Unique overlap in the prerequisites for thrombin inhibition and oral bioavailability resulting in potent oral antithrombotics

Article abstract of DOI:10.1021/jm011110z

Despite intense research over the last 10 years, aided by the availability of X-ray structures of enzyme-inhibitor complexes, only very few truly orally active thrombin inhibitors have been found. We conducted a comprehensive study starting with peptide transition state analogues (TSA). Both hydrophobic nonpeptide analogues as well as hydrophilic peptidic analogues were synthesized. The bioavailability in rats and dogs could be drastically altered depending on the overall charge distribution in the molecule. Compound 27, a tripeptide TSA inhibitor of thrombin, showed an oral bioavailability of 32% in rats and 71% in dogs, elimination half-lives being 58 and 108 min, respectively. The thrombin inhibition constant of compound 27 was 1.1 nM, and in an in vivo arterial flow model, the ED₅₀ was 5.4 nmol/kg^.min, comparable to known non-TSA inhibitors. A molecular design was found that combines antithrombotic efficiency with oral bioavailability at low dosages.

Full text of DOI:10.1021/jm011110z

J . Med. Chem. 2002, 45, 4419-4432
4419
Un iqu e Over la p in th e P r er equ isites for Th r om bin In h ibition a n d Or a l
Bioa va ila bility Resu ltin g in P oten t Or a l An tith r om botics
Anton E. P. Adang,*,† Adrianus P. A. de Man,† Gerard M. T. Vogel,‡ Peter D. J . Grootenhuis,§ Martin J . Smit,‡
Co A. M. Peters,† Arie Visser,| J os B. M. Rewinkel, Theo van Dinther,‡ Hans Lucas,† J an Kelder,#
Sjoerd van Aelst,† Dick G. Meuleman,‡ and Constant A. A. van Boeckel†
Research and Development, Lead Discovery Unit Chemistry Group, NV Organon, Lead Optimization Unit Pharmacology
Department, NV Organon, DuPont Pharmaceuticals Research Laboratories, Lead Discovery Unit Assay Development Group,
NV Organon, Lead Optmization Unit Medicinal Chemistry Department, NV Organon, and Department of Molecular Design
and Informatics, NV Organon
Received November 23, 2001
Despite intense research over the last 10 years, aided by the availability of X-ray structures of
enzyme-inhibitor complexes, only very few truly orally active thrombin inhibitors have been
found. We conducted a comprehensive study starting with peptide transition state analogues
(TSA). Both hydrophobic nonpeptide analogues as well as hydrophilic peptidic analogues were
synthesized. The bioavailability in rats and dogs could be drastically altered depending on the
overall charge distribution in the molecule. Compound 27, a tripeptide TSA inhibitor of
thrombin, showed an oral bioavailability of 32% in rats and 71% in dogs, elimination half-
lives being 58 and 108 min, respectively. The thrombin inhibition constant of compound 27
was 1.1 nM, and in an in vivo arterial flow model, the ED50 was 5.4 nmol/kg‚min, comparable
to known non-TSA inhibitors. A molecular design was found that combines antithrombotic
efficiency with oral bioavailability at low dosages.
In tr od u ction
various X-ray data on thrombin-ligand complexes
became available in the early 1990s, a tremendous
amount of research toward thrombin inhibitors has been
performed.10 However, thrombin turns out to be a very
difficult molecular target for finding orally active anti-
thrombotics. Thrombin expects a peptide in the catalytic
cleft with a highly basic arginine moiety in its specificity
pocket (P1 pocket) and, being a serine protease, expects
a carbonyl function at the cleavage site. The require-
ments for the P2 and P3 binding pockets are less
stringent allowing for turn-inducing hydrophobic moi-
eties of a certain size.11 After a decade of intense
research in many laboratories, only a few orally bio-
available thrombin inhibitors are known. Low anti-
thrombotic potency and side effects in animal models
further reduce the list as can be viewed in recent
reviews.12 Apparently, the quest for the ideal orally
active thrombin inhibitor is still worthwhile to pursue.
An interesting area of drug discovery research deals
with enzymes and receptors geared to recognize peptides
or proteins. To find low molecular weight orally ap-
plicable ligands for these targets, high throughput
screening of compound collections is frequently used.
Alternatively, the peptides and proteins themselves
form the basis for peptidomimetic approaches in which
molecular modification leads to nonpeptide structures.
Screening has been especially successful in finding small
molecules for G-protein-coupled peptide receptors rec-
ognizing, for instance, angiotensin, substance P, chole-
cystokinin, vasopressin, or corticotropin-releasing fac-
tor.1-5 In the area of protease inhibitors, however,
screening has proven less effective, the most important
exception being human immunodeficiency virus (HIV)
protease. Researchers have relied on peptidomimetic
approaches in areas such as elastase, renin, angiotensin-
converting enzyme (ACE), and coagulation enzymes.6-8
The latter class consists of serine proteases such as
thrombin, factor Xa, and factor VIIa/TF. Over the past
years, thrombin inhibitors, together with platelet recep-
tor antagonists, have been the focal point of attention
in pharmaceutical research aimed at finding new orally
active antithrombotic agents.9 Because thrombin plays
a pivotal role in the blood coagulation cascade and
Recently, we reported on several series of potent
thrombin inhibitors with isosteric replacements for the
highly basic S1 binding moieties present in earlier leads
and seen as prohibitive for intestinal absorption.
Acylguanidines, aminoisoquinolines, and piperidines
replaced guanidines, benzamidines, and primary amines,
respectively, showing that the P1 pocket can also ac-
commodate more drug friendly entities.13-15 Now, we
wish to report on a comprehensive lead optimization
study starting with peptide analogues that include
transition state mimics and describe the stepwise
improvement toward potent orally active antithrombot-
ics. During this optimization process, publications ap-
peared on the perceived utility of this class of thrombin
inhibitors. Concerns were expressed that inhibition with
transition state analogues (TSA) would result in low
* To whom correspondence should be addressed. Tel: +31-412-
662069. Fax: +31-412-662514. E-mail: ton.adang@organon.com.
† Research and Development, Lead Discovery Unit Chemistry
Group, NV Organon.
‡ Lead Optimization Unit Pharmacology Department, NV Organon.
§ DuPont Pharmaceuticals Research Laboratories.
| Lead Discovery Unit Assay Development Group, NV Organon.
Lead Optmization Unit Medicinal Chemistry Department, NV
Organon.
#
Department of Molecular Design and Informatics, NV Organon.
10.1021/jm011110z CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/30/2002

4420 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Sch em e 1. Design of First Generation Peptidomimetics
Adang et al.
selectivity over related proteases and, especially, in low
kon rates due to slow tight binding kinetics. These points
were viewed as being incompatible with the intended
use of the inhibitors, i.e., blocking the fast generation
of thrombin at the site of a lesion in acute situations.16,17
Contrary to these concerns, it will be shown that
effective antithrombotics, applicable in arterial throm-
bosis situations, can be generated from this class of
transition state mimics.
earlier work on D-Phe-Pro-Arg aldehydes18 and boronic
acids19 and work on other serine protease inhibitors.6
Compound 1 is a natural product called cyclotheon-
amide with a reported Ki (final) of 1.0 nM, and 2 is a
transition state mimic that came out of our earlier
research showing a Ki of 28 nM.20,21 Although very
potent in vitro, no oral bioavailability has ever been
reported or found for 1 and 2. The first design cycle
started with a hybrid molecule 3 as a putative new
inhibitor, and we evolved this into potent orally bio-
available thrombin inhibitors.
Starting points for the present series of compounds
were peptidic structures 1 and 2 (see Scheme 1) next to

Potent Oral Antithrombotics
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4421
Sch em e 2. General Strategy for the Synthesis of Thrombin Inhibitors Containing an Active Ketone Speciesa
a
Panel A describes keto-heterocyclic peptidomimetics and panel
B describes the diketo series. Series A: (i) TBTU, O,N-
dimethylhydroxylamine‚HCl, Et3N, CH2Cl2. (ii) (1) n-BuLi/ether, 2-bromothiazole, -78 °C; (2) THF, -78 °C. (iii) 50% TFA/CH2Cl2. (iv)
DCCI, HOBt, DMF, Et3N. (v) TFA/thioanisole ) 10/1 v/v%. (vi) Purification preparative HPLC. (vii) DiBAL-H, CH2Cl2, -78 °C. (viii)
2-(Trimethylsilyl)oxazole. (ix) 80 °C; Dess-Martin. Series B: (i) TBTU, CH2Cl2/CH3OH, Et3N. (ii) DiBAL-H, CH2Cl2, -78 °C. (iii) NaCN,
TEBACcat, CH2Cl2/water. (iv) 3 M HClgas in diethyl ether/CH3OH, 4 °C. (v) Boc2O, DMF, Et3N. (vi) H2, 10% Pd/C, DMF-1 N HCl. (vii)
DCCI, HOBt, DMF, Et3N. (viii) Dioxane/water, pH 10. (ix) Dess-Martin. (x) TFA/CH2Cl2. (xi) Purification preparative HPLC.
Ch em istr y
using a phase transfer catalyst; the hydroxyl group,
finally, was acetylated to allow for high yield work up
procedures. The crucial step was the two step Pinner
hydrolysis (III to IV) followed by an in situ Boc protec-
tion step. After the required peptide coupling steps, the
tripeptides were oxidized using Dess-Martin periodi-
nane reagent.25 In our hands, this was one of the few
reagents that was capable of oxidizing secondary hy-
droxyl groups while retaining the stereochemistry R to
the oxidation site. Finally, deprotection and preparative
HPLC purification yielded the products in this series.
In accordance with Brady et al.,16 it was found that the
S configuration of the lysine moiety remained intact
during the synthesis and in solution at low pH. At pH
levels above 7.5, time-dependent epimerization could be
detected. Biological assays were carried out using strict
protocols and time limitations in order to minimize
epimerization.
Chemistry generally involved the stereoselective syn-
thesis of peptide structures with an activated ketone
at the C terminus allowing for a transition state type
of interaction with thrombin’s cleavage site. Essential
in these syntheses is the preparation of the amino acid
monomers with either a heterocycle or an additional
ketone attached to the C terminus of these intermedi-
ates. Scheme 2A shows the conversion of an amino acid
like lysine to a keto-(2-thiazolyl) derivative using Don-
doni type alkylations:22 first the N-methoxymethyl-
amide23 was prepared followed by reaction with 2-lithio-
thiazole at -78 °C. Next, the R amino protective group
was removed with trifluoroacetic acid (TFA) in dichlo-
romethane followed by standard peptide coupling reac-
tions to provide fully protected end molecules. After the
remaining protective groups were removed and pre-
parative high-performance liquid chromatography
(HPLC), the final products were obtained. The oxazole
derivates were obtained via the aldehyde, prepared from
Boc-Lys(Cbz)-OMe (Scheme 2A). Reaction of the alde-
hyde with 2-trimethylsilyloxazole and subsequent oxi-
dation of the resulting alcohol yielded the keto-(2-
oxazolyl) derivative. This ketone was applied in peptide
chain elongation in the same way as the keto-(2-
thiazolyl).
Resu lts a n d Discu ssion
In an effort to combine structural features of the
thrombin inhibitors 1 and 2, tripeptide 3 became our
initial target (Scheme 1). However, 3 was extremely
difficult to synthesize and turned out to be unstable due
to the presence of both the reactive keto-carboxylate
moiety and the guanidine group. During synthesis, the
compound cyclizes spontaneously resulting in the previ-
ously unknown heterocyclic compound 4. The open and
closed form appeared to be in a pH- and temperature-
controlled creatine-creatinine equilibrium with each
other. To avoid this side reaction, we stabilized the
Scheme 2B shows a Pinner type reaction24 sequence
to allow for the introduction of additional ketones in the
end products. Thus, Z-Lys(Boc)-OMe was first trans-
formed to an aldehyde using diisobutylaluminumhy-
dride (DiBAL-H) followed by a cyanohydrin formation

4422 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Adang et al.
Although 6 did not show an improvement in oral
bioavailability, the half-life of 46 min in rats was an
improvement as compared to arginine derivative 5 (see
Figure 1). Liver perfusion experiments revealed that 5
was cleared much faster than the corresponding lysine
analogue 6, which remained in circulation for several
hours. However, compound 6 exhibited a Ki of 146 nM,
warranting further optimization.
First, we resorted to keto-carboxylate derivative 7,
which turned out to be more active than analogue 6 (Ki
of 1.4 nM) and also more stable than its arginine
counterpart 3. It is acknowledged that at the time our
patents29,30 were published describing these lysine
derivatives, a number of publications appeared on
related heterocycle-activated ketone derivatives and
R-keto acids16,17,31 and derivatives of the D-Phe-Pro-Arg
motif.26,27,32-34 Some of these papers describe the sub-
stitution of a lysine for an arginine group. More publica-
tions appeared in recent years on this subject focusing
on potent thrombin inhibition and selectivity but with
limited data on pharmacokinetic behavior.35-37 To fur-
ther improve oral bioavailability, two approaches were
initiated starting from 6 and 7. In the first approach,
we aimed at improvement of passive diffusion, while in
the second approach we incorporated structural features
of compounds that are known to be absorbed by active
or paracellular transport. The first approach involved
the transformation of a tripeptide into a nonpeptide
while the second route still allowed for peptidic char-
acteristics.
F igu r e 1. (A) Pharmacokinetics after intravenous adminis-
tration to rats. Line graph showing dissappearance patterns
of the anti-IIa activity of compound 5 (O) and compound 6 (0)
after administration of 10 µmol/kg IV in rats. Data are
expressed in percent remaining concentration as compared to
T ) 1′ and represent means of three separate experiments (
SEM. (B) Rate of disappearance in rat liver perfusion experi-
ment. The disappearance patterns of the antithrombin activity
of the same compounds (5 (O) and 6 (0)) during liver perfusion
are shown. Data are expressed in percent remaining concen-
tration as compared to preperfusion level and represent means
of two separate experiments.
Following the first approach, we took into account the
Lipinski rules for passive diffusion.38 Reducing the
number of H bond donors and acceptors, minimizing the
number of rotatable bonds, and lowering the molecular
weight, etc. could be achieved. However, we could not
obtain a potent thrombin inhibitor with improved Caco
2 permeability (see Table 1 for a representative set of
compounds and results). The left panel of Table 1 shows
lysine-ketothiazole derivatives with peptide character
and therefore rather poor Lipinski scores but with
nanomolar potency. The right panel depicts some key
compounds such as 13 and 14 that may be described as
nonpeptides with permeabilities of 16 and 81 nm/sec,
respectively, but without any appreciable potency left.
molecule by replacing the reactive keto-carboxylate
moiety by a keto-thiazole. Indeed, 5 was chemically
stable and still an active thrombin inhibitor with a Ki
of less than 1 nM. The N-methyl analogue of 5 was
reported by Costanzo et al. showing a Ki of 3.6 nM for
thrombin and no trypsin selectivity.26,27 Because it was
the new lead compound for further optimization studies,
analogue 5 was subjected to a more elaborate pharma-
cological analysis. This resulted in the findings that the
Ki ratios of related coagulation proteases over thrombin
were generally high except for trypsin. Furthermore,
enzyme kinetics revealed no slow binding characteristics
of 5. Unfortunately, in pharmacokinetic studies, less
than 5% oral bioavailability was observed and the half-
life was only about 5 min (see Figure 1). It was
anticipated that efficient hepatic clearance had occurred
as has been observed frequently with thrombin inhibi-
tors bearing basic guanidine or benzamidine entities.
In addition, haemodynamic side effects have been
described for thrombin inhibitors comprising these
highly basic groups.28
In the second approach, we compared the structural
properties of our lead compounds with hydrophilic drugs
absorbed at least in part through active transport. To
this end, we scrutinized the structural properties of
â-lactam antibiotics and ACE inhibitors (e.g., lisinopril).
For these drugs, it has been reasoned that the presence
of amine and carboxylate groups at specific distances
as well as their resemblance to di- and tripeptides
render them good substrates for peptide carriers in the
intestine.39,40 In addition, cyclic structures such as a
pyrrolidine appear to be required and provide some
constraint in these molecules. As shown in Figure 2, the
resemblance between lead compound 7 and lisinopril is
striking. To create an even better structural overlap, it
was decided to introduce a carboxylate group in lead
compounds 6 and 7. In this approach, it was not
considered to be necessary to lower the pKa of the S1
amine any further. A carboxylate scan was employed
to introduce this acidic group at all synthetically acces-
sible positions (see Table 2 for key compounds). Com-
In the next round of optimization, we replaced the
guanidine group in 5 with less basic or even neutral
moieties. An additional advantage of these modifications
is the increase in lipophilicity. A typical example is 6,
which utilizes a lysine instead of an arginine side chain.

Potent Oral Antithrombotics
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4423
Ta ble 1. Results of the Nonpeptide Approach with In Vitro Thrombin Inhibitory Potency and Transport Across a Monolayer of Caco
II Cells
Ta ble 2. Carboxylate Scan of Compound 15 to Arrive at the
Second Generation Thrombin Inhibitorsa
F igu r e 2. Structural resemblance between compound 7 and
lisinopril.
pound 18 was our first thrombin inhibitor that displayed
significant oral bioavailability in rats at oral dosages
ranging from 5 to 20 mg/kg. Further optimization of this
breakthrough compound on potency and bioavailablility
led to the diketo-carboxylate analogue 20. Introduction
of the N-methyl carboxylate moiety led to an increase
in oral bioavailability in both the keto-(2-thiazolyl) and
the diketo-carboxylate series: compare 15 with 18,
respectively, <5 and 23%, and compare 16 with 20,
respectively, 9 and 39% in rats. On the other hand, as
exemplified by 17 and 19, introduction of a carboxylate
group in the molecule did not always result in enhanced
oral bioavailability.
Additional pharmacokinetic studies with nephrecto-
mized rats using the improved lead 20 showed virtually
no plasma disappearance indicating that the clearance
was almost exclusively renal. Liver perfusion experi-
a
ments (3 h, 5 mg/kg iv) with 20 revealed a clearance of
Data are inhibition constants toward thrombin and pharma-
cokinetics as established in male Wistar rats.
0.025 L/hr, which is very low by all standards. Table 3

4424 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Ta ble 3. Results of Further Optimizations of Compound 20a
Adang et al.
a
Two series are shown in this table: one targetting the proline residue and one the diketo moiety. T1/2 values were obtained after IV
administration. %F values were obtained by dividing the AUC after PO administration by the AUC after IV administration normalized
for the dose.
shows the elimination half-life of 20 in rat and dog after
intravenous administration (t1/2 rat ) 19 min, t1/2 dog
) 78 min).
replacement of the proline moiety by known mimics
(e.g., compounds 21-26) and studied other keto transi-
tion state mimics including the keto carboxylate iso-
propyl ester 27 and keto-phenylethyl amide 28. Al-
though the inhibitory activity of most analogues is
similar, their oral bioavailability differs markedly. Ap-
parently, the requirements for intestinal absorption
could be met with 20 and some closely related ana-
To further guide the lead optimization process, several
crystal structures of enzyme-inhibitor complexes were
obtained. Key compounds such as 18 and 20 could be
crystallized with thrombin showing numerous interac-
tions between the active site of thrombin and the
inhibitor structure (see Figure 3 for X-ray on 20). The
extended binding mode observed with these compounds
is in accordance with previously published crystal-
lographic data. Some of the lysine and the N-methyl-
carboxylate interactions with thrombin, revealing for
instance alternative P1 interactions through water
molecules and backbone amide bonds, were less known.
As the crystal structure did not provide a special clue
for further structural modifications, we dwelled on
logues. Only the S′ site of this class of inhibitors may
1
be modified considerably without losing oral bioavail-
ability, e.g., 2-thiazole, carboxylate, isopropyl, and phen-
ylethylamine. In particular, the ester derivative of 20
(i.e., compound 27) showed oral bioavailabilities of 32
and 71% in rats and dogs, respectively (see Figure 4).
In Caco-2 experiments, we could not measure any
degree of membrane transport with representatives of
this series. In addition, we performed competition

Potent Oral Antithrombotics
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4425
F igu r e 5. Pharmacophore model for intestinal absorption of
polar thrombin inhibitors.
Ta ble 4. Summary of Inhibitory Constants toward Thrombin
and Antithrombotic Potencies in the Aortic Flow Model in Rats
of Selected Thrombin Inhibitors and Reference Compoundsa
F igu r e 3. X-ray crystal structure of compound 20 bound to
human R-thrombin at 1.59 Å.
ED50 (mean (95% CI))
example/ref compd
Ki (nM)
(nmol/kg‚min)
18
20
27
2.6
0.29
1.1
19 (13; 25)
7.5 (4.7; 11)
5.4 (4.0; 6.6)
27 (19; 36)
argatroban
melagatran
1949
2.050
5.2 (4.2; 6.5)
a
KI values in nanomolar; ED50 values in nmol/kg‚min; SD was
25% on average; and data represent means of n ) 4-6.
compounds to enter the body. Studies aimed at estab-
lishing Caco-2 cell transport of ACE inhibitors showed
low proton-linked transport of lisinopril and only limited
competition with known peptide carrier ligands. Appar-
ently, although clearly absorbed after oral administra-
tion in humans, lisinopril is a poor substrate for the
peptide carrier present in Caco-2 cells.41 It remains
unclear by which mechanism these hydrophilic com-
pounds are absorbed. Perhaps intramolecular ion-pair
formation of such molecules occurs resulting in less
hydrophilic and more compact molecules. Despite the
fact that we do not fully understand the underlying
mechanism, a “pharmacophore” for absorption of these
thrombin inhibitors can be envisaged (Figure 5). It is a
very restricted window of opportunity best described as
carboxy-Pro-Lys with some flexibility at the hydrophobic
S3 moiety and the transition state S′ moiety.
1
Finally, we were anxious to compare the in vivo
antithrombotic properties with known thrombin inhibi-
tors (see Table 4). Compounds 18, 20, and 27 display
antithrombotic activities in an aortic flow model in the
rat in the same range as melagatran and argatroban,
generally viewed as the most advanced drug candidates
in this area. The aortic flow model is thought to be
predictive for thrombus formation in fast-flowing ves-
sels, i.e., for arterial thrombosis-related diseases.
F igu r e 4. Pharmacokinetics of compound 27 in rats (A) and
dogs (B). Line graphs showing disappearance patterns of the
anti-IIa activity of compound 27 after i.v. (4) or oral (2)
administration of 10 µmol/kg. Data are expressed in nmol/mL
and represent means of two (dogs) or three (rats) separate
experiments ( SEM.
In vitro data on selectivity profiles of key compounds
in the above-described series have been collected. In a
typical selectivity screen, activities of a number of
human-derived serine proteases were measured. Gener-
ally, it consisted of fXa, trypsin, fVIIa/TF, plasmin, tPA,
activated protein C, and protrombinase. Most com-
pounds are selective (50-1000) toward the coagulation
factors but show no or only limited selectivity (<1-10)
experiments with dipeptides and cephalexin but could
not find any evidence of active transport in Caco-2 cell
layers. Furthermore, we did not find evidence for
paracellular transport that sometimes allows polar

4426 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Adang et al.
Pak C18 (radial compression) prep columns and various
gradients (solvent A: 0.5 M phosphate buffer, pH 2.1; solvent
B: water; solvent C: acetonitril:water, 6:4 v/v%). All (pro-
tected) dipeptides were prepared by synthesis to conventional
toward the digestive enzyme trypsin. No correlation
could be found between oral absorption and trypsin
activity.
Although the enzyme kinetics will not be reported in
great detail in this paper, it must be stated that some
of the compounds, more specifically those with a diketo
moiety (Table 3), showed slow binding characteristics.
methodology: coupling of Boc-protected N-terminal L or
D
amino acid with either L or D amino acid benzyl ester or L or
D amino acid methyl ester, followed by hydrogenation or
saponification, respectively. HPLC elution of the main peak
usually followed by a small second peak with identical mass,
in combination with NMR techniques, warranted the diaster-
eomeric purity of the end products.
For instance, 20 had a Ki of 0.29 nM and a K/ of 0.03
i
nM. The slow binding effects were obvious only in the
first 10-20 min of the kinetic experiments. The results
as depicted in Table 4 lead to the conclusion that
transition state mimic inhibitors are efficient in arterial
models. The slope of the dose response curve in the
aortic flow model of a nontransition state mimic like
melagatran is not different from compounds such as 20
and 27. Presumed differences in kon rates appear hardly
relevant for the ultimate in vivo effect. The concern that
activated ketones in TSAs all show slow tight binding
inhibition that may impair successful blocking of throm-
bin in acute situations appears unwarranted for this
series.
In several papers,16,17 it was mentioned that it is
extremely difficult to find the balance between lipophi-
licity, needed for oral absorption, and antithrombotic
potential, measured for instance in an APTT experi-
ment. Failures have been reported in which the lipo-
philicity was increased to get the desired oral absorption
but at the expense of antithrombotic efficacy. Massive
protein binding and/or large volumes of distribution may
occur, which, for this specific class of compounds that
exert their action on circulating blood enzymes, is very
unfavorable. Clearly, the current hydrophilic series
distinguishes itself in that oral absorption was obtained
without compromising the antithrombotic action of the
compounds. In that respect, this series offers an alter-
native to the main stream passive diffusion approaches
and may become more important as we learn more
about the fate of the leading drug candidates in the
clinical arena and why so many drug candidates fail in
clinical trials.
P r ep a r a tion of N8(D-P h en yla la n yl-p r olyl)-3,8-d ia m in o-
6,7,8,8a -t e t r a h yd r o-8a -h yd r oxyim id a zo[1,5a ]p yr id in -
1(5H)-on e (4). Nr,Nδ,N-Tr iben zyloxyca r bon yl-L-a r gin in e
m eth ylester . The pH of a solution of 40 g (66.7 mmol) of
NR,Nδ,N-tri-benzyloxycarbonyl-L-arginine, prepared as de-
scribed previously,42 and 22.4 g (69.8 mmol) of TBTU in a
mixture of CH2Cl2 (1080 mL) and MeOH (120 mL) was
adjusted at pH 8 with Et3N. The reaction mixture was stirred
for 1 h at room temperature, after which the solution was
successively washed with a 5% NaHCO3 solution and water,
dried on Na2SO4, and evaporated. The solid residue was
crystallized from MeOH to give 35 g of Z-Arg(Z)2-OMe (yield
88.8%). TLC (CH2Cl2:MeOH, 9:1) Rf ) 0.60. 1H NMR (360
MHz, CDCl3): δ 1.70 (m, 4H), 3.63 (s, 3H), 3.95 (dt, 2H), 4.37
(dd, 1H), 5.07 (s, 2H), 5.12 (s, 2H), 5.22 (s, 2H), 5.52 (d, 1H),
7.30-7.42 (m, 15H), 9.24 (bs, 1H), 9.42 (bs, 1H).
2-Acet oxy-3-(b en zyloxyca r b on yla m in o)-6-(d ib en zyl-
oxyca r bon ylgu a n id in o)h exa n en itr il. A solution of 180 mL
(180 mmol) of diisobutylaluminumhydride was added dropwise
at -78 °C to a stirred solution of 30 g (47.4 mmol) of Z-Arg-
(Z)2-OMe in 700 mL of CH2Cl2. The mixture was stirred for 1
h at -78 °C, poured in a solution of citric acid in water, and
extracted with CH2Cl2. The combined extracts were used in
the next step. To the solution of the aldehyde was added a
solution of 28 g (0.57 mol) of NaCN and 35 g (0.15 mmol) of
triethylbenzylammonium chloride in 700 mL of water and 14
mL of acetic anhydride at 0 °C. The mixture was stirred for
30 min at 0-5 °C. The organic layer was separated, washed
with water, dried on Na2SO4, and evaporated in vacuo.
Purification of the residue by chromatography (CH2Cl2:MeOH,
9:1) afforded the title compound as a solid (17 g, 57.0%). TLC
1
(CH2Cl2:MeOH, 9:1) Rf ) 0.76. H NMR (360 MHz, CDCl3): δ
1.98 and 1.99 (s, 3H), 3.90 (m, 1H), 4.10 (m, 1H), 4.15 (m, 1H),
5.05 and 5.07 (s, 2H), 5.09 and 5.10 (s, 2H), 5.23 and 5.24 (s,
2H), 5.38 and 5.44 (d, 1H), 5.79 (d, 1H), 7.32 (m, 15H), 9.29
(br s, 1H), 9.48 (b s, 1H).
Exp er im en ta l Section
3-(Ben zyloxyca r b on yla m in o)-6-(d ib en zyloxyca r b on -
ylgu a n id in o)-2-h yd r oxy-h exa n oic Acid . At -78 °C, hydro-
gen chloride gas was passed through a solution of 6 g (9.30
mmol) of 2-acetoxy-3-(benzyloxycarbonylamino)-6-(dibenzyl-
oxycarbonylguanidino) hexanenitril in 140 mL of a mixture
of Et2O:MeOH, 3:1, until a 3 M solution was obtained. The
mixture was stirred for 16 h at 5 °C, poured into water, and
extracted with CH2Cl2. The combined extracts were washed
with water and 5% NaHCO3 solution and water and dried on
Na2SO4. The solvent was removed in vacuo yielding the title
compound as a gum (6.1 g, quantitative). TLC (CH2Cl2:MeOH,
9:1) Rf ) 0.48. 1H NMR (360 MHz, CDCl3): δ 1.65 (m, 4H),
3.65 and 3.68 (s, 3H), 4.22 and 4.27 (d, 1H), 5.03 and 5.06 (AB,
2H), 5.10 (s, 2H), 5.20 and 5.22 (s, 2H), 7.33 (m, 15H), 9.27 (b
s,1H), 9.43 (b s,1H).
3-(Be n zyloxyca r b on yla m in o)-6-(d ib e n zoylca r b on -
ylgu a n id in o)-2-oxo-h exa n oic Acid Meth ylester . To a solu-
tion of 1.3 g (2.05 mmol) of 3-(benzyloxycarbonylamino)-6-
(dibenzyloxycar-bonylguanidino)-2-hydroxy-hexanoic acid in
130 mL of acetone was added 1 mL of 8 N chromic acid solution
in aqueous sulfuric acid at 0 °C. The mixture was stirred for
1 h at 0 °C and then poured into water. The precipitate was
filtered off, washed with water, and dried in vacuo to give a
white solid (1.15 g, 90.7%). TLC (CH2Cl2:EtOAc, 8:2) Rf ) 0.8.
1H NMR (360 MHz, CDCl3): δ 1.66 (m, 4H), 3.76 (s, 2H), 3.78
(s, 3H), 4.9 (m, 1H), 5.04 (s, 2H), 5.10 (s, 2H), 5.22 (s, 2H),
5.83 (d, 1H), 7.32 (m, 15H), 9.26 (b s, 1H), 9.42 (b s, 1H).
The abbreviations used are as follows: TBTU, O-(benzo-
triazol-1-yl)-N,N,N,N′-tetramethyluronium tetrafluoroborate;
Cbz or Z, benzyloxycarbonyl; DMF, N,N-dimethylformamide;
Boc, tert-butyloxycarbonyl; THF, tetrahydrofuran; DCU, di-
cyclohexylurea; DiPEA, N,N-diisopropylethylamine; Et3N, tri-
ethylamine; TBAF, tetrabutylammonium fluoride; HOBt, 1-hy-
droxybenzotriazole; DCCI, dicyclohexylcarbodiimide; DMSO,
dimethyl sulfoxide; Pmc, 2,2,5,7,8-pentamethylchroman-6-
sulfonyl; MS, mass spectrometry; FAB, fast atom bombart-
ment; ESI, electron spray ionization.
Reagent-grade solvents were used without further purifica-
tion. Evaporation means removal of solvent by use of a Bu¨chi
rotary evaporator at 40-50 °C in vacuo. Silica gel used for
chromatography was Kieselgel-60 (230-400 mesh). Thin-layer
chromatography (TLC) plates coated with silica gel 60 F254
(Merck) were used; detection was by UV (254 nm), by USUI,
or by treatment with Cl2/TMB. NMR spectra (1H NMR and
13C NMR) were recorded on either a Bruker AC 200 (200 MHz),
Bruker AM 360 (360 MHz), or a Bruker DRX 400 (400 MHz)
spectrometer; δ values in parts per million relative to tetra-
methylsilane are given. Mass spectra were recorded with a
Finnigan MAT 90 mass spectrometer (FAB) or a Perkin-Elmer
Sciex API-165 apparatus (ESI). Analytical HPLC was done on
a Hewlett Packard 1090M or 1100 liquid chromatograph.
Reversed phase preparatory HPLC purification was performed
on a Waters Prep 4000, using a Supelcosil LC-18-DB or Delta

Potent Oral Antithrombotics
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4427
3,8-Dia m in o-6,7,8,8a -t et r a h yd r o-8a -h yd r oxyim id a zo-
[1,5a ]p yr id in -1(5H)-on e. Hydrogen gas was passed through
a mixture of 1.5 g (2.37 mmol) of 3-(benzyloxycarbonyl-amino)-
6-(dibenzoylcarbonylguanidino)-2-oxo-hexanoic acid methyl-
ester, 150 mg of Pd/C 10%, and 2.44 mL of 1 N HCl in DMF
(60 mL). Hydrogen was passed through until the reaction was
complete. The catalyst was filtered off, and the filtrate was
evaporated in vacuo to give a white foam (0.8 g, quantitative).
TLC (n-BuOH:pyridine:HOAc:water, 10:1:1:2) Rf ) 0.2. 1H
NMR (200 MHz, CDCl3): δ 1.65 (m, 1H), 1.83 (m, 1H), 1.85
(m, 1H), 3.32 (d t, 1H), 3.63 (d d, 1H), 3.78 (d d, 1H).
(Cbz)-(2-thiazolyl) as a clear oil (8.6 g, 99.0%). TLC (EtOAc:
heptane, 3:1) Rf ) 0.77. H NMR (200 MHz, CDCl3): δ 1.38-
1.76 (m, 6H), 1.42 (s, 9H), 3.13-3.25 (m, 2H), 4.83-4.97 (m,
1H), 5.08 (s, 2H), 5.37-5.48 (m, 2H), 7.27-7.40 (m, 5H), 7.70
(d, 1H), 8.03 (d, 1H).
Boc-D-P h e-P r o-Lys-(2-th ia zolyl). Boc-Lys(Cbz)-(2-thiaz-
olyl) (500 mg, 1.12 mmol) was dissolved in 50% TFA/CH2Cl2
(5 mL) and stirred for 1 h at room temperature. The crude
H-Lys(Cbz)-(2-thiazolyl).TFA was isolated in quantitative yield
after removal of the solvent by evaporation and used im-
mediately in the next step. TLC (EtOAc:pyridine:HOAc:H2O,
63:20:6:11) Rf ) 0.25.
1
N8(D-P h en ylalan yl-pr olyl)-3,8-diam in o-6,7,8,8a-tetr ah y-
d r o-8a -h yd r oxyim id a zo[1,5a ]p yr id in -1(5H)-on e (4). A so-
lution of Boc-D-Phe-Pro-OH (190 mg, 0.52 mmol), HOBt (108
mg, 0.80 mmol), and DCCI (121 mg, 0.58 mmol) in DMF (10
mL) was stirred for 15 min at 0 °C. 3,8-Diamino-6,7,8,8a-
tetrahydro-8a-hydroxyimidazo[1,5a]pyridin-1(5H)-one (300 mg,
0.47 mmol) was added, and the mixture was stirred for 16 h
at room temperature. The reaction mixture was evaporated
in vacuo to a small volume and diluted with n-BuOH. The
solution was washed with water, 5% NaHCO3 solution, and
brine and evaporated in vacuo. The residue was dissolved in
12.5 mL of 90% TFA and 0.36 mL of anisole, and the reaction
mixture was stirred for 2.5 h at room temperature. The
volatiles were removed, and the residue was dissolved in
t-BuOH:water, 1:1. Dowex-2X-8 (acetate form) was added in
portions until the pH of the solution was raised to 5-6. The
ion exchange resin was filtered off, and the filtrate was
lyophilized to give 500 mg of crude 4. An amount of 190 mg of
the residue was purified by RP-18 preparatory HPLC. The
fractions containing the desired product were collected, de-
salted, and lyophilized to yield 4 (42 mg). TLC (n-BuOH:
pyridine:HOAc:H2O, 6:1:1:2) Rf ) 0.7. 1H NMR (360 MHz,
D2O): δ 1.60 (m, 1H), 1.65 (m, 1H), 1.85 (m, 3H), 2.10 (m, 2H),
2.76 (m, 1H), 3.15 (m, 1H), 3.24 (m, 1H), 3.25 (m, 1H), 3.51
(m, 1H), 3.77 (dd, 1H), 3.92 (dd, 1H), 4.36 (dd, 1H), 4.54 (dd,
1H). 13C NMR (360 MHz, D2O): δ 190.6, 176.0, 170.7, 167.7,
87.4, 63.5, 56.1, 53.5, 50.8, 40.9, 39.5, 32.4, 27.8, 26.9, 26.4. Rt
(LC): 22.13 min, 20% A/80% B to 20% A/40% B/40% C in 40
min; purity 94.6% (LC). MS m/z (FAB+): 429 [M + H], 451 [M
+ Na], 411 [M - H2O]; (FAB-): 427 [M - H], 409 [M - H -
H2O].
To a cold (0 °C) solution of Boc-D-Phe-Pro-OH (350 mg, 1
mmol) in DMF (5 mL) were successively added HOBt (145 mg,
1.1 mmol), DCCI (220 mg, 1.1 mmol), and H-Lys(Cbz)-(2-
thiazolyl)‚TFA (445 mg, 1.1 mmol). The pH was adjusted to
pH 8.5 by adding Et3N. The reaction mixture was stirred at 0
°C for 1 h and then kept overnight at room temperature. The
mixture was cooled to -20 °C, and DCU was removed by
filtration. The filtrate was evaporated to dryness. The residue
was dissolved in EtOAc and washed successively with 5%
NaHCO3 solution, 5% KHSO4 solution, and brine, dried over
Na2SO4, and concentrated in vacuo. Purification of the residue
by chromatography (CH2Cl2:MeOH, 95:5) afforded Boc-D-Phe-
Pro-Lys-(2-thiazolyl) (272 mg, 40.9%). TLC (CH2Cl2:MeOH,
1
9:1) Rf ) 0.80. H NMR (200 MHz, CDCl3): δ 0.80-2.50 (m,
10H), 1.41 (s, 9H), 2.70-2.78 (d, 2H), 3.00-3.60 (m, 5H), 4.27-
4.35 (m, 1H), 4.51-4.60 (m, 1H), 5.13 (s, 2H), 5.55-5.67 (m,
2H), 5.59-6.04 (m, 1H), 7.08-7.50 (m, 10H), 7.70 (d, 1H), 8.03
(d, 1H).
H-D-P h e-P r o-Lys-(2-th ia zolyl) (6). Boc-D-Phe-Pro-Lys-(2-
thiazolyl) (260 mg, 0.38 mmol) was treated with TFA:thioani-
sole, 10:1 v/v (2.75 mL), for 4 h at room temperature. The
reaction mixture was concentrated in vacuo, and the residue
was dissolved in water. The aqueous phase was washed
extensively with diethyl ether. The water layer, containing
H-D-Phe-Pro-Lys-(2-thiazolyl), was charged onto a preparative
HPLC Supelcosil LC-18-DB column using a gradient elution
system of 20% A/80% B to 20% A/30% B/50% C over 40 min,
at a flow rate of 20 mL/min. (A: 0.5 M sodium phosphate
buffer, pH 2.1; B: water; C: acetonitril/water; 3/2 v/v).
Fractions containing 6 were desalted using a 0.1 M HCl
solution and lyophilized. Yield, 117.5 mg of H-D-Phe-Pro-Lys-
(2-thiazolyl) (6, 68.3%). 1H NMR (360 MHz, D2O): δ 1.58-
1.70 (m, 3H), 1.74-1.94 (m, 5H), 2.11-2.22 (m, 2H), 2.78-
2.86 (dt, 1H), 3.06 (t, 2H), 3.17-3.35 (m, 2H), 3.50-3.59 (m,
1H), 4.48 (m, 1H), 4.60 (m, 1H), 5.53 (m, 1H), 7.30-7.52 (m,
5H), 8.14 (d, 1H), 8.18 (d, 1H). Rt (LC): 25.67 min, 20% A/80%
B to 20% A/20% B/60% C in 40 min; purity 97.1% (LC). MS
m/z (FAB+): 458 [M + H].
Gen er a l P r oced u r e for th e P r ep a r a tion of P ep tid yl-
Lys-(2-th ia zolyl) Der iva tives (II, Sch em e 2A). P r ep a r a -
tion of Boc-Lys(Cbz)-(2-th ia zolyl) (Ia , Sch em e 2A). Boc-
Lys(Cbz)-(N-Me)OMe. Boc-Lys(Cbz)-OH‚DCHA (10 g, 17.8
mmol) was suspended in CH2Cl2 (200 mL). The suspension was
washed with cold 0.1 N HCl solution twice. TBTU (6.0 g, 18.7
mmol) and O,N-dimethyl-hydroxylamine hydrochloric acid
(1.82 g, 18.7 mmol) were added to the resulting organic phase,
and the pH was adjusted to pH 8 by adding Et3N. The reaction
mixture was stirred for 1 h at room temperature. The mixture
was washed successively with cold 2 N HCl solution, water,
5% NaHCO3 solution, and water. The organic layer was dried
over Na2SO4, filtered, and evaporated. Purification of the
residue by chromatography (CH2Cl2:MeOH, 1:1) afforded Boc-
Lys(Cbz)-NMeOMe as a clear oil (7.2 g, 95.5%). TLC (CH2Cl2:
In an analogues way to the synthesis of 6, compounds 8-11,
13, and 15 were prepared from the corresponding dipeptides.
Eth ylSO2-D-Ch a -P r o-Lys-(2-th ia zolyl) (8). 1H NMR (400
MHz, DMSO-d6): δ 0.8-2.12 (m, 23H), 1.09 (t, 3H), 2.7-2.81
(m, 2H), 2.82-2.93 (m, 2H), 3.42 (m, 1H), 3.59 (dt, 1H), 4.10
(dt, 1H), 4.38 (dd, 1H), 5.26-5.37 (m, 1H), 8.19 (d, 1H), 8.28
(d, 1H). Rt (LC): 25.66 min 20% A/60%B/20% C to 100% C in
40 min; purity 99.3% (LC). MS m/z (ESI+): 556 [M + H], 578
[M + Na], 538 [MH - H2O].
1
MeOH, 95:5) Rf ) 0.55. H NMR (200 MHz, CDCl3): δ 1.38-
1.76 (m, 6H), 1.42 (s, 9H), 3.13-3.25 (m, 2H), 3.20 (s, 3H), 3.76
(s, 3H), 4.58-4.74 (m, 1H), 4.84-4.98 (m, 1H), 5.09 (s, 2H),
5.19-5.29 (d, 1H), 7.28-7.40 (m, 5H).
1
H-D-Ch a -(N-Cyclop r op yl)-Gly-Lys-(2-th ia zolyl) (9). H
NMR (400 MHz, D2O): δ 0.72-1.81 (m, 22H), 1.97-2.06 (m,
1H), 2.79-2.85 (m, 1H), 2.92 (t, 2H), 4.13-4.25 (m, 2H), 4.79
(dd, 1H), 5.41 (dd, 1H), 8.01 (d, 1H), 8.06 (d, 1H). Rt (LC): 30.48
min 20% A/80%B to 20% A/20% B/60% C in 40 min; purity
99.9% (LC). MS m/z (FAB+): 464.1 [M + H].
Boc-Lys(Cbz)-(2-th ia zolyl) (Ia ). To a cold (-78 °C), stirred
solution of n-butyllithium (63.9 mmol) in diethyl ether (58 mL)
was added dropwise a solution of 2-bromothiazole (10.5 g, 63.9
mmol) in diethyl ether (30 mL). The solution was stirred at
-78 °C for 30 min, after which a solution of Boc-Lys(Cbz)-
NMeOMe (8.2 g, 19.4 mmol) in dry THF (75 mL) was added
slowly. The mixture was stirred at -78 °C for 1 h, and then,
5% NaHCO3 solution was added. The mixture was allowed to
warm to room temperature, and the layers were separated.
The aqueous layer was extracted with diethyl ether. The
combined organic layers were washed with water, dried over
Na2SO4, filtered, and evaporated. Purification of the residue
by chromatography (EtOAc:heptane, 3:1) afforded Boc-Lys-
1
In d a n e-(N-Cyclop r op yl)-Gly-Lys-(2-th ia zolyl) (10). H
NMR (400 MHz, CD3OD): δ 0.93 (broad m, 2H), 1.10 (broad,
2H), 1.58 (m, 2H), 1.66-1.88 (m, 3H), 2.08-2.17 (m, 1H), 2.94
(m, 2H), 3.05 (broad, 1H), 3.27-3.35 (m, 2H), 3.40-3.49 (m,
2H), 4.08 (broad, 2H), 4.49 (m, 1H), 5.58 (dd, 1H), 7.20-7.30
(m, 4H), 8.08 (d, 1H), 8.13 (d, 1H). Rt (LC): 31.71 min 20%
A/80%B to 20% A/20% B/60% C in 40 min; purity 99.7% (LC).
MS m/z (FAB+): 427 [M + H], 853 [2M + H]; (FAB-): 426
[M-], 426 [M + Cl].

4428 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Adang et al.
3,3-Dip h en ylp r op ion yl-P r o-Lys-(2-th ia zolyl) (11). 1H
NMR (400 MHz, D2O): δ 1.47-2.32 (m, 10H), 2.72-2.89 (m,
1H), 2.90-3.00 (t, 1H), 3.02-3.18 (m, 1H), 3.29-3.40 (m, 1H),
3.40-3.57 (m, 1H), 3.60-3.70 (m, 1H), 4.31 (dd, 1H), 4.44-
4.55 (m, 1H), 5.37-5.47 (m, 1H), 7.21-7.45 (m, 10H), 8.07 (d,
1H), 8.11 (d, 1H). Rt (LC): 32.89 min 20% A/60%B/20% C to
20% A/80% C in 40 min; purity 97.9% (LC). MS m/z (FAB+):
519 [M + H], 611 [M + glycerol + H], 1037 [2M + H]; (FAB-):
517 [M - H].
were successively added HOBt (7.09 g, 52.5 mmol), DCCI (7.61
g, 36,8 mmol), H-Pro-OBzl.HCl (9.31 g, 38.5 mmol), and Et3N
(6 mL, 43.4 mmol). The mixture was stirred at 0 °C for 1 h
and then kept at room temperature overnight. The mixture
was cooled to -20 °C, and DCU was removed by filtration.
The filtrate was evaporated to dryness. The residue was
dissolved in EtOAc and washed successively with 5% NaHCO3
solution, water, 3% citric acid solution, and brine, dried over
Na2SO4, and concentrated in vacuo. Purification of the residue
by chromatography (heptane:EtOAc, 3:1) afforded N-(tert-
butyloxycarbonylmethyl)-N-Boc-D-Cha-Pro-OBzl (15 g, 75%).
TLC (EtOAc:heptane, 1:1) Rf ) 0.7. 1H NMR (200 MHz,
CDCl3): δ 0.80-2.35 (m, 35H), 3.28-4.03 (m, 4H), 4.45-5.25
(m, 4H), 7.30-7.41 (m, 5H).
N-(ter t-Bu t yloxyca r b on ylm et h yl)-N-Boc-D-Ch a -P r o-
OH. An amount of 10% palladium on charcoal (750 mg) was
added to a solution of N-(tert-butyloxy-carbonylmethyl)-N-Boc-
D-Cha-Pro-OBzl (15 g, 26.2 mmol) in MeOH (150 mL). The
mixture was hydrogenated at atmospheric pressure at room
temperature for 1 h. The palladium catalyst was removed by
filtration, and the solvent was removed by evaporation at
reduced pressure yielding 11.2 g of N-(tert-butyloxycarbonyl-
methyl)-N-Boc-D-Cha-Pro-OH (89%). TLC (EtOAc:pyridine:
HOAc:H2O, 213:20:6:11) Rf ) 0.65. 1H NMR (200 MHz,
CDCl3): δ 0.77-2.39 (m, 35H), 3.30-4.04 (m, 4H), 4.48-5.18
(m, 2H).
7-Met h oxy-2-n a p h t h ylsu lfon yl-Lys-(2-t h ia zolyl) (13).
1H NMR (400 MHz, CD3OD): δ 1.55-1.91 (m, 6H), 2.85-2.97
(m, 2H), 3.88 (s, 3H), 5.20-5.25 (m, 1H), 7.09 (m, 1H), 7.22
(m, 1H), 7.55 (m, 1H), 7.73-7.82 (m, 3H), 8.00 (d, 1H), 8.11
(d, 1H). Rt (LC): 34.06 min 20% A/60%B/20% C to 20% A/80%
C in 40 min; purity 88.5% (LC). MS m/z (FAB+): 434 [M +
H], 526 [M + glycerol + H], 618 [M + 2glycerol + H].
1
H-D-Ch a -P r o-Lys-(2-th ia zolyl) (15). H NMR (360 MHz,
DMSO-d6): δ 0.80-2.13 (m, 24H), 2.77 (t, 2H), 3.68-3.89 (m,
1H), 4.14 (m, 1H), 4.41 (dd, 1H), 5.28 (m, 1H), 8.18 (d, 1H),
8.27 (d, 1H). Rt (LC): 30.59 min 20% A/80% B to 20% A/20%
C/60% C in 40 min; purity 97.8% (LC). MS m/z (FAB+): 464
[M + H]; (FAB-): 463 [M-], 498 [M + Cl].
P r ep a r a tion of HOOC-CH2-D-Ch a -P r o-Lys-(2-th ia zolyl)
(18). P r ep a r a tion of th e Dip ep tid e N-(ter t-Bu tyloxyca r -
bon ylm eth yl)-N-Boc-D-Ch a -P r o-OH. H-D-Ch a -OMe‚HCl.
To cold (-20 °C) and dry MeOH (195 mL) was added dropwise
thionyl chloride (28 mL, 38.6 mmol). H-D-Cha-OH‚HCl (40 g,
19.3 mmol) was added, and the reaction mixture was heated
under reflux for 5 h. The mixture was concentrated in vacuo
and coevaporated with MeOH (three times). The residue was
crystallized from MeOH/Et2O yielding H-D-Cha-OMe‚HCl as
a crystalline powder. Yield, 40.9 g (96%). TLC (n-BuOH:HOAc:
In an analogues way to the synthesis of 6, compound 18
was prepared from the corresponding dipeptide.
HOOC-CH2-D-Ch a -P r o-Lys-(2-th ia zolyl) (18). 1H NMR
(400 MHz, D2O): δ 0.59-2.31 (m, 23H), 2.79-2.90 (m, 2H),
3.37-3.65 (m, 4H), 4.11-4.40 (m, 2H), 5.23-5.36 (m, 1H),
7.49-8.00 (m, 2H). Rt (LC) ) 32.67 min 20% A/80% B to 20%
A/20% B/60% C in 40 min; purity 97.1% (LC). MS m/z (ESI+):
522.1 [M + H].
1
H2O, 10:1:3) Rf ) 0.66. H NMR (200 MHz, CDCl3): δ 0.85-
2.00 (m, 13H), 3.84 (s, 3H), 4.07 (dd, 1H).
N-(ter t-Bu tyloxyca r bon ylm eth yl)-D-Ch a -OMe. tert-Bu-
tyl-bromo acetate (36 g, 0.185 mol) was added to a stirred
solution of H-D-Cha-OMe‚HCl (40.9 g, 0.185 mol) in 400 mL
of acetonitrile. The pH of the mixture was adjusted to 8.5 with
DiPEA. The mixture was stirred for 16 h at room temperature
and evaporated in vacuo. The residue was dissolved in CH2-
Cl2, and the solution was washed with water, dried over Na2-
SO4, and evaporated in vacuo. Purification of the residue by
chromatography (heptane:EtOAc, 9:1) gave 64 g of N-(tert-
butyloxycarbonylmethyl)-D-Cha-OMe (115% yield). TLC (EtOAc:
P r ep a r a tion of P ep tid yl-Ar g-(2-th ia zolyl) Der iva tive.
P r ep a r a tion of Boc-Ar g(P m c)-(2-th ia zolyl) (Ib, Sch em e
2A). In an analogues way to the synthesis of Boc-Lys(Cbz)-
(2-thiazolyl), Ib was prepared starting from Boc-Arg(PMc)-OH.
TLC (CH2Cl2:MeOH, 9:1) Rf ) 0.82. 1H NMR (400 MHz,
CDCl3): δ 1.30 (s, 6H), 1.40 (s, 9H), 1.60-1.73 (bm, 4H), 1.80
(t, 2H), 2.09 (s, 3H), 2.53 (s, 3H), 2.54 (s, 3H), 2.51 (t, 2H),
3.22 (m, 1H), 3.38 (b, 1H), 5.40 (bt, 1H), 5.64 (d, 1H), 6.23 (bs,
1H), 7.71 (d, 1H), 8.03 (d, 1H). 13C NMR (400 MHz, CDCl3): δ
191.9, 164.2, 164.1, 156.2, 153.5, 145.3, 135.5, 134.8, 133.5,
127.0, 124.0, 117.9, 80.3, 73.6, 40.6, 32.9, 28.3, 26.8, 25.4, 21.4,
18.5, 17.4, 12.1. MS m/z (ESI+): 608.4 [M + H], 552.2 [M -
tBu + H], 508 [M - Boc + H], 1215.8 [2M + H]; (ESI-): 606.4
[M - H].
1
heptane, 1:1) Rf ) 0.25. H NMR (200 MHz, CDCl3): δ 0.84-
1.96 (m, 22H), 3.16-3.40 (m, 3H), 3.72 (s, 3H).
N-(ter t-Bu t yloxyca r b on ylm et h yl)-N-Boc-D-Ch a -OMe.
The pH of a solution of N-(tert-butyloxycarbonylmethyl)-D-Cha-
OMe (64 g, 0.185 mol) and di-tert-butyl dicarbonate (40.3 g,
0.185 mol) in 500 mL of DMF was adjusted to 8.5 with DiPEA.
The mixture was stirred for 16 h at room temperature. The
solvent was removed in vacuo. CH2Cl2 and water were added
to the residue. The organic layer was separated and washed
with cold 1 N HCl solution, water, 5% NaHCO3 solution, and
water. The organic layer was dried over Na2SO4, and the
filtrate was evaporated to afford an amorphous solid of N-(tert-
butyloxycarbonylmethyl)-N-Boc-D-Cha-OMe. Yield, 59.6 g (70%).
TLC (EtOAc:heptane, 1:1) Rf ) 0.50. 1H NMR (200 MHz,
CDCl3): δ 0.78-1.90 (m, 31H), 3.55-3.99 (m, 5H), 4.64-5.03
(m, 1H).
In an analogous way to the synthesis of 6, compound 5 was
prepared starting from Boc-D-Phe-Pro-OH.
H-D-P h e-P r o-Ar g-(2-th ia zolyl) (5). 1H NMR (400 MHz,
D2O): δ 1.60-2.30 (complex, 10H), 2.80-2.90 (dt, 2H), 3.20-
3.40 (m, 3H), 3.55-3.65 (m, 1H), 4.52 (m, 1H), 4.65 (m, 1H),
5.48 (m, 1H), 7.28-7.57 (m, 5H), 8.18 (d, 1H), 8.23 (d, 1H). Rt
(LC) ) 13.04 min 20% A/60% B/20% C to 20% A/80% C in 40
min; purity 94.0% (LC). MS m/z (FAB+): 486 [M + H].
P r ep a r a tion of P ep tid yl-(E-h yd r oxy)-Nle-(2-th ia zolyl)
Der iva tive. P r ep a r a tion of Boc-(E-h yd r oxy)-Nle-(2-th ia z-
olyl) (Ic, Sch em e 2A). In an analogues way to the synthesis
of Boc-Lys(Cbz)-(2-thiazolyl), Ic was prepared starting from
Boc-(ꢀ-hydroxy)-Nle-OH. TLC (EtOAc:heptane, 4:1) Rf ) 0.50.
1H NMR (200 MHz, CDCl3): δ 1.40-2.16 (m, 6H), 1.42 (s, 9H),
3.60-3.69 (t, 2H), 5.36-5.52 (b, 2H), 7.72 (d, 1H), 8.05 (d, 1H).
MS m/z (FAB+): 315.0 [M + H], 258.9 [M - tBu + H], 214.9
[M - Boc + H]; (FAB-): 313.9 [M - H].
N-(ter t-Bu tyloxyca r bon ylm eth yl)-N-Boc-D-Ch a -OH. A
solution of N-(tert-Butyloxycarbonylmethyl)-N-Boc-D-Cha-OMe
(59.6 g, 0.15 mol) in 900 mL of dioxane/water ) 9/1 (v/v) was
treated with sufficient 6 N NaOH solution to keep the pH at
12 for 6 h at room temperature. After it was acidified, the
mixture was poured into water and was extracted with CH2-
Cl2. The organic layer was washed with water and was dried
over Na2SO4. The filtrate was evaporated and yielded 54 g of
N-(tert-butyloxycarbonylmethyl)-N-Boc-D-Cha-OH (93%). TLC
In an analogues way to the synthesis of 6, compound 12
was prepared from the corresponding dipeptide.
3,3-Dip h en ylp r op ion yl-P r o-(E-h yd r oxy)-Nle-(2-t h ia z-
olyl) (12). 1H NMR (400 MHz, CD3OD): δ 1.45-2.19 (m, 10H),
2.83-3.27 (m, 2H), 3.32-3.61 (m, 2H), 3.53 (t, 2H), 4.32-4.40
(m, 1H), 4.53-4.60 (m, 1H), 5.47-5.56 (m, 1H), 7.13-7.29 (m,
10 H), 8.00 (d, 1H), 8.07 (d, 1H). Rt (LC): 39.43 min 20%
A/60%B/20% C to 20% A/80% C in 40 min; purity 100% (LC).
1
(CH2Cl2:MeOH, 9:1) Rf ) 0.6. H NMR (200 MHz, CDCl3): δ
0.82-2.05 (m, 31H), 3.48-4.02 (m, 2H), 4.27-4.50 (m, 1H).
N-(ter t-Bu t yloxyca r b on ylm et h yl)-N-Boc-D-Ch a -P r o-
OBzl. To a cold (0 °C) solution of N-(tert-butyloxycarbonyl-
methyl)-N-Boc-D-Cha-OH (13.5 g, 35 mmol) in DMF (150 mL)

Potent Oral Antithrombotics
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4429
MS m/z (FAB+): 520 [M + H], 612 [M + glycerol + H], 1039
[2M + H]; MS (FAB-): 519 [M-].
sively with cold 1 N HCl solution, water, 5% NaHCO3 solution,
and water and dried over Na2SO4. The filtrate was evaporated,
and the residue was chromatographed (heptane:EtOAc, 4:1).
The fractions containing Cbz-Lys(Boc)-OMe were pooled and
evaporated to afford 29.1 g of the title compound (99%). TLC
P r ep a r a tion of P ep tid yl-Lys-(2-oxa zolyl) Der iva tive.
P r ep a r a tion of Boc-Lys(Cbz)-(2-oxa zolyl) (Id , Sch em e
2A). Boc-Lys(Cbz)ψ[CHOH]-(2-oxa zolyl). To a solution of
975 mg (2.47 mmol) of Boc-Lys(Cbz)-OMe in 25 mL of CH2Cl2
at -78 °C under a nitrogen atmosphere was added 6 mL of a
1 M DiBAL-H solution in hexane. After 15 min, the reaction
was completed, and the mixture was poured into 150 mL of
2% citric acid solution and filtered. The organic layer was
separated, washed with water and brine, dried over MgSO4,
and concentrated. The residue was coevaporated with toluene
to give 920 mg of Boc-Lys(Cbz)-H. This aldehyde (890 mg, 2.44
mmol) was dissolved in 1.4 mL of toluene and 900 mg (6.37
mmol) of 2-(trimethylsilyl)oxazole, prepared according to
Edwards et al,43 was added, and heated at 80 °C. After 60 h,
the reaction mixture was concentrated, the residue was
dissolved in 5 mL of THF, treated with 3 mL of a 3 M TBAF
in THF solution, and stirred at room temperature for 2 h. The
mixture was concentrated, dissolved in EtOAc, washed with
3% NaHCO3 solution and brine, dried over MgSO4, and
evaporated. Purification of the residue by chromatography
(gradient of EtOAc:CH2Cl2, 2:1 to ethyl acetate) afforded an
oil that was rechromatographed (gradient of EtOAc:heptane,
1:1 to EtOAc:heptane, 1:3) to give the title compound (220 mg,
18%). TLC (EtOAc) Rf ) 0.7. 1H NMR (200 MHz, CDCl3): δ
0.82-1.71 (m, 15H), 3.07-3.24 (m, 2H), 4.0 (br. s, 1H), 4.77-
5.40 (m, 6H), 7.03 (s, 0.75H) and 7.06 (s, 0.25H), 7.23-7.38
(m, 5H), 7.61 (s, 1H).
1
(EtOAc:heptane, 3:1) Rf ) 0.85. H NMR (200 MHz, CDCl3):
δ 1.42 (s, 9 H), 3.10 (br q, 2 H), 3.74 (s, 3 H), 4.38 (br dd, 1 H),
4.57 (br s, 1 H), 5.11 (s, 2 H), 5.38 (br d, 1H), 7.36 (s, 5 H).
Cbz-Lys(Boc)E[cya n oa ceta te] (III, Sch em e 2B). To a
cold (-78 °C) solution of 29.1 g (74 mmol) of Cbz-Lys(Boc)-
OMe in 800 mL of dry CH2Cl2 was added dropwise 222 mL of
1 M solution of DiBAL-H in hexane keeping the reaction
temperature below -70° C. The resulting solution was stirred
at -78° C for 1 h, and 600 mL of an 5% citric acid solution
was added to the reaction mixture. The two layer mixture was
stirred at room temperature for 10 min, the layers were
separated, and the aqueous layer was extracted twice with
CH2Cl2. The combined CH2Cl2 layers were washed with water,
dried over Na2SO4, and filtered. The filtrate was stirred under
a nitrogen atmosphere and cooled on an icewater bath. A
solution of 36.3 g of NaCN (740 mmol) and 4.2 g of triethyl-
benzylammonium chloride (18.4 mmol) in 600 mL of water was
added. Under vigorous stirring, acetic anhydride was added
portionwise (2 × 9 mL) over a period of 30 min. The organic
layer was separated, and the aqueous layer was extracted
twice with CH2Cl2. The combined CH2Cl2 layers were washed
with water, dried over Na2SO4, filtered, and evaporated in
vacuo. Purification of the residue by chromatography (EtOAc:
heptane, 1:1) afforded 26.3 g of Cbz-Lys (Boc)Ψ[cyanoacetate]
(82%). TLC (CH2Cl2:EtOAc, 7:3) Rf ) 0.6. 1H NMR (200 MHz,
CDCl3): δ 1.42 (s, 9 H), 2.09 and 2.12 (2 × s, 3 H, diastereo-
mers), 3.10 (complex, 2 H), 4.03 (complex, 1 H), 4.59 (br s, 1
H), 5.11 (s, 2 H), 5.20 (complex, 1 H), 5.45 (br dd, 1H), 7.36 (s,
5 H).
Boc-Lys(Cbz)-(2-oxa zolyl) (Id ). To a solution of 0.22 g
(0.51 mmol) of Boc-Lys(Cbz)Ψ[CHOH]-(2-oxazolyl) in 10 mL
of CH2Cl2 was added 0.22 g (0.52 mmol) of periodinane (Dess-
Martin reagent). After it was stirred at room temperature for
1.5 h, 10 mL of aqueous 5% Na2S2O4 solution was added and
the mixture was stirred for 15 min at room temperature. The
organic layer was separated, washed with water, 5% NaHCO3
solution, and brine, dried over MgSO4, and concentrated.
Purification of the residue by chromatography (heptane:EtOAc,
1:1) yielded 162 mg of the title compound (79% yield). TLC
Cb z-Lys(Boc)E[CHOHCO]-OMe (IV, Sch em e 2B). A
solution of 26.3 g (62 mmol) of Cbz-Lys(Boc)Ψ[cyanoacetate]
in 600 mL of Et2O:MeOH, 3:1, was cooled to -20° C under a
nitrogen atmosphere, and 66 g of gaseous hydrogen chloride
was introduced keeping the temperature below -5° C. The
reaction mixture was kept at 4° C overnight. Water (100 mL)
was added dropwise to the reaction mixture keeping the
temperature below 5° C. After it was stirred for 16 h at room
temperature, the organic layer was separated and washed with
water. The aqueous layer was saturated with NaCl and
extracted with sec-BuOH:CH2Cl2, 3:2. The organic phase was
washed with brine, dried over Na2SO4, filtered, and evaporated
in vacuo to give 25.4 g of the crude amine. This was treated
with di-tert-butyl dicarbonate (16 g, 73 mmol) in 400 mL of
DMF in the presence of Et3N (pH 8) at room temperature
overnight. The solvent was removed by evaporation at reduced
pressure. The residue was dissolved in EtOAc, washed with
water and brine successively, dried over Na2SO4, filtered, and
evaporated in vacuo. Purification of the crude product by
chromatography (heptane:EtOAc, 6:4) gave Cbz-Lys(Boc)Ψ-
[CHOHCO]-OMe (15.8 g, 60%). TLC (EtOAc:pyridine:HOAc:
H2O, 63:20:6:11) Rf ) 0.75. 1H NMR (200 MHz, CDCl3): δ 1.42
(s, 9 H), 3.10 (complex, 3 H), 3.74 and 3.80 (2 × s, 3 H,
diastereomers), 4.08 (complex, 1 H), 4.18 and 4.33 (2 × dd, 1
H, diastereomers), 4.56 (br s, 1 H), 4.97 (br d, 1 H), 5.05 and
5.11 (2 × s, 2 H, diastereomers), 7.35 (m, 5H).
1
(EtOAc:heptane, 3:1) Rf ) 0.5. H NMR (200 MHz, CDCl3): δ
1.33-2.12 (m, 15H), 3.13-3.26 (m, 2H), 4.84-5.47 (m, 6H),
7.29-7.42 (m, 6H), 7.84 (d, 1H).
In an analogues way to the synthesis of 6, compound 19
was prepared from the corresponding dipeptide.
HOOC-CH2-D-Ch a -P r o-Lys-(2-oxa zolyl) (19). 1H NMR
(400 MHz, D2O): δ 0.83-2.56 (m, 23H), 3.04-3.17 (2H, m),
3.62-3.91 (m, 4H), 4.32-4.65 (m, 2H), 5.31-5.46 (m, 1H),
7.29-8.28 (m, 2H). Rt (LC): 28.46 min 20% A/80% B to 20%
A/20% B/60% C in 40 min; purity 99.3% (LC). MS m/z (FAB+):
506 [M + H], 524 [M + H2O + H], 598 [M + glycerol + H],
1011 [2M + H], 1029 [2M + H2O + H]; (FAB-): 504 [M - H],
540 [M + Cl].
P r ep a r a t ion of 7-Met h oxyn a p h t h ylsu lfon yl-Lys-P i-
p er id yla m id e. (S)-1-[6-Am in o-2-[[(7-m eth oxy-2-n a p h th a -
len e)su lfon yl]am in o-1-oxoh exyl]-4-m eth yl-piper idin e (14).
The procedure described for compound 6 was used. A 300 mg
amount of 7-methoxy-2-naphthylsulfonyl-Lys(Boc)-(4-meth-
ylpiperidyl) afforded, after chromatography (CH2Cl2:MeOH, 98:
2) and exchange on Dowex-Cl-, 118 mg of the title compound
1
(44% yield). H NMR (400 MHz, CD3OD): δ -0.24-0.99 (2×
Boc-D-P h e-P r o-Lys(Boc)ψ[CHOHCO]-OMe. A solution of
Cbz-Lys(Boc)Ψ[CHOHCO]-OMe (5.52 g, 13 mmol), 0.55 g of
10% palladium on charcoal, and 6.5 mL of 1 N HCl solution
in 50 mL of DMF was hydrogenated for 2 h at room temper-
ature. The catalyst was filtered off, and the filtrate was
evaporated to dryness in vacuo. The crude amine was directly
used in the coupling with Boc-D-Phe-Pro-OH.
d, 2× m, 5H), 1.14-2.33 (m, 10H), 2.66-2.96 (m, 3H), 3.56-
4.28 (m, 3H), 3.94 (2× s, 3H), 7.31 (m, 1H), 7.41 (broad, 1H),
7.61-7.67 (m, 1H), 7.88 (dd, 2H), 8.30 (s, 1H). Rt (LC): 31.17
min 20% A/60%B/20% C to 20% A/80% C in 40 min; purity
98.7% (LC). MS m/z (FAB+): 448 [M + H], 895 [2M + H];
(FAB-): 446 [M - H], 482 [M + Cl], 518 [M + 2HCl - H], 929
[2M + Cl].
Gen er a l P r oced u r e for th e P r ep a r a tion of P ep tid yl-
Lysψ[COCO]-X Der iva tives (V, Sch em e 2B). P r ep a r a tion
of Cbz-Lys(Boc)E[CHOHCO]-OMe (IV, Sch em e 2B). Cbz-
Lys(Boc)-OMe. A solution of 28 g (74 mmol) of Cbz-Lys(Boc)-
OH and 23.6 g (74 mmol) of TBTU in 500 mL of CH2Cl2:MeOH,
9:1, was adjusted to pH 8 by addition of Et3N and stirred for
2 h at room temperature. The mixture was washed succes-
A solution of Boc-D-Phe-Pro-OH (362 mg, 1 mmol) in 3 mL
of DMF was treated with HOBt (175 mg, 1.3 mmol) and DCCI
(265 mg, 1.3 mmol) at 0 °C for 1 h. At this temperature, a
solution of the crude H-Lys(Boc)Ψ[CHOHCO]-OMe (290 mg,
1 mmol) in 2 mL of DMF was slowly added to the reaction
mixture and the pH was set to 8 using Et3N. The reaction was
subsequently stirred for 16 h at room temperature. Next, the

4430 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Adang et al.
mixture was cooled to -20 °C and DCU was removed by
filtration after which the filtrate was concentrated in vacuo.
The residue was dissolved in EtOAc and washed with 2% citric
acid solution, water, 5% NaHCO3 solution, and brine. The
organic layers were dried over Na2SO4 and concentrated.
Purification of the residue by chromatography (CH2Cl2:MeOH,
95:5) afforded 511 mg of Boc-D-Phe-Pro-Lys(Boc)Ψ[CHOHCO]-
OMe (81%). TLC (EtOAc:pyridine:HOAc:H2O, 88:31:18:7) Rf
B/60% C in 40 min; purity 94.5% (LC). MS m/z (ESI+): 541.2
[M + H], 559.2 [M + H2O + H]; (ESI-): 539.2 [M - H].
HOOC-CH2-D-Ch a -P r o-Lysψ[COCO]-OH (20). 1H NMR
(400 MHz, D2O): δ 0.66-2.13 (m, 23H), 2.51-2.77 (2H, m),
3.16-3.55 (m, 4H), 3.82-4.72 (m, 3H). Rt (LC): 23.11 min 20%
A/80% B to 20% A/20% B/60% C in 40 min; purity 99.1% (LC).
MS m/z (FAB+): 483 [M + H], 501 [M + H2O + H], 575 [M +
glycerol + H]; (FAB-): 481 [M - H], 499 [M + H2O - H], 573
[M + glycerol - H].
1
) 0.90. H NMR (200 MHz, CDCl3): δ 1.4-1.5 (2 × s, 18 H),
HOOC-CH2-D-Ch a -Azt-Lysψ[COCO]-OH (21). 1H NMR
(400 MHz, D2O): δ 0.93 (m, 2H), 1.00-1.80 (complex, 16H),
2.22 (complex, 1H), 2.62 (complex, 1H), 2.90 (b s, 2H), 3.65
(m, 2H), 4.09 (m, 1H), 4.18 (m, 1H), 4.30 (m, 1H), 4.81 and
4.88 (2 × dd, 1H), 4.93 (dd, 1H). Rt (LC): 21.09 min 20% A/80%
B to 20% A/20% B/60% C in 40 min; purity 99.3% (LC). MS
m/z (FAB+): 469.2 [M + H], 487.3 [M + H2O + H], 561 [M +
glycerol]; (FAB-): 467.1 [M - H], 485 [M + H2O - H].
HOOC-CH2-D-Ch a -P ec-Lysψ[COCO]-OH (22). 1H NMR
(400 MHz, D2O): δ 0.82-2.20 (complex, 26H), 2.82 (m, 2H),
3.28 (tt, 1H), 3.34-3.55 (complex, 2H), 3,60 (b d, 1H), 4.06 (dd,
1H), 4.51 (m, 1H), 4.78 (dd, 1H), 4.83 and 4.90 (2 × dd, 1H).
Rt (LC): 27.25 min 20% A/80% B to 20% A/20% B/60% C in
40 min; purity 89.2% (LC). MS m/z (ESI+): 497.2 [M + H],
515.4 [M + H2O + H], 479.4 [M - H2O + H]; (ESI-): 495.2
[M - H].
HOOC-CH2-D-Ch a -Deh yd r oP r o-Lysψ[COCO]-OH (23).
1H NMR (400 MHz, D2O): δ 0.94 (m, 2H), 1.15 (m, 4H), 1.20-
1.80 (complex, 22H), 1.92 (m, 1H), 2.91 (m, 2H), 3.38-3.62
(complex, 2H), 4.11 (dd, 1H), 4.27-4.39 (complex, 2H), 4.49
and 4.53 (2 × m, 1H), 4.93 (dd, 1H), 5.10 and 5.15 (2 × m,
1H), 5.78 and 5.83 (2 × m, 1H), 6.04 (m, 1H). Rt (LC): 21.58
min 20% A/80% B to 20% A/20% B/60% C in 40 min; purity
95.9% (LC). MS m/z (ESI+): 481 [M + H], 499 [M + H2O +
H].
2.65 (complex, 1H), 3.00 (d, 2H), 3.56 (complex, 1H), 3.78 and
3.80 (2 × s, 3H), 4.12 (complex, 1H), 4.30 (complex, 1H), 4.53
(b dd, 1H), 7.26 (m, 5H). MS m/z (FAB+): 635.2 [M + H], 535.2
[M - Boc + H], 435.1 [M - 2Boc + H].
Boc-D-P h e-P r o-Lys(Boc)ψ[CH OH CO]-OH . Boc-D-Phe-
Pro-Lys(Boc)Ψ[CHOHCO]-OMe (510 mg, 0.80 mmol) was
dissolved in 20 mL of dioxane:H2O, 7:3, and treated with 0.8
mL of 2 M NaOH solution portionwise over 30 min at room
temperature, keeping the pH at 10-10.5. The reaction mixture
was diluted with water, 2 M HCl solution was added until pH
2.0, and the water layer was extracted with CH2Cl2 (3×). The
combined organic layers were washed with water and brine,
dried over Na2SO4, filtered, and concentrated in vacuo to
obtain 554 mg of Boc-D-Phe-Pro-Lys(Boc)Ψ[CHOHCO]-OH
(100% yield). TLC (EtOAc:pyridine:HOAc:H2O, 63:20:6:11) Rf
) 0.30. 1H NMR (200 MHz, CDCl3): δ 1.41 (s, 9H), 1.44 (s,
9H), 2.56 (complex, 1H), 2.98 (b d, 2H), 3.09 (b s, 2H), 4.52 (b
q, 1H), 7.24 (m, 5H). MS m/z (FAB-): 619 [M - H]; (FAB+):
621.2 [M + H], 521.2 [M - Boc + H], 421.2 [M - 2Boc + H].
Boc-D-P h e-P r o-Lys(Boc)ψ[COCO]-OH. To a solution of
Boc-D-Phe-Pro-Lys(Boc)Ψ[CHOHCO]-OH (554 mg, 0.80 mmol)
in 40 mL of CH2Cl2 was added Dess-Martin periodinane (339
mg, 0.80 mmol). After it was stirred for 1 h at room temper-
ature, 35 mL of 2% Na2S2O4 solution was added and the
mixture was stirred for 30 min at ambient temperature. The
organic layer was separated, washed with water, dried over
Na2SO4, filtered, and concentrated to give 700 mg of Boc-D-
Phe-Pro-Lys(Boc)Ψ[COCO]-OH (quantitative yield). TLC
(EtOAc:pyridine:HOAc:H2O, 94:20:11:6) Rf ) 0.15. 1H NMR
(200 MHz, CDCl3): δ 1.41 (s, 9H), 1.48 (s, 9H), 2.51 (complex,
1H), 3.0 (complex, 2H), 3.11 (complex, 2H), 3.60 (b s, 1H), 4.30-
4.50 (complex, 2H), 4.80 (complex, 1H), 7.25 (m, 5H). MS m/z
(FAB+): 616.9 [M + H].
HOOC-CH2-D-Ch a -3,3-Dm p -Lysψ[COCO]-OH (24). 1H
NMR (400 MHz, D2O): δ 0.86 (s, 3H), 0.96 (s, 3H), 0.80-1.89
(complex, 21H), 2.83 (m, 2H), 3.47-3.76 (m, 2H), 3.47-3.65
(dd, 2H), 3.89 (s, 1H), 4.15 (dd, 1H), 4.28 (m, 1H). Rt (LC):
29.32 min 20% A/80% B to 20% A/20% B/60% C in 40 min;
purity 97.7% (LC). MS m/z (FAB+): 511.4 [M + H], 529.4 [M
+ H2O + H], 603.5 [M + glycerol]; (FAB-): 509.4 [M - H],
601.5 [M + glycerol].
HOOC-CH2-D-Ch a -Oh i-Lysψ[COCO]-OH (25). 1H NMR
(400 MHz, D2O): δ 0.95 (m, 2H), 1.02-2.02 (complex, 29H),
2.12 and 2.19 (2 × dt, 1H), 2.40 (m, 1H), 2.92 (m, 2H), 3.56-
3.70 (m, 2H), 3.76 (m, 1H), 4.17 (dd, 1H), 4.23 (t, 1H), 4.33
and 4.41 (2 × dd, 1H), 4.89 (dd, 1H) (doubling of signals
indicates the presence of both Ohi regioisomers). Rt (LC): 33.03
min 20% A/80% B to 20% A/20% B/60% C in 40 min; purity
92.3% (LC). MS m/z (ESI+): 537 [M + H], 555 [M + H2O +
H].
H OOC-CH 2-D-Ch a -(N-cyclop en t yl)-Gly-Lysψ[COCO]-
OH (26). 1H NMR (400 MHz, D2O): δ 0.86 (m, 2H), 1.20-
1.71 (complex, 6H), 1.78 (complex, 2H), 2.83 (m, 2H), 3.47-
3.73 (complex, 2H), 3.90 (dd, 1H), 4.08 (complex, 1H), 4.55
(complex, 1H), 4.85 (m, 1H). Rt (LC): 29.77 min 20% A/80% B
to 20% A/20% B/60% C in 40 min; purity 92.6% (LC). MS m/z
(FAB+): 511.2 [M + H], 529.2 [M + H2O + H], 603.3 [M +
glycerol + H]; (FAB-): 509.1 [M - H], 601.1 [M + glycerol -
H].
HOOC-CH2-D-Ch a -P r o-Lysψ[COCO]-OiP (27). 1H NMR
(400 MHz, D2O): δ 0.92 (m, 2H), 1.10 (m, 2H), 1.80 (d, 3H),
1.21 (d, 3H), 1.21-1.82 (complex, 12H), 1.96 (m, 1H), 2.20 (m,
1H), 2.89 (m, 2H), 3.40-3.61 (m, 2H), 3.69 (m, 1H), 4.21 (dd,
1H), 4.31-4.37 (complex, 2H), 4.90 (m, 1H). Rt (LC): 30.68
min 20% A/80% B to 20% A/20% B/60% C in 40 min; purity
96.6% (LC). MS m/z (ESI+): 525.4 [M + H], 543.4 [M + H2O
+ H], 507.6 [M - H2O + H]; (ESI-): 523.4 [M - H], 541.2 [M
+ H2O - H].
HOOC-CH2-D-Ch a -P r o-Lysψ[COCO]-NH-(CH2)2P h (28).
1H NMR (400 MHz, D2O): δ 1.08 (m, 2H), 1.20-2.00 (complex,
11H), 2.09 (m, 2H), 2.31 (m, 1H), 2.93 (m, 2H), 3.03 (m, 2H),
3.44-3.87 (complex, 6H), 4.15 (dd, 1H), 4.43-4.56 (m, 2H),
5.14 (dd, 1H), 7.33-7.49 (complex, 5H). Rt (LC): 38.61 min
H-D-P h e-P r o-Lysψ[COCO]-OH (7). Boc-D-Phe-Pro-Lys-
(Boc)Ψ[COCO]-OH (700 mg, 0.80 mmol) was treated with 7
mL of 90% TFA/CH2Cl2 for 4 h at room temperature. The
reaction mixture was concentrated in vacuo, and the residue
was dissolved in water and directly charged on a preparative
HPLC DeltaPak RP-C18 column using a gradient elution
system of 20% A/80% B to 20% A/50% B/30% C in 45 min at
a flow rate of 80 mL/min (A: 0.5 M phosphate buffer pH 2.1;
B: water; C: acetonitrile/water ) 6:4). The fractions contain-
ing 7 were isolated, desalted, and lyophilized (71 mg, 21.2%).
1H NMR (400 MHz, D2O): δ 1.10-1.60 (complex, 6H), 1.67
(m, 2H), 1.92 (complex, 1H), 2.51 (m, 1H), 2.82 (m, 2H), 3.00
(dd, 1H), 3.09 (dd, 1H), 3.36 (m, 1H), 4.04 (dd, 1H), 4.19 and
4.25 (2 x dd, 1H), 4.39 (dd, 1H), 7.15 (m, 2H), 7.26 (m, 3H). Rt
(LC): 16.22 min 20% A/80% B to 20% A/20% B/60% C in 40
min; purity 92.4% (LC). MS m/z (FAB+): 419.1 [M + H], 437.3
[M + H2O + H], 511.1 [M + glycerol]; (FAB-): 417.0 [M - H],
509.1 [M + glycerol].
In an analogues way to the synthesis of 7, compounds 16,
17, and 20-28 were prepared from the corresponding dipep-
tides.
H-D-Ch a -P r o-Lysψ[COCO]-OH (16). 1H NMR (400 MHz,
DMSO-d6): δ 0.80-2.13 (m, 24H), 2.77 (t, 2H), 3.70-3.98 (m,
1H), 4.14 (m, 1H), 4.41 (m, 1H), 4.85 (m, 1H). Rt (LC): 22.14
min 20% A/80% B to 20% A/20% B/60% C in 40 min; purity
97.4% (LC). MS m/z (FAB+): 425 [M + H], 517 [M + glycerol];
(FAB-): 423 [M - H].
Bn SO2-Asp -P r o-Lysψ[COCO]-OH (17). 1H NMR (400
MHz, D2O): δ 1.12-1.86 (complex, 9H), 2.02-2.15 (m, 1H),
2.41 (m, 1H), 2.66 (m, 1H), 2.84 (t, 2H), 3.14 (m, 1H), 3.36 (m,
1H), 3.99 (dd, 1H), 4.16 (m, 2H), 4.27-4.42 (dd, 2H), 7.27-
7.38 (m, 5H). Rt (LC): 21.91 min 20% A/80% B to 20% A/20%

Potent Oral Antithrombotics
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20 4431
20% A/80% B to 20% A/20% B/60% C in 40 min; purity 97.8%
(LC). MS m/z (ESI+): 586.0 [M + H], 605.0 [M + H2O + H],
568.9 [M - H2O + H].
7.4 (composition [mmol/L]: Tris, 50; NaCl, 100), and further
with equally prediluted pooled rat (dog) plasma to obtain
samples with an antithrombin content within the calibration
curve () range of 10-90% of factor IIa inactivation). After 100
µL of diluted sample with 50 µL of 0.5 U/mL human thrombin
(Kordia Laboratory Supplies, Leiden, The Netherlands) was
incubated during 2 min, 100 µL S-2238 (1.56 mg/mL) was
added and anti-IIa activities were read from a calibration
curve, by measuring the difference in optical density at 405
nm after 2 and 22 min (∆OD405), thus quantifying the residual
activity of factor IIa. The calibration curve was obtained from
activity measurements for the test compound itself, spiked in
prediluted pooled rat (dog) plasma. Anti-IIa activities were
expressed as nmol/mL.
The pharmacokinetic parameters were analyzed from the
plasma concentration vs time curve using the computerized
iterative procedure of MW/Pharm (Medi/ware, Groningen, The
Netherlands), based on the Simplex method. Subsequently, the
elimination half-lives were calculated using the model of
relative error independent of the concentration, the area under
the curve (AUC) was determined with the trapezium rule.
Assuming linear kinetics, the percent bioavailability was
calculated by dividing the AUC obtained after po administra-
tion by the mean expected normalized AUC after iv adminis-
tration of that dose (× 100%).
Deter m in a tion of In h ibition Con sta n ts. The thrombin
assay conditions are described in J etten et al. Human throm-
bin was used, and inhibition constants were calculated from
plots using nine different inhibitor concentrations and always
at least in duplicate.19 Assays for factor Xa, factor VIIa/TF,
trypsin, protein C, tPA, plasmin, etc. were performed according
to published methods. Values for all serine proteases are the
average of the least two determinations, where the variation
in the assay is (10%.
Deter m in a tion of CACO-2 Tr a n sp or t. The CACO-2 cell
permeability studies were performed as described in Walter
et al.44 with the following modifications: Cell passages 25-
40 were used; seeding density 6.3 × 104 cells/cm2 on collagen-
coated filters 3.0 µm pore size; three 1 h incubation intervals;
and sample analysis by antithrombin activity or specific HPLC
method.
Deter m in a tion of Hep a tic Clea r a n ce in P er fu sion
Exp er im en ts. The liver of a rat was excised, and circular
perfusion was performed during 120 min with 160 mL of a
2.5% bovine serum albumin solution in Krebs-Henseleit
buffer, pH 7.4, at 37 °C, containing test compound in an initial
concentration of 5 µM. At indicated time points, samples of
buffer were taken and concentration of the compound was
measured using the chromogenic substrate assay described for
the determination of the antithrombin activity. The concentra-
tion of the test compound during this 120 min period was
compared to the concentration in the buffer at the start of the
experiment.
Deter m in a tion of IV/P O Kin etics in Ra ts a n d Dogs.
An im a ls a n d An a esth etics. Male Wistar Hsd/Cpb WU rats
(body weight, 200-500 g) were obtained from Harlan (Horst,
The Netherlands) and anaesthetized intraperitoneally with
sodium pentobarbital (Nembutal, 60 mg/kg; Sanofi, Toulouse,
France) or methohexital sodium (40 g/L stock solution; 200
µL/100 g body wt Brietal, Eli Lilly). Body temperature was
maintained at 37.5 ( 1 °C. Female beagle dogs (6-11 kg) were
obtained from Marshall Europe, Lyon, France. All procedures
were approved by the Ethics Committee of Animal Welfare in
accordance with Dutch guidelines.
Deter m in a tion of IV/P O Kin etics in Ra ts a n d Dogs.
One day prior to administration of the serine protease inhibi-
tors, rats (body weight 275-325 g) were anaesthetized by
injection of methohexital sodium. The right jugular vein was
cannulated with a PE-50 cannula (Clay Adams) filled with
saline and led subcutaneously to the neck to be exteriorized.
After a recovery period of at least 16 h, the tested compound
was administered intravenously via the cannula after which
the cannula was rinsed with saline. Depending on the way of
administration (intravenously or a orally by gavage) and on
the expected half-life of the tested compound, at least six time
points were selected for blood sampling. Both ways of admin-
istration were carried out in separate animals. Blood was
collected from the applied cannula in 0.05 volume 0.2 M Na2-
EDTA‚2H2O and centrifuged at 125 000 N/kg for 2 min after
which the plasma was stored below -20 °C until used for
determination of the plasma concentration. In case the kinetic
study was performed in dogs, both ways of administration were
carried out in the same dog with a wash-out period of at least
2 weeks. Blood was collected from the jugular vein by means
of a siliconized multisample needle and artificial evacuated
blood collection tubes containing 5.8 mg of K2EDTA for 3 mL
blood samples (Terumo, Leuven, Belgium).
Deter m in a tion of An tith r om botic P oten cy in a n Aor ta
F low Mod el in Ra ts. Thrombus formation on a silk thread
positioned in the aorta of a rat was induced as previously
described.46 Rats (350-500 g body wt) were weighed. A
catheter (19 cm PT-46, Portex, Portland Plastics, Hythe, U.K.)
containing a silk thread (size 4/0, uncoated USP, Pfrimmer,
Erlangen, Germany) with two knots, tied at the end of the
thread and protruding 3 mm, was inserted into the left
common carotid artery. Subsequently, the catheter with thread
was led via the carotid artery and the aorta into the left iliac
artery. The silk thread was then fixed in position with a small
bulldog clamp over the knots, which were just visible through
the intact vessel wall. The protruding seal was then cut off
the catheter, which was retracted, leaving the silk thread in
place. Finally, the carotid artery was clamped, the iliac clamp
was removed, and the abdominal cavity was covered provision-
ally. After 10 min of blood circulation, the silk thread covered
with thrombus was carefully removed and gently rinsed.
Thrombus weight and length were determined. Details of this
well-established procedure have been published elswhere.45,46
X-r a y Cr ysta llogr a p h y. Crystallization was essentially
performed using the hirugen [53-65]-thrombin exchange pro-
cedure at room temperature with the sitting drop vapor
diffusion technique. X-ray diffraction data were collected at
100 K as described previously.47 Molecular replacement was
carried out with the aid of the EPMR program and 1.59 Å
resolution could be obtained after refinement.48 The coordi-
nates have been deposited at the Protein Data Bank.
Ack n ow led gm en t. We thank Ms. Y. Diepeveen for
the difficult job of the chemical characterization of the
compounds. In addition, we thank Dr. A. Riboldi-
Tunnicliffe of the University of Glasgow, U.K., for the
crystallization and subsequent structure determination
of human R-thrombin with compound 20.
Refer en ces
(1) Burnier, M.; Brunner, H. R. Angiotensin II AT1 receptor
antagonists: clinical development and future perspectives.
Therapie 1998, 53, 279-284.
The residence time and percent bioavailability of the direct
thrombin inhibitors were measured by determination of the
anti-IIa activity in plasma. In view of its selectivity of the
examined protease inhibitors, inhibition of thrombin linearly
relates with the concentration of the measured protease
inhibitor. Plasma anti-IIa activity was measured amidolyti-
cally with S-2238 (Chromogenix, Cromogenics Ltd., Molndal,
Sweden) based on the method of Teien and Lie.45 All samples
were initially diluted with 3 volumes of tris-NaCl buffer, pH
(2) Freidinger, R. M. Nonpeptide ligands for peptide and protein
receptors. Curr. Opin. Chem. Biol. 1999, 3, 395-406.
(3) De Tullio, P.; Delarge, J .; Pirotte, B. Recent advances in the
chemistry of cholecystokinin receptor ligands (agonists and
antagonists). Curr. Med. Chem. 1999, 6, 433-455.
(4) Burrell, L. M.; Risvanis, J .; J ohnston, C. I.; Naitoh, M.; Balding,
L. C. Vasopressin receptor antagonism-a therapeutic option in
heart failure and hypertension. Exp. Physiol. 2000, 85, 259S-
265S.

4432 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 20
Adang et al.
(5) Owens, M. J .; Nemeroff, C. B. Corticotropin-releasing factor
antagonists in affective disorders. Exp. Opin. Invest. Drugs 1999,
8, 1849-1858.
(6) Veale, C. A.; Damewood, J . R., J r.; Steelman, G. B.; Bryant, C.;
Gomes, B.; Williams, J . Non-peptide inhibitors of human leu-
kocyte elastase. 4. Design, synthesis, and in vitro and in vivo
activity of a series of beta-carbolinone-containing trifluoromethyl
ketones. J . Med. Chem. 1995, 38, 86-97.
(7) Luther, R. R.; Stein, H. H.; Glassman, H. N.; Kleinert, H. D.
Renin inhibitors: specific modulators of the renin-angiotensin
system. Arzneim.-Forsch. 1989, 39, 1-5.
(8) Weitz, J . I.; Hirsh, J . New antithrombotic agents. Chest 1998,
114, 715S-727S.
(9) Turpie, A. G. Clinical potential of antithrombotic drugs in
coronary syndromes. Am. J . Cardiol. 1998, 82, 11L-14L.
(10) Wiley, M.; Fisher, M. J . Small-molecule direct thrombin inhibi-
tors. Exp. Opin. Ther. Pat. 1997, 7, 1265-1282.
(11) Rewinkel, J . B. M.; Adang, A. E. P. Strategies and progress
towards the ideal orally active thrombin inhibitor. Curr. Pharm.
Des. 1999, 5, 1043-1075.
(12) Adang, A. E. P.; Rewinkel, J . B. M. A new generation of orally
active antithrombotics: comparing strategies in the GPIIb/IIIa,
thrombin and factor Xa areas. Drugs Future 2000, 25, 369-383.
(13) Adang, A. E. P.; Lucas, H.; de Man, A. P. A.; Engh, R. A.;
Grootenhuis, P. D. J . Novel acylguanidine containing thrombin
inhibitors with reduced basicity at the P1 moiety. Bioorg. Med.
Chem. Lett. 1998, 8, 3603-3608.
(27) Matthews, J . H.; Krishnan, R.; Costanzo, M. J .; Maryanoff, B.
E.; Tulinsky, A. Crystal structures of thrombin with thiazole-
containing inhibitors: probes of the S1′ binding site. Biophys.
J . 1996, 71, 2830-2839.
(28) 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.
(29) Adang, A. E. P.; Van Boeckel, C. A. A.; Grootenhuis, P. D. J .;
Peters, J . A. M. WO 9717363, 1997.
(30) Van Boeckel, C. A. A.; Adang, A. E. P.; Peters, J . A. M. WO
9807308, 1998.
(31) Lewis, S. D.; Ng, E. A.; Mellott, M. J .; Appleby, S. D.; Brady, S.
F.; Stauffer, K. J .; Sisko, J . T.; Mao, S. S.; Veber, D. F. Inhibition
of thrombin by peptides containing lysyl-alpha-keto carbonyl
derivatives. Thromb. Haemostasis 1995, 74, 1107-1112.
(32) Tamura, S. Y.; Shamblin, B. M.; Brunck, T. K.; Ripka, W. C.
Rational design, synthesis, and serine protease inhibitory activ-
ity of novel P1-argininoyl heterocycles. Bioorg. Med. Chem. Lett.
1997, 7, 1359-1364.
(33) Akiyama, Y.; Tsutsumi, S.; Hatsushiba, E.; Ohuchi, S.; Okonogi,
T. Peptidyl R-keto thiazole as potent thrombin inhibitors. Bioorg.
Med. Chem. Lett. 1997, 7, 533-538.
(34) Semple, J . E.; Rowley, D. C.; Brunck, T. K.; Ripka, W. C.
Synthesis and biological activity of P2-P4 azapeptidomimetic P1-
argininal and P1-ketoargininamide derivatives: A novel class
of serine protease inhibitors. Bioorg. Med. Chem. Lett. 1997, 7,
315-320.
(14) Rewinkel, J . B. M.; Lucas, H.; van Galen, P. J . M. 1-Aminoiso-
quinoline as benzamidine isoster in the design and synthesis of
orally active thrombin inhibitors. Bioorg. Med. Chem. Lett. 1999,
9, 685-690.
(15) Adang, A. E. P.; Peters, C. A. M.; Gerritsma, S.; de Zwart, E.;
Veeneman, G. Solution-phase and solid-phase synthesis of novel
transition state inhibitors of coagulation enzymes incorporating
a piperidinyl moiety. Bioorg. Med. Chem. Lett. 1999, 9, 1227-
1232.
(16) Brady, S. F.; Sisko, J . T.; Stauffer, K. J .; Colton, C. D.; Qiu, H.;
Lewis, S. D.; Ng, A. S.; Shafer, J . A.; Bogusky, M. J .; Veber, D.
F.; Nutt, R. F. Amide and a-keto carbonyl inhibitors of thrombin
based on arginine and lysine: synthesis, stability and biological
characterization. Bioorg. Med. Chem. 1995, 3, 1063-1078.
(17) J ones, D. M.; Atrash, B.; Ryder, H.; Teger-Nilsson, A.-C.;
Gyzander, E.; Szelke, M. Thrombin inhibitors based on ketone
derivatives of arginine and lysine. J . Enzyme Inhib. 1995, 9, 43-
60.
(18) Bajusz, S.; Szell, E.; Bagdy, D.; Barabas, E.; Horvath, G.;
Dioszegi, M.; Fittler, Z.; Szabo, G.; J uhasz, A. Highly active and
selective anticoagulants: D-Phe-Pro-Arg-H, a free tripeptide
aldehyde prone to spontaneous inactivation, and its stable
N-methyl derivative, D.-MePhe-Pro-Arg-H. J . Med. Chem. 1990,
33, 1729-1735.
(35) Boatman, P. D.; Ogbu, C. O.; Eguchi, M.; Kim, H.-O.; Nakanishi,
B. C.; Shea, J . P.; Kahn, M. Secondary structure peptide
mimetics: Design, synthesis, and evaluation of â-strand mimetic
thrombin inhibitors. J . Med. Chem. 1999, 42, 1367-1375.
(36) Charles, R. S.; Matthews, J . H.; Zhang, E.; Tulinsky, A. Bound
structures of novel P3-P1′ â-strand mimetic inhibitors of throm-
bin. J . Med. Chem. 1999, 42, 1376-1383.
(37) Bachand, B.; DiMaio, J.; Siddiqui, M. A. Synthesis and structure-
activity relationship of potent bicyclic lactam thrombin inhibi-
tors. Bioorg. Med. Chem. Lett. 1999, 9, 913-918.
(38) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J .
Experimental and computational approaches to estimate solubil-
ity and permeability in drug discovery and development settings.
Adv. Drug Delivery Rev. 1997, 23, 3-25.
(39) Lee, V. H. L. Membrane transporters. J . Pharm. Sci. 2000, 11,
S41-S50.
(40) Ranadive, S. A.; Chen, A. X.; Serajuddin, A. T. M. Relative
lipophilicities and structural-pharmacological considerations of
various angiotensin-converting enzmyme (ACE) inhibitors. Pharm.
Res. 1992, 9, 1480-1486.
(41) Thwaites, D. T.; Cavet, M.; Hirst, B. H.; Simmons, N. L.
Angiotensin-converting enzyme (ACE) inhibitor transport in
human intestinal epithelial (Caco-2) cells. Br. J . Pharmacol.
1995, 114, 981-986.
(19) Gallemo, R. A., J r.; Fevig, J . M.; Carini, D. J .; Cacciola, J .; Wells,
B. L.; Hillyer, G. L.; Buriak, J ., J r.; Rossi, K. A.; Stouten, P. F.
W.; Alexander, R. S.; Hilmer, R.; Bostrom, L.; Abelman, M. M.;
Lee, S.-L.; Weber, P. C.; Kettner, C. A.; Knabb, R. M.; Wexler,
R. R. (Nacyl-N-alkyl)glycyl borolysine analogues: new class of
potent thrombin inhibitors. Bioorg. Med. Chem. Lett. 1996, 6,
2913-2918.
(20) Lewis, S. D.; Ng, A. S.; Baldwin, J . J .; Fusetani, N.; Naylor, A.
M.; Shafer, J . A. Inhibition of thrombin and other trypsin-like
serine proteinases by cyclotheonamide A. Thromb. Res. 1991,
70, 173-190.
(21) J etten, M.; Peters, C. A. M.; Visser, A.; Grootenhuis, P. D. J .;
van Nispen, J . W. F. M.; Ottenheijm, H. C. J . Peptide-derived
transition state analogue inhibitors of thrombin; synthesis,
activity and selectivity. Bioorg. Med. Chem. 1995, 3, 1099-1114.
(22) Dondoni, A.; Perrone, D.; Merino, P. Chelation- and nonchela-
tion-controlled addition of 2-(trimethylsilyl)thiazole to R-amino
aldehydes: stereoselective synthesis of the â-amino-R-hydroxy
aldehyde intermediate for the preparation of the human immu-
nodeficiency virus proteinase inhibitor Ro 31-8959. J . Org.
Chem. 1995, 60, 8074-8080.
(42) J etten, M.; Peters, C. A. M.; van Nispen, J . W. F. M.; Ottenheijm,
H. C. J . A one-pot N-protection of L-Arginine. Tetrahedron Lett.
1991, 32, 6025-6028.
(43) Edwards, P. D.; Wolanin, D. J .; Andisik, D. W.; Davis, M. W.
Peptidyl a-ketoheterocyclic inhibitors of human neutrophil
eleastase. 2. Effect of varying the heterocyclic ring on in vitro
potency. J . Med. Chem. 1995, 38, 76-85.
(44) Walter, E.; Kissel, T.; Raddatz, P. Transport of peptidomimetic
renin inhibitors across monolayers of a human intestinal cell
line (Caco-2): evidence for self-enhancement of paracellular
transport route. Pharm. Res. 1995, 12, 1801-1805.
(45) Teien, A. N.; Lie, M. Evaluation of an amidolytic heparin assay
method: increased sensitivity by adding purified antithrombin
III. Thromb. Res. 1977, 10, 399-410.
(46) Vogel, G. M. T.; van Amsterdam, R. G. M.; Zandberg, P.; van
Houwelingen, P.; Kop, W. J .; van Mensvoort, F. W. J .; Meule-
man, D. G. Two new closely related rat models with relevance
to arterial thrombosis- efficacies of different antithrombotic
drugs. Thromb. Haemostasis 1997, 77, 183-189.
(47) Riboldi-Tunnicliffe, A.; Hilgenfeld, R. Cryocrystallography with
oil - an old idea revived. J . Appl. Crystallogr. 1999, 32, 1003-
1005.
(48) Kissinger, C. R.; Gehlhaar, D. K.; Fogel, D. B. Rapid automated
molecular replacement by evolutionary search. Acta Crystallogr.
D 1999, 55, 484-491.
(49) Bush, L. R. Argatroban, a selective, potent thrombin inhibitor.
Cardiovasc. Drug Rev. 1991, 9, 247-263.
(23) Nahm, S.; Weinreb, S. M. N-methoxy-N-methylamides as effec-
tive acylating agents. Tetrahedron Lett. 1981, 22, 3815-3818.
(24) Nielsen, D. G. Imidates including cyclic imidates. In The
Chemistry of Amidines and Imidates; Patai, S., Ed.; J ohn Wiley
& Sons: New York, 1975; Vol. I, pp 385-489.
(25) Dess, D. B.; Martin, J . C. Readily accessible 125I oxidant for the
conversion of primary and secondary alcohols to aldehydes and
ketones. J . Org. Chem. 1983, 48, 4155-4156.
(50) Gustafsson, D.; Antonsson, T.; Bylund, R.; Eriksson, U.; Gyz-
ander, E.; Nilsson, I.; Elg, M.; Mattsson, C.; Deinum, J .;
Pehrsson, S.; Karlsson, O.; Nilsson, A.; So¨rensen, H. Effects of
melagatran, a new low-molecular-weight thrombin inhibitor, on
thrombin and fibrinolytic enzymes. Thromb. Haemostasis 1998,
79, 110-118.
(26) Costanzo, M. J .; Maryanoff, B. E.; Hecker, L. R.; Schott, M. R.;
Yabut, S. C.; Zhang, H.-C.; Andrade-Gordon, P.; Kauffman, J .
A.; Lewis, J . M.; Krishnan, R.; Tulinsky, A. Potent thrombin
inhibitors that probe the S1′ subsite: Tripeptide transition state
analogues based on a heterocycle-activated carbonyl group. J .
Med. Chem. 1996, 39, 3039-3043.
J M011110Z