Discovery of novel heterocyclic factor VIIa inhibitors

Article abstract of DOI:10.1016/j.bmcl.2006.01.017

Structure-activity relationships and binding mode of novel heterocyclic factor VIIa inhibitors will be described. In these inhibitors, a highly basic 5-amidinoindole moiety has been successfully replaced with a less basic 5-aminopyrrolo[3,2-b]pyridine sca

Full text of DOI:10.1016/j.bmcl.2006.01.017

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Bioorganic & Medicinal Chemistry Letters 16 (2006) 2270–2273
Discovery of novel heterocyclic factor VIIa inhibitors
Roopa Rai,* Aleksandr Kolesnikov, Paul A. Sprengeler, Steven Torkelson, Tony Ton,
Bradley A. Katz, Christine Yu, John Hendrix, William D. Shrader, Robin Stephens,
Ronnell Cabuslay, Ellen Sanford and Wendy B. Young
Celera, 180 Kimball Way, South San Francisco, CA 94080, USA
Received 9 December 2005; revised 4 January 2006; accepted 6 January 2006
Available online 3 February 2006
Abstract—Structure–activity relationships and binding mode of novel heterocyclic factor VIIa inhibitors will be described. In these
inhibitors, a highly basic 5-amidinoindole moiety has been successfully replaced with a less basic 5-aminopyrrolo[3,2-b]pyridine
scaffold.
Ó 2006 Elsevier Ltd. All rights reserved.
We have reported the development of direct, small-mol-
ecule inhibitors of the factor VIIa-tissue factor complex
as novel anticoagulants.1–4 These and other published
factor VIIa (fVIIa) inhibitors5 contain an amidino
group (the P1 element) which interacts with Asp-189
(in the S1 pocket) via a salt-bridge. The strongly basic
amidino group, while providing a binding anchor to
the enzyme, is also the main limitation to oral bioavail-
ability. Herein, we describe our efforts to replace the
amidine functionality of our fVIIa inhibitors with a less
basic P1 element. Structure–activity relationships
(SARs) establishing in vitro potency for fVIIa and selec-
tivity against factor Xa (fXa), thrombin, and trypsin are
presented.
interactions between our inhibitor and the S1 pocket,
we designed and generated a series of biaryl analogs that
could potentially interact with the Asp-189 and other
residues within the S1 pocket. From this work, the 5-
aminopyrrolo[3,2-b]pyridine 5 proved to be quite prom-
ising. Compound 5 had an inhibition constant of 0.3 lM
for fVIIa and possessed some selectivity over the related
serine proteases (fXa, thrombin, and trypsin). Surpris-
ingly, when the imidazopyridine analog 4 was generated,
it had significantly less activity toward fVIIa than its
analog 5.
In order to gain insight into the binding contributions of
the 5-aminopyrrolo[3,2-b]pyridine heterocycle in fVIIa,
we obtained a crystal structure of 5 bound to fVIIa.8
This structure was then overlaid with the fVIIa crystal
structure of an amidine analog of 5 reported earlier,4
as depicted in Figure 1. The key H-bonding contacts be-
tween inhibitor 5 and fVIIa are represented schematical-
ly in Figure 2.
The potencies of our previously disclosed amidine-con-
taining fVIIa inhibitors 1 and 21 are shown in Table 1.
As a first step toward generating orally bioavailable
fVIIa inhibitors, we aimed to replace the amidino-aryl
group with a less basic heterocycle, while maintaining
critical interactions in the S1 pocket. Such a strategy
has been previously demonstrated with fXa and throm-
bin inhibitors.6 In order to assess the binding contribu-
tion from the amidine group on scaffold 1, we generated
the des-amidino benzimidazole 3. Not surprisingly,
removal of this group results in a >700-fold loss of
potency. Knowing that we needed significant binding
The interactions of the indole N–H and the phenol with
the catalytic residues (Ser195, His57) are similar to those
observed for our amidine-based fVIIa inhibitors.1 This
similarity extends to the distal aryl ring of the molecule
where the 30-phenylnitro group occupies the fVIIa S10
pocket. Major differences between the binding modes
exist in two areas: the S2 region and the S1 pocket.
The indole ring of Trp215 relocates into what would
have been the S2 pocket. This indole now provides an
edge-to-face contact with the phenolic ring of 5. With
amidine-containing inhibitors, Trp215 is not displaced
Keywords: Factor VIIa; Inhibitor; fVIIa; Heterocyclic; 5-aminopyr-
rolo[3,2-b]pyridine; Factor VIIa crystal structure; Anticoagulant.
*
Corresponding author. Tel.: +1 650 596 9410; e-mail: roopa.rai@
hotmail.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.01.017

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
R. Rai et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2270–2273
Table 1. SAR of amidine-containing compounds compared with non-amidino heterocyclic leads
HO2C HO2C
2271
HO2C
NH
H2N
N
N
X
X
H2N
N
H
N
H
N
H
HO
HO
HO
NO2
NO2
NO2
1: X = N
2: X = C
4: X = N
5: X = C
3
Compound
Ki (lM)7
fVIIa
fXa
Thrombin
Trypsin
1
2
3
4
5
0.015
0.004
11
0.32
0.05
>150
>150
3.5
>150
2.9
11
0.40
>150
>150
120
>150
>150
44
39
0.30
and an S2 pocket is clearly defined. The second differ-
ence was seen in the S1 pocket, where the 5-aminopyr-
rolo[3,2-b]pyridine interacts with the protein in a novel
way. The residues from 215 to 220, which define one side
of the S1 pocket, have significantly rearranged from the
position observed for the amidine-based inhibitors.
Some parts of this loop appear to be somewhat disor-
dered. Whereas an amidine makes strong hydrogen
bonds to Asp189, Ser190, and Gly219, the aminopyri-
dine of 5 makes weak interactions with Gly219 and
Ser190 as shown in Figure 2.
With this binding model in hand, we can now offer a
rationale as to why inhibitor 4 was significantly less
potent than 5. We attribute this loss in potency
to an unfavorable entropic contribution of the imid-
azole tautomers combined with a change in pKa
9
(Scheme 1). The calculated pKa for the pyridine
Figure 1. Crystal structure of 5 bound to fVIIa (protein shown in tan,
H-bonds indicated by dotted lines). This is overlaid with a fVIIa crystal
structure bound to an amidino analog (protein shown in orange,
inhibitor not shown). The rearranged residues 215–220 and Trp215 are
shown as ball and stick representations.
nitrogen of a 5-aminopyrrolo[3,2-b]pyridine is 8.79,
while a 5-amino imidazo[4,5-b]pyridine is predicted
to protonate the second imidazole ring nitrogen at
7.98. This pyridine protonation allows for an impor-
tant interaction of 5 with the rearranged Gly219.
Amidino inhibitors 1 and 2 have similar potencies
since the positive charge is localized at the amidine
in both these compounds.
Our next goal was to improve the potency and selectivity
profile of this novel 5-aminopyrrolo[3,2-b]pyridine
5-Aminopyrolo[3,2-b]pyridine
H
+
H2N
N
H2N
N
pKa 8.79
N
H
N
H
This protonation provides an
important binding interaction
5-Aminoimidazo[4,5-b]pyridine
H
+
H2
N
N
H2N
N
N
N
pKa 7.98
N
H
N
H
Scheme 1. Difference in pKa and site of protonation of 5-aminopyrol-
o[3,2-b]pyridine and 5-aminoimidazo[4,5-b]pyridine P1 groups.
Figure 2. Representation of key interactions between 5 and fVIIa.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
2272
R. Rai et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2270–2273
scaffold. We noted that the 60s loop, which forms the
interface between the S2 and S10 pockets of fVIIa, is sig-
nificantly different from that of fXa and thrombin.4
Building off the 30-aryl group of 5 should extend into
this region and might offer some selectivity gains for
fVIIa. Initial attempts to improve binding by incorpora-
tion of 1- and 2-naphthyls as well as ortho- and para-
substituted benzene moieties at the 30 position resulted
in reduced potency. A series of meta- substituted 30-
phenyl ring analogs were examined, several of which
proved to be quite successful. Selected examples are
shown in Table 2. Initially, we found that incorporation
of small groups at the meta position such as a methyl-
sulfone (as in 7), hydroxyl (as in 8), and a urea (as in
9) did not offer any potency gains over the initial lead
5. Next we chose to generate analog 10, whereby the
urea was extended out by one methylene and a hydro-
phobic aryl group was incorporated on the distal nitro-
gen of the urea. Analog 10 proved to have a 15-fold
improvement in binding to fVIIa (Ki = 0.02 lM) as com-
pared to 5. Crystal structures of inhibitors in our ami-
dino series show that a methyl urea can participate in
a specific H-bond with the carbonyl of His57.4 While
His57 is conserved in all serine proteases, the His57 car-
bonyl in fVIIa appears to have a conformation that is
ideal for this H-bond. In addition, it is the only enzyme
in the coagulation cascade with a Lys60a that can
hydrogen bond to the urea carbonyl and thereby im-
prove this interaction. The terminal phenyl substitution
on the urea may provide added thrombin selectivity due
to negative steric interactions with the 60s loop of
thrombin. Replacing the distal phenyl ring with a polar
pyridine or dimethylamino group, as in 11 and 12,
respectively, results in loss of potency, as compared to
10, perhaps due to a loss in hydrophobic interactions
or an unfavorable entropy.
Scheme 2 describes a synthetic route to compound 10
which can be generalized to apply to other compounds
described in Table 2. Protection of the commercially
available 2-amino-5-nitropyridine (13) with BOC-anhy-
dride followed by reduction of the nitro group under
hydrogenation conditions gave 14. Bromination of 14
using N-bromosuccinimide in DMF followed by subse-
quent treatment with BOC-anhydride affords 15 in high
yield. Bromide 15 was then coupled with 1610 producing
alkyne 17. Alkyne 17 is cyclized with TBAF and subse-
quent removal of protecting groups affords 18. Amine
18 is then treated with 19, the succinimidyl ester of phe-
nylcarbamic acid. The crude reaction mixture is then
purified by reverse-phase chromatography to afford
compound 10.
We have described the synthesis and SAR of novel fVIIa
inhibitors, wherein the amidinoindole group has been
successfully replaced with a less basic 5-aminopyrol-
o[3,2-b]pyridine moiety. A comparison of the fVIIa crys-
tal structures of these two classes of compounds reveals
that although the S1 pocket alters its conformation to
accommodate this change, the S10 region of the protein
remains virtually unchanged. We have therefore been
H
N
N
H2N
N
c, d
48%
a, b
57%
Boc
NH2
NO2
14
13
COOMe
H
N
+
Table 2. SAR of 5-aminopyrrolo[3,2-b]pyridine analogs
N
Br
NHBOC
Boc
NHBOC
16
HO2C
H2N
N
15
OMem
N
H
HO
R
Boc
CO2Me
NH
N
f-g
e
Compound R
Selectivity for fVIIa versus
55 %
75 %
NHBoc
MemO
fVII Ki Xa fIIa
(lM)
Trypsin
NHBoc
17
5
6
7
8
NO2
H
0.30
2.4
7.4
2.5
12 147
83
2.5 13
400
67
6
O
H
SO2Me
OH
16
84
MeO2C
HO
N
O
6
116
Ph
N
O
H2N
N
19
O
9
12
2.8 >13
>13
O
NH2
N
H
10
N
H
O
h-i
80 %
Ph
10
0.02
336 >7000 >7000
18
N
H
N
H
NH2
O
Scheme 2. Synthesis of 10. Reagents and conditions: (a) NaHDMS,
THF, Boc-anhydride, 0 °C to rt; (b) 10% Pd–C, THF, MeOH, H2
(1 atm); (c) NBS, DMF 0 °C to rt; (d) BOC-anhydride, BuOH; (e)
Pd(PPh3)2Cl2, triethylamine, CuI, acetonitrile, 80 °C; (f) TBAF, THF;
(g) HCl (4 N in dioxane), MeOH anhyd, (h) 19, triethylamine, DMF; (i)
NaOH (10% aq), MeOH, 50 °C.
3-Py
N
11
12
0.062 90 >2000 >7000
N
N
H
t
H
O
1.4
11 >100 >100
N
N
H
H

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
R. Rai et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2270–2273
2273
Nicholson, N. S.; Clare, M.; Stegeman, R. A.; Stevens,
A. M.; Stallings, W. C.; Kurumbail, R. G.; South, M. S.
J. Med. Chem. 2003, 46, 4050; (c) Klinger, O.; Matter, H.;
Schudok, M.; Donghi, M.; Czech, J.; Lorenz, M.; Nestler,
H. P.; Szillat, H.; Schreuder, H. Bioorg. Med. Chem. Lett.
2004, 14, 3715.
able to adapt key elements of our amidino series of com-
pounds to this new class of fVIIa inhibitors.
Acknowledgments
6. (a) Quan, M. L.; Lam, P. Y. S.; Han, Q.; Pinto, D. J. P.; He,
M. Y.; Li, R.; Ellis, C. D.; Clark, C. G.; Teleha, C. A.; Sun,
J.-H.; Alexander, R. S.; Bai, S.; Luettgen, J. M.; Knabb, R.
M.; Wong, P. C.; Wexler, R. R. J. Med. Chem. 2005, 48,
1729; (b) Sanderson, P. E. J.; Cutrona, K. J.; Dorsey, B. D.;
Dyer, D. L.; McDonough, C. M.; 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., Jr.; Lyle, E. A.;
Lynch, J. J., Jr.; Stranieri, M. T.; Vastag Shafer, J. A.;
Vacca, J. P. Bioorg. Med. Chem. Lett. 1998, 8, 817.
7. Inhibition assays for factor Xa and thrombin were
performed as described Cregar, L.; Elrod, K. C.; Putnam,
D.; Moore, W. R. Arch. Biochem. Biophys. 1999, 366, 125,
with the pH adjusted to 7.4. The trypsin and fVIIa assays
were performed and analyzed as in the above reference
with the following additional details. Factor VIIa (Enzyme
Research) was incubated at 7 nM and CH3SO2-D-CHA-
But-Arg-pNA (Centerchem) was used as the substrate.
The buffer for the factor VIIa assay was supplemented
with 11 nM relipidated tissue factor and 5 mM CaCl2.
Trypsin (Athens Research Institute) was incubated at
10 nM with variable concentrations of inhibitor in 50 mM
Tris (pH 7.4), 150 mM NaCl, 1.5 mM EDTA, 0.05%
Tween 20, and 10% DMSO. The reaction was initiated
with substrate, Tosyl-Gly-Pro-Lys-pNA (Centerchem),
supplied at the Km (25 lM).
The authors thank Mike Green and James Janc for their
contributions.
References and notes
1. Young, W. B.; Kolesnikov, A.; Rai, R.; Sprengeler, P. A.;
Leahy, E. M.; Shrader, W. D.; Sangalang, J.; Burgess-
Henry, J.; Spencer, J.; Elrod, K.; Cregar, L. Bioorg. Med.
Chem. Lett. 2001, 11, 2253.
2. Shrader, W. D.; Costerison, J.; Hendrix, J.; Hu, H.;
Kolesnikov, A.; Kumar, V.; Leahy, E.; Rai, R.; Shaghafi,
M.; Ton, T.; Torkelson, S.; Wesson, K.; Young, W. B.;
Katz, B. A.; Sprengeler, P. A.; Yu, C.; Cabuslay, R.;
Gjerstad, E.; Janc, J.; Sanford, E. Abstracts of Papers,
229th ACS National Meeting, San Diego, CA, United
States, March 13–17, 2005, MEDI-251.
3. Kolesnikov, A.; Rai, R.; Young, W. B.; Torkelson, S.;
Shrader, W. D.; Leahy, E. M.; Katz, B. A.; Sprengeler, P.
A.; Liu, L.; Mordenti, J.; Gjerstad, E.; Janc, J. Abstracts of
Papers, 229th ACS National Meeting, San Diego, CA,
United States, March 13–17, 2005, MEDI-250.
4. Young, W. B.; Mordenti, J.; Torkelson, S.; Hanson, S. R.;
Marzec, U. M.; Shrader, W. D.; Rai, R.; Kolesnikov, A.;
Liu, L.; Hu, H.; Leahy, E.; Sprengeler, P.; Katz, B.; Janc,
J. W. Bioorg. Med. Chem. Lett. in press.
8. PDB Entry Code 2F9B.
9. Calculations done using ACD Labs Product version 8.07.
10. Alkyne 16 was prepared from the corresponding aldehyde
by its reaction with dimethyl-1-diazo-2-oxopropylphosph-
onate and K2CO3 in methanol. The 3-aryl-substituted
aldehyde was prepared from previously reported [3-
bromo-5-formyl-4-(2-methoxy-ethoxymethoxy)-phenyl]-acetic
acid methyl ester[1] via a Suzuki reaction with BOC-
protected aminomethyl 3-boronic acid.
5. (a) Olivero, A. G.; Eigenbrot, C.; Goldsmith, R.; Robarge,
K.; Artis, D. R.; Flygare, J.; Rawson, T.; Sutherlin, D. P.;
Kadkhodayan, S.; Beresini, M.; Elliott, L. O.; DeGuzman,
G. G.; Banner, D. W.; Ultsch, M.; Marzec, U.; Hanson, S.
R.; Refino, C.; Bunting, S.; Kirchhofer, D. J. Biol. Chem.
2005, 280, 9160; (b) Parlow, J. J.; Case, B. L.; Dice, T. A.;
Fenton, R. L.; Hayes, M. J.; Jones, D. E.; Neumann, W.
L.; Wood, R. S.; Lachance, R. M.; Girard, T. J.;