Bicyclic pyridones as potent, efficacious and orally bioavailable thrombin inhibitors

Article abstract of DOI:10.1016/S0960-894X(00)00170-0

A new class of conformationally constrained thrombin inhibitors is described. These compounds contain a unique bicyclic pyridone scaffold which serves as a P3P2 dipeptide surrogate. The synthesis and antithrombotic activity of these inhibitors is reported. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00170-0

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1069±1072
Bicyclic Pyridones as Potent, Ecacious and Orally Bioavailable
Thrombin Inhibitors
Craig A. Coburn, a,* Diane M. Rush, a Peter D. Williams, a Carl Homnick, a
Elizabeth A. Lyle, b S. Dale Lewis, c Bobby J. Lucas, Jr., c Jillian M. Di Muzio-Mower, c
Marylou Juliano, c Julie A. Krueger, c Kari Vastag, d I-Wu Chen d and Joseph P. Vacca a
aDepartment of Medicinal Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
bDepartment of Pharmacology, Merck Research Laboratories, West Point, PA 19486, USA
cDepartment of Biological Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
dDepartment of Drug Metabolism, Merck Research Laboratories, West Point, PA 19486, USA
Received 17 January 2000; accepted 9 March 2000
AbstractÐA new class of conformationally constrained thrombin inhibitors is described. These compounds contain a unique
bicyclic pyridone sca€old which serves as a P3P2 dipeptide surrogate. The synthesis and antithrombotic activity of these inhibitors
is reported. # 2000 Elsevier Science Ltd. All rights reserved.
Thrombus formation is the culmination of a series of
enzymatic reactions in which the serine protease throm-
bin plays an essential and multifaceted role. For exam-
ple, thrombin acts directly to promote clot formation by
catalyzing the hydrolysis of the soluble plasma protein
®brinogen to insoluble ®brin. Thrombin is also a potent
stimulator of platelet aggregation and is required for the
biosynthesis of other coagulation factors. Defects in the
balance between these coagulation factors and their
endogenous inhibitors can have serious consequences,
such as the formation of thrombi in blood vessels.1
These thromboembolic events can contribute to a num-
ber of health conditions such as angina, ischemia, myo-
cardial infarction, and stroke. For these reasons,
inhibition of thrombin has been viewed as an attractive
mechanism for the treatment and prevention of both
venous and arterial thrombosis. Over the last 10 years
there has been considerable progress in the area of
thrombin inhibitor research.2 Initial ventures focused on
the synthesis of peptide inhibitors based on a Phe-Pro-
Arg template. Recent work has focused on nonpeptide
inhibitors as a strategy to improve pharmacological and
pharmacokinetic properties. In this context, several
groups have reported inhibitors which incorporate a
pyridone or pyrazinone as a P2 amino acid surrogate
within the tripeptide sca€old.3 6 This work has led to the
emergence of potent, ecacious and bioavailable throm-
bin inhibitors such as 15 and has spawned interest in more
novel structures. As part of our program aimed at devel-
oping structurally diverse thrombin inhibitors, we have
investigated a number of compounds which contain a
novel bicyclic pyridone P3P2 skeleton. This communica-
tion describes the design, synthesis, and selected biological
results of these unique dipeptide mimics.
In the context of a thrombin inhibitor containing a
bicyclic pyridone relating to inhibitor 1, we envisioned
the constraint of the P3 benzyl sulfonamide side chain
by cyclization to C-4 of the pyridone ring. Topographic
considerations within the thrombin active site required
an R con®guration for the newly formed chiral center
adjacent to the lactam carbonyl group in order to access
to the distal hydrophobic S3 binding pocket occupied
by the phenyl ring in structure 1. Therefore, template 2
became an attractive target for synthesis since its rigid
structure would preorganize the inhibitor for optimal
interaction between its lipophilic P3 ligand and the
enzyme while maintaining the essential hydrogen bond-
ing elements required for interaction with the Gly 216
residue present in the thrombin backbone (Fig. 1).
The chemistry employed for the synthesis of this new
class of compounds is illustrated in Scheme 1. The suc-
cess of the reaction scheme hinged on our ability to
*Corresponding author. Tel.: +1-9215-652-5511;
e-mail: craig_coburn@merck.com
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00170-0

1070
C. A. Coburn et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1069±1072
recyclizations of the unwanted glycolate byproducts to
the back to pyridones 10 using the dealkylation proce-
dure described above. Chiral preparative HPLC proved
to be a successful method for separating the enantio-
mers of lactams 11.9 Hydrogenolysis of benzyl esters 11
a€orded carboxylic acids 12 in quantitative yield. The
resulting acids were coupled to 2-amino-5-aminomethyl-
6-methylpyridine6 to provide target compounds 2.
Figure 1. Conformational restriction of pyridone 1 to generate bicyclic
template 2.
prepare the versatile nitroaldehyde 7, in order to max-
imize analogue production. The desired aldehyde was
synthesized in ®ve steps from commercially available
ethyl nitroacetate (3). Reaction of ester 3 with ethanolic
ammonia a€orded nitroacetamide 4. Construction of the
pyridone nucleus was accomplished by heating an aque-
ous solution of acetamide 4 with commercially available
ethyl 2,4-dioxovalerate in the presence of piperdinium
acetate. This procedure reproducibly a€orded nitro ester
5 in 28% yield and was routinely run in 100 g batches.
Protection of the pyridone nucleus was necessary as the
highly electron de®cient pyridone was susceptible to
electron transfer reactions when exposed to a number of
organometallic reagents. Therefore, pyridone 5 was
masked as its methoxypyridine analogue by reaction
with Meerwein's salt in warm dichloromethane.7 The
oxidation state of the C-4 carbonyl group was adjusted
via a standard two-step reduction/oxidation protocol to
a€ord the pivotal aldehyde 7 in 20% overall yield from
ester 3. Horner±Emmons ole®nation of aldehyde 7 with
an appropriately substituted phosphonoacetate deriva-
tive provided mixtures of cis and trans trisubstituted
enoates 8, which bore the latent P3 substituent.
For the synthesis of a,a-dialkylated lactam derivatives,
the lactam NH of intermediates 9 was protected using
Boc anhydride and catalytic DMAP to a€ord 13
(Scheme 2). Initial attempts to directly alkylate 9 with
various bases resulted in low yields of lactams 14.
Moreover, treatment of protected lactams 13 with a
number of di€erent bases resulted in products which
arose from deprotonation of the pyridine methyl group.
This problem was circumvented by using potassium
tert-butoxide as the base in the presence of equimolar
sodium hydride. Employment of these reaction condi-
tions a€orded, after work up, good yields of the depro-
tected, dialkylated lactam 14. Presumably sodium
t
hydride acts to irreversibly trap the BuOH which is
formed in the reaction. Exclusion of NaH in favor of
excess KOtBu or NaOtBu resulted in no reaction and
complete recovery of 13.
Scheme 2. (a) Boc2O, DMAP, MeCN; (b) NaH, BuOK, TBAI, R0X,
t
The key step in the reaction (Scheme 1) was a one-pot,
double reductive lactamization of nitroenoates 8, which
provided access to the bicyclic lactam core structures.
The pyridone rings were then regenerated upon brief
exposure of methoxypyridines 9 to a melt of pyridinium
hydrochloride.8 Alkylation of the newly formed bicyclic
pyridones 10 with one equivalent of benzyl bromoace-
tate a€orded equal mixtures of N- and O-alkylated pro-
ducts which were easily separated by column
chromatography. Attempts to bias product distribution
in favor of N-alkylation using standard strategies were
unsuccessful. However, we were able to perform iterative
THF, 50 ꢀC; (c) Scheme 1, steps (g)±(j).
The inhibition constants (Ki values) found for bicyclic
pyridones 2 versus thrombin and the closely related
protease, trypsin, are given in Table 1. The compounds
listed in this table are potent inhibitors of thrombin and
are selective over trypsin. These compounds do not
inhibit other serine proteases such as factor Xa, plasmin
and tissue plasminogen activator (Ki>8 0 mM). Nano-
molar binding anities are easily obtained by incor-
porating small alkyl groups at P3 as demonstrated by
Scheme 1. (a) NH3, EtOH, 80 ꢀC, 3 h; (b) ethyl 2,4-dioxovalerate, piperdinium acetate, H2O; (c) Me3O+BF4 , DCM; (d) (i) Dibal-H, DCM; (ii)
(ClCO)2, DMSO, Et3N; (e) (MeO)2P(O)CHRCO2Me, NaH, THF; (f) H2, Pd(C), MeOH, 55 ꢀC; (g) pyridinium HCl, 155 ꢀC, 5 min; (h) (i) Cs2CO3,
BrCH2CO2Bn, DMF, ii. Chiralcel-OD column; (i) H2, Pd(C), MeOH; (j) EDC, HOBT, NMM, 2-amino-5-aminomethyl-6-methylpyridine.

C. A. Coburn et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1069±1072
1071
Table 1. Inhibition constants (Ki) versus thrombin and trypsin, in vitro anticoagulant potency (2 aptt) and in vivo antithrombotic ecacy.
Compd
R
R0
Ki
(nM)
2ÂAPTT
rat
FeCl3
Final plasma
conc (nM)
b
(nM)a
Thrombin
Trypsin
2a
2b
2c
2d
2e
2f
2g
2h
2i
H
Et
n-Pr
H
c-PrCH2
c-BuCH2
PhCH2
i-Bu
i-Bu
i-Bu
H
H
H
n-Pr
H
H
H
H
Me
i-Bu
350
101,000
4200
510
22,000
770
420
12,700
580
860
ndc
310
180
nd
150
190
200
220
160
390
1.6
0.33
36
0.21
0.084
0.36
0.40
0.24
0.27
(1/6)
0/6
185
596
0/6
255
2j
1500
aHuman plasma.
bOcclusions after iv infusion at 10 mg/kg/min. Parenthesis infusion at 3 mk/kg/min.
cnd=not determined.
the ethyl substituted lactam 2b (Ki=1.6 nM). Further
SAR about the P3 substituent reveals that binding a-
nity is enhanced by the incorporation of small
cycloalkyl ligands (e.g., 2e,f). Molecular modeling of the
n-propyl substituted lactam 2c in the enzyme active site
shows a preference for a single enantiomer to achieve
optimal binding. This bias is supported experimentally
when comparing the Ki value of lactam 2c (Ki=0.33
nM) to the Ki value of its antipode 2d (Ki=36 nM). It is
interesting to note that alkylated lactams bearing an S-
con®guration are still 10-fold more potent than the
unsubstituted parent lactam 2a. This result suggests that
additional binding energy is gained by substituting at
the b face of these inhibitors. Indeed, the addition of
either a methyl (2i) or isobutyl (2j) group a-to the lac-
tam carbonyl in 2h results in an increase potency. Fur-
thermore, these dialkylated lactams display better
trypsin selectivity than the corresponding mono-
alkylated analogues.
Table 2. Oral absorption kinetics for compounds 1, 2h and 2i in
dogsa
Compd
Cmax
(mM)
tmax
(min)
t1=2
AUC
(mM h)
1
2h
2i
0.99
0.18
1.37
75
70
30
154
141
90
4.42
0.75
2.97
aDose of 1 mg/kg; n=2.
2. A brief comparison is given to illustrate that subtle
changes at P3 have a pronounced e€ect on the overall
absorption pro®le of a compound in this series without
greatly a€ecting the potency or ecacy. For example,
lactam 2h displays inferior absorption properties relative
to 1, but maintains a comparable half-life. On the other
hand, the a-methyl homologue, 2i, shows similar plasma
levels as 1 but a shorter terminal half-life.
In summary, we have rationally designed a new series of
thrombin inhibitors based on the 3-amino-2-pyridone
acetamide template 1. These compounds include a unique
conformationally constrained peptidomimetic sca€old.
Since the synthesis described is versatile, it is amenable to
SAR work at both the P3 and P1 regions of the molecule.
Second and third generation non-lactam containing
bicyclic inhibitors which show improved pharmacoki-
netics have been synthesized and the results of the studies
will be disclosed in subsequent publications.
Table 1 illustrates the remarkably low 2Â APTT values
inherent to these bicyclic inhibitors. This in vitro assay
measures the concentration of inhibitor that is required
to double clotting times and is dependent upon both the
potency and protein binding of the inhibitor. The pro-
tein binding of 2e and 2f in human plasma is 59 and
81%, respectively. Since cyclopropyl analogue 2e has a
higher free fraction, it is understandable that it has a
lower 2Â APTT value despite being 2.5-fold less potent
than its homologue 2f. Table 1 also lists the results from
the rat ferric chloride model of arterial thrombosis.10 As
suggested by the intrinsic potencies of these bicyclic
inhibitors, rats dosed with 10 mg/kg/min iv of 2f or 2h
experienced no arterial clotting. In fact, when 2e was
challenged at a 3 mg/kg/min iv dose, good ecacy was
maintained and the ®nal plasma concentrations
remained above the 2Â APTT concentration.
Acknowledgements
The authors thank Matthew Zrada, Kenneth Anderson
and Patrice Ciecko-Steck for analytical support.
References and Notes
Pharmacokinetic parameters for compounds 1, 2h and
2i after oral administration in dogs are shown in Table
1. Goldsack, N. R.; Chambers, R. C.; Dabbag, K.; Laurent, J.
Int. J. Biochem. Cell Biol. 1998, 30, 641.

1072
C. A. Coburn et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1069±1072
2. Fevig, J. M.; Wexler, R. R. Ann. Rep. Med. Chem. 1999, 34,
81.
3. Sanderson, P. E. J.; Dyer, D. L.; Naylor-Olsen, A. M.;
Vacca, J. P.; Gardell, S. J.; Lewis, S. D.; Lucas, B. J.; Lyle, E.
A.; Lynch, J. J.; Mulichak, A. M. Bioorg. Med. Chem. Lett.
1997, 7, 1497.
4. Tamura, S. T.; Semple, J. E.; Reiner, J. E.; Goldman, E. A.;
Brunck, T. K.; Lim-Wilby, M. S.; Carpenter, S. H.; Rote, W.
E.; Oldeshulte, G. L.; Richard, B. M.; Nutt, R. F.; Ripka, W.
C. Bioorg. Med. Chem. Lett. 1997, 7, 1543.
5. Sanderson, P. E. J.; Cutrona, K. J.; Dorsey, B. D.; Dyer, D.
L.; McDonough, C. J.; Naylor-Olsen, A. M.; Chen, I.-W.;
Chen, Z.; Cook, J. J.; Gardell, S. J.; Krueger, J. A.; Lewis, S.
D.; Lin, J. H.; Lucas, B. J.; Lyle, E. A.; Lynch, J. J.; Stranieri,
M. T.; Vastag, K.; Shafer, J. A.; Vacca, J. P. Bioorg. Med.
Chem. Lett. 1998, 8, 817.
6. Sanderson, P. E. J.; Lyle, T. A.; Cutrona, K. J.; Dyer, D. L.;
Dorsey, B. D.; McDonough, C. J.; Naylor-Olsen, A. M.;
Chen, I.-W.; Chen, Z.; Cook, J. J.; Cooper, C. M.; Gardell, S.
J.; Hare, T. R.; Krueger, J. A.; Lewis, S. D.; Lin, J. H.; Lucas,
B. J.; Lyle, E. A.; Lynch, J. J.; Stranieri, M. T.; Vastag, K.;
Yan, Y.; Shafer, J. A.; Vacca, J. P. J. Med. Chem. 1998, 41,
4466.
7. Clive, D. L. J.; Sedgeworth, J. J. Heterocycl. Chem. 1987,
24, 509.
8. Ho€man, J. M.; Wai, J. S.; Thomas, C. M.; Levin, R. B.;
O'Brien, J. A.; Goldman, M. E. J. Med. Chem. 1992, 35, 3784.
9. Separation of the enantiomers of these bicyclic compounds
was performed on a Chiracel OD column (9:1 hexane:EtOH
containing 0.1% diethylamine).
10. Kurz, K. D.; Main, B. W.; Sandusky, G. E. Thromb. Res.
1990, 60, 269.