Potent, Orally Bioavailable, and Efficacious Macrocyclic Inhibitors of Factor XIa. Discovery of Pyridine-Based Macrocycles Possessing Phenylazole Carboxamide P1 Groups

Article abstract of DOI:10.1021/acs.jmedchem.9b01768

Factor XIa (FXIa) inhibitors are promising novel anticoagulants, which show excellent efficacy in preclinical thrombosis models with minimal effects on hemostasis. The discovery of potent and selective FXIa inhibitors which are also orally bioavailable has been a challenge. Here, we describe optimization of the imidazole-based macrocyclic series and our initial progress toward meeting this challenge. A two-pronged strategy, which focused on replacement of the imidazole scaffold and the design of new P1 groups, led to the discovery of potent, orally bioavailable pyridine-based macrocyclic FXIa inhibitors. Moreover, pyridine-based macrocycle 19, possessing the phenylimidazole carboxamide P1, exhibited excellent selectivity against relevant blood coagulation enzymes and displayed antithrombotic efficacy in a rabbit thrombosis model.

Full text of DOI:10.1021/acs.jmedchem.9b01768

Subscriber access provided by BIU Pharmacie | Faculté de Pharmacie, Université Paris V
Article
Potent, Orally Bioavailable and Efficacious Macrocyclic
Inhibitors of Factor XIa. Discovery of Pyridine-Based
Macrocycles Possessing Phenylazole Carboxamide P1 Groups
James R. Corte, Donald J. P. Pinto, Tianan Fang, Honey Osuna, Wu Yang, Yufeng Wang, Amy Lai,
Charles Clark, Jung-Hui Sun, Richard A. Rampulla, Arvind Mathur, Mahammed Kaspady, Premsai
Rai Neithnadka, Yi-Xin Li, Karen A Rossi, Joseph E Myers, Steven Sheriff, Zhen Lou, Timothy W.
Harper, Christine S. Huang, Joanna J Zheng, Jeffrey M. Bozarth, Yiming Wu, Pancras C Wong, Earl
Crain, Dietmar A Seiffert, Joseph M Luettgen, Patrick Lam, Ruth R. Wexler, and William R Ewing
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01768 • Publication Date (Web): 13 Dec 2019
Downloaded from pubs.acs.org on December 13, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Potent, Orally Bioavailable and Efficacious
Macrocyclic Inhibitors of Factor XIa. Discovery of
Pyridine-Based Macrocycles Possessing Phenylazole
Carboxamide P1 Groups
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a,
a,
a
a
a
a
James R. Corte , Donald J. P. Pinto , Tianan Fang , Honey Osuna , Wu Yang , Yufeng Wang ,
a
a
b
b
b
Amy Lai , Charles G. Clark , Jung-Hui Sun , Richard Rampulla , Arvind Mathur , Mahammed
c
c
a
a
b
Kaspady , Premsai Rai Neithnadka , Yi-Xin Cindy Li , Karen A. Rossi , Joseph E. Myers Jr. ,
b
d
d
d
d
Steven Sheriff , Zhen Lou , Timothy W. Harper , Christine Huang , Joanna J. Zheng , Jeffrey
d
d
d
d
d
M. Bozarth , Yiming Wu , Pancras C. Wong , Earl J. Crain , Dietmar A. Seiffert , Joseph M.
Luettgen , Patrick Y. S. Lam , Ruth R. Wexler and William R. Ewing
d
a
a
a
aBristol-Myers Squibb Company, Research and Development, 350 Carter Road, Hopewell, New
Jersey 08540, United States
b
Bristol-Myers Squibb Company, Research and Development, US Rt. 206 & Province Line
Road, Princeton, New Jersey 08540, United States
c
Bristol-Myers Squibb Research Center, Syngene International Pvt. Ltd., Biocon Park, Plot No. 2
&
3, Bommasandra–Jigani Road, Bangalore 560 100, India
d
Bristol-Myers Squibb Company, Research and Development, 311 Pennington-Rocky Hill Road,
Pennington, New Jersey 08543, United States
KEYWORDS. Factor XIa; FXIa; Factor XIa Inhibitors; FXIa Inhibitors; Thrombosis; Activated
Partial Thromboplastin Time; aPTT; Macrocycle; Anticoagulant; Antithrombotic.
ABSTRACT
Factor XIa inhibitors are promising novel anticoagulants which show excellent efficacy in
preclinical thrombosis models with minimal effects on hemostasis. The discovery of potent and
selective FXIa inhibitors which are also orally bioavailable has been a challenge. Here, we
describe optimization of the imidazole-based macrocyclic series and our initial progress towards
meeting this challenge. A two-pronged strategy, which focused on replacement of the imidazole
scaffold and the design of new P1 groups, led to the discovery of potent, orally bioavailable
pyridine-based macrocyclic FXIa inhibitors. Moreover, pyridine-based macrocycle 19,
possessing the phenylimidazole carboxamide P1, exhibited excellent selectivity against relevant
blood coagulation enzymes and displayed antithrombotic efficacy in a rabbit thrombosis model.

Corresponding author. Tel.: +1-609-466-5030; fax: +1-609-818-3331; e-mail: james.corte@bms.com

Corresponding author. Tel.: +1-609-466-5068; fax: +1-609-818-3331; e-mail: donald.pinto@bms.com
ACS Paragon Plus Environment
1

Page 3 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Morbidity and mortality as a result of thrombosis continues to be an area of unmet
1
medical need. Significant advances have been made by the introduction of novel oral
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
anticoagulants that target key serine proteases in the blood coagulation cascade such as thrombin
(
e.g. dabigatran) or factor Xa (FXa, apixaban, rivaroxaban, edoxaban and betrixaban).2-3 The risk
4
of bleeding remains a concern for all anticoagulants. Factor XIa (FXIa), another critical serine
protease in the blood coagulation cascade, has emerged as an attractive target for the prevention
and treatment of thrombosis with the potential of reducing the bleeding risk, thus leading to an
even safer anticoagulant.5
Factor XIa (FXIa), the activated form of the zymogen factor XI (FXI), amplifies the
generation of thrombin, the last enzyme in the blood coagulation cascade that leads to fibrin clot
formation. Humans with a genetic deficiency in FXI (hemophilia C) display a minor bleeding
tendency even though they show a prolongation in clotting time in the activated partial
6
thromboplastin time (aPTT) clotting assay. Epidemiological studies revealed that individuals
with a severe FXI deficiency have a lower risk of ischemic stroke and deep vein thrombosis
(
DVT).7-9 On the contrary, elevated levels of FXI are a risk factor for DVT and myocardial
infarction.10-11 Based on this evidence, it is hypothesized that inhibition of either FXI or FXIa
can reduce thrombosis and preserve hemostasis thus reducing the risk of bleeding that is seen
with other anticoagulants.12
Multiple studies have shown that therapeutic inhibition of either FXI [antisense
oligonucleotides (ASO) 13-14 or neutralizing antibodies15-17] or FXIa (irreversible18-20 and
reversible21-25 small molecule inhibitors) provides antithrombotic efficacy with a low risk of
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
bleeding in a variety of preclinical animial thrombosis models. During the course of our
preclinical FXIa program, a FXI-ASO entered the clinic and provided a proof-of-concept for
targeting this key coagulation enzyme. Specifically, in a Phase II clinical trial in total knee
arthroplasty, ASO IONIS-FXIRx was shown to be superior to enoxaparin in reducing the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
6
incidence of DVT. In addition, the rate of clinically revelant bleeding was similar between the
2
7
IONIS-FXIRx and enoxaparin treatment arms. A number of monoclonal antibodies (BAY
2
8
29
30
1
213790, Xisomab 3G3, and MAA-868 ), as well as, small molecule FXIa inhibitors (BMS-
3
1
32
9
62212 and EP-7041 ) have also entered clinical trials. However, all of these clinical agents
require a parenteral route of administration. Orally administered anticoagulants are highly
desirable especially for treating chronic thromboembolic diseases. Despite significant effort, the
discovery of potent, selective, and orally bioavailable FXIa inhibitors has been a challenge until
recently.5
We have previously reported on the discovery of a novel series of potent imidazole-based
33
macrocyclic FXIa inhibitors represented by 12-membered macrocycle 1 (Figure 1). Despite the
excellent FXIa affinity (FXIa Ki = 1.0 nM) and potent in vitro anticoagulant activity as measured
3
4
by the aPTT clotting assay (EC = 0.30 M), macrocycle 1 was not orally bioavailable (F =
1
.5x
rat
0
%). In replacing the chlorophenyltetrazole cinnamide P1 with either a benzamide P1 or cyclic
3
5
carbamate P1, we were able to reduce the polar surface area (PSA) and improve the oral
3
6-37
bioavailability for 13-membered macrocycles 2 (F = 41%) and 3 (F = 59%).
Despite the
rat
rat
good oral bioavailability, these P1 modifications led to an unacceptable loss in both FXIa affinity
and aPTT potency, as well as, poor metabolic stability in human liver microsomes (T = 17
1
/2
min) and poor permeability in the Caco-2 assay (Caco-2 A to B < 15 nm/sec with a high efflux
ACS Paragon Plus Environment
3

Page 5 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
38
ratio) for compound 2. Discovering potent, orally bioavailable FXIa inhibitors which possess
good ADME properties remained as our goal.
P1 prime
O
N N
NH
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
N
N
HN
F
O
F
NHCO Me
2
N
O
13
NHCO Me
12
2
N
N
H
N
H
HN
HN
Me
Cl
Cl
P2 prime
Scaffold
1
2
P1
FXIa Ki = 16 nM
HLM T1/2 = 17 min
aPTT EC1.5x = 6.6 M Caco-2 AB/BA <15/200 nm/sec
FXIa Ki = 1.0 nM
aPTT EC1.5x = 0.30 M Caco-2 AB/BA <15/67 nm/sec
PSA = 169 Å2
%Frat = 0
HLM T1/2 = 90 min
PSA = 125 Å2
%Frat = 41
NH
O
N
NHCO Me
13
2
O
N
HN
O
Cl
FXIa Ki = 17 nM
aPTT EC1.5x = 5.9 M Caco-2 AB/BA = NTa
PSA = 126 Å2
%Frat = 59
3
HLM T1/2 = NTa
a
Figure 1. Imidazole-based 12- and 13-membered macrocyclic FXIa inhibitors discovered at Bristol-Myers Squibb. NT = not tested.
Herein, we describe our initial progress toward achieving this goal by following a two-
pronged optimization strategy that focused on 1) replacement of the imidazole scaffold and 2)
design of new P1 groups. These efforts culminated in the discovery of pyridine-based
macrocycles possessing novel phenylazole carboxamide P1 groups.39-40 Specifically, pyridine-
based macrocycle 19 is a potent, orally bioavailable FXIa inhibitor that exhibited good in vitro
ADME properties, excellent selectivity against relevant blood coagulation enzymes, and was
efficacious in a rabbit model of arterial thrombosis.
RESULTS AND DISCUSSION
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
We had previously disclosed the evolution of our FXIa inhibitors from the acyclic phenyl
imidazole series to the novel imidazole-based macrocyclic series.33 Earlier, we had shown that
oral bioavailability could be achieved in the acyclic phenyl imidazole series by replacing the
imidazole scaffold with various 6-membered ring heterocycles.41 Based on these findings, we
reasoned that replacing the imidazole scaffold in the macrocyclic series with a pyridine ring also
had the potential to lead to potent FXIa inhibitors with improved oral bioavailability. To test this
proposal, several key pyridine-based 12-membered macrocycles were explored (Table 1). The
corresponding imidazole-based macrocycles (1, 4, and 5) which were disclosed previously are
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
included for comparison.33, 36
Table 1. 12-Membered Imidazole- and Pyridine-Based Macrocycles
R
Me
O
O
N N
N
HN
NHCO2Me
HN
N
O
F
O
F
NHCO Me
2
N
H
N
H
X
N
Me
Cl
9
b
c
Entry
X
R
FXIa Kia
nM)
aPTT EC1.5x
HLM T1/2
(min)
PSA
(Å )
Caco-2
AB/BA
2
(
(M)
(nm/sec)
H
N
1
4
5
6





H
Me
Et
H
1.0
0.33
0.39
0.52
0.60
90
169
169
169
153
<15/67d
N
H
N
NTe
NTe
0.39
0.18
1.2
N
H
N
<15/70d
<15/76d

32
40
N


N




0
0
.18
.20
0.62
0.47
7
8
9
Me
Et
--
39
40
24
153
153
109
38/66
<15/84
<15/331
N
N
--
45
27
a
Ki values were obtained from purified human enzyme at 37 C and were averaged from multiple determinations (n ≥ 2). baPTT (activated
o
partial thromboplastin time) in vitro clotting assay was performed in human plasma. EC1.5x refers to the plasma concentration, not the final
c
d
concentration after the addition of aPTT reagent and calcium. HLM = human liver microsome stability. Poor recovery in the Caco-2
permeability assay. NT = Not tested.
e
ACS Paragon Plus Environment
5

Page 7 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Pyridine-based macrocycle 6 displayed excellent FXIa binding affinity (FXIa Ki = 1.2
nM) and potent anticoagulant activity (aPTT EC1.5x = 0.60 M) which compared favorably to
imidazole-based macrocycle 1. The change in scaffold led to a modest reduction in PSA (PSA =
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
1
53 Å for 6 vs PSA = 169 Å for 1), but this did not improve permeability in the Caco-2 assay.
As we reported previously, alkyl groups on the macrocyclic linker, represented in imidazole-
based macrocycles 4 and 5, increased FXIa binding affinity by interacting with a lipophilic
region in the S1 prime (S1') pocket.36 Therefore, incorporation of a methyl moiety in the
macrocyclic linker of the pyridine-based macrocycle led to 7 which resulted in a 6-fold
improvement in FXIa affinity (FXIa Ki = 0.18 nM) and maintained the good anticoagulant
activity in the aPTT clotting assay (EC1.5x = 0.62 M). Larger alkyl groups on the macrocyclic
linker in the pyridine-based macrocycles did not lead to additional improvements in FXIa
affinity, as the ethyl substituted linker 8 was equipotent with the methyl substituted linker 7.
Pyridine-based macrocycles 7 and 8 and imidazole-based macrocycle 5 were equipotent and
showed similar metabolic stability in human liver microsomes. Encouraged by the measurable
but low permeability in the Caco-2 assay, macrocycle 7 was evaluated in a rat pharmacokinetic
(PK) study (Table 5). Macrocycle 7 exhibited a low clearance, moderate volume of distribution,
and a long half-life but unfortunately oral exposure in rat was negligible. As the result, replacing
the imidazole scaffold with a pyridine ring, while maintaining excellent FXIa binding affinity
and potent in vitro anticoagulant activity in the aPTT clotting assay, did not improve the oral
bioavailability for the macrocyclic series. Next, the chlorophenyltetrazole cinnamide P1 was
replaced with the benzamide P1 which afforded 9. As with the imidazole-based macrocycles,
incorporation of the benzamide P1 in 9 led to a significant loss in FXIa affinity (250-fold loss),
aPTT potency (43-fold loss), and metabolic stability with no improvement in permeability
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
Caco-2 A to B < 15 nm/sec). At this juncture, neither the chlorophenyltetrazole cinnamide P1
nor the benzamide P1 groups were acceptable for advancement of the macrocyclic series and our
focus turned to the design and discovery of new P1 groups. A new P1 group that could not only
provide potent, orally bioavailable macrocycles but also maintain good ADME properties.
DESIGN AND DISCOVERY OF NEW P1 GROUPS
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The design of new P1 groups was guided by a detailed analysis of the key interactions of
the chlorophenyltetrazole cinnamide P1 in the S1 pocket of FXIa (Figure 2). The chlorophenyl
portion fills the S1 pocket with the chlorine atom forming a -Cl interaction with Tyr228. The
tetrazole ring extends out of the S1 pocket and interacts with the Cys191-Cys219 disulfide
42
43
bridge. In addition, the tetrazole N4 nitrogen forms a H-bond, via ethylene diol, to Gly216.
The cinnamide portion is in an s-trans orientation with the alkene twisted out of plane (60º) with
the phenyl ring. The NH of the amide forms a H-bond to Ser214 via a water molecule, and the
amide carbonyl interacts with the oxyanion hole and forms a H-bond with the NH of Gly193 and
another H-bond with the NH of Ser195.
Figure 2. X-ray crystal structure of imidazole-based macrocycle 1 in Factor XIa, focusing on the key interactions of the chlorophenyltetrazole
cinnamide P1 in the S1 pocket. The red spheres depict water molecules, and the dotted lines depict hydrogen bonds with distances in Angstroms.
Ethylene diol is an artifact of the flash-cooling procedure.
ACS Paragon Plus Environment
7

Page 9 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Our strategy for modifying the chlorophenyltetrazole cinnamide and designing new P1
groups, represented by structure A, is described in Figure 3. In order to reduce PSA, the
tetrazole ring was removed. Both the chlorophenyl and carboxamide groups were retained in
order to preserve the key interactions with Tyr228 and the oxyanion hole, respectively. Finally,
new linkers (L), which could fill the S1 pocket and have the potential to engage additional
interactions in the FXIa active site, were explored as replacements for the alkene of the
cinnamide.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0

N
N
HN

N

HN
3) Explore new linkers (L) as
replacements for the alkene
1) Remove tetrazole ring
N
HN
L
O
O
O
2
) Retain both chlorophenyl
and carboxamide groups
Cl
Cl
Cl
A
Figure 3. Strategy for designing new P1 groups.
Towards this end, a small set of saturated and unsaturated 5-membered ring heterocyclic
4
4
linkers were surveyed (Table 2; Note the FXIa Ki values were determined at 25 ℃). Poor FXIa
binding affinities were seen for pyrrolidinones 10 and 11, as well as, pyrrolidine 12. By
comparison, pyrazole linked P1 13 was a double-digit nanomolar FXIa inhibitor (FXIa Ki = 47
nM @ 25 ℃). The FXIa Ki was also determined at 37 ℃ (FXIa Ki = 166 nM) for comparison
with subsequent analogs.45 The FXIa binding affinity of 13 was encouraging and it indicated
that an unsaturated, five-membered ring heterocycle could serve as a suitable replacement for the
alkene linker of the cinnamide.
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2. Survey of 5-Membered Ring P1 Linkers in Pyridine-Based Macrocycles
Me
O
HN
NHCO Me
2
O
L
N
H
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
Cl
Entry
L
FXIa Kia
nM)
(
O


10
N
N
N
>13,000

(diastereomer A)
O
11
1
2
,390
,890

(diastereomer B)

12
(diastereomeric mixture)


N
4
166)
7
1
3
N
b
(

a
Ki values were obtained from purified human enzyme at 25 C and were averaged from multiple determinations (n ≥ 2). bKi value was obtained
o
o
at 37 C and were averaged from multiple determinations (n ≥ 2).
To identify areas on the phenylpyrazole carboxamide P1 that could impact FXIa affinity,
additional substitutions were explored (Table 3). As previously shown, the addition of a fluorine
adjacent to the chlorine on the phenyl ring had provided a 2- to 8-fold improvement in FXIa
affinity.25, 37, 46 In this case, incorporation of a fluorine group adjacent to the chlorine in the
phenylpyrazole carboxamide P1 resulted in a dramatic (30-fold) increase in FXIa affinity. For
example, compound 14 was a single-digit nanomolar FXIa inhibitor (FXIa Ki = 5.5 nM) and it
also displayed good metabolic stability (HLM T = 59 min) and good permeability in the Caco-
1
/2
2
assay [Caco-2 AB/BA = 89/561 nm/sec; efflux ratio = 6.3]. The replacement of the fluorine
with a chlorine as in 15, led to a loss in FXIa affinity compared to 14. Substitution on the
pyrazole ring was explored next. The introduction of an amino group at the C5-position of the
pyrazole (16) resulted in a 15-fold improvement in FXIa affinity compared to 13. Combining the
ACS Paragon Plus Environment
9

Page 11 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ortho fluoro and the amino pyrazole groups in 17 led to a similar FXIa binding affinity (FXIa Ki
=
3.4 nM) with a significant 6-fold increase in aPTT potency (aPTT EC1.5x = 1.2 M) observed.
Importantly, 17 showed both good Caco-2 permeability, with a low efflux ratio of 4, and HLM
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
stability.
Table 3. Phenylpyrazole Carboxamide P1 Groups in Pyridine-Based Macrocycles
Me
O
HN
NHCO Me
2
O
N
H
N
N
N
R2
R1
Cl
b
c
Entry
R1
R2
FXIa Kia
aPTT EC1.5x
HLM T1/2
PSA
(Å )
Caco-2 AB/BA
2
(nM)
(M)
(min)
(nm/sec)
d
[ER]
1
3
H
F
H
H
166
5.5
38
NTe
7.9
NTe
127
127
127
153
153
NTe
14
59
89/561 [6.3]
NTe
NTe
1
1
5
6
Cl
H
F
H
>40
4.8
f
NH
2
2
11
97
50
<15/134 [>9]
17
NH
3.4
1.2
40/161 [4.0]
a
Ki values were obtained from purified human enzyme at 37 C and were averaged from multiple determinations (n ≥ 2). baPTT (activated
o
partial thromboplastin time) in vitro clotting assay was performed in human plasma. EC1.5x refers to the plasma concentration, not the final
c
d
e
concentration after the addition of aPTT reagent and calcium. HLM = human liver microsome stability. ER= efflux ratio. NT = Not tested.
Poor recovery in the Caco-2 permeability assay.
f
An X-ray crystal structure of 17 bound to FXIa (2.23 Å resolution, R-work was 0.185
and R-free was 0.226 , Figure 4A) was obtained and the ligand occupies the S1, S1', and S2
prime (S2') binding pockets. The macrocyclic linker fills the S1' region with the methyl
substituent interacting with a lipophilic pocket. The amide NH in the macrocyclic linker forms a
H-bond with the carbonyl of Leu41 (2.9 Å). The methyl N-phenyl carbamate occupies the S2'
ACS Paragon Plus Environment
1
0

Journal of Medicinal Chemistry
Page 12 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
pocket. The NH of the carbamate forms a H-bond with the carbonyl of His40 (2.9 Å) and the
carbonyl of the carbamate forms a H-bond through a water molecule to Ile151 (2.6 Å, 3.0 Å
respectively). The pyridine nitrogen in the scaffold forms a H-bond via a water molecule to the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH of Lys192 (3.0 Å, 3.0 Å respectively). Figure 4B highlights the key interactions of the
2
phenylpyrazole carboxamide P1 in the S1 pocket. The chlorophenyl maintains the -Cl
interaction with Tyr228. The ortho fluoro, which led to a significant increase in FXIa affinity in
comparing 13 and 14, makes van der Waals contact with the side chain of Thr213. The pyrazole
ring is twisted 80º out of plane with the phenyl ring. This large dihedral angle is due to both the
steric clash between the ortho fluoro on the phenyl and the C5-NH group on the pyrazole, as
2
well as, electrostatic repulsion between the ortho fluoro and the lone pair of the pyrazole N2. The
pyrazole N2 forms a long H-bond to Gly216 (3.5 Å), and the C5-NH on the pyrazole forms a H-
2
bond to the carbonyl of Cys191 (2.8 Å). The NH of the carboxamide forms a H-bond to Ser214
via ethylene diol (3.3 Å, 2.7 Å respectively) and the amide carbonyl forms a H-bond with only
the NH of Gly193 (3.0 Å) in the oxyanion hole. This is in contrast to the chlorophenyltetrazole
cinnamide P1 (vide supra), where the carbonyl of the amide interacts with both Gly193 and
Ser195. An overlay of the phenylpyrazole carboxamide P1 in 17 and the chlorophenyltetrazole
P1 in 1 is provided in Figure 4C and highlights a similar trajectory for the different P1 groups
(see Supporting Information for an overlay of the entire molecules of both 17 and 1).
ACS Paragon Plus Environment
1
1

Page 13 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(A)
(B)
(C)
Figure 4. (A) X-ray crystal structure of 17 bound to Factor XIa with omit mFo-DFc electron density contoured at 4.0 r.m.s.d (gray). The red
spheres depict water molecules, and the dotted lines depict hydrogen bonds. Ethylene diol is an artifact of the flash-cooling procedure. (B) Key
interactions of the phenylpyrazole carboxamide P1 in the S1 pocket with the distances in Angstroms. (C) Overlay of phenylpyrazole carboxamide
P1 (yellow) in 17 with chlorophenyltetrazole cinnamide P1 (cyan) in 1. PDB deposition number for 17 is 5QTT.
ACS Paragon Plus Environment
1
2

Journal of Medicinal Chemistry
Page 14 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Based on the X-ray crystal structure of 17, the C5-NH group on the pyrazole ring is not
2
only forming a H-bond with the carbonyl of Cys191, but it is also contributing to the large
dihedral angle due to the steric clash with the ortho fluoro on the phenyl ring. In order to
determine which is the major contributor to FXIa binding affinity, the amino was replaced with a
methyl group. The methyl pyrazole 18 and amino pyrazole 17 were equipotent indicating that the
H-bond with Cys191 had little influence on FXIa binding affinity. Importantly, replacing the C5-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
NH with the methyl group led to a reduction in PSA (PSA = 153 Å for 17 vs PSA = 127 Å for
2
1
8). Next we explored N-linked imidazole and triazole rings as replacements for the pyrazole
linker. We found that phenylimidazole carboxamide 19 showed a 2-fold loss in FXIa affinity
(FXIa Ki = 4.1 nM) compared to 18 but was equipotent in the in vitro clotting assay. This result
indicated that the H-bond between the N2 of the pyrazole in 17 (or 18) and Gly216 (Figure 4B)
was not a critical interaction in the S1 pocket as 19 was devoid of this interaction. An X-ray
crystal structure of 19 bound to FXIa showed the phenylimidazole carboxamide P1 and the
phenylpyrazole carboxamide 17 bind in a very similar orientation (see Supporting Information
for more details). Compound 19 exhibited poor permeability in the Caco-2 assay with a high
efflux ratio. Phenyltriazole carboxamide 20 resulted in a 3-fold loss in both FXIa affinity and
aPTT clotting activity compared to 18. C-linked phenylazoles, such as isoxazoles 21 and 22,
were also explored but were less potent than the N-linked phenylazole analogs 18-20. Having
discovered these new phenylazole carboxamide P1 groups, we went back and examined them in
the imidazole-based macrocycle and 23 is provided as a representative example. In general,
imidazole-based macrocycles were comparable to the corresponding pyridine-based analogs in
terms of FXIa affinity, aPTT potency and Caco-2 permeability but suffered from poor metabolic
stability in cynomolgus liver microsomes (CLM T1/2 ≤ 17 min).
ACS Paragon Plus Environment
1
3

Page 15 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 4. Phenylazole Carboxamide P1 Groups in Pyridine- and Imidazole-Based Macrocycles
Me
Me
O
O
HN
O
HN
NHCO Me
NHCO Me
2
2
O
N
N
H
N
N
L
F
N
H
HN
N
Me
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
F
23
Cl
Cl
Entry
L
FXIa Kia
aPTT EC1.5x
b
LM T1/2
H, R, D, C
(min)
c
PSA
(Å )
Caco-2 AB/BA
2
(nM)
(M)
(nm/sec)
[ER]
d

N
1
1
2
2
8
9
0
1
N
2.2
4.1
7.0
34
1.4
2.0
81, >120, >120, 54
55, 68, 73, 48
>120, >120, 118, 75
NTe
127
127
140
135
135
143
22/348 [16]
<15/367 [>24]
37/338 [9]
NTe

Me
N

Me

N

N
O
N
N
4.4
N

Me

Me

N
NTe
NTe

O

NTe
NTe
22
231
2.1
Me
--
2
3
1.6
66, 26, 56, 17
37/441 [12]
a
Ki values were obtained from purified human enzyme at 37 C and were averaged from multiple determinations (n ≥ 2). baPTT (activated
o
partial thromboplastin time) in vitro clotting assay was performed in human plasma. EC1.5x refers to the plasma concentration, not the final
c
d
concentration after the addition of aPTT reagent and calcium. LM = liver microsome stability; H = human; R = rat; D = dog; C = cyno. ER=
e
efflux ratio. NT = Not tested.
Pharmacokinetic Profiles.
The PK profile for several macrocyclic compounds were evaluated in rat, using a discrete
dosing protocol (Table 5). As mentioned earlier, 7, a pyridine-based macrocycle containing the
chlorophenyltetrazole P1, exhibited a low clearance, a long half-life but no exposure in rat. By
contrast, replacing the chlorophenyltetrazole P1 with the phenylazole carboxamide P1 groups in
the pyridine-based macrocycles provided moderate clearance compounds, with half-lives ranging
from 1- to 2 hours, and low to modest oral bioavailability ranging from 2-42%. Interestingly,
compounds 14 and 17, which possessed good permeability with low efflux ratios, displayed poor
oral exposures. On the contrary, compounds 18-20, which possessed higher Caco-2 efflux ratios,
ACS Paragon Plus Environment
1
4

Journal of Medicinal Chemistry
Page 16 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
exhibited improved oral exposures, albeit variable upon retest (see Supporting Information for
details on specific runs).47
Table 5. Pharmacokinetic Profile In Rat For Selected Macrocycles.a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
AUCe
(nM*h)
g
Entry
Cl
T
(h)
1/2
Vdss
(L/kg)
F
(%)
Caco-2 AB/BA
RLM T1/2
Rat Protein
Binding
(% bound)
(mL/min/kg)
(nm/sec)
(min)
f
[ER]
1
10
4.6
1.0
11.5
0.3
3.4
0
0
24
0
<15/67 [>4.5]
38/66 [1.7]
76
54
98.4
NTh
7b
1
1
1
1
4
7
8
9
25
1.8
2.9
2
24
89/561 [6.3]
40/161 [4.0]
22/348 [16]
<15/367 [>24]
37/338 [9]
114
36
NTh
NTh
45
1.0
2.8
4
25
13-28d
6-29d
9-21d
1.2-1.8
1.0-1.4
2.2-4.2
1.3-1.9
0.5-2.0
2.1-3.5
6-18
7-25
13-42
69-475
68-1231
213-1150
>120
68
96.7
97.4
94.8
2
0c
>120
a
Compounds were dosed as TFA salts in rat in a discrete format. Dose: 0.70 mpk for iv arm and 1.40 mpk for po arm. Vehicle for both iv and po:
b
7
4
0% polypropylene glycol; 20% water; 10% ethanol. Dose: 0.50 mpk for iv arm and 1.0 mpk for po arm. Vehicle for both iv and po: 70% PEG-
c
d
00; 20% water; 10% ethanol. Dose: 0.55 mpk for iv arm and 1.10 mpk for po arm. The compound was evaluated in three separate PK studies
e
f
g
and the ranges are given here, please see Supporting Information for data on the individual runs. Oral AUC. ER= efflux ratio. RLM = rat liver
microsome stability. NT = Not tested. Note: Compound 1 was evaluated in an n-in-1 study. Dose: 1.13 mpk for iv arm and 2.27 mpk for po
h
arm. Vehicle for both iv and po: 70% PEG-400; 20% water; 10% ethanol.
Pyridine-based macrocycles 18-20 were also evaluated in dog and cynomolgus PK
studies (Table 6). The phenylpyrazole carboxamide 18 exhibited a moderate clearance and
modest oral bioavailability in dog (%F = 35). Compound 18 was also a moderate clearance
compound in cyno but unfortunately showed poor oral bioavailability (%F = 4). The
phenylimidazole carboxamide 19 was only dosed iv in dog, but showed a similar PK profile to
1
8 and was a moderate clearance compound with a long half-life (T = 5 h) and a high volume
1/2
distribution. In contrast to 18, compound 19 displayed low clearance and modest oral
bioavailability in cyno (%F = 19). The phenyltriazole carboxamide 20, exhibited low clearance,
a long half-life and modest oral bioavailability in both dog (%F= 34) and cyno (%F = 15).
ACS Paragon Plus Environment
1
5

Page 17 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 6. Pharmacokinetic Profile In Dog and Cyno For Pyridine-Based Macrocycles 18, 19, and 20.a
Dose
iv/po
Protein
Binding
(% bound)
AUCb
(nM*h)
LMc
1/2 (min)
Cl
T
(h)
1/2
Vdss
(L/kg)
F
(%)
Entry
Species
Dog
(mL/min/kg)
T
(mpk)
16.4
20
2.9
4.0
3.6
6.0
3.2
2.4
2.9
35
4
266
69
0.40/0.57
>120
54
95.0
96.4
95.7
95.8
96.2
94.5
1
1
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Cyno
Dog
3.3
5.0
3.7
4.8
4.0
0.72/1.33
0.80/NT
0.80/1.58
0.40/0.62
0.71/1.52
15.6
14.5
5.4
IV only
19
IV only
509
73
Cyno
Dog
48
34
922
118
75
20
Cyno
9.9
15
584
a
Compounds were dosed as TFA salts in dog and as HCl salts in cyno in a discrete format. Vehicle for both iv and po: 70% PEG-400; 20%
b
c
water; 10% ethanol. Oral AUC. LM = liver microsome stability.
A cohort of potent, pyridine-based macrocyclic FXIa inhibitors (18-20) exhibiting good
ADME properties and improved, yet modest, oral bioavailability had been discovered. With this
achievement in hand, our focus turned to the evaluation of a pyridine-based macrocycle in a
rabbit thrombosis model (vide infra). As a prelude to this efficacy study, both rabbit FXIa
binding affinity and rabbit aPTT assays were measured for compounds 18-20 (Table 7). Only a
2
-fold loss in binding affinity was seen in rabbit when compared to human. However, a more
dramatic loss in potency in the rabbit aPTT clotting assay was seen which was presumably due to
a combination of lower binding affinity and higher protein binding for rabbit compared to
human. Macrocycle 19 showed the best balance of potency in both human and rabbit (Table 7)
and desirable PK parameters (Tables 5 and 6). Compound 19 exhibited >600 fold selectivity
against many of the relevant serine proteases (Table 8), except for plasma kallikrein (1.5-fold),
with no significant inhibition (IC > 20 M) toward cytochrome P450 enzymes or the hERG ion
50
channel (26% inhibition at 30 M). As a result, 19 was selected for further evaluation in a rabbit
model of arterial thrombosis.
ACS Paragon Plus Environment
1
6

Journal of Medicinal Chemistry
Page 18 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 7. Comparison of Human and Rabbit In Vitro Potency for Macrocycles 18-20.
Entry
Human
FXIa Ki
Rabbit
FXIa Ki
(nM)
Human
aPTT EC1.5x
Rabbit
aPTT EC1.5x
a
b
c
d
(nM)
(M)
(M)
1
8
2.2
4.6
9.6
1.4
30
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
19
4.1
2.0
35
2
0
7.0
15.2
4.4
42
a
o
Ki values were obtained from purified human enzyme at 37 C and were averaged from multiple determinations (n ≥ 2).
b
c
o
Ki values were obtained from purified rabbit enzyme at 37 C and were averaged from multiple determinations (n ≥ 2).
aPTT (activated partial thromboplastin time) in vitro clotting assay was performed in human plasma.
d
aPTT (activated partial thromboplastin time) in vitro clotting assay was performed in rabbit plasma. For compounds 18-20, the rabbit free
fraction is ~1% while free fraction in human plasma is 3 to 4%.
Table 8. Human Serine Protease Selectivity Profile for Compound 19.
Human Enzyme
Ki (nM)a
Selectivity (fold)
Factor XIa
Factor VIIab
Factor IXa
Factor Xa
4.1
>13,300
>27,100
>13,300
2,530
NAc
>3,100
>6,600
>3,100
620
Factor XIIa
Thrombin
>11,500
>10,000
6.3
>2,800
>2,400
1.5
Trypsin
Plasma Kallikrein
Activated Protein C
Plasmin
>21,500
>15,000
2,940
>5,200
>3,600
720
TPA
Urokinase
>15,100
>3,600
a
o
Ki values in nM were obtained using human purified enzymes at 37 C.
FVlla Ki values in nM were obtained using human purified enzyme at 25 C, please see experimental section for details.
NA = Not applicable.
b
c
o
RABBIT ECAT EFFICACY MODEL
The rabbit electrically induced carotid arterial thrombosis (ECAT) model has been used
to evaluate a number of marketed anticoagulants (warfarin, dabigatran, and apixaban) and
antiplatelet agents (clopidogrel and prasugrel).22, 48-49 In this model, thrombosis was induced by
electrical stimulation and carotid blood flow was measured to monitor thrombosis-induced
ACS Paragon Plus Environment
1
7

Page 19 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
occlusion. Compound 19 was evaluated in this model and produced a dose-dependent increase in
blood flow of the injured carotid artery (Figure 5A). In addition, a dose-dependent reduction in
thrombus formation was demonstrated with an estimated ED of 1.6 mg/kg + 1 mg/kg/h (Figure
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
B).
(A)
(B)
Figure 5. Antithrombotic effect of compound 19 in rabbit ECAT model. (A) Carotid blood flow versus time following arterial injury. CBF =
carotid blood flow. (B) Antithrombotic dose-response. TWR = thrombus weight reduction. Vehicle = 10% dimethylacetamide : 90%
cyclodextrin.
CHEMISTRY
A ring-closing olefin (RCM) metathesis strategy was employed as the key step in the
5
0
synthesis for both the pyridine- and imidazole-based macrocycles. The preparation of the RCM
precursors for the 12-membered pyridine-based macrocycles is described in Scheme 1.
5
1
Condensation of aldehyde 24 with (S)-2-methylpropane-2-sulfinamide in the presence of
anhydrous CuSO and Cs CO afforded sulfinimine 25. Using a modified procedure described
4
2
3
52
by Sun et al., an allylindium species, prepared by transmetallation of allylmagnesium bromide
with indium(III) chloride, was added to sulfinimine 25, which gave sulfinamide 26 with
excellent diastereoselectivity (>19:1 diastereomeric excess).5354 Protecing group interconversion
ACS Paragon Plus Environment
1
8

Journal of Medicinal Chemistry
Page 20 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
was accomplished in two steps to give 27. Suzuki-Miyaura coupling between 4-chloropyridine
2
7 and boronic ester 28 provided aniline 29. Aniline 29 was coupled with 3-butenoic acid using
T3P and DIPEA which gave diene 30. Reduction of the nitro to the aniline intermediate followed
by reaction with methyl chloroformate afforded RCM precursor 31. An alternative sequence was
used to prepare RCM precursors 35 and 36. Reduction of the nitro to the bis-aniline intermediate
and selective introduction of the carbamate at -78 ℃ gave aniline 32. Aniline 32 was then
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
coupled with (R)-2-methylbut-3-enoic acid (33) using T P and pyridine which gave 35. The
3
choice of pyridine base was critical in preventing epimerization in the amide coupling step with
chiral acid 33. Aniline 32 was also coupled with (±)-2-ethylbut-3-enoic acid (34) that gave RCM
precursor 36 as a mixture of diastereomers.
Scheme 1. Synthesis of RCM Precursors for the Pyridine-Based Macrocycles.a
O
S
O
S
H
a
b
c, d
Cl
Cl
Cl
O
N
N
H
N
N
N
24
25
26
H2N
NO2
O
O
B
HN
NO2
H2N
NO2
O
28
f
Boc
Cl
Boc
Boc
N
N
H
N
H
H
N
N
N
e
27
29
30
g, h
g, h
O
R1
OH
O
HN
R1
3, R1
Me
3
3
=
=
H2N
NHCO2Me
4, R1
NHCO2Me
Et
Boc
Boc
N
H
N
for 33, i
for 34, f
H
N
N
1, R1 = H
32
3
3
5, R1
=
=
Me
Et
36, R1
aReagents and conditions: (a) (S)-2-methylpropane-2-sulfinamide, CuSO
bromide, InCl
two steps; (e) PdCl
ethylbut-3-enoic acid (34), T
, DCM, 94%; (b) allylmagnesium
O, TEA, CH CN, 100% over
O, 90 C, 78%; (f) 3-butenoic acid or (±)-2-
P, DIPEA, EtOAc, -10 C to rt, 74-87%; (g) Zn, NH Cl, MeOH, 98%; (h) methyl
P, pyridine, EtOAc, -10 C to rt, 97%.
4
2 3
, Cs CO
3
, THF, EtOH, 106%, >19:1 de; (c) 4M HCl in dioxane, MeOH; (d) Boc
2
3
2
(dppf)-CH Cl , boronate 28, K PO , DMSO, H
2
2
3
4
2
3
4
chloroformate, DCM, -78 C, 88%; (i) (R)-2-methylbut-3-enoic acid (33), T
3
ACS Paragon Plus Environment
1
9

Page 21 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The synthesis of the pyridine-based 12-membered macrocycles possessing the
chlorophenyltetrazole and benzamide P1 groups is described in Scheme 2. RCM precursors 31
and 35 were treated with pTsOH to form the pyridinium salt and then cyclized at 40 ℃ using
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
5
Grubbs (II) catalyst which provided macrocycles 37 and 38. A modified procedure was used
for the macrocyclization of 36. RCM precursor 36 was subjected to Grubbs (II) catalyst in a
microwave at 120 ℃, followed by separation of diastereomers by normal phase chromatography,
afforded macrocycle 39. Hydrogenation of 37-39 over palladium on carbon, followed by
deprotection with either TFA or 4M HCl in dioxane yielded the saturated macrocyclic amines
4
0-42. The macrocyclic amines 40-42 were coupled with (E)-2,5-dioxopyrrolidin-1-yl 3-(5-
4
1
chloro-2-(1H-tetrazol-1-yl)phenyl)acrylate or 2,6-difluoro-4-methylbenzoic acid, which gave
the corresponding amides 6-9.
Scheme 2. Synthesis of Pyridine-Based Macrocycles 6-9 from Table 1.a
R1
R1
O
O
O
R1
HN
NHCO2Me
HN
NHCO2Me
HN
NHCO2Me
a for 31 and 35
b for 36
c or d
Boc
Boc
N
N
H2N
H
H
then e or f
N
N
N
3
3
3
1, R1 = H
37, R1 = H
38, R1 = Me
39, R1 = Et
40, R1 = H
5, R1
6, R1
=
=
Me
Et
1, R1 = Me
2, R1 = Et
4
4
R1
N
N
F
O
N
N

HN
NHCO2Me
g or h

O
R=
Me
F
R
N
H
N
Cl
, R1 = Me
9
6
, R = H
7, R1 = Me
, R1 = Et
8
aReagents and conditions: (a) Grubbs II catalyst (20 mol%), pTsOH-H
(
O, DCM, reflux, for 37: 53%, for 38: 73%;
b) Grubbs II catalyst (20 mol%), DCE, microwave, 120 ℃, diastereomers were separated by normal phase
2
2 2
chromatography, 16%; (c) 10% Pd/C, H , MeOH, 100%; (d) 10% Pd/C, H (55 psi), MeOH, THF, 90%; (e) 4M HCl
ACS Paragon Plus Environment
2
0

Journal of Medicinal Chemistry
Page 22 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
in dioxane, 90%; (f) TFA, DCM, 100%; (g) (E)-2,5-dioxopyrrolidin-1-yl 3-(5-chloro-2-(1H-tetrazol-1-
yl)phenyl)acrylate, DIPEA, DMF, 41-74% (h) 2,6-difluoro-4-methylbenzoic acid, EDC, HOBt, DIPEA, DMF, 58%.
Scheme 3 describes the syntheses of the pyridine-based macrocycles 10-22 from Tables 2,
3 and 4. Macrocyclic amine 41 was coupled with the various phenylpyrrolidinone-,
phenylpyrrolidine- and phenylazole carboxylic acids, which afforded analogs 10-22. The
synthesis of the imidazole-based macrocycle 23 followed a similar sequence as described for the
pyridine-based macrocycles (see Supplemental Information Scheme S1).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 3. Synthesis of Pyridine-Based Macrocycles 10-22 in Tables 2, 3 and 4.a
Me
Me
O
O
O
HN
NHCO2Me
HN
NHCO2Me
O
R
OH
H2N
R
N
H
N
a or b or c
N
41




Z

Z

O
N
N
N
N
Y
Y
N
N
N
NH2
Me
Me
R =
R1
R1
F
F
Cl
Cl
Cl
Cl
Cl
Cl
13 R1 = H
14 R1 = F
15 R1 = Cl
16 R1 = H
17 R1 = F
18
Y = N, Z = CH
21 Y = N, Z = O
22 Y = O, Z = N
1
1
0 (diastereomer A)
1 (diastereomer B)
12
19 Y = CH, Z = N
Y and Z = N
20
aReagents and conditions: (a) phenylpyrrolidinone carboxylic acid 43, phenylpyrrolidine carboxylic acid 45, or
phenylazole carboxylic acid 48a and 48c-d, EDC, HOBt, TEA, DMF, 50 C, 17-55%; (b) phenylazole carboxylic
acids 48b, 48e, 50, 52, 54, or 57, EDC, HOBt, DIPEA, DMF, rt, 28-77%; (c) phenylisoxazole carboxylic acid 60,
EDC, HOBt, DIPEA, DMF, rt to 55 C, 34%.
Scheme 4 describes the synthesis of phenylpyrrolidinone carboxylic acid 43 and
phenylpyrrolidine carboxylic acid 45. Condensation of 3-chloroaniline and 2-methylenesuccinic
acid at elevated temperature afforded phenylpyrrolidinone carboxylic acid 43.56 Acid 43 was
converted to the methyl ester 44. Reduction of the lactam carbonyl with BH -THF complex
3
followed by hydrolysis of the ester provided phenylpyrrolidine carboxylic acid 45.
ACS Paragon Plus Environment
2
1

Page 23 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The syntheses for the phenylpyrazole carboxylic acid intermediates are described in
Scheme 5. Appropriately substituted phenylhydrazines 46a-c were condensed with (E)-ethyl 2-
cyano-3-ethoxyacrylate, which provided aminopyrazoles 47a-c. Treatment of 47a-c with
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
57
isoamyl nitrite in THF at elevated temperatures, followed by hydrolysis of the ester, afforded
the unsubstituted phenylpyrazole carboxylic acids 48a-c. Hydrolysis of 47a and 47b gave C5-
amino pyrazole carboxylic acids 48d and 48e. Phenylhydrazine 46b was condensed with ethyl 2-
((dimethylamino)methylene)-3-oxobutanoate to make C5-methyl pyrazole 49. Hydrolysis of the
ester then afforded C5-methyl pyrazole carboxylic acid 50.
Scheme 4. Synthesis of Phenylpyrrolidinone and Phenylpyrrolidine Carboxylic Acid
a
Intermediates.
O
O
O
NH2
a
N
c, d
N
OR
OH
Cl
Cl
Cl
45
4
3 R = H
b
44 R = Me
aReagents and conditions: (a) 2-methylenesuccinic acid, 120 ℃, then H
to rt, 98%; (c) BH -THF, THF, rt, 69%; (d) NaOH (aq), MeOH, rt, 91%.
O, 110 ℃, 52%; (b) SOCl , MeOH, rt, 0 ℃
2
2
3
Scheme 5. Synthesis of Phenylpyrazole Carboxylic Acid Intermediates.a
O
O
N
N
N
N
c, d
or d
OH
OEt
R2
NH2
R1
R1
a
Cl
Cl
48a R1 = H, R2 = H
48b R1 = F, R2 = H
48c R1 = Cl, R2 = H
4
7a R1 = H
47b R1 = F
H
N
NH2
7c R1 = Cl
4
R1
48d
48e R1 = F, R2 = NH2
R1 = H, R2 = NH2
Cl
46a R1 = H
b
4
4
6b R1 = F
6c R1 = Cl
O
N
N
O
N
N
d
OEt
OH
Me
F
Me
F
Cl
Cl
49
50
ACS Paragon Plus Environment
2
2

Journal of Medicinal Chemistry
Page 24 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
aReagents and conditions: (a) (E)-ethyl 2-cyano-3-ethoxyacrylate, K
cyano-3-ethoxyacrylate, NaOAc, AcOH, H O, rt to 100 ℃, 74%; or (E)-ethyl 2-cyano-3-ethoxyacrylate, TEA,
EtOH, reflux, 34%; (b) ethyl 2-((dimethylamino)methylene)-3-oxobutanoate, TEA, EtOH, rt, 45%; (c) isoamyl
nitrite, THF, 65 ℃ or reflux, 80-86%; (d) NaOH, H O, MeOH, rt or 50 ℃, 85-96%; or NaOH, H O, EtOH, reflux,
87-88%; or LiOH, H O, THF, MeOH, rt, 81%.
2
3
CO , EtOH, reflux, 66-85%; or (E)-ethyl 2-
2
2
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The syntheses for phenylimidazole carboxylic acid 52 and phenyltriazole carboxylic acid
4 are described in Scheme 6. Addition of 3-chloro-2-fluoroaniline to ethyl 3-ethoxy-2-nitrobut-
5
2-enoate gave 51. Reacting 51 with triethyl orthoformate and platinum on carbon under a
hydrogen atmosphere at 75 ℃ afforded, following hydrolysis of the ester, phenylimidazole
carboxylic acid 52.58 Diazotization of 3-chloro-2-fluoroaniline and displacement with NaN3
provided the corresponding arylazide intermediate that was condensed with methyl acetoacetate
in the presence of piperidine to give C5-methyl triazole 53.59 Hydrolysis of 53 afforded
phenyltriazole carboxylic acid 54.
Scheme 6. Synthesis of Phenylimidazole and Phenyltriazole Carboxylic Acid Intermediates.a
N
O
NO2
H
N
b, c
N
F
O
OH
OEt
Me
a
F
Cl
Cl
51
52
NH2
F
Cl
N
O
N
O
N
N
N
d, e
f
N
OMe
OH
Me
Me
F
F
Cl
Cl
53
54
aReagents and conditions: (a) ethyl 3-ethoxy-2-nitrobut-2-enoate, EtOH, rt, 36%; (b) 10% Pt/C, H
triethyl orthoformate, 75 C, 61%; (c) NaOH, MeOH, rt, 81%; (d) TFA, H O, NaNO , then NaN , 0 ℃ to rt; (e)
methyl acetoacetate, piperidine, DMSO, H O, rt, 81% over two steps; (f) NaOH, MeOH, H O, 80 C, 53%.
(balloon),
2
2
2
3
2
2
Scheme 7 describes the syntheses of the phenylisoxazole carboxylic acids 57 and 60.
Condensation of 3-chloro-2-fluorobenzaldehyde with hydroxylamine gave the corresponding 3-
chloro-2-fluorobenzaldehyde oxime. Oxidation of the oxime with MagtrieveTM generated the
nitrile oxide in situ which undergoes a 1,3-dipolar cycloaddition with ethyl but-2-ynoate to give
60
isoxazole 56. Hydrolysis of 56 provided the phenylisoxazole carboxylic acid 57. The
ACS Paragon Plus Environment
2
3

Page 25 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
regioisomeric phenylisoxazole carboxylic acid 60 was prepared in three steps from ketone 58.61
Deprotonation of ketone 58 with LiHMDS and reaction with diethyl oxalate provided diketoester
59. Treatment of diketoester 59 with hydroxylamine led to the formation of the corresponding
isoxazole and subsequent hydrolysis gave phenylisoxazole carboxylic acid 60.
Scheme 7. Synthesis of Phenylisoxazole Carboxylic Acid Intermediates.a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
O
O
O
N
N
CHO
F
a, b
c
OEt
OH
Me
Me
F
F
Cl
5
Cl
Cl
57
56
5
O
F
O
O
N
O
O
F
d
OEt
e, c
OH
O
Me
F
Cl
5
Cl
Cl
60
8
59
aReagents and conditions: (a) NH
O, EtOH, rt, 92%; (b) ethyl but-2-ynoate, MagtrieveTM
O, diethyl oxalate, -78 C, 6%; (e)
2
OHHCl, NaOH, H
O, MeOH, rt, 100%; (d) LiHMDS, Et
2
2
,
CH
NH
3
CN, 80 C, 2%; (c) NaOH, H
OHHCl, EtOH, 90 C, 39%.
2
2
CONCLUSION
We described the optimization of imidazole-based macrocyclic FXIa series, which
focused on replacing the imidazole scaffold with a pyridine ring system as well as the design of
new P1 groups. These efforts led to the discovery of potent, orally bioavailable pyridine-based
macrocycles containing phenylazole carboxamide P1 groups. The pyridine-based macrocyclic
series, exemplified by compound 19, exhibited excellent selectivity against relevant blood
coagulation enzymes and is the first reported macrocyclic inhibitor of FXIa to display potent
antithrombotic efficacy in the rabbit ECAT model. Further optimization of the pyridine-based
macrocyclic FXIa inhibitors, which led to additional improvements in oral bioavailability, will
be reported in due course.
ACS Paragon Plus Environment
2
4

Journal of Medicinal Chemistry
Page 26 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
EXPERIMENTAL
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General Chemistry Methods. All reactions were carried out using commercial grade
reagents and solvents. Solution ratios express a volume relationship, unless stated otherwise.
NMR chemical shifts (δ) are reported in parts per million relative to internal TMS, CDCl3,
CD OD, or DMSO-d . Normal phase chromatography was carried out on ISCO CombiFlash
3
6
systems using prepacked silica cartridges and eluted with gradients of the specified solvents.
Preparative reverse phase high pressure liquid chromatography (HPLC) was carried out on C18
HPLC columns using methanol/water gradients containing 0.1% trifluoroacetic acid unless
otherwise stated. Purity of all final compounds was determined to be ≥95% by analytical HPLC
using the following conditions: SunFire C18 column (3.5 μm C18, 3.0 x 150 mm); Gradient
elution (0.5 mL/min) from 10-100% Solvent B for 12 min and then 100% Solvent B for 3 min.
Solvent A is (95% water, 5% acetonitrile, 0.05% TFA) and Solvent B is (5% water, 95%
acetonitrile, 0.05% TFA); monitoring UV absorbance at 220 and 254 nm. Marvin Sketch
1
6.12.19 was used to name all macrocycles.
Methyl
N‐[(14S)‐14‐[(2E)‐3‐[5‐chloro‐2‐(1H‐1,2,3,4‐tetrazol‐1‐yl)phenyl]prop‐2‐enamido]‐9‐oxo‐8,1
2
,7
6
‐diazatricyclo[13.3.1.0 ]nonadeca‐1(19),2(7),3,5,15,17‐hexaen‐5‐yl]carbamate,
Trifluoroacetic acid salt (6). A clear, pale pink solution of 40 (0.070 g, 0.12 mmol) and (E)-2,5-
41
dioxopyrrolidin-1-yl 3-(5-chloro-2-(1H-tetrazol-1-yl)phenyl)acrylate (0.038 g, 0.11 mmol) in
DMF (1.20 mL) and DIPEA (0.11 mL, 0.60 mmol) was stirred at rt for 4 h. Then, H O (4 mL)
2
was added dropwise to the vigorously stirred reaction to give a suspension. The solid was
ACS Paragon Plus Environment
2
5

Page 27 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
collected by filtration to give a pink solid. The cloudy filtrate was extracted with 10%
IPA/CHCl (3x). The combined organic layers were washed with brine and then concentrated to
3
give a pink residue which was combined with the pink solid. Purification by reverse phase
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chromatography gave, following concentration and lyophilization, 6 (62.3 mg, 74%) as a pale
+
1
yellow solid. MS (ESI) m/z: 587.3 (M+H) . H NMR (500 MHz, DMSO-d ) δ 9.97 (s, 1H), 9.84
6
(
s, 1H), 9.81 (s, 1H), 8.76 (d, J=6.3 Hz, 1H), 8.68 (d, J=5.2 Hz, 1H), 7.93 (d, J=1.9 Hz, 1H), 7.75
br dd, J=8.5, 2.2 Hz, 1H), 7.73 - 7.70 (m, 2H), 7.56 - 7.52 (m, 2H), 7.50 (dd, J=8.5, 1.9 Hz, 1H),
(
7
3
.39 (d, J=1.7 Hz, 1H), 6.95 (d, J=15.7 Hz, 1H), 6.84 (d, J=15.7 Hz, 1H), 5.06 - 4.99 (m, 1H),
.70 (s, 3H), 2.42 - 2.35 (m, 1H), 1.96 - 1.85 (m, 1H), 1.81 - 1.64 (m, 3H), 1.58 - 1.43 (m, 1H),
1
3
1.27 - 1.11 (m, 1H), 0.65 - 0.47 (m, 1H). C NMR (126 MHz, DMSO-d ) δ 170.6, 163.5, 158.3
6
(q, J=35.8 Hz, 1C), 156.8, 153.8, 150.3, 147.0, 145.2, 141.0, 136.2, 135.8, 133.0, 130.92,
1
30.86, 130.1, 129.9, 129.0, 128.4, 127.5, 126.9, 122.5 (overlap of two carbons based on
HMQC), 117.0, 116.4, 115.8 (q, J=292.5 Hz, 1C), 52.3, 51.8, 35.9, 31.9, 22.8, 22.6.
Methyl
N‐[(10R,14S)‐14‐[(2E)‐3‐[5‐chloro‐2‐(1H‐1,2,3,4‐tetrazol‐1‐yl)phenyl]prop‐2‐enamido]‐10‐
2
,7
methyl‐9‐oxo‐8,16‐diazatricyclo[13.3.1.0 ]nonadeca‐1(19),2(7),3,5,15,17‐hexaen‐5‐yl]carba
mate, Trifluoroacetic acid salt (7). To a solution of 41, 2TFA (0.028 g, 0.047 mmol) in DMF
(
(
1 mL) was added (E)-2,5-dioxopyrrolidin-1-yl 3-(5-chloro-2-(1H-tetrazol-1-yl)phenyl)acrylate41
0.016 g, 0.047 mmol) and DIPEA (0.041 mL, 0.24 mmol). The rxn was stirred at rt for 18 h.
Then the reaction was diluted with MeOH and purification by reverse phase chromatography
gave, following concentration and lyophilization, 7 (0.020 g, 59%) as a white solid. MS (ESI)
+
1
m/z: 601.0 (M+H) . H NMR (500 MHz, CD OD) δ 9.66 (s, 1H, NH), 9.48 (s, 1H), 8.69 (d,
3
J=6.3 Hz, 1H), 8.18 (d, J=1.7 Hz, 1H), 7.97 (d, J=2.2 Hz, 1H), 7.90 (dd, J=6.2, 1.8 Hz, 1H), 7.66
ACS Paragon Plus Environment
2
6

Journal of Medicinal Chemistry
Page 28 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
dd, J=8.5, 2.5 Hz, 1H), 7.61 (d, J=8.5 Hz, 1H), 7.58 - 7.54 (m, 2H), 7.48 (dd, J=8.5, 2.2 Hz,
1
H), 7.10 (d, J=15.4 Hz, 1H), 6.79 (d, J=15.4 Hz, 1H), 5.15 (dd, J=11.0, 6.1 Hz, 1H), 3.77 (s,
3H), 2.78 - 2.71 (m, 1H), 2.19 - 2.09 (m, 1H), 1.97 - 1.83 (m, 2H), 1.67 - 1.44 (m, 2H), 0.95 (d,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
J=6.9 Hz, 3H), 0.55 - 0.40 (m, 1H).
Methyl
N‐[(10R,14S)‐14‐[(2E)‐3‐[5‐chloro‐2‐(1H‐1,2,3,4‐tetrazol‐1‐yl)phenyl]prop‐2‐enamido]‐10‐e
2
,7
thyl‐9‐oxo‐8,16‐diazatricyclo[13.3.1.0 ]nonadeca‐1(19),2(7),3,5,15,17‐hexaen‐5‐yl]carbama
te, Trifluoroacetic acid salt (8). Using a procedure analogous to that which was used to prepare
6
, compound 42 (0.010 g, 0.016 mmol) was coupled with (E)-2,5-dioxopyrrolidin-1-yl 3-(5-
chloro-2-(1H-tetrazol-1-yl)phenyl)acrylate (5.70 mg, 0.016 mmol) to give 8 (0.0050 g, 41%) as a
+
+ 1
yellow solid. MS (ESI) m/z: 615.4 (M+H) and 617.4 (M+2+H) . H NMR (500 MHz, CD OD)
3
δ 9.67 (s, 1H, NH), 9.49 (s, 1H), 8.69 (d, J=6.1 Hz, 1H), 8.14 (d, J=1.7 Hz, 1H), 7.97 (d, J=2.2
Hz, 1H), 7.88 (dd, J=6.1, 1.9 Hz, 1H), 7.67 (dd, J=8.5, 2.5 Hz, 1H), 7.62 (d, J=8.5 Hz, 1H), 7.59
-
7.55 (m, 2H), 7.49 (dd, J=8.5, 2.2 Hz, 1H), 7.10 (d, J=15.7 Hz, 1H), 6.78 (d, J=15.7 Hz, 1H),
.13 (dd, J=11.1, 5.9 Hz, 1H), 3.77 (s, 3H), 2.50 - 2.43 (m, 1H), 2.20 - 2.10 (m, 1H), 1.95 - 1.78
5
(
m, 2H), 1.67 - 1.49 (m, 2H), 1.48 - 1.38 (m, 1H), 1.31 - 1.21 (m, 1H), 0.85 (t, J=7.3 Hz, 3H),
.66 - 0.54 (m, 1H).
Methyl N‐[(10R,14S)‐14‐(2,6‐difluoro‐4‐methylbenzamido)‐10‐methyl‐9‐oxo‐8,16‐
0
2
,7
diazatricyclo[13.3.1.0 ]nonadeca‐1(19),2(7),3,5,15,17‐hexaen‐5‐yl]carbamate,
Trifluoroacetic acid salt (9). A clear, pink solution of 41, 2HCl (0.015 g, 0.034 mmol), 2,6-
difluoro-4-methylbenzoic acid (5.85 mg, 0.034 mmol), EDC (9.77 mg, 0.051 mmol), and HOBT
(7.81 mg, 0.051 mmol) in DMF (0.34 mL) and DIPEA (0.030 mL, 0.17 mmol) was stirred at rt.
After 17 h, the reaction was partitioned between EtOAc and water and the layers were separated.
ACS Paragon Plus Environment
2
7

Page 29 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The aqueous layer was extracted with EtOAc (2x). The organic layers were combined, washed
with sat. NaHCO , brine, dried over Na SO , filtered and concentrated to give an off-white solid.
3
2
4
The solid was dissolved in 1:1 DMF/MeOH and one drop of TFA. Purification by reverse phase
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chromatography gave, after concentration and lyophilization, 9 (0.0126 g, 58%) as a white solid.
+
1
MS (ESI) m/z: 523.3 (M+H) . H NMR (500 MHz, CD OD) δ 9.65 (s, 1H, NH), 8.75 (d, J=6.1
3
Hz, 1H), 8.16 (d, J=1.7 Hz, 1H), 7.88 (dd, J=5.9, 1.8 Hz, 1H), 7.64 (d, J=8.5 Hz, 1H), 7.57 (d,
J=1.9 Hz, 1H), 7.51 (dd, J=8.5, 2.2 Hz, 1H), 6.90 (d, J=8.8 Hz, 2H), 5.29 (dd, J=11.3, 6.1 Hz,
1
1
H), 3.77 (s, 3H), 2.79 - 2.69 (m, 1H), 2.38 (s, 3H), 2.20 - 2.08 (m, 1H), 2.02 - 1.84 (m, 2H),
1
9
.69 - 1.56 (m, 1H), 1.54 - 1.41 (m, 1H), 0.95 (d, J=6.9 Hz, 3H), 0.58 - 0.42 (m, 1H). F NMR
(471 MHz, CD OD) δ -77.3, -116.2.
3
Methyl
N‐[(10R,14S)‐14‐[1‐(3‐chlorophenyl)‐5‐oxopyrrolidine‐3‐amido]‐10‐methyl‐9‐oxo‐8,16‐diaz
2
,7
atricyclo[13.3.1.0 ]nonadeca‐1(19),2(7),3,5,15,17‐hexaen‐5‐yl]carbamate, Trifluoroacetic
acid salt (10, Diastereomer A) and Methyl
N‐[(10R,14S)‐14‐[1‐(3‐chlorophenyl)‐5‐oxopyrrolidine‐3‐amido]‐10‐methyl‐9‐oxo‐8,16‐diaz
2
,7
atricyclo[13.3.1.0 ]nonadeca‐1(19),2(7),3,5,15,17‐hexaen‐5‐yl]carbamate, Trifluoroacetic
acid salt (11, Diastereomer B). A solution of 41, 2TFA (0.015 g, 0.025 mmol), 43 (9.0 mg,
0
0
.038 mmol), EDC (9.64 mg, 0.050 mmol), HOBT (7.70 mg, 0.050 mmol) and TEA (0.018 mL,
.13 mmol) in DMF (1 mL) was warmed to 50 °C. After 2 h, the reaction was cooled to rt and
then it was concentrated. Purification by reverse phase chromatography (solvent A: 10% ACN,
0% H O, 0.1% TFA and solvent B: 90% ACN, 10% H O, 0.1% TFA) gave 10 (first eluting
9
2
2
diastereomer, diastereomer A, 6 mg, 34%) as a yellow solid. Diastereomer B (second eluting
diastereomer) was obtained but contained impurities. Purification by reverse phase
ACS Paragon Plus Environment
2
8