Heterocyclic core analogs of a direct thrombin inhibitor

Article abstract of DOI:10.1016/j.bmcl.2014.01.002

Thrombin is a serine protease that plays a key role in blood clotting. Pyrrolidine 1 is a potent thrombin inhibitor discovered at Merck several years ago. Seven analogs (2-8) of 1 in which the pyrrolidine core was replaced with various heterocycles were p

Full text of DOI:10.1016/j.bmcl.2014.01.002

Bioorganic & Medicinal Chemistry Letters 24 (2014) 1111–1115
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Heterocyclic core analogs of a direct thrombin inhibitor
b
b
b
b
Timothy A. Blizzard a, , Sanjay Singh , Basanagoud Patil , Naresh Chidurala , Venukrishnan Komanduri ,
Samarpita Debnath b, Sergei Belyakov b, Alejandro Crespo a, Alice Struck a, Marc Kurtz a, Judyann Wiltsie a,
Xun Shen a, Lisa Sonatore a, Marta Arocho a, Dale Lewis a, Martin Ogletree a, Tesfaye Biftu a
⇑
a Merck Research Laboratories, Rahway, NJ, USA
b Albany Molecular Research, Singapore Research Center, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history:
Thrombin is a serine protease that plays a key role in blood clotting. Pyrrolidine 1 is a potent thrombin
inhibitor discovered at Merck several years ago. Seven analogs (2–8) of 1 in which the pyrrolidine core
was replaced with various heterocycles were prepared and evaluated for activity against thrombin, clot-
ting factors VIIa, IXa, Xa, and XIIa, and trypsin. The thiomorpholine analog 6 was the most active, essen-
tially matching the thrombin inhibitory activity of 1 with slightly improved selectivity over trypsin.
Ó 2014 Elsevier Ltd. All rights reserved.
Received 18 November 2013
Revised 1 January 2014
Accepted 3 January 2014
Available online 10 January 2014
Keywords:
Thrombin inhibitor
DTI
Thrombosis
Anti-thrombotic
Thrombotic events (stroke, heart attack, etc.) are an important
cause of morbidity and mortality, especially in older patients.1
thrombin and Factor Xa inhibitors are in various stages of
development.9
A
safe and effective drug that could prevent thrombotic events
would clearly have a significant positive impact on life expectancy
and quality of life. Fortunately, thrombogenesis is a complicated
biological process involving a number of enzymes that provide
attractive targets for potential new therapies.2 Several approved
drugs, including Warfarin (a vitamin K epoxide reductase inhibi-
tor), Dabigatran (a thrombin inhibitor), Rivaroxaban and Apixaban
(factor Xa inhibitors) are currently used prophylactically to reduce
the risk of thrombotic events.3 Unfortunately, current therapy suf-
fers from significant side effects. Bleeding, in particular, remains a
significant concern. Thus, there remains an unmet medical need for
improved antithrombotic agents that would not have the bleeding
liability of current drugs. This is a very active field of research and,
in addition to those drugs already on the market, numerous com-
pounds employing a variety of mechanisms are in develop-
ment.4–6
Thrombin (Factor IIa) is a serine protease that plays a key role in
blood clotting. Selective thrombin inhibition is an established
mechanism for thromboembolism prevention.7,8 Three small mol-
ecule thrombin inhibitors have reached the market. The first,
Ximelagatran, was subsequently withdrawn due to unacceptable
liver toxicity.7 Two newer direct thrombin inhibitors, Dabigatran
and Argatroban, are currently on the market. Numerous other
H2N
OH
N
NH
O
CO2H
H2N
N
H
O
NH
O
S
H
NH
N
N
O
O
HN
CO2Et
Argatroban
Ximelagatran
O
N
N
H2N
CO2Et
NH
nC6H13
O
O
N
N
N
Dabigatran
Pyrrolidine 1 is a potent thrombin inhibitor discovered10 at
Merck several years ago as part of a program directed at finding
clinically useful thrombin inhibitors.10–13
⇑
Corresponding author.
E-mail address: timblizzard@comcast.net (T.A. Blizzard).
0960-894X/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2014.01.002

1112
T. A. Blizzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 1111–1115
BOC
N
BOC
N
BOC
N
Cl
Cl
i
ii
iii
Cl
H
N
O
O
O
N
N
H
N
H2N
N
O
CBZ OCH3
OH
CBZ OH
21
O
AcO
NH2
O
• TFA
16
22
NHBOC
OH
9
10
H
1
N
N
N
•TFA
O
N
iv
v
O
O
N
N
H
CBZ OCH3
23
CBZ OCH3
In their original report, Morissette et al. noted that reducing the
core ring size (i.e. replacement of pyrrolidine with azetidine) re-
sulted in reduced thrombin inhibitory activity.10 However, the ef-
fect of ring expansion was unknown. In an attempt to discover
OCH3
24
25
N
N
vi
vii
viii
Cl
O
O
N
N
OCH3
OH
O
O
OAc
OH
26
Cl
27
i
ii
OH
OH
N
N
H
N
N
N
N
ix
FMOC
O
O
H
H
N
11
12
N
Cl
Cl
O
O
O
NHBOC
O
NH2
• 2 TFA
H
N
iii
iv
H
N
OH
OH
28
4
N
N
H
O
FMOC
O
Scheme 3. Reagents and conditions: (i) CBZ-Cl, NaHCO3, 1,4-dioxane, H2O, room
temp, 16 h, 82%; (ii) K2CO3, CH3I, DMF, room temp, 2 h, 74%; (iii) TFA, CH2Cl2, room
temp, 30 m, 97%; (iv) paraformaldehyde, Et3N, NaCNBH3, CH3OH, AcOH, 0 °C to
room temp, 2 h, 70%; (v) H2 (1 atm), 10% Pd/C, CH3OH, room temp, 3 h, 71%; (vi) 10,
Et3N, CH2Cl2, room temp, 30 m, 40%; (vii) LiOH, THF, H2O, CH3OH, room temp, 2 h,
59%; (viii) 9, EDC, HOBT, Et3N, DMF, room temp, 16 h, 32%; (ix) TFA, CH2Cl2, room
temp, 1 h, 73%.
NHBOC
Cl
NHBOC
Cl
13
14
H
N
H
N
v, vi
N
N
O
O
NH2
O
O
NHBOC
• TFA
OH
OAc
an improved thrombin inhibitor, and to more fully elucidate the
SAR of the heterocyclic core of 1, we prepared and report herein
a series of six-membered heterocyclic core analogs of 1.
We initially targeted the piperidine analog 2. Starting with acid
11,14 2 was readily synthesized as outlined in Scheme 1. Treatment
of 11 with FMOC-Cl afforded the protected aminoacid 12 in good
yield. The right-hand side chain was then introduced by EDC-med-
iated coupling of 12 with amine 9,10 to afford intermediate amide
15
2
Scheme 1. Reagents and conditions: (i) FMOC-Cl, Na2CO3, 1,4-dioxane, water, 16 h,
room temp, 52%; (ii) 9, EDC, HOBT, DMF, 16 h, room temp, 70%; (iii) piperidine,
CH2Cl2, 1 h, room temp, 80%; (iv) 10, Et3N, THF, À10 °C to room temp, 1 h, 65%; (v)
K2CO3, CH3OH, room temp, 30 m; (vi) TFA, CH2Cl2, room temp, 2 h, 30% for two
steps.
BOC
N
BOC
N
i
ii
O
O
N
i, ii
iii
OH
OH
N
OH
N
H
16
OH
N
FMOC
O
O
FMOC
O
BOC O
17
30
29
BOC
N
Cl
Cl
BOC
N
Cl
Cl
O
O
H
N
iii
iv
BOC
H
N
iv
v
H
H
N
N
N
H
N
H
N
N
O
O
FMOC
O
BOC
FMOC
O
N
H
NHBOC
Cl
N
H
NHBOC
Cl
18
19
31
32
Cl
H
N
BOC
N
Cl
O
N
O
N
v, vi
H
N
H
H
N
vi, vii
H
N
N
N
N
O
O
O
O
O
BOC
NH2
O
NH2
O
O
NHBOC
• TFA
N
H
• 2 TFA
OH
OH
OAc
OAc
20
3
33
5
Scheme 2. Reagents and conditions: (i) FMOC-Cl, Na2CO3, 1,4-dioxane, water, 0 °C
to room temp, 16 h, 44%; (ii) 9, EDC, HOBT, DMF, room temp, 16 h, 78%; (iii)
piperidine, CH2Cl2, room temp, 2 h, 73%; (iv) 10, Et3N, THF, À5 °C to room temp, 1 h,
51%; (v) K2CO3, CH3OH, room temp, 30 m; (vi) TFA, CH2Cl2, room temp, 2 h, 40% for
two steps.
Scheme 4. Reagents and conditions: (i) TFA, CH2Cl2, room temp, 5 h; (ii) FMOC-Cl,
K2CO3, 1,4-dioxane, H2O, room temp, 3 h, 84% for two steps; (iii) 9, EDC, HOBT, DMF,
0 °C to room temp, 5 h, 80%; (iv) piperidine, CH2Cl2, room temp, 1 h, 62%; (v) 10,
Et3N, CH2Cl2, 0 °C to room temp, 16 h, 71%; (vi) K2CO3, CH3OH, room temp, 30 m;
(vii) TFA, CH2Cl2, 0 °C, 30 m, 48% for two steps.

T. A. Blizzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 1111–1115
1113
S
The corresponding piperazine analog 3 was similarly prepared
from the protected piperazine acid 1614 and reagents 9 and 10 as
outlined in Scheme 2.
S
N
i, ii
iii
OH
OH
N
FMOC
O
The synthesis of the N-methyl piperazine analog 4 from 16
(Scheme 3) was more complicated due to the protecting group
manipulations required to install the N-methyl substituent. Treat-
ment of 16 with CBZ-Cl readily afforded the bis-protected pipera-
zine 21. Protection of the acid as the methyl ester followed by
removal of the BOC protecting group and reductive methylation
of amine 23 provided the N-methyl intermediate 24 in good overall
yield. Removal of the CBZ group afforded amine 25. Analog 4 was
readily prepared from 25 by the synthetic route employed for 2.
The morpholine analog 5 was prepared in a similar manner
from the protected amino acid 29,14 as outlined in Scheme 4.
The thiomorpholine analog 6 was prepared from the protected
amino acid 34 by an analogous sequence, as outlined in Scheme 5.
The corresponding sulfoxide 7 was readily prepared by MCPBA
oxidation of intermediate 38 followed by deprotection to afford 7
as outlined in Scheme 6.
BOC O
35
34
Cl
Cl
S
S
H
N
iv
v
H
N
N
H
N
O
FMOC
O
NHBOC
Cl
NHBOC
Cl
36
37
S
N
S
N
H
N
H
N
vi, vii
O
O
NH2
O
O
NHBOC
• TFA
OH
OAc
38
6
Scheme 5. Reagents and conditions: (i) TFA, CH2Cl2 room temp, 2 h; (ii) FMOC-Cl,
NaHCO3, 1,4-dioxane, H2O, room temp, 16 h, 67% for two steps; (iii) 9, EDC, HOBT,
Hunig’s base, DMF, room temp, 16 h, 71%; (iv) piperidine, CH2Cl2, room temp, 2 h,
83%; (v) 10, Et3N, CH2Cl2, À10 °C to room temp, 1 h, 63%; (vi) K2CO3, CH3OH, room
temp, 30 m; (vii) TFA, CH2Cl2, room temp, 2 h, 27% for two steps.
Finally, sulfone 8 was prepared by MCPBA oxidation of sulfide
intermediate 38 under stronger conditions followed by deprotec-
tion (Scheme 7).
The new analogs were generally less potent inhibitors of throm-
bin than the lead pyrrolidine 1 (Table 1). The piperidine analog 2,
although differing from 1 only in the addition of a single methylene
unit, was about five-fold less active than 1. The piperazines 3 and 4,
with an additional basic nitrogen in the ring, were even weaker
inhibitors. The morpholine analog 5 was comparable in activity
to the piperazines. Interestingly, the thiomorpholine analog 6
was the most active of the seven analogs, exhibiting thrombin inhi-
bition comparable to that of the pyrrolidine 1. In addition, 6 dem-
onstrated slightly improved selectivity for thrombin over trypsin
(>13,000X for 6 vs ꢀ3700X for 1), although both compounds were
highly selective. Finally, the sulfoxide analog 7 and the sulfone 8
were significantly less active than the thiomorpholine 6.
Cl
Cl
O
S
S
N
i, ii, iii
H
N
H
N
N
O
O
NHBOC
NH2
O
O
• TFA
OAc
OH
38
7
Scheme 6. Reagents and conditions: (i) MCPBA (0.88 equiv), CH2Cl2, room temp,
30 m, 98%; (ii) K2CO3, CH3OH, room temp, 30 m; (vii) TFA, CH2Cl2, room temp, 2 h,
35% for two steps.
Pyrrolidine 1 (Fig. 1) and the most active analog, thiomorpho-
line 6 (Fig. 2), were docked into a published crystal structure of
thrombin with a close analog of 1 bound in the active site
(2zgb.pdb).16 As expected, 1 and 6 bind by placing the Cl-phenyl
group into the P1 pocket, the core heterocyclic group into the P2
pocket, with the hydroxyl group interacting with G216 in P3 and
the tert-butyl group in the P4 pocket. Thrombin activity differences
for compounds 1–8 can thus be explained by the differential inter-
action of the heterocyclic core with the hydrophobic residues
(shown as sticks in Figs. 1 and 2) framing the P2 pocket (H57,
L99, Y60A and particularly W60D, which caps the pocket). The
hydrophobic pyrrolidine 1, piperidine 2, and thiomorpholine 6
are the more potent analogs. The hydrophilic piperazine 3,
morpholine 5, sulfoxide 7 and sulfone 8 are less active. Although
N-methyl piperazine analog 4 is relatively hydrophobic due to
N-methylation, the additional methyl group introduces a steric
clash with the W60D side chain, accounting for its reduced activity.
Cl
Cl
O
O
S
N
S
i, ii, iii
H
N
H
N
N
O
O
NHBOC
NH2
O
O
• TFA
OAc
OH
38
8
Scheme 7. Reagents and conditions: (i) MCPBA (2.5 equiv), CH2Cl2, room temp, 4 h,
66%; (ii) K2CO3, CH3OH, room temp, 30 m; (vii) TFA, CH2Cl2, room temp, 2 h, 29% for
two steps.
13. Removal of the FMOC group, followed by introduction of the
left-hand side chain by reaction of amine 14 with acid chloride
1015 afforded the penultimate intermediate 15. Finally, protecting
group removal completed the synthesis of piperidine analog 2.
Table 1
Enzyme inhibition data (Ki, nM) for heterocyclic core analogs 1–8
Compound
Thrombina
Factor VIIaa
Factor IXaa
Factor Xaa
Factor XIIaa
Trypsina
1
2
3
4
5
6
7
8
1.5
7.3
63
34
32
1.1
78
52
>15,000
15,000
15,000
15,000
15,000
15,000
nt
>3000
nt
15,000
nt
15,000
nt
nt
1400
1800
4400
3700
3900
900
>15,000
15,000
15,000
15,000
15,000
15,000
nt
5600
10,600
15,000
15,000
15,000
9400
nt
3100
nt
15,000
15,000
nt
15,000
a
Values are means of two experiments (nt = not tested).

1114
T. A. Blizzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 1111–1115
Figure 1. Suggested binding mode of pyrrolidine 1 against thrombin (2zgb.pdb).
Figure 2. Suggested binding mode of thiomorpholine 6 against thrombin (2zgb.pdb).
Consistent with this binding hypothesis, the observed inhibition
lead pyrrolidine 1 (Table 2). This is consistent with the reduced
dog hepatocyte stability of 6 (76% remaining at t = 90 m) versus
1 (93% remaining). None of the analogs inhibited CYP 3A4, CYP
data correlates reasonably well with the hydrophobicity of the het-
erocyclic core. Figure 3 shows the correlation of thrombin activity
with clogD (compounds are colored by PSA).17 It is clear that the
most potent compounds 1, 2 and 6 are more hydrophobic, whereas
the less active analogs 3, 5, 7 and 8 are more polar. Interestingly,
although compound 4 could be assigned to the less polar group
based on clogD, its activity is only around 30 nM. As noted above,
modeling suggests that this reduced activity is due to an unfavor-
able steric interaction.
2D6, or CYP 2C9 at concentrations up to 50
lM. Due to its poor
pharmacokinetics relative to pyrrolidine 1, thiomorpholine 6 was
not evaluated further.
In conclusion, the novel heterocyclic analogs 2–8 reported here-
in had no significant advantage over 1 as direct thrombin inhibi-
tors. The best analog, thiomorpholine 6, was comparable in
activity to 1 and was slightly more selective but, unfortunately,
had poorer pharmacokinetic properties. Thus, ring expansion of
the lead pyrrolidine did not lead to the improved oral thrombin
inhibitor that were the goal of this research program. However,
The most interesting of the novel analogs, thiomorpholine 6
was further evaluated in
a dog pharmacokinetics study (IV
dosing). Disappointingly, 6 exhibited poorer PK in dogs than the

T. A. Blizzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 1111–1115
1115
Figure 3. Correlation plot between thrombin activity (Ki, nM) and clogD (color by PSA).
Lyle, E. A.; Wallace, A. A.; Leonard, Y. M.; Welsh, D. C.; Lynch, J. J.; McMasters, D.
R. Bioorg. Med. Chem. Lett. 2004, 14, 4161.
Table 2
Dog PK data for pyrrolidine 1 and thiomorpholine 6a
11. 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.
12. 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.
#
t
(h)
AUC (
lM h)
Cl (mL/min/kg)
Vd (L/kg)
½
1
6
6.8
2.7
10.0
1.0
3.5
12.8
1.9
2.3
a
IV dosing (0.3 mpk).
13. 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.
the medical need for improved prophylactic antithrombotic agents
remains. Additional results in this area will be reported in future
publications from this laboratory.
14. Purchased from Aldrich Chemical Company.
15. (R)-1-Chloro-3,3-dimethyl-1-oxobutan-2-yl acetate (10) was freshly prepared
using the following sample procedure: To the neat hydroxy acid12 (86.2 mg,
0.653 mmol) was added acetyl chloride (0.15 mL, 2.11 mmol) and mixture was
warmed with hot gun. Vigorous evolution of bubbles was observed. After
cessation of bubbles, the mixture was heated at 60 °C for 20 min. Completion
of the reaction was monitored by TLC (1:1 ethyl acetate/hexane), excess acetyl
chloride was distilled off and oil was dried under high vacuum for 10 min.
Thionyl chloride (0.15 mL, 2.06 mmol) was added to the residual oil and the
reaction mixture was refluxed for 1.5 h. After completion of the reaction (TLC;
1:1 ethyl acetate/hexane), excess thionyl chloride was distilled off and the
resulting oil was dried under high vacuum. The crude product was used
without purification.
References and notes
1. Di Minno, M. N. D.; Momi, S.; Di Minno, A.; Russolillo, Anna Curr. Drug Targ.
2013, 14, 3.
2. Versteeg, H. H.; Heemskerk, J. W. M.; Levi, M.; Reitsma, P. H. Physiol. Rev. 2013,
93, 327.
3. Arsenault, K. A.; Hirsh, J.; Whitlock, R. P.; Eikelboom, J. W. Nat. Rev. Cardiol.
2012, 9, 402.
4. Weitz, J. I.; Eikelboom, J. W.; Samama, M. M. Chest 2012, 141, e120S.
5. Weitz, J. I. Am. J. Hematol. 2012, 87, S133.
6. Young, R. J. Bioorg. Med. Chem. Lett. 2011, 21, 6228.
7. Eriksson, B. I.; Quinlan, D. J.; Eikelboom, J. W. Ann. Rev. Med. 2011, 62, 41.
8. O’Brien, P. J.; Mureebe, L. J. Cardiovasc. Pharmacol. Ther. 2012, 17, 5.
9. Winstanley, L.; Chen, R. CNS Neurol. Disorders–Drug Targets 2013, 12, 242.
10. Morrissette, M. M.; Stauffer, K. J.; Williams, P. D.; Lyle, T. A.; Vacca, J. P.;
Krueger, J. A.; Lewis, S. D.; Lucas, B. J.; Wong, B. K.; White, R. B.; Miller-Stein, C.;
16. Docking calculations for 1 and 6 were performed against thrombin structure in
complex with
a close analog of pyrrolidine 1 (2zgb.pdb) using Glide
(Schrödinger 2013R1) docking software with standard parameters and XP
precision.
17. Physical properties were computed with ACD/Labs release 11. The correlation
plot was generated using Spotfire.