Low molecular weight thrombin inhibitors with excellent potency, metabolic stability, and oral bioavailability

Article abstract of DOI:10.1016/j.bmcl.2004.06.030

Modification of lead compound 1 by reducing lipophilicity in the P3 group produced a series of low molecular weight thrombin inhibitors with excellent potency in functional assays, metabolic stability, and oral bioavailability. These modifications led to

Full text of DOI:10.1016/j.bmcl.2004.06.030

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4161–4164
Low molecular weight thrombin inhibitors with excellent potency,
metabolic stability, and oral bioavailability
Matthew M. Morrissette,a,* Kenneth J. Stauffer,a Peter D. Williams,a Terry A. Lyle,a
Joseph P. Vacca,a Julie A. Krueger,b S. Dale Lewis,b Bobby J. Lucas,b Bradley K. Wong,c
Rebecca B. White,c Cynthia Miller-Stein,c Elizabeth A. Lyle,d Audrey A. Wallace,d
Yvonne M. Leonard,d Denise C. Welsh,d Joseph J. Lynchd and Daniel R. McMasterse
aDepartment of Medicinal Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
bDepartment of Biological Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
cDepartment of Preclinical Drug Metabolism, Merck Research Laboratories, West Point, PA 19486, USA
dDepartment of Pharmacology, Merck Research Laboratories, West Point, PA 19486, USA
eDepartment of Molecular Systems, Merck Research Laboratories, West Point, PA 19486, USA
Received 16 April 2004; accepted 9 June 2004
Available online
Abstract—Modification of lead compound 1 by reducing lipophilicity in the P3 group produced a series of low molecular weight
thrombin inhibitors with excellent potency in functional assays, metabolic stability, and oral bioavailability. These modifications led
to the identification of two optimized compounds, 14 and 16.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
only), and warfarin (slow onset of action, chronic
patient monitoring required).
The serine protease, thrombin, occupies a central role in
the coagulation cascade. The primary actions of
thrombin are the activation of platelets and cleavage of
fibrinogen to fibrin, which together constitute the pri-
mary components of a vascular thrombus. Inhibitors of
thrombin have long been recognized as potential ther-
apeutic agents for the treatment of a variety of throm-
botic disorders, for example, deep vein thrombosis,
pulmonary embolism, and thromboembolic stroke. In-
deed, intravenous and oral thrombin inhibitors have
shown promising results in human clinical trials1;2
resulting in the recent approval of oral ximelagatran.3
Our goal has been to identify a thrombin inhibitor with
a convenient dosing regimen, that is, once or twice daily
oral dosing, predictable pharmacokinetics, and no food
or drug–drug interactions. Such a compound would
potentially overcome the limitations associated with the
two most widely employed antithrombotic agents, low
molecular weight heparin (parenteral administration
An interesting thrombin inhibitor lead from our labo-
ratories, 1, resulted by combining an optimized P3–P2
portion of a compound from a proline P2-based series of
thrombin inhibitors4 with a recently discovered potency-
enhancing P1 group.5 Compound 1 was found to have
good potency as a thrombin inhibitor in an isolated
enzyme assay (Ki ¼ 0:5 nM), good potency in a coagu-
lation assay in human plasma (2 · APTT ¼ 0.37 lM),
and an extremely favorable pharmacokinetic profile in
dogs (F ¼ 39%, iv t1=2 ¼ 13 h). A shortcoming with 1,
however, was its poor efficacy in a FeCl3-induced arte-
rial thrombosis model in rats. The lack of efficacy in this
animal model was attributed to the high plasma protein
binding of 1.
P2
Cl
H
P3
P1
NH2
N
N
O
O
O
H
* Corresponding author. Tel.: +1-215-652-9102; fax: +1-215-652-3971;
e-mail: matthew_morrissette@merck.com
1
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.06.030

4162
M. M. Morrissette et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4161–4164
2. Chemical synthesis
Cl
Br
Cl
a, b, c, d, e
f, g
Methyl 2-bromo-5-chloro-benzoate 2 was converted to
2-(N-Boc-aminomethyl)-5-chlorobenzylamine 35;6 in five
steps, and standard peptide coupling and deprotection
O
H2N
O
NH-Boc
Cl
2
3
reactions with Fmoc-L-proline or Fmoc-L-azetidine
carboxylic acid provided multi-gram quantities of P2–P1
intermediates 4 and 5, respectively (see Scheme 1).
Standard peptide coupling and deprotection procedures
using 4 and 5 and commercially available hydroxy acids
provided compounds 1 and 8–15 in Table 1.6 The hy-
droxy acid used for the synthesis of compound 16 in
Table 1 was obtained from methyl 3,3-dimethyl-4-
pentenoate 6. The olefin in 6 was cyclopropanated, and
this product was hydroxylated using the camphor-based
oxaziridine reported by Davis et al.7 The resulting 4:1
mixture of R and S hydroxy ester enantiomers was
hydrolyzed and the hydroxy acids were separated by
chiral HPLC to give 7. Standard peptide coupling of
hydroxy acid 7 and amine 5 followed by deprotection
provided 16. All final compounds were purified by re-
verse phase HPLC and isolated as amorphous TFA salts
with purities greater than 95%.
Cl
R2
( )n
( )n
h, i
H
H
N
R1
HO
N
HN
N
O
O
O
NH2
NH-Boc
4 (n=1)
5 (n=0)
8-16
j, k, l, m
CO2H
OH
CO2CH3
6
7
Scheme 1. Reagents and conditions: (a) Zn(CN)2, Pd(PPh3)4, DMF,
60 ꢁC; (b) LiAlH4, THF; (c) Boc2O, THF; (d) DPPA, toluene; (e) PPh3,
H2O, THF; (f) Fmoc-amino acid, EDCI, HOBT, i-Pr2NEt, DMF; (g)
piperidine, DMF; (h) hydroxy acid, EDCI, HOBT, i-Pr2NEt, DMF; (i)
HCl gas, EtOAc, 0 ꢁC; (j) CH2N2, cat. Pd(OAc)2, ether; (k) LDA,
THF, )78 ꢁC, Davis oxaziridine; (l) NaOH, H2O, MeOH; (m) chiral
HPLC separation.
3. Biological assays
In this report we describe an approach to decreasing
plasma protein binding, which involves reducing the size
of the lipophilic fluorenyl P3 group in 1. This approach
has resulted in the production of potent, low molecular
weight thrombin inhibitors with excellent efficacy,
selectivity, metabolic stability, and pharmacokinetic
properties.
Inhibition constants (Ki values) were determined for test
compounds in an assay using human derived thrombin,
trypsin, urokinase, plasmin, kallikrein, t-PA, and chy-
motrypsin with the corresponding fluorescent or chro-
mogenic substrate under steady state conditions as
previously described.8;9 The concentration of test
compound required to double the activated partial
Table 1. Biological results for compounds 1, 8–16
Cl
R2
( )n
H
R1
N
N
O
O
HO
NH2
Compd
n
R1
R2
Kia (nM)
2 · APTTb
(lM)
Plasma protein
binding (% free)
In vivo ratc
(occlusions)
Dog 1 mg/kg pod
Human
Rat
AUC
(lM h)
t1=2 (h)
1
1
1
1
1
1
1
1
1
0
0
––
Me
Ph
H
––
Ph
0.50
4.7
0.37
0.47
0.33
0.14
1
9
4/5
1/6
30
21
14
1.4
13
8
3.1
3.4
2.0
9
Me
Ph
2.7
0.90
80
15
24
10
11
12
13
14
15
16
Ph
H
H
c-Hex
i-Pr
t-Bu
0.27
5.6
0.15
0.37
0.23
0.37
0.28
4
63
24
21
11
36
60
54
26
17
0/6
2/6
0/6
0/5
0/5
1.7
4.9
10
2.0
4.2
5.4
3.4
4.7
H
H
2.1
5.0
H
t-Bu
–C(Me)2c–Pr
3.5
11
H
3.7
a Inhibition constants for human thrombin.
b Concentration of test compound required for doubling the activated partial thromboplastin time in human plasma.
c Test compounds administered at 10 lg/kg/min prior to FeCl3-induced clot formation on the carotid artery, results are expressed as number of vessel
occlusions in five or six animals.
d Test compounds dosed as TFA salts in 1% methocel suspension, results are expressed as area under the concentration versus time curve (AUC) and
plasma half life (t1=2).

M. M. Morrissette et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4161–4164
4163
thromboplastin time of human plasma, the 2 · APTT
value, was determined under conditions involving a
threefold dilution of human plasma as previously
reported.9 The antithrombotic activities of intravenously
administered test compounds were determined in an
anesthetized rat model of topical FeCl3-induced arterial
thrombosis.10 Test compounds were administered by
intravenous infusion at a dose of 10 lg/kg/min for
180 min, FeCl3 was applied to the carotid artery at
120 min post dose, and carotid artery blood flow was
monitored for 60 min. Results are expressed as number of
animals, which had occluded vessels (n ¼ 5 or 6). For
plasma pharmacokinetic assays, amorphous TFA salts of
test compounds were dosed orally in 1% methocel solu-
tion to beagle dogs (n ¼ 2) at 1 mg/kg. Blood samples
were obtained over 8 h post dose, and a final blood sample
was taken at 24 h. At each time point, test compound was
isolated from the plasma fraction by solid phase extrac-
tion and quantified by LC–MS/MS against a standard
curve. Area under the curve (AUC) and plasma half life
(t1=2) were determined from analysis of the concentration
versus time curve. Plasma protein binding was determined
as previously described.11
Figure 1. Compounds 9 (green) and 11 (orange) docked in the
thrombin active site, shown as a transparent surface.
compared to lead compound 1 in both human and rat
plasma. The increased free fraction of 9 in human
plasma explains the observation that, even though 9 is
five times less potent than 1 in the enzyme inhibition
assay, the two compounds are very close in potency in
the 2 · APTT assay, which is run in the presence of
human plasma. Also, the increased free fraction of 9
compared to 1 in rat plasma is presumably responsible
for the improved performance of 9 in the in vivo
thrombosis model in rats. Both 8 and 9 were well ab-
sorbed after a 1 mg/kg oral dose in dogs, but the plasma
half life of both compounds was significantly diminished
relative to 1. Absorption and plasma half life in dogs
continued to diminish with the phenylhydroxyacetamide
P3 analog 10, so attention was turned to investigating
P3 analogs containing nonaromatic substituents. The c-
hexyl derivative 12 proved to be very potent in both the
enzyme and 2 · APTT assays, with full efficacy in the in
vivo thrombosis model in rats. The pharmacokinetic
behavior of 12 in dogs, however, was not improved over
that of 10. Additional studies with 12 had shown that it
was rapidly degraded in the presence of dog hepatic
microsomal preparations, with oxidation occurring
predominantly on the c-hexyl ring.
4. Results and discussion
The first two analogs of 1 that were synthesized in order
to reduce lipophilicity in P3 were the phenylhydroxy-
propionamide diastereomers 8 and 9 (see Table 1).
Interestingly, diastereomers 8 and 9 were of similar po-
tency in the enzyme assay, with the S isomer 9 being
slightly more potent. This is in contrast to previously
established structure–activity relationships for P3 dia-
stereomers in which one of R1 and R2 is hydrogen. For
these compounds, there has generally been observed a
large difference in potency between R and S isomers with
the S isomer being less potent.12 Indeed, diastereomers
10 and 11 with methyl replaced by hydrogen were found
to exhibit the usual profile in the enzyme assay with the S
isomer 11 being significantly less potent. Molecular
modeling provided a rationale for these results. The
lowest energy conformation13 of 11 places the P3 phenyl
group away from the S3 pocket of the enzyme. Rotating
about the C@O, Ca carbon–carbon bond in the P3 group
of 11 to bring the phenyl group into the proper orien-
tation for binding to the S3 pocket raises the energy by
3.6 kcal/mol. The lowest energy conformation of 9, on
the other hand, places the P3 phenyl group in an ideal
orientation for binding to the S3 pocket of the enzyme.
Interestingly, however, this low energy conformation of
9 also places the hydroxyl group out toward solvent and
unable to hydrogen bond to Gly-216 such as is the usual
paradigm for P3 group interactions in peptide-based
thrombin inhibitors. Corroboration of this proposed
binding mode of 9 awaits verification by X-ray crystal-
lographic analysis of the enzyme–inhibitor complex.
Figure 1 shows an overlay of 9 and 11 in their low energy
conformations docked into the enzyme active site.
Removing potential sites of metabolism in 12 by trun-
cating the c-hexyl group to an i-propyl group (13) indeed
had a beneficial effect on dog hepatic microsome sta-
bility and also improved the oral pharmacokinetic
profile in dogs. Potency and in vivo efficacy in rats,
however, were reduced with the smaller P3 side chain in
13. Increasing the size of P3 side chain to a t-butyl group
provided a compound (14), which was found to possess
a very favorable combination of properties: good
potency in the 2 · APTT assay, full efficacy in the in vivo
thrombosis model in rats, little inhibition of several
other serine proteases (Ki ¼ 5:9 lM against trypsin;
Ki >10 lM against urokinase, plasmin, kallikrein, t-PA,
chymotrypsin), oral bioavailability in three species with
a good plasma half life (dog: F ¼ 81%, iv t1=2 ¼ 3:9 h;
rhesus monkey: F ¼ 46%, iv t1=2 ¼3.5 h; rat: F ¼ 37%,
As was anticipated, plasma protein binding experiments
with 9 showed a large increase in the free fraction

4164
M. M. Morrissette et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4161–4164
iv t1=2 ¼ 2:0 h), and no inhibition of any of the major
human cytochrome P450 isozymes at concentrations up
to 100 lM. The potency of 14 in the isolated enzyme and
2 · APTT assays compares very favorably with that of
melagatran, the active form of the double prodrug
ximelagatran,3 measured in our assays (Ki ¼ 1:1 nM,
2 · APTT ¼ 0.28 lM).
References and notes
1. Kaplan, K. L.; Francis, C. W. Seminars in Hematology
2002, 39, 187–196.
2. Crowther, M. A. Fundam. Clin. Cardiol. 2003, 46, 265–
271.
3. Salam, A. M.; Al-Mousa, E. N. Expert Opin. Pharmac-
other. 2004, 1423–1430.
4. Brady, S. F.; Stauffer, K. J.; Lumma, W. C.; Smith, G. M.;
Ramjit, H. G.; Lewis, S. D.; Lucas, B. J.; Gardell, S. J.;
Lyle, E. A.; Appleby, S. D.; Cook, J. J.; Holahan, M. A.;
Stranieri, M. T.; Lynch, J. J.; Lin, J. H.; Chen, I. W.;
Vastag, K.; Naylor-Olsen, A. M.; Vacca, J. P. J. Med.
Chem. 1998, 41, 401–406.
5. Rittle, K. E.; Barrow, J. C.; Cutrona, K. J.; Glass, K. L.;
Krueger, J. A.; Kuo, L. C.; Lewis, S. D.; Lucas, B. J.;
McMasters, D. R.; Morrissette, M. M.; Nantermet, P. G.;
Newton, C. L.; Sanders, W. M.; Yan, Y.; Vacca, J. P.;
Selnick, H. G. Bioorg. Med. Chem. Lett. 2003, 13, 3477–
3482.
6. A more efficient larger scale synthesis of 14 has been
reported: Nelson, T. D.; LeBlond, C. R.; Frantz, D. E.;
Matty, L.; Mitten, J. V.; Weaver, D. G.; Moore, J. C.;
Kim, J.; Boyd, R.; Kim, P.-Y.; Gbewonyo, K.; Brower,
M.; Sturr, M.; McLaughlin, K.; McMasters, D. R.; Kress,
M. H.; McNamara, J. M.; Dolling, U. H. J. Org. Chem.
2004, 69, 3620–3627.
In vitro metabolism studies with 14 using human hepatic
microsomes had shown a small amount of oxidative
metabolism involving the P2 proline ring and the t-butyl
side chain in P3. To address the former issue, the az-
etidine P2 analog 15 was investigated. Indeed, in vitro
metabolism studies with 15 indicated that no metabo-
lism had occurred on the azetidine ring. However, the
smaller azetidine ring did result in a loss of potency.
Potency could be recovered by increasing the size of the
P3 side chain. For example, 16, the analog of 15 in
which one of the methyl groups of the t-butyl P3 side
chain is replaced with a c-propyl group, had favorable
potency and oral pharmacokinetic properties in dogs
similar to those of 14. Additionally, very little oxidative
metabolism of 16 was observed in vitro using human
hepatic microsomes.
7. Davis, F. A.; Haque, M. S.; Ulatowski, G. T.; Towson, J.
C. J. Org. Chem. 1986, 51, 2402–2404.
8. Lewis, S. D.; Ng, A. S.; Baldwin, J. J.; Fusetani, N.;
Naylor, A. M.; Shafer, J. A. Thromb. Res. 1993, 70, 173.
9. Lewis, S. D.; Ng, A. S.; Lyle, E. A.; Mellott, M. J.;
Appleby, S. D.; Brady, S. F.; Stauffer, K. S.; Sisko, J. T.;
Mao, S. S.; Veber, D. F.; Nutt, R. F.; Lynch, J. J.; Cook,
J. J.; Gardell, S. J.; Shafer, J. A. Thromb. Haemostasis
1995, 74, 1107–1112.
5. Conclusion
In summary, reducing the size of the lipophilic P3 flu-
orenyl group in lead compound 1 gave a series of potent,
low molecular weight thrombin inhibitors, which had
reduced plasma protein binding and improved anti-
thrombotic activity in rats. Compound 14 was found to
possess an excellent combination of properties including
high 2 · APTT potency, high selectivity for inhibiting
thrombin versus a panel of other serine proteases, and
good pharmacokinetic properties in three animal spe-
cies. Identification of sites of metabolism in 14 led to the
design of 16, which was found to possess very good in
vitro metabolic stability as well as good potency and
pharmacokinetic properties. Compounds such as 14 and
16 offer significant potential as drug development can-
didates for the treatment of various thrombotic dis-
orders.
10. Schumacher et al. J. Pharmacol. Exp. Ther. 1993, 267,
1237–1242.
11. Tucker, T. J.; Lumma, W. C.; Lewis, S. D.; Gardell, S. J.;
Lucas, B. J.; Baskin, E. P.; Woltmann, R.; Lynch, J. J.;
Lyle, E. A.; Appleby, S. D.; Chen, I. W.; Dancheck,
K. B.; Vacca, J. P. J. Med. Chem. 1997, 40, 1565–
1569.
12. Sanderson, P. E. J. S.; Naylor-Olsen, A. M. Curr. Med.
Chem. 1998, 5, 289–304.
13. All energy calculations were performed with Batchmin
Mohamadi, F.; Richards, N. G. J.; Guida, W. C.;
Liskamp, R.; Caufield, C.; Chang, G.; Hendrickson, T.;
Still, W. C. J. J. Comput. Chem. 1990, 11, 440, using the
MMFFs force field and a distance-dependent dielectric of
4r. Low-energy P3 conformations were determined by
performing a two-torsion energy drive over the two bonds
on either side of the P3 sp3-hybridized carbon atom [N–
C(@O)–C–O and C(@O)–C–O–H]. The docked confor-
mations depicted in Figure 1 were generated by minimiz-
ing the low-energy conformers in the rigid thrombin active
site.
Acknowledgements
We would like to express our gratitude to Carl F.
Homnick for his skillfull chromatographic separations.