Discovery of Phenylglycine Lactams as Potent Neutral Factor VIIa Inhibitors

Article abstract of DOI:10.1021/acsmedchemlett.6b00282

Inhibitors of Factor VIIa (FVIIa), a serine protease in the clotting cascade, have shown strong antithrombotic efficacy in preclinical thrombosis models with minimal bleeding liabilities. Discovery of potent, orally active FVIIa inhibitors has been largely unsuccessful because known chemotypes have required a highly basic group in the S1 binding pocket for high affinity. A recently reported fragment screening effort resulted in the discovery of a neutral heterocycle, 7-chloro-3,4-dihydroisoquinolin-1(2H)-one, that binds in the S1 pocket of FVIIa and can be incorporated into a phenylglycine FVIIa inhibitor. Optimization of this P1 binding group led to the first series of neutral, permeable FVIIa inhibitors with low nanomolar potency.

Full text of DOI:10.1021/acsmedchemlett.6b00282

Letter
pubs.acs.org/acsmedchemlett
Discovery of Phenylglycine Lactams as Potent Neutral Factor VIIa
Inhibitors
Nicholas R. Wurtz,*,† Brandon L. Parkhurst,† Wen Jiang,† Indawati DeLucca,† Xiaojun Zhang,†
Vladimir Ladziata,† Daniel L. Cheney,† Jeffrey R. Bozarth,‡ Alan R. Rendina,‡ Anzhi Wei,‡
Joseph M. Luettgen,‡ Yiming Wu,‡ Pancras C. Wong,‡ Dietmar A. Seiffert,‡ Ruth R. Wexler,†
and E. Scott Priestley†
†Bristol-Myers Squibb R&D, 350 Carter Road, Hopewell Township, New Jersey 08540, United States
‡Bristol-Myers Squibb R&D, 311 Pennington Rocky Hill Road, Pennington, New Jersey 08534, United States
S
* Supporting Information
ABSTRACT: Inhibitors of Factor VIIa (FVIIa), a serine protease in the clotting cascade, have shown strong antithrombotic
efficacy in preclinical thrombosis models with minimal bleeding liabilities. Discovery of potent, orally active FVIIa inhibitors has
been largely unsuccessful because known chemotypes have required a highly basic group in the S1 binding pocket for high
affinity. A recently reported fragment screening effort resulted in the discovery of a neutral heterocycle, 7-chloro-3,4-
dihydroisoquinolin-1(2H)-one, that binds in the S1 pocket of FVIIa and can be incorporated into a phenylglycine FVIIa
inhibitor. Optimization of this P1 binding group led to the first series of neutral, permeable FVIIa inhibitors with low nanomolar
potency.
KEYWORDS: TF-FVIIa inhibitor, anticoagulant, fragment-based drug design, neutral P1
Inhibition of the TF-FVIIa complex may also have therapeutic
hromboembolic disorders remain a major cause of
Tmorbidity and mortality in the developed world. While potential in other areas such as cancer or inflammation due
both to its antithrombotic effect and to other cellular signaling
pathways.6 A parenterally dosed FVIIa inhibitor has prevented
tumor growth in mouse models and was in clinical trials for
treatment of pancreatic cancer.7,8
anticoagulants such as the vitamin K antagonist, warfarin, and
heparins have been the standard of care for many years, they
possess significant limitations. Warfarin has highly variable
effects in patients caused by significant drug and food
interactions and therefore requires blood tests to maintain
the desired therapeutic effect without excessive bleeding.1
Heparins require parenteral administration, which prevents
their widespread use in long-term therapy. Several decades of
research devoted to the discovery of new therapies have led to
new oral anticoagulant drugs that target thrombin and FXa,
with improved bleeding profiles.2 Despite overcoming many of
the limitations of warfarin including having a wider therapeutic
index, research to find safe, effective oral anticoagulants to
complement existing therapies continues.
An early target in the clotting cascade that has attracted
significant interest is the tissue factor-Factor VIIa complex (TF-
FVIIa).3,4 The complex initiates the extrinsic pathway in the
clotting cascade by activating Factors X and IX. Studies show
that inhibition of the TF-FVIIa complex by a small molecule in
preclinical models demonstrates strong antithrombotic efficacy
and a wide therapeutic index with respect to bleeding time.5
Despite years of study, the discovery of a potent, orally
available FVIIa inhibitor remains a challenge to the field. FVIIa,
like other trypsin-like serine proteases, has a deep S1 pocket9
with Asp189 positioned at one end. The adjacent 190 position
contains a serine, which makes the FVIIa S1 pocket smaller and
more hydrophilic than other serine proteases such as FXa, and
thrombin, which have alanine at this position.10 The majority of
reported FVIIa inhibitors contain a highly basic moiety such as
an amidine4,11 that forms a bidentate salt-bridge with Asp189
for high affinity binding. Optimizing these structural classes
into orally available inhibitors is difficult without use of
prodrugs because amidines and other highly basic groups exist
as charged species at physiological pH (7.4) and lack
Received: July 22, 2016
Accepted: September 16, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acsmedchemlett.6b00282
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
permeability. The basic P1 groups can also preclude selectivity
over other serine proteases, which have Asp189 and similar S1
pockets. Many of these shortcomings can be overcome with
weakly basic P1 groups such as aminoisoquinoline and
pyrrolopyridine.12−16 Furthermore, examples from the discov-
ery of oral FXa inhibitors demonstrate that there are advantages
of moving toward neutral groups that bind in the P1 pocket.2
There have been a few reports of neutral P1 groups with
modest FVIIa potency. One of the successful neutral P1 groups
is the meta-benzamide.15,18 In addition to forming a hydrogen
bond with Asp189, the primary amide makes two interactions
with the backbone carbonyl of Gly219: A benzamide N−H
hydrogen bonds with the oxygen of the Gly219 carbonyl, while
the carbonyl of the amide forms a dipolar interaction19 with the
permeability and selectivity. Conversely, the corresponding
aminoisoquinoline 2 (which is expected to be protonated under
physiologic conditions) provides a potent inhibitor (FVIIa Ki =
1.8 nM with insufficient permeability and low selectivity for
enzymes such as tissue kallikrein, which promotes adverse
cardiac remodeling when inhibited.20 Both compounds have
good activity in the Factor VII deficient prothrombin time assay
(FVII def PT), which measures the time to prolong blood
coagulation by 2-fold (Figure 1).
In order to search for a P1 that combined potency,
permeability, and selectivity, we carried out a small molecule
fragment screening effort,21 which involved a protocol starting
with virtual screening, followed by NMR and enzymatic
screening, and ending with crystallography. One high priority
Figure 1. Comparison of previously discovered phenyl pyrrolidine
FVIIa inhibitors with basic and neutral P1 groups.5,17
Figure 3. Crystal structures of compounds 3 (orange) and 8 (white) at
2.2 and 2.5 Å, respectively, bound in the active site of FVIIa.21 Crystal
structures of compounds 3 and 8 bound in the active site of FVIIa.
Only key residues in the S1 pocket are shown.
hit identified during our efforts was the dihydroisoquinolin-
1(2H)-one 3, which binds with a 5-fold higher affinity than
benzamide 4 and is comparable in potency to 1-aminoisoquino-
line 5 (Figure 2). The structure led us to consider a similar
binding orientation to the meta-benzamide P1. Initially this
looked unlikely because the aliphatic region of the lactam
overlapped with the same hydrophilic region where benzami-
dines form the charged bidentate hydrogen bond with Asp189.
Herein we report our results for incorporation of this small
molecule screening hit into the phenylglycine scaffold and
subsequent optimization of the resulting FVIIa inhibitors.
In order to explore the various regions of the phenylglycine
chemotype, a modular synthesis was pursued. Limited success
in coupling aryl bromides to phenylglycine amines using
various palladium- and copper-N-arylation protocols led us to
pursue another synthetic route. The Petasis three-component
coupling16,22 with the neutral P1 fragments, dimethoxyphenyl
boronic acid, and glyoxylic acid (Scheme 1) yielded the
phenylglycine scaffold. Neutral P1 fragments were synthesized
or purchased as the corresponding aniline (see Supporting
Information). Pyrrolidine 6 was coupled to the resulting
phenylglycine carboxylic acids and the diastereomers were
separated using preparatory HPLC to give the target
compounds. The stereochemistry of the active diastereomer
was assigned based on crystal structures of related compounds
bound to FVIIa (not shown). While this chemistry sequence
did not allow introduction of the P1 groups in the final
synthetic step, the two step sequence provided a robust route
Figure 2. Fragments that bind in the S1 pocket of FVIIa.
a
Scheme 1. General Synthesis of Phenylglycine Lactams
a
Reagents and conditions: (a) acetonitrile/DMF (4:1), microwave,
100 C, 20−80%; (b) EDCI, 6, DIEA, HOAt, DCM, 20−80%; (c)
HPLC separation.
carbon in the Gly219 carbonyl. In our phenylglycine series,16
we synthesized a library of substituted phenyl P1 groups and
discovered that the only active compound was meta-benzamide
1, which provides a compound with modest potency, but good
B
DOI: 10.1021/acsmedchemlett.6b00282
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Table 1. Activity and Permeability of FVIIa Inhibitors with
Neutral P1 Substituents
Table 2. Activity and Permeability of Substituted
Isoquinolinones
of the side chain of Asp189, placing the aliphatic moiety next to
the polar acid, which is unprecedented to our knowledge.
To determine if an unsaturated ring would bind preferentially
in the narrow pocket, isoquinolin-1(2H)-one 10 was
synthesized, resulting in a 10-fold potency increase in the
binding assay and a 2-fold increase in the clotting assay.
Incorporation of nitrogen in the C-4 position to produce
quinazolin-4(3H)-one 11 maintained potency, but lost
permeability. Moving the nitrogen to the three position in
the phthalazin-1(2H)-one 12 resulted in a 60-fold loss in
potency, likely caused by the electrostatic repulsion with
Asp189. The carbonyl in the C-4 position in 2,3-dihydroph-
thalazine-1,4-dione 13 recaptured some of the binding affinity
but still only resulted in a compound with moderate potency.
Like the isoquinolin-1(2H)-one, the flat isoindolin-1-one 14
produced a potent compound, albeit with little permeability.
The introduction of the hydrogen bond donor in the 1H-
indazol-3(2H)-one 15 resulted in a loss of potency.
FVIIa crystal structures contain a conserved water molecule
above the Tyr228 side chain. Since targeting this binding site
and displacement of a similar water molecule by FXa inhibitors
resulted in significant potency increase, we decided to target
this region by substitution of the isoquinolin-1(2H)-one at the
C-4 position. All C-4 halogenated analogues 16−18 gave
equipotent inhibitors in the binding assay with the expected
trend of increasing permeability with the size of the halides
(Table 2). The clotting assay potency decreased for the Cl and
to access all the desired compounds. The route also allowed
facile SAR exploration via alteration of the boronic acid and
phenylpyrrolidine components in the reaction sequence.
As reported previously,21 the 3,4-dihydroisoquinolin-1(2H)-
one was spliced into the phenylglycine FVIIa inhibitor series,
producing compound 8 with similar potency in the FVIIa
binding assay and increased permeability relative to the
corresponding P1 meta-benzamide 1. Overlaying the crystal
structures of the fragment 3 and phenylglycine 8 shows a close
overlay of the dihydroisoquinolinone atoms in the S1 binding
pocket (Figure 3). Both structures maintain the interactions
with the Gly218 observed with the meta-benzamide. Further
optimization of the P1 group is shown in Table 1. The open
form of the lactam, methyl benzamide 9, is inactive,
demonstrating the importance of tying back the lactam to
maintain key interactions with Gly218. Interestingly, the
aliphatic region of the lactam is in approximate vDW contact
C
DOI: 10.1021/acsmedchemlett.6b00282
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Table 3. Activity and Permeability of Substituted
Isoindolinones
Figure 4. Potency, permeability, and selectivity of compounds 10 and
24 for other serine proteases. Ki values were averaged from two
experiments. TF/FVIIa assays were performed with recombinant
human enzyme, while other enzyme assays were performed with
purified human enzymes. All assays were run at 25 °C except tissue
kallikrein 1, which was run at 37 °C. Detailed descriptions of the
enzyme assay protocols have been reported.5
of the isoquinolin-1(2H)-one may be a useful discovery in the
design of FVIIa P1 groups.
Substitution around isoindolin-1-one 14 was largely not
tolerated. The exception was fluorination of the C-2 position,
which resulted in equipotent compound 24 with significantly
increased permeability (Table 3). However, fluorination of the
C-6 and C-3 position caused a 300-fold and ∼30-fold decrease
in binding affinity in compounds 25 and 26, respectively.
Methyl substitution at the C-3 position to give 27 proved to be
5-fold less potent than fluorine. C-8 methyl substitution on the
lactam ring with either chirality resulted in more permeable
compounds but a 4−6-fold loss in potency. A carbonyl at the
same position in compound 30 resulted in an even larger loss of
potency.
Overall, isoquinolin-1(2H)-one 10 and 2-fluoro isoindoli-
none 24 provided the best balance of in vitro permeability and
potency in the phenylglycine series. A broader look at binding
to other relevant serine proteases for these two compounds
(Figure 4) shows excellent selectivity (>200-fold) over all other
serine proteases tested, which is representative of the selectivity
observed for all compounds in this series with low nanomolar
potency. Both compounds had high clearance when dosed IV in
dog, so they were not advanced into further studies (See
Supporting Information).
A dihydroisoquinolin-1(2H)-one that binds in the S1 pocket
of FVIIa was identified in a fragment screen and spliced into a
phenylglycine FVIIa inhibitor series to produce a neutral,
potent, selective, and permeable FVIIa inhibitor. Subsequent
optimization of the lactam P1 group with more planar
heterocycles that maintained the amide−protein interactions
led to nanomolar inhibitors of FVIIa with an array of neutral P1
binding groups, which represents the first reported series of
potent neutral Factor VIIa inhibitors. This collection of P1
binding groups creates a toolbox of neutral P1 groups that will
allow modulation of molecular properties, such as permeability,
clotting activity, and selectivity while maintaining potency, and
should be broadly useful in the design of orally active FVIIa
inhibitors.
Br, likely due to increased protein binding. Methyl and ethyl
substitution at the C-4 position in 19 and 20 resulted in ∼10×
loss in potency, while cyclopropyl analogue 21 resulted in
∼300-fold loss in potency. The difference in potency between
the halo and alkyl substituted series may be due to differential
desolvation effects and the favorable electrostatic interactions of
the aryl halides with the S1 pocket, a subject of current
investigation. Methylation of the 3,4-dihydroisoquinolin-
1(2H)-ones at a similar position, compounds 22 and 23,
resulted in moderate losses in potency. While the expected
increase in potency for the Tyr228 interaction was not
observed, the increase in permeability arising from halogenation
D
DOI: 10.1021/acsmedchemlett.6b00282
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
(12) Miura, M.; Seki, N.; Koike, T.; Ishihara, T.; Niimi, T.; Hirayama,
F.; Shigenaga, T.; Sakai-Moritani, Y.; Tagawa, A.; Kawasaki, T.;
Sakamoto, S.; Okada, M.; Ohta, M.; Tsukamoto, S. Design, synthesis
and biological activity of selective and orally available TF/FVIIa
complex inhibitors containing non-amidine P1 ligands. Bioorg. Med.
Chem. 2007, 15 (1), 160−73.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acsmedchem-
lett.6b00282.
■
S
(13) Trujillo John, I.; Huang, H.-C.; Neumann William, L.; Mahoney
Matthew, W.; Long, S.; Huang, W.; Garland Danny, J.; Kusturin, C.;
Abbas, Z.; South Michael, S.; Reitz David, B. Design, synthesis, and
biological evaluation of pyrazinones containing novel P1 needles as
inhibitors of TF/VIIa. Bioorg. Med. Chem. Lett. 2007, 17 (16), 4568−
74.
(14) Riggs, J. R.; Hu, H.; Kolesnikov, A.; Leahy, E. M.; Wesson, K. E.;
Shrader, W. D.; Vijaykumar, D.; Wahl, T. A.; Tong, Z.; Sprengeler, P.
A.; Green, M. J.; Yu, C.; Katz, B. A.; Sanford, E.; Nguyen, M.;
Cabuslay, R.; Young, W. B. Novel 5-azaindole factor VIIa inhibitors.
Bioorg. Med. Chem. Lett. 2006, 16 (12), 3197−200.
Syntheses and characterization data for new compounds
(PDF)
AUTHOR INFORMATION
Corresponding Author
*(N.R.W.) Tel: 609-466-5099. E-mail: nicholas.wurtz@bms.
■
com.
Notes
The authors declare no competing financial interest.
(15) Bolton, S. A.; Sutton, J. C.; Anumula, R.; Bisacchi, G. S.;
Jacobson, B.; Slusarchyk, W. A.; Treuner, U. D.; Wu, S. C.; Zhao, G.;
Pi, Z.; Sheriff, S.; Smirk, R. A.; Bisaha, S.; Cheney, D. L.; Wei, A.;
Schumacher, W. A.; Hartl, K. S.; Liu, E.; Zahler, R.; Seiler, S. M.
Discovery of nonbenzamidine factor VIIa inhibitors using a biaryl acid
scaffold. Bioorg. Med. Chem. Lett. 2013, 23 (18), 5239−43.
(16) Zhang, X.; Jiang, W.; Jacutin-Porte, S.; Glunz, P. W.; Zou, Y.;
Cheng, X.; Nirschl, A. H.; Wurtz, N. R.; Luettgen, J. M.; Rendina, A.
R.; Luo, G.; Harper, T. M.; Wei, A.; Anumula, R.; Cheney, D. L.;
Knabb, R. M.; Wong, P. C.; Wexler, R. R.; Priestley, E. S. Design and
Synthesis of Phenylpyrrolidine Phenylglycinamides As Highly Potent
and Selective TF-FVIIa Inhibitors. ACS Med. Chem. Lett. 2014, 5 (2),
188−92.
(17) Kansy, M.; Senner, F.; Gubernator, K. Physicochemical high
throughput screening: parallel artificial membrane permeation assay in
the description of passive absorption processes. J. Med. Chem. 1998,
41, 1007−1010.
(18) Miura, M.; Seki, N.; Koike, T.; Ishihara, T.; Niimi, T.; Hirayama,
F.; Shigenaga, T.; Sakai-Moritani, Y.; Kawasaki, T.; Sakamoto, S.;
Okada, M.; Ohta, M.; Tsukamoto, S. Potent and selective TF/FVIIa
inhibitors containing a neutral P1 ligand. Bioorg. Med. Chem. 2006, 14
(23), 7688−705.
(19) Paulini, R.; Muller, K.; Diederich, F. Orthogonal multipolar
interactions in structural chemistry and biology. Angew. Chem., Int. Ed.
2005, 44 (12), 1788−805.
(20) Pons, S.; Griol-Charhbili, V.; Heymes, C.; Formes, P.; Heudes,
D.; Hagege, A.; Loyer, X.; Meneton, P.; Giudicelli, J.; Samuel, J.;
Alhenc-Gelas, F.; Richer, C. Tissue kallikrein deficiency aggravates
cardiac remodelling and decreases survival after myocardial infarction
in mice. Eur. J. Heart Failure 2008, 10, 343−351.
(21) Cheney, D. L.; Bozarth, J. M.; Metzler, W. J.; Morin, P. E.;
Mueller, L.; Newitt, J. A.; Nirschl, A. H.; Rendina, A. R.; Tamura, J. K.;
Wei, A.; Wen, X.; Wurtz, N. R.; Seiffert, D. A.; Wexler, R. R.; Priestley,
E. S. Discovery of novel P1 groups for coagulation factor VIIa
inhibition using fragment-based screening. J. Med. Chem. 2015, 58 (6),
2799−808.
(22) Petasis, N. A.; Zavialov, I. A. A New and Practical Synthesis of
alpha-Amino Acids from Alkenyl Boronic Acids. J. Am. Chem. Soc.
1997, 119, 445−446.
ACKNOWLEDGMENTS
We would like to thank Atsu Apedo for carrying out separations
of final molecules.
■
ABBREVIATIONS
■
TF-FVIIa, tissue factor−factor VIIa; FVII def PT, factor VII
deficient prothrombin assay; FIXa, factor IXa; FXa, factor Xa;
FXIa, factor XIa; FXIIa, factor XIIa; aPC, active protein C;
tissue kal, human tissue kallikrein 1; TPA, tissue plasminogen
activator; EDCI, 1-ethyl-3-(3-diemethylaminopropyl)-
carbodiimide; HOAt, 1-hydroxy-7-azabenzotriazole
REFERENCES
■
(1) Ansell, J.; Hirsh, J.; Hylek, E.; Jacobson, A.; Crowther, M.;
Palareti, G. Pharmacology and management of the vitamin K
antagonists. Chest 2008, 133, 160S−198S.
(2) Pinto, D. J.; Smallheer, J. M.; Cheney, D. L.; Knabb, R. M.;
Wexler, R. R. Factor Xa inhibitors: next-generation antithrombotic
agents. J. Med. Chem. 2010, 53 (17), 6243−74.
(3) Girard, T. J.; Nicholson, N. S. The role of tissue factor/factor
VIIa in the pathophysiology of acute thrombotic formation. Curr. Opin.
Pharmacol. 2001, 1 (2), 159−163.
(4) Priestley, E. S. Tissue factor-fVIIa inhibition: update on an
unfinished quest for a novel oral antithrombotic. Drug Discovery Today
2014, 19 (9), 1440−4.
(5) Wong, P. C.; Luettgen, J. M.; Rendina, A. R.; Kettner, C. A.; Xin,
B.; Knabb, R. M.; Wexler, R.; Priestley, E. S. BMS-593214, an active
site-directed factor VIIa inhibitor: enzyme kinetics, antithrombotic and
antihaemostatic studies. Thromb. Haemostasis 2010, 104 (2), 261−9.
(6) Rao, L. V.; Pendurthi, U. R. Tissue factor-factor VIIa signaling.
Arterioscler., Thromb., Vasc. Biol. 2005, 25 (1), 47−56.
(7) Prescott, J.; Thiemann, P.; Loury, D. PCI-27483, a small
molecule inhibitor of Factor VIIa, inhibits growth of BxPC3 pancreatic
adenocarcinoma xenograft tumors. Cancer Res. 2008, 68, 5669.
(8) Young, W. B.; Mordenti, J.; Torkelson, S.; Shrader, W. D.;
Kolesnikov, A.; Rai, R.; Liu, L.; Hu, H.; Leahy, E. M.; Green, M. J.;
Sprengeler, P. A.; Katz, B. A.; Yu, C.; Janc, J. W.; Elrod, K. C.; Marzec,
U. M.; Hanson, S. R. Factor VIIa inhibitors: chemical optimization,
preclinical pharmacokinetics, pharmacodynamics, and efficacy in an
arterial baboon thrombosis model. Bioorg. Med. Chem. Lett. 2006, 16
(7), 2037−41.
(9) Schechter, I.; Berger, A. On the size of the active site in proteases.
I. Papain. Biochem. Biophys. Res. Commun. 1967, 27 (2), 157−62.
(10) Banner, D. W.; D’Arcy, A.; Chene, C.; Winkler, F. K.; Guha, A.;
Konigsberg, W. H.; Nemerson, Y.; Kirchhofer, D. The crystal structure
of the complex of blood coagulation factor VIIa with soluble tissue
factor. Nature 1996, 380 (6569), 41−6.
(11) Shirk, R. A.; Vlasuk, G. P. Inhibitors of Factor VIIa/tissue factor.
Arterioscler., Thromb., Vasc. Biol. 2007, 27 (9), 1895−900.
E
DOI: 10.1021/acsmedchemlett.6b00282
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX