Synthesis and structure-activity relationships of potent thrombin inhibitors: Piperazides of 3-amidinophenylalanine

Article abstract of DOI:10.1021/jm960668h

Thrombin is the key enzyme in the blood coagulation system, and inhibitors of its proteolytic activity are of therapeutic interest since they are potential anticoagulants. The most potent inhibitor of the benzamidine type is N(α)-[(2-naphthylsulfonyl)glycyl]-4-amidinophenylalanylpiperidide (NAPAP). However, NAPAP and other benzamidine derivatives do not show favorable pharmacological properties; above all, they have very low systemic bioavailability after oral administration. The goal of designing new compounds was to obtain potent inhibitors with improved pharmacokinetic properties. Piperazide derivatives of 3-amidinophenylalanine as the key building block were synthesized. The piperazine moiety opened the possibility to introduce quite different substituents on the second nitrogen using common synthetic procedures. Some of the newly synthesized compounds are potent inhibitors of thrombin and offer an approach to study structure-function relationships for inhibition of thrombin and related enzymes and for the improvement of their pharmacokinetic properties.

Full text of DOI:10.1021/jm960668h

J . Med. Chem. 1997, 40, 3091-3099
3091
Syn th esis a n d Str u ctu r e-Activity Rela tion sh ip s of P oten t Th r om bin In h ibitor s:
P ip er a zid es of 3-Am id in op h en yla la n in e
J o¨rg Stu¨rzebecher,* Dagmar Prasa, J o¨rg Hauptmann, Helmut Vieweg,† and Peter Wikstro¨m†
Klinikum der Friedrich-Schiller-Universita¨t J ena, Zentrum fu¨r Vaskula¨re Biologie und Medizin, Nordha¨user Str. 78, D-99089
Erfurt, Germany, and Pentapharm Ltd., CH-4002 Basel, Switzerland
Received September 25, 1996X
Thrombin is the key enzyme in the blood coagulation system, and inhibitors of its proteolytic
activity are of therapeutic interest since they are potential anticoagulants. The most potent
inhibitor of the benzamidine type is NR-[(2-naphthylsulfonyl)glycyl]-4-amidinophenylalanylpi-
peridide (NAPAP). However, NAPAP and other benzamidine derivatives do not show favorable
pharmacological properties; above all, they have very low systemic bioavailability after oral
administration. The goal of designing new compounds was to obtain potent inhibitors with
improved pharmacokinetic properties. Piperazide derivatives of 3-amidinophenylalanine as
the key building block were synthesized. The piperazine moiety opened the possibility to
introduce quite different substituents on the second nitrogen using common synthetic
procedures. Some of the newly synthesized compounds are potent inhibitors of thrombin and
offer an approach to study structure-function relationships for inhibition of thrombin and
related enzymes and for the improvement of their pharmacokinetic properties.
In tr od u ction
rivatives.15,16 However, NAPAP, its derivatives, and
other benzamidine-derived inhibitors do not yet fulfill
the pharmacological requirements for an ideal antico-
agulant.17,18 In the case of NAPAP, most of the struc-
tural variations resulted in a drastic loss of antithrom-
bin activity;19 however, reduction of the basicity of the
amidino moiety and introduction of a carboxyl group led
to improved pharmacokinetics.9 From the X-ray crystal
structure it was deduced that in the NAPAP-thrombin
complex there is no space available for additional
substituents.20,21 However, X-ray crystal structure
analysis led to other promising structures, such as NR-
tosylated 3-amidinophenylalanylpiperidide (3-TAPAP).14
The thrombin-3-TAPAP complex indicated more space
available for substituents both at the toluene ring and
the piperidine moiety.22,23 A considerable number of
new derivatives of 3-amidinophenylalanine were syn-
thesized.24,25 In this series, the NR-naphthylsulfony-
lated 4-methylpiperidide shows high antithrombin ac-
tivity equal to that of NAPAP (Ki 6.2 and 6.0 nM,
respectively). Although substituents could be intro-
duced into the molecule which should theoretically
improve the absorption after oral application and de-
crease elimination from the circulation without a dra-
matic loss of activity, low absorption rates were found
for these compounds.9 Therefore, we looked for alterna-
tive derivatives.
The currently used anticoagulants meet only some of
the criteria for an ideal antithrombotic drug.1 It has
been shown that direct thrombin inhibitors have a
number of advantages over the therapeutically used
anticoagulants.2-4 From this point of view, both the
naturally occurring thrombin inhibitor hirudin5 and the
synthetic peptide hirulog6 are promising anticoagulants.
However, due to their peptide structure, these agents
can only be administered by the parenteral route. In
contrast to large peptides and recombinant proteins,
synthetic, low molecular weight inhibitors may be
absorbed after oral administration. Thus, a synthetic
inhibitor selective toward thrombin and with an ap-
propriate half-life after oral administration could be the
ideal anticoagulant. Absorption after oral administra-
tion has been reported for peptide aldehydes,7 boronic
acid derivatives,8 and benzamidine-derived inhibitors;9
however, systemic bioavailability is very low.10 Rapid
elimination and low oral bioavailability of several ben-
zamidine- and arginine-derived thrombin inhibitors
were shown to be caused by hepatic uptake and biliary
excretion.11,12 For the first time, relatively high oral
bioavailability was shown for agmatine derivatives of
D-Phe-Pro-Arg type inhibitors containing a N-terminal
acetic acid.13 However, anticoagulant-active compounds
derived from benzamidine with improved pharmacoki-
netic properties have not yet been described.
The X-ray crystal structure of the complex of throm-
bin with the NR-naphthylsulfonylated 3-amidinopheny-
lalanyl-4-methylpiperidide shows clearly that the 4-m-
ethyl group located on the piperidine ring does not
optimally fill the space in the cavity formed by the side
chains of His57, Tyr60A, and Trp 60D in thrombin and
by the piperidine moiety,26 so that larger substituents
should be tolerated. However, the piperidine moiety
does not allow introduction of a wide variety of substit-
uents. In the 3-amidinophenylalanine-type inhibitors
we exchanged the piperidine moiety for a piperazine
ring to facilitate the incorporation of different substit-
uents on the second nitrogen of the piperazine moiety.27
We have designed and systematically synthesized
several types of low molecular weight inhibitors derived
from benzamidine-containing amino acids. Amidinophe-
nylalanine, in which the benzamidine moiety imitates
the guanidinoalkyl side chain of arginine, was used as
a key structural element for the development of inhibi-
tors.14 The most potent thrombin inhibitors of the
benzamidine type are NR-[(2-naphthylsulfonyl)glycyl]-
4-amidinophenylalanylpiperidide (NAPAP) and its de-
† Pentapharm Ltd.
X Abstract published in Advance ACS Abstracts, August 1, 1997.
S0022-2623(96)00668-1 CCC: $14.00 © 1997 American Chemical Society

3092 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19
Stu¨rzebecher et al.
Ta ble 1. Inhibition of Thrombin, Factor Xa, Plasmin, and Trypsin by N′-Substituted Piperazides of NR-2-Naphthylsulfonylated
3-Amidinophenylalanine
Ki (µM)a
no.
R1
thrombin
F Xa
plasmin
trypsin
1
2
3
4
5
6
7
8
H
CH3
C2H4OH
C6H5
CO-H
CO-CH3
0.50 (0.09)
0.036 (0.008)
0.058 (0.006)
0.52 (0.05)
0.037 (0.010)
0.023 (0.004)
1.04 (0.10)
6.5 (0.6)
36 (9)
7.6 (0.7)
0.46 (0.11)
0.019 (0.005)
0.67 (0.06)
17 (4)
0.0088 (0.0021)
0.51 (0.10)
0.0033 (0.0008)
0.044 (0.004)
0.28 (0.04)
21 (5)
30 (6)
35 (4)
27 (7)
19 (2)
48 (4)
80 (11)
19 (1)
58 (4)
16 (1)
7.3 (0.5)
77 (18)
54 (6)
43 (6)
24 (2)
12 (2)
22 (1)
24 (1)
22 (6)
18 (2)
24 (3)
24 (8)
25 (6)
38 (1)
41 (9)
7.9 (0.7)
10.2 (1.0)
55 (7)
42 (2)
0.22 (0.02)
1.3 (0.2)
1.1 (0.3)
2.6 (0.6)
6.3 (1.3)
8.4 (1.1)
8.6 (2.1)
2.6 (0.1)
10.2 (2.3)
7.2 (0.4)
3.0 (0.3)
2.8 (0.3)
2.5 (0.4)
2.3 (0.1)
6.1 (1.2)
14 (3)
0.027 (0.004)
0.22 (0.03)
0.14 (0.03)
0.049 (0.012)
0.064 (0.013)
0.072 (0.019)
0.14 (0.03)
0.19 (0.05)
0.020 (0.005)
0.022 (0.003)
0.028 (0.006)
0.17 (0.04)
0.22 (0.06)
0.19 (0.02)
0.19 (0.03)
0.21 (0.02)
0.18 (0.02)
0.33 (0.07)
0.22 (0.05)
0.40 (0.13)
0.33 (0.05)
0.14 (0.02)
0.69 (0.04)
CO-n-C3H7
CO-n-C5H11
CO-c-C6H11
CO-C6H5
CO-â-C10H7
CO-OCH3
CO-OC2H5
CO-OC6H5
CO-N(CH3)2
CO-N(C2H5)2
SO2-CH3
SO2-C2H5
SO2-n-C3H7
SO2-n-C4H9
SO2-C6H5
SO2-C6H4-4-CH3
SO2-â-C10H7
Ppd
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
12 (2)
3.5 (0.9)
7.8 (0.1)
11 (3)
19 (4)
15 (1)
0.35 (0.09)
0.041 (0.010)
0.022 (0.002)
0.024 (0.009)
0.065 (0.011)
0.0062 (0.0014)
0.0060 (0.0006)
11 (2)
24b
7.0 (0.4)
5.2 (0.3)
30 (4)
25b
Ppd-4-CH3
NAPAP
a
b
Ki-values were calculated according to Dixon32 using a linear regression program. Mean values ((SD, n ) 3-5). These reference
compounds contain a C-terminal piperidine ring (Ppd) instead of the piperazine moiety.
We now report that this approach led to the design of
novel, potent, and selective thrombin inhibitors.
enzyme with Ki-values in the low micromolar range. In
contrast, antitrypsin activity is enhanced slightly with
enlargement of the alkyl residue (compounds 7-9) but
not with aryl residues (compounds 10 and 11). Surpris-
ingly, with the â-naphthyl derivative 11, the antithrom-
bin activity increases again. A similar trend is found
among the alkoxy- and phenoxycarbonyl derivatives.
Only the methyl carbamate 12 exerts remarkable an-
tithrombin activity, while the inhibitory power de-
creases drastically with introduction of an ethyl or
phenyl residue (compounds 13 and 14). However, this
variation does not influence the high antitrypsin activ-
ity. A remarkable compound is the dimethylcarbamoyl
derivative 15 with further enhanced antithrombin but
lower antitrypsin activity. The variation from the
dimethyl derivative 15 to the diethyl compound 16
resulted in a drastic loss of antithrombin activity.
Substitution of the piperazide nitrogen with alkyl- or
arylsulfonyl moieties is tolerated in several compounds
(Table 1). With a Ki of 3.3 nM the methylsulfonyl
derivative 17 is the most potent thrombin inhibitor
derived from 3-amidinophenylalanine. As shown with
the isosteric acyl, alkoxycarbonyl, and carbamoyl de-
rivatives, antithrombin activity is reduced with increas-
ing length of the alkyl residue. However, the N′-
arylsulfonylated piperazides (21-23) are again potent
Resu lts a n d Discu ssion
Va r ia tion of th e N′-Su bstitu en t of th e P ip er a -
zin e Moiety. In previous work, we synthesized several
derivatives of the NR-(naphthylsulfonyl)-3-amidinophe-
nylalanylpiperidide.24,25 Substitution in the 4-position
of the piperidine moiety was preferred for inhibition of
thrombin. However, only a few 4-piperidide derivatives
could be obtained due to the cumbersome synthesis.
Therefore, the piperazide derivatives 1, 2, and 4 (Table
1) were designed opening the possibility for introducing
different substituents. The N′-methylpiperazide 2 ex-
erts antithrombin activity equal to that of the piperidide
24; however, it is 10 times less potent than the isosteric
4-methylpiperidide 25. Despite this, we have synthe-
sized several derivatives with varying N′-substituents.
Table 1 shows selected 3-amidinophenylalanine deriva-
tives with a C-terminal piperazide motif which inhibit
the enzymes tested in a competitive manner. What was
already evident with the alkyl and aryl derivatives (2-
4) has been confirmed with the acyl derivatives: only
residues with one or two carbons are compatible with
high inhibitory strength toward thrombin. Thus, the
formyl and acetyl derivatives (5 and 6) inhibit the
thrombin inhibitors with Ki-values within the 10-8
M
range.

3-Amidinophenylalanine-Derived Inhibitors
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19 3093
Ch a r t 1. Structure of the Arylsulfonyl Residues R2 a
of 10-100 µM; structural variations do not influence the
anti-F Xa activity. Plasmin as well is inhibited less
strongly than thrombin, the Ki-values being mostly near
10 µM.
Va r ia tion of th e Nr-P r otectin g Gr ou p . The N′-
methylsulfonyl piperazide of 3-amidinophenylalanine
(compound 17, Table 1) was chosen as the most potent
inhibitor to investigate the influence of the NR-arylsul-
fonyl residue on the inhibitory activity. In general,
inhibition of thrombin by the respective derivatives is
more pronounced than inhibition of F Xa, plasmin, and
trypsin (Table 2). However, substitution of the R-ni-
trogen with an arylsulfonyl residue other than the
â-naphthylsulfonyl protecting group (Chart 1) did not
lead to an increase in antithrombin activity (Table 2).
In most cases, it resulted in a drastic loss of inhibitory
potency. In contrast, the antitrypsin and antiplasmin
activities are influenced only to a slight extent by the
variations. Only in the case of F Xa did exchange of
âNAPS for other NR-substituents result in some increase
in affinity; however, potent inhibitors of F Xa were not
found.
Ster eoselectivity. First it was proven by X-ray
crystal structure analysis that for trypsin as well as
thrombin20,22 the insertion of the amidinophenylalanine
moiety into the specificity pocket required the L-confor-
mation of the central phenylalanine residue in inhibitors
derived from 3-amidinophenylalanine but the D-confor-
mation in inhibitors of the NAPAP type. As a rule, the
racemates of new compounds were synthesized, as was
the case with the inhibitors listed in Tables 1 and 2.
However, we were able to synthesize also the pure
enantiomers, and a few examples are shown in Table
3. It is evident that the L-derivatives inhibit thrombin
and trypsin much more potently than the corresponding
D-derivatives. The Ki-values differ by a factor between
50 and 100. In the case of F Xa and plasmin, which
are inhibited to a lower extent, the differences are much
smaller. Stereoselectivity of inhibition by the piperazide
derivatives is found to be similar, as was observed for
the derivatives with a C-terminal piperidide structure.25
Selectivity of In h ibition . For proposed in vivo use
as an antithrombotic drug, the inhibitors should be
a
PhS ) phenylsulfonyl; BUPS ) (4-tert-butylphenyl)sulfonyl;
TIPPS ) (2,4,6-triisopropylphenyl)sulfonyl; Mtr ) (4-methoxy-
2,3,6-trimethylphenyl)sulfonyl; RNAPS ) naphthalen-1-ylsulfonyl;
âNAPS ) naphthalen-2-ylsulfonyl; Pmc ) (2,2,5,7,8-pentameth-
ylchroman-6-yl)sulfonyl; CS ) camphor-10-ylsulfonyl.
From the X-ray crystal structure of the complexes of
thrombin with inhibitors of the 3-amidinophenylalanine
type, the C-terminal piperazine moiety is expected to
bind in the S2 site, which in thrombin is bounded by
residues His57, Ser214, Leu99, Tyr60A, and Trp60D
and where medium-sized residues, such as piperidine22
or 4-methylpiperidine,26 are preferred. In trypsin, this
site is much more open, due to the lack of the insertion
loop Tyr60A-Trp60D and should therefore be less
specific and less sensitive to variations. Thus, the
piperazides with a small N′-residue and Ki-values
between 10-8 and 10-9 M should bind optimally into the
S2 site of thrombin. However, compounds 21-23 with
an aromatic moiety at the piperazide nitrogen, which
one would expect to be too large to fit in the S2 cavity,
also inhibit thrombin within the 10-8 M range. Pre-
sumably, the N-terminal naphthylsulfonyl group and
the hydrophobic aryl moiety form a compact structure
that occupies the S2 cavity and the hydrophobic aryl-
binding site.26,28 In contrast, variations at the piper-
azide nitrogen have little influence on antitrypsin
activity, so that the thrombin/trypsin selectivity ratio
is improved with increasing thrombin affinity.
Among the other enzymes, inhibition of F Xa29 by the
piperazides is very low. The Ki-values are in the range
Ta ble 2. Inhibition of Thrombin, Factor Xa, Plasmin, and Trypsin by NR-Substituted N′-(Methylsulfonyl)piperazides of
3-Amidinophenylalanine
Ki (µM)a
no.
R2
thrombin
F Xa
plasmin
trypsin
17
26
27
28b
29
30b
31
32
33
34
âNAPS
BUPS
Pmc
(-)CS
RNAPS
(+)CS
Mtr
PhS
TIPPS
H
0.0033 (0.0008)
0.0043 (0.013)
0.0074 (0.013)
0.012 (0.003)
0.026 (0.006)
0.031 (0.002)
0.032 (0.003)
0.22 (0.06)
22 (1)
6.8 (1.1)
75 (17)
1.1 (0.1)
9.7 (0.5)
6.7 (1.8)
120 (10)
29 (5)
12 (2)
14 (2)
30 (4)
14 (1)
5.0 (1.1)
29 (5)
0.19 (0.02)
0.16 (0.06)
0.76 (0.09)
0.22 (0.08)
0.32 (0.05)
0.44 (0.12)
1.8 (0.5)
1.0 (0.1)
0.98 (0.22)
29 (1)
18 (4)
29 (6)
0.28 (0.08)
13 (2)
7.4 (2.1)
> 1000
6.4 (1.6)
> 1000
a
b
Ki-values were calculated according to Dixon32 using a linear regression program. Mean values ((SD, n ) 3-5). L-Enantiomer.

3094 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19
Stu¨rzebecher et al.
Ta ble 3. Inhibition of Thrombin and Trypsin by the
Enantiomers of Selected NR-2-Naphthylsulfonylated
3-Amidinophenylalanylpiperazides
Ta ble 5. Inhibition of Thrombin and Prolongation of Clotting
Times of Human Plasma by Selected
NR-2-Naphthylsulfonylated 3-Amidinophenylalanylpiperazides
and Established Thrombin Inhibitors
concentration
(IC50, µM)a for doubling of
no.
Ki (nM)a
TT
aPTT
PT
3, L
5, L
2, L
6, L
15, L
17, L
37 (6)
23 (6)
16 (3)
12 (1)
0.14 (0.01)
0.56 (0.13)
1.1 (0.3)
0.7 (0.1)
1.8 (0.5)
0.76 (0.14)
0.72 (0.04)
0.39 (0.11)
0.66 (0.06)
1.0 (0.2)
0.063 (0.002) 0.35 (0.07)
0.22 (0.04) 1.2 (0.4)
0.062 (0.009) 0.28 (0.08)
4.0 (0.7) 0.042 (0.005) 0.39 (0.05)
2.1 (0.5) 0.037 (0.008) 0.26 (0.11)
Ki (µM)a
argatroban 9.6 (0.5) 0.062 (0.003) 0.42 (0.06)
NAPAP
25, L
no. enantiomer
R1
thrombin
trypsin
6.0 (0.6) 0.045 (0.002) 0.50 (0.05)
2.5 (0.6) 0.24 (0.05) 0.7 (0.1)
1.2 (0.2)
2
D,L
L
CH3
CH3
CH3
0.036 (0.008)
0.016 (0.003)
7.1 (1.4)
1.3 (0.2)
0.47 (0.13)
73 (19)
hirudin
0.000027 0.011 (0.001) 0.092 (0.018) 0.27 (0.04)
a
Mean values ((SD, n ) 3-5).
D
3
D,L
L
C2H4OH
C2H4OH
C2H4OH
CO-H
CO-H
CO-H
CO-CH3
CO-CH3
CO-CH3
0.058 (0.006)
0.037 (0.006)
5.6 (1.0)
0.037 (0.010)
0.023 (0.006)
1.6 (0.2)
0.023 (0.004)
0.012 (0.001)
1.9 (0.3)
1.1 (0.3)
0.60 (0.08)
39 (4)
0.22 (0.03)
0.073 (0.015)
11.1 (2.1)
0.14 (0.03)
0.044 (0.002)
17 (3)
fibrinolytic enzymes, such as plasmin and the plasmi-
nogen activators UK and tPA. In a plasma system,
representative derivatives of the series at concentrations
higher than 100 µM only impair the clot lysis induced
by streptokinase, UK, or tPA.31 Of the enzymes tested,
besides thrombin only trypsin is inhibited with a
relatively high inhibitory activity giving Ki-values in the
low micromolar range. High potency in trypsin inhibi-
tion could lead to side effects especially in the case of
oral administration.10,30 Thus, with the exception of
trypsin only, the compounds are highly selective inhibi-
tors of thrombin. Therefore, selectivity with respect to
trypsin has to be improved for this class of inhibitor,
especially if oral administration is provided.
D
5
D,L
L
D
6
D,L
L
D
15
17
D,L
L
CO-N(CH3)2 0.0088 (0.0021)
CO-N(CH3)2 0.0040 (0.0007)
CO-N(CH3)2 0.35 (0.04)
0.17 (0.04)
0.097 (0.004)
8.5 (1.9)
0.19 (0.02)
0.067 (0.011)
13 (2)
D
D,L
L
SO2-CH3
SO2-CH3
SO2-CH3
0.0033 (0.0008)
0.0021 (0.0005)
0.49 (0.01)
D
a
Ki-values were calculated according to Dixon32 using a linear
regression program. Mean values ((SD, n ) 3-5).
An ticoa gu la n t Activity. Some of the new thrombin
inhibitors with a C-terminal piperazine moiety exert
high anticoagulant activity if they have nanomolar Ki-
values (Table 5). Concentrations of 20-60 nM are
needed for doubling TT and 0.1-1 µM for doubling aPTT
and PT. The anticoagulant potency is higher, or at least
in the same range, as that of NAPAP and argatroban.
The compounds are similar in potency to hirudin (e.g.,
compound 17). For doubling the clotting times the
thrombin added (25 nM in the TT assay) or generated
(200-400 nM after extrinsic or intrinsic activation in
PT and aPTT assays32) must be inhibited by nearly 50%.
Tight-binding reversible inhibitors need a Ki lower than
highly selective. It is especially important that they do
not inhibit enzymes of the fibrinolytic system and
protein Ca (APC) which are counterbalancing pro-
thrombotic stimuli.2,10,30 Therefore, we have determined
the inhibitory activity toward different serine protein-
ases of selected compounds of the series presented.
Compared with thrombin, the inhibitory activity toward
the clotting enzymes F Xa and F XIIa is low. The same
is true for plasma and glandular kallikreins. The
anticoagulant-effective enzyme APC is not influenced
remarkably. In most cases the Ki-values are more than
3 orders of magnitude higher than that for inhibition
of thrombin. The same is true of the inhibition of
Ta ble 4. Inhibition of Various Trypsin-like Enzymes by the L-Enantiomers of Selected NR-Substituted
3-Amidinophenylalanylpiperazides
Ki (µM)a
no.
R1
R2
thrombin
F Xa
F XIIa
APC
PK
GK
plasmin
UK
sc-tPA
trypsin
2, L
3, L
5, L
CH3
C2H4OH
CO-H
CO-CH3
CON(CH3)2
SO2-CH3
SO2-CH3
SO2-CH3
âNAPS
âNAPS
âNAPS
âNAPS
âNAPS
âNAPS
Pmc
0.016
0.037
0.023
0.012
0.0040
0.0021
0.0053
0.012
19
22
9.7
34
16
19
41
1.1
280
400
53
200
61
94
39
10.3
12
13.5
15.2
180
33
35
27
3.3
15
7.3
58
97
22
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
30
21
110
140
34
32
28
28
58
49
>1000
180
0.47
0.60
3.3
5.2
4.9
9.0
10.3
14
110
410
0.073
0.044
0.097
0.067
0.32
6, L
15, L
17, L
27, L
28, L
>1000
52
190
>1000
470
20
(-)CS
55
0.22
a
Ki-values were calculated according to Dixon32 using a linear regression program. Standard deviation did not exceed 25%.

3-Amidinophenylalanine-Derived Inhibitors
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19 3095
Ta ble 6. Plasma Level of the Acetylpiperazide 6 (L-Enantiomer) and NAPAP in Anesthetized Rats after iv Administration of 1 mg/kg
and Oral Administration of 50 mg/kg
concentration in plasma (µg/mL)a
compound 6, L
NAPAP
time after
injection (min)
iv
oral
iv
oral
nd
nd
nd
0.10 (2)
nd
0.08 (2)
nd
0.08 (2)
0.09 (2)
0.09 (2)
2
5
15
30
45
60
90
120
180
240
3.55 ( 1.22 (3)b
2.08 ( 0.97 (4)
0.61 ( 0.27 (4)
0.27 ( 0.14 (4)
0.17 ( 0.09 (4)
0.13 ( 0.08 (4)
0.10 ( 0.05 (4)
0.09 ( 0.04 (3)
<0.04 (3)
ndc
2.43 ( 0.10 (3)
0.98 ( 0.08 (3)
0.24 ( 0.02 (3)
0.07 ( 0.02 (3)
0.04 (2)
0.23 ( 0.15 (3)
0.41 ( 0.14 (3)
0.45 ( 0.11 (3)
0.41 ( 0.14 (3)
0.38 ( 0.10 (3)
0.27 ( 0.09 (3)
0.26 (2)
<0.04 (3)
<0.04 (3)
<0.04 (3)
0.30 (2)
<0.04 (3)
<0.04 (3)
nd
<0.04 (3)
a
b
Data are mean ( SD. Number of experiments in parentheses. c Not determined.
20 nM for nearly total blockade of the thrombin active
site.33 Therefore, if the inhibitor reacts exclusively with
the thrombin active site, a concentration of 10-20 nM
is necessary for doubling TT and 100-200 nM for
doubling aPTT and PT. Corresponding effective inhibi-
tor concentrations (IC50) in this range were found with
hirudin and the potent synthetic inhibitors. The me-
thylpiperazide 2 and the methylpiperidide of 3-amidi-
nophenylalanine 25 are exceptions to this rule. Prob-
ably, binding to plasma proteins of these more
hydrophobic compounds is the reason for their higher
effective concentrations.
L-form. The enantiomers were obtained from the race-
mate of the acetyl derivative via acidic deacetylation
using acylase.
Conversion with the corresponding sulfonyl chloride
(R2-SO2Cl) led to the sulfonylated cyanophenylalanine
compound IV with a free carboxyl group. However,
because of the low yield the methyl ester II was
prepared, sulfonylated to III, and hydrolyzed yielding
the sulfonylated cyanophenylalanine IV. Conversion of
IV into V was carried out via a general peptide-coupling
reaction between a reactive intermediate (such as the
corresponding acid chloride or an activated ester with
DCC/HOBt) and a piperazide substituted at the 4-ni-
trogen with different residues (R1). The conversion of
the cyano function into the amidino function was
achieved using one of two known routes. Addition of
hydrogen sulfide in the presence of triethylamine in
pyridine gave a thioamide (VI) which was then S-
methylated with iodomethane to give the thioimidate
hydrochloride V which was further reacted with am-
monium acetate in methanol to give the crude hydroio-
dide of the end product IX. The end product IX was
generally purified by chromatography over silica gel or
Sephadex LH-20 to give homogeneous products followed
by ion exchange to give the more soluble hydrochlorides
of the end product IX. Another route to the amidino
compounds was achieved by converting the correspond-
ing cyano compound VI to the corresponding imidate
hydrochloride VIII followed by treatment with ethanolic
ammonia solution to give the end products IX. The
syntheses of esters of derivatives containing a carboxyl
group were achieved by esterification of the amidino
compounds with the corresponding alcohols in the
presence of gaseous hydrogen chloride or p-toluene-
sulfonic acid. Preparation of the NR-(2-naphthylsulfo-
nyl)-3-amidino-L-phenylalanyl-4-(2-hydroxyethyl)piper-
azide (3) and the NR-(2-naphthylsulfonyl)-3-amidino-L-
phenylalanyl-4-(methylsulfonyl)piperazide hydrochloride
(17) is exemplified in detail in the Experimental Section.
P h a r m a cok in etic Stu d ies. The plasma levels of
selected inhibitors were followed after intravenous
administration at a dose of 1 mg/kg of body weight to
anesthetized rats. Compound 6 was studied in com-
parison to NAPAP. Table 6 shows the plasma level over
a period of 2 h. The plasma level of NAPAP rapidly
declines reaching the detection limit 60 min after
administration. There is almost no distinction between
distribution and elimination phases. For 6, there are
somewhat higher values in the initial distribution phase
followed by a slower decline of the plasma level. For
comparison, on a molar basis, 6 shows plasma levels 5
times that of NAPAP after 30 and 45 min, whereas the
difference in dose was 1.26-fold. The time course of the
plasma level derived from a clotting time assay (experi-
ments in a separate group of animals, data not shown)
is in accord with those of the HPLC assay, although the
numerical values are lower by a factor of about 1.4.
HPLC did not give evidence of metabolites in plasma.
The results in rats on NAPAP confirm the finding of a
very short half-life first found in rabbits.17 The present
results show a comparatively prolonged time course of
plasma levels of 6 in rats. After oral administration, 6
reaches higher plasma levels than NAPAP. More
detailed studies are under way on the oral bioavailabil-
ity and metabolism of 6 and other compounds of this
series.
Ch em istr y
Con clu sion s
The syntheses of NR-sulfonylated 3-amidinophenyl-
alanylpiperazides with the general structure IX
(Scheme 1) were achieved via the corresponding cyano
compounds V which were synthesized in the following
way.27 Starting from N-acetyl-D,L-3-cyanophenylala-
nine, 3-cyanophenylalanine (I) was synthesized and
used for further syntheses in either the racemic, D-, or
The structure-activity relationships presented for the
new thrombin inhibitors are in agreement with the
predictions from the X-ray crystal structure analysis of
thrombin-inhibitor complexes. From the structure of
the NAPAP-thrombin complex it was deduced that the
NAPAP molecule is not a suitable candidate for deriva-

3096 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19
Stu¨rzebecher et al.
Sch em e 1. Syntheses of NR-Arylsulfonylated 3-Amidinophenylalanylpiperazides
tizations since NAPAP fits so ideally to thrombin that
there is no further space for additional substituents.20,21
Therefore, we used the NR-tosylated 3-amidinopheny-
lalanylpiperidide (3-TAPAP) as a prototype for fur-
ther synthesis because the corresponding inhibitor
complex indicates more space for substituents.22 Indeed,
within a first series of compounds, we found with the
NR-naphthylsulfonylated 4-methylpiperidide of 3-ami-
dinophenylalanine 25 a thrombin inhibitor whose an-
tithrombin activity is equal to that of NAPAP (Ki 6.2
and 6.0 nM, respectively);27 however, the compounds do
not show improved pharmacokinetics.9 With the pip-
erazide derivatives described in this paper, potent
thrombin inhibitors with sufficient selectivity were
found. Some of the compounds possess remarkable
anticoagulant activity. As is shown with the acetylpip-
erazide 6, the piperazides are eliminated more slowly
than NAPAP and 3-amidinophenylalanine derivatives
containing a C-terminal piperidide. Furthermore, first
results indicate enhanced enteral absorption. There-
fore, we are engaged in establishing structure-function
relationships of the pharmacokinetic properties of de-
rivatives containing quite different substituents of the
second nitrogen of the piperazine ring in order to
improve the oral bioavailability. Results will be re-
ported in due course.
Exp er im en ta l Section
Ch em istr y. Reagent-grade solvents were used without
further purification. Evaporation refers to removal of solvent
by use of a Bu¨chi rotary evaporator at 40-50 °C in vacuo. All
organic extracts were dried over Na2SO4. TLC plates coated
with silica gel 60 F254 used were from E. Merck AG (Darmstadt,
Germany); detection was by UV (254 nm). Normal-phase silica
gel used for flash chromatography was Kieselgel-60 (230-400
mesh); ion-exchange resin used was Amberlite IRA-420 (Cl-
form, 16-50 mesh); both were supplied by Fluka (Buchs,
Switzerland). Optical rotations [R]20 were determined with
D
a Perkin-Elmer 243B polarimeter, c in g/100 mL. All 1H NMR
spectra were recorded on a Bruker AMX 600 spectrometer.
Fast atom bombardment mass spectra (FAB/MS; Finnigan
4516) and TOF/MS (Voyager RP) were recorded.
N r-(2-Na p h t h ylsu lfon yl)-3-c ya n o -L -p h e n y la la n in e
Meth yl Ester (III in Sch em e 1, R2 ) â-n a p h th yl) (3a ).
3-Cyano-L-phenylalanine methyl ester hydrochloride (24.1 g,
0.1 mol) was suspended in 200 mL of absolute dioxane, 20.6 g
(0.204 mol) of NMM was added under stirring, and a solution
of 23.6 g (0.104 mol) of 2-naphthalenesulfonyl chloride in 200
mL of ethyl acetate was added dropwise. The mixture was
stirred for 16 h at room temperature, the precipitated N-

3-Amidinophenylalanine-Derived Inhibitors
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19 3097
methylmorpholine hydrochloride was filtered off, and the
solvent was evaporated under reduced pressure. The formed
precipitate crystallized from methanol/water: yield 35.6 g
(90%); 1H NMR (DMSO) δ 8.60 (1 H, d, NH), 8.15 (1 H, m,
phenyl-H), 8.05-7.20 (10 H, m, phenyl-H, naphthyl-H), 4.11
(1 H, m, CR-H), 3.31 (s, 3 H, CH3), 3.00/2.75 (dd, 2 H, CH2-
phenyl); TOF/MS m/ z (M + 2H)+ 396.73 (C21H18N2O4S, Mr )
394.45); mp 118 °C; [R]D ) +11.8°, c ) 1 (MeOH). Anal.
(C21H18N2O4S) C, H, N.
piperazide-H); FAB/MS m/ z (M + H)+ 510.4 (C26H31N5O4S
formula 509, Mr ) 509.633); mp 219 °C; [R]D ) +15.4°, c ) 1
(MeOH). Anal. (C26H33N5O4SCl2‚1.5 H2O‚0.75NH4Cl) C, H,
N, O, S, Cl.
Nr-(2-Naph th ylsu lfon yl)-3-cyan o-L-ph en ylalan yl-4-(m e-
th ylsu lfon yl)p ip er a zid e (V in Sch em e 1, R2 ) â-n a p h th yl,
R1 ) SO2CH3) (17a ). Compound 3b (5.4 g, 14.2 mmol), 2.32
g (15.6 mmol) of HOBt, and 3.22 g (15.6 mmol) of DCC were
dissolved in 70 mL of DMF and stirred for 1 h. A mixture of
3.12 g (15.6 mmol) of 1-(methylsulfonyl)piperazide hydrochlo-
ride and 1.72 mL (15.6 mmol) of NMM in 30 mL of DMF was
added, and the mixture was stirred for another 20 h at room
temperature. The precipitated dicyclohexylurea was filtered
off and the solvent evaporated; the resulting residue was
chromatographed over silica gel with chloroform as the elu-
Nr-(2-Na p h th ylsu lfon yl)-3-cya n o-L-p h en yla la n in e (IV
in Sch em e 1, R2 ) â-n a p h th yl) (3b). Methyl ester 3a (15
g) was allowed to reflux for 2 h in a mixture of 75 mL each of
1 N HCl and acetic acid. Upon cooling to room temperature,
water was added until a turbid solution resulted and crystal-
lization took place. The crystals were filtered off, washed with
30% acetic acid in water and vacuum-dried over KOH: yield
ent: yield 6.4 g (85%); 1
H NMR (DMSO) δ 8.35 (1 H, bs, NH),
1
8.15 (1 H, m, phenyl-H), 8.09-7.32 (10 H, m, phenyl-H,
naphthyl-H), 4.50 (1 H, m, CR-H), 3.00/2.85 (2 H, dd, CH2-
phenyl), 3.50-3.35 (3 H, m, piperazide-H), 3.15 (1 H, piper-
azide-H), 2.75-2.68 (3 H, m, piperazide-H), 2.72 (s, 3 H, SO2
CH3), 2.32 (1 H, m, piperazide-H); TOF/MS m/ z (M + 2H)+
528.08 (C25H26N4O5S2, Mr ) 526.64); mp 157 °C; [R]D ) +47.3°,
c ) 1 (MeOH). Anal. (C25H26N4O5S2) C, H, N.
13.3 g (92%); H NMR (DMSO) δ 12.90 (1 H, s, COOH), 8.30
(1 H, d, NH), 8.17 (m, 1 H, phenyl-H), 8.05-7.18 (10 H, m,
phenyl-H, naphthyl-H), 4.00 (m, 1 H, CR-H), 3.02/2.75 (2 H,
dd, CH2-phenyl); TOF/MS m/ z (M + 2H)+ 382.52 (C20H16N2O4S,
Mr ) 380.43); mp 192-193 °C; [R]D ) +11.9°, c ) 1 (MeOH).
Anal. (C20H16N2O4S) C, H, N.
Nr-(2-Na p h th ylsu lfon yl)-3-cya n o-L-p h en yla la n yl-4-(2-
h yd r oxyeth yl)p ip er a zid e (V in Sch em e 1, R2 ) â-n a p h -
th yl, R1 ) CH2CH2OH) (3c). NR-(2-Naphthylsulfonyl)-3-
cyano-L-phenylalanyl chloride was prepared from compound
3b (5 g) by addition to 15 g of thionyl chloride and refluxed
for 1 h. The corresponding acid chloride was precipitated by
the addition of petroleum ether; 5.1 g (12.8 mmol) of the acid
chloride was dissolved in 75 mL of THF and added dropwise
over 15 min to a solution of 3.5 g (26.8 mmol) of 1-(2-
hydroxyethyl)piperazide in 45 mL of THF. It was stirred for
1 h and filtered and the solvent evaporated under reduced
pressure. The residue was dissolved in 30 mL of methanol,
water was added until turbidity set in, and the mixture was
allowed to stand overnight. The free base of 3c precipitated
out as an oil. The solvent was decanted and the oil dissolved
in 150 mL of ethyl acetate. The organic phase was washed
with saturated NaCl solution and dried over MgSO4, and the
solution was concentrated to ca. 100 mL and acidified by the
addition of 2 N HCl in ethyl acetate. The hydrochloride of 3c
was precipitated by the addition of diethyl ether, filtered off,
Nr-(2-Na p h th ylsu lfon yl)-3-a m id in o-L-p h en yla la n yl-4-
(m eth ylsu lfon yl)piper azide Hydr och lor ide (IX in Sch em e
1, R1 ) â-n a p h th yl, R2 ) SO2CH3) (17). The cyano com-
pound 17a (5.8 g) was dissolved in 70 mL of pyridine, 1 mL of
TEA was added, and the solution was saturated by a 10-min
introduction of gaseous H2S. The mixture was allowed to
stand at room temperature for 2 days. The solvent was
evaporated under reduced pressure and the residue dissolved
in ethyl acetate, whereby the thioamide crystallized out: yield
5.9 g (95%). The obtained thioamide (5.9 g, 10.5 mmol) was
dissolved in 8 mL of warm DMF and diluted with 100 mL of
acetone, 14.2 g (0.1 mol) of methyl iodide was added, and the
mixture was kept in the dark for 20 h at room temperature.
The product was precipitated by the addition of diethyl ether,
filtered off, and dried: yield 6.5 g (88%). A 9.3-mmol portion
of methylthioimidate hydroiodide was dissolved in 220 mL of
methanol, 1.2 g (15.5 mmol) of ammonium acetate was added,
and the mixture was heated for 3 h at 60 °C in a water bath.
The solvent was evaporated under reduced pressure; the
residue (amidine hydroiodide) was dissolved in methanol and
converted to the corresponding hydrochloride by passing the
solution over Amberlite IRA-410 (Cl- form). The resulting
hydrochloride salt was precipitated from methanol by the
addition of diethyl ether. The crude product was chromato-
1
and dried: yield 4.95 g (73%); H NMR (DMSO) δ 8.35 (1 H,
d, NH), 8.15 (1 H, m, phenyl-H), 8.09-7.32 (10 H, m, phenyl-
H, naphthyl-H), 5.35 (1 H, bs, OH), 4.58 (1 H, m, CR-H), 4.35-
4.00 (2 H, m, ethyl-H), 3.75 (2 H, m, ethyl-H), 3.50-3.25 (3 H,
m, piperazide-H), 3.10 (1 H, m, piperazide-H), 2.95/2.75 (2 H,
dd, CH2-phenyl), 2.75-2.65 (3 H, m, piperazide-H), 2.30 (1 H,
m, piperazide-H); TOF/MS m/ z (M + 2H)+ 494.1 (C26H28N4O4S,
Mr ) 492.60); mp 155 °C; [R]D ) -5.4°, c ) 1 (MeOH). Anal.
(C26H28N4O4S) C, H, N.
1
graphed over Sephadex LH-20: yield 3.78 g (71%); H NMR
(DMSO) δ 9.25 (1 H, bs, Imin-H), 8.12 (1 H, m, phenyl-H),
8.15-7.32 (10 H, m, phenyl-H, naphthyl-H), 4.58 (1 H, m, CR-
H), 3.45-3.35 (3 H, m, piperazide-H), 3.10 (1 H, m, piperazide-
H), 3.00/2.85 (2 H, dd, CH2-phenyl), 2.75-2.65 (3 H, m,
piperazide-H), 2.72 (s, 3 H, SO2CH3), 2.31 (1 H, m, piperazide-
H); FAB/MS m/ z (M + H)+ 544.6 (C25H29N5O5S2 formula 543,
Mr ) 543.67); mp 220 °C; [R]D ) +70.0°, c ) 1 (MeOH). Anal.
(C25H30N5O5S2Cl‚0.6H2O) C, H, N, O, S, Cl.
Deter m in a tion of In h ibition Con sta n ts. The measure-
ments were carried out on a microplate reader (MR 5000,
Dynatech, Denkendorf, Germany) at 25 °C. The test medium
consisted of 200 µL of Tris buffer (0.05 M; 0.154 M NaCl, 5%
ethanol, pH 8.0), 25 µL of aqueous substrate solution, and 50
µL of enzyme solution. Two concentrations of the substrate
and five concentrations of the inhibitor were used. Three
minutes after the addition of the enzyme, 25 µL of acetic acid
(50%) was added to quench the reaction, and the optical
density was measured at 405 nm. The Ki-values were calcu-
lated according to Dixon34 using a linear regression program.
If not otherwise indicated, the Ki-values are those of the
racemates. The Ki-values reported are means from at least
three determinations.
Nr-(2-Na p h th ylsu lfon yl)-3-a m id in o-L-p h en yla la n yl-4-
(2-h yd r oxyet h yl)p ip er a zid e Dih yd r och lor id e (IX in
Sch em e 1, R2 ) â-n a p h th yl, R1 ) CH2CH2OH) (3). Cyano
compound 3c (4.2 g, 7.95 mmol) was dissolved in a mixture of
22 mL of absolute methanol and 30 mL of absolute dioxane
and cooled to 0 °C, and 15.6 g (0.43 mol) of dried HCl gas was
introduced. The reaction solution was kept at 2 °C for 4 days
and poured into 400 mL of diethyl ether, and the resulting
precipitate was triturated with 100 mL of THF until crystal-
lization took place. The crystals were washed with diethyl
ether and dried; 4.32 g (7.24 mmol) of the methylimidate
dihydrochloride was suspended in 90 mL of methanol, and
ethanolic ammonia solution was added under stirring until a
clear solution was obtained (pH 8.7). The mixture was heated
to 60 °C for 4 h in a water bath. The solvent was evaporated,
the residue was dissolved in 45 mL of absolute ethanol and
filtered, and 3 mL of 2 N HCl in ethyl acetate was added. The
amidine dihydrochloride was precipitated with addition of
diethyl ether and chromatographed over Sephadex LH-20:
yield 2.8 g (73.5%); 1H NMR (DMSO) δ 9.25 (1 H, bs, Imin-H),
8.35 (1 H, d, NH), 8.15 (1 H, m, phenyl-H), 8.09-7.32 (10 H,
m, phenyl-H, naphthyl-H), 4.61 (1 H, m, CR-H), 4.40-3.95 (2
H, m, ethyl-H), 3.72 (2 H, m, ethyl-H), 3.52-3.25 (3 H, m,
piperazide-H), 3.10 (1 H, m, piperazide-H), 2.95/2.75 (2 H, dd,
CH2-phenyl), 2.75-2.65 (3 H, m, piperazide-H), 2.31 (1 H, m,
En zym es a n d Su bstr a tes for Ki Deter m in a tion . The
following enzymes and the respective substrates were used at
the final concentrations indicated: bovine thrombin prepared
according to Walsmann35 (2262 U/mg, final concentration 0.45
U/mL), substrate MeSO2-D-hexahydrotyrosyl-Gly-Arg-pNA (fi-
nal concentrations 0.18 and 0.09 mM); bovine F Xa (5 U/vial,

3098 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19
Stu¨rzebecher et al.
a novel class of bivalent peptide inhibitors of thrombin. Bio-
chemistry 1990, 29, 7095-7101.
0.11 U/mL; Diagnostic Reagents Ltd., Thame, U.K.), substrate
MeSO2-D-Nle-Gly-Arg-pNA (0.36 and 0.18 mM); human F XIIa
(0.2 U/vial, 0.0023 U/mL; RD Laboratorien, Martinsried,
Germany), substrate H-D-hexahydrotyrosyl-Gly-Arg-pNA (0.36
and 0.18 mM); human APC (1 mg/vial, 1.13 µg/mL; Kordia Lab.
Supplies, Leiden, The Netherlands), substrate H-D-Lys(Cbo)-
Pro-Arg-pNA (0.36 and 0.18 mM); human PK (1 U/vial, 0.0088
U/mL; RD Laboratorien, Martinsried, Germany), substrate Bz-
Pro-Phe-Arg-pNA (0.36 and 0.18 mM); GK isolated from
porcine pancreas (500 KU/vial, 1.5 KU/mL; RD Laboratorien,
Martinsried, Germany), substrate H-D-Val-cyclohexylalanyl-
Arg-pNA (0.18 and 0.09 mM); human plasmin (0.67 CTA-U/
mg, 0.06 CTA-U/mL; Behringwerke AG, Marburg, Germany),
substrate Tos-Gly-Pro-Lys-pNA (0.18 and 0.09 mM); human
UK (500 000 U/vial, final concentration 150 U/mL; Ribosep-
harm GmbH, Haan, Germany), substrate Bz-âAla-Gly-Arg-
pNA (0.18 and 0.09 mM); sc-tPA purified from CHO cells36 (4.1
mg/mL, 0.0031 µg/mL), substrate MeSO2-D-hexahydrotyrosyl-
Gly-Arg-pNA (0.54, 0.27, and 0.145 mM); bovine pancreatic
trypsin (42 U/mg, 0.0038 U/mL; Serva, Heidelberg, Germany),
substrate MeSO2-D-hexahydrotyrosyl-Gly-Arg-pNA (0.18 and
0.06 mM). The substrates were supplied by Pentapharm Ltd.,
Basel, Switzerland.
Clottin g Assa ys. For determination of prothrombin time
(PT), 0.1 mL of thromboplastin (Dade, Unterschleissheim,
Germany) and 0.1 mL of inhibitor dissolved in CaCl2 (0.025
M, 5% ethanol) were incubated at 37 °C for 2 min. Coagulation
was started by the addition of 0.1 mL of citrated human
plasma. For determination of activated partial thromboplastin
time (aPTT), citrated human plasma (0.1 mL) was incubated
at 37 °C with 0.1 mL of PTT reagent (Boehringer Mannheim
GmbH, Mannheim, Germany). After 3 min, 0.1 mL of inhibitor
dissolved in CaCl2 (0.025 M, 5% ethanol) was added. For
determination of thrombin time (TT), citrated human plasma
(0.1 mL) was mixed with 0.05 mL of inhibitor dissolved in NaCl
(0.154 mM, 5% ethanol), and coagulation was started by the
addition of 0.05 mL of thrombin (10 U/mL). Clotting times
were determined in duplicate using the coagulometer Throm-
botrack 8 (Immuno GmbH, Heidelberg, Germany). Inhibitor
concentrations required to double the respective clotting times
(IC50) were read from lin/log graphs of clotting times versus
inhibitor concentrations.
(7) Bagdy, D.; Barabas, E.; Szabo, G.; Bajusz, S.; Szell, E. In vivo
anticoagulant and antiplatelet effect of D-Phe-Pro-Arg-H and
D-Me-Phe-Pro-Arg-H. Thromb. Haemostas. 1992, 67, 357-365.
(8) Hussain, M. A.; Knabb, R.; Aungst, B. J .; Kettner, C. Antico-
agulant activity of a peptide boronic acid thrombin inhibitor by
various routes of administration in rats. Peptides 1991, 12,
1153-1154.
(9) Stu¨rzebecher, J .; Prasa, D.; Bretschneider, E.; Bode, W.; Bauer,
M.; Brandstetter, H.; Wikstro¨m, P.; Vieweg, H. New develop-
ments in the field of benzamidine-derived thrombin inhibitors.
In DIC - Pathogenesis, Diagnosis and Therapy of Disseminated
Intravascular Fibrin Formation; Mu¨ller-Berghaus, G., Madlener,
K., Blomba¨ck, M., ten Cate, J . W., Eds.; Excerpta Medica:
Amsterdam, London, New York, Tokyo, 1993; pp 183-190.
(10) Kimball, S. D. Challenges in the development of orally bioavail-
able thrombin active site inhibitors. Blood Coag. Fibrinol. 1995,
6, 511-519.
(11) Hauptmann, J .; Kaiser, B.; Paintz, M.; Markwardt, F. Biliary
excretion of synthetic benzamidine-type thrombin inhibitors in
rabbit and rats. Biomed. Biochim. Acta 1987, 46, 445-453.
(12) Hauptmann, J .; Kaiser, B. In vitro and in vivo comparison of
arginine- and benzamidine-derived highly potent synthetic
thrombin inhibitors. Pharmazie 1991, 46, 57-58.
(13) Eriksson, U. G.; Renberg, L.; Vedin, C.; Strimfors, M. Pharma-
cokinetics of inogatran, a new low molecular weight thrombin
inhibitor, in rats and dogs. Thromb. Haemostas. 1995, 73, 1605.
(14) Markwardt, F.; Wagner, G.; Stu¨rzebecher, J .; Walsmann, P. NR-
Arylsulfonyl-ω-(4-amidinophenyl)-R-aminoalkylcarboxylic acid
amides - novel selective inhibitors of thrombin. Thromb. Res.
1980, 17, 425-431.
(15) Stu¨rzebecher, J .; Markwardt, F.; Voigt, B.; Wagner, G.; Wals-
mann, P. Cyclic amides of NR-arylsulfonylaminoacylated 4-ami-
dinophenylalanine - tight binding inhibitors of thrombin. Thromb.
Res. 1983, 29, 635-642.
(16) Reers, M.; Koschinsky, R.; Dickneite, G.; Hoffmann, D.; Czech,
J .; Stu¨ber, W. Synthesis and characterisation of novel thrombin
inhibitors based on 4-amidinophenylalanine. J . Enzyme Inhib.
1995, 9, 61-72.
(17) Kaiser, B.; Hauptmann, J .; Weiss, A.; Markwardt, F. Pharma-
cological characterization of a new highly effective synthetic
thrombin inhibitor. Biomed. Biochim. Acta 1985, 44, 1201-1210.
(18) Dickneite, G.; Seiffge, D.; Diehl, K. H.; Reers, M.; Czech, J .;
Weinmann, E.; Hoffmann, D.; Stu¨ber, W. Pharmacological
characterization of a new 4-amidinophenylalanine thrombin
inhibitor (CRC 220). Thromb. Res. 1995, 77, 357-368.
(19) Stu¨rzebecher, J .; Horn, H.; Walsmann, P.; Voigt, B.; Markwardt,
F.; Wagner, G. Synthetische Inhibitoren der Serinproteinasen.
34. Mittg.: Einfluss der Abwandlung der Amidinofunktion bei
NR-substituierten 4-Amidinophenylalaninamiden auf die In-
hibitoraktivita¨t. (Synthetic inhibitors of serine proteinases. Part
34: Influence of modification of the amidino function of NR-
substituted 4-amidinophenylalanine amides on the inhibitory
activity.) Pharmazie 1988, 43, 782-783.
(20) Bode, W.; Turk, D.; Stu¨rzebecher, J . Geometry of binding of
benzamidine and arginine-based inhibitors NAPAP and MQPA
to human R-thrombin: X-ray crystallographic determination of
the NAPAP-trypsin complex and modeling of NAPAP-thrombin
and MQPA-thrombin. Eur. J . Biochem. 1990, 193, 175-182.
(21) Bauer, M.; Brandstetter, H.; Turk, D.; Stu¨rzebecher, J .; Bode,
W. Crystallographic determination of thrombin complexes with
small synthetic inhibitors as starting point for the receptor-based
design of antithrombotics. Semin. Thromb. Hemostas. 1993, 19,
352-360.
(22) Turk, D.; Stu¨rzebecher, J .; Bode, W. Geometry of binding of the
NR-tosylated piperidides of m-amidino-, p-amidino- and p-
guanidino phenylalanine to trypsin and thrombin complexes.
FEBS Lett. 1991, 287, 133-138.
(23) Brandstetter, H.; Turk, D.; Hoeffken, H. W.; Grosse, D.; Stu¨r-
zebecher, J .; Martin, P. D.; Edwards, B. F. P.; Bode, W. Refined
2.3 Å X-ray crystal structure of bovine thrombin complexes
formed with the benzamidine and arginine-based thrombin
inhibitors NAPAP, 4-TAPAP and MQPA. J . Mol. Biol. 1992, 226,
1085-1099.
In Vivo Elim in a tion Stu d y. NAPAP and compound 6
were administered to anesthetized (urethane, 1.5 g/kg ip) rats
in aqueous solution intravenously in a dose of 1 mg/kg and
orally (by gastric tube) in a dose of 50 mg/kg. Citrated blood
was drawn at various times after administration from the
cannulated carotid artery and centrifuged, and the plasma was
poured onto Chromabond C18 solid-phase exctraction columns
(Macherey-Nagel, Du¨ren, Germany). The concentrations in
plasma were measured by HPLC (2700 HPLC Systems, Bio-
Rad, Mu¨nchen, Germany) using Nucleosil 7 C18 columns
(Macherey-Nagel, Du¨ren, Germany). The mobile phase was
acetonitrile/water/perchloric acid (30/70/0.04; flow rate 1 mL/
min); the compounds were detected by fluorescence (λex ) 232
nm, λem ) 343 nm).
Ack n ow led gm en t. We would like to express our
thanks to Ute Altmann, Christa Bo¨ttner, and Gabriele
Riesener for their excellent technical assistance.
Refer en ces
(1) Sixma, J . J .; de Groot, P. G. The ideal anti-thrombotic drug.
Thromb. Res. 1992, 68, 507-512.
(24) Stu¨rzebecher, J .; Vieweg, H.; Wikstro¨m, P. WO 92/08709, 1991.
(25) Stu¨rzebecher, J .; Prasa, D.; Wikstro¨m, P.; Vieweg, H. Structure-
activity relationships of inhibitors derived from 3-amidinophe-
nylalanine. J . Enzyme Inhib. 1995, 9, 87-99.
(26) Bergner, A.; Bauer, M.; Brandstetter, H.; Stu¨rzebecher, J .; Bode,
W. The X-ray crystal structure of thrombin in complex with
NR-2-naphthylsulfonyl-L-3-amidinophenylalanyl-4-methyl-pip-
eridide: the beneficial effect of filling out an empty cavity. J .
Enzyme Inhib. 1995, 9, 101-111.
(27) Stu¨rzebecher, J .; Vieweg, H.; Wikstro¨m, P. WO 94/18185, 1993.
(28) Bode, W.; Turk, D.; Karshikov, A. The refined 1.9 Å X-ray crystal
structure of D-Phe-Pro-Arg chloromethylketone-inhibited human
R-thrombin: Structure analysis, overall structure, elecctrostatic
(2) Powers, J . C.; Kam, C.-M. Synthetic substrates and inhibitors
of thrombin. In Thrombin. Structure and Function; Berliner, L.
J ., Ed.; Plenum Press: New York, 1992; pp 117-158.
(3) Harker, L. A.; Hanson, S. R.; Kelly, A. B. Antithrombotic benefits
and hemorrhagic risks of direct thrombin antagonists. Thromb.
Haemostas. 1995, 74, 464-472.
(4) Stone, S. Thrombin inhibitors - A new generation of antithrom-
botics. Trends Cardiovasc. Med. 1995, 5, 134-140.
(5) Markwardt, F. The development of hirudin as an antithrombotic
drug. Thromb. Res. 1994, 74, 1-23.
(6) Maraganore, J . M.; Bourdon, P.; J ablonski, J .; Ramachandran,
K. L.; Fenton, J . W., II Design and characterization of hirulogs:

3-Amidinophenylalanine-Derived Inhibitors
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19 3099
properties, detailed active-site geometry, and structure-function
plasma independent of thrombin decay processes. Thromb.
Haemostas. 1986, 56, 9-17.
(33) Prasa, D.; Svendsen, L.; Stu¨rzebecher, J . The ability of thrombin
inhibitors to reduce the thrombin activity generated in plasma
on extrinsic and intrinsic activation. Thromb. Haemostas. 1997,
77, 498-503.
(34) Dixon, M. Determination of enzyme inhibitor constants. Bio-
chem. J . 1953, 55, 170-171.
(35) Walsmann, P. U¨ ber die Reinigung von Thrombinpra¨paraten. (On
the purification of thrombin preparations.) Pharmazie 1968, 23,
401-402.
(36) Stu¨rzebecher, J .; Neumann, U.; Kohnert, U.; Kresse, G.-B.;
Fischer, S. Mapping of the catalytic site of CHO-t-PA and the
t-PA variant BM 06.022 by synthetic inhibitors and substrates.
Protein Sci. 1992, 1, 1007-1013.
relationships. Protein Sci. 1992, 1, 426-471.
(29) Abbreviations:
F Xa, factor Xa; F XIIa, factor XIIa; APC,
activated protein C; PK, plasma kallikrein; GK, glandular
kallikrein; UK, urokinase; sc-tPA, single-chain tissue-type plas-
minogen activator; NMM, N-methylmorpholine; HOBt, hydroxy-
benzotriazole; DCC, N,N′-dicyclohexylcarbodiimide; DMF, N,N′-
dimethylformamide; TEA, triethylamine; THF, tetrahydrofuran.
(30) Deadman, J . J .; Elgendy, S.; Goodwin, C. A.; Green, D.; Baban,
J . A.; Patel, G.; Skordalakes, E.; Chino, N.; Claeson, G.; Kakkar,
V. V.; Scully, M. F. Characterization of a class of peptide
boronates with neutral P1 side chains as highly selective
inhibitors of thrombin. J . Med. Chem. 1995, 38, 1511-1522.
(31) Authors, unpublished results.
(32) Hemker, H. C.; Willems, G. M.; Be´guin, S. A computer assisted
method to obtain the prothrombin activation velocity in whole
J M960668H