(20) Maraganore, J . M.; Bourdon, P.; J ablonski, J .; Ramachandran,
K. L.; Fenton, J . W. II. Design and Characterization of Hiru-
logs: A Novel Class of Bivalent Peptide Inhibitors of Thrombin.
Biochemistry 1990, 29, 7095-7101.
(21) DiMaio, J .; Gibbs, B.; Munn, D.; Lefebvre, J .; Ni, F.; Konishi, Y.
Bifunctional Thrombin Inhibitors Based on the Sequence of
Hirudin45-65. J . Biol. Chem. 1990, 265, 21698-21703.
(22) Wittig, J . I.; Bourdon, P.; Maraganore, J . M.; Fenton, J . W. II.
Hirulog-1 and -B2 Thrombin Specificity. Biochem. J . 1992, 287,
663-664.
(23) Kline, T.; Hammond, C.; Bourdon, P.; Maraganore, J . M. Hirulog
Peptides with Scissile Bond Replacements Resistant to Throm-
bin Cleavage. Biochem. Biophys. Res. Commun. 1991, 177,
1049-1055.
(24) DiMaio, J .; Ni, F.; Gibbs, B.; Konishi, Y. A New Class of Potent
Thrombin Inhibitors that Incorporates a Scissile Pseudopeptide
Bond. FEBS Lett. 1991, 282, 47-52.
(25) DiMaio, J .; Gibbs, B.; Lefebvre, J .; Konishi, Y.; Munn, D.; Yue,
S. Y.; Hornberger, W. Synthesis of a Homologous Series of
Ketomethylene Arginyl Pseudopeptides and Application to Low
Molecular Weight Hirudin-like Thrombin Inhibitors. J . Med.
Chem. 1992, 35, 3331-3341.
(26) Krishnan, R.; Tulinsky, A.; Vlasuk, G. P.; Pearson, D.; Vallar,
P.; Bergum, P.; Brunck, T. K.; Ripka, W. C. Synthesis, Structure,
and Structure-activity Relationships of Divalent Thrombin
Inhibitors Containing an R-keto-amide Transition-state Mimetic.
Protein Sci. 1996, 5, 422-433.
(45) Stone, S. R.; Betz, A.; Hofsteenge, J . Mechanistic Studies on
Thrombin Catalysis. Biochemistry 1991, 30, 9841-9848.
(46) Fersht, A. R. The Hydrogen Bond in Molecular Recognition.
Trends Biochem. Sci. 1987, 12, 301-304.
(47) Szewczuk, Z.; Gibbs, B. F.; Yue, S. Y.; Purisima, E.; Zdanov, A.;
Cygler, M.; Konishi, Y. Design of
a Linker for Trivalent
Thrombin Inhibitors: Interaction of the Main Chain of the
Linker with Thrombin. Biochemistry 1993, 32, 3396-3404.
(48) Bodanszky, M.; Bodanszky, A. The Practice of Peptide Synthesis;
Springer-Verlag: Berlin-Heidelberg, 1984.
(49) Szewczuk, Z.; Gibbs, B. F.; Yue, S. Y.; Purisima, E. O.; Konishi,
Y. Conformationally Restricted Thrombin Inhibitors Resistant
to Proteolytic Digestion. Biochemistry 1992, 31, 9132-9140.
(50) Segel, I. H. Enzyme Kinetics: Behavior and Analysis of Rapid
Equilibrium and Steady-State Enzyme Systems; J ohn Wiley &
Sons: New York, 1975; pp 100-160.
(27) Meyer, R. F.; Essenburg, A. D.; Smith, R. D.; Kaplan, H. R.
Angiotensin Converting Enzyme Inhibitors: Modifications of a
Tripeptide Analogue. J . Med. Chem. 1982, 25, 996-999.
(28) Natarajan, S.; Gordon, E. M.; Sabo, E. F.; Godfrey, J . D.; Weller,
H. N.; Pluscek, J .; Rom, M. B.; Cushman, D. W. Ketomethyl-
dipeptides I. A New Class of Angiotensin Converting Enzyme
Inhibitors. Biochem. Biophys. Res. Commun. 1984, 124, 141-
147.
(51) Morrison, J . F.; Stone, S. R. Approaches to the Study and
Analysis of the Inhibition of Enzymes by Slow- and Tight-binding
Inhibitors. Comments Mol. Cell. Biophys. 1985, 2, 347-368.
J M9807297