Chlorothiophenecarboxamides as P1 surrogates of inhibitors of blood coagulation factor Xa

Article abstract of DOI:10.1016/j.bmcl.2004.09.043

Neutral chlorothiophenecarboxamides bearing an amino acid and a substituted aniline were synthesized and investigated for their factor Xa inhibitory activity in vitro. From selected 2-methylphenyl morpholinones the solution properties were determined. The most soluble and active compounds were then investigated in different animal species to compare the pharmacokinetic parameters. This led to a potent, water soluble and orally bioavailable candidate for further development: EMD 495235.

Full text of DOI:10.1016/j.bmcl.2004.09.043

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5817–5822
Chlorothiophenecarboxamides as P1 surrogates of inhibitors
of blood coagulation factor Xa
a,
a
Werner W. K. R. Mederski,a,* Bertram Cezanne, Christoph van Amsterdam,
*
Karl-Ulrich Buhring,b Dieter Dorsch,a Johannes Gleitz,a
¨
Joachim Ma¨rza and Christos Tsaklakidisa
aMerck KGaA, Preclinical Pharmaceutical Research, 64271 Darmstadt, Germany
bInstitute for Pharmacokinetics and Metabolism, 85567 Grafing, Germany
Received 23 June 2004; revised 2 September 2004; accepted 17 September 2004
Available online 5 October 2004
Abstract—Neutral chlorothiophenecarboxamides bearingan amino acid and a substituted aniline were synthesized and investigated
for their factor Xa inhibitory activity in vitro. From selected 2-methylphenyl morpholinones the solution properties were deter-
mined. The most soluble and active compounds were then investigated in different animal species to compare the pharmacokinetic
parameters. This led to a potent, water soluble and orally bioavailable candidate for further development: EMD 495235.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
have discovered a series of potent nonamidines as exem-
plified by compound 13 (Fig. 1), which contains a 4-
chlorophenyl-carbamoyl P1 ligand, a central D-amino
acid linker and a phenyl morpholinone P4 residue.
The search for further novel benzamidine mimics led
to the discovery of achiral chlorothiophene benzimidaz-
oles such as compound 24 (Fig. 1), which incorporated
an alkylamide tether at the 5(6)-position of the benzimi-
dazole ringtogether with a substituted phenyl morpholi-
none as P4 ligand. Although both compounds displayed
fXa inhibitory activity in the mono-digit nanomolar
range, their physicochemical parameters such as solubil-
ity (<10lg/mL in pH7 phosphate buffer) and oral
pharmacokinetic profiles (bioavailabilty <10% in rats)
have to be improved.
The serine protease factor Xa (fXa) plays a crucial role
in the coagulation cascade and has, therefore, been identi-
fied as an attractive target for the development of anti-
thrombotic agents.1 Inhibition of this central enzyme
within this cascade has emerged as a particularly active
area of research.
At the outset of the research obtaininginhibitors of
fXa the reported compounds contained mono- or
bis-(benz)amidine functionalities.2 However, (benz)-
amidine-containingmolecules displayed undesired poor
pharmacokinetic properties. In an effort to develop
fXa inhibitors with more favourable oral profiles, we
Me
O
Cl
O
H
N
O
N
Cl
O
N
N
S
N
N
H
H
H
O
F
N
N
H
O
O
1 IC50 fXa = 5.6 nM
2 IC50 fXa = 5.9 nM
Figure 1. FXa bindingaffinities of chlorophenyl urea 1 and chlorothiophene benzimidazole 2.
Keywords: Anticoagulants; Factor Xa inhibitors; Chlorothiophenecarboxamides; Pharmacokinetics; EMD 495235.
*
Correspondingauthors. E-mail addresses: mederski@merck.de; bertram.cezanne@merck.de
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.09.043

5818
W. W. K. R. Mederski et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5817–5822
To overcome this gap between the pharmacodynamics
and pharmacokinetics an alternative route was envis-
aged starting from chlorothiophene.
N-Boc-O-methyl-D-serine 4 was prepared in one step
from the corresponding D-serine 3 by usingmethyl
iodide and sodium methoxide in tetrahydrofuran.6
Couplingof the resultingacid
methyl-phenyl)-morpholin-3-one under standard pep-
tide-couplingtechniques egnerated the amide 5. The
N-Boc protectinggroup was subsequently removed by
treatment with trifluoroacetic acid in dichloromethane
to give the amine 6. Finally, intermediate 6 was
converted with 5-chloro-thiophene-2-carboxylic acid
again under peptide-coupling conditions to the targeted
compound EMD 495235.
4 and 4-(4-amino-2-
A superimposition of representative fXa inhibitors of
both classes of compounds based on the conformation
found by co-crystallization of those inhibitors with
fXa4 showed, that the chlorothiophene residue is a good
isoster for the substitution of the chlorophenyl ringin
the S1 pocket (e.g., compound 1). However, the urea
moiety seemed to be not the optimal linkingelement
between P1 residue and amino acid part. Computer
modellingstudies showed, that a methanone rgoup
could be a good replacement for the carbamoyl part
within the chlorophenyl derivatives giving rise to
chlorothiophenecarboxamides.
Scheme 2 illustrates a representative synthesis of proto-
typic P4 morpholinone buildingblock 10. Commercially
available 4-nitro-phenylamine 7 was subjected to a
standard acylation procedure with (2-chloro-ethoxy)-
acetyl chloride7 to give anilide 8. Ringclosure of this
intermediate with potassium carbonate in acetonitrile
afforded the correspondingmorpholinone 9. Finally,
the nitro group of compound 9 was reduced under
hydrogenation conditions to the desired anilino morpho-
linone 10.
Recently, the 5-chlorothiophene–methanone fragment
as a surrogate for the P1 moiety has also been described
in the literature, although with different central scaffolds
and P4 residues.5
Herein, we report on structure–activity relationship
studies around variations on the amino acid core as
central scaffold and the P4 residue. Further studies of
selected compounds on water solubility and pharmaco-
kinetic parameters are presented resultingin the discov-
ery of EMD 495235.
O
a.
O2N
N
H
O
O2N
NH2
Cl
8
7
b.
2. Chemistry
O
N
O
N
c.
Startingwith commercially available N-tert-butoxy-
carbonyl (Boc) protected amino acids the targeted
compounds 11a–x bearinga P1 chlorothiophenecarbox-
amide, an amino acid as core structure and the respec-
tive P4 anilino side chain were synthesized in three or
four steps dependingon the employed amino acid. A
representative synthesis of the modified inhibitors 11a–
x is described in Scheme 1, exemplified by morpholinone
11r (EMD 495235).
O2N
O
H2N
O
9
10
Scheme 2. General synthesis of the P4 morpholinones, illustrated at 4-
(4-amino-phenyl)-morpholin-3-one 10. Reagents and conditions: (a)
(2-chloro-ethoxy)-acetyl chloride, toluene, 95ꢁC, 5h; (b) K2CO3,
MeCN, rt, 8h; (c) H2, Pd–C-5%, MeOH, 1.0bar, 35ꢁC, 1h.
Me
O
Me
O
OH
OH
b.
a.
O
O
O
H
N
Me
N
OH
O
N
H
O
N
H
O
N
H
O
O
O
4
O
3
5
O
c.
Me
O
Me
O
O
d.
H
H
N
N
Me
Me
S
H2N
N
H
Cl
O
O
N
N
O
O
6
EMD 495235 = 11r
O
O
Scheme 1. General synthesis of the final compounds 11a–x, illustrated at EMD 495235 (=11r). Reagents and conditions: (a) NaOMe, MeI, THF, rt,
24h; (b) 4-(4-amino-2-methyl-phenyl)-morpholin-3-one, EDC, HOBt, DMF, rt, 18h; (c) TFA, DCM, rt, 2h; (d) 5-chloro-thiophene-2-carboxylic
acid, EDC, HOBt, DMF, rt, 24h.

W. W. K. R. Mederski et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5817–5822
5819
3. Results and discussion
ties within our different series. Therefore, the phenyl
morpholinone became the focus of further efforts. To
probe additional interactions in the S4-pocket the effects
of substitution on the phenyl ringrelative to the mor-
pholinone were investigated. The results of ortho-substi-
tution are shown in Table 1, which gave the best
improvements concerningfXa binding. The initial com-
pound 11m in this series contained the isobutyl group in
C-2 position of the central moiety and the ortho-methyl
on the phenyl ring, which directly led to a 2-fold increase
in potency compared to 11f. The replacement of the
isbobutyl in compound 11f by a propyl (11n), an ethyl
(11o) or a methyl (11p) resulted in slightly diminished
bindingaffinities. However, introducingmore polar
C-2 side-chains based on oxygen further potency
enhancement was observed. Compared to isobutyl 11f
the hydroxymethyl analogue 11q displayed a 2-fold de-
creased affinity in binding, whereas the corresponding
methylserine 11r showed enhanced potency representing
the most potent chlorothiophenecarboxamide inhibitor.
Isosteric replacement of the methyl group in 11r with a
trifluoromethyl (11s), a chloro (11t) or a fluoro (11u)
resulted in decreased bindingaffinities. Larger ortho-al-
kyl substituents than methyl were not prepared due to
synthetic reasons. Further C-2 variations of the chain
length based on the oxygen atom had a negative effect
on fXa binding. Elongation of serine 11q at the hydroxyl
function with ethyl (11v), propyl (11w) or methoxyethyl
(11x) decreased potency up to 2-fold.
The structure–activity relationships surroundingthe
inhibitors incorporatingthe different side chains of the
amino acids and the modified P4 ligands are outlined
in Table 1. Compounds 11a–x were assayed against
human fXa, thrombin and trypsin as previously
described.8 Briefly protease activity was monitored in
vitro usingprotease-specific chromoegnic substrates.
Antiprotease activity of drugs was calculated from the
OD ratio of drugand vehicle containingassay. For
thrombin and trypsin all compounds displayed
IC50 > 10lM.
2-Amino-acetamide 11a containingthe unsubstituted
phenyl morpholinone 10 and a chlorothiophene–metha-
none was initially evaluated as an inhibitor of factor Xa.
Although already submicromolar in binding affinity,
this achiral compound delivered the weakest potency
within our series. However, this result encouraged us
to explore a variety of different substituents at the C-2
position of the central amino acid unit. Based on the
findings within our chlorophenylurea series D-amino
acids were employed at the beginning. A selection of this
chosen derivatives is summarized in Table 1. Introduc-
tion of aliphatic C-2 side-chains markedly increased po-
tency. The n-propyl substituent (11d) was the most
effective, which displayed a 58-fold potency improve-
ment compared to the parent 11a. The corresponding
C-2 methyl (11b) and ethyl (11c) side-chain derivatives
were less active than the norvaline 11d. Steric crowding
closer to the C-2 branchingpoint ( 11e) resulted in de-
creased fXa inhibitory activity. Replacement of the iso-
propyl group of 11e with isobutyl 11f regained potency,
indicating that a larger group more apart from the
branchingpoint is still tolerated. Leucine 11f is equipo-
tent to the correspondingnorvaline 11d.
Next, the physicochemical profile of the more potent 2-
methylphenyl morpholinones was examined. The re-
sults of some selected inhibitors 11m, 11o–r, 11w and
11x are summarized in Table 2. The compounds are ar-
ranged in descending order of the octanol–water distri-
bution coefficient (logD at pH7). A decreasinglog D
correlates with an increase in solubility. Calculated
and experimental values are given in columns 5 and 6
of Table 2. Oral absorption (passive diffusion) depends
on the permeability of the compound and the solubi-
lity, which can be a limitation for the concentration
gradient through the intestine wall. The permeability
depends on lipophilicity (logD), the degree of ioniza-
tion and the molecular size. The factor Xa inhibitors
showed a good permeability but the solubility of some
derivatives was not good enough for a quantitative
absorption at higher doses. Therefore, the maximum
absorbable dose concept10 (MAD) was used for the
fXa inhibitors to avoid problems with the formulation
of the compounds and the dosingin toxicoloigcal
experiments. The target dose should be at least 10mg/
kg. To achieve this a water solubility of >300lg/mL
is necessary. This holds true for the compounds 11q,
11r (EMD 495235) and 11x.
Replacement of the alkyl groups in 11a–f with an aro-
matic phenyl ringresulted in the two phenlylgycines
11g and 11i, respectively, dependingon the employed
chiral amino acid. Both R- and S-enantiomers were
equipotent but showed a slight loss in binding compared
to leucine 11f. However, there is a preference for R-con-
figuration at C-2 in aliphatic substituted inhibitors. The
S-configurated isomer 11i is about nine times less potent
than R-leucine 11f.
In order to support our assumption that a methanone is
more favourable as a linker in 2-chlorothiophenes than
a carbamoyl moiety, the urea analogue of D-norvaline
11d was synthesized. Bindingaffinity dropped about
16-fold compared to that of the correspondingmetha-
none 11d (data not shown in Table 1), 9 which confirmed
our initial hypothesis.
For EMD 495235 the inhibition constant (Ki value)
for human Factor Xa was determined (Ki = 6.8nM).
EMD 495235 was also evaluated for his anticoagulant
activity in the standard coagulation assay for activated
thromboplastin time (APTT) and prothrombin time
(PT). The concentration required to double plasma
clottingtime of APTT and PT was one micromolar,
respectively.
To explore P4 modification morpholinone in 11f was
replaced with different carbonyl bearingheterocycles.
Two promisingderivatives are shown in Table 1. With
pyrazinone 11k we observed a 2-fold loss in binding
affinity, whereas pyridinone 11l is equipotent to the par-
ent 11f. In general, pyridones could not compete with
morpholinones with regard to pharmacokinetic proper-

5820
W. W. K. R. Mederski et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5817–5822
Table 1. Factor Xa bindingaffinity data of selected chlorothiophenecarboxamides 11
O
R
H
N
R1
R2
S
N
H
Cl
O
Compd
R
H
R1
R2
IC50 fXa (nM)
700.0
N
O
O
O
O
11a
H
H
N
O
CH3
11b
59.0
N
O
CH2CH3
11c
11d
11e
11f
11g
11h
11i
H
20.0
12.0
66.0
15.0
29.0
26.0
130.0
39.0
16.0
7.2
N
O
O
O
O
O
O
O
N
CH2CH2CH3
CH(CH3)2
H
N
O
H
N
O
CH2CH(CH3)2
C6H5
H
N
O
H
N
O
C6H5
H
N
O
CH2CH(CH3)2
CH2CH(CH3)2
CH2CH(CH3)2
CH2CH(CH3)2
H
N
O
11k
11l
H
N
O
H
N
O
O
O
O
O
O
11m
CH3
N
O
11n
11o
11p
11q
CH2CH2CH3
CH2CH3
CH3
CH3
CH3
CH3
CH3
9.5
11.0
15.0
29.0
N
O
N
O
N
O
CH2OH

W. W. K. R. Mederski et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5817–5822
5821
Table 1 (continued)
Compd
R
R1
R2
IC50 fXa (nM)
5.5
N
O
O
11r (EMD495235)
CH3
CH2OCH3
CH2OMe
O
N
11s
CF3
14.0
O
N
O
N
O
O
O
CH2OCH3
CH2OCH3
11t
Cl
F
13.0
22.0
11u
N
O
O
O
O
CH2OCH2CH3
11v
11w
11x
CH3
CH3
CH3
10.0
12.0
11.0
N
O
CH2O(CH2)2CH3
CH2O(CH2)2OCH3
N
O
Table 2. Physicochemical parameters and the estimated maximum absorbable dose of selected 2-methylphenyl morpholinones
R
O
H
N
Me
N
S
N
H
Cl
O
O
O
Compd
R
MW
logD pH7a
Solubility (lg/mL)a
Solubility (lg/mL)b
MADc (mg/kg)7
11m
11w
CH2CH(CH3)2
CH2O(CH2)2CH3
CH3
464
480
422
436
438
452
496
2.8
1.7
1.4
1.9
1.0
0.7
0.4
5
37
29
92
0.6
1.9
11p
11o
130
41
134
308
520
510
1054
2.8
6.5
CH2CH3
CH2OH
11q
11r (EMD 495235)
11x
260
380
340
11.0
11.0
22.0
CH2OCH3
CH2O(CH2)2OCH3
a Calcd: WSKowWin 1.40 (Syracuse Res. Corp.).
b Meas.: shake flask, T = 37ꢁC, phosphate buffer pH7.
c Maximum absorbable dose.
Orientatingstudies of the pharmacokinetics of compound
EMD 495235 were investigated in Wistar rats, Beagle
dogs and Cynomolgus monkeys after treatment with
single intravenous and oral doses. A compilation of the
pharmacokinetic parameters is presented in Table 3.
The outcome of these studies will be presented elsewhere
in due course.
4. Conclusion
EMD 495235 is characterized by rapid absorption from
the gastrointestinal tract, plasma elimination half-lives
of 0.57–2.3h, a clearance between 0.25 and 1.3L/h/kg
and absolute bioavailability of 60–80%.
Based on the structural information from co-crystalliza-
tion studies of different lead series chlorothiophenecarb-
oxamides have been identified. From these
investigations
a promising drug candidate EMD
495235 has been characterized as a compound with good
in vitro activities, acceptable water solubility and high
absolute bioavailability in different animal models.
EMD 495235 met our criteria so far so that in vivo
investigations in different animal models were initiated.

5822
W. W. K. R. Mederski et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5817–5822
Table 3. Pharmacokinetic parameters of EMD 495235
Rat; male (N = 3)
Dog; female (N = 2)
iv
0.15
Monkey; female (N = 2)
Administration
Dose (mg/kg)
Cmax (ng/mL)
tmax (h)
iv
0.2
po
0.5
po
0.5
iv
0.2
po
0.5
270
2.5
92
0.5
161
0.5
AUC (ng/mL · h)
0–6h
271
211
1100
0–1
212
117
803
t1/2 (h)
1.1
0.57
1.3
2.3
CL (L/h/kg)
Vss (L/kg)
Bioavailability (%)
0.94
1.6
0.25
0.81
0.86
ꢀ60
ꢀ60
ꢀ80
Chem. Abstr. 2002, 137, 125096; (b) Mederski, W. W. K.
R.; Germann, M. Bioorg. Med. Chem. Lett. 2003, 13,
3715.
EMD 495235 was chosen as a candidate for further in
vivo investigations as an anticoagulating drug.
4. Mederski, W. W. K. R.; Dorsch, D.; Anzali, S.; Gleitz, J.;
Cezanne, B.; Tsaklakidis, C. Bioorg. Med. Chem. Lett.
2004, 14, 3763.
Acknowledgements
5. Bauer, S. M.; Goldman, E. A.; Huang, W.; Su, T.; Wang,
L.; Woolfrey, J.; Wu, Y.; Zuckett, J. F.; Arfsten, A.;
Huang, B.; Kothule, J.; Lin, J.; May, B.; Sinha, U.; Wong,
P. W.; Hutchaleelaha, A.; Scarborough, R. M.; Zhu, B.-Y.
Bioorg. Med. Chem. Lett. 2004, 14, 4045.
6. For N-Boc-O-methyl-L-serine, see: Chen, F. M. F.;
Benoiton, N. L. J. Org. Chem. 1979, 44, 2299.
7. Salmi, E. J.; Leimu, R.; Kallio, H. Suom. Kemistil. B 1944,
17.
For their skilful experimental work or their valuable
support, we would like to thank Christine Altwein,
Dr. Hans-Markus Bender, Gabriele Czmok, Herbert
Glas, Christina Heiner, Gabriele Mahr, Hy-May
Nguyen, Philipp Roth, Yvonne Schmidt, and Man-
fred-Hans Tabola.
References and notes
8. Hara, T.; Yokoyama, A.; Ishihara, H.; Yokoyama, Y.;
Nagahara, T.; Iwamoto, M. Thromb. Haemostasis 1994,
71, 314.
9. FXa bindingaffinity of 2-[3-(5-chlorothiophen-2-yl)-ureido]-
3-methoxy-N-[3-methyl-4-(3-oxo-morpholin-4-yl)-phenyl]-
propionamide, R-enantiomer: IC50 = 190.0nM.
10. Curatolo, W. Pharm. Sci. Technol. Today 1998, 1,
387.
1. Adang, A. E. P.; Rewinkel, J. B. M. Drugs Future 2000,
25, 369.
2. Ewing, W. R.; Pauls, H. W.; Spada, A. P. Drugs Future
1999, 24, 635.
3. (a) Dorsch, D.; Mederski, W.; Tsaklakidis, C.; Cezanne,
B.; Gleitz, J; Barnes, C. PCT Int. Appl. WO 0257236,